Iron Nutrition, Hematologic Status and Infections in Tanzan

Abstract

Background Disordered iron metabolism underlies the pathogenesis of iron deficiency, anemia of inflammation and elevated iron status. These iron disorders are common and they are associated with important consequences including mortality. The co-occurrence of infections with these disorders complicates their epidemiology, either as a key component of their etiology or by making their diagnosis or control challenging. The central aim of this dissertation was to examine the interactions of infection risk and inflammation with iron nutrition among pregnant women at risk of malaria and people living with HIV (PLWHIV) initiating highly active antiretroviral therapy (HAART) in Tanzania. Methods We conducted three posthoc analyses using data from recently completed randomized controlled trials in Tanzania. The first was among 1,450 HIV-negative pregnant women without severe iron deficiency in Tanzania. Participants were followed up from antenatal clinic enrollment till delivery and 6-weeks postpartum. The second analysis was among 400 PLWHIV initiating HAART in the context of test-and-treat for HIV. Iron status biomarkers were measured at HAART initiation, 1, 6 and 12 months of follow-up. In addition, the incidence of virologic outcomes and weight loss were assessed. The third analysis was among participants 2,293 PLWHIV initiating HAART on account of clinical severity. Participants had dietary assessment at baseline using a food frequency questionnaire and the incidence of mortality, pulmonary tuberculosis, severe anemia and weight loss were evaluated during follow-up. Results In the first analysis, we found that both maternal iron deficiency anemia (IDA) and non-iron deficiency anemia (NIDA) among pregnant women were associated with adverse maternal and neonatal outcomes, compared to non-anemia. There were statistically significant differences in the mean differences for delivery hemoglobin among baseline IDA (vs. non-anemia) compared to baseline NIDA (vs. non-anemia). There were also statistically significant differences in the mean differences for delivery hemoglobin in the iron supplemented group compared to placebo across the anemia groups. These statistically significant differences were however small in magnitude, and unlikely to be clinically meaningful. In the second analysis, we found that the prevalence of both iron deficiency and elevated iron status were substantial among PLWHIV initiating HAART, whether inflammation was accounted for using the higher cutoff, Thurnham-correction or regression correction methods. The prevalence of iron deficiency (ID) worsened considerably during follow-up. The shape of the trajectories of iron status were otherwise similar across the inflammation correction methods. Elevated iron status was associated with the risk of high viral load and incident underweight during follow-up based on the higher cutoff method, but not the other inflammation correction methods. In the third analysis, we found that overall iron intake from animal-source foods as well as intake of iron-rich food groups such as red meat and legumes were associated with the risk of all-cause and AIDS-related mortality among PLWHIV initiating HAART. In addition, we found relationships of red meat intake with severe anemia, and fish intake with pulmonary tuberculosis. Conclusion By exploring how the occurrence and consequences of iron disorders differ by infection status, we may be able to improve public health measures to address them, including when targeted approaches may be beneficial or unnecessary, and how targeting may be effectively accomplished.