Risk factors for adolescent mental disorders

Abstract

Background: The epidemic of adolescent mental disorder is of increasing public health concern, as a leading cause of disability with long lasting effects throughout life. Action is needed to protect vulnerable adolescents worldwide. However, most of the evidence concerning adolescent mental disorders comes from Western settings where socio-economic position is often associated with both potential determinants and the outcomes of mental disorders and with adolescent mental disorders, making these results susceptible to residual confounding. Evidence from a non-Western setting with a different confounding structure and Mendelian randomization can be valuable in disentangling whether these associations are contextually specific or biologically based potential targets of intervention. Objectives: This thesis took advantage of a large (n=8,327), population-representative Chinese birth cohort from Hong Kong to examine the association of modifiable risk factors, including birth weight, life course body mass index (BMI), gestational age and age at onset of puberty, with adolescent mental disorders indicated by depressive symptoms, self-esteem and emotional and behavioral problems. Methods: Mental disorders were assessed by self-reported Patient Health Questionnaire-9 item at ~ 13 years for depressive symptoms, self-reported Culture-Free Self-Esteem Inventory at ~11 years and parent-reported Rutter scores for emotional and behavioral problems at 7 years and ~ 11 years. Partial least squares regression was used to examine the adjusted association of birth weight and life course adiposity with adolescent mental disorders. Multivariable linear and logistic regression was used to examine the adjusted association of gestational age and age at onset of puberty with adolescent mental disorders. To verify these findings a Mendelian randomization study using a large publically available case-control study was meta-analyzed with all previous such studies to estimate the effect of BMI on adult depression. Results: Birth weight, life course adiposity and gestational age were largely unrelated to adolescent mental disorders, although lower birth weight and very/moderate preterm birth were associated with higher Rutter sub scores for specifically hyperactivity. Consistent with these findings in adults genetically determined BMI was unassociated with depression, odds ratio (OR) 1.07, confidence interval (CI) 0.86 to 1.34. The association of onset of puberty with depressive symptoms varied by sex (interaction p value <0.05). Earlier onset of breast development was associated with higher risk of the presence of depression (OR 0.83, 95% CI [0.70 to 0.98]). Discussion: In this population-representative birth cohort from a developed non-Western setting with little social patterning of birth weight, life course adiposity, gestational age or mental disorders, birth weight, gestational age and life course adiposity do not appear to be factors in the development of adolescent mental disorders. Mendelian randomization confirmed the lack of association of BMI with depression. Inconsistencies with studies from Western settings suggest setting-specific unmeasured confounding may underlie some of the observed associations. Early onset of puberty in girls was associated with depressive symptoms. Whether these findings concerning age at pubertal onset and depressive symptoms are indicators of the effects of hormones or transient effects of social pressures remains to be determined. Overall, causes of adolescent mental disorders should be sought elsewhere than life course adiposity.