Cross-sectional studies have shown that vegetarians and vegans are leaner than omnivores. Longitudinal data on weight gain in these groups are sparse.
We investigated changes in weight and body mass index (BMI) over a 5-year period in meat-eating, fish-eating, vegetarian, and vegan men and women in the UK.
Self-reported anthropometric, dietary and lifestyle data were collected at baseline in 1994–1999 and at follow-up in 2000–2003; the median duration of follow-up was 5.3 years.
A total of 21 966 men and women participating in Oxford arm of the European Prospective Investigation into Cancer and Nutrition aged 20–69 years at baseline.
The mean annual weight gain was 389 (SD 884) g in men and 398 (SD 892) g in women. The differences between meat-eaters, fish-eaters, vegetarians and vegans in age-adjusted mean BMI at follow-up were similar to those seen at baseline. Multivariable-adjusted mean weight gain was somewhat smaller in vegans (284 g in men and 303 g in women, P<0.05 for both sexes) and fish-eaters (338 g, women only, P<0.001) compared with meat-eaters. Men and women who changed their diet in one or several steps in the direction meat-eater → fish-eater → vegetarian → vegan showed the smallest mean annual weight gain of 242 (95% CI 133–351) and 301 (95% CI 238–365) g, respectively.
During 5 years follow-up, the mean annual weight gain in a health-conscious cohort in the UK was approximately 400 g. Small differences in weight gain were observed between meat-eaters, fish-eaters, vegetarians and vegans. Lowest weight gain was seen among those who, during follow-up, had changed to a diet containing fewer animal food.
The prevalence of obesity is growing worldwide and the resulting increased risk for diseases like type 2 diabetes, coronary heart disease, stroke, gall bladder disease and some cancers, and the ensuing economic costs to society, are of major concern.1 Obesity is difficult to treat; a previous meta-analysis indicated that the average weight loss, 5 years after completing structured weight-loss programs, was less than 3% of initial body weight.2 To control the obesity epidemic, prevention of people gaining weight during adulthood is therefore crucial.
Cross-sectional studies have shown that, at least among westernized populations, people who follow a vegetarian diet have lower body weight and body mass index (BMI) than omnivores.3, 4, 5, 6, 7 In the Oxford arm of the European Prospective Investigation into Cancer and Nutrition (EPIC-Oxford), which intentionally recruited a high proportion of people who did not eat meat, mean BMI was highest in meat-eaters, lowest in vegans, and intermediate in fish-eaters and vegetarians.5 It was also found that in the groups that did not eat meat, mean BMI was lower in those who had adhered to their diet for 5 or more years than among those who had followed their diet for a shorter period.8 We are aware of only one previous longitudinal study investigating the possible role of a vegetarian diet in preventing weight gain. This study of Seventh-day Adventists indicated a small but non-significantly lower weight gain between 1960 and 1976 in vegetarians below age 50 compared with non-vegetarians,9 although among older subjects the finding was reversed. However, these analyses were not adjusted for possible confounding factors. In the present paper, we assessed changes in weight over 5 years in the late 1990s in meat-eating, fish-eating, vegetarian, and vegan men and women participating in EPIC-Oxford.
Materials and methods
Subjects and study design
Subjects were selected from the EPIC-Oxford. The EPIC-Oxford cohort consists of 65 500 participants aged ⩾20 years and living in the UK between 1993 and 1999.10 The aim was to recruit participants with a wide range of diets by targeting vegetarians and vegans, as well as the general UK population. Participants were recruited through collaborating general practitioners or by post via vegetarian and vegan societies, vegetarian and health food magazines, or were friends or relatives of other participants. A detailed questionnaire that included questions on diet, height and weight, physical activity, medical history and current health, marital and job status, education and age at menarche in women was completed by 57 498 participants. Approximately 5 years after recruitment, a follow-up questionnaire was sent to those who had completed the baseline questionnaire and who had not died or been lost to follow-up; the response rate at follow-up was 68%. The study protocol was approved by a Multi-Centre Research Ethics Committee. The current study is based on subjects who completed the follow-up questionnaire and had no prevalent malignant neoplasm at baseline (n=36 956). Subjects were excluded if weight and height were measured and not self-reported (n=1389), if anthropometric data were missing or if change in body weight between baseline and follow-up exceeded 20 kg, because we considered that such a change was more likely to be the result of reporting error (n=2267), if subjects were aged ⩾70 years or had suffered a heart attack, stroke, angina or diabetes at baseline (n=4625), if the diet group was unclear at baseline and/or follow-up (529), or if values of any of the other variables in the analyses were missing (6180). This left a total of 21 966 subjects who were included in the analyses.
Participants were asked to report their body weight in stones and pounds or in kilograms, and their height in feet and inches or in centimetres. Data reported in stones and pounds were converted to kilograms; data reported in feet and inches were converted to centimetres. Body mass index was calculated as weight in kilograms divided by the square of height in metres. Anthropometric data at baseline and follow-up were used to calculate change in body weight and BMI. Annual weight gain during follow-up was calculated as the gain in weight from baseline to follow-up divided by the number of months of follow-up, multiplied by 12. The validity of self-reported height and weight at baseline was previously evaluated in a sample of 4808 EPIC-Oxford participants in whom height and weight were both self-reported and measured by a nurse.11 This study showed that self-reported and measured height and weight, respectively, were highly correlated (r>0.9). Height was overestimated by a mean of 1.23 cm in men and 0.60 cm in women, and the extent of overestimation was greater in older men and women, shorter men and heavier women. Weight was underestimated by a mean of 1.85 kg in men and 1.40 kg in women, and in both men and women there was a trend towards greater underestimation with increasing weight. Assuming that any difference in an individual's true weight and reported weight did not vary from baseline to follow-up, systematic misestimation of weight is unlikely to bias the results because the current study concerns changes in weight.
Classification of diet groups
Classification of diet groups was based on four questions: ‘Do you eat any meat (including bacon, ham, poultry, game, meat pies, sausages)? (Yes/No)’, ‘Do you eat any fish? (Yes/No)’, ‘Do you eat any eggs? (Yes/No)’ and ‘Do you eat any dairy products (including milk, cheese, butter, yogurt) (Yes/No)?’. Based on these data at baseline and follow-up, six diet groups were defined: subjects who had not changed their diet during the follow-up period were classified as ‘meat-eater’, ‘fish-eater’ (who did not eat meat but ate fish), ‘vegetarian’ (who did not eat meat or fish but ate eggs and/or dairy products) or ‘vegan’ (who did not eat any food of animal origin), subjects who during the follow-up period had changed their diet in one or more steps in the direction vegan → vegetarian → fish-eater → meat-eater were classified as ‘reverted’ and subjects who had changed their diet by one or more steps in the opposite direction were classified as ‘converted’. In addition to the questions used to classify the diet group, the participants' dietary intake during the previous 12 months was assessed by a 130-item food frequency questionnaire.
Classification of physical activity was based on questions relating to participants' physical activity at work and leisure time, averaged over summer and winter. Participants were assigned to one of four levels of physical activity at work: sedentary occupation, standing occupation, manual work and heavy manual work, with participants who were not currently employed assigned to the sedentary category. The average number of hours per week of cycling and other exercise, such as sports, jogging and aerobics/keep fit, was calculated and participants were assigned to one of four levels: 0, <3.5, 3.5–<7 and ⩾7 h per week. Data on physical activity at work and leisure time were combined and a physical activity index with four categories was created (1=inactive, 2=low activity, 3=moderately active and 4=very active). This index was shown to rank participants according to energy expenditure measured via heart rate monitoring.12 Based on physical activity index at baseline and follow-up, six groups were defined: subjects who had not changed their physical activity level during the follow-up period were classified into groups of ‘inactive’, ‘low activity’, ‘moderately active’ and ‘very active’, subjects who during the follow-up period had changed their physical activity level by one or more steps in the direction inactive → low activity → moderately active → very active were classified as ‘more active’, and subjects who had changed their physical activity index by one or more steps in the opposite direction were classified as ‘less active’.
Additional variables described below were also included in the analyses since preliminary analyses had indicated that they were each associated with weight gain. Information on smoking, marital status and current paid job were obtained from the questionnaires at baseline and follow-up. Based on these data, for each variable, subjects who had not changed their status during the follow-up period were classified as ‘yes–yes’ or ‘no–no’, and subjects who had changed their status from baseline to follow-up were classified as ‘yes–no’ or ‘no–yes’; for example, subjects who smoked at baseline but not at follow-up were classified as ‘yes–no’. Information on age, height, age at menarche and age at leaving school were obtained through the baseline questionnaire. Age at recruitment was categorized as 5-year age bands, age at menarche was categorized into groups of <12, 12, 13, 14 and ⩾15 years, and age at leaving school was dichotomized into groups of ⩽16 and >16 years.
At baseline, participants were also asked if they had modified their diet in the past year owing to overweight/obesity and if they were receiving long-term treatment for any illness. Subset analyses were performed excluding subjects who answered yes to these questions. The baseline questionnaire also provided information on the age at which the subjects became a fish-eater, vegetarian and vegan, respectively, and also on the participants' body weight at the age of 20. These data were used to test the hypothesis that young non-vegetarians who become vegetarian are leaner than young people who do not subsequently become vegetarian.
Data for men and women were analysed separately. Age-adjusted mean BMI at baseline and follow-up was calculated for meat-eaters, fish-eaters, vegetarians and vegans using multiple linear regression. The association between weight gain and diet group was analysed using multiple linear regression, adjusting for physical activity, smoking, marital status, current paid job, age at leaving school, age at menarche, and age, height and weight at baseline. For each variable, the means in the different groups were compared with the mean in the reference group using standard t-tests arising from the regression model. Heterogeneity in means across categories was assessed using F-tests. Adjusted mean values for BMI and weight gain and their 95% confidence intervals were calculated from the fitted values arising from the regression models. All analyses were performed using version 8.1 of the STATA statistical package (Stata Statistical Software: Release 8.0, College Station, TX, Stata Corporation, 2003).
The study included 5373 men and 16 593 women (Table 1). The year of recruitment ranged from 1994 to 1999 and the year of follow-up ranged from 2000 to 2003; the median duration of follow-up was 5.3 years (range 3.2–9.1 years). During the follow-up period, men and women gained approximately 400 g/year in body weight (P<0.001 for a null hypothesis of zero weight gain for both men and women). The mean annual increase in BMI was 0.12 kg m−2 in men and 0.15 kg m−2 in women. From baseline to follow-up, the proportion of overweight subjects (25 kg m−2⩽BMI<30 kg m−2) increased from 29.4 to 34.9% in men and from 19.1 to 24.2% in women, and the proportion of obese subjects (BMI⩾30 kg m−2) increased from 4.0 to 6.9% in men and from 5.7 to 8.4% in women.
The increase in age-adjusted BMI among subjects who were in the same diet group at baseline and follow-up is illustrated in Figure 1. The differences in mean BMI between the diet groups at follow-up were similar to those seen at baseline; mean age-adjusted annual increases in BMI in meat-eaters, fish-eaters, vegetarians and vegans were 0.12, 0.12, 0.12 and 0.10 kg m−2 in men (P for heterogeneity=0.556), and 0.15, 0.12, 0.15 and 0.12 kg m−2 in women (P for heterogeneity=0.017), respectively.
Multivariable-adjusted annual weight gains in all six diet groups (meat-eaters, fish-eaters, vegetarians, vegans, reverted and converted) are presented in Table 2. P for heterogeneity in weight gain between the diet groups was <0.05 for men and <0.001 for women. Compared with meat-eaters, mean annual weight gain was lower in vegans (P<0.05 for both men and women) and fish-eaters (women only, P<0.001). The lowest mean weight gains were seen in men and women classified as converted, in whom the mean annual weight gain was 40 and 29% smaller, respectively, compared with the mean annual weight gain in meat-eaters (P<0.001 for both sexes). The highest weight gains were seen in men and women classified as reverted, but these values were not significantly different from the mean weight gains in meat-eaters.
Table 2 also shows the associations between weight gain and the non-dietary variables. For each variable, P for heterogeneity between categories was ⩽0.01 for both men and women. For subjects who did not change their physical activity level during follow-up, there was a linear inverse relationship between physical activity and weight gain; in very active men and women the mean annual weight gain was 35 and 34% lower, respectively, compared with that seen in inactive subjects (P<0.01 and <0.001, respectively). Subjects who became more active during the follow-up period also gained less weight compared with subjects who were inactive at both baseline and follow-up (P<0.001 for both men and women). Smokers gained less weight compared with non-smokers, whereas subjects who quit smoking during follow-up gained more weight (P<0.001 in each case for both men and women). Women who started to smoke during follow-up gained less weight compared with non-smoking women (P<0.01). Compared with subjects who were married or living as married, mean annual weight gain was greater for unmarried women (P<0.01) and for men and women who got married (P<0.05 and <0.01, respectively) and smaller for men and women who were no longer married (P<0.01 and <0.001, respectively). Men who lost or quit their job gained less weight than men who had a job at both baseline and follow-up (P<0.01), whereas women who obtained work gained less weight than women working at both baseline and follow-up (P<0.01). Inverse associations between weight gain and age at leaving school (P<0.01 for men and P<0.001 for women) and age at menarche were seen. Age at baseline was inversely associated with weight gain in both sexes. With the exception of young men aged 20–24 years in whom a very large weight gain was seen, the influence of age was most pronounced at the age of 50 and above in both men and women. Weight gain was positively associated with baseline height and inversely associated with baseline weight (data not shown).
The association between diet group and weight gain was also analysed in subsets, excluding subjects who at baseline reported that they had modified their diet in the past year owing to overweight or obesity and excluding subjects receiving long-term treatment for any illness at baseline. The results were similar to those reported in Table 2: in the first subset, compared with meat-eaters, the mean annual weight gain was smaller in fish-eaters (women only, P<0.01), vegetarians (women only, P<0.05), vegans (men only, P<0.05) and the converted group (P<0.05 for both sexes). In the second subset, compared with meat-eaters, the mean annual weight gain was smaller in fish-eaters (women only, P<0.01), vegans (women only, P<0.05) and the converted group (P<0.01 for both sexes).
Restricting the analyses to subjects who were non-vegetarians at the age of 20 years (4521 women and 4706 men), we calculated BMI at the age of 20 years in two subgroups. In men who had become a vegetarian or vegan at recruitment to the study the mean BMI at the age of 20 years was 21.8 kg m−2, whereas in men who were not a vegetarian or vegan at recruitment the mean BMI at the age of 20 years was 22.2 kg m−2 (P for difference <0.001). In women, the corresponding figures were 21.4 and 21.5 kg m−2, respectively (P=0.02). These differences remained significant after adjusting for age at baseline, age at leaving school and age at menarche (data not shown).
It has been observed in cross-sectional studies that vegetarians are leaner than omnivores.3, 4, 5, 6, 7 This effect could be partly mediated by the intake of fibre, which is higher in a vegetarian diet10, 13, 14 and is a dietary factor that has been shown to prevent weight gain.1, 15, 16, 17 Further, a vegetarian diet may be lower in energy-dense micronutrient-poor food, which is also an established risk factor for obesity.1 In the current study, we assessed weight gain over 5 years during the late 1990s in meat-eating, fish-eating, vegetarian, and vegan men and women in the EPIC-Oxford cohort. The average weight gain in this population was approximately 400 g/year. Differences were seen between the diet groups; a statistically significant smaller weight gain was seen in the vegans, and in women also in fish-eaters, compared with the meat-eaters. However, in these groups the weight gain was still approximately 300 g/year. As obesity is associated with increased risk for many chronic diseases, the World Health Organization recommends that people should not gain more than 5 kg weight during adulthood.18 At a weight gain rate of 400 g/year, it would take less than 15 years to achieve this.
The small differences in annual weight gain between the meat-eaters and the vegetarians during the 5 years of follow-up are surprising, given the considerable difference in mean BMI between these groups (Figure 1). This finding raises the question of when the differences in mean BMI between the diet groups develop. In the current study, the lowest weight gain was seen in people who during the follow-up period had changed to a diet containing fewer animal food. This suggests that people who become ‘more vegetarian’ reduce or stop gaining weight for a period and that this could give rise to the subsequent differences in BMI. Another possibility is that people who become vegetarians are more health conscious and therefore likely to be relatively slim before they change their diet. Our data indicate that this could be a contributing factor, at least among men in whom those who became vegetarian or vegan after the age of 20 years had a 0.4 kg m−2 lower mean BMI at the age of 20 years compared with subjects who did not become vegetarian or vegan later in life. However, in these analyses we were not able to control for all possible confounding factors and the finding should therefore be interpreted with caution. Further research is needed to determine when and how the differences in BMI arise.
The optimal diet for weight control is hotly debated. Low-carbohydrate/high protein diets are popular at present, although their long-term efficacy and safety are not established.19, 20, 21 In the current study, among people who had not changed their diet during follow-up, the largest weight gain was seen in the meat-eaters, whereas the smallest weight gain was seen in the vegans. As reported previously, at baseline the percent of energy intake from protein was lowest among the vegans and highest among the meat-eaters, whereas the opposite was seen for carbohydrates.10 Thus, contrary to the hypothesis that high intakes of carbohydrates and low intake of protein may exacerbate weight gain, we observed the lowest weight gain in the group with the highest intake of carbohydrates and lowest intake of protein.
The association between weight gain and diet group was adjusted for a number of variables and we have also presented data for these variables. The inverse association between physical activity and weight gain observed in both men and women is consistent with what several other large prospective studies have shown.15, 22, 23, 24, 25, 26, 27, 28 These data support encouraging physical activity to help control the rising prevalence of obesity. However, in the current study, even in men and women who were classified as being very physically active at both baseline and follow-up, a mean weight gain of around 300 g/year was seen. The effect on weight gain of becoming more active during the follow-up period was similar to the effect of changing to a diet including fewer animal foods, suggesting that obesity may be best prevented if both diet and physical activity are considered for intervention.
Compared with non-smokers, the mean annual weight gain was lower in smokers and considerably higher in men and women who stopped smoking during follow-up. The unfavourable effect of smoking cessation on body weight is well known and is probably caused by a combination of increased energy intake, decreased resting metabolic rate, decreased physical activity and increased lipoprotein lipase activity.29 However, although the risk of gaining weight is of concern for people who want to stop smoking, an increase in body weight due to smoking cessation is of minor concern compared with the hazard of continuing smoking.
Our finding that people who separated or became widowed gain less weight and that people who entered a marriage gain more weight compared with people who stayed married during follow-up agrees with findings from other studies.30, 31, 32, 33, 34, 35 This may be explained by changes in health behaviour; for example, marriage seems to promote an increased consumption of vegetables, whereas marriage termination has been associated with worsening smoking, drinking and dietary habits, possible more so in men than in women.31, 35 Although entering a marriage is associated with weight gain, married people are generally healthier and have lower mortality compared with unmarried people.36, 37, 38
The association between job status and weight gain indicates that this variable affects diet and physical activity in ways beyond those represented by the diet group and physical activity variables. In our study, the influence of change in job status differed with gender: men who left employment and women who gained employment during follow-up had lower weight gain compared with those who had a paid job at both the beginning and end of follow-up. Leaving school after age of 16 was associated with decreased weight gain in both men and women, as also seen in other studies.39, 40 Education, income and occupation are used as indicators of socioeconomic status. A recent review of longitudinal studies showed that among non-blacks in developed countries there was a relatively consistent inverse association between weight gain and occupation, and, to a somewhat lesser degree, education, whereas the association between weight gain and income was inconsistent.41
Several studies have reported that a younger age at menarche is associated with increased risk of being overweight as an adult.42, 43, 44, 45 This association was shown to be partly explained by childhood obesity.46 Our study suggests that age at menarche is also associated with adult weight gain, as reported in other studies.39, 40
The EPIC-Oxford cohort differs from other studies in that it involves subjects who are more health conscious than the general UK population. Using data from the two recent British national diet and nutrition surveys,47, 48 we estimated the mean annual weight gain among persons aged 35–49 in the earlier survey to be approximately 700 g in men and 450 g in women. In our study, the annual weight gain among subjects of the same age was approximately 460 and 470 g in men and women, respectively. However, the estimates were derived from surveys conducted in 1986 and 2001, so the time period differed from that in our study. The current study showed significant increases in body weight during follow-up in all subgroups analysed, except for women aged >64 years. Even among the vegans body weight increased by approximately 300 g/year on average. These findings demonstrate that gaining weight is a general trend and support the view that the obesity epidemic seen over the past decades1, 49 is caused by an increasingly ‘obesogenic’ environment in which the easiest choices are those leading to high energy intake and reduced physical activity.50
Swinburn BA, Caterson I, Seidell JC, James WPT . Diet, nutrition and the prevention of excess weight gain and obesity. Public Health Nutr 2004; 7: 123–146.
Anderson JW, Konz EC, Frederich RC, Wood CL . Long-term weight-loss maintenance: a meta-analysis of US studies. Am J Clin Nutr 2001; 74: 579–584.
Appleby PN, Thorogood M, Mann J, Key TJ . Low body mass index in non-meat eaters: the possible roles of animal fat, dietary fibre and alcohol. Int J Obes Relat Metab Disord 1998; 22: 454–460.
Fraser GE . Associations between diet and cancer, ischemic heart disease, and all-cause mortality in non-Hispanic white California Seventh-day Adventists. Am J Clin Nutr 1999; 70 (Suppl): 532S–538S.
Spencer EA, Appleby PN, Davey GK, Key TJ . Diet and body mass index in 38 000 EPIC-Oxford meat-eaters, fish-eaters, vegetarians and vegans. Int J Obes Relat Metab Disord 2003; 27: 728–734.
Key TJ, Fraser GE, Thorogood M, Appleby PN, Beral V, Reeves G et al. Mortality in vegetarians and nonvegetarians: detailed findings from a collaborative analysis of 5 prospective studies. Am J Clin Nutr 1999; 70 (3 Suppl): 516S–524S.
Newby PK, Katherine LT, Wolk A . Risk of overweight and obesity among semivegetarian, lactovegetarian, and vegan women. Am J Clin Nutr 2005; 81: 1267–1274.
Key T, Davey G . Prevalence of obesity is low in people who do not eat meat. Br Med J 1996; 313: 816–817.
Fraser GE . Diet, Life Expectancy, and Chronic Disease. Studies of Seventh-day Adventists and Other Vegetarians. Oxford University Press: New York, 2003. pp 132–134.
Davey GK, Spencer EA, Appleby PN, Allen NE, Knox KH, Key TJ . EPIC-Oxford: lifestyle characteristics and nutrient intakes in a cohort of 33 883 meat-eaters and 31 546 non meat-eaters in the UK. Public Health Nutr 2003; 6: 259–269.
Spencer EA, Appleby PN, Davey GD, Key TJ . Validity of self-reported height and weight in 4808 EPIC-Oxford participants. Public Health Nutr 2001; 5: 561–565.
Wareham NJ, Jakes RW, Rennie KL, Shuit J, Mitchell J, Hennings S et al. Validity and repeatability of a simple index derived from the short physical activity questionnaire used in the European Prospective Investigation into Cancer and Nutrition (EPIC) study. Public Health Nutr 2003; 6: 407–413.
Cade JE, Burley VJ, Greenwood DC, and the UK Women's Cohort Study Steering Group. The UK Women's Cohort Study: comparison of vegetarians, fish-eaters and meat-eaters. Public Health Nutr 2004; 7: 871–878.
Haddad EH, Tanzman JS . What do vegetarians in the United States eat? Am J Clin Nutr 2003; 78 (Suppl): 626S–632S.
Koh-Banerjee P, Chu N-F, Spiegelman D, Rosner B, Colditz G, Willett W et al. Prospective study of the association of changes in dietary intake, physical activity, alcohol consumption, and smoking with 9-y gain in waist circumference among 16 587 US men. Am J Clin Nutr 2003; 78: 719–727.
Liu S, Willett WC, Manson JE, Hu FB, Rosner B, Colditz G . Relation between changes in intakes of dietary fiber and grain products and changes in weight and development of obesity among middle-aged women. Am J Clin Nutr 2003; 78: 920–927.
Ludwig DS, Pereira MA, Kroenke CH, Hilner JE, Slattery ML, Jacobs Jr DR . Dietary fiber, weight gain, and cardiovascular disease risk factors in young adults. JAMA 1999; 282: 1539–1546.
WHO. Diet nutrition and the prevention of chronic diseases: report of a joint WHO/FAO expert consultation. WHO Technical Report Series 916. Geneva, Switzerland, 2003.
Katz DL . Pandemic obesity and the contagion of nutritional nonsense. Public Health Rev 2003; 31: 33–44.
Bravata DM, Sanders L, Huang J, Krumholz HM, Olkin I, Gardner CD . Efficacy and safety of low-carbohydrate diets. A systematic review. JAMA 2003; 289: 1837–1850.
Astrup A, Larsen TM, Harper A . Atkins and other low-carbohydrate diets: hoax or an effective tool for weight loss? Lancet 2004; 364: 897–899.
Sternfeld B, Wang H, Quesenberry CP, Abrams B, Everson-Rose SA, Greendale GA et al. Physical activity and changes in weight and waist circumference in midlife women: findings from the Study of Women's Health Across the Nation. Am J Epidemiol 2004; 160: 912–922.
Schmitz KH, Jacobs Jr DR, Leon AS, Schreiner PJ, Sternfeld B . Physical activity and body weight: associations over ten years in the CARDIA study. Coronary Artery Risk Development in Young Adults. Int J Obes Relat Metab Disord 2000; 24: 1475–1487.
Williamson DF, Madans J, Anda RF, Kleinman JC, Kahn HS, Byers T . Recreational physical activity and ten-year weight change in a US national cohort. Int J Obes Relat Metab Disord 1993; 17: 279–286.
Owens JF, Matthews KA, Wing RR, Kuller LH . Can physical activity mitigate the effects of aging in middle-aged women? Circulation 1992; 85: 1265–1270.
Sherwood NE, Jeffery RW, French SA, Hannan PJ, Murray DM . Predictors of weight gain in the Pound of Prevention study. Int J Obes Relat Metab Disord 2000; 24: 395–403.
Taylor CB, Jatulis DE, Winkleby MA, Rockhill BJ, Kraemer HC . Effects of life-style on body mass index change. Epidemiology 1994; 5: 599–603.
Klesges RC, Klesges LM, Haddock CK, Eck LH . A longitudinal analysis of the impact of dietary intake and physical activity on weight change in adults. Am J Clin Nutr 1992; 55: 818–822.
Filozof C, Fernandes Pinilla MC, Fernandes-Cruz A . Smoking cessation and weight gain. Obes Rev 2004; 5: 95–103.
Jeffery RW, Rick AM . Cross-sectional and longitudinal associations between body mass index and marriage-related factors. Obes Res 2002; 10: 809–815.
Eng PM, Kawachi I, Fitzmaurice G, Rimm EB . Effects of marital status on changes in dietary and other health behaviours in US male health professionals. J Epidemiol Commun Health 2005; 59: 56–62.
Sobal J, Rauschenbach B, Frongillo EA . Marital status changes and body weight changes: a US longitudinal analysis. Soc Sci Med 2003; 56: 1543–1555.
Kahn HS, Williamson DF . The contributions of income, education and changing marital status to weight change among US men. Int J Obes Relat Metab Disord 1990; 14: 1057–1068.
Umberson D . Gender, marital status and the social control of health behaviour. Soc Sci Med 1992; 34: 907–917.
Lee S, Cho E, Grodstein F, Kawachi I, Hu FB, Colditz GA . Effects of marital transitions on changes in dietary and other health behaviours in US women. Int J Epidemiol 2005; 34: 69–78.
Lund R, Holstein BE, Osler M . Marital history from age 15–40 years and subsequent 10-year mortality: a longitudinal study of Danish males born in 1953. Int J Epidemiol 2004; 33: 389–397.
Ebrahim S, Wannamethee G, McCallum A, Walker M, Shaper AG . Marital status, change in marital status, and mortality in middle-aged British men. Am J Epidemiol 1995; 142: 834–842.
Johnson NJ, Backlund E, Sorlie PD, Loveless CA . Marital status and mortality: the national longitudinal mortality study. Ann Epidemiol 2000; 10: 224–238.
Lahmann PH, Lissner L, Gullberg B, Berglund G . Sociodemographic factors associated with long-term weight gain, current body fatness and central adiposity in Swedish women. Int J Obes Relat Metab Disord 2000; 24: 685–694.
Wen W, Gao YT, Shu XO, Yang G, Li HL, Jin F et al. Sociodemographic, behavioural, and reproductive factors associated with weight gain in Chinese women. Int J Obes Relat Metab Disord 2003; 27: 933–940.
Ball K, Crawford D . Socioeconomic status and weight change in adults: a review. Soc Sci Med 2005; 60: 1987–2010.
Laitinen J, Power C, Jarvelin M-R . Family social class, maternal body mass index, childhood body mass index, and age at menarche as predictors of adult obesity. Am J Clin Nutr 2001; 74: 287–294.
Okasha M, McCarron P, McEven J, Smith GD . Age at menarche: secular trends and association with adult anthropometric measures. Ann Hum Biol 2001; 28: 68–78.
vanLenthe FJ, Kemper CG, Mechelen W . Rapid maturation in adolescence results in greater obesity in adulthood: the Amsterdam Growth and Health Study. Am J Clin Nutr 1996; 64: 18–24.
Adair LS, Gordon-Larsen P . Maturational timing and overweight prevalence in US adolescent girls. Am J Public Health 2001; 91: 642–644.
Freedman DS, Khan LK, Serdula MK, Dietz WH, Srinivasan SR, Berenson GS . The relation of menarcheal age to obesity in childhood and adulthood: the Bogalusa heart study. BMC Pediatr 2003, http://www.biomedcentral.com/1471-2431/3/3 (accessed Jan 2005).
Social Survey Division of OPCS with Dietary and Nutritional Evaluations by the Ministry of Agriculture, Fisheries and Food and the Department of Health. The Dietary and Nutritional Survey of British Adults. HMSO: London, UK, 1990.
Food Standards Agency and the Departments of Health by the Office for National Statistics and Medical Research Council Human Nutrition Research. The National Diet and Nutrition Survey: adults aged 19 to 64 years. HMSO: Norwich, UK, 2004.
Hill JO, Wyatt HR, Reed GW, Peters JC . Obesity and the environment: where do we go from here? Science 2003; 299: 853–855.
Swinburn B, Egger G . The runaway weight train: too many accelerators, not enough brakes. Br Med J 2004; 329: 736–739.
We thank all participants in this study and the EPIC-Oxford study staff at the Cancer Research UK Epidemiology Unit. EPIC-Oxford is supported by Cancer Research UK and the Medical Research Council. MR was supported by a grant from the Swedish Council for Working Life and Social Research.
About this article
The American Journal of Clinical Nutrition (2019)
Proceedings of the Nutrition Society (2019)
Clinical Nutrition ESPEN (2019)