The demographic and laboratory characteristics of 80 HIV-infected patients and 80 age- and sex-matched controls are shown in Table 1. In both groups, the mean age of the patients was 39 ± 8 years, and all were male. The two groups did not differ in terms of smoking (p = 0.08), alcohol consumption (p = 1.0), or presence of diabetes mellitus (p = 0.24) or hypertension (p = 0.12). BMI was significantly lower in HIV-infected patients than in the control group (22.9 ± 3.0 vs. 24.9 ± 3.3, p < 0.001). HIV-infected patients had a greater history of previous fractures (12.5% [n=10] vs. 1% [n=1], p = 0.009) and presence of dyslipidemia (10.0% [n=10] vs. 0%, p = 0.028).
Among HIV-infected patients, the current mean CD4 T-cell count was 669 ± 271 cells/µl and the nadir CD4 T-cell count was 349 ± 183 cells/µl. Seventy-eight patients (97.5%) were receiving ART, and 93.7% had plasma HIV-1 RNA <40 copies/ml. The mean duration of ART was 677.9 ± 411.2 days. The ART regimens used were non-nucleoside reverse transcriptase inhibitor (NNRTI)-based (n = 28, 35.8%), protease inhibitor (PI)-based (n = 29, 37.1%), or integrase inhibitor-based (n = 21, 26.9%). Forty-one (51.3%) patients had had exposure to tenofovir disoproxil fumarate (TDF).
TBS and BMD results
In the HIV group, four (5.0%) patients had an intermediate TBS fracture risk and two (2.5%) patients had a high TBS fracture risk; in the matched control group, seven (8.8%) subjects had an intermediate TBS fracture risk. The fracture risk assessed via TBS was not different between the HIV and control groups (p = 0.244). However, the mean TBS value in HIV-infected patients was 1.41 ± 0.07, which was significantly lower than the mean value for the matched control group (1.45 ± 0.07, p = 0.008) (Fig 1A).
The proportion of subjects with low BMD was significantly higher in the HIV group than in the control group (21.3% [17/80] vs. 8.8% [7/80], p = 0.027). The proportion of subjects with low BMD at the lumbar spine was significantly different between the two groups (21.3% [17/80] vs. 6.3% [5/80], p = 0.006), but the proportion with low BMD at either the femoral neck or total hip did not differ between the groups (0% [0/80] vs. 2.5% [2/80], p = 0.155). Figure 1B shows the BMD values for the lumbar spine, femoral neck, and total hip for the HIV group and the matched control group. The mean femoral neck BMD values (0.95 ± 0.14 vs. 1.02 ± 0.14, p = 0.009) and total hip BMD values (0.98 ± 0.13 vs. 1.06 ± 0.14, p < 0.001) were significantly lower in the HIV group than in the matched control group; however, BMD values for the lumbar spine did not differ between the two groups (1.16 ± 0.16 vs. 1.19 ± 0.18, p = 0.214).
There was discordance between the BMD and TBS results in 11 patients in the HIV group and 13 subjects in the control group; these proportions were not significantly different (p = 0.66).
In both groups, TBS values were positively correlated with BMD at the lumbar spine, femoral neck, and total hip. However, TBS values were not correlated with BMI in either group (Table 2). In the HIV group, TBS values were negatively correlated with the duration of tenofovir exposure (p = 0.04) and trended toward a negative correlation with the duration of HIV diagnosis (p = 0.07) (Table 2, Figure S1).
Laboratory tests of bone metabolism in HIV-infected patients
In the group of 80 HIV-infected male patients, the mean 25[OH]D level was 21.2 ± 8.08 ng/ml, with 45% of patients (n = 36) measuring below 20 ng/ml. The quartile distribution of 25[OH]D values in patients was as follows: <10 ng/ml (2.6%), 10–19.99 (43.6%), 20–29.99 (44.8%), 30–39.99 (3.8%), ≥40 (5.1%). Mean calcium, phosphorus, and alkaline phosphatase levels were 9.14 ± 0.32 mg/dl, 3.33 ± 0.51 mg/dl, and 64.7 ± 17.95 U/l, respectively. Mean osteocalcin and CTX levels were 18.91 ± 8.4 µg/ml and 0.39 ± 0.22 µg/ml, respectively.
Relationship between clinical characteristics and TBS in HIV-infected patients
Table 3 shows the demographic, laboratory, and disease-related parameters in HIV-infected patients with normal TBS values and HIV-infected patients with low TBS values. BMI, current smoking status, and presence of diabetes mellitus did not differ between the normal-TBS and low-TBS groups. Serum creatinine levels were significantly lower in HIV-infected patients with low TBS (p = 0.003); however, calcium (p = 0.21), phosphorus (p = 0.19), alkaline phosphatase (p = 0.38), 25[OH]D (p = 0.80), osteocalcin (p = 0.41), and CTX (p = 0.08) levels did not differ between the groups. Duration since HIV diagnosis, nadir CD4 T-cell count, TDF exposure, duration of TDF treatment, duration of ART, and ART regimen type did not differ between the normal-TBS and low-TBS groups.