Prevalence of and risk factors for low bone mineral density in untreated HIV infection: a substudy of the INSIGHT Strategic Timing of AntiRetroviral Treatment (START) trial

Abstract

Objectives: HIV infection is associated with a higher prevalence of low bone mineral density (BMD) and fractures than that found in the general population. There are limited data in HIV-positive adults, naïve to antiretroviral therapy (ART), with which to estimate the relative contribution of untreated HIV infection to bone loss.

Methods: The primary objective of the Strategic Timing of AntiRetroviral Treatment (START) Bone Mineral Density Substudy is to compare the effect of immediate versus deferred initial ART on bone. We evaluated traditional, demographic, HIV-related and immunological factors for their associations with baseline hip and lumbar spine BMD, measured by dual-energy X-ray absorptiometry, using multiple regression.

Results: A total of 424 ART-naïve participants were enrolled at 33 sites on six continents; the mean age was 34 years [standard deviation (SD) 10.1 years], 79.0% were nonwhite, 26.0% were women, and 12.5% had a body mass index (BMI) < 20 kg/m(2) . Mean (SD) Z-scores were -0.41 (0.94) at the spine and -0.36 (0.88) for total hip; 1.9% had osteoporosis and 35.1% had low BMD (hip or spine T-score < -1.0). Factors independently associated with lower BMD at the hip and spine were female sex, Latino/Hispanic ethnicity, lower BMI and higher estimated glomerular filtration rate. Longer time since HIV diagnosis was associated with lower hip BMD. Current or nadir CD4 cell count and HIV viral load were not associated with BMD.

Conclusions: In this geographically and racially diverse population of ART-naïve adults with normal CD4 cell counts, low BMD was common, but osteoporosis was rare. Lower BMD was significantly associated with traditional risk factors but not with CD4 cell count or viral load.

Trial registration: ClinicalTrials.gov NCT00867048.

Keywords: ART-naïve; HIV; antiretroviral therapy; bone mineral density.

© 2015 British HIV Association.

Figures

Figure 1

Associations of baseline factors with BMD measured at the lumbar spine (L1-L4), total hip and femoral neck. Estimated coefficients from multivariable linear regression models are presented with 95% confidence intervals; factors with p-values ≤ 0.05 are independently associated with BMD. Abbreviations: BMD=bone mineral density; BMI=body mass index; CI=confidence interval; coef.=regression coefficient; eGFR=estimated glomerular filtration rate; NSAID=nonsteroidal anti-inflammatory drug. Footnote: Multivariable models were determined by backwards variable selection, and factors with p-values >0.10 were excluded from the final models; the following factors were investigated but were not significant (excluded by the variable selection procedure or p>0.05): scanner type (Hologic or GE Lunar); season of enrolment; current smoking, liver disease, hepatitis B or C, alcohol or recreational drug use; use of vitamin D3, hormone replacement therapy, proton pump inhibitors; CD4 cell count, nadir CD4 cell count, CD8 cell count, log10 plasma HIV viral load; total cholesterol, haemoglobin, alkaline phosphatase, and proteinuria by urine dipstick. Age, sex, menopause status, race/ethnicity and geographical region were included a priori in all models. In addition to the displayed variables, the model for total hip BMD was adjusted for baseline CD4 and nadir CD4 cell counts (p=0.09 for each) and baseline eGFR (p=0.07).