Date of Award


Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)

First Advisor

Michele Barry


Objective: Although highly active antiretroviral therapy (HAART) resolves a significant proportion of anemia among HIV-patients in Western cohorts, outcomes may vary in developing countries, due to a higher prevalence of nutritional deficiencies, intestinal parasites, tuberculosis, and opportunistic infections. The purpose of this study is to describe the prevalence of, factors associated with, and influence of HAART on, anemia among HIV-infected individuals in South India. Methods: To examine factors associated with anemia, the first-recorded hemoglobin values for adults who visited an HIV tertiary care center in Chennai, India between January 1996 and April 2006 were collected (n=7069). Univariate and multivariate regression analyses were performed to examine associations between anemia and stage of HIV disease, co-morbidities, and medications. To examine the influence of HAART use on anemia on a smaller subset of patients (n=401), the mean of baseline hemoglobin values measured within 3 months prior to HAART initiation was compared to the mean of follow-up hemoglobin values collected between 3-12 months after HAART initiation. A similar analysis based on the time of AIDS onset was performed for a control group of patients with clinical/immunological AIDS who never received HAART (n=77). Results: The prevalence of anemia in the overall cohort was 40% by the WHO definition. While 22% of patients with CD4 counts >500 cells/μL had anemia, this increased to 60% for those with CD4 counts <100 cells/μL (p<0.001). In multivariate analysis, CD4 count <100 cells/μL, underweight body mass index, female sex, active tuberculosis, and lack of cotrimoxazole prophylaxis had significant associations with anemia. In the analysis of the influence of HAART on anemia, median baseline CD4 count and mean time to follow-up hemoglobin value were similar for both the HAART and control arms. For patients who initiated HAART, the mean baseline hemoglobin of 10.7g/dL significantly increased to 12.4g/dL during the year of follow-up (p<0.001). While 46.6% of patients were not anemic at HAART initiation, this increased significantly to 78.1% during follow-up (p<0.001). Severe grades 3 and 4 anemia decreased from 12.5% to 2% for those on HAART (p<0.001). For the control arm of patients who never received HAART, the increase in mean baseline hemoglobin from 10.50 to 11.10g/dL did not reach significance (p=0.06). While the percent of non-anemic patients increased from 40.3 to 54.5% on follow-up in this group, this was not significant (p=0.291). Conclusion: Anemia is strongly correlated with the severity of immunosuppression in this population seen at an HIV tertiary referral center. Endemic malnutrition and tuberculosis further exacerbate the level of anemia. Anemia resolved in a large proportion of patients within the first year of HAART use. Therefore, antiretroviral therapy, nutritional supplementation, and aggressive tuberculosis treatment should be the cornerstones of anemia management in this setting. The high prevalence of anemia among patients with immunological AIDS complicates the roll-out of antiretroviral regimens containing zidovudine, a drug which may exacerbate anemia, and highlights the need for increased access to alternative nucleoside reverse transcriptase inhibitors in developing countries.