Skip to main content

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.

Changes in diet and physical activity in the 1990s in a large British sample (1958 birth cohort)

Abstract

Objectives: To investigate whether adults studied in 1991 and 1999 (at ages 33 and 42 y) improved their diet and their physical activity level, in the direction of recommendations issued during the same period.

Design: Longitudinal 1958 British birth cohort study.

Setting: England, Scotland and Wales.

Participants: All births, 3rd–9th March, 1958. A minimum of 11 341 participants provided data at 33 y, 11 361 at 42 y.

Main outcome measures: Frequency of leisure time activity and consumption of (i) fried food, (ii) chips, (iii) wholemeal bread and (iv) fruit and salad/raw vegetables, at 33 and 42 y.

Results: Most people changed their physical activity and dietary habits over the 8-y period. About a third of men and women increased, and a third decreased their activity frequency. Findings for fried food consumption were similar. A significantly greater proportion of cohort members decreased their chips consumption (32%), rather than increased it (17%) and increased their fruit and salad consumption (30%), rather than decreased it (25%). In all, 26% of men and 33% of women consistently ate, or switched to eating mostly wholemeal bread, while 56% of men and 48% of women consistently ate less or switched to eating less. Social gradients were seen for activity and diet in 1991, but associations between social factors or body mass index and change in activity or diet were inconsistent.

Conclusions: Lifestyle habits such as dietary intake and physical activity are slow to change. Current health promotion strategies may need to be supplemented with additional methods to affect the desired change in these habits.

Sponsorship: TJ Parsons holds a Medical Research Council Special Training Fellowship in Health Services and Health of the Public Research.

Introduction

The 1990s was an interesting period in the UK in terms of health messages relating to diet and physical activity. In 1991, The Committee of Medical Aspects of Food Policy (COMA) published its report on dietary reference values for energy and nutrient requirements, and for the first time there were clear goals for population average intakes (Department of Health, 1991). Specific recommendations included a reduction in population average total fat intake (to 35% of food energy intake) and an increase in fibre intake (from 13 to 18 g/day) (Department of Health, 1991) (these recommendations existed previously, but in a less specific form (Department of Health and Social Security, 1984)). The COMA recommendations enabled the Health of the Nation document to set a target for fat intake; the proportion of the population deriving less than 35% of their food energy from total fat should be at least 50% by 2005 (Department of Health, 1992). In 1986/87, only 12% of men and 15% of women derived 35% or less of their energy from fat (The Dietary and Nutritional Survey of British Adults, 1990). The recommendations were reinforced by the ‘Balance of Good Health’ (1994), depicting a plate of different foods, translating dietary recommendations into proportions of food types making up a healthy diet (Food Standards Agency, 2002). Eight simple food-related guidelines accompanied the food plate including, for example, ‘don’t eat too many foods that contain a lot of fat’ and ‘eat plenty of foods rich in starch and fibre’. In 1994, a COMA report on diet and heart disease recommended that fruit and vegetable consumption in the UK should increase by 50%, from an average of three, to at least five portions per day (Department of Health, 1994). The ‘5-a-day’ message to promote fruit and vegetable consumption, already in use in the US since 1991, was adopted in the UK, and the ‘Balance of Good Health’ guidelines were modified to include this message. From 1998, the British Dietetic Association ran ‘Give me 5’, an educational campaign on fruit and vegetable consumption. (http://www.bda.uk.com/faw01.html).

The accepted guidelines for physical activity in the early 1990s were to take part in vigorous activity at least three times per week, for 20 min or more on each occasion (American College of Sports Medicine, 1990; Allied Dunbar Health and Fitness Survey, 1992). In 1995, the government's consultation paper ‘More people more active more often’ (Department of Health, 1995) followed new US recommendations (Pate et al, 1995) and encouraged people to take 30 min of moderate activity on a daily basis (at least 5 days of the week). This activity could be accumulated, for example in 10 min bouts throughout the day. Between 1996 and 1998, the Health Education Authority ran the ACTIVE for LIFE campaign to promote the new recommendations, using among other tools, advertising via the mass media (Hillsdon et al, 2001).

This study investigates, in a large British population sample, the extent to which adults changed their physical activity and dietary patterns over the period 1991–1999 (from age 33 to 42 y), and whether any changes were in the same direction as the health promotion messages advised. We looked specifically at whether leisure time activity increased and, using consumption frequencies of fried food, chips, wholemeal bread, and fruit and salad combined as proxies, whether fat intake was reduced, and fibre and fruit and vegetable intake increased. We also looked at the characteristics of those who changed their activity or eating habits in terms of social class, education and body mass index (BMI).

Participants and methods

Study population

Data are from the 1958 British birth cohort, studied in 1991 (at age 33 y) and 1999 (at 42 y). The original sample included all children born in England, Scotland and Wales from 3rd–9th March 1958; surviving children were contacted at ages 7, 11, 16, 23, 33 and 42 y (Butler & Bonham, 1963). The target samples at 33 and 42 y were 15 600 and 16 460 subjects, respectively (Introduction to the National Child Development Study (NCDS) 2002). Sample attrition has resulted in a slight under-representation of those who are most disadvantaged, but the remaining sample is generally representative of the original sample (Ferri, 1993). We looked at the data from 42 y; the distribution of social class for those with diet and activity data of 42 y was very similar to the distribution for those with social class data at birth.

Measures

Physical activity

In 1991 and 1999, participants were asked how often they took part in any activity, from most days to less than twice per month (six categories). The activity show-card listed: any competitive sports, ‘keep fit’ or aerobics classes, circuit training, weight training or other repeated exercises (at home or in the gym) (listed in 1991 only), running or jogging, swimming, cycling, going for walks, taking part in water sports, outdoor sports, dancing, any other sport or leisure activity that involves physical exercise.

Diet

In 1991, participants were asked how often they ate fresh fruit in summer, salads or raw vegetables in winter, chips, and fried food, from more than once a day to never (six categories). In 1999, the questions were similar except that season was not specified for fruit or salad/raw vegetables, and food fried in oil and hard fat were separate questions. Since 90% of men and 95% of women consumed food fried in hard fat on less than 1 day per week, we used food fried in oil as our fried food variable in 1999. In 1991, participants indicated what sort of bread they usually ate most of, from a list of eight options, including wholemeal bread and wholemeal pitta bread, and in 1999, how often they ate (a) wholemeal bread and (b) other types of bread. We calculated the proportion of people eating mostly wholemeal bread at both time-points.

We combined fruit and vegetable consumption to obtain an indicator of overall fruit and vegetable consumption, and used fried food and chips as indicators of fat intake. Four categories were constructed for each variable.

Other measures

Social class was assessed at birth, based on father's occupation, according to the UK 1951 General Registrar's classification. Four categories are used here: (i) classes I and II (professional and managerial), (ii) III-NM (skilled non-manual), (iii) III-M (skilled manual), and (iv) IV and V (semiskilled and unskilled manual), and those recorded as having ‘no male head of household’. Qualifications; in 1991 (age 33 y) cohort members were asked to report qualifications achieved to date. We use five categories: (i) none, (ii) less than O level, (iii) O levels (were usually taken at 16 y in school), (iv) A levels (usually taken at 18 y in school), and (v) higher qualifications (postschool education). BMI (kg/m2) at 33 y was calculated from height measured to the nearest centimetre and weight measured with indoor clothing, without shoes, to the nearest 0.1 kg.

Data analysis

All analyses were performed using SPSS for Windows, version 10.0, and carried out for males and females separately. A Wald-type test (which allows for within-individual dependency) was used to test for a change in distribution of diet or activity between 1991 and 1999. For further assessment of the pattern of changes, we used a McNemar test (proportions of individuals increasing and decreasing their frequency of activity or diet) and log-linear models (Agresti, 1990). The extent to which people maintained their frequency level between time-points was measured by Kappa. Trends in the likelihood (percentage) of people increasing/decreasing their activity (or consumption of chips, fried food or fruit and salad) compared with remaining stable across the (i) social class, (ii) educational level and (iii) BMI quartile were assessed using a χ2-test.

Results

Physical activity

The distribution of physical activity was similar in 1991 (at age 33 y) and 1999 (at 42 y) for men and women (Table 1). Although the cross-sectional distribution suggests little change over the period, the longitudinal comparison shows only a modest level of stability, similar for men and women, with 39 and 37%, respectively, maintaining the same frequency of activity (Kappa=0.18 for men, 0.15 for women) (Table 2). About a third of men and women increased their frequency and a third decreased it; in men the pattern was symmetrical where the number changing from activity frequency i to activity frequency j was similar to that changing from j to i (Agresti, 1990). Expressing our results in the context of the current UK recommendations, about four in 10 men and women either maintained their activity at four–seven times per week, or demonstrated any increase in their activity level. Thus, the majority (about six in 10) either reduced their activity level or maintained it below the recommendation. About 15% of the sample were inactive at both time-points.

Table 1 Leisure activity and dietary habits in 1991 (33 y) and 1999 (41 y)
Table 2 Changes in frequency of leisure activity and diet from 1991 (33 y) to 1999 (41 y)

Diet

The frequency distributions for chips consumption were significantly different in 1991 and 1999 for both men and women (Table 1). In all, 50% of men and 52% of women maintained their consumption frequency (Kappa=0.23 and 0.24, respectively) while 17% of men and 18% of women increased their consumption and 33% of men and 30% of women decreased it (Table 2). For fried food the distributions in 1991 and 1999 were significantly different in women but not men, and there was no clear trend (Table 1). In all, 38% of men and 40% of women maintained their consumption frequency (Kappa=0.11 in both genders), and similar proportions (about 30%) increased and decreased their consumption (Table 2).

The distributions of fruit/salad/raw vegetable consumption were significantly different in 1991 and 1999, with a tendency in both sexes towards increasing consumption (Table 1). Longitudinally, about 45% of men and women maintained the same consumption frequency (Kappa=0.23 and 0.21, respectively), while a significantly greater proportion increased (30%) rather than decreased (25%) their consumption (Table 2). Although 36% of men and 44% of women either maintained their consumption at more than once a day or improved it: more than half the study population maintained an intake of once a day or less, with four in 10 men and two in 10 women reporting eating fruit/salad/raw vegetables less than daily on both occasions.

In 1991, 21% of men and 25% of women were eating mostly wholemeal bread, that increased to 26 and 33%, respectively, in 1999 (data not presented). In all, 26% of men and 33% of women either consistently ate mostly wholemeal bread at both time-points or switched to consuming mostly wholemeal bread, while 56% of men and 48% of women ate less wholemeal than other types of bread on both occasions, or switched from eating mostly to less wholemeal bread.

Patterns of activity and diet with social class, education and BMI

Physical activity and dietary habits in 1991 (33 y) by education are presented in Table 3. All habits showed some patterning by education, with greater proportions of the less educated being inactive and consuming fruit/salad/raw vegetables infrequently than the more educated, and greater proportions of the more educated consuming chips and fried food infrequently (Table 3). The trend for those less educated to be less active was not evident for the high-frequency category in men or women (data not presented). Similar results were observed for social class, except that activity and fruit/salad/raw vegetable consumption showed no pattern in men. Relationships were also seen with BMI; those more active had lower BMIs, most vs least active 25.35 vs 25.81 kg/m2 (P=0.001) in men, 24.45 vs 24.88 kg/m2 (P=0.014) in women. Women (not men) who ate more chips had higher BMIs, high vs rare frequency 25.19 vs 24.53 kg/m2 (P=0.03), (25.43 vs 25.73 kg/m2, P=0.30 in men). Similarly women (not men) who ate more fried food had higher BMIs, high vs rare frequency 25.52 vs 24.59 kg/m2 (P=0.003), (25.70 vs 25.76 kg/m2, P=0.78 in men). Men and women who ate more fruit and salad had higher BMIs, high vs rare frequency in men 25.83 vs 25.40 kg/m2 (P=0.05), in women 24.90 vs 24.13 kg/m2 (P=0.02).

Table 3 Physical activity and dietary habits in 1991 (33 y) by education at 33 y

The proportions of men and women who increased or decreased their frequency of activity were broadly similar in terms of social class at birth, educational level at 33 y or BMI at 33 y (Table 4). The proportions of men and women increasing their consumption of chips were similar across the social classes, educational levels and BMI quartiles, but there was a trend for a greater proportion of those from the lower social classes and educational levels to decrease their chips consumption. Among women, the proportion who decreased their chips consumption also increased slightly with increasing BMI, so that a greater proportion of those with a higher BMI decreased their consumption (Table 4). Patterns for fried food showed that greater proportion of those from higher social and educational BMI groups, and in women, BMI groups, increased their fried food consumption, while, as for chips, greater proportions from the lower social/educational/BMI groups decreased their consumption. Among men and women who increased their fruit/salad/raw vegetable consumption, there was little difference by social/educational/BMI group. Among those who decreased their fruit/salad/raw vegetable consumption, men of lower social class and education were more likely to decrease consumption, as were women of lower education level or BMI group (Table 4).

Table 4 Percentages of those who changed their frequency of physical activity or diet from 1991 (33 y) to 1999 (41 y) by social class, qualifications and BMI

Discussion

Within the UK, the 1990s saw major revisions of the physical activity recommendations—to be physically active on at least 5 days per week (American College of Sports Medicine, 1990; Department of Health, 1995), reinforcement of the recommendations to reduce fat and increase fibre intake (Department of Health and Social Security, 1984; Department of Health, 1991) and new recommendations for fruit and vegetable consumption—to eat five portions per day (Department of Health, 1994). Concurrently, greater recognition was given to the potential role of these lifestyle factors in the prevention of noncommunicable disease (Department of Health, 1994; 1998) No reference was made to the dietary or physical activity recommendations in the questions asked to the 1958 birth cohort members, but this study allows an insight into national patterns of diet and activity change over the same period that the recommendations received widespread attention.

There are certain limitations to our data. We lack information on occupational physical activity, and on duration of physical activity, but believe that frequency is a useful measure in itself, particularly as the current recommendations state that activity can be accumulated throughout the day, and shorter episodes will be more likely to be forgotten. Although food frequency questions rely on memory, they allow day-to-day variation to be taken into account, and do not heavily burden subjects or influence their eating habits. Information on amounts of food eaten was not available in our study, but others have found that most of the variation in food intake is explained by consumption frequency rather than portion size (Willett, 1998). We would not claim that our food items encompass all dimensions of diet but importantly, that they might be markers for different types of diet. Diets rich in fruit, vegetables and high fibre foods (eg wholemeal bread) and lower in fat, sometimes termed a ‘prudent’ or ‘health-conscious’ diet, and also physical activity have been found to be associated with better self-rated health (Whichelow & Prevost, 1996; Osler et al, 2001). A low prudent diet score has been found to independently predict increased mortality after adjusting for other possible confounders (Osler et al, 2001). Although our study is restricted to a narrow age in adulthood, this group is an important group to focus on, not only in terms of their own health but because of their influence on young children.

We cannot distinguish between time and ageing trends, but an advantage of longitudinal data is that changes for individuals can be assessed, which may be obscured in cross-sectional information. For example, we found the distribution of physical activity for men and women to be very similar at both time-points, but this hides the fact that two-thirds of men and women changed their activity level between 1991 and 1999, a situation also reported in a longitudinal study of the ACTIVE for LIFE campaign (Hillsdon et al, 2001).

In our study, as in others (Health Survey for England, 1999), participation in physical activity was strikingly low; the recommendation is to participate in moderate activity on at least five, preferably all days of the week (Department of Health, 1995), yet at both time-points a third of the population participated less than once a week. About 40% of people maintained their activity level at 4–7 days per week or increased their level in this direction, but importantly, 60% did not. The Allied Dunbar Health and Fitness Survey identified several barriers to physical exercise, namely lack of time, needing to relax in spare time, not being the sporty type and having an injury or disability (Allied Dunbar Health and Fitness Survey, 1992). The most common reasons for stopping exercise were work reasons, loss of interest and needing time to do other things. These barriers were identified before the change in recommendations, which shifted from emphasising more vigorous and sporting activities to moderate and lifestyle activities, and may or may not be the same today. However, it is also worth noting that participation in physical activity tends to decrease with age (Health Survey for England, 1999) and the fact that we found little change in activity distribution, particularly in men, could be viewed as encouraging.

In respect of diet, our study showed a reduction in consumption frequency of chips, little change in fried food consumption (although the high-frequency category in women showed an increase) and some suggestions of improvement in consumption of fruit/salad/raw vegetables, and wholemeal bread. Cross-sectional data from the National Food Survey suggest that although during the 1990s, fruit consumption increased, total fresh vegetable consumption remained almost constant and wholemeal bread consumption declined slightly (Department for Environment, Food and Rural Affairs, 2001).

It is possible that the increase in fruit and salad consumption is due to wording differences at the two time-points. Substantially, more salad is eaten in summer (Whichelow, 1993), and therefore the year-round average (1999) would be expected to be higher than an estimate for winter only (1991), which we found in our data (not shown). We would also expect year-round fruit consumption (1999) to be lower than in summer only (1991). In fact, we saw more people at the extremes of the distribution at 42 y, that is, consuming fruit more than once a day or less than once a week (data not shown). The increase in people consuming fruit less than once a week may be due to seasonal changes, but the increase in people consuming fruit more frequently than once a day is unlikely to be so, and therefore may be attributed to behaviour change.

We found that in 1991 (at baseline), physical activity and dietary habits were related to the educational level, and to a lesser extent, to social class and BMI; similar cross-sectional relationships have been found in other studies (Whichelow & Prevost, 1996; Health Survey for England, 1999; National Diet and Nutrition Survey, 2002). However, such a relationship between a lifestyle habit and social class (eg), does not mean that change in that habit will necessarily be related to social class. For example, we found that the proportions changing their frequency of physical activity (either increasing or decreasing), did not vary by social class, education or BMI group. We did, however, find relationships for chips and fried food, and the patterns were unexpected; although similar proportions of men and women from each social class, education level or BMI group increased their chips consumption, greater proportions from the lower social and educational groups, and in women, higher BMI groups, decreased their consumption. For fried food, the higher social, educational or BMI groups were more likely to increase consumption and the lower groups to decrease consumption. These patterns would act to weaken the cross-sectional associations between chips or fried food and social class or education over time. In contrast, greater proportions of men from the lower social classes and men and women from the lower educational groups decreased their fruit/salad/raw vegetable consumption, patterns that would act to strengthen the cross-sectional relationships between fruit/salad/raw vegetable consumption and social class or education over time.

In summary, we found that most people changed the frequency of their physical activity level and dietary intake of fried food, fruit and salad/raw vegetables and wholemeal bread over an 8-y period suggesting that activity and dietary habits are amenable to change. However, between 1991 and 1999, there was no overall improvement in physical activity or consumption frequency of fried food, and some improvement in consumption of fruit and salad/raw vegetables, wholemeal bread, and chips consumption. The changes in consumption of fried food and chips suggest that those from lower social groups were more likely to improve their lifestyles; however the changes in consumption of fruit/salad/raw vegetable consumption suggest the opposite—that it was those from higher social groups or with a lower BMI who made positive lifestyle changes. This suggests that health messages may be taken up differentially by different social groups, for reasons that would be worth exploring, but perhaps related to issues such as food preferences, availability, and cost. For the population as a whole, it is evident that changes in diet and physical activity habits in the direction recommended by health promotion messages is slow, and perhaps additional strategies will be needed if this rate of progress is to be increased.

References

  1. Agresti A (1990): Categorical Data Analysis. New York: Wiley.

    Google Scholar 

  2. Allied Dunbar Health and Fitness Survey (1992): A report on activity patterns and fitness levels. London: Sports Council and Health Education Authority.

  3. American College of Sports Medicine (1990): American College of Sports Medicine position stand. The recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness in healthy adults. Med. Sci. Sports Exerc. 22, 265–274.

  4. Butler NR & Bonham DG (1963): Perinatal Mortality. Edinburgh: Churchill Livingstone.

    Google Scholar 

  5. Department for Environment, Food and Rural Affairs (2001): National Food Survey 2000. London: The Stationary Office http://www.defra.gov.uk/esg/Work_htm/publications/cf/nfs/current/nfs.htm.

  6. Department of Health (1991): Dietary reference values for food energy and nutrients for the United Kingdom. Report of the Panel on Dietary Reference Values of the Committee on Medical Aspects of Food Policy. London: The Stationary Office.

  7. Department of Health (1992): The Health of the Nation: a Strategy for Health in England. London: The Stationary Office.

  8. Department of Health (1994): Nutritional aspects of cardiovascular disease. Report of the cardiovascular review group of the Committee on Medical Aspects of Food Policy. London: The Stationary Office.

  9. Department of Health (1995): More people more active more often. Physical Activity in England: A Consultation paper. London: The Stationary Office.

  10. Department of Health (1998): Nutritional aspects of the development of cancer. Report of the working group on diet and cancer of the Committee on Medical Aspects of Food and Nutrition Policy. London: The Stationary Office.

  11. Department of Health and Social Security (1984): Diet and cardiovascular disease. Committee on Medical Aspects of Food Policy. Report of the Panel on Diet in Relation to Cardiovascular Disease. London: The Stationary Office.

  12. Ferri E (1993): Life at 33: the Fifth Follow-up of the National Child Development Study. London: National Children's Bureau.

    Google Scholar 

  13. Food Standards Agency (2002): The Balance of Good Health. http://www.nutrition.org.uk/information/dietandhealth/balanceddiet.html.

  14. Health Survey for England (1999): Cardiovascular disease ‘98. London: The Stationary Office http://www.doh.gov.uk/public/hse98.htm.

  15. Hillsdon M, Cavill N, Nanchahal K, Diamond A & White IR (2001): National level promotion of physical activity: results from England's ACTIVE for LIFE campaign. J. Epidemiol.Community Health 55, 755–761.

    CAS  Article  Google Scholar 

  16. Introduction to the National Child Development Study (NCDS) (2002): http://www.cls.ioe.ac.uk/Cohort/Ncds/Documentation/maindocs.htm.

  17. National Diet and Nutrition Survey (2002): Adults aged 19 to 64 years, Vol 1. London: The Stationary Office. http://www.foodstandards.gov.uk/science/101717/ndnsdocuments/printedreportpage.

  18. Osler M, Heitmann BL, Hoidrup S, Jorgensen LM & Schroll M (2001): Food intake patterns, self rated health and mortality in Danish men and women. A prospective observational study. J. Epidemiol. Community Health 55, 399–403.

    CAS  Article  Google Scholar 

  19. Pate RR, Pratt M, Blair SN, Haskell WL, Macera CA, Bouchard C, Buchner D, Ettinger W, Heath GW, King AC, Kriska A, Leon AS, Marcus BH, Morris J, Paffenbarger Jr RS, Patrick K, Pollock ML, Rippe JM, Sallis J & Wilmore JH (1995): Physical activity and public health. A recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine. JAMA 273, 402–407.

    CAS  Article  Google Scholar 

  20. The Dietary and Nutritional Survey of British Adults (1990): The Dietary and Nutritional Survey of British Adults. London: HMSO.

  21. Whichelow MJ (1993): Changes in dietary habits. In The Health and Lifestyle Survey: Seven Years On ed. BD Cox, FA Huppert & MJ Whichelow, pp 197–220. Aldershot: Dartmouth Publishing Company.

    Google Scholar 

  22. Whichelow MJ & Prevost AT (1996): Dietary patterns and their associations with demographic, lifestyle and health variables in a random sample of British adults. Br. J. Nutr. 76, 17–30.

    CAS  Article  Google Scholar 

  23. Willett W (1998): Nutritional Epidemiology. In Modern Epidemiology ed. KJ Rothman & S Greenland, pp. 623–642. Philadelphia: Lippincott Williams & Wilkins.

    Google Scholar 

Download references

Acknowledgements

Data obtained from the UK Data Archive, University of Essex (files: National Child Development Study, SN 3148, SN 4396). Data providers: Centre for Longitudinal Studies, Institute of Education and National Birthday Trust Fund, National Children's Bureau, City University Social Statistics Research Unit (original data producers).

Author information

Affiliations

Authors

Corresponding author

Correspondence to T J Parsons.

Additional information

Guarantors: TJ Parsons and C Power.

Contributors: TJP, CP and OM developed the hypotheses for the study, and participated in writing this paper. TJP and OM did the analyses.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Parsons, T., Manor, O. & Power, C. Changes in diet and physical activity in the 1990s in a large British sample (1958 birth cohort). Eur J Clin Nutr 59, 49–56 (2005). https://doi.org/10.1038/sj.ejcn.1602032

Download citation

Keywords

  • diet
  • physical activity
  • cohort study

Further reading

Search

Quick links