Neonatal erythropoiesis and subsequent anemia in HIV-positive and HIV-negative Zimbabwean babies during the first year of life: a longitudinal study

Abstract

Background: Anemia is common in HIV infection and independently associated with disease progression and mortality. The pathophysiology of HIV-related anemia is not well understood especially in infancy.

Methods: We conducted a longitudinal cohort study nested within the Zimbabwe Vitamin A for Mothers and Babies Project. We measured hemoglobin, erythropoietin (EPO), serum transferrin receptor (TfR) and serum ferritin at 6 weeks, 3 and 6 months of age and hemoglobin at 9 and 12 months in 3 groups of randomly selected infants: 136 born to HIV-negative mothers, and 99 born to HIV-positive mothers and who were infected themselves by 6 weeks of age, and 324 born to HIV-positive mothers but who did not become infected in the 6 months following birth.

Results: At one year of age, HIV-positive infants were 5.26 (adjusted odds ratio, P < 0.001) times more likely to be anemic compared to HIV-negative infants. Among, HIV-negative infants, EPO was or tended to be inversely associated with hemoglobin and was significantly positively associated with TfR throughout the first 6 months of life; TfR was significantly inversely associated with ferritin at 6 months; and EPO explained more of the variability in TfR than did ferritin. Among infected infants, the inverse association of EPO to hemoglobin was attenuated during early infancy, but significant at 6 months. Similar to HIV-negative infants, EPO was significantly positively associated with TfR throughout the first 6 months of life. However, the inverse association between TfR and ferritin observed among HIV-negative infants at 6 months was not observed among infected infants. Between birth and 6 months, mean serum ferritin concentration declined sharply (by approximately 90%) in all three groups of babies, but was significantly higher among HIV-positive compared to HIV-negative babies at all time points.

Conclusion: HIV strongly increases anemia risk and confounds interpretation of hematologic indicators in infants. Among HIV-infected infants, the EPO response to anemia is attenuated near the time of infection in the first weeks of life, but normalizes by 6 months.

Figure 1

Model of the relationships between indicators or iron status and erythropoiesis. The model is based on the assumptions that hypoxia determines EPO production which in turn stimulates erythropoietic activity that drives the need for iron and corrects the hypoxia. Abbreviations: Hb, hemoglobin; EPO, erythropoietin; TfR, transferrin receptor; SF, serum ferritin.

Figure 2

Hemoglobin concentration during the first year of life in Zimbabwean babies by HIV status. Values are means with 95% confidence intervals, adjusted for birthweight. The asterisk indicates a statistically significant difference in mean hemoglobin among the groups at a certain time point (6 weeks, F = 3.54, P = 0.030; 9 months, F = 4.37, P = 0.005; 12 months, F = 11.34, P < 0.001). Abbreviations: Nn, mother HIV negative, baby HIV negative; Pn, mother HIV positive, baby HIV negative at 6 months; Pp, mother HIV positive, baby HIV positive at 6 weeks.