The present study expands the existing literature and supplements today's knowledge on the relationship between personal, psychosocial and lifestyle factors and depressive symptoms among adolescents. The study aimed to investigate the variety of depressive symptoms predictors-personal resources, adverse school and family, health, lifestyle-related (sense of coherence, self-esteem, school involvement, negative acts at school, family stress and violence, psychosomatic health complaints, physical activity, smoking, alcohol) as well as gender, employing hierarchical linear regression analysis in a large representative sample of adolescents (N = 2212) in Kaunas, Lithuania. Four blocks of predictors were employed in hierarchical linear regression analysis. In the final model 64.9% of depressive symptoms were explained by all the predictors. Sense of coherence was the strongest predictor of depressive symptoms (standardized regression coefficient β = -605, p < 0.001 in the first model and β = -263, p < 0.001 in the final model after adjustment for all other independent variables) and accounted for 36.6% of variance. In conclusion, this study supports the notion that depressive symptoms among adolescents have multifactorial origins with many predictors showing significant effect seizes. Therefore, high sense of coherence and self-esteem, school involvement, higher levels of physical activity would be protective and influence lower levels of depressive symptoms among adolescents. Exposure to negative acts at school and negative experiences in the family, psychosomatic health complaints, smoking would increase the probability of depressive symptoms. Girls are more prone to depression as compared to boys.
Keywords: adolescents; depressive symptoms; hierarchical linear regression analysis; predictor; sense of coherence.