Psychological morbidity and parenting stress in mothers of primary school children by timing of acquisition of HIV infection: a longitudinal cohort study in rural South Africa

Abstract

Longitudinal maternal mental health data are needed from high HIV prevalence settings. The Siyakhula Cohort (SC) is a population-based cohort of HIV-positive and negative mothers (n=1506) with HIV-negative children (n=1536) from rural South Africa. SC includes 767 HIV-negative mothers; 465 HIV-positive in pregnancy; 272 HIV-positive since pregnancy (n=2 missing HIV status). A subgroup (n=890) participated in a non-randomized breastfeeding intervention [Vertical Transmission Study (VTS)]; the remaining (n=616) were resident in the same area and received antenatal care at the time of the VTS, but were not part of the VTS, instead receiving the standard of care Prevention of Mother-to-Child Transmission (PMTCT) Programme. In secondary analysis we investigated the prevalence of, and factors associated with, psychological morbidity amongst mothers who were still the primary caregiver of the child (1265 out of 1506) at follow-up (7–11 years post-birth). We measured maternal depression (Patient Health Questionnaire-9), anxiety (General Anxiety Disorder Scale-7) and parenting stress (Parenting Stress Index-36), using standardized cut-offs and algorithms. In total, 75 (5.9%) mothers met criteria for depression, 37 (2.9%) anxiety and 134 (10.6%) parenting stress. Using complete case logistic regression (n=1206 out of 1265 mothers) as compared to being HIV-negative, testing HIV-positive in pregnancy doubled odds of depression [adjusted odd ratios (aOR)=1.96 [1.0–3.7] P=0.039]. Parenting stress was positively associated with acquisition of HIV after pregnancy (aOR=3.11 [1.9–5.2] P<0.001) and exposure to household crime (aOR=2.02 [1.3–3.2] P=0.003); negatively associated with higher maternal education (aOR=0.29 [0.1–0.8] P=0.014), maternal employment (aOR=0.55 [0.3–0.9] P=0.024). Compared with the standard of care PMTCT, VTS mothers had reduced odds of parenting stress (aOR=0.61 [0.4–0.9] P=0.016). Integrating parental support into mostly bio-medical treatment programmes, during and beyond pregnancy, is important.