Relationship between pre-pregnancy maternal body mass index and infant weight trajectories in HIV-exposed and HIV-unexposed infants

Abstract

Background: Maternal HIV and antiretroviral therapy (ART) exposure in utero may influence infant weight, but the contribution of maternal y body mass index (BMI) to early life overweight and obesity is not clear.

Objective: To estimate associations between maternal BMI at entry to antenatal care (ANC) and infant weight through approximately 1 year of age and to evaluate whether associations were modified by maternal HIV status, maternal HIV and viral load, breastfeeding intensity through 6 months or timing of entry into ANC.

Methods: We followed HIV-uninfected and -infected pregnant women initiating efavirenz-based ART from first antenatal visit through 12 months postpartum. Infant weight was assessed via World Health Organization BMI and weight-for-length z-scores (WLZ) at 6 weeks, 3, 6, 9 and 12 months. We used multivariable linear mixed-effects models to estimate associations between maternal BMI and infant z-scores over time.

Results: In 861 HIV-uninfected infants (454 HIV-exposed; 407 HIV-unexposed), nearly 20% of infants were overweight or obese by 12 months of age, regardless of HIV exposure status. In multivariable analyses, increasing maternal BMI category was positively associated with higher infant BMIZ and WLZ scores between 6 weeks and 12 months of age and did not differ by HIV exposure status. However, HIV-exposed infants had slightly lower BMIZ and WLZ trajectories through 12 months of age, compared with HIV-unexposed infants across all maternal BMI categories. Differences in BMIZ and WLZ scores by HIV exposure were not explained by timing of entry into ANC or maternal viral load pre-ART initiation, but z-scores were slightly higher for HIV-exposed infants who were predominantly or exclusively versus partially breastfed.

Conclusions: These findings suggest maternal BMI influences early infant weight gain, regardless of infant HIV exposure status. Intervention to reduce maternal BMI may help to address growing concerns about obesity among HIV-uninfected children.

Keywords: BMI; HIV; HIV-exposed; South Africa; infant growth; infant weight.

Figure 1.

Study Population

Figure 2.. Multivariable association between maternal body mass index (BMI) category at entry into antenatal care and infant body mass index BMIZ and weight for length (WLZ) z-scores between 6 weeks and approximately 12 months of age.

Estimates from linear (continuous BMIZ and WLZ) and modified Poisson (BMIZ and WLZ >2 SD) multivariable mixed effects models with random intercepts adjusted for baseline measures of maternal poverty (categorical); marital status (categorical); age (spline); gravidity (spline); depressive symptoms (binary); alcohol use (binary); and maternal HIV/viral load status at entry into antenatal care (categorical) and food insecurity (categorical) at 12 months postpartum.

Figure 3.. Multivariable association between maternal BMI category at entry into antenatal care and infant BMIZ and WLZ scores between 6 weeks and approximately 12 months of age – stratified by maternal HIV status and disease severity.

Estimates from linear multivariable mixed effects models with random intercepts adjusted for baseline measures of poverty (categorical); marital status (categorical); maternal age (spline); gravidity (spline); depressive symptoms (binary); alcohol use (binary) and postpartum food security (categorical). VL: viral load, measured by number of log₁₀ HIV RNA copies/mL and assessed at entry into antenatal care.