Adolescent Mental Health Problems and Adult Human Capital: Findings From the South African Birth to Twenty Plus Cohort at 28 Years of Age

Abstract

Purpose: We investigated associations between adolescent internalizing and externalizing problems and adult human capital in a non-Western setting. Little is known about adolescent mental health problems and adult outcomes in low- and middle-income countries, many of which are characterized by high levels of adversities.

Methods: Data came from the Birth to Twenty Plus cohort, started in Soweto, Johannesburg, South Africa, in 1990. We estimated associations of internalizing and externalizing problems at the age of 14 years with self-reported educational, employment, welfare receipt, psychosocial (psychological distress, criminality, substance use), interpersonal (social isolation, intimate partner violence, partnership status), and HIV outcomes at the age of 28 years.

Results: Adolescents with high internalizing problems were less likely to have completed secondary school or be formally employed and more likely to report psychological distress. Those with high levels of externalizing problems were more likely to report adulthood criminal activity and substance use. We found significant associations between internalizing and externalizing problems and intimate partner violence. There was no association between adolescent mental health problems and welfare receipt, HIV, social isolation, or partnership status. Men were more likely to report incomplete secondary education, no formal employment, criminality and substance use, social isolation, and no serious relationship, whereas women were more likely to experience psychological distress and be in receipt of welfare.

Conclusions: Adolescent mental health problems are associated with long-term negative adult functioning under varying socioeconomic conditions. Interventions to recognize and address youth mental health problems in low- and middle-income countries are needed to avert serious adverse adult and societal consequences.

Keywords: Adolescent; Birth cohort; Externalizing problems; Human capital; Internalizing problems; South Africa.

Figure 1

Distribution of internalizing and externalizing scores among men and women. The figure shows the distribution of internalizing (A) and externalizing (B) problem scores in men (blue) and women (red). Differences are expressed as effect size (Cohens' D), and significance was tested using t-tests. ∗∗∗p < .001.

Figure 2

Adjusted associations between adolescent internalizing and externalizing problems and adult human capital outcomes. The figure shows in the y-axis, the adjusted odds ratio for the association of high psychological problems (vs. low) with adult outcomes for the whole sample and for men (blue) and women (red) separately. The central dot represents the estimated odds ratio, and the bar represents the 95% confidence intervals. The dotted line represents the null association. We reported fully adjusted estimates (see Supplementary Table 2 for the exact numbers). In addition, the figure shows associations for the imputed sample (round dot) and for the available sample (triangle) to show consistency of the estimates across these two methods for handling missing data. IPV = intimate partner violence.