Previous research has focused on overall associations between work stress and body mass index (BMI) ignoring the possibility that stress may cause some people to eat less and lose weight and others to eat more. Using longitudinal data, we studied whether work stress induced weight loss in lean individuals and weight gain in overweight individuals.
Prospective cohort study.
A total of 7965 British civil servants (5547 men and 2418 women) aged 35–55 at study entry (The Whitehall II study).
Work stress, indicated by the job strain model and measured as job control, job demands and job strain, was assessed at baseline and BMI at baseline and at 5-year follow-up.
In men, the effect of job strain on weight gain and weight loss was dependent on baseline BMI (P⩽0.03). In the leanest quintile (BMI<22 kg/m2) at baseline, high job strain and low job control were associated with weight loss by follow-up, whereas among those in the highest BMI quintile (>27 kg/m2), these stress indicators were associated with subsequent weight gain. No corresponding interaction was seen among women.
Inconsistent findings reported by previous studies of stress and BMI have generally been interpreted to indicate the absence of an association. In light of our results, the possibility of differential effects of work stress should also be taken into account.
Work stress and high body mass index (BMI) are both risk factors for cardiovascular disease,1, 2, 3, 4 but the extent to which they are associated with each other remains unclear. According to a dominant theoretical model, persistent work stress is generated by a combination of high job demands and low control at work: job strain.5 While some studies have found an association between high strain,6, 7 high demands,6, 8, 9 low control,3, 10 and increased BMI, other studies have reported no associations between these psychosocial characteristics of the work environment and BMI,11, 12, 13, 14 and, in some samples, high strain or low control were associated with lower BMI.15, 16 Associations of job strain and its components with BMI have been found to vary by sex, but not in a consistent manner17 and stress has also been proposed to be linked with BMI beyond the work context.18
Mixed findings on work stress and BMI may reflect a failure to take into account the possibility that stress may cause some people to eat more, but others to eat less.19 Dalman et al.20 demonstrated that eating high fat and carbohydrate caloric content ‘comfort’ food may help in reducing biological stress system activity and concomitant negative emotions during chronic stress. The tendency to eat such energy-dense foods under stress is likely to contribute to weight gain.18, 21 Both physical inactivity and lack of time to prepare healthy meals further increase weight gain among people with chronic stress.18, 22 On the other hand, chronic stress may also trigger physiological processes which lead to weight loss,20, 23 in particular, among individuals for whom chronic work stress is associated with suppressed appetite and increased physical activity.18 As sex is a factor that appears to confer differential sensitivity to stress and may also affect choice of coping strategy,24, 25, 26, 27, 28 it is important to examine stress effects on eating and weight change separately for men and women.
In this report from the Whitehall II study, we examined the association between work stress and BMI with particular attention to the hypothesis that chronic work stress may induce weight gain in some individuals but weight loss in others. If such individual differences are stable and accumulate, then the tendency to weight gain under stress is likely to be more common among people with high BMI while among those with low BMI the tendency will be towards further weight loss.
The target population for the Whitehall II study is all London-based office staff working in 20 Civil Service departments during recruitment in 1985–1988.29 With a participation rate of 73%, the baseline cohort consisted of 10 308 civil servants: 6895 men and 3413 women aged 35–55 years. The true participation rate was probably higher, however, because around 4% of those invited had, in fact, moved employment before the study and were not eligible for inclusion. Baseline screening and a follow-up screening in 1991–1993 included the assessment of BMI. This study focused on those 5547 men and 2418 women with data on job strain at baseline and BMI at baseline and follow-up. The 2343 employees excluded had slightly higher BMI (25.0 vs 24.5 kg/m2), were older (45.0 vs 44.3 years), more likely to be women (42 vs 36%) and in the lowest employment grade (36 vs 19%) (all P-values<0.001).
Assessment of work stressors was based on a modified Job Content Questionnaire comprised of the job demand scale (4 items, Cronbach's α=0.67) and job control scale (15 items, α=0.84).4 Job demand and job control scores are standardized mean scores from the respective scales. Job strain is a standardized score derived from the equation: job demand score – job control score,30 a continuous variable of highly predictive validity for coronary heart disease in the Whitehall II study.31 We repeated the analyses with a more commonly used binary variable defining job strain as a job demand score above the median combined with a job control score below the median. Other baseline characteristics were sex, age (years), employment grade (1=administrative, 2=professional, 3=clerical) and BMI, calculated as weight in kilograms divided by height in meters squared. Weight was measured with all items of clothing removed except underwear. A Soehnle scale was used to read weight to the nearest 0.1 kg. If the reading alternated between two readings >0.1 kg apart with the participant standing still, the higher reading was recorded. Height was measured to the nearest mm using a stadiometer with the participant standing completely erect with the head in the Frankfurt plane.32 Assessment of BMI was repeated at follow-up. Weight gain was indicated if BMI increased and weight loss if BMI decreased between the measurements.
The associations between each work stressor (high strain, high demands and low control) and BMI at follow-up were examined with separate linear regression models including age, grade and BMI at baseline as covariates. The associations between each work stressor and weight gain or loss were examined by using logistic regression analysis. Adjustments were made for age, grade and BMI at baseline. We tested the effect of an interaction between each stressor and baseline BMI on weight gain and weight loss by including the corresponding interaction term in regression models that already included the main effects. To illustrate the form of significant interactions, the associations between stressors and weight gain/weight loss were calculated separately for the top and bottom quintiles of baseline BMI in addition to the combined three middle quintiles. A significant interaction supports our hypothesis that the work stressor predicts weight gain in the top quintile (i.e., in overweight subjects), weight loss in the bottom quintile (in lean subjects), and is not associated with weight change in the middle quintiles. In addition, we tested potential interaction between employment grade and baseline BMI on weight gain and weight loss, as employment grade is associated with metabolic syndrome in the Whitehall II study.33 All analyses were conducted separately for men and women. All analyses were performed with the use of SAS software, version 8.2 (SAS Institute).
Table 1 presents sample characteristics by sex. On average, the men were younger than the women and a greater proportion were employed in higher-grade jobs. Men also had higher job demands, higher job control, lower job strain, and lower BMI at baseline and follow-up (P<0.001). In both sexes, BMI increased over the 5-year-period between the two Phases (P<0.0001), 0.6 kg/m2 among men and 1.1 kg/m2 among women.
Table 2 shows the associations between work stressors at baseline and BMI at follow-up after adjustment for age, grade and BMI at baseline. In men, no association between stressors and subsequent BMI was found. In women, higher job demands were weakly associated with higher BMI at follow-up.
Table 3 presents results of the tests of interaction between stressors and baseline BMI on weight gain and weight loss (as only one woman and four men had exactly the same BMI at baseline and at follow-up, the P-values were similar for the tests examining interaction for weight gain and weight loss and are therefore presented only once). In men, the interactions of baseline BMI with job strain, treated as a continuous variable, and job control were statistically significant. No interaction was found for job demands in men or for any work stressors among women. Repeating the analysis with the cruder dichotomous job strain variable also failed to provide significant interaction (P=0.25 in men and P=0.60 in women).
To illustrate the significant interactions, the effects of high strain and low job control as predictors of weight gain and weight loss are presented by baseline BMI level in Table 4. The pattern was the opposite to that of the regression towards mean. Among the 986 men who already had high BMI at baseline (>27 kg/m2, the top quintile), higher job strain and lower job control were associated with greater likelihood of weight gain, a 1 s.d. increase in job strain or decrease in job control being related to an odds ratio of 1.2 (P=0.006 and 0.10, respectively). In contrast, higher strain and lower control were associated with a higher likelihood of weight loss among the 1042 men with low baseline BMI (<22 kg/m2, the bottom quintile). The corresponding odds ratios were 1.1 (P=0.07) and 1.3 (P=0.002), respectively. Among the 3519 men with intermediate BMI at baseline (between 22 and 27 kg/m2), no association between job strain, job control, weight gain and weight loss was found.
There was no interaction between employment grade and baseline BMI on weight gain or weight loss (P=0.96 in men; P=0.09 in women).
This prospective study suggests that work stress, as indicated by job strain and low job control, increases the likelihood of weight gain among men with a higher BMI, but seems to predict weight loss among lean men who have no need for weight reduction. In contrast to these bidirectional effects, there was some evidence of a prospective association between work stress and increased BMI in women. Our investigation is based on a large well-characterized cohort of British employees and a prospective study design with an average follow-up of 5 years.
A recent review concluded that evidence does not support a consistent association between job strain and BMI and this conclusion seems also to apply to studies that were not included in the review.17 Data from employees in New York City public and private sector worksites revealed no association between change in job demands or job control, the components of job strain, and change in overweight.14 Similarly, a Finnish study of industrial employees found no association between job strain, job demands and subsequent BMI, but low job control predicted a modest increase in BMI during a 10-year follow-up.3
Our findings of bidirectional effects for work stress in men raise the possibility that the overall null associations between work stress and BMI observed in previous studies represent the canceling out of these opposing relationships. Existing physiological models, which describe plausible mechanisms indicate that work stress may result in either weight gain or weight loss. On one hand, a stress-induced increase in glucocorticoid levels is assumed to increase ingestion of comfort food,20 but, on the other hand, stress may inhibit appetite through activation of the sympathetic nervous system which suppresses upper gastrointestinal motility and stimulates energy substrate mobilization.18
Animal studies, which offer the opportunity to examine the effect of stress on food intake while controlling for confounding factors, have generally reported hypophagia (reduced food intake) in relation to chronic stress models.23 However, systematic manipulations of stressor intensity suggest that hypophagia is induced at higher intensities, but that at lower intensities hyperphagia (increase in food intake) is seen.21 In human beings, bidirectional effects are widely recognized in relation to depression, another consequence of chronic work stress.8, 34, 35 Indeed, the American Psychiatric Association's Diagnostic and Statistical Manual includes opposite pairs, such as weight gain/weight loss and hyperphagia/hypophagia, among criteria for a diagnosis of depression.
Although the determinants of body weight are not well understood, interactions between genes and environmental factors are likely.36 Stable individual differences may determine which reaction pattern, that is, weight gain, weight loss or neither, would prevail under stressful conditions for each individual. A field study of middle-aged men and women, who kept stress diaries and at the same time recorded their dietary intake, identified three reactions to stress.37 Some subjects consistently ate more during stressful periods than other times, some consistently ate less and there were also subjects with no stress-related change in eating patterns. In line with this, a study of university students found both a self-reported tendency for some students consistently to eat more during examination periods while others consistently ate less during these periods.19 A follow-up revealed greater weight gain among stress eaters. Our study suggests that BMI may be a marker (and a long-term consequence) of such stable individual differences. In addition to eating patterns, stress could contribute to weight change through changes in physical activity level or metabolic rate. However, tendency toward weight gain during work stress seems to characterize employees with a higher BMI whereas a tendency to weight loss during chronic stress was found to be dominant among employees with a lower BMI.
It is unclear why the bidirectional effect of work stress was not seen among the women. There are apparent sex differences in physiological responses to experimental and real-life stressors24, 25, 26, 27, 28 and the links between secretion of cortisol (a stress hormone) and central adiposity varied between men and women in a Whitehall II subsample study.26 These sex differences are potential contributors to our findings, though small scale studies do not report consistent differences in stress eating patterns between men and women.19, 37 The assessment of stress in the occupational setting is another potential explanation of the null finding. For women the stress of unpaid work at home (e.g. child care) may be more important because women still carry a larger share of the responsibility for doing domestic duties.38, 39 A Swedish study showed that a combination of stressful conditions at work and at home predicted perceived symptoms in white-collar women whereas for men symptoms were more strongly determined by work stress alone.40 It is therefore possible that by also including stress beyond work life an association between stress and BMI among women would be revealed.
Study limitations and implications for further research
When interpreting these findings, at least five issues warrant consideration. First, 20% of the baseline cohort was lost to follow-up, the loss being slightly greater among women, clerical workers, older participants and those with higher BMI. This may not be a major source of bias because all analyses were carried out separately for men and women and all estimates were adjusted for age and employment grade. However, as our target population was London-based civil servants, further research is needed to evaluate generalizability of the present findings to other working populations.
Second, the interaction effect among men was seen with continuous measures of job control and job strain, but not with a commonly used binary job strain variable. Substantially larger sample sizes are needed to detect interactions than main effects and we believe that the null finding for a binary job strain variable was attributable to a cruder operationalization with information loss due to dichotomization. However, replications with larger cohorts are needed to confirm this.
Third, we illustrated the interaction between work stress and initial BMI with subgroups defined by BMI categories with approximately 20% of men in the lowest and highest categories (cut-points >27, 22–27 and <22 kg/m2). Among these two extreme groups, we found a higher proportion of men who, respectively, eat less or more in response to stress, a pattern opposite to that of the regression towards mean. For the 60% of men in the middle BMI category, we expected the bidirectional effects of stress on weight change to cancel out, as in the whole cohort. The contrasting effects of stress are not solely attributable to the specific BMI categories used in this study, since the stress-BMI interaction was found for continuous BMI. In a small-scale questionnaire survey the proportions who reported eating either more or less in response to stress were between 25 and 35%.19 Further research is needed to determine whether use of self-reported stress eating patterns would provide more contrasting effects on weight change than the initial BMI categories used in our study.
The fourth issue requiring consideration is that longstanding illness is a common source of weight loss and thus a potential confounding factor for the stress effects seen among men with low BMI. In our results for lean men, such confounding is unlikely, because additional adjustment for longstanding illness attenuated the association between low control and weight loss by <2%.
Lastly, further research is needed to link the epidemiological evidence with the neuroendocrine, metabolic and behavioral mechanisms that may underlie the effects of work stress (or its absence) on BMI among men and women. As disturbances of hypothalamic–pituitary–adrenocortical axis function are assumed to contribute in particular to central obesity,26, 41, 42 determining stress effects on change in waist circumference or waist/hip ratio, in addition to BMI, would be important.
Prospective evidence suggests a bidirectional effect of work stress on BMI among men. While stress seems to result in weight gain among overweight and obese men employees, weight loss is more likely among their lean colleagues. Inconsistent findings reported by previous studies of stress and BMI have generally been interpreted to indicate the absence of an association. In the light of the present findings, the possibility of differential effects of work stress should also be taken into account.
Kopelman PG . Obesity as a medical problem. Nature 2000; 404: 635–643.
Hemingway H, Marmot M . Evidence based cardiology: psychosocial factors in the aetiology and prognosis of coronary heart disease. Systematic review of prospective cohort studies. BMJ 1999; 318: 1460–1467.
Kivimäki M, Leino-Arjas P, Luukkonen R, Riihimäki H, Vahtera J, Kirjonen J . Work stress and risk of cardiovascular mortality: prospective cohort study of industrial employees. BMJ 2002; 325: 857–861.
Kuper H, Marmot M . Job strain, job demands, decision latitude, and risk of coronary heart disease within the Whitehall II study. J Epidemiol Community Health 2003; 57: 147–153.
Karasek R, Theorell T . Healthy Work: Stress, Productivity, and the Reconstruction of Working Life. Basic Books: New York, 1990.
Hellerstedt WL, Jeffery RW . The association of job strain and health behaviours in men and women. Int J Epidemiol 1997; 26: 575–583.
Wamala SP, Wolk A, Orth-Gomer K . Determinants of obesity in relation to socioeconomic status among middle-aged Swedish women. Prev Med 1997; 26: 734–744.
Niedhammer I, Goldberg M, Leclerc A, David S, Bugel I, Landre MF . Psychosocial work environment and cardiovascular risk factors in an occupational cohort in France. J Epidemiol Community Health 1998; 52: 93–100.
Lallukka T, Laaksonen M, Martikainen P, Sarlio-Lähteenkorva S, Lahelma E . Psychosocial working conditions and weight gain among employees. Int J Obes Relat Metab Disord 2005; 29: 909–915.
Steptoe A, Cropley M, Griffith J, Joekes K . The influence of abdominal obesity and chronic work stress on ambulatory blood pressure in men and women. Int J Obes Relat Metab Disord 1999; 23: 1184–1191.
Brisson C, Larocque B, Moisan J, Vezina M, Dagenais GR . Psychosocial factors at work, smoking, sedentary behavior, and body mass index: a prevalence study among 6995 white collar workers. J Occup Environ Med 2000; 42: 40–46.
Reed DM, LaCroix AZ, Karasek RA, Miller D, MacLean CA . Occupational strain and the incidence of coronary heart disease. Am J Epidemiol 1989; 129: 495–502.
Jonsson D, Rosengren A, Dotevall A, Lappas G, Wilhelmsen L . Job control, job demands and social support at work in relation to cardiovascular risk factors in MONICA 1995, Goteborg. J Cardiovasc Risk 1999; 6: 379–385.
Landsbergis PA, Schnall PL, Deitz DK, Warren K, Pickering TG, Schwartz JE . Job strain and health behaviors: results of a prospective study. Am J Health Promot 1998; 12: 237–245.
Theorell T, Ahlberg-Hulten G, Jodko M, Sigala F, de la Torre B . Influence of job strain and emotion on blood pressure in female hospital personnel during workhours. Scand J Work Environ Health 1993; 19: 313–318.
Amick III BC, Kawachi I, Coakley EH, Lerner D, Levine S, Colditz GA . Relationship of job strain and iso-strain to health status in a cohort of women in the United States. Scand J Work Environ Health 1998; 24: 54–61.
Overgaard D, Gyntelberg F, Heitmann BL . Psychological workload and body weight: is there an association? A review of the literature. Occup Med (Lond) 2004; 54: 35–41.
Wardle J, Gibson EL . Impact of stress on diet: processes and implications. In: Stansfeld S, Marmot M (eds). Stress and the Heart: Psychosocial Pathways to Coronary Heart Disease. BMJ Books: London, 2002. pp 124–149.
Epel E, Jimenez S, Brownell K, Stroud L, Stoney C, Niaura R . Are stress eaters at risk for the metabolic syndrome? Ann NY Acad Sci 2004; 1032: 208–210.
Dallman MF, Pecoraro N, Akana SF, La Fleur SE, Gomez F, Houshyar H et al. Chronic stress and obesity: a new view of ‘comfort food’. Proc Natl Acad Sci USA 2003; 100: 11696–11701.
Robbins TW, Fray PJ . Stress-induced eating: fact, fiction or misunderstanding? Appetite 1980; 1: 103–133.
Kouvonen A, Kivimäki M, Elovainio M, Virtanen M, Linna A, Vahtera J . Job strain and leisure-time physical activity in female and male public sector employees. Prev Med 2005; 41: 532–539.
Willner P . Validity, reliability and utility of the chronic mild stress model of depression: a 10-year review and evaluation. Psychopharmacology (Berlin) 1997; 134: 319–329.
Katz JR, Taylor NF, Perry L, Yudkin JS, Coppack SW . Increased response of cortisol and ACTH to corticotrophin releasing hormone in centrally obese men, but not in post-menopausal women. Int J Obes Relat Metab Disord 2000; 24 (Suppl 2): S138–S139.
Chandola T, Kuper H, Singh-Manoux A, Bartley M, Marmot M . The effect of control at home on CHD events in the Whitehall II study: gender differences in psychosocial domestic pathways to social inequalities in CHD. Soc Sci Med 2004; 58: 1501–1509.
Steptoe A, Kunz-Ebrecht SR, Brydon L, Wardle J . Central adiposity and cortisol responses to waking in middle-aged men and women. Int J Obes Relat Metab Disord 2004; 28: 1168–1173.
Lundberg U . Stress hormones in health and illness: the role of work and gender. Psychoneuroendocrinol 2005; 30: 1017–1021.
Faraday MM, Blakeman KH, Grunberg NE . Strain and sex effects of stress and nicotine on feeding, body weight, and HPA axis hormones. Pharmacol Biochem Behav 2005; 80: 577–589.
Marmot MG, Smith GD, Stansfeld S, Patel C, North F, Head J et al. Health inequalities among British civil servants: the Whitehall II study. Lancet 1991; 337: 1387–1393.
Landsbergis PA, Schnall PL, Warren K, Pickering TG, Schwartz JE . Association between ambulatory blood pressure and alternative formulations of job strain. Scand J Work Environ Health 1994; 20: 349–363.
Kivimäki M, Ferrie JE, Brunner E, Head J, Shipley MJ, Vahtera J et al. Justice at work and reduced risk of coronary heart disease among employees: The Whitehall II Study. Arch Intern Med 2005; 165: 2245–2251.
Ezekiel A, Toon L . Whitehall II study: Manual for Phase 5 Screening Examination. The Department of Epidemiology and Public Health, University College London: London, 2002.
Brunner EJ, Marmot MG, Nanchahal K, Shipley MJ, Stansfeld SA, Juneja M et al. Social inequality in coronary risk: central obesity and the metabolic syndrome. Evidence from the Whitehall II study. Diabetologia 1997; 40: 1341–1349.
Stansfeld SA, Fuhrer R, Head J, Ferrie J, Shipley M . Work and psychiatric disorder in the Whitehall II study. J Psychosom Res 1997; 43: 73–81.
Stansfeld SA, Fuhrer R, Shipley MJ, Marmot MG . Work characteristics predict psychiatric disorder: prospective results from the Whitehall II Study. Occup Environ Med 1999; 56: 302–307.
Goodrick GK, Poston III WS, Foreyt JP . Methods for voluntary weight loss and control: update 1996. Nutrition 1996; 12: 672–676.
Stone A, Brownell KD . The stress-eating paradox: multiple daily measurements in adult males and females. Psychol Health 1994; 9: 425–436.
Lundberg U, Mardberg B, Frankenhaeuser M . The total workload of male and female white collar workers as related to age, occupational level, and number of children. Scand J Psychol 1994; 35: 315–327.
Väänänen A, Kevin MV, Ala-Mursula L, Pentti J, Kivimäki M, Vahtera J . The double burden of and negative spillover between paid and domestic work: Associations with health among men and women. Women Health 2004; 40: 1–18.
Krantz G, Berntsson L, Lundberg U . Total workload, work stress and perceived symptoms in Swedish male and female white-collar employees. Eur J Public Health 2005; 15: 209–214.
Pasquali R, Vicennati V . Activity of the hypothalamic-pituitary-adrenal axis in different obesity phenotypes. Int J Obes Relat Metab Disord 2000; 24 (Suppl 2): S47–S49.
Björntorp P, Rosmond R . Neuroendocrine abnormalities in visceral obesity. Int J Obes Relat Metab Disord 2000; 24 (Suppl 2): S80–S85.
The work for this paper was supported by the Health and Safety Executive. The Whitehall II study has been supported by grants from the Medical Research Council; British Heart Foundation; Health and Safety Executive; Department of Health; National Heart Lung and Blood Institute (HL36310), US, NIH: National Institute on Aging (AG13196), US, NIH; Agency for Health Care Policy Research (HS06516); and the John D and Catherine T MacArthur Foundation Research Networks on Successful Midlife Development and Socio-economic Status and Health. MK, also working at the University of Helsinki, Finland, and JV were supported by the Academy of Finland (projects 104891 and 105195), the Finnish Environment Fund and the Yrjö Jahnsson Foundation, JEF is supported by the MRC (Grant number G8802774), MJS by a grant from the British Heart Foundation, and MGM by an MRC Research Professorship. We thank all participating Civil Service departments and their welfare, personnel, and establishment officers; the Occupational Health and Safety Agency; the Council of Civil Service Unions; all participating civil servants in the Whitehall II study; all members of the Whitehall II study team.
About this article
Cite this article
Kivimäki, M., Head, J., Ferrie, J. et al. Work stress, weight gain and weight loss: evidence for bidirectional effects of job strain on body mass index in the Whitehall II study. Int J Obes 30, 982–987 (2006). https://doi.org/10.1038/sj.ijo.0803229
- psychosocial factors
- work stress
- body mass index
- weight gain
- weight loss
The relative contributions of behavioral, biological, and psychological risk factors in the association between psychosocial stress and all-cause mortality among middle- and older-aged adults in the USA
International Journal of Public Health (2020)
Psychological Studies (2019)
Occupational stress is associated with major long-term weight gain in a Swedish population-based cohort
International Archives of Occupational and Environmental Health (2019)
Increased intake of energy-dense diet and negative energy balance in a mouse model of chronic psychosocial defeat
European Journal of Nutrition (2018)