Testosterone (T) and estradiol (E2) are poorly associated to the reduction of bone mineral density (BMD) in Young/Middle Aged Men with Human immunodeficiency virus (HIV).

Background: Osteopenia and osteoporosis, as well as hypogonadism, are common findings in men with HIV- infection and they occur at a younger age than healthy subjects. The reduction of BMD is due to both HIV-related and HIV-unrelated factors. Previous studies suggest that T deficiency is not or poorly associated with reduced BMD in HIV context. On the other hand, estrogens are consid- ered more important than androgens for bone health in general population, but data about their role in HIV- infected men are still scanty. Objective: To investigate the relationship between BMD and circulating sex steroids assessed by Liquid Chro- matography tandem Mass Spectrometry (LC-MS/MS) in a cohort of young/middle aged HIV-infected men. Methods: Prospective, cross-sectional, observational study on 233 consecutive HIV-infected male patients with ongo- ing Highly Active Antiretroviral Therapy (HAART), attend- ing the Multidisciplinary Metabolic Clinic of Modena. Body composition and BMD at total body, lumbar spine (L1 to L4) and total hip were measured using a Hologic QDR-2000 densitometer (DXA). LC-MS/MS was used for hormonal assays. Statistical analysis: The nonparametric Mann–Whitney U test was used for group comparisons because variables were not normally distributed at the Kolmogorov-Smirnov test. Correlations were performed using linear regression models. Results: Two hundred and thirty-three HIV-infected patients were enrolled (mean age 45.29 ! 5.33 years) with average duration of HIV-infection of 190.8 ! 102.8 months. Eight patients (3.4%) had hypogonadism, defined as total T serum levels below 300 ng/dL. Considering results at DXA examination, BMD was normal in 36.5% and reduced in 63.5% (55.8% osteopenia, 7.7% osteoporo- sis). Both total T and E2 did not significantly differ com- paring patients with normal BMD to patients with reduced BMD. Body and lumbar BMD did not show any significant difference between eugonadal patients and patients with low T and/or low E2, while both femoral BMD and femoral T-score were significantly higher in patients with E2 above 20 pg/mL than in those with E2 below 20 pg/mL (p = 0.043 and p = 0.033, respectively). At linear and step- wise multiple regression analyses, BMD was positively associated with total lean mass (R2 = 0.154, p < 0.0001); apart from it, neither T nor E2 correlated with BMD and T- score at any site. Conclusion: Classical factors associated to BMD as E2 and T seem to be less relevant in this model of male osteoporo- sis. Other specific HIV-related factors, such as changes in body composition and consequent lipodystrophy, could be more deeply involved than sex steroids as potential mecha- nisms in bone loss in this setting. Finally, we confirm the high prevalence of reduced BMD in young/middle aged HIV-infected men, representing one of the clinical hallmarks of the premature aging process related to HIV infection.