Search tips
Search criteria 


Logo of ageingLink to Publisher's site
Age Ageing. 2016 May; 45(3): 389–395.
Published online 2016 March 13. doi:  10.1093/ageing/afw020
PMCID: PMC4846792

The long-arm of adolescent weight status on later life depressive symptoms


Background: given the increase in worldwide obesity among children and adolescents, the long-term consequences of childhood obesity on the risk of adverse health outcomes in later life has garnered increased attention. Much of the work on earlier life weight status and later life health has focused on cardiovascular-related outcomes in mid- to late-adulthood; however, little is known about the later life mental health consequences of adolescent body weight.

Methods: data came from the Wisconsin Longitudinal Study. We estimated gender-stratified logistic regression models to characterise the relationship between adolescent weight status using standardised relative body mass ascertained from high school photograph portraits in 1957 and depressive symptoms at age 65 using the Center for Epidemiologic Studies Depression Scale measured in 2004.

Results: women who were overweight in adolescence were significantly more likely to experience depressive symptoms in later adulthood than their normal weight counterparts (odds ratio [OR] = 1.740) when the full set of controls was included. This relationship was not observed among men. The relationship between women's adolescent weight status and later life depressive symptoms was moderated by childhood socioeconomic status, and adolescent overweight was more predictive of later life depressive symptoms for women who were raised in low- and middle-income families (OR = 2.568 and OR = 2.763) than in high-income families (OR = 1.643).

Conclusion: these findings provide further evidence for the wide range of long-term consequences of adolescent overweight on later life well-being and are notable for the gender differences in the connection between early life circumstances and later life mental health.

Keywords: obesity, mental health, longitudinal studies, lifecourse/childhood circumstances, ageing, older people


Increasingly, studies of later life health have adopted a lifecourse approach to understand the downstream influences of earlier life factors. Prior research suggests that adverse early life circumstances are strongly linked to a broad range of later life health outcomes, including subsequent risk of hypertension, cardiovascular disease, diabetes and mortality [16]. Likewise, there is a strong relationship between early life circumstances and later life social status and general well-being [710].

Given the dramatic increase in worldwide obesity among children and adolescents [11, 12], the long-term consequences of earlier obesity for the risk of poor health in adulthood are particularly salient. Several studies have focused on the link between earlier life obesity and adult cardiovascular disease risk [1315]. Fewer studies have examined the association between earlier life (defined as adolescence in this paper) weight status and non-cardiovascular-related outcomes, but there is evidence to believe that such relationships exist. For instance, overweight in adolescence was observed to confer some disadvantage on educational attainment and occupational standing [16], as well as later life physical function, chronic health conditions and mortality [17, 18].

From this growing body of work, there is evidence that the long-term consequences of adolescent weight status may also be associated with mental health in adulthood. A national Australian school survey found that childhood overweight and obesity increased the risk of young adulthood mood disorder [19]. While this study did not explicitly examine adolescent weight status (average age was 10 years) or later life well-being, the findings provide support for future studies examining the connection between adolescent weight and later life mental health.

Likewise, a study using The Nurses' Health Study II cohort indicated that retrospective reports of childhood and adolescent obesity was a risk factor for any adult depressive symptoms or depression medication usage for women [20]. To our knowledge, no study has investigated the link between earlier life weight status and later life depression for both men and women in the United States. Exploring both genders may be important given that the consequences of overweight and obesity throughout the lifecourse may differ by gender, as previously observed in studies of adolescent weight status and later life physical function [18].

This paper addresses two additional gaps in knowledge: (i) whether the influence of earlier life overweight on mental health extends beyond young and middle adulthood and into older adulthood; and (ii) whether the relationship between adolescent weight status and later life mental health in the United States varies based on socioeconomic status (SES). Childhood SES is an established predictor of later life mental health, with lower status children experiencing more depressive symptoms later in life [21]. Our study aims to investigate the longitudinal association between high school (HS) body weight status and depressive symptoms at age 65 for both men and women, and whether this relationship changes based on adolescent and contemporary SES, health status, and behaviours. To accomplish this, we examined the link between adolescent weight status using a novel photograph-based indicator and later life depressive symptoms using the Wisconsin Longitudinal Study (WLS).



The WLS is a longitudinal study that follows ~10,000 class of 1957 high school graduates in Wisconsin [22]. Follow-up surveys were conducted in 1964, 1975, 1993 and 2004. In 2004, respondents were about 65 years of age, and the outcome measure, depressive symptoms, was obtained from this wave.


Depressive symptoms

The WLS used the modified Center for Epidemiologic Studies Depression Scale (CES-D) to measure depressive symptoms. Respondents answered 20 questions, and each scale item asked about the way the respondent felt or behaved during the past week with four response options: rarely or never (<1 day); some or a little (1–2 days); occasionally or moderate (3–4 days) and most or all (5–7 days). The CES-D score ranged from 0 to 60. Scores of 16 and higher are the standard cut point for mild to moderate depression [23], and we used this standard to create a dichotomous measure of depressive symptoms.

Standardised relative body mass

The independent variable of interest was standardised relative body mass (SRBMI) from HS yearbook photographs of the graduates in 1957 to create adolescent weight status. Facial adiposity ratings based on photographs of university students were found to be highly correlated with BMI in recent studies [24, 25]. Photos were collected in 2000 and were analysed based on an 11-point scale to estimate the SRBMI measure by six coders. The SRBMI scale has demonstrated high inter-rater reliability (α = 0.91), and the values represent standard deviation differences from the mean scale value [17]. We constructed four SRBMI categories to represent HS weight based on previous research using the WLS SRBMI measure [16, 18]. The categories were underweight (SRBMI ≤ −1), normal weight (−1 < SRBMI < 0), risk of overweight (0 < SRBMI < 1) and overweight (SRBMI ≥ 1).


We created income terciles representing low, middle and high income for ease of interpretation for family income in 1957. Head of household education in 1957 was categorised as: <8th grade, Grade 8 to 11, high-school diploma (12 years) and some college (13 or more years). Retrospective child health was dichotomised where excellent/very good health translated to good health and good/fair/poor health translated to poor health [26]. Body mass index (BMI) at age 65 was based on the standard cut points for normal weight (BMI 18.5–24.9), overweight (BMI 25–29.9) and obese (BMI ≥30) [27]. Since all respondents obtained at least a HS diploma, adult educational attainment was categorised into three categories: HS diploma, some college or vocational training, and bachelor's degree or higher. Current household income was categorised into high-, middle- or low-income tercile. Marital status at age 65 was categorised as either currently married or not married. Current self-reported health was dichotomised into two categories: fair/poor health and excellent/very good/good health. Current drinking behaviour was coded according to NIAAA guidelines for heavy drinking versus normal drinking [28]. Smoking was categorised into three categories: never smoked, used to smoke and current smoking. Finally, we created a dichotomous variable for recommended level of physical activity for age 65 based on the amount of time devoted to strenuous or light physical activity in a week [29].

Data analysis

We calculated mean scores of the CES-D scale by each SRBMI category for both women and men. We then estimated gender-stratified logistic regression models to calculate odds ratios (OR) using the dichotomous CES-D measure including covariates representing adolescent and current SES and health circumstances. Gender-stratified results were important given prior evidence of gender differences in the long-term consequences of weight status [18]. Stata SE 13 was used for all analyses [30].


The sample characteristics (presented in Table Table1)1) illustrate that the most common SRBMI weight status category among the HS seniors was normal weight (42.8%) followed by at risk of overweight (36.3%). Similar proportions of the sample were overweight (10.9%) and underweight (10.1%). The head of household education levels in 1957 were noticeably lower than the distributions among parents today [31]. Only ~14% of households had a parent with a college degree, and attending school through Grade 11 was most common (41.9%), followed by obtaining a high-school diploma (25.4%). Adult BMI varied by gender. While about one-third of both men and women were obese at age 65, only 17.8% of men were normal weight, while 32.1% of women were normal weight. All respondents have attained at least a high-school diploma, but more men than women obtained a bachelor's degree or higher (34.8 versus 20.6%). The majority of the sample reported being in current good health, with low rates of heavy drinking, current smoking or lack of physical activity.

Table 1.
Descriptive characteristics

The mean CES-D scores are presented in Table Table2.2. The scores for women who were overweight in HS were significantly higher than those women who were normal weight. This relationship also exists for men, though it was not statistically significant. None of the depressive symptom scores for the other adolescent weight categories were significantly different than the scores for normal weight adolescents.

Table 2.
Mean CES-D score

Table Table33 presents the results from the logistic regression analysis for depressive symptoms at age 65 on SRBMI (adolescent weight status). Although there was no relationship between adolescent weight status and later life depressive symptoms for men, among women, the relationship between adolescent overweight and depressive symptoms in later life was statistically significant in Models 4–6. Overweight women were twice as likely to experience depressive symptoms in later life as their normal weight counterparts (OR = 2.372). This association was statistically significant even when controlling for adult BMI status and current SES and health status (OR = 1.740). Adult obesity at age 65 was also significantly associated with depressive symptoms for women (OR = 1.592), but not for men (OR = 1.305).

Table 3.
Odds ratios of the effect of adolescent weight status on later life depression (CES-D)

It also appeared that adolescent SES moderated the relationship between adolescent overweight and later life depressive symptoms for women in analysis stratified by family income in 1957 ( Supplementary data, Table S1, available in Age and Ageing online). No such relationship was found for men (available upon request). There was no significant association between adolescent overweight and later life depressive symptoms for women who grew up in a high-income household in Model 7 (OR = 1.643), but this association was statistically significant for women from low-income households in Model 1 (OR = 2.568) and those from middle-income households in Model 4 (OR = 2.763). The inclusion of current SES and health behaviours reduced the significance of the association between adolescent overweight and later life depressive symptoms for women who grew up in low- and middle-income homes (Models 3 and 6). Similar patterns were found using education as a measure of SES (available upon request).

In Supplementary data, Appendix S1, available in Age and Ageing online, we interacted adolescent weight status with adult weight status to understand how weight throughout the lifecourse influenced depressive symptoms (see Supplementary data, Appendix S1, available in Age and Ageing online). Once again, we found no association between adolescent weight and depressive symptoms for men (Models 1–3). However, we found several significant differences in the lifecourse pattern of the relationship between weight status and depressive symptoms for women in Models 4–6. In Model 6, where the full set of controls was included, we found that women who were in the heaviest categories at both age periods—overweight in HS and obese at age 65—were about three times as likely to experience depression than women who were normal weight (OR = 3.125). Women at risk of overweight in HS and obese at age 65 were also more likely to experience later life depressive symptoms (OR = 1.630) than women who were normal weight at both time periods, though this association was on the margin of statistical significance.


This study found that overweight adolescent women had poorer mental health in later life compared with their normal body weight counterparts. Our findings extend previous research and suggest that: (i) no association between adolescent weight status and later life depressive symptoms was observed for men; (ii) the link between adolescent weight status and adulthood depressive symptoms extends into later life adulthood for women; (iii) the association between earlier life weight status and adulthood depressive symptoms is most pronounced for women who grew up in a low SES household and (iv) a synergistic link between overweight in adolescence and in later life place women at a greater odds of having later life depressive symptoms.

Our finding of an association between adolescent overweight and later life mental health among women aligns with that of prior work in Australia, the United States and Britain [1920, 32]. In Australia, this relationship was observed among both men and women; however, after adjusting for adulthood weight status (as completed in our analytic models), the association between adolescent weight and later life mood disorder was observed in only women [19]. In the US Nurses' Health Study II, women in the two highest categories of body shape at age 10 had higher prevalence and incidence of adult depression, and these associations remained when considering body shape at age 20 and BMI at age 18. In Britain, obesity was associated with common mental disorders over a 19-year period [32]. Taken together, these previous findings and our current results suggest a robust relationship between earlier life weight (in childhood and adolescence) and later life mental health.

Conversely, some studies have reported an absence of an association between obesity and depression. Most notably, a systematic review of epidemiological studies examining this relationship reported an overall weak support for the link between obesity and the incidence of depressive symptoms [33]. However, the heterogeneity of the studies, the paucity of high-quality cohort studies and the majority of cross-sectional studies examined in the systematic review suggests that this relationship may not accurately depict the true association between obesity and depression, if the association does exist.

The association between adolescent overweight and later life depressive symptoms observed in women may be driven, in part, by the development of self-esteem in early life. Cross-sectional and longitudinal analyses report an inverse association between body weight and self-esteem among children [34–36]. For example, the longitudinal relationship between childhood obesity corresponds with self-esteem in Canadian children, such that children (up to ages 11) who were obese at baseline had almost twice the odds of reporting low self-esteem 4 years later relative to their non-obese counterparts [37]. However, the association between early life weight status and later self-esteem may be context specific. This is evidenced by an absence of an association between lifecourse BMI and adolescent self-esteem in the children from Hong Kong, a developed, non-western environment [38].

The literature on body weight and self-esteem suggests that the consequences of earlier life obesity on subsequent low self-esteem are apparent and may have important downstream consequences for later life health. This is supported by the link between low self-esteem (or negative self-regard) and subsequent mental health concerns, including stress, loneliness, anxiety and increased depressive symptoms [39–41]. The mechanism through which overweight feeds into self-esteem may include social stigma of teasing by peers [42–46], which may be more damaging to girls than boys [47], and in particular, girls from lower SES families [48].

Some caveats about this study are important to mention. First, our use of photographs as an indicator of body mass does not reflect a wholly objective measure; however, the coders' use of a relative body mass (RBM) scale set a standardised algorithm for categorically quantifying body mass among raters [17]. Further support for this novel use of high-school photographs is evidenced by previous studies suggesting that face and neck characteristics provide important information on general adiposity and are associated with body mass [24, 25, 49]. Second, although the CES-D is a well-established screening measure for depressive symptoms among diverse groups of people, including older adults [50], it does not provide a clinical diagnosis of depression. We also do not have an available measure of medication use for depression, which may vary by SES. Third, we are unable to determine the directionality of the association between earlier life obesity and later life depressive symptoms. The mechanisms that feed into earlier life obesity are multi-faceted, with one study from Scotland suggesting that childhood behaviour is associated with the development of later life obesity [51], a link we are not able to investigate with the WLS. Last, we acknowledge that WLS is not reflective of the general US population. Future work would benefit greatly from examining the relationship between adolescent weight status and later life depressive symptoms within more diverse populations, particularly given the increase in adolescent obesity in current cohorts.

One of the strengths of this paper's contribution is the unique opportunity presented by the use of WLS respondents' yearbook photos to analyse the lifecourse impact of adolescent weight status. Since concerns surrounding childhood and adolescent obesity emerged following the 1980s, and established longitudinal studies had not incorporated questions regarding body weight prior to adulthood, using weight status determined from HS photographs is a novel solution. This measure is unique in that it is does not rely on retrospective self-reporting of body weight. Hence the weight measure from photographs was not confounded by potential biases in respondents' recall of current and previous health, as is a limitation in previous research [20].

Ultimately, the depressive symptoms associated with overweight in women, and the particular vulnerability of those from lower SES families suggest an area ripe for future research. Improving our understanding of additional later life health outcomes that may be influenced by early life weight status, including cognitive health and overall life expectancy, will greatly contribute to our ability to address the needs of today's youth. Moreover, continued efforts to target improvements in childhood and adolescent weight status, and consequently improving self-esteem, may mitigate the later life health effects that are influenced by such critical early life health circumstances.

Key points

  • We found no association between adolescent weight status and later life depressive symptoms for men.
  • For women, the link between adolescent weight status and later life depressive symptoms was significant.
  • The association between adolescent weight status and adulthood depressive symptoms is most notable in women who grew up in a low SES household.
  • Being overweight at both adolescence and in later life placed women at a greater odds of having later life depressive symptoms.

Conflicts of interest

None declared.


Partial support for this research came from a Eunice Kennedy Shriver National Institute of Child Health and Human Development research infrastructure grant, R24 HD042828, to the Center for Studies in Demography & Ecology at the University of Washington, the National Institute on Aging T32AG023480, and the Charles A. King Trust Postdoctoral Research Fellowship Program, Bank of America, N.A., Co-Trustee.

Supplementary Material

Supplementary Data:


Only the most important are listed here and are represented by bold type throughout the text. The full list of references is available as Supplementary data, available in Age and Ageing online.

1. Barker DJP. Mothers, Babies, and Health in Later Life. London, UK: Churchill Livingstone, 1998.
2. Blane D, Hart CL, Davey Smith G, Gillis CT, Hole DJ, Hawthorne VM Association of cardiovascular disease risk factors with socioeconomic position during childhood and during adulthood. Br Med J 1996; 313: 1434–8. [PMC free article] [PubMed]
3. Cruick-Shank JK, Mzayek F, Liu L et al. Origins of the black/white difference in blood pressure: Roles of birth weight, postnatal growth, early blood pressure, and adolescent body size. Circulation 2005; 111: 1932–7. [PubMed]
4. Doblhammer G, Vaupel JW Lifespan depends on month of birth. Proc Natl Acad Sci USA 2001; 98: 2934–9. [PubMed]
5. Huxley RRR, Shiel AW, Law CM The role of size of birth and postnatal catch-up growth in determining systolic blood pressure: a systematic review of the literature. J Hypertens 2000; 18: 815–31. [PubMed]
6. Lawlor DA, Smith GD Early life determinants of adult blood pressure. Nephrol Hypertension 2005; 14: 259–64. [PubMed]
7. Alwin DF, Wray LA A life-span developmental perspective on social status and health. Gerontologist 2005; 60: 7–14. [PubMed]
8. Dannefer D. Cumulative advantage/disadvantage and the life course: cross-fertilizing age and social science theory. J Gerontol B Psychol Sci Soc Sci 2003; 58: 327–37. [PubMed]
9. Hayward MD, Gorman BK The long arm of childhood: the influence of early-life social conditions on men's mortality. Demography 2004; 41: 87–107. [PubMed]
10. Montez JK, Hayward MD Cumulative childhood adversity, educational attainment, and active life expectancy among U.S. adults. Demography 2014; 51: 413–35. [PMC free article] [PubMed]
11. Troiano RP, Flegal KM, Kuczmarski RJ, Campbell SM, Johnson CL Overweight prevalence and trends for children and adolescents: the National Health and Nutrition Examination Surveys, 1963 to 1991. Arch Pediatr Adolesc Med 1995; 149: 1085–91. [PubMed]
12. Ogden CL, Carroll MD, Kit BK, Flegal KM Prevalence of obesity and trends in body mass index among US children and adolescents, 1999–2010. JAMA 2012; 307: 483–90. [PubMed]
13. Barker DJP, Winter PD, Osmond C, Margetts B, Simmonds SJ Weight in infancy and death from ischaemic heart disease. Lancet 1989; 2: 577–80. [PubMed]
14. Gunnell DJ, Frankel SJ, Nanchahal K, Peters TJ, Smith GD Childhood obesity and adult cardiovascular mortality: a 57-y follow-up study based on the Boyd Orr cohort. Am J Clin Nutr 1998; 67: 1111–8. [PubMed]
15. Stein CE, Fall CHD, Kumaran K, Osmond C, Cox V, Barker DJP Fetal growth and coronary heart disease in south India. Lancet 1996; 348: 1269–73. [PubMed]
16. Glass CM, Hass SA, Reither EN The skinny on success: body mass, gender and occupational standing across the life course. Soc Forces 2010; 88: 1777–806. [PMC free article] [PubMed]
17. Reither EN, Hauser RM, Swallen KC Predicting adult health and mortality from adolescent facial characteristics in yearbook photographs. Demography 2009; 46: 27–41. [PMC free article] [PubMed]
18. Vasunilashorn S, Martinson ML Weight status in adolescence is associated with later life functional limitations. J Aging Health 2013; 25: 758–77. [PMC free article] [PubMed]
19. Sanderson K, Patton GC, McKercher C, Dwyer T, Venn AJ Overweight and obesity in childhood and risk of mental disorder: a 20-year cohort study. Aust N Z J Psychiatry 2011; 45: 384–92. [PubMed]
20. Sanchez-Villegas A, Field A, O'Reilly E et al. Perceived and actual obesity in childhood and adolescence and risk of adult depression. J Epidemiol Commun Health 2013; 67: 81–6. [PubMed]
21. Luo Y, Waite LJ The impact of childhood and adult SES on physical, mental, and cognitive well-being in later life. J Gerontol Ser B Psychol Sci Soc Sci 2005; 60: S93–101. [PMC free article] [PubMed]
22. Sewell WH, Hauser RM, Springer KW, Hauser TS As we age: a review of the Wisconsin Longitudinal Study, 1957–2001. In: Leicht KT, editor. , eds. Research in Social Stratification and Mobility, Volume 20 Oxford: Elsevier, 2004; 3–114.
23. Radloff LS. CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Measurement 1977; 1: 385–401.
24. Coetzee V, Perrett DI, Stephen ID Facial adiposity: a cue to health? Perception 2009; 38: 1700. [PubMed]
25. Tinlin RM, Watkins CD, Welling LL, DeBruine LM, Al-Dujaili EA, Jones BC. Perceived facial adiposity conveys information about women's health. Br J Psychol 2013; 104: 235–48. [PubMed]
26. Smith JP. Reconstructing childhood health histories. Demography 2009; 46: 387–403. [PMC free article] [PubMed]
27. National Institutes of Health. Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults: The Evidence Report. Bethesda, MD: National Institutes of Health, U.S. Department of Health and Human Services, 1998. [PubMed]
28. National Institute on Alcohol Abuse and Alcoholism. Drinking Levels Defined. (15 October 2015, date last accessed).
29. Centers for Disease Control and Prevention. How Much Physical Activity Do Older Adults Need: Guidelines. (15 October 2015, date last accessed ).
30. StataCorp. Stata: Release 13. Statistical Software. College Station, TX: StataCorp LP, 2013.

Articles from Age and Ageing are provided here courtesy of Oxford University Press