Session 147-Poster Abstracts
Low Bone Density: Etiology, Associations, and Role of ART
Thursday, 2-4 pm; Poster Hall
Paper #875
The Association between Body Composition, Soluble and Cellular Immune Activation Biomarkers, and BMD in ART-naïve HIV-1+ Persons: Baseline Results of ACTG Study A5260s
Todd Brown*1, J Currier2, Y Chen3, H Ribaudo3, J Touw4, J Rothenberg4, M Dube5, R Murphy6, J Stein7, and G McComsey8
1Johns Hopkins Univ, Baltimore, MD, US; 2Univ of California, Los Angeles, David Geffen Sch of Med, US; 3Harvard Sch of Publ Hlth, Boston, MA, US; 4Social & Sci Systems, Inc, Silver Spring, MD, US; 5Univ of Southern California, Keck Sch of Med, Los Angeles, US; 6Northwestern Univ, Chicago, IL, US; 7Univ of Wisconsin Sch of Med and Publ Hlth, Madison, US; and 8Case Western Reserve Univ Sch of Med, Cleveland, OH, US

Background:  In untreated HIV+ persons, immune activation and alterations in body composition have been hypothesized to contribute to low bone mineral density (BMD). We performed a cross-sectional analysis to examine associations with BMD in ART-naïve persons enrolled into a randomized clinical trial (ACTG A5257).

Methods:  Subjects underwent dual energy x-ray absorptiometry (DXA), single-slice abdominal computed tomography (CT), and laboratory assessment for metabolic and immune activation markers and hepatitis C serostatus. DXA-measured BMD at the lumbar spine, total hip, and femoral neck were the outcomes of interest, expressed as a z-score (number of standard deviations away from an age-, race-, sex-matched reference population). Total lean mass and fat mass were measured by DXA, and visceral adipose tissue (VAT) by CT. Soluble biomarkers included adipocytokines (leptin, adiponectin), inflammatory markers (hsCRP, IL-6), and markers related to bone metabolism (osteoprotegerin), receptor activator of NF-κB ligand (RANKL)). Percentage CD8+CD38+DR+ was the primary cellular immune activation marker of interest. Univariate associations related to demographics, HIV disease, body composition, and biomarkers with z-score at each of the 3 sites were evaluated with non-parametric k-sample tests. Adjusted analyses used linear regression.

Results:  The 331 subjects had a median (Q1, Q3) age of 36 (28 to 45) years; 89% male; and 44% white. The median CD4 cell count was 349 (207,455)/mm3 and HIV-1 RNA was 4.5 (4.0 to 5.1) log10 copies/mL. The median z-scores were –0.4 (–1.2 to 0.4) at the lumbar spine, –0.1 (–0.6 to 0.6) at the total hip, and –0.1 (–0.7 to 0.5) at the femoral neck. In univariate analyses, there were no associations between z-scores and smoking, alcohol use, hepatitis C serostatus, CD4 cell count, HIV-1 RNA, hsCRP, Il-6, RANKL, or percentage CD8+CD38+DR+ (p ≥0.1). In a linear model adjusting for age, gender, race, and total fat mass, lower lumbar spine z-scores were associated with lower total lean mass (p = 0.002), higher serum adiponectin (p = 0.08), and  lower osteoprotegerin (p = 0.003); there was no independent association with VAT or leptin, (p ≥0.19). Analyses of the total hip or femoral neck showed similar results.

Conclusions:  Among ART-naïve HIV+ persons, lower lumbar spine, femoral neck, and total hip BMD were associated with lower lean mass, higher adiponectin, and lower osteoprotegerin, but not HIV disease variables or any of the inflammation or immune activation markers.