Preliminary study on the causes of anemia in newborns delivered by human immunodeficiency virus infected mothers / 中华传染病杂志

ABSTRACT

Objective:To investigate the causes of anemia in newborns delivered by human immunodeficiency virus (HIV) infected mothers.Methods:This was a retrospective study. Forty-two newborns delivered by HIV infected mothers during January 2010 and May 2019 in Beijing Ditan Hospital Affiliated to Capital Medical University were selected. According to the hemoglobin levels of newborns on the days of their birth, newborn cases were divided into two groups, anemia group and non-anemia group. The clinical data including gestational ages, birth weight, maternal anemia status during pregnancy, using of antiviral drugs during pregnancy, percentages of HIV RNA positivity in early pregnancy/pre-treatment and before delivery, maternal percentage of different CD4 + T lymphocyte counts in early pregnancy/pre-treatment and before delivery between two groups were compared. The efficacies of relative indicators for prediction of anemia in newborns were evaluated by the area under receiver operating characteristic curve (AUROC). Differences between groups were compared by chi-square test. Results:Among 42 cases of newborns, 14 cases were in anemia group and 28 cases in non-anemia group. There were no statistical differences in gestational ages, birth weight, maternal anemia status during pregnancy and positive percentage of HIV RNA before delivery between two groups ( χ2=2.211, 1.025, 1.362 and 3.783, respectively, P=0.283, 0.763, 0.181 and 0.092, respectively). In anemia group, 11 mothers took zidovudine during pregnancy, which was 12(42.86%) in non-anemia group. The difference was statistically significant ( χ2=4.359, P=0.037). Eight cases of mothers with HIV RNA positive in early pregnancy/pre-treatment in the anemia group, which was 11(39.29%) in the non-anemia group. The difference was statistically significant ( χ2=6.490, P=0.011). The number of CD4 + T lymphocyte count ≤500/μL was 13 in early pregnancy/pre-treatment in anemia group, which was 20(71.43%) in the non-anemia group. The difference was statistically significant ( χ2=16.396, P<0.01). The number of CD4 + T lymphocyte ≤0.28 was 13 in early pregnancy/pre-treatment in the anemia group, which was 19(67.86%) in the non-anemia group ( χ2=19.908, P<0.01). The number of CD4 + T lymphocyte count ≤500/μL was 14 before delivery, which was 15(53.37%) in the non-anemia group ( χ2=9.536, P=0.008). The number of CD4 + T lymphocyte ≤0.28 before delivery was 14 in anemia group, which was 15(53.37%) in the non-anemia group ( χ2=9.750, P=0.006). According to the receiver operating characteristic curve results, the AUROC, optimal cut-off value, sensitivity and specificity of CD4 + T lymphocyte count before delivery in predicting neonatal anemia were 0.708, 476.0/μL, 100.0% and 50.0%, respectively. The AUROC, optimal cut-off value, sensitivity and specificity of maternal CD4 + T lymphocyte percentage before delivery in predicting neonatal anemia were 0.719, 0.275, 100.0% and 53.6%, respectively. Conclusion:Low CD4 + T lymphocyte level in HIV-infected mothers before delivery, HIV positive in early pregnancy/pre-treatment and using of zidovudine during pregnancy may be associated with neonatal anemia.