Symptoms of Depression In Middle School Teenagers Linked to Risky Behavior

D. Hollander

First published online:

Middle school students who report symptoms of depression are at risk for subsequently engaging in potentially harmful sexual behavior, a study based on data from two waves of the National Longitudinal Study of Adolescent Health (Add Health) shows.1 Both males and females categorized as having a high level of depressive symptoms at the time of the first interview had significantly elevated odds of reporting at least one of four specified sexual risk behaviors about one year later. Likewise, for both genders, in calculations using a continuous measure of depressive symptoms, the odds of engaging in at least one risky behavior rose with symptom level. Despite the similarities in findings for the overall measures, however, associations between depressive symptom level and individual behaviors varied by gender.

Add Health followed a cohort of youth who were in grades 7–12 at baseline, in 1995; the study of the relationships between depressive symptoms and risky sexual behavior used data collected then and in the subsequent wave, roughly a year later. Respondents were eligible for inclusion in the analytic sample if they were unmarried at the time of the second interview, had had sex between survey waves, provided information on the sexual risk behaviors being studied and, at baseline, had completed at least 16 items on a 19-item scale assessing depressive symptoms in the week before the interview. For the main analyses, the researchers categorized participants' level of depressive symptoms as high, moderate or low; in secondary analyses, they substituted a continuous measure of depressive symptoms. All analyses were conducted separately for males and females.

Some 1,921 males and 2,231 females were included in the sample; the majority were white, aged 15–17 and nonpoor. Four in 10 participants of each gender lived with both biological parents, two in 10 in a stepfamily and the rest in some other arrangement; about one-quarter reported a high degree of religiosity. Small proportions reported same-sex attraction or behavior (2–3%) or sexual experience at age 10 or earlier (7% of males and 2% of females). Nine percent of males and 16% of females scored high on the depressive symptoms scale at baseline (a statistically significant difference); 42–43% of each registered moderate scores. In the first interview, close to half of youth of each gender reported at least one of the following risk behaviors: not having used a condom at last sex, not having used a contraceptive at last sex, having used alcohol or drugs at the time of last sex and having had three or more sexual partners. At follow-up, 63% of males and 55% of females reported having engaged in at least one of these behaviors between interviews.

In regression analyses controlling for background and risk-related characteristics, males with a high level of depressive symptoms at baseline were significantly more likely than those with a low level to say at follow-up that they had engaged in at least one of the specified risk behaviors between interviews (odds ratio, 1.7). They had elevated odds of reporting nonuse of condoms and contraceptives at last sex (1.6 and 1.8, respectively), and of using alcohol or drugs at that time (2.5). Males with a moderate level of depressive symptoms at baseline were not at increased risk of subsequently engaging in these behaviors.

Multivariate analyses using the continuous measure of depressive symptoms yielded similar results. Males' odds of having engaged in any risk behavior during the follow-up period increased by 3% for every one-point increase on the depressive symptoms scale; the odds of nonuse of condoms, nonuse of contraceptives and substance use at last sex each rose by 2–4% per one-point increase in the level of depressive symptoms.

Among females, a high level of depressive symptoms at baseline was associated with a significantly elevated likelihood of engaging in risky behavior between interviews (odds ratio, 1.5), but not with the odds of reporting any particular behavior. A moderate depressive symptoms score was marginally associated with an elevated risk, and the odds of substance use at last sex were significantly raised for females with scores in this category (1.4). Additionally, the analysts observe that the odds ratios for a moderate level of symptoms were consistently between those for low and high levels, a pattern that they say may suggest a dose-response relationship between depressive symptoms and the likelihood of risky behavior among females.

The analyses using the continuous measure of depressive symptoms revealed differences not detected with use of the categorical measure. Females' odds of reporting any risk behavior and each behavior except substance use at last sex increased by 2% with each one-point increase in their depressive symptoms score. The analysts speculate that differences in statistical power may explain the differences between models in results for females.

In the analysts' view, the associations identified in this study should be of interest to primary care providers, mental health practitioners and sexual health care providers who counsel young people and can help to ensure that they get all of the services they need. Furthermore, the analysts note that "increased understanding of gender-specific mechanisms for the association between depressive symptoms and [sexual risk behavior] would be helpful for intervention design." Finally, they press for "expanded, population-based efforts in mental health promotion, prevention, and care for adolescents," and suggest that schools are a "natural setting" for such efforts.—D. Hollander


1. Lehrer JA et al., Depressive symptoms as a longitudinal predictor of sexual risk behaviors among US middle and high school students, Pediatrics, 2006, 118(1):189–200.