Our intervention did not have an overall effect on reducing the severity of depressive symptoms or increasing the use of coping strategies. Among the 51 participants who entered the study with clinically meaningful depressive symptoms, 22 did not have such symptoms at follow-up. A decrease in symptomatology reduces the likelihood of substance abuse, interpersonal problems, and other behavioral concerns (22
At lower levels of stress at baseline, the number of depressive symptoms decreased as stress increased. At a certain point, however, the number of depressive symptoms increased as stress increased. This finding is consistent with the "challenge" model of resilience (23
), which proposes that the association between risk factors (eg, stressful events) and a health or well-being outcome (eg, depressive symptoms) is curvilinear: exposure to both low and high levels of a risk factor are more strongly associated with negative outcomes, but moderate levels of stress are associated with less negative outcomes. According to the challenge model, participants who had moderate levels of stress learned how to overcome these risks but were not exposed to so much stress that it was impossible to overcome its effects.
The use of active coping decreased for every year increase in age until 21 and then increased among participants aged 22 to 24. This pattern of decreasing active coping among younger members contradicts some research that identified increases in active coping during adolescence (24
). Our study participants may have been exposed to more pervasive and acute stress than other adolescents, or they may have felt more frustrated or hopeless about the efficacy of active coping strategies in helping them to manage stressful life situations. These young people may have turned to other types of coping — eg, support seeking, distraction — if active coping strategies were not perceived as effective, although our data do not show whether our participants actually did this. Research shows that adolescents tend to use less active coping strategies when they perceive stressful situations to be particularly threatening or uncontrollable (25
); our participants faced these kinds of situations.
Participants aged 18 or younger showed increases and older participants reported decreases in support-seeking coping as they got older. People living in stressful urban environments may seek out multiple sources of support as they grow older and are faced with more life stressors — including, in this population, disengagement from school and the workplace. However, as people mature, they may recognize the limitations of the people on whom they rely for social support and less frequently use support-seeking strategies. African American adolescents tend to rely on the same people (eg, extended family networks) to provide emotional, tangible, and informational support (26
Some research suggests that adolescents and young adults use more support-seeking coping in school and peer contexts (27
). The participants in our study did not attend school, so they lacked the school-based peer network that would have encouraged support seeking. The limited peer networks that exist for out-of-school youth validate the importance of peer-based approaches such as our depression-prevention intervention; peer-based approaches may be effective in encouraging older adolescents and young adults to actively seek and use various types of social support. Our finding that active and support-seeking coping strategies were used less frequently as participants aged also signals the importance of adapting intervention components for different age groups. For example, young adults may need encouragement and instruction to realize that just because past attempts to use active or support-seeking strategies did not work does not mean that future attempts will not work.
Our study had several limitations. We had a small sample, and depression can be cyclical, so findings on changes from baseline to follow-up should be interpreted cautiously. We did not use an experimental design to assess intervention effect; we believed that it was premature to use such a design because of the small amount of research available on the integration and evaluation of mental health services in employment training settings. The lack of a comparison group limits the interpretation of findings on the effect of the intervention on mental health outcomes. Our study was implemented at a single employment training program. Future studies could be implemented at multiple programs, thereby increasing the number of study participants and allowing greater generalizability of study results.
Study investigators and community partners learned several lessons, which have shaped the development and implementation of a more comprehensive mental health intervention currently being assessed through a quasi-experimental design. First, thorough training is needed to ensure that paraprofessionals — in our case, peer leaders — feel adequately prepared to deliver the intervention content. In our new intervention (28
), peer leaders were observed by a research team member and found to have done an excellent job in both fully covering intervention content and delivering the content as it was presented in the instructor manual. Second, mental health services need to be integrated into regular employment training activities so participants are "touched" by mental health activities daily and health activities are not viewed as "special" services for a certain group of people but ongoing and accepted activities for all participants. Third, because we could not show overall intervention effects on depressive symptoms or coping strategies in this study, we developed a more targeted approach to providing mental health services in our new intervention. We now use baseline depressive symptom scores to triage participants according to 3 levels of depressive symptoms (low, moderate, and high). Clinicians work more with participants who have more depressive symptoms.
This the first study to our knowledge to examine the effects of an intervention aimed at improving the mental health of adolescents and young adults in employment training programs — a population shown to exhibit worse health and greater health risk than similarly aged in-school peers (27
). This study suggests that alternative strategies — specially modeled for young people or targeted toward specific mental health concerns — may be needed to decrease the severity of depressive symptoms and increase use of coping strategies among adolescents and young adults in employment training programs. Future evaluations of such interventions should use more rigorous quasi-experimental or experimental designs to provide clearer evidence of intervention effect.