BROWN, D. R. and C. J. BLANTON. Physical activity, sports participation, and suicidal behavior among college students. Med. Sci. Sports Exerc., Vol. 34, No. 7, pp. 1087–1096, 2002.
Purpose: To evaluate the relationship between physical activity, sports participation, and suicidal behavior among college students (N = 4,728).
Methods: Data from the 1995 National College Health Risk Behavior Survey were analyzed. Students were classified as engaging in frequent vigorous activity 6–7 d·wk−1, vigorous activity 3–5 d·wk−1, moderate activity, low activity, or no activity. Sports participation was dichotomized into “yes” or “no” participation. Suicidal behavior was defined as thoughts about, plans for, or attempts at suicide during the 12 months before completing the survey. Data were stratified by sex and multivariable logistic regression modeling, calculated odds ratios (ORs) (adjusted for age, race/ethnicity, Body Mass Index/weight perception, cigarette smoking, episodic heavy alcohol use, drug use, and either activity level or sport participation) for suicidal behavior as associated with physical activity and sports participation.
Results: Adjusted ORs show that men in the “low activity” group were at almost half the odds (adjusted OR = 0.54;P < 0.015) of reporting suicidal behavior than men in the “not active” group. Women who engaged in moderate or frequent vigorous activity were at greater odds of reporting suicidal behavior compared with inactive women; OR = 1.76 (P < 0.035) and 1.99 (P < 0.034) respectively. Sports participation was protective against suicidal behavior. Adjusted ORs show that men who did not participate in sports were 2.5 times (P < 0.0003) more likely to report suicidal behavior than men who were sports participants. Women not participating in sports had 1.67 times the odds of reporting suicidal behavior than women sports participants (P < 0.05).
Conclusions: Associations were found between sports participation/selected patterns of physical activity and suicidal behavior. Causal factors mediating the relationships need to be identified.