Building on prior research, this paper addressed the question of whether adolescents with a history of depressive disorders remain at risk for negative health outcomes in young adulthood even when current depression is taken into account. The analyses suggested that both objectively- and self-rated general health during the transition to adulthood may be compromised among individuals who have experienced depressive disorders, and that these detriments to health are associated with higher medical costs and work impairment. However, depression by age 15 did not predict physical functioning or reported chronic disease status.
Early depressive diagnosis was associated with lower ratings of health by objective interviewers, as well as the youths themselves, even after the effects of current depressive diagnoses were controlled. These results suggest that the health consequences associated with a history of early-onset depression are not simply the result of concurrent depression or negative reporting biases in individuals inclined towards depression. Were age 20 depression not controlled, the effects of early depression would have been stronger across all measured variables, and associations with chronic disease status would have been significant. Given the bidirectionality of the relationship between depression and health, it is likely that these analyses represent a conservative test of the effects of early depression. The fact that current depression was a strong contributor to most of the measured health outcomes highlights the importance of controlling for concurrent diagnoses when testing the long-term health outcomes of adolescents with psychiatric morbidity.
In contrast, our findings did not support an association between adolescent depression and self-reported limitations to physical functioning or chronic illness. Although these findings may initially seem inconsistent with the conclusions on general health, they are not entirely at odds with the idea that depression has negative health consequences. While adolescents with a history of depression have poorer overall health, possibly reflecting higher rates of short-lived illnesses, they did not report higher rates of chronic illnesses that would likely impede their daily physical functioning. The lack of association with chronic disease at age 20 may be accounted for by the relatively early age at which these outcomes were measured. Many of the chronic illnesses on which depression has been shown to have the greatest impact, such as heart disease, are unlikely to be fully developed by early adulthood. However, the precursor conditions to some of these diseases, such as arthrosclerosis, are believed to begin during adolescence [29
], and it is possible that early depression has an effect on chronic illness and associated impairment in physical functioning that will be apparent when later adulthood health is measured. Longitudinal studies following adolescents into later adulthood are needed to fully address these issues.
Perhaps most importantly, the data provided support for the hypothesis that adolescent depression is associated with medical and social costs in young adulthood. Even after the effects of concurrent depressive diagnoses were considered, young adults who had experienced depression in early adolescence reported more visits to medical professionals and greater impairment in work functioning due to their physical health. These results are similar to those obtained among adult populations [2
] and suggest that screening and treatment for depression may be a cost-effective method of limiting medical expenses [2
]. However, more explicit tests of this hypothesis are needed.
These findings echo those established among adult populations, which have shown that depression is a risk factor for onset and recurrence of a variety of health problems [4
], as well as for mortality and impairment among ill populations [31
]. Katon [7
] proposed a theoretical model for the association between depression and health, suggesting that early adversity, biobehavioral factors, genetics, life stress, poor self-care, and the burden of chronic illness all contribute to a bidirectional loop between depression and disease. This model recognizes that factors predisposing individuals towards depression, such as genetics, childhood adversity, and life stress, may also predispose individuals towards other pathways to poor health outcomes, including negative health behaviors, difficulty collaborating with physicians, and maladaptive physical environments. In addition, medical illness may contribute to the onset or maintenance of depression through direct physiological mechanisms or illness-related decreases in quality of life. Similarly, depression may lead to the onset or maintenance of physical illness through pathophysiological or biobehavioral pathways. Although a full test of this model has not yet been completed, each component has received some empirical support among adults. It is critical that future research tests the applicability of this model to adolescents so that appropriate interventions can be developed.
Research examining potential mechanisms of the association between depression and health among younger populations has been more limited. There is evidence to suggest that adverse health behaviors could be one mechanism through which adolescent depression influences adulthood health. Correlational and longitudinal studies have shown that depression is associated with higher rates of smoking, alcohol abuse, unhealthy eating, and infrequent exercise [12
]. Some research has also suggested that direct physiological mechanisms may be involved. A 6-year longitudinal study of younger adolescents found that depressed mood during middle school predicted obesity in high school, even after controlling for behavioral and environmental mechanisms (i.e. obesity at baseline, parental obesity, socioeconomic status, and reported physical activity). Further research is needed to examine the relative contributions of environmental, behavioral, and physiological mechanisms that may explain the connection between depression in early adolescence and negative health consequences in adulthood.
Although the present study indicated a prospective relationship between prior depressive disorders and later health status, the direction of causality is doubtless bidirectional. Just as there is evidence that depression leads to adverse health outcomes, there is also evidence that adolescents with medical problems are at an increased risk of developing depression [15
]. Unfortunately, the nature of this dataset did not allow for a full test of the effects of health problems on the development of future depressive episodes. This reverse direction of association is important to consider, particularly given the limitations to our measure of health conditions prior to our measurement of depression. The current findings should be interpreted with caution until replicated by studies using more stringent controls of pre-existing health conditions. It should also be noted that causality in either direction cannot be definitively established, given that this area of research does not lend itself to randomized trials.
Several limitations to the present analyses are noted. First, the majority of the data was obtained through self-report. Although efforts were made to include observer-rated data, it was not possible to corroborate health status through examination of medical records. However, the inclusion of a control for depression at the time of follow-up limits the concern that our results were influenced by systematic depressive biases in self-reporting. Given the strong association between age 20 depression and several of the health outcomes of interest, it is likely that inclusion of a statistical control for depression at age 20 has yielded conservative estimates of the effects of early depression. In addition, measurements of youth health status at age 15 were not available, and the control measure of prior health was taken from childhood. It is possible that this early childhood measurement did not adequately capture the illness status of the youths prior to the initial depressive episode. Furthermore, it was not possible to verify the reported childhood illnesses by examining medical records. Additionally, the amount of variance accounted for in some of the models was small, highlighting the fact that depression is only one of many important risk factors for negative health outcomes. Finally, the sample was oversampled for maternal depression, so estimates of depression and health problems in our sample should not be taken as general population estimates.
The present study makes a contribution to the study of health and depression by using a prospective, longitudinal design to test the consequences that clinically significant depression in early adolescence has for health in young adulthood. The inclusion of a control for concurrent depression, as well as multiple measurements of health outcomes, are additional strengths of the present study. Overall, our findings suggest that early adolescent depression may be related to poorer health, although the exact mechanism remains to be clarified. These results highlight the importance of building more complex pathophysiological models of depression and illness among adolescent populations. The finding that adolescent depression is associated with higher medical costs and work impairment in young adulthood also suggests that screening and treatment for depression during adolescence may be a cost-effective method of preventing poor health outcomes and limiting the costs associated with depression.