Association of adolescent risk behaviors with mental health symptoms in high school students

Purpose: To examine the hypothesis that self-reported symptoms of depression and stress may be associated with other risk behaviors. Methods: A secondary data analysis of the 1992 Massachusetts Adolescent Health Survey involving a representative sample of 2224 ninth and twelfth grade students was performed. The dichotomous dependent variable was positive if the adolescent reported feeling depressed or stressed for 10 or more days in the past month. Potential independent variables examined were age, gender, race/ethnicity, and 14 risk or protective behaviors: each scored on a seven point scale representing increasing frequency of a behavior in the past month. A four-level sexual risk variable was constructed as well. Associations were assessed using Chi-square, and phi/contingency coefficients, and logistic regression analyses to predict the odds of reporting depression/stress. Results: The mean age of the sample was 16.2 +/-1.6 years; 52% males; 78% were white, 9% black, 6% Latino, 2% Asian, and 4% other racial/ethnic heritage; 35% reported feeling depressed/stressed greater than or equal to 10 days in the past month. A logistic regression model found that feelings of depression/stress were associated with increasing age (OR = 1.09 with each additional year [95% Cl, 1.02-1.18]), female gender (3.28 [2.62-4.12]); increasing levels of tobacco use (1.07 [1.01-1.12]), physical fights (1.19 [1.111.28]); and non-use of birth control compared with never having been sexually active (1.81 [1.31-2.49]). Independent variables of reporting depression/stress for males included increasing age (1.15 [1.03-1.28]), and physical fights (1.20 [1.10-1.30]), and non-use of birth control compared with never sexually active (1.91 [1.28-2.92]). Independent risk and protective factors for females included tobacco use (1.10 [1.02-1.19]), healthy diet (0.89 [0.83- 0.96]), and always (1.49 [1.03-2.28]) or sometimes used birth control (1.56 [1.03-1.28]) compared with never sexually active. Conclusions: Female gender had greater than threefold increased odds of reporting depression/stress. Other associations, with some gender differences, include older age, physical fights, non-use of birth control, lack of a healthy diet, and use of tobacco.