Nearly half of this community-based sample (40% - 54%) met criteria for at least one of these costly and impairing health-related conditions from age 21 to age 30, and 8% to 12% reported comorbid problems at some point during this period. Unique patterns of comorbid problems emerged for women and men, although many more significant associations within and across time were shown for women. For women, multiple relationships emerged between these three disorders, though the associations found were not consistent from age 24 to 27 and age 27 to 30, suggesting that the timing of these conditions may matter. For men, by and large, we found that these disorders are not co-related over time.
This study suggests that depression, obesity, and alcohol use disorders are more related for women than for men. These results are consistent with other psychological theories and research. For example, Susan Nolen-Hoeksema has theorized that women’s vulnerability to “the toxic triangle” of disordered eating, problem drinking, and depression is caused by their greater tendency, compared with men, to respond to stress with rumination, a form of self-focused coping [41
]. While we did not measure rumination in the current study, some research supports her theory. For example, she found that both men and women who scored high on rumination were more depressed and more likely to binge eat or turn to alcohol when they were upset; and, overall, women were more prone to rumination [56
]. Additionally, adolescent girls who scored high in rumination were more likely to develop more symptoms of depression, bulimia, and substance abuse over a three-year period compared to adolescent girls who scored low in rumination [57
]. In addition, these disorders may share predisposing factors such as a common biological pathway, dysfunctional family, or poor self-control [58
Throughout the measured period of young adulthood, the prevalence of comorbidity between conditions increased only for women. Consistent with prior research [12
], depression and obesity were associated concurrently for women at two of the three time points. In contrast, among men in the study, depression and obesity were not concurrently associated. Depression and alcohol use disorders showed cross-sectional associations for both women and men, although more so at later ages (ages 27 and 30 for women; age 30 for men), with generally stronger associations for women. This age trend is consistent with studies showing increases in drinking as a means of coping with stress from the mid-20s to the early 30s [59
]. Alcohol use disorders were not associated with obesity concurrently for women or men.
The longitudinal results also showed gender-specific patterns, though only one of the pathways showed a difference in direction, namely the relation between obesity and later depression. We found evidence of a positive prospective relation among females and a negative relation among males, consistent with other research that has identified gender as a moderator of the association between obesity and depression [9
]. The ‘jolly fat’ hypothesis suggests that being overweight and the dietary habits associated with obesity may protect older adults against the experience of depression [60
]. This hypothesis has received some support in other studies, but only among men [15
]. In our study, we found that even as early as young adulthood there may be some protective effect of obesity for males in the development of depression. Our results are also consistent with the negative association between young adult depression and BMI in males observed by Pine and colleagues [46
The presence of an alcohol use disorder at age 24 increased the risk of developing obesity for women at age 27. Further research is needed to understand whether this association is directly related to the ingestion of excessive alcohol, or is the result of other processes. One potential explanation is that the caloric energy of alcohol contributes to subsequent weight gain. A number of controlled studies have also found a short-term stimulation of appetite following alcohol ingestion [33
], which, when coupled with calories ingested as alcohol, may lead to obesity. Another possibility is that hypersensitivity to reward may be a common vulnerability to both binge eating and chronic alcohol abuse [37
The presence of a major depressive episode at age 27 increased the risk of developing an alcohol use disorder at age 30 among women only. Khantzian’s “self-medication hypothesis” theorizes that affective disturbances may increase the risk for the onset and maintenance of substance use. In particular, Khantzian suggests that “although they are not good antidepressants, alcohol and related drugs create the illusion of relief because they temporarily soften rigid defenses and ameliorate states of isolation and emptiness that predispose to depression” [61, p. 233]. Reviews of the self-medication hypothesis have reported inconsistent findings, and potential gender differences have not been thoroughly addressed by the research [62
]. However, studies using experience-sampling methods, such as examining sadness-to-drinking intervals, to evaluate the self-medication hypothesis have suggested some gender differences in the consequences of self-medication. In a daily study of college students, women who reported more alcohol-related problems and were motivated to drink as a way to cope, drank more following sadness in comparison to men. The author suggests that attempted self-medication with alcohol may have more dire consequences for women because it enhances ruminative coping. Findings of the current study are consistent with others showing that substance use is particularly linked to stress-related processes, such as depression and post-traumatic stress, for women [64
]. The finding that females experiencing depression are at higher risk for developing an alcohol use disorder suggests that there are potential gender differences in attempts to self-medicate with alcohol, and that this may explain women’s drinking more than men’s.
Consistent with literature showing the role of socioeconomic factors as risks for these disorders, our analyses showed income was negatively related to obesity for women (age 24) and was significantly negatively associated with depression for men (age 27). An association between increased BMI and lower social class specific to women has been consistently found in epidemiological studies, with early pregnancy cited as one factor suspected to play a role [46
]. Future research should continue to account for socioeconomic factors, as failure to do so may overestimate disorder co-occurrence that could otherwise be accounted for by shared risk factors such as low income and social status.
Limitations, Strengths, and Future Directions
There are limitations of the current study. The study was conducted with a community sample in Seattle, Washington, with participants from higher crime areas somewhat overrepresented. Thus, caution should be used in generalizing both prevalence rates and associations between conditions to other communities. However, many findings are consistent with those found in the general population, such as the gender-specific relation between depression and obesity [46
], and we know of no specific studies to indicate that utilizing an urban population would undermine our findings. In addition, while the study included three time points, analyses did not allow an examination of comorbidity of these conditions continuously during the study time frame. This may be particularly important for more episodic conditions such as depression. The use of path analysis presents limitations in that the size of the cross-lagged loadings are affected by both the true longitudinal relations and the reliability among measures. Further, obesity was calculated on the basis of self-reported weight and height; however, previous studies have suggested that the sensitivity and specificity of obesity measurement based on self-reported weight and height data is very high, indicating that resulting error is likely to be very low [66
]. Additionally, intentional under- or over-reporting these measures may have been deterred due to the fact that the interviews were conducted in person.
The strengths of this study include the use of a nonclinical community sample, reducing the potential biases that can arise when narrowly defined samples are selected from treatment settings. Moreover, this study is one of the first to examine comorbidity with longitudinal tracking between these three important health conditions over time during young adulthood. We also used statistical analyses that controlled for income and stability over time.
Results indicate that depression, obesity, and alcohol use disorders are interrelated health conditions for women more so than for men, with implications for future research on prevention and treatment relevant to psychiatry and primary care. First, more research exploring the impact of interventions that can lead to the amelioration of more than one of these conditions among women, such as increasing physical activity and learning effective stress management strategies, is warranted. For some women, there may be some common pathogenic processes underlying these conditions, as they can all be triggered by stress and involve behavior in their maintenance. Randomized controlled trials of intervention programs for these conditions that measure broader outcomes than the target problem of interest can provide even more stringent tests of etiological pathways. Second, the stability of each disorder during young adulthood reiterates the importance of conducting prevention earlier in the life course, including childhood and adolescence. There are already a number of preventions for these disorders that target these early developmental stages, but further attention to early prevention may be warranted [38
]. Third, the relevance of income for obesity and depression suggests that targeting low-income populations to prevent these problems may be important. Fourth, addressing alcohol use among young women may have the additional benefit of reducing obesity for some participants. Overall, a greater understanding of the mechanisms underlying these linkages could lead to improved efforts to reduce these common but critical preventable public health problems.