HIV / STI
HIV exposure and cardiometabolic health outcomes among South African children in a birth cohort study: a mixed effects modeling assessment of blood pressure over time Chelsie Cintron* Chelsie Cintron Maresa Botha Lesley Workman Tiffany Burd Jennifer Pellowski Toby Mansell David Burgner Heather J Zar Angela M. Bengtson
Children exposed but uninfected with HIV (HEU) in utero have altered immune function, which may affect cardiometabolic health, including blood pressure (BP). Few data on BP trajectories in HEU and HIV-unexposed (HU) children exist beyond early life (<5 years). To address this gap, we explored the association between HIV-exposure status and BP trajectories in early childhood in a birth cohort of South African children. Maternal participants were enrolled from two communities during pregnancy and their children were followed from birth through 8 years with annual anthropometry and BP (from 5 years). We considered both continuous BP percentiles for child age, sex and height and a binary measure of elevated ( ≥90th percentile) BP. We used mixed effects models with random intercepts and interaction terms between HIV exposure status and time to assess BP trajectories and the probability of elevated BP over time. Models were adjusted for maternal factors in pregnancy: BMI, alcohol use, psychological distress, smoking, and socioeconomic status and child factors at birth (gestational age, birthweight, being breastfed, age started solid foods), weight at 5-8 years, and physical activity at 7 and 8 years. Among 810 children, (HU= 648; HEU=162) 52% were female, 15% born preterm, 58% exclusively breastfed, median birthweight was 3080g, 5% had an obese bmi zscore from 5 to 8 years, and 88% reported low physical activity. HEU children had a lower probability of elevated BP compared to HU children apart from at age 7 years (Fig 1A).HEU children had marginally lower systolic and higher diastolic BP percentiles at 5, 6, and 8 years and similar levels as HU children at 7 (Fig 1B). Confidence intervals were wide and overlapped across model results (Fig 1 A and B). With similarities in BP by HIV status, our results are largely reassuring. Future studies should investigate BP trajectories into later childhood and factors that mediate or moderate BP trajectories in this population.
