Table 1.
Prevalence of low bone mineral density, osteopenia and osteoporosis in HIV-infected individuals in low–middle-income countries versus high-income countries
| Prevalence* of | |||||||||
| Country (region) | Reference | Type of study | Patients | % Women | Age, mean or median (SD or IQR) | Low BMD | Osteopenia | Osteoporosis | Key findings or Remarks (DXA machine) |
| Low–middle-income countries (LMICs) | |||||||||
| Uganda (SSA) | Wandera et al., 21st CROI (EARNEST trial) [45▪] | Cross-sectional analysis of prospective study, 2010–2014 | 181 HIV-infected patients failing their first line ART (duration of first line ART 3.7 (2.7–5) years, 16.7% had used TDF, CD4 67 cells/μl (35–151) | 69 (12.8% postmenopausal) | 35 (31, 41) | LS 50.9%, TH 24.8% | LS 42.9%, TH 23.7% | LS 8%, TH 1.1% | Low BMD at LS was associated with both low BMI and use of TDF in first-line regimen. At TH, a low BMI was predictive of low BMD. (Discovery Hologic.) |
| India (SA) | Dravid et al. [46] | Cross-sectional study, June to December 2013 | 536 patients: | 34 | 42 | Age, low BMI, current smoking, and menopause were associated with low BMD. Choice of ART use (TDF vs. non-TDF, PI vs. no PI) did not influence loss of BMD. (Lunar Prodigy.) | |||
| 496 HIV-infected patients on ART | LS or TH 80.4% | LS or TH 51% | LS or TH 29.4% | ||||||
| 40 HIV-infected ART-naïve patients | LS or TH 67% | LS or TH 47.2% | LS or TH 19.8% | ||||||
| Nigeria (SSA) | Alonge et al. [47▪] | Cross-sectional study, September to December 2010 | 1005 HIV-positive patients (78.1% on ART; PI 12.6%), median CD4 371 cells/μl, median VL 200 copies/ml | 72 | 41.3 ± 10 | n/a | Lt. distal radius 46.6% | Lt. distal radius 31.9% | Osteoporosis was higher in those aged >40 years, women, single, and underweight. There was no difference in BMD of those with or without PI containing medications and treatment-naïve patients. (Specific DXA machine was not indicated.) |
| Indonesia (EA/P) | Masyeni et al. [48] | Cross-sectional study, January to June 2012 | 73 HIV-positive ART-naïve patients (mean CD4 144.7 cells/μl and mean VL 272.3 copies/ml | 32.9 | 33.1 ± 8.3 | LS or FN 43.8% | LS or FN 35.6% | LS or FN 8.2% | Low BMD was correlated with HIV stage (r = 0.337; P < 0.001) (Lunar DPX.) |
| South Africa (SSA) | Hamill et al. [49] | Cross-sectional study, February 2010 to July 2010 | 75 HIV-infected ART-naïve patients with low CD4 counts (200 cells/μl) | 100 | 33.4 ± 6.5 | No significant differences in BMD at LS, TH, or FN. Proportions with osteopenia or osteoporosis were not reported. | HIV-positive women did not have lower BMD compared to HIV-negative controls, despite the pre-ARV group being lighter with lower BMI. (Hologic QDR 4500A.) | ||
| 74 HIV-positive ART-naïve with relatively preserved CD4 cell counts (>350 cells/μl) | 100 | 33.5 ± 6.1 | |||||||
| 98 healthy controls | 100 | 30.0 ± 8.1 | |||||||
| Turkey (EU/CA) | Aydin et al. [50] | Cross-sectional study, June 2010 to May 2011 | 126 HIV-infected patients (63.5% on ART; either AZT/3TC or TDF/FTC, LPV/r or EFV), mean CD4 313.8 cells/μl | 16 | 40.1 ± 11.3 | LS or TH 77.7% | LS or TH 53.9% | LS or TH 23.8% | Neither NNRTI nor PI containing regimens was associated with low BMD. High VL, using and duration of ART were associated with bone loss. (Norland) |
| China (EA/P) | Zhang et al. [51] | Prospective study, April 2007 to March 2011 | 40 HIV-infected ART-naïve patients | 12.5 | 37.3 ± 9.9 | At baseline, LS BMD of HIV-infected patients was lower than controls (1.195 ± 0.139) vs (1.138 ± 0.112). Proportions with osteopenia or osteoporosis were not reported. | With ART initiation, LS, FN, and TH BMD reduced significantly in HIV-infected patients (annual percentage decline 1.78–3.28%). (Lunar Prodigy Advance PA + 300388.) | ||
| 40 healthy controls | 14.3 | 37.2 ± 10.3 | |||||||
| China (EA/P) | Wang et al. [52] | Cross-sectional study, January 2010 to May 2014 | 21 ART-naïve patients with acute HIV infection (mean CD4 420 ± 152 cells/μl, MSM 95.2%) | 0 | 31.1 ± 6.9 | LS or TH 33.3% | LS or TH 33.3% | LS or TH 0% | TH and FN BMD in patients with chronic HIV infection were lower than the other 2 groups. HIV infection, older age, lower BMI and MSM were associated with low BMD. (MEDI LINK Osteocore.) |
| 55 ART-naïve patients with chronic HIV infection (mean CD4 286 ± 168 cells/μl, MSM 81.7%) | 0 | 31.6 ± 5.9 | LS or TH 63.4% | LS or TH 56.3% | LS or TH 7.1% | ||||
| 71 healthy controls | 0 | 33.7 ± 5.7 | LS or TH 45.1% | LS or TH 38% | LS or TH 7.1% | ||||
| Israel (ME/NA) (Ethiopian and Caucasian origin) | Shahar et al. [53] | Cross-sectional study, Summer 2009 | 43 HIV-infected Ethiopians (mean CD4 233 cells/μl, 82% on ART, 20 study participants on PI) | 100 | 35.9 ± 8.2 | LS 85%, TH 55%, FN 65% | n/a | n/a | Low BMD in HIV-infected patients was associated with duration of HIV infection and ART use. (Lexxos, France.) |
| 32 HIV-infected Caucasians (mean CD4 264 cells/μl, 64% on ART, 21 study participants on PI) | 100 | 34.8 ± 8.7 | LS 40%, TH 13.3%, FN 39.3% | ||||||
| Thailand (EA/P) | Wattanachanya et al. [54] | Cross-sectional analysis of prospective study, 2010–2011 | 220 HIV-positive ART-naïve patients (mean CD4 348 cells/μl) | 46.8 | Male 39 ± 6; female 39 ± 4 | n/a | LS, TH or FN 20.9% (men), 23% (women) | LS, TH or FN 0% (men), 2.3% (women) | No difference in BMD was found between HIV-positive patients and controls (Hologic QDR4500.) |
| 233 healthy controls | 52.4 | Male 40 ± 6; female 41 ± 5 | n/a | 16.9% (men), 26.7% (women) | 0.7% (men), 0.8% (women) | ||||
| South Africa, India, Thailand, Malaysia, Argentina | Martin et al. [12]; Haskelberget al. (second-line trial) [55] | Prospective study, 2010–2011 | 210 HIV-infected patients failing their first line ART with current CD4 202 (104–307) cells/μl; median duration of ART use 3.4 years, AZT 34%, d4T 48%, TDF 17% at baseline | 52 | 38.8 (32.9–44.2) | n/a | LS 31.3%, TH 19.7% (at baseline) | LS 5%, TH 1.5% (at baseline) | Reduced BMD was associated with longer duration of TDF and low BMI. An NRTI-sparing ARV regimen of LPV/r and raltegravir is associated with less bone loss than a LPV/r regimen containing NRTIs. BMD decrease was greatest at 48 weeks with stabilization to week 96, but no recovery. (Lunar-India, Malaysia, Argentina, Thailand or Hologic-Thailand, South Africa.) |
| High-income countries (HICs) | |||||||||
| USA1 (NA) | Battalora et al. (HOPS and SUN study) [56▪] | Cross-sectional analysis of prospective study, 2004–2012 | 1006 HIV-infected patients with median CD4+ 461 cells/μl (96.6% on ART, 67% non-Hispanic White) | 17 | 43 (36, 49) | FN 40% | FN 36% | FN 4% | During 4,068 person-years of observation, 85 incident fractures occurred, predominantly rib/sternum, hand, foot, and wrist. Osteoporosis and current/prior tobacco use were associated with incident fracture. (Lunar or Hologic, reference standard: NHANES III database.) |
| UKa (EU/CA) | Short et al. [57] | Cross-sectional study, May to August 2008 | 168 HIV-infected patients (63% on ART; PI 27%, NNRTI 45%, NRTI included TDF 47%) | 0 | 45 (38, 51) | n/a | LS, TH, or FN 58% | LS, TH, or FN 12% | Number of fractures since HIV diagnosis was increased among those with osteoporosis. Duration of infection >13 years was associated with osteoporosis. (Hologic QDR4500C.) |
| Italy (EU/CA) | Mazzotta et al. [58] | Cross-sectional study, April 2009 to March 2011 | 163 HIV-infected patients (79.7% on ART; PIs 59.2%, TDF 70%) | 29.4 | 44.2 ± 10 | LS or TH; 63.2% | LS or TH; 49.7% | LS or TH; 13.5%; LS or TH 19.6% (Z-score ≤ −2) | Low BMD was associated with lower BMI, AIDS diagnosis, HCV co-infection, ART, and nontraumatic fractures (NTBFs). Prevalence of NTBFs was 27.0%, predictors; male sex, HCV co-infection, lower FN Z-scores. (Hologic QDR 4500A.) |
| The Nether-lands (EU/CA) | Kooij et al. [59] | Cross-sectional study, 2010–2012 | 581 HIV-positive patients (94.7% on ART; NRTI/TDF 96.4/77.1%, PI 43.6%, NNRTI/NVP 60.4/ 30.2%) | 11.5 | 52.7 (48.3, 59.4) | n/a | LS 34%, FN 43%, TH 29% | LS 11%, FN 4%, TH 2% | Low BW was negatively associated with BMD in HIV-positive persons. Regardless of HIV status, younger MSM had lower BMD than older MSM, heterosexual men, and women. (Hologic QDR 4500 W, the reference standard: NHANES database.) |
| 520 HIV-negative controls | 15.2 | 52.0 (47.9, 58.0) | n/a | LS 35%, FN 34%, TH 16% | LS 6%, FN 1%, TH 0% | ||||
| Japan (EA/P) | Kinai et al. [60] | Cross-sectional study, February 2012 to June 2013 | 184 HIV-infected men (93% on ART; PIs 64%, TDF 62%), median CD4 493 cells/μl | 0 | 43 (38, 51) | n/a | LS 46% FN 54% | LS 10% FN 12% | Low BMD was associated with long-term treatment with a PI and a low BMI. Patients who discontinued PI had a higher BMD than those who currently use PI at LS but not at FN. (Hologic QDR 4500 W.) |
| USA2b (NA) | Overton et al. [61] | Cross-sectional analysis of prospective study, September 2011 to February 2012 | 165 HIV-infected patients on EFV/FTC/TDF regimen (33% non-Hispanic Black; median CD4 341 cells/μl) | 9.7 | At baseline | Authors evaluated vitamin D3 (4000 IU daily) plus calcium (1000 mg calcium carbonate daily) supplementation on bone loss associated with ART initiation. BMD loss in the first year after ART initiation may be minimized by calcium and vitamin D supplementation. (Specific DXA machine was not indicated.) | |||
| Pre vitamin D/Calcium group (n = 79) | 36 (28, 47) | LS 9%; TH 5% | n/a | n/a | |||||
| Placebo group (n = 86) | 31 (25, 44) | LS 10%; TH 6% | |||||||
| USA3c (NA) | Cotter et al. [62] | Cross-sectional analysis of prospective study February 2011 to July 2012 | 210 HIV-positive patients (40% African) | 41 | 39 (33, 46) | At baseline; LS 24.3%; TH 13.8%; FN 23.8% | n/a | n/a | HIV was independently associated with lower BMD at femoral neck, total hip and lumbar spine. Lunar Prodigy DXA (GE Medical Systems, Madison, Wisconsin, USA.) |
| 264 HIV-negative controls (25% African) | 56 | 42 (34, 49) | LS 12.5%; TH 5.7%; FN 11.7% | n/a | n/a | ||||
| Mixed LMICs and HICs | |||||||||
| Australia, Belgium, Brazil, India, Ireland, Peru, South Africa, Spain, Thailand, UK, US | Carr et al. (STARTBone Mineral Density Substudy) [63] | Cross-sectional analysis of prospective study, June 2011 to June 2013 | 424 ART-naïve participants with mean CD4 688 ± 152 cells/μl | 26 | 34 ± 10.1 | LS, TH or FN 35.1%, FN 18.8% | n/a | LS, TH or FN 1.9%, FN 0.5% | Lower BMD was associated with female sex, Latin/ Hispanic ethnicity, lower BMI and higher estimated GFR. Longer time since HIV diagnosis was associated with lower TH BMD, but not with CD4 cell count or viral load. [Lunar or Hologic, reference standard: NHANES III database (hip) and Hologic's reference data (spine).] |
Data shown in the table include published articles and abstracts related to prevalence of low BMD in HIV-infected adolescents or adults from RLS in 2014 and 2015 plus articles of special interest from 2013. However, for RRS, only the articles published in mid-2014 to 2015 and had more than 100 HIV-infected participants were included. BMD, bone mineral density; FN, femoral neck; HICs, high-income countries; LMICs, low–middle-income countries; LS, lumbar spine; n/a, not available; TH, total hip; VL, viral load.
*BMD was assessed by DXA either central or peripheral sites. In most studies, low BMD, osteopenia, and osteoporosis were defined as T-score < −1, T-score between −2.5 and −1, and T-score less than −2.5, respectively.
aIn this study, low BMD was defined as Z-score < −2 at LS or TH.
bIn this study, osteopenia and osteoporosis were defined as T- or Z-score < −1 and T- or Z-score < −2.5, respectively.
cIn this study, low BMD was defined as T-score <1 in those older than 40 years or Z <2.0 in those younger than 40 years, respectively.