Association between high body mass index and adverse birth outcomes by HIV and ART status in Cape Town, South Africa

Abstract

Background: Tested independently, studies report that obesity and HIV infection and/or ART use in pregnancy are associated with adverse birth outcomes. However, there is limited data on the combined impact of these maternal factors on adverse birth outcomes. Given the high prevalence of obesity and HIV infection in Sub-Saharan Africa (SSA), understanding these associations is important. This study examined the association of the double burden of high maternal body mass index and HIV infection/ART use in pregnancy with adverse birth outcomes. Methods: Part A of this mini-dissertation presents the study protocol which outlines the rationale, aim and objectives of the study; the research methodology, analysis plan and ethical considerations. Part B is the literature review of studies conducted in SSA which investigated the relationship between BMI and HIV infection and adverse birth outcomes of interest. Part C is the journal-formatted manuscript which presents the results and discussion of the study findings in relation to other scholars. The referencing style used for the whole thesis is Vancouver as required by the journal chosen for the formatting of the manuscript. We used data collected from a large observational Prematurity Study that enrolled HIV-infected and HIV-uninfected women seeking antenatal care at Gugulethu MOU in Cape Town between April 2015 and October 2016. A subset of HIV-infected women who booked early (≤24weeks) was prospectively followed through delivery and was used to study gestational weight gain (GWG) and adverse birth outcomes. Data was obtained from review of medical records and study questionnaires. Logistic regression was used to compare birth outcomes by BMI status: preterm delivery (PTD), low/high birthweight (LBW/HBW) and small/large gestational age (SGA/LGA) between HIV-uninfected and -infected women; and between HIV-infected women who initiated ART before pregnancy and those who initiated ART during pregnancy. Using the subset of HIV-infected women who booked early (≤24weeks), we compared the adverse birth outcomes between low, adequate and high GWG. Results: Of the 2779 participants included in the analysis, 20% had normal BMI, 29% were overweight, 51% were obese and 39% were HIV-infected. Overall, there was no association between obese BMI and PTD (aOR 1.06, 95% CI 0.75-1.49). Instead, obese BMI was negatively associated with LBW (aOR 0.53; CI: 0.39-0.72) and SGA infants (aOR 0.55, 95% CI 0.41-0.75) compared to normal BMI women. Stratifying by HIV infection showed similar results for LBW (aOR 0.54; CI: 0.35-0.83) and SGA (aOR 0.60, 95% CI 0.38-0.94) in obese HIV-infected women compared to corresponding women with normal BMI. However, comparison of obese HIV-uninfected and obese HIV-infected women showed a higher incidence of LBW and SGA infants in obese HIV-infected women (12% vs 8%). The association of obese BMI and LBW and SGA in HIV-infected women did not differ by timing of ART initiation. In terms of HBW and LGA, overall, obese BMI was positively associated with HBW (aOR 2.00; CI: 1.13-3.57) and LGA infants (aOR 1.98, 95% CI 1.40-2.80) compared to normal BMI women. Stratifying by HIV infection also showed a positive association between obese BMI and HBW (aOR 2.54; CI: 1.17-5.53) and LGA (aOR 2.30; CI: 1.46-3.62) in HIV-uninfected women. Although a similar positive association was also obtained in obese HIV-infected women, the strength of this association was weaker for both HBW (aOR 1.41; CI: 0.59-3.34) and LGA (aOR 1.58; CI: 0.91-2.72). When the analysis was restricted to HIV-infected women by timing of ART initiation we found that obese women who initiated ART during pregnancy had 3-fold likelihood of having LGA infants (aOR 3.26; CI: 1.32-8.09) and those who initiated ART before pregnancy had a reversed effect (aOR 0.87; CI: 0.43-1.78) compared to respective normal BMI women. However, restricting the analysis to obese HIV-infected women only revealed a counter effect of the two conditions where the frequencies of both LGA and SGA are high. Abnormal gestational weight gain had no association with PTD, LBW, HBW and SGA. However, we showed that GWG lower than the IOM recommended values reduced the likelihood of having LGA infants (aOR 0.29; CI: 0.12-0.70) compared to adequate GWG. Conclusions: Obese HIV-infected women appear to be cushioned by their BMI against LBW and SGA when compared to normal BMI. However, comparison of these outcomes amongst women who are either obese or HIV-infected reveal a higher burden of both SGA and LGA infants in obese HIV-infected women, regardless of ART initiation status.