Association between maternal HIV infection and low birth weight and prematurity: a meta-analysis of cohort studies

Peng-Lei Xiao,Yi-Biao Zhou,Yue Chen,Mei-Xia Yang,Xiu-Xia Song,Yan Shi,Qing-Wu Jiang
DOI: https://doi.org/10.1186/s12884-015-0684-z
2015-01-01
BMC Pregnancy and Childbirth
Abstract:Background To assess the association between maternal human immunodeficiency virus (HIV) infection and low birth weight (LBW)/prematurity (PTD), we conducted a meta-analysis of cohort studies of HIV infected and uninfected women. Methods Several English and Chinese databases were searched (updated to May 2015) to find the studies reporting infant outcomes associated with exposure to maternal HIV infection during pregnancy. Relevant articles were manually selected based on several inclusion and exclusion criteria. Results Fifty-two cohort studies including 15,538 (for LBW) and 200,896 (for PTD) HIV infected women met the inclusion criteria. There was significant heterogeneity among studies for maternal HIV infection associated with LBW/PTD (I 2 = 71.7 %, P < 0.05, and I 2 = 51.8 %, P < 0.05 for LBW and PTD, respectively). The meta-analysis demonstrated that the maternal HIV infection was significantly associated with both LBW (pooled odds ratio (OR): 1.73, 95 % confidence interval (CI): 1.64, 1.82, P < 0.001) and PTD (pooled OR: 1.56, 95 % CI: 1.49, 1.63, P < 0.001). No significant difference in the relationship between maternal HIV infection and adverse pregnancy outcomes was detected among the groups of different study periods. HIV infected women were at slightly higher risk of LBW in developing countries compared with women in developed countries (OR: 2.12 (95 % CI: 1.81, 2.48) vs. 1.75 (95 % CI: 1.44, 2.12)). Antiretroviral drugs usage did not significantly change the associations of maternal HIV exposure with LBW and PTD. Conclusions HIV infected women were at higher risk of having a low birth weight infant or a preterm delivery infant compared with uninfected women. Such associations did not change significantly over time or were not significantly affected by the usage of antiretroviral drugs.
What problem does this paper attempt to address?