Background
HIV-Exposed Uninfected (HEU) children have worse cardiometabolic health outcomes, compared to HIV-Unexposed uninfected (HU) children, but the effect of maternal cardiometabolic health in pregnancy on child outcomes is not clear. Our study investigated the impact of maternal gestational lipid biomarkers on child cardiometabolic health indicators in a birth cohort of South African HEU and HU children, the Drakenstein Child Health Study (DCHS).
Methods
The DCHS is a longitudinal birth cohort in a peri-urban area of South Africa (N=1143 live born infants). Our study randomly sampled 260 mother-child pairs from DCHS (N= 100 HEU & N= 160 HU). We used multivariable linear regression to assess associations between maternal lipid biomarkers from 2nd trimester of pregnancy (20-28 weeks’ gestation) with child cardiometabolic indicators at 5-8 years of age and explored potential effect modification by HEU status in stratified analyses. Child cardiometabolic indicators included lipid and glucose metabolism, blood pressure, and body composition.
Results
Associations were found between maternal total cholesterol (TC) and child TC and low-density lipoprotein (LDL) [TC; mean difference [MD] 0.19, 95% CI 0.09, 0.30] & [LDL; MD 0.21, 95% CI 0.12, 0.31]. When stratified by HIV exposure status, the relationship between maternal TC and child TC was stronger in HEU children than in HU children [HEU; MD 0.29, 95% CI 0.13, 0.45 vs HU; MD 0.13, 95% CI -0.01, 0.27]. The relationship between maternal TC and child LDL was also observed to be stronger in the HEU child [HEU; MD 0.26, 95% CI 0.13, 0.40 vs HU MD 0.18, 95% CI 0.05, 0.30]. Maternal LDL had similar associations with child TC and LDL (table 1). Maternal high-density lipoprotein (HDL) influenced the child HDL levels [MD 0.13, 95% CI 0.03, 0.23]; when stratified, this association was only demonstrated in the HU child [HU; MD 0.21, 95% CI 0.08, 0.34 vs HEU; MD 0.02, 95% CI -0.15, 0.19]. No significant associations were found between maternal lipids and child HbA1c, insulin resistance, systolic or diastolic blood pressure or body composition.
Conclusions
In a birth cohort of South African HEU and HU children, maternal gestational lipid profiles influenced child lipid metabolism, and this relationship may differ based on in-utero HIV exposure status. These findings suggest optimizing maternal lipid profiles during pregnancy may improve child cardiometabolic indicators, particularly in the context of HIV exposure and should be evaluated in larger cohorts.