Comparative growth of children who are HIV-exposed and uninfected with those who are HIV-unexposed and uninfected at age 1 and 2 years in South Africa: a prospective cohort study.

Background: The effects of maternal HIV infection and antiretroviral therapy (ART) on growth of children who are HIV-exposed and uninfected (CHEU) are uncertain. We aimed to explore the association between in utero exposure to maternal HIV and child growth in the context of ART for all pregnant women living with HIV regardless of CD4 count. Methods: We compared growth at 1 and 2 years of age between CHEU and children who are HIV-unexposed and uninfected (CHUU) in a prospective cohort study (INTERBIO-21st), conducted in Soweto, South Africa, with recruitment of pregnant women between 21 May 2013 and 21 December 2015. Length-for-age (LAZ), weight-for-age (WAZ), weight-for-length (WLZ) and head circumference-for-age (HCAZ) Z-scores, based on the World Health Organization (WHO) Child Growth Standards, were compared longitudinally using age-specific and mixed linear regression. Prevalence of stunting, underweight, wasting and overweight were compared at 1 and 2 years. Findings: At 1 and 2 years of age, 398 (143 CHEU, 255 CHUU) and 286 (82 CHEU, 204 CHUU) children, respectively, were followed up. CHEU and CHUU had LAZ and WAZ scores considerably below the WHO standardised median at 1 year (CHEU: mean (SD) LAZ -1.13 (1.39), WAZ -0.29 (1.33); CHUU: LAZ -0.87 (1.33), WAZ -0.15 (1.20)) and 2 years (CHEU: LAZ -1.00 (1.24), WAZ -0.52 (1.05); CHUU: LAZ -0.94 (1.33), WAZ -0.34 (1.19)), whereas WLZ scores were above the median at year 1 only (CHEU 0.34 (1.27), CHUU 0.42 (1.28). There were no statistically significant differences in growth between CHEU and CHUU, although there was a trend to lower LAZ (adjusted mean difference [aMD] -0.19 [95% CI: -0.43, 0.06]), WAZ (aMD -0.09 [95% CI: -0.30, 0.15]), WLZ (aMD -0.06 [95% CI: -0.30, 0.18]) and HCAZ (aMD -0.14 [95% CI: -0.36, 0.08]) scores in CHEU. Stunting prevalence was high in both groups at both timepoints (1 year CHEU 29.3% [95% CI: 22.4%, 37.3%], CHUU 19.4% [95% CI: 15.0%, 24.8%]; 2 years CHEU 18.8% [95% CI: 11.7%, 28.7%], CHUU 20.5% [95% CI: 15.5%, 26.6%]). The risk of stunting, underweight and wasting was higher in CHEU at 1 year, but not significant in analyses adjusted for maternal age, height, weight, nulliparity, smoking, alcohol use, socioeconomic status, education, marital status, child sex and age (stunting adjusted odds ratio [aOR] 1.56 [95% CI: 0.91, 2.68], underweight aOR 1.69 [95% CI: 0.71, 4.03], wasting aOR 3.64 [95% CI: 0.62, 31.2]), and risks were similar between the groups at 2 years (stunting aOR 0.77 [95% CI: 0.35, 1.61], underweight aOR 1.01 [95% CI: 0.34, 2.70], wasting aOR 0.93 [95% CI: 0.11, 5.39]). At 2 years, CHEU had lower risk of overweight (adjusted OR 0.12 [95% CI: 0.00, 0.89]). Interpretation: No significant differences in growth were observed between CHEU and CHUU children up to 2 years. The prevalence of stunting was high in both groups. Key limitations of this study include limited data on maternal ART and infant breastfeeding, which may have introduced residual confounding. Future studies should aim to identify children that are most at risk of growth impairment to enable the development of targeted interventions for optimising growth and healthy development. Funding: Gates Foundation, Oxford-MRC Doctoral Training Partnership.