research questions about healthy eating

Will Healthy Eating Make You Happier? A Research Synthesis Using an Online Findings Archive

• Open access
• Published: 14 August 2019
• Volume 16 , pages 221–240, ( 2021 )

Cite this article

You have full access to this open access article

• Ruut Veenhoven   ORCID: orcid.org/0000-0002-5159-393X 1 , 2  

28k Accesses

10 Citations

128 Altmetric

13 Mentions

Explore all metrics

Healthy eating adds to health and thereby contributes to a longer life, but will it also add to a happier life? Some people do not like healthy food, and since we spend a considerable amount of our life eating, healthy eating could make their life less enjoyable. Is there such a trade-off between healthy eating and happiness? Or instead a trade-on , healthy eating adding to happiness? Or do the positive and negative effects balance? If there is an effect of healthy eating on happiness, is that effect similar for everybody? If not, what kind of people profit from healthy eating happiness wise and what kind of people do not? If healthy eating does add to happiness, does it add linearly or is there some optimum for healthy ingredients in one’s diet? I considered the results published in 20 research reports on the relation between nutrition and happiness, which together yielded 47 findings. I reviewed these findings, using a new technique. The findings were entered in an online ‘findings archive’, the World Database of Happiness, each described in a standardized format on a separate ‘findings page’ with a unique internet address. In this paper, I use links to these finding pages and this allows us to summarize the main trends in the findings in a few tabular schemes. Together, the findings provide strong evidence of a causal effect of healthy eating on happiness. Surprisingly, this effect is not fully mediated by better health. This pattern seems to be universal, the available studies show only minor variations across people, times and places. More than three portions of fruits and vegetables per day goes with the most happiness, how many more for what kind of persons is not yet established.

Similar content being viewed by others

How the Exposure to Beauty Ideals on Social Networking Sites Influences Body Image: A Systematic Review of Experimental Studies

Health, Health-Related Quality of Life, and Quality of Life: What is the Difference?

The Psychology of Mukbang Watching: A Scoping Review of the Academic and Non-academic Literature

Avoid common mistakes on your manuscript.

Introduction

Healthy eating, in particular a diet rich in fruit and vegetables (FV) adds to our health; primarily because it reduces our chances of contracting a number of eating related diseases (Oyebode et al. 2014 ; Bazzano et al. 2002 ; Liu et al. 2000 ). Since good health adds to happiness, it is likely that healthy diets will also add to happiness, but a firm connection has not been established.

In recent years, the relationship between obesity and mental states has begun to attract serious research interest (Becker et al. 2001 ; Rooney et al. 2013 ), as has the relationship between specific micro-nutrients and psychological health (Stough et al. 2011 ). As yet, there is little research on the relationship between nutrition and happiness.

It is worth knowing to what extent our eating habits affect our happiness. One reason is that most people are concerned about their happiness and look for ways to increase it. Most determinants of happiness are beyond our control, but what we eat is largely in our own hands. In this context, we would like to know whether there is a trade-off between healthy eating and happy living. Gains in length of life due to healthy eating may be counterbalanced by loss of satisfaction with life, as is argued in the debate on the benefits of drinking alcohol (Baum-Baicker 1985 ). If so, healthy eating may mean that we live longer, but not happier.

Empirical assessment of the effects of healthy eating on happiness is fraught with complications. One complication is that the effect of nutrition is probably not the same for everybody. Hence, we must identify what food pattern is optimal for what kind of person. A second problem is that happiness can influence nutrition behaviour, for example unhappiness can lead to the consumption of unhealthy comfort foods. Cause and effect must be disentangled. If a healthy diet does appear to add to happiness, then a third question arises: Is eating more healthy food always better or is there an optimum amount one should eat? For instance, is one apple a day enough to make us feel happy? Or will we feel better with four daily portions of fruit? How about small sins, such as a bar of chocolate or a daily glass of wine?

Research Questions

Is there a trade-of or between healthy eating and happiness? Or rather a trade-on , healthy eating adding to happiness? Or do the positive and negative effects balance?

Is this effect of healthy eating on happiness similar for everybody? If not, what kind of people profit from healthy eating and what kind of people do not?

Is the shape of the relationship between healthy eating and happiness linear? The healthier one’s diet, the happier one is? Or is there an optimum?

I explored answers to these three questions in the available research literature and took stock of the findings obtained in quantitative studies on the relation between healthy eating and happiness. I applied a new technique for research reviewing, that takes advantage of an on-line findings archive, the World Database of Happiness (Veenhoven 2018a ), which allows us to present a lot of findings in a few easy to oversee tabular schemes.

To my knowledge, the research literature on this subject has not been reviewed as yet. One review has considered the observed effect of eating fruit and vegetables on psychological well-being (Rooney et al. 2013 ), however, this review does not really deal with happiness, as will be defined in “ Happiness ” section, but is about mental disorders, such as depression and anxiety.

Structure of the Paper

The remainder of this paper is organized as follows. I define the key concepts in “ Concepts and Measures ” section; healthy eating and happiness and give a short account of happiness research. Next, I describe the new review technique in more detail: how the available research findings were gathered and how these are presented in an easy to overview way ( Methods section). Then I discuss what answers the available findings have provided for our research questions ( Results section). I found a clear answer to the first research question, but no clear answers to the second and third question. I discuss these findings in “ Discussion ” section and draw conclusions in “ Conclusions ” section.

Concepts and Measures

There are different view on what constitutes ‘healthy eating’ and ‘happiness’; for this reason, a delineation of these notions is required.

Healthy Eating

I follow the WHO ( 2018 ) characterization of a ‘healthy diet’ as involving’: 1) a varied diet, 2) rich in fruit and vegetables 3) a moderate amount of fats and oil and 4) less salt and sugar than usual these days. The typical Mediterranean diet is considered to fit these demands well. Unhealthy foods are considered to be rich in sugar and fat, such as processed meat, fast foods, sweets, cakes, sodas, deserts, alcohol and other foods high in calories, but low in nutritional content.

Throughout history, the word happiness has been used to denote different concepts that are loosely connected. Philosophers typically used the word to denote living a good life and often emphasize moral behaviour. ‘Happiness’ has also been used to denote good living conditions and associated with material affluence and physical safety. Today, many social scientists use the word to denote subjective satisfaction with life , which is also referred to as subjective well-being (SWB).

Definition of Happiness

In that latter line, I defined happiness as the degree to which an individual judge the overall quality of his/her life-as-a-whole favourably Footnote 1 (Veenhoven 1984 ) and in a later paper distinguished this definition of happiness from other notions of the good life (Veenhoven 2000 ). In this paper, I follow this conceptualization as it is also the focus of the World Database of Happiness (Veenhoven 2018a ) from which the data reported in this paper are drawn.

Components of Happiness

Our overall evaluation of life draws on two sources of information: a) how well one feels most of the time and b) to what extent one perceives one is getting from life what one wants from it. I refer to these sub-assessments as ‘components’ of happiness, called respectively ‘hedonic level of affect’ and ‘contentment’ (Veenhoven 1984 ). The affective component tends to dominate in the overall evaluation of life (Kainulainen et al. 2018 ).

The affective component is also known as ‘affect balance’, which is the degree to which positive affective (PA) experiences outweigh negative affective (NA) experiences Positive experience typically signals that we are doing well and encourages functioning in several ways (Fredrickson 2004 ) and protects health (Veenhoven 2008 ). As such, this aspect of happiness was particularly interesting for this review of effects of healthy eating.

Difference with Wider Notions of Wellbeing

Happiness in the sense of the ‘subjective enjoyment of one’s life-as-a-whole’, should not be equated with satisfaction with domains of life, such as satisfaction with one’s life-style, one’s diet in particular. Likewise, happiness in the sense of the ‘subjective enjoyment of one’s life’ should not be equated with ‘objective’ notions of what is a good life, which are sometimes denoted using the same term. Though strongly related to happiness, mental health is not the same; one can be pathologically happy or be happy in spite of a mental condition.

Differences in wider notions of well-being are discussed in more detail in Veenhoven (15).

Measurement of Happiness

Since happiness is defined as something that is on our mind, it can be measured using questioning. Various ways of questioning have been used, direct questions as well as indirect questions, open questions and closed questions and one-time retrospective questions and repeated questions on happiness in the moment.

Not all questions used fit the above definition of happiness adequately, e.g. not the question whether one thinks one is happier than most people of one’s age, which is an item in the Subjective Happiness Scale (Lyobomirsky and Lepper 1999 ). Findings obtained using such invalid measures are not included in the World Database of Happiness and hence were not considered in this research synthesis. Further detail on the validity assessment of questions on happiness is available in the introductory text to the collection Measures of Happiness of the World Database of Happiness (Veenhoven 2018b ) chapter 4. Some illustrative questions deemed valid for archiving in the WDH are presented below.

Question on overall happiness:

Taking all together, how happy would you say you are these days?

Questions on hedonic level of affect:

Would you say that you are usually cheerful or dejected?

How is your mood today? (Repeated several days).

Question on contentment:

How important are each of these goals for you?

How successful have you been in the pursuit of these goals?

Happiness Research

Over the ages, happiness has been a subject of philosophical speculation and in the second half of the twentieth century it also became the subject of empirical research. In the 1960’s, happiness appeared as a side-subject in research on successful aging (Neugarten et al. 1961 ) and mental health (Gurin et al. 1960 ). In the 1970’s happiness became a topic in social indicators research (Veenhoven 2017 ) and in the 1980s in medical quality of life research (e.g. Calman 1984 ). Since the 2000’s, happiness has become a main subject in the fields of ‘Positive psychology’ (Lyubomirsky et al. 2005 ) and ‘Happiness Economics’ (Bruni and Porta 2005 ). All this has resulted in a spectacular rise in the number of scholarly publications on happiness and in the past year (2017) some 500 new research reports have been published. To date (May 2018), the Bibliography of Happiness list 6451 reports of empirical studies in which a valid measure of happiness has been used (Veenhoven 2018c ).

Findings Archive: The World Database of Happiness

This flow of research findings on happiness has grown too big to oversee, even for specialists. For this reason, a findings archive has been established, in which quantitative outcomes are presented in a uniform format and are sorted by subject. This ‘World Database of Happiness’ is freely available on the internet at https://worlddatabaseofhappiness.eur.nl

Its structure is shown on Fig.  1 and a recent description of this novel technique for the accumulation of research findings can be found with Veenhoven ( 2019 ).

Start page of the World Database of Happiness, showing the structure of this findings archive

One of the subject categories in the collection of correlational findings is ‘Happiness and Nutrition’ (Veenhoven 2018c ). I draw on that source for this paper.

A first step in this review was to gather the available quantitative research findings on the relationship between happiness and healthy eating. The second step was to present these findings in an uncomplicated form.

Gathering of Research Findings

In order to identify relevant papers for this synthesis, I inspected which publications on the subject of healthy eating were already included of the Bibliography of World Database of Happiness, in the subject sections ‘ Health behaviour’ and consumption of ‘ Food ’. Then to further complete the collection of studies, various databases were searched such as Google Scholar, EBSCO, ScienceDirect, PsycINFO, PubMed/Medline, using terms such as ‘ happiness ’, ‘ life satisfaction ’, ‘ subjective well-being ’, ‘ well-being ’, ‘ daily affect ’, ‘ positive affect ’, ‘ negative affect ’ in connection with terms such as ‘ food ’, ‘ healthy food ’, ‘ fruit and vegetables ’, ‘ fast food ‘and ‘ soft drinks ’ in different sequences.

All reviewed studies had to meet the following criteria:

A report on the study should be available in English, French, German or Spanish.

The study should concern happiness in the sense of life-satisfaction (cf. Healthy Eating section). I excluded studies on related matters, such as on mental health or wider notions of ‘flourishing’.

The study should involve a valid measure of happiness (cf. Happiness section). I excluded scales that involved questions on different matters, such as the much-used Satisfaction With Life Scale (Diener et al. 1985 ).

The study results had to be expressed using some type of quantitative analysis.

Studies Found

Together, I found 20 reports of an empirical investigation that had examined the relationship between healthy eating and happiness, of which two were working papers and one dissertation. None of these publication s reported more than one study . Together, the studies yielded 47 findings.

All the papers were fairly recent, having been published between 2005 and 2017. Most of the papers (44.4%) were published in Medical Journals, including the International Journal of Behavioural Medicine, Journal of Health Psychology, The Journal of Nutrition, Health & Aging, The Journal of Nutrition, Health & Aging, The Journal of Psychosomatic Research, The International Journal of Public Health, and Social Psychiatry & Psychiatric Epidemiology.

People Investigated

Together, the studies covered 149.880 respondents and 27 different countries. The publics investigated in these studies, included the general population in countries and particular groups such as students, children, veterans and medical patients. The majority of respondents belonged to a general public group (50%), students made up 27.8%, with children and veterans each forming 11.1%.

Research Methods Used

Most of the studies were cross-sectional 64.4%, longitudinal and daily food diaries accounted for 22% and 10.2% of the total number of studies respectively, and one experimental study accounted for 3.4%.

I present an overview of all the included studies, including information about population, methods and publication in Table 1 .

Format of this Research Synthesis

As announced, I applied a new technique of research reviewing, taking advantage of two technical innovations: a) The availability of an on-line findings-archive (the World Database of Happiness) that holds descriptions of research findings in a standard format and terminology, presented on separate finding pages with a unique internet address. b) The change in academic publishing from print on paper to electronic text read on screen, in which links to that online information can be inserted.

Links to Online Detail

In this review, I summarize the observed statistical relationships as +, − or 0 signs. Footnote 2 These signs link to finding pages in the World Database of Happiness, which serves as an online appendix in this article. If you click on a sign, one such a finding page will open, on which you can see full details of the observed relationship; of the people investigated, sampling, the measurement of both variables and the statistical analysis. An example of such an electronic finding page is presented in Fig.  2 . This technique allows me to present the main trends in the findings, without burdening the reader with all the details, while keeping the paper to a controllable size, at the same time allowing the reader to check in depth any detail they wish.

Example of an online findings page

Organization of the Findings

I first sorted the findings by the research method used and these are presented in three separate tables. I distinguished a) cross-sectional studies, assessing same-time relationships between diet and happiness (Table 2 ), b) longitudinal studies, assessing change in happiness following changes in diet (Table 3 ), and c) experimental studies, assessing the effect of induced changes in diet on happiness (Table 4 ).

In the tables, I distinguish between studies at the micro level, in which the relation between diet and happiness of individuals was assessed and studies at the macro level, in which average diet in nations is linked to average happiness of citizens.

I present kinds of foods consumed vertically and horizontally two kinds of happiness: overall happiness (life-satisfaction) and hedonic level of affect.

Presentation of the Findings

The observed quantitative relationships between diet and happiness are summarized using 3 possible signs: + for a positive relationship, − for a negative relationship and 0 for a non-relationship. Statistical significance is indicated by printing the sign in bold . See Appendix . Each sign contains a link to a particular finding page in the World Database of Happiness, where you can find more detail on the checked finding.

Some of these findings appear in more than one cell of the tables. This is the case for pages on which a ‘raw’ (zero-order) correlation is reported next to a ‘partial’ correlation in which the effect of the control variables is removed. Likewise, you will find links to the same findings page at the micro level and the macro level in Table 2 ; on this page there is a time-graph of sequential studies in Russia from which both micro and macro findings can be read.

Several cells in the tables remain empty and denote blanks in our knowledge.

Advantages and Disadvantages of this Review Technique

There are pros and cons to the use of a findings-archive such as the World Database of Happiness and plusses and minuses to the use of links to an on-line source in a text like this one.

Use of a Findings-Archive

Advantages are: a) efficient gathering of research on a particular topic, happiness in this case, b) sharp conceptual focus and selection of studies on that basis, c) uniform description of research findings on electronic finding pages, using a standard format and a technical terminology, d) storage of these finding pages in a well searchable database, e) which is available on-line and f) to which links can be made from texts. The technique is particular useful for ongoing harvesting of research findings on a particular subject.

Disadvantages are: a) the sharp conceptual focus cannot easily be changed, b) considerable investment is required to develop explicit criteria for inclusion, definition of technical terms and software, Footnote 3 c) which pays only when a lot of research is processed on a continuous basis.

Use of Links in a Review Paper

The use of links to an on-line source allows us to provide extremely short summaries of research findings, in this text by using +, − and 0 signs in bold or not, while allowing the reader access to the full details of the research. This technique was used in an earlier research synthesis on wealth and happiness (Jantsch and Veenhoven 2019 ) and is described in more detail in Veenhoven ( 2019 ). Advantages of such representation are: a) an easy overview of the main trend in the findings, in this case many + signs for healthy foods, b) access to the full details behind the links, c) an easy overview of the white spots in the empty cells in the tables, and d) easy updates, by entering new sign in the tables, possibly marked with a colour.

The disadvantages are: a) much of the detailed information is not directly visible in the + and – signs, b) in particular not the effect size and control variables used, and c) the links work only for electronic texts.

Differences with Traditional Reviewing

Usual review articles cannot report much detail about the studies considered and rely heavily on references to the research reports read by the reviewer, which typically figure on a long list at the end of the review paper that the reader can hardly check. As a result, such reviews are vulnerable to interpretations made by the reviewer and methodological variation can escape the eye.

Another difference is that the conceptual focus of many traditional reviews in this field is often loose, covering fuzzy notions of ‘well-being’ rather than a well-defined concept of ‘happiness’ as used here. This blurs the view on what the data tell and involves a risk of ‘cherry picking’ by reviewers. A related difference is that traditional reviews of happiness research often assume that the name of a questionnaire corresponds with its conceptual contents. Yet, several ‘happiness scales’ measure different things than happiness as defined in “ Healthy Eating ” section, e.g. much used Life Satisfaction Scale (Neugarten et al. 1961 ), which measures social functioning.

Still another difference is that traditional narrative reviews focus on interpretations advanced by authors of research reports, while in this quantitative research synthesis I focus on the data actually presented. An example of such a difference in this review, is the publication by Connor & Brookie (Conner et al. 2015 ) who report no effect of healthier eating on mood in the experimental group, while their data show a small but significant gain in positive affect and a small but insignificant reduction of negative effect (Table 3 ), which together denote a positive effect on affect balance.

Difference with Traditional Meta-Analysis

Though this research synthesis is a kind of meta-analysis, it differs from common meta-analytic studies in several ways. One difference is the above- mentioned conceptual rigor; like narrative reviews many meta-analyses take the names given to variables for their content thus adding apples and oranges. Another difference is the direct online access to full detail about the research findings considered, presented in a standard format and terminology, while traditional meta-analytic studies just provide a reference to research reports from which the data were taken. A last difference is that most traditional meta-analytic studies aim at summarizing the research findings in numbers, such as an average effect size. Such quantification is not well possible for the data at hand here and not required for answering our research questions. My presentation of the separate findings in tabular schemers provides more information, both of the general tendency and of the details.

Let us now revert to the research questions ( Structure of the Paper section) and answer these one by one.

Is there a Trade-Of between Healthy Eating and Happiness?

Or does healthy eating rather add to happiness or do the positive and negative effects balance.

This question was addressed using different methods, a) same-time comparison of diet and happiness (cross-sectional analysis) b) follow-up of change in happiness following change in diet (longitudinal) and c) assessing the effect on happiness of induced change in diet (experimental). The results are summarized in, respectively, Tables  2 , 3 and 4 .

Cross-Sectional Findings

Together I found 42 correlational findings, which are presented in Table 2 . Of these findings 14 concerned raw correlations, while 28 reflected the results of a multivariate analysis. In Table 2 I see only micro level studies.

There were 16  +  signs, which indicates that people who eat healthy tend to be happier than people who do not. A few (3) – signs were linked to unhealthy eating habits, i.e. fast food, soft drinks and sweets, and as such support this pattern.

Not all the findings supported the view that healthy eating goes with greater happiness. Consumption of soft-drinks was positively related to overall happiness, though not significantly, while the correlation with affect balance was significantly negative. A high intake of high caloric protein and fat is generally deemed to be unhealthy but appeared in one case to go with greater overall happiness, a study among medical patients in Arkhangelsk in Russia, where the medical conditions and cold climate may have require a higher intake of such foods.

The findings were mixed with respect to the relation of happiness with consumption of animal products, dairy and meat. For these foods a positive relation with overall happiness was found and a negative relation with affect level, in the case of milk products both relations were insignificant.

Several studies report both raw correlations and partial ones for the same population. Controls reduced the effect size somewhat but did not change the direction of the correlation. Importantly, the control for health and other health behaviours in 8 studies Footnote 4 did not change the direction of the correlation.

Longitudinal Findings

The findings of two studies that assessed the change in happiness following change in diet are presented in Table 3 , one study at the micro level among students and another study at the macro-level among the general population in Russia. Both studies found positive correlations, indicating that healthier eating adds to one’s happiness. The effects of greater consumption of meat and milk were not significant. No control variables were used in these studies. The relationship between healthy eating and affect level was not investigated longitudinally.

Experimental Study

To date, there is only one study on the effect of induced change to a healthier diet on an individual’s happiness. In this study people were randomly assigned to an experimental group and stimulated in various ways to consume more fruit and vegetables (FV), among other things by providing vouchers for health foods and sending e-mail reminders. After 2 weeks of increased FV consumption, the participant’s mood level had increased more than those of the control group.

Together, these findings provide a clear answer to our first research question. The net effect of healthy eating on happiness tends to be positive. If there is any trade-off at all, this is apparently more than compensated by the trade-on . The positive relationship is robust across research methods and measures of happiness.

Is this Effect of Healthy Eating on Happiness Similar for Everybody?

If not, what kind of people profit from healthy eating and what kind of people do not.

The 19 studies reported here cover a wide range of populations, the general public in several parts of the world, children, students, church members, medical patients and elderly war veterans. No great differences in the correlation between diet and happiness appear in these findings, though children seem to be happier when allowed to consume sweets and soft drinks. The cross-national study by Grant et al. ( 2009 ) observed some differences in strength of the correlation between healthy eating and happiness across part of the world, but no difference in direction of the correlation. The micro-level studies by Pettay ( 2008 ) and Warner et al. ( 2017 ) found no differences between males and females, while Ford et al. ( 2013 ) found a slightly bigger negative effect of unhealthy eating among women than among men.

The observed positive effect of healthy eating on happiness seems to be universal. Possible differences in what diet provides the most happiness for whom have not (yet) been identified.

Is the Shape of The Relationship Linear; the Healthier One’s Diet, the Happier One Is?

Or is there an optimum, if so what is optimal for whom.

Two studies find a linear relation between happiness and the number of portions fruits and vegetables per day, Lesani et al. ( 2016 ) among students in Iran and Blanchflower et al. ( 2013 ) among the general public in the UK, the latter study up to 7–8 portion a day. Another study observed an optimum at the lower level of 3–4 portions a day among female Iranian students (Fararouei et al. 2013 ). These thee studies suggest that the optimum is at least beyond three portions a day. As yet the focus of research has been on particular kinds of food, while the relationship between happiness and total diet composition has not been investigated.

Together, our findings leave no doubt that healthy eating ads to happiness, frequent consumption of fruit and vegetables in particular.

Causal Effect

Though happiness may influence nutrition behaviour, happier people being more inclined to follow a healthy diet, there is strong evidence for a causal effect of healthy eating on happiness. Spurious correlation is unlikely to exist, since correlations remain positive after controlling for many different variables. Causality is strongly suggested by 3 out of the 4 longitudinal findings and the experimental study.

This is not to say that healthy eating will always add to the happiness of everybody, but the trend is sufficiently universal and strong to be used in policies that aim at greater happiness for a greater number of people, such as in happiness education.

Causal Paths

Healthy eating will add to good health and good health will add to happiness. An unexpected finding is that the effect of healthy eating on happiness is not fully mediated by better health. As mentioned in “ Is there a Trade-Of between Healthy Eating and Happiness? ” section, significant positive correlations remain when health is controlled. This means that healthy eating also affects happiness in other ways. As yet I can only speculate about what these ways are. Possibly effects are that healthy eaters attract nicer people or that intake of fruit and vegetables has a direct effect on mood.

Limitations

This first synthesis of the research on happiness and healthy eating draws on 20 empirical studies, which together yielded 47 findings. Though these results provide strong indications of a positive effect of healthy eating on happiness, we need more research to be sure. This research synthesis limits to happiness defined as the subjective enjoyment of one’s life as a whole and measure that matter adequately. This conceptual focus has a piece, we came to know more about less. The available research findings do not allow a traditional meta-analysis, both because of the limited numbers and their heterogeneity. Hence, we cannot yet compute effect sizes or test statistical significance of differences.

Topics for Further Research

Although we now know that healthy eating tends to make one’s life more satisfying, we do not know in much detail what particular diets are the most conducive to the happiness of what kinds of people. We are also largely in the dark about the causal mechanisms involved. The focus of current research is very much on particular food items, consumption of fruit and vegetables in particular. Future research should pay more attention to the effect of total diets on happiness.

Conclusions

Healthy eating adds to happiness, not just by protecting one’s health but also in other, as yet unidentified, ways. This finding deserves to be drawn to the public’s attention. People should know that changing to a healthier diet will not be at the cost of their happiness but will add to it. Faulty beliefs and misleading advertisements should be counter-balanced by this established fact.

Likewise, Diener (26) defined ‘life satisfaction’ as an overall judgement of one’s life.

The technique also allows summarization in a number, which can be presented in a stem-leaf diagram, or in short verbal. Statements, such as ‘U shaped relationship’

The archive can be easily adjusted for other subjects. The software is Open Source

Blanchflower et al. ( 2013 ); Fararouei et al. ( 2013 ); Ford et al. ( 2013 ); Huffman and Rizov ( 2016 ); Lesani et al. ( 2016 ); Lengyel et al. ( 2009 ) and Kye and Park ( 2014 )

Studies Included in this Research Synthesis Are Marked with a Link below the Reference. The Links Lead to a Standardized Description of that Study in the World Database of Happiness. The Codes Denote Place and Year of the Study

Averina, M. M., Brox, J. & Nilsson, O. (2005) Social and lifestyle determinants of depression, anxiety, sleeping disorders and self-evaluated quality of life in Russia. Social Psychiatry and Psychiatric Epidemiology 40: 511–518 Study RU Archangelsk 1999

Baum-Baicker, C. (1985). The psychological benefits of moderate alcohol consumption: a review of the literature. Drug & Alcohol Dependence, 15 (4), 305–322.

Article   Google Scholar  

Bazzano, L. A., He, J., Ogden, L. G., Loria, C. M., Vupputuri, S., Myers, L., & Whelton, P. K. (2002). Fruit and vegetable intake and risk of cardiovascular disease in US adults: the first National Health and Nutrition Examination Survey Epidemiologic Follow-up. American Journal of Clinical Nutrition, 76 (1), 93–99.

Becker, E. S., Margraf, J., Türke, C., Soeder, U., & Neumer, S. (2001). Obesity and mental illness in a representative sample of young women. International Journal of Obesity Related Metabolic Disorders, 25 (Suppl. 1), S5–S9.

Blanchflower, D. G., Oswald, A. J., & Stewart-Brown, S. (2013). Is psychological well-being linked to the consumption of fruit and vegetables? Social Indicators Research, 114 , 785–801 Study GB Wales 2007-2010 .

Breslin, G., Donnelly, P., & Nevill, A. M. (2013). Socio-demographic and behavioural differences and associations with happiness for those who are in good and poor health. International Journal of Happiness and Development, 1 , 142–154 Study GB 2009 .

Bruni, L., & Porta, P. L. (2005). Economics and happines . UK: Oxford University Press.

Book   Google Scholar  

Caligiuri, S., Lengyel, C. O., & Tate, R. B. (2012). Changes in food group consumption and associations with self-rated diet, health, life satisfaction, and mental and physical functioning over 5 years in very old Canadian men: The Manitoba follow-up study. The Journal of Nutrition, Health & Aging, 16 (8), 707–712 Study CA 2000-2005 .

Calman, A. C. (1984). Quality of life in cancer patients--a hypothesis. Journal of Medical Ethics, 10 , 124–127.

Chang, H. H., & Nayga, R. M., Jr. (2010). Childhood obesity and unhappiness: The influence of soft drinks and fast food consumption. Journal of Happiness Studies, 11 , 261–275 Study TW 2001 .

Conner, T.S. & Brookie, K.L. (2017). Let them eat fruit! The effect of fruit and vegetable consumption on the psychological well-being in young adults: A randomized controlled trial. PLOS one , February. 3201 Study NZ Auckland 2015

Conner, T.S., Brookie, K.L. & Richardson, A.C. (2015). On carrots and curiosity: Eating fruit and vegetables is associated with greater flourishing in daily life. British Journal of Health Psychology, 20: 413–444. Study NZ Auckland 2013

Diener, E., Emmons, R. A., Griffin, S., & Larsen, R. J. (1985). The Satisfaction with Life Scale. Journal of Personality Assessment, 49 , 71–75.

Fararouei, M., Akbartabar Toori, M. & Brown, I.J. (2013). Happiness and health behaviour in Iranian adolescent girls. Journal of Adolescence, 36: 1187–1192. Study IR 2008

Ford, P.A., Jaceldo-Siegl, K. & Lee, J.W. (2013). Intake of mediterranean foods associated with positive and low negative affect. Journal of Psychosomatic Research , 74 (2013) 142–148. Study ZZ Anglo-America 2002

Fredrickson, B. L. (2004). The broaden - and - build theory of positive emotions . Philosophical Transactions, Biological Sciences , 359: 1367–1377.

Grant, N.; Steptoe, A.; Wardle, J. (2009). The relationship between life satisfaction and health behaviour: a cross-cultural analysis of young adults. International Journal of Behavioural Medicine , 16: 259–268 Study ZZ 1999-2001

Gschwandtner, A., Jewell, S. & Kambhampati, U. (2015). On the relationship between lifestyle and happiness in the UK. Paper for 89th Annual Conference of AES, 2015, 1–33, Warwick, England. Study GB 2012

Gurin, G., Feld, S. & Veroff, J. (1960). Americans view their mental health. A nationwide interview survey. Basic Books, New York, USA (Reprint in 1980, Arno Press, New York, USA).

Honkala, S., Al Sahli, N., & Honkala, E. (2006). Consumption of sugar products and associated life- and school- satisfaction and self-esteem factors among schoolchildren in Kuwait. Acta Odonatological Scandinavia, 64 , 79–88 Study KW 2002 .

Huffman, S.K.; Rizov, M. (2016). Life satisfaction and diet: Evidence from the Russian longitudinal monitoring survey. Paper prepared for presentation at the Agricultural & Applied Economics Association Annual Meeting, 2016, 1–25, Boston, Massachusetts. Study RU 1994-2005

Jantsch, A & Veenhoven, R. (2019). Private wealth and happiness: A research synthesis using an online findings archive . In Gael Brule & Christian Suter (Eds.), “Wealth(s) and Subjective Well-Being Springer/Nature, pp 17–50.

Kainulainen, S., Saari, J., & Veenhoven, R. (2018). Life-satisfaction is more a matter of how well you feel, than of having what you want . International Journal of Happiness and Development, 4 (3), 209–235.

Kye, S.K. & Park, K. (2014). Health-related determinants of Happiness in Korean Adults. International Journal of Public Health, 59: 731–738. Study KR 2009

Lengyel, C. O., Oberik Blatz, A. K., & Tate, R. B. (2009). The relationships between food group consumption, self-rated health, and life satisfaction of community-dwelling canadian older men: The manitoba follow-up study. Journal of Nutrition for the Elderly, 28 , 158–173 Study CA 2000 .

Lesani, A., Javadi, M., & Mohammadpoorasl, A. (2016). Eating breakfast, fruit and vegetable intake and their relation with happiness in college students. Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity, 21 , 645–651 Study IR Tehran 2011 .

Liu, S., Manson, J. E., Lee, I. M., Cole, S. R., Hennekens, C. H., Willett, W. C., & Buring, J. E. (2000). Fruit and vegetable intake and risk of cardiovascular disease: The Women’s Health Study. American Journal of Clinical Nutrition, 72 (4), 922–928.

Lyobomirsky, S., & Lepper, H. S. (1999). A measure of subjective happiness: preliminary reliability and construct validation. Social Indicators Research, 46 , 137–155.

Lyubomirsky, S., Diener, E., & King, L. A. (2005). The benefits of frequent positive affect: Does happiness lead to success? Psychological Bulletin, 131 , 803–855.

Mujcic, R., & Oswald, A. (2016). Evolution of well-being and happiness after increases in consumption of fruit and vegetables. American Journal of Public Health, 106 (8), 1504–1510 Study AU 2007-2009 .

Neugarten, B.L., Havighurst, R.J. & Tobin, S. S. (1961). The measurement of life satisfaction . Journal of Gerontology , 16: 134–143.

Oyebode, O., Gordon-Dseagu, V., Walker, A., & Mindell, J. S. (2014). Fruit and vegetable consumption and all-cause, cancer and CVD mortality: analysis of Health Survey for England data. Journal of Epidemiological Community Health, 68 (9), 856–862.

Pettay, R.S. (2008). Health behaviours and life satisfaction in college students . PhD Thesis, Kansas State University, USA. Study US 2006

Rooney, C., McKinley, M.C. & Woodside, J.V. (2013). The Potential Role of Fruit and Vegetables in aspects of Psychological Well-Being: A Review of the Literature and Future Directions. Proceedings of the Nutrition Society, 72: 420–432.

Stough, C., Scholey, A., Lloyd, J., Spong, J., Myers, S., & Downey, L. A. (2011). The effect of 90-day administration of a high dose vitamin B-complex on work stress. Human Physio-pharmacology, 26 (7), 470–476.

Google Scholar  

Veenhoven, R. (1984). Conditions of happiness . Reidel (now Springer), Dordrecht, Netherlands.

Veenhoven, R. (2000). The four qualities of life. Ordering concepts and measures of the good life . Journal of Happiness Studies, 1 , 1–39.

Veenhoven, R. (2008). Healthy happiness: Effects of happiness on physical health and the consequences for preventive health care . Journal of Happiness Studies, 9 , 449–464.

Veenhoven, R. (2017). Co-development of happiness research: Addition to “fifty years after the social indicator movement . Social Indicators Research , 135: 1001–1007.

Veenhoven, R. (2018a) World database of happiness: Archive of research findings on subjective enjoyment of life . Erasmus University Rotterdam, The Netherlands.

Veenhoven, R. (2018b). Measures of happiness . World Database of Happiness, Erasmus University Rotterdam.

Veenhoven, R. (2018c). Bibliography of happiness . World Database of Happiness, Erasmus University Rotterdam.

Veenhoven, R. (2019). World database of happiness: A ‘findings archive’ . Chapter in Handbook of Wellbeing, Happiness and the Environment. Editors: Heinz Welsch, David Maddison and Katrin Rehdanz, Edward Elgar Publishing (forthcoming).

Warner, R.M., Frye, K. & Morrell, J. S. (2017). Fruit and vegetable intake predicts positive affect. Journal of Happiness Studies , 18: 809–826. Study US New England 2013

WHO (2018) Healthy diet. http://www.who.int/news-room/fact-sheets/detail/healthy-diet

Download references

Author information

Authors and affiliations.

Erasmus Happiness Economics Research Organization EHERO, Erasmus University Rotterdam, POB 1738, 3000DR, Rotterdam, Netherlands

Ruut Veenhoven

Optentia Research Program, North-West University, Vanderbijlpark, South Africa

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Ruut Veenhoven .

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Signs used in Finding Tables

positive correlation, statistically significant

positive correlation, not statistically significant

positive and negative correlations, depending on control variables used

no correlation

negative correlation, not statistically significant

negative correlation, statistically significant

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Reprints and permissions

About this article

Veenhoven, R. Will Healthy Eating Make You Happier? A Research Synthesis Using an Online Findings Archive. Applied Research Quality Life 16 , 221–240 (2021). https://doi.org/10.1007/s11482-019-09748-7

Download citation

Received : 20 December 2018

Accepted : 20 June 2019

Published : 14 August 2019

Issue Date : February 2021

DOI : https://doi.org/10.1007/s11482-019-09748-7

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Health behaviour
• Research synthesis
• Find a journal
• Publish with us
• Track your research

Log in using your username and password

• Search More Search for this keyword Advanced search
• Latest content
• BMJ NPH Collections
• BMJ Journals More You are viewing from: Google Indexer

You are here

• Online First
• Qualitative research study on addressing barriers to healthy diet among low-income individuals at an urban, safety-net hospital
• Article Text
• Article info
• Citation Tools
• Rapid Responses
• Article metrics

other Versions

• You are currently viewing an earlier version of this article (January 17, 2023).
• View the most recent version of this article

• Erin Cahill 1 ,
• Stacie R Schmidt 2 ,
• Tracey L Henry 2 ,
• http://orcid.org/0000-0002-2791-9960 Gayathri Kumar 3 ,
• Sara Berney 4 ,
• Jada Bussey-Jones 2 and
• Amy Webb Girard 1
• 1 Emory University School of Public Health , Atlanta , Georgia , USA
• 2 Division of General Medicine and Geriatrics , Emory University School of Medicine , Atlanta , Georgia , USA
• 3 Emory University School of Medicine , Atlanta , Georgia , USA
• 4 North Carolina State University School of Public and International Affairs , Raleigh , North Carolina , USA
• Correspondence to Tracey L Henry, General Medicine and Geriatrics, Emory University, Atlanta, GA 30322, USA; henrytracey{at}hotmail.com

Background Some American households experience food insecurity, where access to adequate food is limited by lack of money and other resources. As such, we implemented a free 6-month Fruit and Vegetable Prescription Program within a large urban safety-net hospital .

Methods 32 participants completed a baseline and postintervention qualitative evaluation about food-related behaviour 6 months after study completion. Deductive codes were developed based on the key topics addressed in the interviews; inductive codes were identified from analytically reading the transcripts. Transcripts were coded in MAXQDA V.12 (Release 12.3.2).

Results The information collected in the qualitative interviews highlights the many factors that affect dietary habits, including the environmental and individual influences that play a role in food choices people make. Participants expressed very positive sentiments overall about their programme participation.

Conclusions A multifaceted intervention that targets individual behaviour change, enhances nutritional knowledge and skills, and reduces socioeconomic barriers to accessing fresh produce may enhance participant knowledge and self-efficacy around healthy eating. However, socioeconomic factors remain as continual barriers to sustaining healthy eating over the long term. Ongoing efforts that address social determinants of health may be necessary to promote sustainability of behaviour change.

• nutritional treatment
• nutrition assessment
• malnutrition

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:  http://creativecommons.org/licenses/by-nc/4.0/ .

https://doi.org/10.1136/bmjnph-2020-000064

Statistics from Altmetric.com

Request permissions.

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Most US households have consistent, reliable access to enough food for active, healthy living. 1 Some American households, however, experience food insecurity, which is defined by the US Department of Agriculture (USDA) as a lack of consistent access to enough food for an active, healthy life. 1 In 2016, an estimated one in eight Americans were food insecure, equating to 42 million people. 1 Food insecurity can be influenced by a number of factors including income, employment and disability (Healthy People 2020). The prevalence of food insecurity varies across subgroups of the US population; some groups are more likely to be food insecure than others. The distribution of food insecurity across residence areas shows that the majority of food-insecure households are in metropolitan areas, with income as one of the primary characteristics associated with food insecurity. 2 Lower income households have a higher prevalence of food insecurity compared with higher income households. 2 Furthermore, food insecurity may increase the risk for obesity and chronic diseases. 3

Food assistance programmes such as the Women, Infants and Children programme and the Supplemental Nutrition Assistance Program (SNAP) help address barriers to accessing healthy food and may reduce food insecurity. 4 5 Interventions implemented within healthcare settings—such as onsite food pantries and mobile food distributions—that serve food-insecure populations have also been effective. 3

Our hospital is a large, urban safety-net hospital in metro Atlanta that provides care to low income and other vulnerable populations. In 2015, an assessment of food insecurity was conducted in the hospital’s primary care centre, where 323 patients completed a questionnaire that included questions regarding age, sex, race, household income, number of people in the household, zip code, diabetes status, the USDA two-item food security screener and SNAP utilisation. The study revealed that 55% of low-income patients receiving outpatient care were food-insecure. 6 To address this issue, we implemented a free 6-month Fruit and Vegetable Prescription Program (FVRx) within a primary care clinic at the hospital in 2016.

Eligible participants had a Body Mass Index (BMI)>30 and at least one associated chronic condition, such as diabetes. Components of the FVRx programme included 4 weeks of fruit and vegetable prescriptions to be redeemed for fresh fruits and vegetables packaged locally, monthly interactive groups classes on nutrition, and monthly cooking classes providing evidence-based nutrition and cooking skills education.

On completion of the programme, we conducted a postintervention qualitative evaluation among participants of the FVRx programme to assess (1) constraints on programme participation, (2) barriers to maintaining a healthy diet among participants, (3) participant capacity to sustain behaviour change during and after completion of the programme, in an effort to identify strategies that could improve participant retention and satisfaction with future programmes. This paper describes the results of this evaluation.

This evaluation incorporated a qualitative research study design. A telephone interview script was used to ask questions about patients’ experience with the FVRx programme, grocery shopping habits and the patient’s current fruit and vegetable consumption (see online supplementary appendix A ). Interview questions were developed to address the main goals of the evaluation, which were to investigate constraints on programme participation, barriers to maintaining a healthy diet among participants postintervention and strategies to improve participant retention.

Supplemental material

Enrolment of the 32 patients into the FVRx study occurred in June 2016; participation in the FVRx programme by the 32 participants took place from July 2016 to December 2016. Participants were referred to the programme by their primary care provider if they had a BMI>30 and at least one diet-related illness. All 32 participants had access to a phone rather their own or a family member’s phone.

The first author contacted the original 32 patients who participated in the FVRx programme by phone in June 2017, approximately 6 months after completing the programme. Six of 32 participants did not answer but had a working voicemail, for which a maximum of two messages were left. Additionally, the team encountered the wrong number for three participants, and full mailboxes for two numbers. Two numbers went unanswered (no voicemail) and one number was disconnected. Thus, of the 32 participants, 18 were reached by phone and verbally consented to participate in follow-up evaluation. Seven participants completed the programme while 11 participants attended a few classes but dropped out. None of the FVRx participants contacted refused to be interviewed.

Interviews were recorded using the TapeACall app and transcribed verbatim. Four interviews were not recorded due to technical difficulties with the app. In these instances, detailed notes were taken and were used in analyses in lieu of verbatim transcripts. A codebook was developed consisting of deductive and inductive codes. Deductive codes were developed based on the key topics addressed in the interviews; inductive codes were identified from analytically reading the transcripts. Transcripts were uploaded to and coded in MAXQDA V.12 (Release 12.3.2). Constant comparative analysis was used to compare experiences and perspectives between those who graduated and those who dropped out. This comparison was undertaken to understand how capabilities, motivations and opportunities changed over the course of their participation, and how this ultimately influenced programme retention.

Participant data on demographic and socioeconomic characteristics were collected at baseline ( table 1 ).

• View inline

Demographics of the FVRx participants*

Overall participant perspectives of the program

When asked about their main motivation for enrolling in the programme, most participants reported the desire to eat healthier and the desire to lose weight. ‘Motivation to enrol’ was one of the codes used in MAXQDA for the analysis, with subcodes of lose weight, eat healthy or doctor recommended. Of the 18 people interviewed, 8 or 44% mentioned enrolling in the programme to lose weight, and 11 or 61% mentioned enrolling to learn to eat healthy. When asked about the most useful thing they learnt in the programme, nearly all the respondents mentioned an improvement in their knowledge of nutrition, such as learning correct portion sizes or reading nutrition labels. Other participants reported enjoying meeting new people and having a sense of camaraderie and support from the group. Additionally, over half of the participants, including those who did not finish the classes, said they would like to enrol in the programme again if given the opportunity.

Participant capacity to sustain behaviour change

When asked about fruit and vegetable consumption since the programme ended, most respondents reported they continue to eat a good amount of fruits and vegetables ( Excerpts: ‘I’m beginning to start to like broccoli and been doing some kale’ and ‘Yes, I do a lot of salads and fruits…I am loving the fresh fruits’). The majority of participants reported that they continue to use the lessons they learnt in the healthy living and cooking classes when making food choices.

Of the 18, 15 or 83% respondents mentioned nutrition knowledge as a positive takeaway from the programme, and 15 of the 18 or 83% respondents also mentioned continuing to consume fruits and vegetables.

Constraints on program participation

Two participants mentioned that even though they were getting free food with the vouchers, it was still expensive ( Excerpt: ‘I had to pay a co-pay each time, and it just got too expensive…’). Others reported challenges in having transportation to attend the Healthy Living Classes ( Excerpt: ‘I wasn’t able at that time to have the transportation to go to all of them’). Another participant with mobility limitations had difficulty picking up their packaged fresh produce. Those who did not graduate cited their own or a family member’s poor health; out of pocket costs (ie, copays); lack of affordable transportation or parking; and/or inconvenient scheduling of the sessions. ‘Dropout/Missed Sessions Reasons’ code had a subcode of transportation/mobility, and four of the 18 or 22% of the respondents mentioned lack of transportation as their reason for not attending classes.

Barriers to maintaining a healthy diet among participants

When asked what they believe the biggest barrier to healthy eating is, the most commonly reported answer was cost (n=6) ( Excerpt: ‘…for people like me, that have so many medical bills…it’s easier to get the cheaper, unhealthy things…’). Another participant explained that her family often gets groceries from the food pantry, where the healthy options such as fresh produce are limited. Another reported barrier was finding the time to cook healthy meals, especially when working or caring for children. Over half of the respondents mentioned shopping at multiple stores in order to obtain the lowest prices ( Excerpt: ‘I shop at the cheapest store I can get it (fruits and vegetables) at’).

This evaluation reveals that most participants of the FVRx programme reported improved knowledge of nutrition and continue to consume fresh fruits and vegetables months after completion of the programme. However, FVRx participants continue to encounter barriers to maintaining a healthy diet with the most commonly reported barriers being the cost of fresh produce and competing priorities such as child care which prohibited time dedicated to healthy food preparation.

Lifestyle change interventions have been shown to be effective in the treatment and prevention of diet-related illnesses such as diabetes. 7 Similarly, other research has shown the use of goal setting and small groups to be promising tools in dietary behaviour modification, both of which are used in FVRx. 8 However, lifestyle change initiatives and health education may be ineffective in increasing healthy food consumption if they do not take into consideration other factors such as neighbourhood segregation, market strategies and poverty as important modifiers of accessibility. 9 In order to address the food insecurity in these low-income patients, we have to find ways to tackle the cost barriers they face when it comes to accessing healthier foods. Our FVRx programme attempts to integrate both health education and monetary incentives through vouchers, enabling improvement in participant knowledge of healthy eating and addressing any socioeconomic barriers to eating fresh fruits and vegetables during the intervention.

Without access to free fruits and vegetables through vouchers, consumption of fruits and vegetables continued to be met with challenges such as their cost and competing priorities that precluded time for healthy food preparation. This highlights the importance of incorporating strategies that equip participants with the knowledge and self-efficacy to continue healthy behaviours, even after the programme has ended. While the healthy living curriculum and cooking classes work to provide participants with those tools, conducting follow-up with participants at various intervals, via phone calls or hosting alumni events to serve as booster sessions, could be useful strategies to increase likelihood of continued behaviour.

There are a few limitations to this study. One limitation is that qualitative data were collected from a small sample of participants of the programme. However, this study was intended to be an evaluation of a pilot programme, and results will be used to inform expansion of the FVRx programme within our hospital.

Given the poverty status of many of our patients (figure 1), it is expected that many would have transient housing, possibly leading to the wrong number for three participants, and a disconnected telephone numbers for one another participant. Such social determinants might have also affected the ability to afford transportation to and from classes, as well as copays for the classes. We suspect these factors contributed to the high dropout rate (n=11) and the 44% non-response rate when calling patients 6 months postcompletion of the programme. This is potentially supported by our findings among the six respondents who mentioned cost as the biggest barrier; five of those were individuals who did not finish the programme. The interviews show that nearly all 18 of the respondents had the same motivation for starting the programme: to learn to eat better; however for those that did not ‘graduate’ (n=11), they reported current life circumstances as preventing them from completing the programme. This included health issues (their own or that of a family member), scheduling or difficulty with transportation to the programme site were reported by respondents as reasons for dropping out. These types of variables are not able to be addressed through the FVRx programming in the pilot phase of the programme, but should be researched and addressed in larger studies moving forward

Our multifaceted FVRx pilot programme enhanced participants’ nutritional knowledge and skills and continued consumption of fresh produce months after completion of the programme. However, socioeconomic factors remain as continual barriers to sustaining healthy eating. Additional efforts may be necessary to promote sustained healthy eating, such as skill building around gardening and growing fresh produce in the home. Using these types of innovative approaches may empower lower income populations to overcome barriers to healthy behaviour change. Efforts to improve participant retention in the programme, expand the programme to more participants and promote sustained behaviour change on programme completion are underway.

Acknowledgments

We would like to acknowledge Wholesome Wave Georgia, Project Open Hand, and The Common Market for their contributions to the FVRx program at our site. We are appreciative to Grady Memorial Hospital and the Primary Care Center for their innovative role in implementing systems change by supporting patient-centred group classes and FVRx prescriptions at our site.

• Coleman-Jensen A ,
• Rabbitt MP ,
• Gregory CA , et al
• Rabbitt M ,
• Laraia BA ,
• Oddo VM , et al
• Kreider B ,
• Pepper JV ,
• Ratcliffe C ,
• McKernan S-M ,
• Knowler WC ,
• Barrett-Connor E ,
• Fowler SE , et al
• Ammerman AS ,
• Lindquist CH ,
• Lohr KN , et al
• Azétsop J ,

Contributors All authors listed have contributed sufficiently to the project to be included as authors, and all those who are qualified to be authors are listed in the author byline. Authors’ contributions: EC conducted the study and the analysis for the study, and helped to write up the study. SRS (MD) gave idea for study and helped plan and conduct the study and helped write up the study. TLH (MD, MPH, MS, FACP) helped plan, developed and conducted the study along with helping write up the study. SB helped plan, developed and conducted the study along with helping to write up the study. GK (MD) helped plan the study and write up the study. JB-J (MD, FACP) helped plan and developed the study. AWG (PhD) supervised and assisted EC in conducting the study and analysing the study and helped to write up the study.

Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

Competing interests None declared.

Patient consent for publication Not required.

Ethics approval All study protocols, informed consent documents and tools were reviewed and approved by the hospital review board and deemed exempt from review by Emory University Institutional Review Board. All participants gave verbal informed consent to participate and provided permission to record the call.

Provenance and peer review Not commissioned; externally peer reviewed.

Read the full text or download the PDF:

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 18 January 2022

Large-scale diet tracking data reveal disparate associations between food environment and diet

• Tim Althoff   ORCID: orcid.org/0000-0003-4793-2289 1 ,
• Hamed Nilforoshan 2 ,
• Jenna Hua 3 , 4 &
• Jure Leskovec   ORCID: orcid.org/0000-0002-5411-923X 2 , 5  

Nature Communications volume  13 , Article number:  267 ( 2022 ) Cite this article

22k Accesses

31 Citations

114 Altmetric

Metrics details

• Epidemiology
• Risk factors

An unhealthy diet is a major risk factor for chronic diseases including cardiovascular disease, type 2 diabetes, and cancer 1 , 2 , 3 , 4 . Limited access to healthy food options may contribute to unhealthy diets 5 , 6 . Studying diets is challenging, typically restricted to small sample sizes, single locations, and non-uniform design across studies, and has led to mixed results on the impact of the food environment 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 . Here we leverage smartphones to track diet health, operationalized through the self-reported consumption of fresh fruits and vegetables, fast food and soda, as well as body-mass index status in a country-wide observational study of 1,164,926 U.S. participants (MyFitnessPal app users) and 2.3 billion food entries to study the independent contributions of fast food and grocery store access, income and education to diet health outcomes. This study constitutes the largest nationwide study examining the relationship between the food environment and diet to date. We find that higher access to grocery stores, lower access to fast food, higher income and college education are independently associated with higher consumption of fresh fruits and vegetables, lower consumption of fast food and soda, and lower likelihood of being affected by overweight and obesity. However, these associations vary significantly across zip codes with predominantly Black, Hispanic or white populations. For instance, high grocery store access has a significantly larger association with higher fruit and vegetable consumption in zip codes with predominantly Hispanic populations (7.4% difference) and Black populations (10.2% difference) in contrast to zip codes with predominantly white populations (1.7% difference). Policy targeted at improving food access, income and education may increase healthy eating, but intervention allocation may need to be optimized for specific subpopulations and locations.

Similar content being viewed by others

Population mobility data provides meaningful indicators of fast food intake and diet-related diseases in diverse populations

Integrating human activity into food environments can better predict cardiometabolic diseases in the United States

Assessing household lifestyle exposures from consumer purchases, the My Purchases cohort

Introduction.

Dietary factors significantly contribute to risk of mortality and chronic diseases such as cardiovascular diseases, type 2 diabetes and cancer globally 1 , 2 , 3 . Emerging evidence suggests that the built and food environment, behavioral, and socioeconomic factors significantly affect diet 7 . Prior studies of the food environment and diet have led to mixed results 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , and very few used nationally representative samples. These mixed results are potentially attributed to methodological limitations of small sample size, differences in geographic contexts, study population, and non-uniform measurements of both the food environment and diet across studies. Therefore, research with larger sample size and using improved and consistent methods and measurements is needed 9 , 24 , 25 .

Commercially available and widely used mobile applications allow the tracking of health behaviors and population health 26 , as recently demonstrated in physical activity 27 , 28 , sleep 29 , 30 , COVID-19 pandemic response 31 , 32 , 33 , women’s health 34 , as well as diet research 35 , 36 , 37 , 38 , 39 , 40 . With ever increasing smartphone ownership in the U.S. 41 and the availability of immense geospatial data, there are now unprecedented opportunities to combine various data on individual diets, population characteristics (gender and ethnicity), socioeconomic status (income and educational attainment), as well as food environment at large scale. Interrogation of these rich data resources to examine geographical and other forms of heterogeneity in the effect of food environments on health could lead to the development and implementation of cost-effective interventions 42 . Here, we leverage large-scale smartphone-based food journals of 1,164,926 participants across 9822 U.S. zip codes (Fig.  1 ) and combine several Internet data sources to quantify the independent associations of food (grocery and fast food) access, income and educational attainment with food consumption and body-mass index (BMI) status (Fig.  2 ). This study constitutes the largest nationwide study examining the relationship between the food environment and diet to date.

A choropleth showing the number of participants in each U.S. county. This country-wide observational study included 1,164,926 participants across 9822 U.S. zip codes that collectively logged 2.3 billion food entries for an average of 197 days each. This study constitutes the largest nationwide study examining the relationship between food environment and diet to date (e.g., with 511% more counties represented compared to BRFSS data 93 ).

A set of choropleths showing the main study outcomes of the number of entries that are classified as fresh fruit and vegetables, fast food, and soda consumption as well as the fraction affected by overweight/obesity (BMI > 25) participants across the USA by counties with more than 30 participants. We observe that food consumption healthfulness varies significantly across counties in the United States.

Data validation: diet tracking data correlates with existing large-scale measures

To determine the ability of our dataset to identify relationships between fast food, grocery store access, income, educational attainment and diet health outcomes, we confirmed that this studies’ smartphone-based food logs correlate with existing large-scale survey measures and purchase data. Specifically, the fraction of fresh fruits and vegetables (F&V) that participants logged is correlated with Behavioral Risk Factor Surveillance System (BRFSS) survey data 43 (Fig.  3 a; Pearson Correlation R  = 0.63, p  < 10 −5 ; Two-sided Student’s t -test; Methods). Further, the reported BMI of MyFitnessPal (MFP) participants is correlated with BRFSS survey data 44 (Fig.  3 b; Pearson Correlation R  = 0.78, p  < 10 −5 ; Two-sided Student’s t -test; Methods). Lastly, the digital food logs data replicate previous findings of relative consumption differences in low-income, low-access food deserts based on Nielsen purchase data 45 (Fig.  3 c; Pearson Correlation R  = 0.88, p  < 0.01; Two-sided Student’s t -test; Methods). These results demonstrate that smartphone-based food logs are highly correlated with existing, gold-standard survey measures and purchase data.

a Fraction of fresh fruits and vegetables logged is correlated with BRFSS survey data 43 (Pearson Correlation R  = 0.63, p  < 10 −5 ; Two-sided Student’s t -test; Methods). b Body-mass index of smartphone cohort is correlated with BRFSS survey data 44 (Pearson Correlation R  = 0.78, p  < 10 −5 ; Two-sided Student’s t -test; Methods). Lines in a , b show best linear fit along with shaded 95% bootstrap confidence intervals. c Digital food logs replicate previous findings of relative consumption differences in low-income, low-access food deserts based on Nielsen purchase data 45 (Pearson Correlation R  = 0.88, p  < 0.01; Two-sided Student’s t -test; Methods). These results demonstrate that smartphone-based food logs are highly correlated with existing, gold-standard survey measures and purchase data.

Associations between food environment, demographics and diet

Using these data across all 9822 U.S. codes, we found that high income, high educational attainment, high grocery store access, and low fast food access were independently associated with higher consumption of fresh F&V, lower consumption of fast food and soda, and lower prevalence of BMI levels categorized as overweight or obesity (Fig.  4 ; BMI > 25). The only exception to this pattern was a very slight (0.6%) positive difference in BMI levels categorized as overweight or obesity associated with income. Specifically, in zip codes of above median grocery store access participants logged 3.4% more F&V, 7.6% less fast food, 6.4% less soda and were 2.4% less likely to be affected by overweight or obesity (all P  < 0.001). In zip codes of below median fast food access participants logged 5.3% more F&V, 6.2% less fast food, 13.3% less soda and were 1.5% less likely to be affected by overweight or obesity (all P  < 0.001). In zip codes of above median education, participants logged 9.2% more F&V, 8.5% less fast food, 13.8% less soda and were 13.1% less likely to be affected by overweight or obesity (all P  < 0.001). Finally, in zip codes of above median household income (referred to as higher income below), participants logged 3.3% more F&V, 6.8% less fast food, 8.6% less soda (all P  < 0.001), but had a 0.6% higher likelihood of being affected by overweight or obesity ( P  = 0.006). Note that the reported effect size are based on comparing above and below median zip codes for any given factor. We found a general pattern of consistent, and in many cases higher effect sizes when comparing top versus bottom quartiles (Supplementary Fig.  2 ), suggesting the possibility of a dose-response relationships across most considered variables. We found that zip codes with high educational attainment levels compared to low educational attainment levels had the largest relative positive differences across F&V, fast food, soda, and BMI levels categorized as overweight or obesity.

Independent contributions of high income (median family income higher than or equal to $70,241), high educational attainment (fraction of population with college education 29.8% or higher), high grocery store access (fraction of population that is closer than 0.5 miles from nearest grocery store is greater than or equal to than 20.3%), and low fast food access (less than or equal to 5.0% of all businesses are fast-food chains) on relative difference in consumption of a fresh fruits and vegetables, b fast food, c soda, and d relative difference in fraction affected by overweight or obesity (BMI > 25). Cut points correspond to median values. Y -axes are oriented such that consistently higher is better. Estimates are based on matching experiments controlling for all but one treatment variable, across N  = 4911 matched pairs of zip codes (Methods). Bar height corresponds to mean values; error bars correspond to 95% bootstrap confidence intervals (Methods). While the most highly predictive factors vary across outcomes, only high educational attainment was associated with a sizeable difference of 13.1% in the fraction affected by overweight or obesity.

Significant differences across zip codes with predominantly Black, Hispanic, and Non-Hispanic white populations

We separately repeated our data analyses within zip codes with predominantly Black (3.7%), Hispanic (5.6%), and non-Hispanic white populations (78.4%) (Fig.  5 ). Results within zipcodes with predominantly non-Hispanic white populations closely resembled results within the overall population, since most zip codes in this study had predominantly white populations (78.4%; not unlike the overall U.S. population at 61.3%) 46 . However, restricting our analyses to zip codes with predominantly Black and Hispanic populations led to significantly different findings. Specifically, within zip codes with predominantly Black populations we found associations of higher income in the inverse direction of the population average and towards low healthful food consumption, across four out of four outcome variables, resulting in lower F&V consumption (−6.5%), higher fast food consumption (5.5%), and higher likelihood of BMI levels categorized as overweight or obesity (8.1%). Higher income was also associated with higher soda consumption (14.1%) but was not statistically significant ( P  = 0.061). On the other hand, low fast food access and high educational attainment access were generally associated with higher diet health, with low fast food access correlating with the highest significant negative difference in fast food consumption (−12.0%) and high educational attainment with the highest positive difference in fresh fruit and vegetable consumption (11.2%), although lower fast food access was associated with worse outcomes for one of the outcome variables. Specifically, lower fast food access was associated with a slightly higher likelihood of being affected by overweight or obesity (3.1%). Higher grocery store access had a positive association with diet health across all outcome variables in zip codes with predominantly Black populations, and was associated with higher F&V consumption (10.2%), lower fast food consumption (12.6%), lower likelihood of BMI levels categorized as overweight or obesity (9.0%), and lower soda consumption (5.3%), although the association with soda consumption was not statistically significant ( P  = 0.060).

Independent contributions of high income (median family income higher than or equal to $70,241), high educational attainment (fraction of population with college education 29.8% or higher), high grocery store access (fraction of population that is closer than 0.5 miles from nearest grocery store is greater than or equal to than 20.3%), and low fast food access (less than or equal to 5.0% of all businesses are fast-food chains) on relative difference in consumption of a fresh fruits and vegetables, b fast food, c soda, and d relative difference in fraction affected by overweight or obesity (BMI > 25). Cut points correspond to median values. Y -axes are oriented such that consistently higher is better. Estimates are based on matching experiments controlling for all but one treatment variable, across N  = 4277, 4102, 3510, 3205 matched pairs of non-Hispanic white-majority zip codes, treated on income, educational attainment, fast food access, grocery store access respectively; N  = 42, 74, 259, 259 matched pairs of Black-majority zip codes, treated on income, educational attainment, fast food access, grocery store access respectively; N  = 67, 61, 297, 471 matched pairs of Hispanic-majority zip codes, treated on income, educational attainment, fast food access, grocery store access respectively (Methods). Bar height corresponds to mean values; error bars correspond to 95% bootstrap confidence intervals (Methods). We observe significant differences in outcomes between zip codes with predominantly Black, Hispanic, and non-Hispanic white populations.

In contrast, within zip codes with predominantly Hispanic populations we found a significant association between higher, above-median, income and higher F&V consumption (5.7%), but not with the remaining three outcome variables. Zip codes with higher proportion of people with high educational attainment had the most positive association with diet health across all variables. Specifically, higher educational attainment was associated with higher F&V consumption (8.9%), lower fast food consumption (11.9%), lower soda consumption (16.5%), and lower likelihood of BMI levels categorized as overweight or obesity (13.7%). Higher grocery store access and lower fast food access had similar effect sizes as on the overall population for some outcome variables (i.e. similar associations with likelihood of BMI levels categorized as overweight or obesity and fast food consumption). However, in some cases the magnitude of association was higher (i.e. grocery store access was associated with 7.4% higher F&V consumption in areas with predominantly Hispanic population, which is more than twice than the difference within the overall population) and in others, unlike the overall population, there was no significant association (i.e. no significant relationship between fast food access on soda consumption, or between fast food access and F&V consumption).

Few factors were consistently associated with better outcomes across all three subpopulations. Across all three groups, F&V consumption was significantly higher in zip codes with high grocery store access and high educational attainment. Fast food consumption was lower across all potential intervention targets besides higher income. Soda consumption was lowest most with lower fast food access for Black and white-majority zip codes, whereas it was lowest with higher educational attainment in Hispanic zip codes. Lastly, BMI levels categorized as overweight or obesity were far lower with higher educational attainment levels compared to all other intervention targets, across all three groups.

Commercially available and widely used mobile applications and devices enable the individuals to track their own health, and in aggregate may inform our understanding of population health. These emerging data sources capture health behaviors from millions of participants 26 and have uniquely enabled large-scale research studies, including in physical activity 27 , 28 , sleep 29 , 30 , COVID-19 pandemic response 31 , 32 , 33 , women’s health 34 , as well as diet research 35 , 36 , 37 , 38 , 39 , 40 .

While many of our results were consistent with previous studies 47 , 48 , 49 , importantly, we found that zip codes with higher proportion of people with high educational attainment had the largest relative difference in the likelihood of BMI levels categorized as overweight or obesity (13.1% lower). It is well established that social determinants of health are linked to obesity 50 , 51 , 52 . As an important component of social determinants of health, our study suggests that having higher educational attainment is the most predictive of reduced overweight and obesity for all ethnicities.

When we restricted our analyses to zip codes with predominantly Black, Hispanic, and non-Hispanic white populations, we found the independent associations of food access, income and educational attainment with food consumption and BMI status varied significantly across these three groups. These findings suggest that tailored intervention strategies are needed based on neighborhood population distributions, assets and contexts.

Within zip codes with predominantly Black populations, the association between having higher income and diet health was negative. Having higher income was associated with lower F&V consumption, higher fast food and soda consumption, and higher likelihood of overweight and obesity. This could be explained by the “diminishing return hypothesis”, which suggests that Black people receive fewer protective health benefits from increases in SES than white people 53 , 54 . A combination of factors, including neighborhood economic disadvantage 55 , 56 , racial/ethnic discrimination 57 , 58 , and stress associated with educational attainment and mobility 59 , may prevent Black people from higher SES backgrounds from achieving their fullest health potential relative to white people 60 .

Within zip codes with predominantly Hispanic populations, higher income was not associated with lower likelihood of BMI levels categorized as overweight or obesity. The absence of a relationship between higher income and BMI, compared to in zip codes with predominantly Black and non-Hispanic white populations, could be partially explained by the “Hispanic health paradox” and “Hispanic health advantage” 61 , 62 , 63 , 64 , 65 , 66 . The Hispanic health paradox suggests that even though the first-generation Hispanic people have lower SES, they experience better health outcomes including lower prevalence of cardiovascular diseases, asthma, diabetes and cancer compared to those who were U.S.-born 61 , 62 , 63 . Hispanic health advantage suggests that Hispanic people have lower rates of harmful health behaviors, such as smoking, which in turn positively influence other health outcomes compared to non-Hispanic white people 61 , 64 , 65 , 66 . Additionally, through acculturation or adopting American culture, Hispanic immigrants may engage in less healthy behaviors, which in turn put themselves at higher risk for chronic diseases 61 , 62 , 63 , 67 , 68 , 69 , 70 , 71 .

While it is challenging to close the education and income gaps, establishing more grocery stores and limiting fast food restaurant access may help improve diet health across the population. Previous reviews suggested that government policies that addressing food affordability and purchase, such as the Healthy Food Financing Initiative (HFFI), increasing food stamp (SNAP) benefit and provide incentives to create healthy retail food environment have been effective in reducing food insecurity and dietary behaviors 72 , 73 , 74 , 75 , 76 , 77 . While several studies showed that the establishments of new supermarkets had little improvement in BMI 78 , 79 , 80 ; however, the investments in the new supermarkets have improved economic opportunity and social cohesion 81 , 82 , 83 . Our results showed that higher grocery store access was associated with 2–3 times higher fresh fruit and vegetable consumption and lower fast food consumption more for Black people than for white people. Although previous literature has shown null effects of grocery store access 84 , 85 , these studies have focused on the general population, which is white-skewed. Therefore, policies and strategies in increasing grocery store access and decreasing fast food access could potentially be the most effective approaches in changing dietary habits among locations with predominantly Black populations.

Furthermore, having more grocery store access and lower fast food access, in the food environment may work in synergistic ways that may lead to even lower obesity prevalence and obesity-related lifestyle and behavior changes. This is demonstrated in a recent study by Cantor et al. that HFFI boosted the effects of SNAP participation on improving food security and healthy food choices in food desserts 86 . This synergy could be multiplied when combining with effective education programs that could potentially lower obesity prevalence further by increasing individuals’ SES (e.g., income and educational attainment) 87 , 88 , health literacy and behaviors 87 , 88 , 89 , 90 , 91 , as well as sense of control and empowerment 92 .

Due to the cross-sectional nature of the study, we were not able to make any causal inferences between SES, food environment variables, dietary behavior, and BMI, as unobserved neighborhood and individual demographic and social characteristics could lead to confounding. However, we used a matching-based approach to mimic a quasi-experimental design to disentangle the individual associations of income, educational attainment and food access with participants’ food consumption. Our analysis did not include other demographic variables such as gender and age, as both variables were naturally balanced across treatment and control groups and we observed minimal zipcode-level correlations between age/gender and any of our four outcome measures (Supplementary Table  13 ). In addition, we confirmed that results were virtually identical (Pearson Correlation R  = 0.95), when explicitly controlling for age and gender in our matching-based approach. However, we jointly considered the potential impacts of neighborhood income, neighborhood educational attainment and food environment access on participants’ food consumption with consistent measures across the U.S., whereas previously published studies examined one or a few at a time. Our study population, based on a sample of MFP users, is an imperfect representation of the United States national population. Comparing our study population to nationally representative survey data, we found that our study population had significant overlap with the U.S. national population in terms of population demographics, educational attainment and BMI status, but that it was skewed towards women and higher income (Supplementary Table  3 ). We used individuals’ food loggings to estimate their consumption (specifically, the number of food entries as the logged amount consumed varied highly across foods without standardization; e.g., specifying weight, volume, or number). Food loggings may not capture what individuals actually ate and participants may be particularly motivated or care about their diet and weight. Importantly, we conducted multiple validation experiments through comparisons with high quality and highly representative datasets which demonstrated high correlations to gold-standard approaches (Fig.  3 ). The majority of food environment studies used screeners, food frequency questionnaires or 24-h recalls for dietary assessment, and very few used diaries 9 . In contrast, our participants logged their food intakes for an average of 197 days each. We also harnessed other large datasets such as Yelp to examine participants’ food environments. Considering both the strengths and limitations of this study, more research is needed especially based on longitudinal study design and detailed individual level data to enable causal inference and precise interpretation of the results.

In conclusion, we analyzed 2.3 billion food intake logs and BMIs from 1.2 million MFP smartphone app participants over 7 years across 9822 zip codes in relation to educational attainment, ethnicity, income, and food environment access. Our analyses indicated that higher access to grocery stores, lower access to fast food, higher income and educational attainment were independently associated with higher consumption of fresh F&V, lower consumption of fast food and soda, and lower likelihood of being affected by overweight or obesity, but that these associations varied significantly across zip codes with predominantly Black, Hispanic and white subpopulations. Policy targeted at improving food access, income and education may increase healthy eating. However, intervention allocation may need to be optimized for specific subpopulations and locations.

Study design and population

We conducted a United States countrywide cross-sectional study of participants’ self-reported food intake and BMI in relation to zip code level demographic (educational attainment, ethnicity), socioeconomic (income), and food environment factors (grocery store and fast food access) by combining datasets from MFP, US Census, USDA and Yelp.

Overall, this cross-sectional matching-based study analyzed 2.3 billion food intake logs from U.S. smartphone participants over 7 years across 9822 zip codes, which is 24% of overall USA zip codes (U.S. has a total of 41,692 zip codes). Participants were users of the MFP app, a free application for tracking caloric intake. We analyzed anonymized, retrospective data collected during a 7-year observation period between 2010 and 2016 that were aggregated to the zip code level. Comparing our study population to nationally representative survey data, we found that our study population had significant overlap with the U.S. national population in terms of population demographics, educational attainment and BMI status, but that it was skewed towards women and higher income (Supplementary Table  3 ). Our matching-based statistical methodology controls for observed biases between comparison groups in terms of income, educational attainment, grocery store access, and fast food access (Methods: Statistical Analysis). Data handling and analysis was conducted in accordance with MFP policies and with the guidelines of the Stanford University Institutional Review Board.

Study data: MyFitnessPal

We compute outcome measures of food consumption and BMI status from 2.3 billion food intake logs by a sample of 1,164,926 U.S. participants of the MFP smartphone application to quantify food consumption across 9822 zip codes. The scale and geographic distribution of our study participants, as well as our outcome measures, are illustrated in Figs.  1 and 2 respectively. To ensure participant privacy as well as reliability of our measures, we decided to only include zip codes in which we had access to 30 or more participant food logs, which reduced the dataset size from 27,027 zip codes (spanning 3117 counties) to the final 9822 zip codes (spanning 1730, or 55% of all counties in the United States). Nevertheless, the geographical breadth of this dataset far exceeds existing food surveys. For example, our final dataset contained 511% more counties than the BRFSS survey of 283 counties, with 370% more participants per county on average 93 . While size and coverage compare favorably to BRFSS, it is important to understand what is not covered by our study. Figure  1 illustrates that our study lacks representation in the Midwest of the USA as well as in Alaska. In our study data, we further observed under-representation of zip codes with majority non-white population (Supplementary Table  3 ) and rural zip codes (RUCA codes 7 through 10 94 , 95 ), as well as over-representation of high-income zip codes (median family income higher than $70,241).

During the observation period from January 1, 2010 to November 15, 2016, the average participant logged 9.30 entries into their digital food journal per day. The average participant used the app for 197 days. All participants in this sample used the app for at least 10 days. We classified the 2.3 billion food intake entries into three categories of public health interest, fresh F&V, fast food, and sugary non-diet soda, and excluded them from analysis if they did not match these categories. Our classification method is consistent with USDA MyPlate with one divergence of the exclusion of juices. The healthiness of juice as a fruit and vegetable serving is contested due to its sugar content and limited nutritional profile 96 , 97 , 98 . For more details on the definition of a food entry, our classification method, and the choice of outcome measure, see Details on outcome measures subsection in Methods.

We intentionally use a cross-sectional rather than longitudinal study design, since fine-grained and large-scale temporal data on changes in the food environment were not available.

Study data: demographic and socioeconomic factors

We obtained data on demographic and socioeconomic factors from CensusReporter 99 . Specifically, for each zip code in our data set we obtained median family income, fraction of population with college education (Bachelor’s degree or higher), and fraction of population that is white (not including Hispanic), Black, or Hispanic from the 2010 to 2014 American Community Survey’s census tract estimates 99 . While data were available only on zip code level, previous studies have shown that area-level income measures are meaningful for health outcomes and describe unique socioeconomic inequities 100 .

Study data: grocery store and fast food access

Grocery store access was defined as the fraction of population that is more than 0.5 miles away from a grocery store following the food desert status definitions from the USDA Food Access Research Atlas 101 . Contrary to the USDA definition, we found evidence that even in rural zipcodes, the fraction of population greater than 0.5 miles away from grocery stores has the strongest association with food consumption (compared to 10 and 20 miles away), and thus we used 0.5 miles as the threshold across rural and urban zipcodes (Methods: Details on food environment measures). We measured fast food access through the fraction of restaurants that are fast food restaurants within a sample from Yelp, querying the nearest 1000 businesses from the zip code’s center, up to a maximum radius of 40 km (25 miles). See subsections Data Validation and reproducing State-of-the-art Measures using Population-scale Digital Food Logs for details and validation of these objective food environment measures.

We release all data aggregated at zipcode level in order to enable validation, follow-up research, and use by policy makers.

Details on food environment measures

We obtained data on grocery store access (fraction of population that is more than 0.5 miles away from grocery store) and food desert status from the USDA Food Access Research Atlas 101 . A census tract is considered a food desert by the USDA if it is both low-income (defined by Department of Treasury’s New Markets Tax Credit program) and low-access, meaning at least 500 people or 30 percent of residents live more than 0.5 miles from a supermarket in urban areas (10 miles in rural areas) 45 .

Although the USDA uses different thresholds for urban and rural areas (0.5 and 10 miles respectively), we found that even in rural zipcodes (defined as USDA rural-urban continuum RUCA scores of 7 through 10 94 , 95 ), the fraction of population that is farther than 0.5 miles from grocery stores had the highest correlation to fruit and vegetable consumption (Pearson Correlation R  = −0.20), compared to 1 miles (Pearson Correlation R  = −0.17), 10 miles ( R  = −0.05), and 20 miles (Pearson Correlation R  = 0.03). This suggests that the fraction of the population farther than 0.5 miles from a grocery store has the strongest relationship with healthy food consumption, even in rural zipcodes . Hence, we decided used 0.5 miles distance as a standard measure of grocery store access for rural and urban zip codes, contrary to the USDA definition. We subsequently sanity checked for any downstream confounding of urbanicity in our primary matching experiment of above/below median grocery store access, and found a negligible difference (Standardized Mean Difference (SMD) of 0.18) in urbanicity between control and treatment, suggesting that the effect size was not due to grocery store distance functioning as a proxy for urbanicity, but rather directly due to differential grocery store access.

We aggregated these data from a census tract level to a zip code level using USPS Crosswalk data, which provides a list of all census tracts which overlap with a single zip code 102 . We related these data on census tract level to the zip code level by taking the weighted average of each census tract food environment measure (both grocery store access and food desert status), weighted by the number of people in the tract 102 . For instance, if zip code A overlapped with Census Tract A (2500 people, food desert) and Census Tract B (7500 people, not a food desert), the food desert measure of zip code A would be estimated as 25%. We defined the binary threshold for food desert, used in Fig.  3 , as 50% or higher.

We measured fast food access through the fraction of restaurants in a zip-code that are fast food restaurants. Data on local restaurants and businesses were obtained through the Yelp API 103 . For each zip code, we consider up to 1000 restaurant businesses that are nearest to the zip code center up to a distance of 40 km (67.8% of zip code queries resulted in 1000 restaurant businesses within 40 km; Yelp API results are restricted to 1000 results). This resulted in a varying sample radius depending on urbanicity. For example, Urban zipcodes (RUCA code of 1) had an average effective centroid size of 15 miles, which we calculated by taking the distance from the zipcode center to the furthest restaurant returned by Yelp. We further used Yelp-based environment variables that we expected not to influence food consumption, such as the availability of waterproofing services, countertop installers, or electronic stores, as null experiments to demonstrate discriminant validity of our statistical analysis pipeline (see Supplementary Fig.  3 ).

Details on outcome measures (food consumption and BMI status)

We used 2.3 billion food intake logs by a sample of 1,164,926 U.S. participants of the MFP smartphone application to quantify food consumption across 9822 zip codes. During the observation period from January 1, 2010 to November 15, 2016, the average participant logged 9.30 entries into their digital food journal per day. The average participant used the app for 197 days. All participants in this sample used the app for at least 10 days.

Clustering of food consumption observations within individuals and zip codes was handled through multiple levels of aggregation. First we aggregated within participant and day (i.e., someone eating a banana at breakfast and another for dinner), then we aggregated across all days with tracking within each participant, and then across all participants within one zip code. We computed non-parametric confidence intervals and p -values through bootstrapping with 1000 replications on zip code level (last level of aggregation) 104 .

The unit of analysis for each zipcode was the average number of daily entries per person. An entry is a single food consumption event logged in the app MFP. Each entry contains a separate food component (e.g., banana, yogurt, hamburger, ...), brand name (e.g., “Campbells”), description (e.g., “Chicken Soup”), serving size unit (e.g., “cup”), and number of servings (e.g., “1”). Supplementary Fig.  1 shows the application interface for logging a food entry (e.g., 1 Banana from Whole Foods). We decided to use entries based on the observation that there was little variance in the number of servings per food category logged by participants in a single entry, and since the amount consumed varied highly across foods without standardization (e.g., specifying weight, volume, or number). Participants typically log “standard portion sizes” of each food individually (e.g., one bowl of cereal, one banana) on the MFP app. For example, for participants that logged a banana, and listed the serving size as “Banana”, the median entry was for 1 banana, the mean was for 0.88 bananas, and 95% of food entries were for between 0.5 to 1.5 bananas. The MFP app strongly encourages this behavior through a large library of foods to log that follows these standard portion sizes.

We classified all entries into three categories using brand name and description, and three separate binary classifiers: fresh F&V (through a proprietary classifier by MFP which used key words in the brand name and description), fast food (if the brand name contained the name of a fast food chain listed in Supplementary Table  8 , and sugary (non-diet) soda (if the brand name contained the name of a soda drink listed in Supplementary Table  9 and the description did not contain “diet”, “lite”, “light”, or “zero”). In all cases, descriptions, as well fast food and soda drink keywords, were normalized by lower-casing and removing punctuation. Each binary classifier thus took a food entry as input (i.e., “Coca Cola, Diet Cherry Coke, 8oz”) and outputted a binary label (i.e., soda or NOT soda). Entries which were predicted to be in none of the three categories based on all three models were excluded from the study.

Our classification method for fresh F&V is consistent with USDA MyPlate. The only divergence from USDA MyPlate is that we intentionally excluded juices, for which MFP has a separate classifier, which does not separate sweetened juice drinks or sports drinks and 100% juice. For our definition of Fresh F&V, we chose to exclude juices because the healthiness of juice as a fruit and vegetable serving is contested 96 , 98 , as even 100% fruit juices are typically high in sugar and calories, and low in fiber, and vegetable juices are often mixed with other high-sugar ingredients. We thus took a conservative approach to estimating diet healthiness by excluding these food entries.

We evaluated the accuracy of each of the three binary classification model by estimating the precision (# True Positive / # Predicted Positive) from a random sample of 50 entries belonging to each category. Precision estimates are summarized in Supplementary Table  1 , and Supplementary Tables  10 , 11 , and 12 show random samples of 50 food items from all elements predicted to be in each category (where asterisk “*” indicates an incorrect prediction). Note that across 2.3 billion food logs it was not possible to measure recall, but were able to measure precision by manually inspecting the food entry brand and description and assigning it a category.

We then calculated the average number of food entries logged per participant, per day, for each of the F&V, fast food, and soda categories (e.g. average number of F&V logged per participant per day), excluding days in which the participant was inactive (i.e., consistently did not log anything). Finally, we aggregated these participant-level measures to the zip code level by taking the mean of each category’s measure for all participants in each zip code. We further used BMI health in each zip code as a BMI status outcome, specifically the fraction of participants in a zip code which are affected by overweight or obesity (BMI > 25). BMI was self-reported by participants of the smartphone application (99.92% of participants did report BMI). Supplementary Table  2 shows basic summary statistics for the outcome measures used in this study. In our statistical analyses, we compared two sets of zip codes that differ in a dimension of interest (e.g., grocery store access access) as treatment and control group and use the relative difference in F&V consumption, fast food consumption, soda consumption, and BMI health of the treated group relative to the control group. To generate confidence intervals, as well as to compute p -values to test for statistical significance of differences in outcome, we use non-parametric bootstrap resampling with 1000 replications 104 . Specifically, we follow the method proposed by Austin and Small 105 , which is to draw bootstrap samples post-matching from the matched pairs in the propensity-score-matched sample after the Genetic Matching stage 106 . We confirmed the validity of this method empirically by also calculating t -tests for each experiment, which gave qualitatively similar results. We note that we perform bootstrapping on zip code level (highest level of aggregation). While, multilevel bootstrapping methods exist, they do not scale well with our dataset size of 2.3 billion food items. However, due to the large number of 9822 zip codes our analyses are well-powered statistically even with bootstrapping at zip code level.

Data validation

We find that our study population has significant overlap with the U.S. national population (Supplementary Table  3 ) but is skewed towards women and higher income. We demonstrate that food consumption measured based on this population are highly correlated with state-of-the-art measures (Fig.  3 ). Smartphone apps such as MFP feature large databases with nutritional information and can be used to track one’s diet over time. Previous studies have compared app-reported diet measures to traditional measures including 24-h dietary recalls and food composition tables. These studies found that both measures tend to be highly correlated 107 , 108 , but that app-reported measures tend to underestimate certain macro- and micronutrients 107 , 108 , especially in populations that were previously unfamiliar with the smartphone applications 109 . In contrast, this study leverages a sample of existing participants of the smartphone app MFP. Yelp data has been used in measures of food environment 110 and a study in Detroit found Yelp data to be more accurate than commercially available databases such as Reference USA 111 . This study uses a combination of MFP data to capture food consumption, Yelp, and USDA data to capture food environment, and Census data to capture basic demographics. As a preliminary, basic test, we investigated correlations between the Mexican food consumption, the fraction of Mexican restaurants, and the fraction of Hispanic people in the population, on a zip code level. We found that Mexican food consumption (entries labeled as Mexican food by a proprietary MFP classifier, logged per participant, per day) was correlated with the fraction of Mexican restaurants (Pearson Correlation R  = 0.72; < 10 −4 ) and the fraction of Hispanic people in the population (Pearson Correlation R  = 0.54; P  < 10 −4 ). Further, the fraction of Mexican restaurants was correlated with the fraction of Hispanic people in the population as well (Pearson Correlation R  = 0.51; P  < 10 −4 ).

Reproducing state-of-the-art measures using population-scale digital food logs

A primary concern in studying diet health via food logs is the bias inherent to the MFP population, which is not a representative sample of the US population. To investigate the applicability of population-scale digital food logs to study the relationship between food environment, income and educational attainment with food consumption, we measured the correlation between our smartphone app-based measures and state-of-the-art measures of food consumption including the BRFSS, based on representative surveys of over 350,000 adults in the United States 43 , 44 , and the Nielsen Homescan data 112 , which is a nationally representative panel survey of the grocery purchases of 169,000 unique households across the United States, based on UPC records of all consumer packaged goods participants purchased from any outlet (Fig.  3 ). We used the latest survey data from BRFSS 43 , 44 available at the county-level. Specifically, we used variables FV5SRV from BRFSS 2011 representing the faction of people eating five or more servings of fresh fruit and vegetables 43 , and BMI5 from BRFSS 2012 representing BMI 44 . We compare against BRFSS rather than National Health and Nutrition Examination Survey (NHANES), since BRFSS is significantly larger than NHANES, it is remotely administered matching our study, and it has much better geographical coverage than NHANES and geographical comparisons are central to our study. Despite these advantages, no reference dataset is without limitations 113 , 114 , 115 , motivating this study’s use of large-scale digital food journals.

Comparing our data to BRFSS on county level, we found moderate to high correlations between the amount of fresh F&V consumed (Fig.  3 a, Pearson Correlation R  = 0.63, p  < 10 −5 ) and BMI (Fig.  3 b, Pearson Correlation R  = 0.78, p  < 10 −5 ). We further compared to published results by the USDA 45 , which used data from the 2010 Nielsen Homescan Panel Survey that captured household food purchases for in-home consumption (but did not capture restaurants and fast food purchases). We attempted to reproduce published findings on the differences in low-income, low-access communities (food deserts) compared to non-low-income, non-low-access communities 45 across categories of fruit, vegetable, sweets, red meat, fish/poultry, milk products, diet drinks, and non-diet drinks (Table 4 in Rahkovsky and Snyder 45 ). We used proprietary MFP classifiers to categorize foods logged into these categories. We found that our app-based food logs were very highly correlated with previously published results (Fig.  3 c, R  = 0.88, p  < 0.01) and that the absolute differences between food deserts and non-food deserts were stronger in the MFP data compared to Nielsen purchase data. Overall, these results demonstrate convergent validity and suggest that the employed non-representative sample of population-scale digital food logs can reproduce the basic dynamics of traditional, state-of-the-art measures, and they can do so at massive scale and comparatively low cost.

Statistical analysis

In this large-scale observational study, we used a matching-based approach 116 , 117 to disentangle contributions of income, educational attainment, grocery store access, and fast food access on food consumption. We considered multiple statistical strategies, including regression modeling and propensity score matching. We decided to employ a full matching on all variables, which avoids parametric assumptions and is a more conservative method for matching than for example propensity score-based techniques 117 . To estimate the treatment effects of each of these factors, we divided all available zip codes into treatment and control groups based on a median split; that is, we estimated the difference in outcomes between matched above-median and below-median zip codes. We created matched pairs of zip codes by selecting a zip code in the control group that is closely matched (i.e., less than 0.25 SMD between the treated and control groups) 117 to the zip code in the treatment group across all factors, except the treatment factor of interest. Since we repeated this matching process for each zip code in the treatment group, this approach estimated the Average Treatment Effect on the Treated (ATT). Through this process, we attempted to eliminate variation of plausible influences and to isolate the effect of interest. We repeated this process for each treatment of interest; for example for the results presented in Fig.  4 , we performed four matchings, one for each of income, educational attainment, grocery store access and fast food access. For the sub-population experiments (Fig.  5 ), we repeated the same method on the subset of the zip codes in which the majority of inhabitants were of a particular ethnic group. Lastly, although we considered controlling for age and gender in the matching procedure, as these are related to diet health at the individual-level, we did not include them in our final analysis after observing (1) minimal zipcode-level correlations between age/gender and any of our four outcome measures (Supplementary Table  13 ; largest Pearson Correlation was 0.12) and (2) virtually identical results (Pearson Correlation R  = 0.95) when comparing before and after controlling for age and gender by adding them the genetic matching algorithm. See subsection on Details on Matching Approach for further details and statistics that demonstrate that treatment and control groups were well-balanced on observed covariates after matching.

We tested discriminant validity of our statistical approach by measuring the effect of null-treatments that should not have any impact on food consumption. We chose examples of null-treatments by selecting variables that had little correlation with study independent variables (income, educational attainment, grocery store access, fast food access) and were plausibly unrelated to food consumption. This selection process lead to use of the fraction of countertop installers, electronics stores, and waterproofing services nearby as measured through Yelp. Applying our analysis pipeline to these null-treatments, we found that all of these null-treatments had zero effect on food consumption. This demonstrated that our statistical analysis approach did not produce measurements that it was not supposed to measure; that is, discriminant validity (Supplementary Fig.  3 and Supplementary Table  7 ).

Details on matching approach

Specifically, we use a one-to-one Genetic Matching approach, 106 with replacement, and use the mean of the SMD between treatment and control groups, across all matched variables, as the Genetic Matching balance metric in order to maximize balance (overlap) between the treated and the control units. Some definitions of SMD use the standard deviation in the overall population before matching 116 . However, we choose the standard deviation in the control group post-matching, which typically is much smaller and therefore gives more conservative estimates of balance between treated and control units 118 .

After matching, we evaluated the quality of balance between the treated and the control units by the SMD across each of the variables that were controlled for and included in the matching process. A good balance between treated and control groups was defined as a SMD of less than 0.25 standard deviations 117 across each variable. By default, we do not enforce a caliper in order to minimize bias in matching process, although in rare cases in which a good balance was not achieved, a caliper was enforced, starting at 2.5 standard deviations between matched and controlled units, and decreased by 0.1 until the matched and control groups had a SMD smaller than 0.25 across all matched variables.

For the vast majority of matching experiments the SMD across all matched variables was well below 0.25, with a mean of 0.040 and median of 0.016 for the four overall population matching experiments. The SMD for the ethnicity-majority zipcode experiments was slightly higher, but still very significantly below 0.25 across all 12 experiments, with a mean of 0.055 and median of 0.036. Thus, no caliper was necessary to ensure a good balance, with the exception of one out of the 12 of sub-population experiments (white, high educational attainment). Detailed balancing statistics for each of the matches are available in the Supplementary Information (Supplementary Tables  14 – 36b ), as well as a supplementary matching experiment in which a top/bottom quartile split was used instead of a median split (Supplementary Fig.  2 ).

Details on the use of zip codes

A zip code is a postal code used by the US Postal Services. Zip codes consist of 5 digits and were introduced in their current form in 1983 in order to provide granular demarcations of US geography for mail purposes 119 . Most previous surveys such as BRFSS aggregate individuals at the less fine-grained levels of granularity: city, county, or MSA (Metropolitan statistical area) level. By contrast, we chose to use zip codes in order to study diet health and obesity at a more fine-grained level of analysis. As a point of reference, there are currently 41,692 zip codes in the USA compared to 3143 counties and county equivalents (i.e., 13.2 zip codes per county on average). Zip codes are on average 91 square miles and contain 7872 people 120 , compared to counties and county-equivalents which are on average 1208 square miles and contain 104,422 people 121 . Neighboring zip codes which may be in the same county have sharply contrasting demographics 122 . A zip code-level analysis better enables us to measure the disparate impacts of educational attainment, income, and food environment on diet health and obesity, and to stratify our analyses by ethnicity.

The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or sponsors.

Reporting summary

Further information on research design is available in the  Nature Research Reporting Summary linked to this article.

Data availability

Data are available at http://snap.stanford.edu/dietdisparities .

Code availability

Code is available at http://snap.stanford.edu/dietdisparities .

Afshin, A. et al. Health effects of dietary risks in 195 countries, 1990–2017: a systematic analysis for the global burden of disease study 2017. Lancet 393 , 1958–1972 (2019).

Article   Google Scholar  

Micha, R. et al. Association between dietary factors and mortality from heart disease, stroke, and type 2 diabetes in the united states. JAMA 317 , 912–924 (2017).

Article   PubMed   PubMed Central   Google Scholar  

Jayedi, A., Soltani, S., Abdolshahi, A. & Shab-Bidar, S. Healthy and unhealthy dietary patterns and the risk of chronic disease: an umbrella review of meta-analyses of prospective cohort studies. Br. J. Nutr. 124 , 1133–1144 (2020).

Article   CAS   PubMed   Google Scholar  

Beaglehole, R. et al. Priority actions for the non-communicable disease crisis. Lancet 377 , 1438–1447 (2011).

Article   PubMed   Google Scholar  

Fleischer, N. L., Roux, A. V. D., Alazraqui, M. & Spinelli, H. Social patterning of chronic disease risk factors in a latin american city. J. Urban Health 85 , 923 (2008).

Sallis, J. F. & Glanz, K. The role of built environments in physical activity, eating, and obesity in childhood. Future Child. 16 , 89–108 (2006).

Fleischhacker, S. E., Evenson, K. R., Rodriguez, D. A. & Ammerman, A. S. A systematic review of fast food access studies. Obes. Rev. 12 , e460–e471 (2011).

Odoms-Young, A., Singleton, C. R., Springfield, S., McNabb, L. & Thompson, T. Retail environments as a venue for obesity prevention. Curr. Obes. Rep. 5 , 184–191 (2016).

Kirkpatrick, S. I. et al. Dietary assessment in food environment research: a systematic review. Am. J. Preventive Med. 46 , 94–102 (2014).

Caspi, C. E., Sorensen, G., Subramanian, S. & Kawachi, I. The local food environment and diet: a systematic review. Health Place 18 , 1172–1187 (2012).

Feng, J., Glass, T. A., Curriero, F. C., Stewart, W. F. & Schwartz, B. S. The built environment and obesity: a systematic review of the epidemiologic evidence. Health Place 16 , 175–190 (2010).

Cummins, S. & Macintyre, S. Food environments and obesity-"neighbourhood or nation? Int. J. Epidemiol. 35 , 100–104 (2006).

Charreire, H. et al. Measuring the food environment using geographical information systems: a methodological review. Public Health Nutr. 13 , 1773–1785 (2010).

Kelly, B., Flood, V. M. & Yeatman, H. Measuring local food environments: an overview of available methods and measures. Health Place 17 , 1284–1293 (2011).

McKinnon, R. A., Reedy, J., Morrissette, M. A., Lytle, L. A. & Yaroch, A. L. Measures of the food environment: a compilation of the literature, 1990–2007. Am. J. Preventive Med. 36 , S124–S133 (2009).

Elinder, L. S. & Jansson, M. Obesogenic environments–aspects on measurement and indicators. Public Health Nutr. 12 , 307–315 (2009).

PubMed   Google Scholar  

Gittelsohn, J. & Sharma, S. Physical, consumer, and social aspects of measuring the food environment among diverse low-income populations. Am. J. Preventive Med. 36 , S161–S165 (2009).

Gustafson, A., Hankins, S. & Jilcott, S. Measures of the consumer food store environment: a systematic review of the evidence 2000–2011. J. Community health 37 , 897–911 (2012).

Lytle, L. A. Measuring the food environment: state of the science. Am. J. Preventive Med. 36 , S134–S144 (2009).

Odoms-Young, A. M., Zenk, S. & Mason, M. Measuring food availability and access in african-american communities: implications for intervention and policy. Am. J. Preventive Med. 36 , S145–S150 (2009).

Ohri-Vachaspati, P. & Leviton, L. C. Measuring food environments: a guide to available instruments. Am. J. Health Promotion 24 , 410–426 (2010).

Sharkey, J. R. Measuring potential access to food stores and food-service places in rural areas in the us. Am. J. Preventive Med. 36 , S151–S155 (2009).

Kamphuis, C. B. et al. Environmental determinants of fruit and vegetable consumption among adults: a systematic review. Br. J. Nutr. 96 , 620–635 (2006).

CAS   PubMed   Google Scholar  

Lytle, L. A. & Sokol, R. L. Measures of the food environment: a systematic review of the field, 2007–2015. Health Place 44 , 18–34 (2017).

Kumanyika, S. K. Environmental influences on childhood obesity: ethnic and cultural influences in context. Physiol. Behav. 94 , 61–70 (2008).

Hicks, J. L. et al. Best practices for analyzing large-scale health data from wearables and smartphone apps. NPJ Digital Med. 2 , 1–12 (2019).

Althoff, T., Hicks, J. L., King, A. C., Delp, S. L. & Leskovec, J. Large-scale physical activity data reveal worldwide activity inequality. Nature 547 , 336–339 (2017).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Althoff, T., White, R. W. & Horvitz, E. Influence of pokémon go on physical activity: study and implications. J. Med. Internet Res. 18 , e315 (2016).

Althoff, T., Horvitz, E., White, R. W. & Zeitzer, J. Harnessing the web for population-scale physiological sensing: a case study of sleep and performance. In Proceedings of the 26th international conference on World Wide Web . 113–122 (International World Wide Web Conferences Steering Committee, 2017).

Althoff, T., Horvitz, E. & White, R. W. Psychomotor function measured via online activity predicts motor vehicle fatality risk. NPJ Digital Med. 1 , 1–2 (2018).

Bento, A. I. et al. Evidence from internet search data shows information-seeking responses to news of local covid-19 cases. Proc. Natl Acad. Sci. 117 , 11220–11222 (2020).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Chang, S. et al. Mobility network models of covid-19 explain inequities and inform reopening. Nature 589 , 82–87 (2021).

Article   ADS   CAS   PubMed   Google Scholar  

Suh, J., Horvitz, E., White, R. W. & Althoff, T. Population-scale study of human needs during the covid-19 pandemic: Analysis and implications. In Proceedings of the 14th ACM International Conference on Web Search and Dat a Mining . 4–12 (Association for Computing Machinery, 2021).

Pierson, E., Althoff, T., Thomas, D., Hillard, P. & Leskovec, J. Daily, weekly, seasonal and menstrual cycles in women’s mood, behaviour and vital signs. Nat. Hum. Behav. 5 , 716–725 (2021).

Aiello, L. M., Schifanella, R., Quercia, D. & Del Prete, L. Large-scale and high-resolution analysis of food purchases and health outcomes. EPJ Data Sci. 8 , 14 (2019).

West, R., White, R. W. & Horvitz, E. From cookies to cooks: Insights on dietary patterns via analysis of web usage logs. In Proceedings of the 22nd international conference on World Wide Web . 1399–1410 (International World Wide Web Conferences Steering Committee, 2013).

Abbar, S., Mejova, Y. & Weber, I. You tweet what you eat: Studying food consumption through twitter. In Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems . 3197–3206 (Association for Computing Machinery, 2015).

De Choudhury, M., Sharma, S. & Kiciman, E. Characterizing dietary choices, nutrition, and language in food deserts via social media. In Proceedings of the 19th ACM conference on computer-supported cooperative work & social computing . 1157–1170 (Association for Computing Machinery, 2016).

Gordon, M., Althoff, T. & Leskovec, J. Goal-setting and achievement in activity tracking apps: a case study of MyFitnessPal. In The World Wide Web Conference . 571–582 (International World Wide Web Conferences Steering Committee, 2019).

Allcott, H. et al. Food deserts and the causes of nutritional inequality. Q. J. Econ. 134 , 1793–1844 (2019).

Article   MATH   Google Scholar  

Sasson, C. et al. American heart association diabetes and cardiometabolic health summit: Summary and recommendations. J. Am. Heart Assoc. 7 , e009271 (2018).

Mason, K. E., Pearce, N. & Cummins, S. Associations between fast food and physical activity environments and adiposity in mid-life: cross-sectional, observational evidence from uk biobank. Lancet Public Health 3 , e24–e33 (2018).

CDC. Behavioral risk factor surveillance system survey questionnaire (2011). https://www.cdc.gov/brfss/smart/smart_2011.htm . Accessed 18 Aug 2017.

CDC. Behavioral risk factor surveillance system survey questionnaire (2012). https://www.cdc.gov/brfss/smart/smart_2012.htm . Accessed 18 Aug 2017.

Rahkovsky, I. & Snyder, S. Food choices and store proximity (2015). https://www.ers.usda.gov/webdocs/publications/45432/53943_err195.pdf?v=42276/ . Accessed 31 Dec 2019.

USCB. Quickfacts (2017). https://www.census.gov/quickfacts/fact/table/US/PST045219 .

Larson, N. I., Story, M. T. & Nelson, M. C. Neighborhood environments: disparities in access to healthy foods in the us. Am. J. Preventive Med. 36 , 74–81 (2009).

Mackenbach, J. D. et al. A systematic review on socioeconomic differences in the association between the food environment and dietary behaviors. Nutrients 11 , 2215 (2019).

Article   PubMed Central   Google Scholar  

Meyer, K. A. et al. Sociodemographic differences in fast food price sensitivity. JAMA Intern. Med. 174 , 434–442 (2014).

Lakerveld, J. & Mackenbach, J. The upstream determinants of adult obesity. Obes. Facts 10 , 216–222 (2017).

Cockerham, W. C., Hamby, B. W. & Oates, G. R. The social determinants of chronic disease. Am. J. Preventive Med. 52 , S5–S12 (2017).

Yusuf, Z. I. et al. Social determinants of overweight and obesity among children in the united states. Int. J. Matern. Child Health AIDS 9 , 22 (2020).

Farmer, M. M. & Ferraro, K. F. Are racial disparities in health conditional on socioeconomic status? Soc. Sci. Med. 60 , 191–204 (2005).

Shuey, K. M. & Willson, A. E. Cumulative disadvantage and black-white disparities in life-course health trajectories. Res. Aging 30 , 200–225 (2008).

Do, D. P. et al. Does place explain racial health disparities? quantifying the contribution of residential context to the black/white health gap in the united states. Soc. Sci. Med. 67 , 1258–1268 (2008).

Williams, D. R. & Collins, C. Us socioeconomic and racial differences in health: patterns and explanations. Annu. Rev. Sociol. 21 , 349–386 (1995).

Bratter, J. L. & Gorman, B. K. Is discrimination an equal opportunity risk? racial experiences, socioeconomic status, and health status among black and white adults. J. Health Soc. Behav. 52 , 365–382 (2011).

Williams, D. R., Neighbors, H. W. & Jackson, J. S. Racial/ethnic discrimination and health: findings from community studies. Am. J. Public Health 93 , 200–208 (2003).

Pearson, J. A. Can’t buy me whiteness: New lessons from the titanic on race, ethnicity, and health. Du Bois Rev.: Soc. Sci. Res. Race 5 , 27–47 (2008).

Boen, C. The role of socioeconomic factors in black-white health inequities across the life course: Point-in-time measures, long-term exposures, and differential health returns. Soc. Sci. Med. 170 , 63–76 (2016).

Lommel, L. L., Thompson, L., Chen, J.-L., Waters, C. & Carrico, A. Acculturation, inflammation, and self-rated health in mexican american immigrants. J. Immigr. Minority Health 21 , 1052–1060 (2019).

Bostean, G. Does selective migration explain the hispanic paradox? a comparative analysis of mexicans in the us and mexico. J. Immigr. Minority Health 15 , 624–635 (2013).

Ruiz, J. M., Steffen, P. & Smith, T. B. Hispanic mortality paradox: a systematic review and meta-analysis of the longitudinal literature. Am. J. Public Health 103 , e52–e60 (2013).

Min, J. W., Rhee, S., Lee, S. E., Rhee, J. & Tran, T. Comparative analysis on determinants of self-rated health among non-hispanic white, hispanic, and asian american older adults. J. Immigr. Minority Health 16 , 365–372 (2014).

Kimbro, R. T., Gorman, B. K. & Schachter, A. Acculturation and self-rated health among latino and asian immigrants to the united states. Soc. Probl. 59 , 341–363 (2012).

Brewer, J. V. et al. Contributors to self-reported health in a racially and ethnically diverse population: focus on hispanics. Ann. Epidemiol. 23 , 19–24 (2013).

Article   MathSciNet   PubMed   Google Scholar  

Ridker, P. M. C-reactive protein: a simple test to help predict risk of heart attack and stroke. Circulation 108 , e81–e85 (2003).

Viruell-Fuentes, E. A. Beyond acculturation: immigration, discrimination, and health research among mexicans in the united states. Soc. Sci. Med. 65 , 1524–1535 (2007).

Rodriguez, F. et al. Association of educational attainment and cardiovascular risk in hispanic individuals: Findings from the cooper center longitudinal study. JAMA Cardiol. 4 , 43–50 (2019).

Lutsey, P. L. et al. Associations of acculturation and socioeconomic status with subclinical cardiovascular disease in the multi-ethnic study of atherosclerosis. Am. J. Public Health 98 , 1963–1970 (2008).

Koya, D. L. & Egede, L. E. Association between length of residence and cardiovascular disease risk factors among an ethnically diverse group of united states immigrants. J. Gen. Intern. Med. 22 , 841–846 (2007).

Moran, A. et al. Financial incentives increase purchases of fruit and vegetables among lower-income households with children. Health Aff. 38 , 1557–1566 (2019).

Moran, A. J. et al. Associations between governmental policies to improve the nutritional quality of supermarket purchases and individual, retailer, and community health outcomes: An integrative review. Int. J. Environ. Res. Public Health 17 , 7493 (2020).

Phipps, E. J. et al. Impact of a rewards-based incentive program on promoting fruit and vegetable purchases. Am. J. Public Health 105 , 166–172 (2015).

Steele-Adjognon, M. & Weatherspoon, D. Double up food bucks program effects on snap recipients’ fruit and vegetable purchases. BMC Public Health 17 , 1–7 (2017).

Wilde, P., Klerman, J. A., Olsho, L. E. & Bartlett, S. Explaining the impact of usda’s healthy incentives pilot on different spending outcomes. Appl. Economic Perspect. Policy 38 , 655–672 (2015).

Polacsek, M. et al. A supermarket double-dollar incentive program increases purchases of fresh fruits and vegetables among low-income families with children: the healthy double study. J. Nutr. Educ. Behav. 50 , 217–228 (2018).

Cummins, S., Flint, E. & Matthews, S. A. New neighborhood grocery store increased awareness of food access but did not alter dietary habits or obesity. Health Aff. 33 , 283–291 (2014).

Dubowitz, T. et al. Changes in diet after introduction of a full service supermarket in a food desert. Health Aff. 34 , 1858 (2015).

Zhang, Y. T. et al. Is a reduction in distance to nearest supermarket associated with bmi change among type 2 diabetes patients? Health Place 40 , 15–20 (2016).

Rogus, S., Athens, J., Cantor, J. & Elbel, B. Measuring micro-level effects of a new supermarket: do residents within 0.5 mile have improved dietary behaviors? J. Acad. Nutr. Dietetics 118 , 1037–1046 (2018).

Chrisinger, B. A mixed-method assessment of a new supermarket in a food desert: contributions to everyday life and health. J. Urban Health 93 , 425–437 (2016).

Giang, T., Karpyn, A., Laurison, H. B., Hillier, A. & Perry, R. D. Closing the grocery gap in underserved communities: the creation of the pennsylvania fresh food financing initiative. J. Public Health Manag. Pract. 14 , 272–279 (2008).

Jack, D. et al. Socio-economic status, neighbourhood food environments and consumption of fruits and vegetables in new york city. Public health Nutr. 16 , 1197–1205 (2013).

Strome, S., Johns, T., Scicchitano, M. J. & Shelnutt, K. Elements of access: the effects of food outlet proximity, transportation, and realized access on fresh fruit and vegetable consumption in food deserts. Int. Q. Community Health Educ. 37 , 61–70 (2016).

Cantor, J. et al. Snap participants improved food security and diet after a full-service supermarket opened in an urban food desert: Study examines impact grocery store opening had on food security and diet of supplemental nutrition assistance program participants living in an urban food desert. Health Aff. 39 , 1386–1394 (2020).

Chandola, T., Clarke, P., Morris, J. & Blane, D. Pathways between education and health: a causal modelling approach. J. R. Stat. Soc.: Ser. A 169 , 337–359 (2006).

Article   MathSciNet   Google Scholar  

Cutler, D. M. & Lleras-Muney, A. Education and health: evaluating theories and evidence. Working Paper Series, Working Paper 12352 National Bureau of Economic Research. http://www.nber.org/papers/w12352 (2006).

Kenkel, D. S. Health behavior, health knowledge, and schooling. J. Political Econ. 99 , 287–305 (1991).

Fletcher, J. M. & Frisvold, D. E. Higher education and health investments: does more schooling affect preventive health care use? J. Hum. Cap. 3 , 144–176 (2009).

Lleras-Muney, A. The relationship between education and adult mortality in the united states. Rev. Economic Stud. 72 , 189–221 (2005).

Mirowsky, J. Education, Social Status, and Health (Routledge, 2017).

CDC. Behavioral risk factor surveillance system survey questionnaire. https://www.cdc.gov/brfss . Accessed 18 Aug 2017.

Weeks, W. B. et al. Differences in health-related quality of life in rural and urban veterans. Am. J. Public Health 94 , 1762–1767 (2004).

Cromartie, J. Rural-urban commuting area codes (2020). https://www.ers.usda.gov/data-products/rural-urban-commuting-area-codes.aspx .

Ferris, H. A., Isganaitis, E. & Brown, F. Time for an end to juice in the special supplemental nutrition program for women, infants, and children. JAMA pediatrics 171 , 509–510 (2017).

Caswell, H. The role of fruit juice in the diet: an overview. Nutr. Bull. 34 , 273–288 (2009).

Guasch-Ferré, M. & Hu, F. B. Are fruit juices just as unhealthy as sugar-sweetened beverages? JAMA Netw. Open 2 , e193109–e193109 (2019).

Bureau, U. C. American community survey 5-year estimates (2014). http://censusreporter.org . Accessed 30 Aug 2017.

Buajitti, E., Chiodo, S. & Rosella, L. C. Agreement between area-and individual-level income measures in a population-based cohort: implications for population health research. SSM-Popul. Health 10 , 100553 (2020).

Rhone, A. Food access research atlas documentation (2015). https://www.ers.usda.gov/data-products/food-access-research-atlas/documentation/ . Accessed 20 Aug 2017.

Din, A. HUD USPS Zip Code Crosswalk Files (2017). https://www.huduser.gov/portal/datasets/usps_crosswalk.html . Accessed 10 Aug 2017.

Yelp. Yelp Fusion V3 API (2017). https://www.yelp.com/dataset/documentation/main/ . Accessed 30 Aug 2017.

Efron, B. & Tibshirani, R. J. An introduction to the bootstrap (CRC press, 1994).

Austin, P. C. & Small, D. S. The use of bootstrapping when using propensity-score matching without replacement: a simulation study. Stat. Med. 33 , 4306–4319 (2014).

Article   MathSciNet   PubMed   PubMed Central   Google Scholar  

Diamond, A. & Sekhon, J. S. Genetic matching for estimating causal effects: a general multivariate matching method for achieving balance in observational studies. Rev. Econ. Stat . 95 , 932–945 (2013).

Teixeira, V., Voci, S. M., Mendes-Netto, R. S. & da Silva, D. G. The relative validity of a food record using the smartphone application MyFitnessPal. Nutr. Dietetics 75 , 219–225 (2018).

Griffiths, C., Harnack, L. & Pereira, M. A. Assessment of the accuracy of nutrient calculations of five popular nutrition tracking applications. Public Health Nutr. 21 , 1495–1502 (2018).

Chen, J., Berkman, W., Bardouh, M., Ng, C. Y. K. & Allman-Farinelli, M. The use of a food logging app in the naturalistic setting fails to provide accurate measurements of nutrients and poses usability challenges. Nutrition 57 , 208–216 (2019).

Nguyen, Q. et al. Social media indicators of the food environment and state health outcomes. Public Health 148 , 120–128 (2017).

Gomez-Lopez, I. N. et al. Using social media to identify sources of healthy food in urban neighborhoods. J. Urban Health 94 , 429–436 (2017).

Einav, L., Leibtag, E. S. & Nevo, A. On the accuracy of Nielsen homescan data. Economic Research Report No. ERR-69. https://www.ers.usda.gov/publications/pub-details/?pubid=46114 (2008).

for Disease Control, C., Prevention et al. Behavioral risk factor surveillance system 2019 summary data quality report. july 16, 2020 (2020).

Mokdad, A. H. The behavioral risk factors surveillance system: past, present, and future. Annu. Rev. Public Health 30 , 43–54 (2009).

Pierannunzi, C., Hu, S. S. & Balluz, L. A systematic review of publications assessing reliability and validity of the behavioral risk factor surveillance system (brfss), 2004–2011. BMC Med. Res. Methodol. 13 , 1–14 (2013).

Austin, P. C. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivar. Behav. Res. 46 , 399–424 (2011).

Stuart, E. A. Matching methods for causal inference: a review and a look forward. Stat. Sci.: A Rev. J. Inst. Math. Stat. 25 , 1 (2010).

Article   MathSciNet   MATH   Google Scholar  

Glass, G. V. Primary, secondary, and meta-analysis of research. Educ. Researcher 5 , 3–8 (1976).

US Zip Codes History (2012). https://www.zip-codes.com/zip-codes-history.asp . Accessed 11 May 2021.

USPS Postal Facts (2021). https://facts.usps.com/42000-zip-codes/ . Accessed 11 May 2021.

US Census 2012 (2012). https://web.archive.org/web/20130707052113/http://www.census.gov/popest/data/counties/totals/2012/CO-EST2012-alldata.html/ . Accessed 11 May 2021.

Rundle, A. et al. Personal and neighborhood socioeconomic status and indices of neighborhood walk-ability predict body mass index in new york city. Soc. Sci. Med. 67 , 1951–1958 (2008).

Download references

Acknowledgements

T.A. and J.L. were supported by a National Institutes of Health (NIH) grant (U54 EB020405, Mobilize Center, NIH Big Data to Knowledge Center of Excellence). T.A. was supported by the SAP Stanford Graduate Fellowship, NSF grant IIS-1901386, NSF grant CNS-2025022, NIH grant R01MH125179, Bill & Melinda Gates Foundation (INV-004841), and the Office of Naval Research (#N00014-21-1-2154). H.N. was supported by NSF REU #1659585 at the Stanford Center for the Study of Language and Information (CSLI). J.H. is supported by Postdoctoral Fellowship in Cardiovascular Disease Prevention (T32) funded by the National Heart, Lung, and Blood Institute (NHLBI) at the National Institutes of Health (NIH). J.L. was supported by DARPA under Nos. FA865018C7880 (ASED), N660011924033 (MCS); ARO under Nos. W911NF-16-1-0342 (MURI), W911NF-16-1-0171 (DURIP); NSF under Nos. OAC-1835598 (CINES), OAC-1934578 (HDR), CCF-1918940 (Expeditions); Stanford Data Science Initiative, Wu Tsai Neurosciences Institute, Chan Zuckerberg Biohub, Amazon, Boeing, Chase, Docomo, Hitachi, Huawei, JD.com, NVIDIA, Dell. We thank MyFitnessPal for donating the data for independent research and Mitchell Gordon for helping with data preparation.

Author information

Authors and affiliations.

Allen School of Computer Science & Engineering, University of Washington, Seattle, WA, USA

Tim Althoff

Department of Computer Science, Stanford University, Stanford, CA, USA

Hamed Nilforoshan & Jure Leskovec

Stanford Prevention Research Center, Department of Medicine, Stanford University School of Medicine, Stanford, CA, USA

Million Marker Wellness Inc., San Francisco, CA, USA

Chan Zuckerberg Biohub, San Francisco, CA, USA

Jure Leskovec

You can also search for this author in PubMed   Google Scholar

Contributions

T.A. and J.L. conceived the study concept and design. T.A. and H.N. conducted the statistical analysis. T.A., H.N., J.H., and J.L. interpreted the data, drafted the manuscript, and critically revised the manuscript for important intellectual content.

Corresponding author

Correspondence to Tim Althoff .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Peer review information

  Nature Communications  thanks Bridget Kelly, Joreintje Mackenbach, Quynh Nguyen and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information, reporting summary, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Althoff, T., Nilforoshan, H., Hua, J. et al. Large-scale diet tracking data reveal disparate associations between food environment and diet. Nat Commun 13 , 267 (2022). https://doi.org/10.1038/s41467-021-27522-y

Download citation

Received : 20 January 2021

Accepted : 18 November 2021

Published : 18 January 2022

DOI : https://doi.org/10.1038/s41467-021-27522-y

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Multimodal digital phenotyping of diet, physical activity, and glycemia in hispanic/latino adults with or at risk of type 2 diabetes.

• Rony Santiago

npj Digital Medicine (2024)

Diabetic Cardiomyopathy and Cell Death: Focus on Metal-Mediated Cell Death

• Kupper Wintergerst

Cardiovascular Toxicology (2024)

Designing and Evaluating a Hierarchical Framework for Matching Food Outlets across Multi-sourced Geospatial Datasets: a Case Study of San Diego County

• Jiue-An Yang
• Marta M. Jankowska

Journal of Urban Health (2024)

Wearable and digital devices to monitor and treat metabolic diseases

• Ayya Keshet
• Lee Reicher

Nature Metabolism (2023)

Nature Communications (2023)

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

The best of EcoWatch right in your inbox. Sign up for our email newsletter!

• About EcoWatch
• Contact EcoWatch
• Terms of Use
• Privacy Policy

Learn About Solar Energy

Flooding and landslides kill at least 85 people in brazil as climate change forms ‘disastrous cocktail’ with severe storms, acoustic recordings reveal possible comeback for critically endangered antarctic blue whales, peaceful bonobos and aggressive chimps new research says it’s more complicated than that, uk installs nearly 6,000 public ev chargers in first 3 months of 2024, florida’s desantis bans lab-grown meat, climate change could lead to major venomous snake migrations, study says, 'it's got to be a fight': author adam welz on surviving climate breakdown and saving species of a 'tarnished eden', sounds right recognizes nature as musician, with royalties going to environmental causes, new pbs documentary focuses on a more hopeful future, 'we have just arrived': author rick bass on writing and activism in montana, new documentary explores climate breakdown and 'protectors' fighting to adapt, 4 must-read graphic nonfiction books about the environment, birding 101: everything you need to know, how to get your garden winter-ready, learn about deregulated energy, learn about windows, 3 research questions could hold the key to sustainable eating.

A woman harvests home-grown lettuce. sanjeri / Getty Images

By Sarah Reinhardt

When it comes to healthy eating, there’s a lot we already know.

Just take a look at the scientific report issued by the 2020 Dietary Guidelines Advisory Committee , the group of scientific experts behind the newly released 2020-2025 Dietary Guidelines for Americans . At 835 pages, the report spans a rigorous review of current research on dozens of topics, from whether eating peanuts early in life reduces the likelihood of peanut allergies (it probably does), to how much added sugar we can eat and still maintain a healthy diet ( way less than what we’re eating now ). It also outlines the broad contours of a healthy diet, which has changed little from past editions of the Dietary Guidelines : it’s typically higher in vegetables, fruits, legumes, whole grains, low- or non-fat dairy, lean meat and poultry, seafood, nuts, and unsalted vegetable oils, and lower in red and processed meats, sugar-sweetened foods and drinks, and refined grains.

But food is more than just nutrition. We also know that food can affect the health of many people before it even reaches our plates. For example, pollution caused by fertilizer runoff from farms have left communities without access to safe drinking water, while many of the people who produce our food don’t earn enough to afford healthy diets themselves. So, is it possible to eat in ways that promote health and produce better outcomes for our collective wellbeing, livelihoods, and natural resources? And what would it take to get us there?

These are complex questions, but critical ones—and if the federal government made it a national priority to answer them, we might have a chance at avoiding otherwise devastating consequences.

That’s why leading experts and scientific bodies, including the 2020 Dietary Guidelines Advisory Committee , are increasingly calling attention to the importance of research on sustainable diets . Though there is no official record of how much federal funding is currently supporting this field, funding for nutrition research across the board has remained stagnant for decades . Research funding at the intersection of food and agriculture issues, in particular, is decidedly inadequate to address the magnitude of the public health challenges before us, including climate change, threats to food security, and persistent poverty and health disparities.

Many, including UCS, are working to change that by advocating for greater government investment in sustainable diets research. Here are three of the most pressing questions that research can answer to enable more sustainable eating for all.

1. When it comes to healthy and environmentally sustainable diets, what exactly is on the menu?

Research on healthy and environmentally sustainable diets has seen extraordinary growth in recent years. According to UCS research , nearly 100 new scholarly articles were published on this topic between 2015 and 2019, including 22 articles focused specifically on US diets. For reference, that’s more than four times the number of articles published on the same topic between 2000 and 2015—in about a quarter of the time.

Much of this research agrees that, in general, healthy diets that are higher in plant-based foods and lower in animal-based foods are more environmentally sustainable—meaning they are often associated with reduced energy and land use, as well as less air and water pollution.

This is an important finding that enables us to take the first steps toward more sustainable diets—and each step is critical, given the urgency and magnitude of dual public health and climate crises. Indeed, a number of other countries already have acted on existing knowledge to incorporate sustainability into national dietary guidance .

But the more we know, the faster (and further) we can move forward.

For example, we need to better understand the impacts of all different kinds of dietary shifts that could improve health and sustainability, and the most effective ways to encourage such shifts at the individual, institutional, and societal levels . We also need to better understand how a wide variety of sustainable diets align with diverse cultures and culinary traditions . Both of these areas of research will be crucial to informing practical public health interventions and policy recommendations that can help all people make meaningful changes.

There are also challenges with available data and models. Many studies on sustainable diets rely on models called Life Cycle Assessments , which help estimate the cumulative environmental impacts of the foods we eat. For example, how much energy is required to process, package, and transport a pound of beef from the farm to the grocery store? How much land, water, and energy was required to raise the cattle? While a Life Cycle Assessment is useful in providing general estimates, it’s often based on averages that don’t account for key differences in production methods, for example, or regions. More data—and more diverse data sources—could make this an even more powerful and practical tool.

These and other outstanding research gaps prompted the Interagency Committee on Human Nutrition Research (ICHNR), an interdisciplinary group of leaders across key federal agencies, to identify sustainable diets as a research priority in the creation of its 2016-2021 National Nutrition Research Roadmap. And though interest in this topic continues to grow, there is evidence that federal funding remains woefully inadequate to support independent research on sustainable diets and food systems.

2. How could more socially and economically sustainable food systems support healthy eating?

While environmental sustainability receives a lot of attention, a truly sustainable diet also takes into account the social and economic conditions that can threaten our health, wellbeing, and the future of our food supply.

Among these considerations are the persistent health and economic disparities embedded in the current US food system, which disproportionately affect Black, Indigenous, and other resilient communities of color who have faced decades of racist and discriminatory practices and policies. Even before the pandemic, the people working throughout the US food system —nearly 40 percent of whom are people of color—experienced greater poverty, poorer healthcare access, and higher occupational health hazards relative to the general population. Many of these same populations are also most vulnerable to the environmental consequences of unsustainable food systems, such as climate change, water pollution, and other environmental impacts. The irony should not be lost on anyone that many Indigenous populations have for generations produced food in harmony with ecological systems—a way of living that, in the US and elsewhere, was deeply and violently disrupted by colonization and genocide.

Pervasive exploitation in the food chain can function to make certain foods cheap, while also preventing workers from being able to afford healthy diets themselves. And unfortunately, this isn’t a challenge unique to workers in the food chain. Many US households are already struggling to afford a healthy —let alone sustainable—diet.

So how can research help?

For one, there are an increasing number of studies that are putting dollar signs on healthier and more sustainable diets. Understanding what these diets might cost consumers, and what segments of the population might be unable to afford them, is an important first step in creating policy and programs that can bring healthier and more sustainable eating into reach for more people. Existing research on programs and initiatives such as the Supplemental Nutrition Assistance Program (also known as SNAP), cash assistance programs, or minimum wage increases will also be useful in identifying the best ways to increase consumer purchasing power, particularly among low-income households. Perhaps most importantly, effective community-based participatory research can support community efforts to resist and address the root causes of social and economic disparities, including racism and exploitative economic systems. Organizations like the HEAL Food Alliance , a multi-racial coalition building a national movement to transform food and farming systems, should be looked to as experts and leaders in shaping the goals and objectives of research that will impact the communities they represent.

3. What are the synergies and tradeoffs when you value both public health and sustainability?

Not every food choice that supports health will support sustainability, and vice versa. In fact, as our research shows , if the US population shifted from our current diet to a healthier diet, greenhouse gas emissions, energy use, and water use could actually increase. This is due in part to the fact that many fruits and vegetables—which most of us don’t eat nearly enough of—are produced in ways that require a lot of water and energy.

Another commonly cited example is the conflict between US dietary recommendations for fish and shellfish and the ability of fisheries to meet this demand in a sustainable way—an issue highlighted by experts in a National Academies of Science workshop and the 2015 Dietary Guidelines Advisory Committee .

None of this is to say that we should abandon our efforts to eat healthfully. Rather, it’s to say that we need to do some problem-solving if we want to eat in ways that are healthy and sustainable for people and the planet, and research can play a key role in helping us get there.

Solving the sustainable food Rubik’s cube

It was more than thirty years ago that federal policy established the first specific directives for prioritizing nutrition research, yet its funding has remained stagnant for decades and coordination across federal agencies remains inconsistent. Meanwhile, many complex factors are fueling the public health challenges embedded in the US food system: rates of diet-related disease and health disparities continue to climb, the degradation of soil and other natural resources has grown more dire, many workers are still underpaid and unsafe, and climate change is now a reality that looms large in our daily lives.

It’s time for the next generation of nutrition research. To fill these gaps in research and provide information we need to make the best possible policy decisions, we must invest in research that is systems-oriented and can help identify healthy diets that deliver the best possible social, economic, and environmental benefits for all populations. Such research must span disciplines, take leadership from communities most affected by health disparities and environmental injustices, and at its core focus on tackling root causes of the greatest threats facing the food system, including the systemic exploitation of people and finite resources.

President Biden’s administration has no shortage of opportunities to prioritize this area of research, which sits at the nexus of some of its foremost policy priorities . US Department of Agriculture Secretary Vilsack is well positioned to play a powerful role in lifting up and improving coordination of nutrition research needs, particularly as he works to rebuild capacity in the department’s research arms. And Congress, for its part, can leverage legislative opportunities such as appropriations and the potential reauthorization of child nutrition programs to ensure that funding is finally available to answer these and other critical nutrition research questions.

Reposted with permission from Union of Concerned Scientists.

• Chipotle Proves Sustainable Food Sourcing Is Profitable - EcoWatch
• How Organic Farming Contributes to a Sustainable Food System ...
• 20 Most Sustainable Food and Health Solutions on the Planet ...
• Beyond Organic: Just What Is Sustainable Food? | Martha Stewart

Common Questions About Fruits and Vegetables

You’ve heard the nutrition prescriptions: “Aim for 5-A-Day” and “Eat a Rainbow.” Few would argue that fruits and vegetables offer health benefits , including reducing the risk of some chronic diseases. But maybe you’ve hesitated to fill your grocery cart with colorful produce because of concerns that pop up in online articles and discussions. Here’s a look at some commonly asked questions about fruits and vegetables, and what the current research says:

“Is fruit bad for me because it contains sugar?”

Sugar gets such a bad rap in the popular press that some people hesitate to eat foods that contain natural sugars, especially fruit. There’s plenty of research exposing the detrimental health effects of eating too much added sugar , but less so with natural sugars. Overall, fruit intake is associated with a lower risk of cardiovascular disease, high blood pressure, and type 2 diabetes.

Another confusing point is that some fruits like watermelon and grapes have a high glycemic index, a tool that measures how quickly a food increases blood sugar. But more accurate than the glycemic index is the glycemic load , which factors the amount of carbohydrate per serving of a food and to what degree it will raise a person’s blood glucose level after eating it. Because watermelon and grapes are fairly low in carbohydrate, their glycemic load is low. In general, most whole fresh or frozen fruits have a relatively low glycemic load.

Furthermore, whole fruits have a built-in stoplight. They are more naturally satisfying, whereas highly processed foods with added sweeteners tend to be easy-to-digest and are therefore easy to overeat! Imagine how quickly one can gulp down a standard 12-ounce can of cola that contains 40 grams of processed sugar. Now think about how long it takes to chew through an apple that contains about 13 grams of natural sugar in addition to various fibers, vitamins, antioxidants, and water. The apple takes longer to eat and is more satiating, so you are less likely to reach for another.

The takeaway: Fruits contain natural sugar and many other beneficial nutrients, so they are a healthful addition to a daily diet.

“Are canned and frozen fruits and vegetables lower in nutrients?”

Refrigeration can slow degradation, but even so, certain highly perishable fruits like berries only last about a week before visible changes in color, texture, and flavor set in. Apples, pears, and squashes are less sensitive and can last for 1-2 months with refrigeration.

For longer storage, canning and freezing can preserve nutrients more effectively than refrigeration. Initially some nutrients are lost during processing. For example, canning uses heat treatment or other methods to destroy bacteria, followed by storage in a liquid medium in an airtight container; freezing entails a prior step of blanching the produce quickly to deactivate enzymes that speed ripening. In both cases, some water-soluble vitamins like C and B may be destroyed or leached into the cooking or storage liquid.

Asparagus loses about 30% of vitamin C during canning but only 10% after blanching and freezing. Generally, the freezing process averages about 50% loss of vitamin C (range 10-90%), and canning treatment causes average losses greater than 60% (range 8-90%). (1) The ranges are wide depending on the exact processing treatment, storage conditions, and type of produce. But regardless of losses during processing, some or many of the nutrients are retained, which upon storage are further preserved through protection from oxygen, heat, and light. Broccoli retains almost 100% of vitamin C after freezing.

Home cooking of fresh or processed produce will cause additional losses due to nutrients destroyed by heat or seeping into the cooking water. Microwave cooking and steaming as opposed to boiling and deep-frying preserves the most nutrients due to quicker cooking times and reduced contact with water.

The takeaway: Unless fruits and vegetables are eaten immediately after harvesting, there is going to be nutrient loss to some degree. Regardless of losses, fruits and vegetables remain a valuable source of a variety of nutrients. In certain cases, frozen or canned produce may contain higher levels of nutrients than fresh produce that has been stored for too long. To obtain the most nutrients, consume fresh produce within a few days, or frozen varieties within a few weeks. When choosing canned or frozen produce, select options that do not include extra sodium, sugar, or other additives.

“Are smoothies and juices just as good as eating whole fruits and vegetables?”

If most or all of the whole fruit and/or vegetable is blended into the beverage (skin, pulp, and flesh), then the nutrients and fiber are preserved, making it nutritionally comparable to eating the ingredients in whole form. In some cases, these “smoothies” may offer high quality nutrition if very fresh produce is used. They may also be easier to digest in a blended texture. However, consuming these same foods in whole form will provide longer term satiety and make it easier to avoid excessive caloric intake. Also, don’t forget the other ingredients that are often added: milk, processed fruit juice, sweeteners, seeds, and protein powders can quickly drive caloric intake upwards of 700-800 calories per serving! Even if only whole produce is used, calories can still be high depending on how much is added (e.g., including several cups of fruits, or using bananas and avocados that are higher in calories).

Juice from extractors and bottled juices (even when labeled 100% fruit juice) lack fiber and some nutrients that are destroyed or removed during processing. Without fiber, they are quickly digested and not as satisfying, and may cause spikes in blood sugar, leaving one feeling hungry soon after drinking them. On a hot day, it is easy to drink several hundred calories in a large glass.

Not surprisingly, higher consumption of fruit juice has been associated with greater weight gain. (2)  Also, a large prospective study found likely adverse effects from juice. Men and women who drank one or more servings of fruit juice each day experienced a higher risk of developing type 2 diabetes by as much as 21%. (3) Whereas eating at least two servings a week of certain whole fruits such as blueberries, grapes, and apples was associated with a lower risk for type 2 diabetes by as much as 23% when compared with those who ate less than one serving per month.

The takeaway: Whole fruits and vegetables are the best choice. Smoothies that blend whole fruits and vegetables without additional sweeteners and are served in appropriate portions may be helpful for some people to consume more of these foods, but should not replace eating them in their whole form. It is best to prepare smoothies at home so that you can control the type and amount of ingredients added to ensure calorie control and optimal nutrients.  Juices are easy to over-consume and, especially if weight control is a concern, should be limited to one small glass per day. 

• Rickman, J.C., Barrett, D.M., Bruhn, C.M. Nutritional comparison of fresh, frozen and canned fruits and vegetables. Part 1. Vitamins C and B and phenolic compounds. J Sci Food Agric . 2007; 87:930–944.
• Mozaffarian, D., et al. Changes in diet and lifestyle and long-term weight gain in women and men.  N Engl J Med 364 (2011): 2392-2404.
• Muraki, I., et al. Fruit consumption and risk of type 2 diabetes: results from three prospective longitudinal cohort studies. BMJ . 2013 Aug 28;347:f5001.

Terms of Use

The contents of this website are for educational purposes and are not intended to offer personal medical advice. You should seek the advice of your physician or other qualified health provider with any questions you may have regarding a medical condition. Never disregard professional medical advice or delay in seeking it because of something you have read on this website. The Nutrition Source does not recommend or endorse any products.

• Open access
• Published: 24 February 2022

Stress eating: an online survey of eating behaviours, comfort foods, and healthy food substitutes in German adults

• Kathrin Gemesi 1 ,
• Sophie Laura Holzmann 1 ,
• Birgit Kaiser 2 ,
• Monika Wintergerst 3 ,
• Martin Lurz 4 ,
• Georg Groh 3 ,
• Markus Böhm 4 , 5 ,
• Helmut Krcmar 4 ,
• Kurt Gedrich 2 ,
• Hans Hauner 1 , 6 &
• Christina Holzapfel 1  

BMC Public Health volume  22 , Article number:  391 ( 2022 ) Cite this article

13k Accesses

7 Citations

12 Altmetric

Metrics details

In many people, stress is associated with changes in eating behaviour. Food products consumed during stress (comfort foods) are often unhealthy. It is rather unknown what comfort foods are consumed in Germany and what healthier food products are considered as alternatives to support stress-eaters in making healthier food choices.

This online survey was conducted in spring 2021 throughout Germany. Participants were digitally recruited by newsletters, homepages, social media, and mailing lists. The survey included a standardized questionnaire with items concerning e.g. sociodemography, stress, and nutrition. Comfort foods were pre-selected through literature search and food substitutes were defined and discussed by experts. Analyses examined comfort food consumption and substitute preferences dependent on sex, age, body mass index (BMI), and being a self-identified stress-eater. The statistical analysis was performed using R.

Survey participants were mostly female (80.6%, 994/1234), had a mean age of 31.4 ± 12.8 years and a mean BMI of 23.4 ± 4.3 kg/m 2 . Participants stated, that the two favourite comfort foods were chocolate (consumed often/very often by 48.3%, 596/1234) and coffee (consumed often/very often by 45.9%, 566/1234). Regarding food substitutes, the most frequently named alternative food for chocolate and cookies was fresh fruits (for chocolate: 74.4%, 815/1096, for cookies: 62.6%, 565/902). Tea without added sugar (64.4%, 541/840) was the preferred substitute for coffee. Almost 50% of participants (48.1%, 594/1234) identified themselves as stress-eaters, of which 68.9% (408/592) stated to eat (very) often more than usual in subjective stress situations.

Conclusions

The results from this work suggest that specific comfort foods and substitutes are preferred by the participants in stressful situations. This knowledge about food choices and substitutes should be investigated in further studies to improve eating behaviour in stressful situations.

Trial registration

The survey was registered in the German Register of Clinical Studies (Registration number: DRKS00023984 ).

Peer Review reports

Introduction

Stress is known to be associated with disturbed sleeping, memory, learning, and attention and can have a negative impact on the immune and the cardiovascular system [ 1 ]. It is also known that stress can have an indirect effect on health, e.g. by changing eating behaviour [ 2 ]. For stress recognition, different stress indicators and assessment methods can be used [ 3 ].

Previous studies have shown that nearly equal numbers of participants (i.e. 50%) responded to stress by eating either more or less [ 4 , 5 , 6 ]. Moreover, stress seems to change food preferences towards unhealthy food products. A systematic review of 16 studies examining food intake and food frequency of women under psychological stress has shown a significant association between stress and an unhealthy diet (e.g. high in fat, sweets, and salt but low in fruits and vegetables) [ 7 ]. In addition, a meta-analysis of five studies with 3471 participants revealed a negative association between stress and diet quality [ 7 ].

Certain mediators seem to have an impact on the relationship between stress and eating. Eating self-regulation was shown to partially influence the relationship of stress and emotional eating [ 8 ]. Van Blyderveen et al. could show in female undergraduated students that impulsive women had a higher susceptibility to stress-induced eating, and that impulsivity and emotional suppression had an influence on the relationship between negative affect and food consumption during stress [ 9 ]. A review by Adam et al. describes the mediator role of the reward system in the relationship between stress and stress-induced eating [ 10 ]. Severe stress can result in higher cortisol levels, leading to activation of the hypothalamus pituitary adrenal axis, which interacts with various hormones having an influence on food intake [ 10 ].

Experiencing stress accompanied by negative emotions is one reason why so called “comfort foods” are consumed [ 11 ]. Comfort foods are suspected to be comforting by having positive emotional effects. An experiment by Wagner et al. showed that consuming your personal comfort food has a positive emotional effect, but not more than other food or no food [ 12 ]. A review summarised that there is not enough evidence for the reasons of consuming comfort foods and their emotional benefits [ 13 ].

The selection of comfort foods is dependent on country-specific popular foods, why comparisons across countries are limited. In studies from Korea [ 14 ], Saudi Arabia [ 15 ], Great Britain [ 5 ], the USA [ 16 , 17 ], and in a study comparing three different European countries including Germany ( N  = 696) [ 18 ] a variation of comfort foods has been described. In addition an association between perceived stress and frequent consumption of unhealthy food products like sweets and fast food could be shown in these studies. Moreover, it is suggested that comfort food preferences are associated with gender [ 5 , 15 , 18 ].

The study examining comfort foods in Germany [ 18 ] was focused on students and used a non-validated food frequency questionnaire to measure food consumption.

A frequent consumption of energy-dense comfort foods during stress could lead to an increase of caloric intake and, as a consequence of a chronically positive energy balance, to increased body weight. The MIDUS study in US adults observed an association between stress-eating and metabolic parameters, which could be attributed to the presence of abdominal obesity [ 19 ]. Increased body weight in turn could lead to weight stigmatization, which could result in increased food intake forming a vicious circle [ 20 ]. It should be investigated whether a positive association between BMI and comfort food consumption frequency exists.

In conclusion, the recommendation of healthier alternatives for comfort foods (substitutes) could be a prerequisite to prevent from unhealthy energy-dense food intake in stressful situations. To put this into practice, concepts of nudging are able to positively influence peoples’ behaviour [ 21 ]. To the best of our knowledge there are no studies that investigated whether stress-eaters would consume substitutes instead of comfort foods if offered to them under stress.

Aim of the present online survey was to collect sex-, age-, BMI-, and stress-eater-specific data on food products, which are chosen in stressful situations, among adults in Germany together with information on food products which are considered as healthier alternatives.

Data were collected by an open online survey performed during the Covid-19 pandemic between January and April 2021 throughout Germany. Participants were digitally recruited by university internal and external channels such as newsletters, homepages, social media accounts (e.g. facebook), and mailing lists using snowball sampling. Interested persons were guided to the online survey by an invitation including a link to the survey platform SoSci Survey (V3.1.06).

A calculation of the response rate was not possible since the survey invitation was delivered electronically and the number of invitations was unknown.

The survey started with an introduction presenting information about the research team, aim of the survey, guidance on answering the questions, and information about data privacy and protection. Before answering the survey questions, participants had to confirm the data privacy statement and to give informed consent prior to participation. Additional inclusion criteria were being 18+ years old and having good German language skills. No incentives were offered to the participants.

Questionnaire

The 38-item questionnaire was developed by an interdisciplinary team of nutritionists, public health experts, and computer scientists. The questionnaire was implemented in SoSci Survey and pretested by the target group with regard to understanding, difficulty, and structure of questions and answers. According to the results from the pretest the questionnaire was shortend and single questions were changed. The final questionnaire comprised questions (closed, open, single or multiple choice) referring to nutrition (one question), stress perception and coping (four questions), stress-eating (17 questions), technical behaviour (four questions), digital applications (apps) detecting stress (three questions), and personality (one question) from which a selection of questions focused on stress-eating behaviour is presented in this work. Socio-demographic and anthropometric data (eight questions) were collected at the end of the survey. Neutral answer options like “sometimes” and “other” were provided if indicated. Each question had to be answered to continue, whereas the survey could be stopped after each question.

Comfort foods

Literature search was performed to pre-select comfort foods and to divide them in four food product categories. Studies about stress-eating and comfort foods were selected [ 5 , 14 , 15 , 16 , 17 , 18 , 22 , 23 , 24 , 25 ]. Most of the studies were conducted in the USA and most of the participants were students. The selected studies had no uniform definition of a comfort food. Therefore, the final set of 13 comfort foods was adapted to German food culture according to experts’ opinion. Figure  1 shows the pre-selected comfort foods, which have been integrated into the questionnaire. In the survey, participants were asked how frequently (5-Likert scale: “never” to “very often”) they consume these comfort foods in stressful situations. Comfort foods other than pre-selected could not be specified by survey participants. This question had been answered by all participants, not only by those who identified themselves as stress-eaters.

Pre-selected comfort foods identified by literature search and adapted to German food culture

Substitutes

Suitable substitutes for each comfort food were discussed and pre-selected by the same experts who pre-selected the comfort foods. The following criteria were considered: substitutes should 1) not be comfort foods, 2) have a similar flavour like the comfort foods, and 3) be “healthier” (e.g. less energy, salt, or sugar) than the comfort foods. Quantity, volume, brand, and packaging of substitutes were not considered. In addition to the pre-selected substitutes, survey participants could name other substitutes. Participants were asked whether they could imagine to consume the offered substitutes for the prevailing comfort foods they stated to consume in stressful situations.

Stress-eating

Participants were asked if they would identify themselves as stress-eaters ( “Do you think you are a stress-eater?” ). Being a stress-eater was defined by eating differently when experiencing stress. Survey participants who stated to be a stress-eater where asked to characterise their stress-eating behaviour (e.g. eating more, less, more often, etc.). Another analysis of this survey by Kaiser et al. used the validated Salzburg Stress Eating Scale to characterize stress-overeaters [ 26 ].

Statistical analysis

As a first step, integrity and plausibility checks were performed. Descriptive data analyses (frequencies, percentages, standard deviation, and mean) were performed using Excel 2016 (Microsoft Corp). Only data from participants who provided sociodemographic data were analysed. Since the dataset contains missing answers and some single answers had to be excluded because of inconsistencies in the answers, sample size differs between questions and answer options. Normality was tested using the Shapiro-Wilk test. Variance homogeneity was checked by using F-test. Differences in mean age and BMI were estimated by performing Welch’s t-test or a two-sample t-test. Pearson’s Chi-squared test or Fisher’s exact test were used to examine sex differences in stress-eating behaviour and differences between self-identified stress-eaters and non-stress-eaters. Associations between sex, age, BMI, and being a self-identified stress-eater and the consumption of comfort foods and the consideration of named substitutes were assessed for total of participants using multinomial logistic regression analysis or binary logistic regression analysis. In addition, a sub-analysis in self-identified stress-eaters was performed. After adjusting for multiple testing by Bonferroni correction, p -values < 0.004 for multinomial and <  0.002 for binary logistic regression models were considered as statistically significant. All analyses were performed using R (V4.1.0).

Characteristics

Characteristics of the survey population are presented in Table  1 . Participants were on average 31.4 ± 12.8 years old and had a mean BMI of 23.4 ± 4.3 kg/m 2 . Women showed a significantly lower BMI compared to men (23.10 vs. 24.60 kg/m 2 , p  = 1.07e − 6 ). Most participants were female (80.6%), single (66.0%), students (53.3%), and had a higher education (82.0%). Regarding stress frequency, 22.9% and 41.8% of participants reported being stressed everyday or more than once a week, respectively. The two most frequently indicated stressors were work (65.2%) and partner, family, and friends (36.1%). Half of the survey participants (48.1%) identified themselves as stress-eaters. Self-identified stress-eaters were predominantly women (88.7%, 527/594) (Table 1 ).

The stress-eating behaviour of self-identified stress-eaters is shown in Table  2 . More than 50% of the participants stated to eat more ( often/very often : 68.9%, 408/592), more often ( often/very often : 67.2%, 399/594), faster ( often/very often : 55.6%, 330/594), or other foods ( often/very often : 54.4%, 323/594) in stressful situations. There were no statistically significant sex differences (Table 2 , Additional File  1 : Supplementary Table 1).

The consumption frequency of 13 pre-selected comfort foods in stressful situations is shown in Table 3 . Regarding all participants, the two most often consumed comfort foods were chocolate ( often/very often : 48.3%) and coffee ( often/very often : 45.9%). The comfort foods, which were stated to be never consumed in stressful situations by at least 50% of the survey participants were energy drinks (88.9%), fried food (62.1%), ice cream (59.3%), sugar-sweetened beverages (58.1%), hamburgers (52.8%), and candies (50.8%). Supplementary Table 2 (Additional File 1 ) shows the sex-specific consumption frequency of comfort foods.

In a sub-analysis of self-identified stress-eaters, chocolate ( often/very often : 72.4%, 430/594) and coffee ( often/very often : 53.3%, 318/594) were again found as the two most often consumed comfort foods (Supplementary Table  2 ). Regarding the consumption frequency of these comfort foods, a statistically significant difference was found between self-identified stress-eaters and non-stress-eaters ( p  <  0.001).

The odds of stating to eat chocolate was three times higher for women ( very often : OR = 3.05, p  <  0.001) than for men. Regarding age and BMI statistically significant differences have been observed (Table  4 ). For each comfort food except from energy drinks and sugar-sweetend beverages, it could be shown that stress-eaters stated up to 25 times more often to consume the comfort food (very) frequently during stress compared to non-stress-eaters ( p  <  0.001) (Table 4 ).

Table  5 shows which substitutes were considered as alternatives for the comfort foods “chocolate”, “cookies”, and “coffee”, that were stated to be consumed often and very often by at least 20% of the total survey population. The most frequently considered alternative food for chocolate was fresh fruits (74.4%) followed by dark chocolate (69.8%). For the comfort food “cookies”, the participants’ most favourite substitutes were fresh fruits (62.6%) and nuts-fruits-mixtures (55.2%). Tea without added sugar (64.4%) followed by water (47.0%) were the two most often stated substitutes for coffee.

The odds of considering chocolate fruits as substitute for chocolate (OR = 2.07, p  <  0.001) was two times higher for women than for men. Age was associated with four different substitutes for chocolate (Table  6 ). No association was found between BMI and considering substitutes for chocolate, cookies, and coffee ( p  > 0.002). Self-identified stress-eaters stated more often to consider light/zero drinks instead of coffee (OR = 2.11, p  <  0.001) compared to non-stress-eaters (Table 6 ).

This survey shows that more than half of the survey participants change often or very often their eating behaviour in response to stress. The two most favourite comfort foods were chocolate and coffee. The consideration of substitutes for chocolate, cookies, and coffee was heterogeneous with different substitutes per comfort food.

According to the literature, about 50% of stress-eaters eat more and about 50% of stress-eaters eat less than normal (e.g. [ 4 ]). In contrast, 68.9% of the self-reported stress-eaters in this survey population stated to be stress-overeaters.

Chocolate and coffee were by far the two most favoured comfort foods, followed by cookies. This finding confirms the result of a study conducted in the USA that identified chocolate as the most frequently mentioned sweet comfort food [ 22 ]. A previous survey completed by students in the USA has reported a significantly positive association of perceived stress with coffee consumption [ 16 ]. The fact that only two out of 13 comfort foods were stated to be consumed often or very often by our survey participants could indicate that stress-eating behaviour does not vary much from person to person. Furthermore, experiencing stress as well as eating behaviour per se is very subjective and it might be difficult for participants to figure out that dietary patterns and food consumption change in response to stress.

The offered substitutes per comfort food were diverse. Some substitutes were very similar to the respective comfort food (e.g. dark chocolate instead of chocolate). Other substitutes were very different from the comfort food (e.g. fresh fruits instead of cookies). The fact that all offered substitutes for the three most favourite comfort foods could be imagined more or less by the survey participants as healthier alternative might indicate that these substitutes may be suitable for practical terms in real life. However, this assumption needs to be examined in an intervention study. We further found that rather healthy substitutes were stated more often compared to others (e.g. fresh fruits instead of chocolate and cookies, tea without added sugar instead of coffee). This fact could be used to address the issue of “stress-eating” comprehensively with nudging, an approach to change food choices. A systematic review showed its ability to promote healthy food choices by changing the order of food products or their proximity [ 27 ]. However, it cannot be ruled out that our survey participants made these healthy substitute choices due to social desirability, as participants were only asked if they could imagine to eat the offered substitutes.

Regarding the role of sex for the reported consumption frequency of comfort foods, we found that chocolate was more frequently consumed by women in stressful situations than by men, which is in line with previous findings [ 5 , 15 , 18 ]. Looking at substitutes, women stated more often to consider chocolate fruits instead of chocolate than men.

Age was partially associated with comfort food consumption. For example older people stated more often to consume frequently coffee during stress. Furthermore, older people stated less often to accept selected substitutes for chocolate, but more often to consider self-named substitutes than younger people. This could be a hint that older people have different preferences regarding substitutes.

A higher BMI was associated with a frequent consumption of fried food, hamburgers, and sugar-sweetend beverages. This result is plausible since individuals with overweight or obesity often experience emotional stress like social exclusion and shame [ 28 ], which could promote emotional eating. Previous studies examing the relationship between perceived stress, eating behaviour, and obesity differ in their study design (e.g. cross-sectional, longitudinal), methods, and research question. In a cross-sectional study by Richardson et al. including 101 American women with children, perceived stress was positively associated with uncontrolled and emotional eating, and stress with severe obesity, but independently from eating behaviour and quality [ 29 ]. A longitudinal, population-based study with almost 6000 participants in Finland could show for women at the age of 31 years that stress-eaters had the highest BMI and that stress-eating was associated with obesity [ 30 ]. In general, more longitudinal studies are needed to be able to assess the causal relationship between stress-eating and obesity.

Because of these findings the issue of “stress-eating” should be tackled to protect people at risk. It has to be mentioned that the present survey has not collected data about the quantity of comfort food consumption and the causality for obesity is, therefore, speculative. The focus of this survey was on collecting data about comfort foods per se and on examing whether and what healthier substitutes the survey participants could imagine to choose.

The findings of this survey on stress-eating behaviour provide new insights into potential new strategies to address this frequent cause of high-caloric and unhealthy food intake. However, there is need to perform prospective and intervention studies to explore how stress-related consumption of popular comfort foods can be replaced by healthier food alternatives, and how approaches like nudging and personalised dietary recommendations can be best employed to achieve the consumption of healthy food during stress.

This online survey is the first examining the topic of “comfort foods” and “suitable healthy substitutes” among adults in Germany. The data has been collected with a standardised questionnaire developed by an interdisciplinary team. Although the sample size is rather large, the validity is limited by the non-representative study population consisting of a majority of females and students, and by the fact that all data are self-reported. In addition, a set of pre-selected comfort foods was presented to the participants and they could not specify any others. In addition, the consideration of specific foods as healthier substitutes was theoretical, which allows no conclusion whether these substitutes would work in real life. Therefore, the knowledge about comfort foods and substitutes should be explored in intervention studies to examine whether stress-eaters consume substitutes and whether eating behaviour in stressful situations can be improved. Lastly, it is noteworthy that the data collection was performed during the Covid-19 pandemic, which is associated with changed food consumption [ 31 ]. This survey did not focus on stress-eating during the pandemic period or collected detailed data on the reason of perceived stress but aimed to collect data about stress and stress-induced eating in general.

According to this survey performed during the Covid-19 pandemic, specific comfort foods and substitutes are preferred in stressful situations. The consumption frequency of comfort foods and the selection of substitutes seem to be associated with sex, age, BMI, and being a self-identified stress-eater. The findings should be confirmed in further studies, especially in intervention studies.

Availability of data and materials

The datasets used and/or analyzed during this survey are available from the corresponding author upon reasonable request.

Abbreviations

Body mass index

Confidence interval

Hapke U. Chronic stress among adults in Germany: results of the German Health Interview and Examination Survey for Adults (DEGS1). Bundesgesundheitsbl Gesundheitsforsch Gesundheitsschutz. 2013;56:749–54. https://doi.org/10.1007/s00103-013-1690-9 .

Article   CAS   Google Scholar  

Rutters F, Nieuwenhuizen AG, Lemmens SGT, Born JM, Westerterp-Plantenga MS. Acute Stress-related Changes in Eating in the Absence of Hunger. Obesity. 2008;17:72–7. https://doi.org/10.1038/oby.2008.493 .

Article   PubMed   Google Scholar  

Kaiser B, Holzmann SL, Hauner H, Holzapfel C, Kurt G. Nutrition and stress: Overview of selected stress indicators and smart measurment techniques. Ernahrungs-Umschau. 2020;67:98–107.

Google Scholar  

Weinstein SE, Shide DJ, Rolls BJ. Changes in food intake in response to stress in men and women: psychological factors. Appetite. 1997;28:7–18. https://doi.org/10.1006/appe.1996.0056 .

Article   CAS   PubMed   Google Scholar  

Oliver G, Wardle J. Perceived effects of stress on food choice. Physiol Behav. 1999;66:511–5. https://doi.org/10.1016/s0031-9384(98)00322-9 .

American Psychological Association. Stress a major health problem in the U.S., warns APA; 2007. Available from: URL: http://www.apa.org/news/press/releases/2007/10/stress.aspx . cited 2021 July 9. 

Khaled K, Tsofliou F, Hundley V, Helmreich R, Almilaji O. Perceived stress and diet quality in women of reproductive age: a systematic review and meta-analysis. Nutr J. 2020;19:92. https://doi.org/10.1186/s12937-020-00609-w .

Article   PubMed   PubMed Central   Google Scholar  

Ling J, Zahry NR. Relationships among perceived stress, emotional eating, and dietary intake in college students: Eating self-regulation as a mediator. Appetite. 2021;163:105215. https://doi.org/10.1016/j.appet.2021.105215 .

van Blyderveen S, Lafrance A, Emond M, Kosmerly S, O'Connor M, Chang F. Personality differences in the susceptibility to stress-eating: The influence of emotional control and impulsivity. Eat Behav. 2016;23:76–81. https://doi.org/10.1016/j.eatbeh.2016.07.009 .

Adam TC, Epel ES. Stress, eating and the reward system. Physiol Behav. 2007;91:449–58. https://doi.org/10.1016/j.physbeh.2007.04.011 .

Dubé L, LeBel JL, Lu J. Affect asymmetry and comfort food consumption. Physiol Behav. 2005;86:559–67.  https://doi.org/10.1016/j.physbeh.2005.08.023 .

Wagner HS, Ahlstrom B, Redden JP, Vickers Z, Mann T. The Myth of Comfort Food. Health Psychol. 2014;33:1552–7. https://doi.org/10.1037/hea0000068 .

Spence C. Comfort food: A review. Int J Gastronomy Food Sci. 2017;9:105–9. https://doi.org/10.1016/j.ijgfs.2017.07.001 .

Article   Google Scholar  

Choi J. Impact of Stress Levels on Eating Behaviors among College Students. Nutrients. 2020;12. https://doi.org/10.3390/nu12051241 .

Mohamed BA, Mahfouz MS, Badr MF. Food Selection Under Stress Among Undergraduate Students in Riyadh, Saudi Arabia. Psychol Res Behav Manag. 2020;13:211–21. https://doi.org/10.2147/PRBM.S236953 .

Errisuriz VL, Pasch KE, Perry CL. Perceived stress and dietary choices: The moderating role of stress management. Eat Behav. 2016;22:211–6. https://doi.org/10.1016/j.eatbeh.2016.06.008 .

Kandiah J, Yake M, Jones J, Meyer M. Stress influences appetite and comfort food preferences in college women. Nutr Res. 2006;26:118–23. https://doi.org/10.1016/j.nutres.2005.11.010 .

Mikolajczyk RT, Ansari WE, Maxwell AE. Food consumption frequency and perceived stress and depressive symptoms among students in three European countries. Nutr J. 2009;8:31. https://doi.org/10.1186/1475-2891-8-31 .

Tsenkova V, Morozink Boylan J, Ryff C. Stress eating and health. Findings from MIDUS, a national study of US adults. Appetite. 2013;69:151–5. https://doi.org/10.1016/j.appet.2013.05.020 .

Tomiyama AJ. Stress and Obesity. Annu Rev Psychol. 2019;70:703–18. https://doi.org/10.1146/annurev-psych-010418-102936 .

Broers VJ, de Breucker C, van den Broucke S, Luminet O. A systematic review and meta-analysis of the effectiveness of nudging to increase fruit and vegetable choice. 2017;27:912–20. https://doi.org/10.1093/eurpub/ckx085 .

Zellner DA, Loaiza S, Gonzalez Z, et al. Food selection changes under stress. Physiol Behav. 2006;87:789–93. https://doi.org/10.1016/j.physbeh.2006.01.014 .

Klatzkin R, Dasani R, Warren M, et al. Negative affect is associated with increased stress-eating for women with high perceived life stress. Physiol Behav. 2019;210:112639. https://doi.org/10.1016/j.physbeh.2019.112639 .

Article   CAS   PubMed   PubMed Central   Google Scholar  

Torres SJ, Nowson CA. Relationship between stress, eating behavior, and obesity. Nutrition. 2007;23:887–94. https://doi.org/10.1016/j.nut.2007.08.008 .

Oliver G, Wardle J, Gibson E. Stress and food choice: a laboratory study. Psychosom Med. 2000;62:853–65. https://doi.org/10.1097/00006842-200011000-00016 .

Kaiser B, Gemesi K, Holzmann SL, et al. Stress-induced hyperphagia: empirical characterization of stress-overeaters. BMC Public Health. 2022;22:1–11. https://doi.org/10.1186/s12889-021-12488-9 .

Bucher T, Collins C, Rollo ME, et al. Nudging consumers towards healthier choices: a systematic review of positional influences on food choice. Br J Nutr. 2016;115:2252–63. https://doi.org/10.1017/S0007114516001653 .

Westermann S, Rief W, Euteneuer F, Kohlmann S. Social exclusion and shame in obesity. Eat Behav. 2015;17:74–6. https://doi.org/10.1016/j.eatbeh.2015.01.001 .

Richardson AS, Arsenault JE, Cates SC, Muth MK. Perceived stress, unhealthy eating behaviors, and severe obesity in low-income women. Nutr J. 2015;14:122. https://doi.org/10.1186/s12937-015-0110-4 .

Laitinen J, Ek E, Sovio U. Stress-Related Eating and Drinking Behavior and Body Mass Index and Predictors of This Behavior. Prev Med. 2002;34:29–39. https://doi.org/10.1006/pmed.2001.0948 .

Janssen M, Chang B, Hristov H, Pravst I, Profeta A, Millard J. Changes in Food Consumption During the COVID-19 Pandemic: Analysis of Consumer Survey Data From the First Lockdown Period in Denmark, Germany, and Slovenia. Front Nutr. 2021;8:635859. https://doi.org/10.3389/fnut.2021.635859 .

Download references

Acknowledgements

The authors thank Elena Pateyuk for her contribution to the questionnaire and Lara Donik for statistical advice. We further thank all participants.

Open Access funding enabled and organized by Projekt DEAL. This work was supported by the enable competence cluster, an interdisciplinary cluster of nutrition research, funded by the German Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung, BMBF; grant number: 01EA1807H, 01EA1807A). The enable publication number is 78.

Author information

Authors and affiliations.

Institute for Nutritional Medicine, School of Medicine, Technical University of Munich, Georg-Brauchle-Ring 62, 80992, Munich, Germany

Kathrin Gemesi, Sophie Laura Holzmann, Hans Hauner & Christina Holzapfel

Research Group Public Health Nutrition, ZIEL – Institute for Food & Health, Technical University of Munich, Freising, Germany

Birgit Kaiser & Kurt Gedrich

Research Group Social Computing, Department of Informatics, Technical University of Munich, Garching, Germany

Monika Wintergerst & Georg Groh

Krcmar Lab, Department of Informatics, Technical University of Munich, Garching, Germany

Martin Lurz, Markus Böhm & Helmut Krcmar

Department of Informatics, University of Applied Sciences Landshut, Landshut, Germany

Markus Böhm

Else Kröner Fresenius Center for Nutritional Medicine, School of Life Sciences, Technical University of Munich, Freising, Germany

Hans Hauner

You can also search for this author in PubMed   Google Scholar

Contributions

SLH, BK, MW and ML developed the survey. KaG and CH performed data analysis and wrote the manuscript. All authors conducted and approved the survey and reviewed the manuscript.

Corresponding author

Correspondence to Christina Holzapfel .

Ethics declarations

Ethics approval and consent to participate.

The Ethical Committee of the School of Medicine at the Technical University of Munich had approved the survey (ethical vote: 729/20 S). All methods were carried out in accordance with relevant guidelines and regulations (e.g. Helsinki declaration). Informed consent was obtained from all participants.

Consent for publication

Not applicable.

Competing interests

KaG, SLH, BK, MW, ML, GG, MB, HK, and KuG declare that they have no competing interest. HH is a member of the scientific advisory board of Oviva AG (Zurich, Switzerland). CH is a member of the scientific advisory board of the 4sigma GmbH (Oberhaching, Germany).

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: supplementary results.

. Suppl. Table 1. Stress-eating behaviour of women and men who identified themselfes as stress-eaters. Suppl. Table 2. Sub-group specific consumption frequency of comfort foods in stressful situations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Gemesi, K., Holzmann, S.L., Kaiser, B. et al. Stress eating: an online survey of eating behaviours, comfort foods, and healthy food substitutes in German adults. BMC Public Health 22 , 391 (2022). https://doi.org/10.1186/s12889-022-12787-9

Download citation

Received : 12 November 2021

Accepted : 16 February 2022

Published : 24 February 2022

DOI : https://doi.org/10.1186/s12889-022-12787-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

BMC Public Health

ISSN: 1471-2458

• Submission enquiries: [email protected]
• General enquiries: [email protected]

• Our Vision Overview
• Why Health Equity
• Focus Areas
• Measuring RWJF Progress
• Confronting Structural Racism to Transform Health
• Grants Overview
• Active Funding Opportunities
• Awarded Grants
• Grantee Stories
• Grant Process
• Grantee Resources
• Insights Overview
• Our Research
• Advocacy And Policy
• About RWJF Overview
• Our Guiding Principles
• How We Work
• Impact Investments
• Staff And Trustees

Healthy Eating Research

Application is now closed.

Application closed as of: April 03, 2024 3:00 PM ET

Award Contact

Direct inquiries to:

Introduction & Purpose

As part of its commitment to building a national Culture of Health, the Robert Wood Johnson Foundation (RWJF) seeks ideas that advance health, equity, and wellbeing. Healthy Eating Research (HER) supports this goal through improving diet quality and nutrition for all Americans, addressing nutrition disparities, and reducing food and nutrition insecurity. HER’s mission is to support and disseminate research on policy, systems, and environmental strategies that promote healthy eating among children and families and advance nutrition security and health equity.

The purpose of this call for proposals (CFP) is to generate evidence on supportive family policies and programs that have strong potential to impact equitable access to nutritious food in communities, nutrition security, diet quality, and improved nutrition and health outcomes. We are especially interested in strategies to improve health outcomes for families with lower incomes and/or racially and ethnically diverse populations experiencing higher rates of diet-related chronic disease and/or health disparities. Findings will be used to guide and inform decisionmaking about policy and system changes that can advance nutrition equity and improve health.

Through this CFP, we seek to learn what works (or not) and why; who benefits the most from these policies and programs; and whether disparities are reduced. We are interested in solution-oriented research that focuses on policy, systems, and environmental (PSE) change at the national, state, local, and tribal levels. The CFP provides opportunities to develop and test innovative approaches to increasing access to healthy foods, such as through child nutrition programs; innovations in technologies; expanding program outreach and eligibility; simplifying program enrollment processes; identifying ways to improve equity within nutrition assistance programs; building the evidence base for these programs’ impacts on diet quality, food insecurity, nutrition, weight, and health; and areas for improvements and scalability.

HER issues CFPs to solicit scientifically rigorous, solution-oriented proposals from investigators representing diverse disciplines and backgrounds. To learn more about the program, we encourage applicants to visit the Healthy Eating Research website at www.healthyeatingresearch.org .

Eligibility & Selection Criteria

• Applicant organizations must be based in the United States or its territories.
• Awards will be made to organizations, not to individuals.
• Preference will be given to applicants that are either public entities or nonprofit organizations that are tax-exempt under Section 501(c)(3) of the Internal Revenue Code and are not private foundations or Type III supporting organizations. The Foundation may require additional documentation. 

Key Dates & Deadlines

• January 31–April 3, 2024  RWJF online system becomes available to applicants to submit concept papers.  
• April 3, 2024 (3 p.m. ET)   Deadline for receipt of concept papers. Those submitted after the deadline will not be reviewed.*  
• May 8, 2024 Applicants notified whether they are invited to submit a full proposal.  
• July 10, 2024 (3 p.m. ET)   Deadline for receipt of invited full proposals. Those submitted after the deadline will not be reviewed.*  
• September 2024 Notification of finalists.  
• November 2024 Awards begin.

*All proposals for this solicitation must be submitted via the RWJF online system. If you have not already done so, you will be required to register at  my.rwjf.org before you begin the application process. All applicants should log in to the system and familiarize themselves with online proposal requirements well before the final submission deadline. Staff may not be able to assist all applicants in the final 24 hours before the submission deadline.

Total Awards

Number of Awards:   Up to 10 awards will be funded. Amount of Awards:   Each award will be up to a maximum of $275,000 with a total of up to $2.5 million to be funded through this CFP. We encourage proposals that request lower budget amounts. Award duration:    Awards will be a maximum of 24 months in duration. We encourage proposals that request shorter periods (e.g., 12-18 months)

Grant Resources

• Preview a sample proposal before submitting PDF DOWNLOAD
• Full Funding Opportunity PDF DOWNLOAD
• FAQs for Applicants
• Program-Specific FAQs
• The Grant Process: What to Expect

Subscribe to receive Funding Alerts & more

Explore the latest in reflection and research from subject matter experts at RWJF and our wide network of partners.

Email address already subscribed. Please check your inbox to manage your subscriptions.

Subscribed!

Thank you. You are now subscribed.

Tell us what type of content you want to receive.

Be informed with our twice a month newsletter updating you with relevant news and research around a Culture of Health, as well as the latest funding opportunities.

Get funded by RWJF: Receive notifications when new funding opportunities are released.

Receive monthly updates on RWJF-sponsored research that informs many robust health policy debates on Capitol Hill, covering topics like health equity, improving access to quality healthcare, equitable housing, and more.

Shop talk for researchers. This monthly newsletter covers research news and opportunities from the Robert Wood Johnson Foundation.

Sometimes we have news, announcements or opportunities that don't quite fit the subscription parameters above. If you're interested, we'll send you this information under "There's more...". *If you've indicated you are an EU resident, we will only send these communications if you intentionally check this box.

Which profession or pursuit best describes you?

• Activist/Advocate
• Grant maker
• Policymaker or Staffer
• Public Health or Healthcare Professional
• Other Other

Area(s) Of Interest

• Birth justice/Maternal Health 1 of 14
• Caregiving 2 of 14
• Community Power 3 of 14
• Disability Rights/Disability Justice 4 of 14
• Economic inclusion 5 of 14
• Entertainment 6 of 14
• Food Systems 7 of 14
• Future of Health 8 of 14
• Global Learning 9 of 14
• Health Policy 10 of 14
• Impact Investing 11 of 14
• Leadership and Leadership Programs 12 of 14
• Physical and Built Environment 13 of 14
• Research 14 of 14

Unsubscribe

Stop receiving all emails from RWJF

CONTACT RWJF

50 College Road East Princeton, NJ 08540-6614

US Toll Free: (877) 843-7953

International: +1 (609) 627-6000

MANAGE YOUR GRANTS

MyRWJF Login

• Twitter link
• Facebook link
• LinkedIn link
• YouTube link

©2001-  

Robert Wood Johnson Foundation. All Rights Reserved. 

• Manage Email
• Privacy Statement
• Terms and Conditions

• Diet & Nutrition
• Dietary Supplements
• Herbs & Spices
• Sleep Centre
• Elderly care
• Sex & Intimacy
• Marriage & partnership
• Mental Health
• Mind Over Matter
• Consciousness
• Mantras & Chants
• Browse all topics
• Book Reviews

20 Common Questions on Healthy Eating

A registered dietician and nutritionist answers the most frequently asked questions about healthy diet and nutrition

When it comes to diet and nutrition, we all have many questions/doubts in our mind. As a registered dietician and nutritionist, I have had to sit on the ‘hot seat’ many times as clients fired away questions like in the ‘20 Questions’ game. Because of being unsure about what to eat and what not to eat, their nutrition needs can suffer.

Here, I present the top 20 common questions I have been often asked, along with their answers. I hope they will help clear many doubts you have regarding what constitutes healthy diet and also help you to improve your health.

Common questions on healthy eating (and their answers)

Question 1 – Are potato, corn, beans, and green peas good for one’s daily vegetable quota? Answer: Unfortunately, these are starches. You need to add vegetables to your daily meals. The more variety in colour you add, the better the variety of vitamins and minerals . Red bell peppers, yellow squash, purple eggplant, green beans, orange pumpkin—try them all. In a salad, soup, with your meat, in a stew, or even with your rice as a pilaf.

Question 2 – I eat well and sleep for 7 – 8 hours. Why do I still feel tired? Answer: Eating well is good. But you need to be hydrating well too. Signs and symptoms of dehydration include fatigue, nausea, headache, and dry skin. Ensure that you are drinking enough water and other healthy fluids, like coconut water , unsweetened juices, soups, lemon juice and herbal teas . The colour of your urine is a fairly reliable marker of your levels of hydration. If it’s darker than pale yellow, you need more fluids.

Question 3 – Will cutting out rice from my diet help me lose weight? Answer: The bottom line is calories. If your total calorie intake exceeds what your body can burn, then you gain weight. Although calories in rice add up fast [1/3 cup cooked rice has almost 100 calories], omitting rice may not be the only way to cut down on calories. Keep a food diary for a week and consult a nutritionist to help identify the hidden calories in your diet.

Question 4 – Is it okay to skip meals if I have a heavy breakfast? Answer: “Eat breakfast like a king, lunch like a prince, and dinner like a pauper” goes the old saying. However, our lifestyles have become much more sedentary since the feudal times. Eating a large breakfast if you are just going to be sitting at your desk the entire day isn’t recommended. It also doesn’t mean skipping lunch if you’ve eaten a lot in the morning. Instead, spread your calorie intake through the day. Large gaps between meals might lead to hyperacidity.

Question 5 – Why should one eat every 2 – 3 hours even if not hungry? Won’t eating so often cause weight gain? Answer: If you wait until you are hungry, you will eat more than you should be eating. Eating wholesome, balanced meals and snacks every 2 – 3 hours helps keep your metabolism high [which is the key to fat loss], blood sugar levels stable, and energy levels high. It is important to spread your overall calorie intake through the day and include all six food groups [starch, fruits, vegetables, meat/protein, dairy, and fats] in your daily meal plan.

Question 6 – Why do we have to eat banana, chikoo ( sapota ), and mango in moderation? Answer: These fruits contain more sugar than some other fruits; so the serving sizes are smaller for these fruits as they contain more calories. Half a medium-size mango and half a banana have the same calories as a tennis-ball size apple. The best time to have a banana , mango, or chikoo is post an aerobic workout such as a run to help replenish energy stores in your muscles.

Question 7 – Besides eating spicy or oily foods, what else could cause acidity? Answer: More than the food itself, what causes hyperacidity is irregular eating patterns. If your stomach remains empty for long due to large gaps between meals or snacks, the acid in your stomach, which aids in digestion, has nothing to digest. This creates an acidic environment in the stomach—what we call acidity. The best way to prevent hyperacidity is to eat small meals/snacks every 2 – 3 hours. You must also drink enough water to help digestion of foods and avoid drinking excess tea or coffee [two cups of tea/coffee are good enough].

Question 8 – Can beans and lentils alone fulfil my daily protein requirement? Answer: Beans and lentils are good sources of protein. If you are a vegetarian, there are other sources of protein that you can add to your diet. These include: paneer [cottage cheese], light tofu, low-fat or skim milk and yoghurt, and soy beans. Add grilled paneer to your salad, blend tofu with skim milk into soups to make thicker and creamier soups, and have buttermilk/ chaas as an afternoon snack.

Question 9 – I hate milk, can I take a calcium supplement instead? Answer: Taking a calcium citrate or calcium carbonate supplement is okay. However, your body is better at absorbing minerals from foods than from supplements. If you don’t like milk, you can try other sources of calcium such as broccoli, spinach, sardines, and fortified soy milk. Vitamin D is required for appropriate absorption of calcium too, so ensure you check vitamin D levels in your body if you do not get at least 10 – 15 minutes of exposure to the sun every day.

Question 10 – Can I cook food in vegetable-fat margarines instead of oil or butter? Answer: Yes, you can use vegetable-fat margarines for cooking. But note that an increased intake of saturated fat [butter, clarified butter/ghee, palm oils] will increase risk of heart disease along with a high fat diet. It is important to know exactly what percentage of your daily calorie intake should come from fat. Different types of cooking oils are fine to use in small amount in conjunction with a low-fat meal plan to reduce the risk of heart disease. The general rule is two teaspoons of oil per day per person. Ask your nutritionist to know exactly what percentage of your total calorie intake should come from fat as per your needs, medical history, and fitness goals.

Question 11 – Is it okay to cook food in extra virgin olive oil? Answer: Extra virgin olive oil is excellent when added in raw form on salads. However, avoid using it for cooking as it has a low smoke point. This means that when put in a hot pan, it smokes sooner than other oils. When oil smokes, it no longer remains healthy and becomes potentially carcinogenic [cancer-causing].

Question 12 – Which is the healthiest cooking oil? Answer: All oils, with the exception of palm oils, are healthy when used in small amounts. It is best to use a variety of oils for cooking—rice bran oil, vegetable oil, soybean oil, groundnut oil. Keep changing your oil every month; that way you can get both the heart healthy fats—monounsaturated and polyunsaturated fats—in your diet.

Question 13 – Is it okay to eat something immediately after exercise? If yes, what to eat? Answer: As per research, consuming a combination of carbohydrates [such as fruit, fruit juice, or a sports drink] and a protein supplement [such as a whey protein isolate powder, which is made from milk] within 20 minutes of your workout allows for best utilisation of the protein by your body. A whey protein supplement could help increase your stamina, strength, and endurance as proteins are the building blocks of muscle tissue. If you are a vegetarian, your overall protein intake might be low. Protein supplements allow your muscles to recover faster. They also allow you to get a larger amount of protein into your body quickly soon after a workout, which helps your body recover better.

Question 14 – Are biscuits and green tea good in-between meals snack options? Answer: Green tea is a great source of antioxidants and two cups a day is fine, if you have to drink it. However, liquids move out of your system faster than solids, leaving you hungry again. And if you are having green tea only for the anti-oxidants then there are many other sources for the same. As biscuits are a starch, the calories add up fast if you snack on them. So if you have to eat biscuits, go for the whole wheat, oat, or multi-grain variety. But fruits, nuts and seeds in small amounts, and low-fat cheese are better snacking options than biscuits.

Question 15 – What are some healthy “eat on the go” snack options? Answer: Home food is the best. Keep at least five small snacks in your office bag that are a combination of protein and carbohydrates. The protein will help keep your stomach full for some time and the carbohydrates act as the main source of energy for the brain. Half a peanut butter sandwich made with one slice of whole wheat bread and half a tablespoon of natural peanut butter, fruit yoghurt made with low-fat yoghurt and half a cup of fruit such as strawberries, pomegranate, or peaches are some examples.

Question 16 – Is it okay to drink water immediately after eating meals? Answer: As per Ayurveda, drinking water immediately after meals as the water dilutes the digestive juices, thereby slowing down digestion. It is best to drink water 30 minutes before or after a meal. Also, drink room temperature water or warm water, as it helps digestion. Cold water, on the other hand, hampers digestion and is best avoided.

Question 17 – What are good midnight snack food options? Answer: Foods with protein are best options for midnight snacks. A cup of skim milk, two-thirds cup of probiotic plain yoghurt, and even casein protein, which is a slow release protein supplement are a few good options.

Question 18 – What are some good sources of fibre in foods to help reduce cholesterol levels? Answer:   Here are some easy ways to add extra fibre into your meal plan:

• One tbsp psyllium husk in one glass warm water pre-bedtime.
• Ground flaxseed ; add to dosa or pancake batter.
• Oatmeal; great morning breakfast option post morning walk/cardio.
• About ¾ cup whole wheat bran flakes cereal with 200ml milk for breakfast.
• A bowl of beans and veggie salad with dinner packs in a great amount of fibre.

Fibre without water is of no use; ensure you continue drinking enough water. Also, minimise your intake of processed foods.

Question 19 – Will avoiding eating meals after 7pm help in weight loss? Answer: Yes, it helps. Most people become less active after 7pm. And also the metabolism is slower towards the end of the day. Hence, the calorie intake should be less as the day comes to a close. So, eating a light snack [ideally an hour before bed time] is okay but skipping meals entirely, especially if you are likely to stay up for several hours past 7pm, might make you hungry as you go to bed, leading you to indulge.

Question 20 – Isn’t peanut butter fattening? Answer: Peanut butter is mostly fats and protein, most of which are heart-healthy monounsaturated fats. If had in small amounts, peanut butter is a good source of fibre. It also contains vitamin B3 [niacin], which helps your body utilise energy out of foods; folate, which helps to form haemoglobin, and many other trace minerals. One serving of peanut butter is half a tablespoon.

Remember, diet is only half the story. A sedentary life will cancel all the benefits of healthy diet. If your activity level is low, include some kind of exercise in your daily routine. This will multiply the goodness of your healthy diet.

• Facebook Messenger

Looking for healthy, pastured meats in southern Ontario? Check out greenergrazing.ca

What is a healthy protein breakfast shake. I’ve heard that some of them have harmful ingredients.

What’s the best meal before you practice a sport that requires a lot of stamina?

In two different categories (11 and 12) you are telling us about oils we should use in our “healthy” diet. All I wanna ask is… should not we use lard over oils? Lard is supposed to be much more healthy, many websites like healthyfitnatural.com confirm this fact so I wanna check this with you. There are already a big number of studies that shows scientific evidences of lard benefits. Thanks!

LEAVE A REPLY Cancel reply

Save my name, email, and website in this browser for the next time I comment.

• Write for us
• List of authors
• Privacy policy
• Terms of use

Institute of Epidemiology & Health Care

• Eating behaviour questionnaires

Eating behaviour questionnaires 

Jump to:  General Nutrition Knowledge Questionnaire (GNKQ)     |     Child Eating Behaviour Questionnaire (CEBQ)     |     Baby Eating Behaviour Questionnaire (BEBQ): concurrent version     |     Baby Eating Behaviour Questionnaire (BEBQ): retrospective version     |     Adult Eating Behaviour Questionnaire (AEBQ)     |     Self-Regulation of Eating Behaviour Questionnaire (SREBQ)     |     Adolescents' Food Habit Checklist (AFHC)     |     Parental Feeding style Questionnaire (PFQ)     |     Food Choice Questionnaire (FCQ)     |     Food Preference Questionnaire (FPQ) for adolescents and adults     |     Food Preference Questionnaire (FPQ) for children     |     Dog Obesity Risk and Appetite Questionnaire (DORA)

General Nutrition Knowledge Questionnaire (GNKQ)

• Questionnaire (pdf)
• The questionnaire with the expected answers and the scoring information (pdf)
• Supplementary material (pdf)
• Development paper (pdf):  Kliemann N, Wardle J, Jonhson F and Croker H  Reliability and validity of a revised version of the General Nutrition Knowledge Questionnaire.  European Journal of Clinical Nutrition , 2016, 1-7

For any enquiries, please contact Dr Helen Croker:  [email protected]

Child Eating Behaviour Questionnaire (CEBQ)

• Questionnaire (pdf)
• Scoring information (pdf)
• Development paper (pdf) : Wardle J, Guthrie CA, Sanderson S and Rapoport L. Development of the Children's Eating Behaviour Questionnaire.  Journal of Child Psychology and Psychiatry , 42, 2001, 963-970.

This questionnaire is also available in Brazilian-Portuguese, Chilean-Spanish, Chinese, Dutch, French, Hebrew, Italian, Malaysian, Polish, Portuguese, and Spanish. For copies, please contact  Dr Clare Llewellyn : [email protected]

Baby Eating Behaviour Questionnaire (BEBQ): concurrent version 

(for parents whose baby is currently 3 months or less and is still fed milk only)

• Questionnaire (concurrent version) (pdf)
• Scoring information (concurrent version) (pdf)
• Development paper (pdf):  Llewellyn, C. H., van Jaarsveld, C. H., Johnson, L., Carnell, S., & Wardle, J Development and factor structure of the Baby Eating Behaviour Questionnaire in the Gemini birth cohort.  Appetite  57.2, 2011, 388-396.

This questionnaire is also available in several other languages.  For more information, please contact Dr Clare Llewellyn : [email protected]

Baby Eating Behaviour Questionnaire (BEBQ): retrospective version 

(for parents whose baby is currently older than 3 months and has already started solid food)

• Questionnaire (retrospective version) (pdf)
• Scoring information (retrospective version) (pdf)
• Development paper (pdf):  Llewellyn, C. H., van Jaarsveld, C. H., Johnson, L., Carnell, S., & Wardle, J Development and factor structure of the Baby Eating Behaviour Questionnaire in the Gemini birth cohort.  Appetite  57.2, 2011, 388-396

This questionnaire is also available in several other languages.  For more information, please contact  Dr Clare Llewellyn : [email protected]

Adult Eating Behaviour Questionnaire (AEBQ)

• Questionnaire and scoring information (pdf)
• Development paper (pdf):  Hunot C, Fildes A, Croker H, Llewellyn CH, Wardle J, Beeken RJ. Appetitive traits and relationships with BMI in adults: development of the adult Eating Behaviour Questionnaire.  Appetite  105, 2016, 356-363.

Self-Regulation of Eating Behaviour Questionnaire (SREBQ)

• Scoring information (pdf)
• Development paper (pdf):  Kliemann N, Beeken B, Wardle J, Johnson F Development and validation of the Self-Regulation of Eating Behaviour Questionnaire for adults.  International Journal of Behavioural Nutrition and Physical Activity , 2016, 13:87.

For any enquiries, please contact Dr Fiona Johnson:  [email protected]

 Adolescents' Food Habits Checklist (AFHC)

• Development paper (pdf):  Johnson F, Wardle J and Griffith J. The Adolescent Food Habits Checklist: reliability and validity of a measure of healthy eating behaviour in adolescents.  European Journal of Clinical Nutrition , 56, 2002, 644-649.
• Key references AFHC (pdf)

Parental Feeding style Questionnaire (PFQ)

• Development paper (pdf):  Wardle J, Sanderson S, Guthrie CA, Rapoport L and Plomin R. Parental feeding style and the inter-generational transmission of obesity risk.  Obesity Research , 10, 2002, 453-462.

Food Choice Questionnaire (FCQ)

• Development paper (pdf):  Steptoe A, Pollard TM, Wardle J. Development of a measure of the motives underlying the selection of food: the Food Choices Questionnaire.  Appetite , 1995, 25, 267-284.

Food preference questionnaire (FPQ) for adolescents and adults

• Questionnaire including scoring information (pdf)

Please contact Andrea Smith ( [email protected] ) for further enquiries about the adolescent and adult food preferences questionnaire. Telephone (UK): +44 (0) 20 7679 1736.

Food Preference Questionnaire (FPQ) for children

• Development paper (pdf) : Fildes A, van Jaarsveld CHM, Llewellyn CH, Fisher A, Cooke L, Wardle J. Nature and nurture in children's food preferences. The American Journal of Clinical Nutrition , 2014, 99, 911-917.

Dog Obesity Risk and Appetite Questionnaire (DORA)

• Development paper (pdf) : Raffan E, Smith SP, O'Rahilly S, Wardle J. Development, factor structure, and application of the Dog Obesity Risk and Appetite (DORA) questionnaire.  PeerJ , 2015, 3:e1278.

Healthy Eating Research

• Program Staff
• National Advisory Committee
• Our Commitment to Health Equity
• Resources for Researchers
• Current Funding Opportunities
• Healthy Communities
• Diet Quality & Healthy Weight
• Early Childhood
• Food Access
• Food Marketing
• Food Retail
• Nutrition Policy & Programs
• Pricing & Economics
• School & After School
• List of All Research & Publications
• Ages 0-2 Feeding Recommendations
• Ages 0-5 Beverage Recommendations
• Ages 2-8 Feeding Recommendations
• Ages 5+ Beverage Recommendations
• Grantee Announcements

Healthy Eating Research Commissioned Research Funding Opportunity

Click here for more information and to download our abstract form.

Healthy Eating Research is excited to share that we have research funds available for small commissioned studies and papers. Commissioned projects are intended to address time-sensitive research questions; all projects should aim to identify and evaluate policies, environmental approaches, and systems changes that advance health and nutrition equity in the areas of healthy food access and security, diet quality, and healthy weight, thus improving inequities in social structures and community conditions that are the main drivers of diet and weight-related health disparities.

Types of Projects

Commissioned Studies:

HER’s commissioned research is intended to fund small studies that produce timely, policy-relevant research. Commissioned analyses/studies will be for a duration of up to 12 months and a maximum of $100,000 (though please note, the average amount of most studies funded is $40,000-50,000). Examples of project activities may include primary data collection and analysis, secondary data analyses, in-depth case studies, qualitative research needs assessments, modeling studies, and policy analyses.

Deliverables could include data analyses and summary reports, manuscripts for publication in a peer-reviewed journal, measurement tools (or their validation), reports with recommendations, policy and/or research briefs, or other deliverables as appropriate.

Commissioned Papers:

HER also commissions papers, including manuscripts for publication in peer-reviewed journals, as well as evidence-based reports and briefs published outside the peer-reviewed literature.

Issue or Research Briefs are intended to distill existing published research into an easily digestible brief format for an advocate, policymaker, or practice audience (funding typically ranges between $1,500 and $2,500; timelines vary based on the time-sensitive nature of the topic).

Manuscripts for publication in peer-reviewed journals typically include Narrative or Scoping Reviews (up to $6,000 for a duration of up to 12 months) and Systematic Reviews (up to $8,000 for a duration of up to 12 months). Researchers are the primary audience for these deliverables.

Proposals should focus on the investigation of policy, systems, and environmental change strategies, and consider nutrition-related health disparities. The goal of Healthy Eating Research is to help all children achieve optimal nutrition and address nutrition-related equity. Therefore, all commissioned research must have the potential to impact groups at highest risk for poor health and well-being, and nutrition-related health disparities. All proposed studies must have a clear impact on children (0-18) and families in the U.S. HER’s Equity for Researchers document offers tools to help researchers understand and incorporate racial equity principles into their research in order to advance health equity. For more information on past studies funded, visit our Research and Publications webpage .

How to Apply

Please submit an abstract following the abstract template via email to Healthy Eating Research ( [email protected] ) with the subject “ATTN: HER Commissioned Research.” The abstract will be reviewed by HER staff and, if considered in-scope for this funding opportunity, the applicant will be invited to submit a full proposal. All abstracts and full proposals will be reviewed on a rolling basis.