Background:We previously reported that single measurements of albumin strongly predict survival in HIV-1-infected women independent of disease-specific markers. We now extend this to the use of serial measurements and single albumin values prior to initiation of highly active antiretroviral therapy.
Design:Prospective cohort study of 1941 women enrolled at six sites in the Women's Interagency HIV Study.
Results:Albumin fell 0.44 g/L/y in 1627 women who survived and at a faster rate in 397 who died (1.54 g/L/y; p < .01). In a time-dependent model adjusting for disease markers, the relative hazard (RH) was fivefold higher in patients with serum albumin <35 g/L compared with patients with serum albumin >42 g/L. The RH of serum albumin <35 g/L in women with CD4+ lymphocyte counts ≥200 cells/μL was 8.2 [95% CI: 4.2-15.8]) versus only 3.8 [95% CI: 2.4-6.1] in those with counts <200 cells/mm3. In a fixed-covariate Cox analysis of patients who started HAART during the study, albumin prior to HAART was associated with a higher RH (7.0 for albumin <35 g/L versus >42 g/L) than were other factors.
Conclusion:Serum albumin is a strong independent predictor of mortality in HIV-1-infected women after adjustment for known disease markers and may be useful for clinical monitoring.
The WIHS is funded by the National Institute of Allergy and Infectious Diseases and the National Institute of Child Health and Human Development, with supplemental funding from the National Cancer Institute, the National Institute on Drug Abuse, and the National Institute of Dental Research (U01-AI-35004, U01-AI-31834, U01-AI-34994, U01-AI-34989, U01-HD-32632, U01-AI-34993, U01-AI-42590, N01-AI-35161).
Address correspondence and reprint requests to J.G. Feldman, Department of Preventive Medicine and Community, SUNY Health Science Center at Brooklyn, 450 Clarkson Avenue, Brooklyn, NY 11203, U.S.A.; e-mail: Joseph.Feldman@Downstate.edu
Manuscript received September 4, 2002; accepted December 2, 2002.
© 2003 Lippincott Williams & Wilkins, Inc.