HIV positive people in the HOPS cohort have 4-fold risk of fracture compared to general population in the US

Simon Collins, HIV i-Base

Over the last ten years HTB has reported numerous studies of lower bone mineral density and higher rates of osteopenia and osteoporosis in HIV positive people compared to rates in age- and gender-matched general population.

While some research groups cautioned that their findings might not translate into higher fracture rates, most others suggested that ageing was likely to compound this risk and that it would be only a matter of time before the additional impact of HIV might be seen.

Two studies at CROI unfortunately suggest that these concerns are likely to be real.

Christine Dao from the US CDC in Atlanta and colleagues presented an analysis of fracture rates from 1994-2008 in over 8,400 HIV-positive patients followed in the HIV Outpatients (HOPS) Cohort and compared this to rates from non-HIV US inpatient (NHDS) and outpatient (NHAMCS) surveys that were weighted to make inferences to the US general population.

Only first fractures were included from HOPS and adjusted Hazard Ratios (AHR) were calculated controlling for age and gender. As fracture data were not comprehensively collected in HOPS prior to 2000, the presentation focused on fractures from around 5800 patients seen at least annually over the last eight years.

Baseline characteristics at first visit included: 79% male; 52% non-Hispanic white, 38% non-Hispanic Black; median (IQR) age 40 years (34-46), median BMI 24.4 (22.3-27.4); median time since diagnosis 5.3 years (1.3-9.9). Three-quarters of patients were treatment-experienced with median viral load (for the cohort) of 1,300 (<400-35,560) copies/mL.

Approximately 4% patients (236/5826) experienced a fracture at median age 45 years (38-51), roughly in proportion to baseline characteristic of race and gender, although only 51% of fractures were in treatment-experienced patients.

Gender-adjusted rates were restricted to patients aged 25-54, representing most HIV-positive individuals and showed not only an increase in fracture rates over time from about 55 to 85/10,000 PY, p=0.01 (compared to stable rates at around 30/10,000 PY in general population outpatient data). The rate in HIV-positive people was stable from 2002-2008 (and was approximately 4-fold higher).

Fractures at fragility sites (wrist, vertebra, femoral head) occurred at higher rate in both HIV-positive men and women compared to the general population, and are detailed in Table 1.

Factors independently associated with increased risk of fracture (adjusted HR: 95%CI) included: age >47 years (1.6; 1.0–2.5, p<0.05); nadir CD4 <200 cells/mm3 (1.6; 1.1–2.3, p=0.04); HCV coinfection (1.6; 1.1–2.3, p=0.01); diabetes (1.6; 1.0–2.6, p<0.05); and history of substance use (1.5; 1.0–2.3, p<0.05).

Limitations of the study included the use of different data collection systems for the HIV-positive and general population groups, no data linking bone mineral density to fracture risk, and whether this increase was due to an increase in events or a potential improved reporting.

Table 1: Percentage of fractures by anatomic site in adults 25-54 years old (HOPS 2000-2008)

Men n=146 1,705,433
Wrist 9% 3% <0.05
Vertebra 10% 1% <0.05
Femoral head 3% 2% NS
Non-fragility site 78% 94% <0.05
Women n=45 n=1,136,788
Wrist 4% 8% NS
Vertebra 18% 4% <0.05
Femoral head 7% 1% <0.05
Non-fragility site 71% 86% <0.05

Importantly, they also stressed that the actual event rate remained low, even though
the relative rate was significantly higher.

Nevertheless, the researchers concluded that HOPS patients experienced higher rates of fracture compared to the general US population, that this rate increased over time and included a higher percentage of fragility fractures; and that in addition to known risk factors, low CD4 nadir was also associated with increased fracture risk.

Julie Womack and colleagues from the Veterans Ageing Cohort Study (VACS) presented results from men in the largely male VA cohort, focusing on the prevalence and incidence of fragility fractures (ie associated with minimal or no trauma, and with low bone mineral density). [2]

The VACS is a prospective observational cohort of about 40,000 HIV-positive veterans matched 1:2 by age, sex and site to around 80,000 HIV-negative controls. This analysis only included first fracture. Wrist fractures and compound fractures were excluded because of the higher likelihood of being related to trauma.

Multivariable models were adjusted for HIV status, race, enrolment date, age, coinfection, BMI and corticosteroid use, with additional adjustment for baseline CD4 and ARV use in HIV-positive patients.

During a median follow-up of 8 years (range 4-11), 952 fractures (644 hip and 308 vertebral) fractures were recorded, with an unadjusted incidence of 16 vs 11/10,000 PYFU in the HIV-positive vs HIV-negative groups. Fractures occurred at a mean age of 55 years (SD+11).

Prevalence results showed that across all ages only hip fractures occurred at a significantly higher rate in the HIV-positive vs HIV-negative groups (p<0.0001), with a difference developing and widening from age 40 onwards. Although vertebra fractures were generally similar in both groups, the rate in HIV-positive men increased significantly in men over 70 years who were HIV-positive.

After adjustment for cofactors (AHR; 95%CI), including Caucasian race (1.79; 1.57–2.03), BMI <19 (2.50; 1.54–4.05), alcohol use (H1.79; 1.47–2.18), pulmonary disease (1.38; 1.10–1.73), cerebrovascular disease (2.16; 1.54–3.02), and peripheral vascular disease (1.64; 1.10–2.44), the HIV effect was reduced, though still significant (1.38; 1.18–1.60). Similar ratios applied for the full model and HIV-positive only model.

The presentation also discussed management and this was picked up in the question session afterwards. Prevention is stressed for all patients. Clinical assessment for fracture risk was recommended for patients aged 40-50 and DEXA indicated when HIV is not the only risk factor.

Several members of the audience commented that they would encourage broader use of DEXA scans, especially given the high rates of other risk factors, including higher rates of smoking, alcohol use and low testosterone.

Although ARV use was not found to be significant in this study, another question from the audience suggested that this was unlikely to be a reliable conclusion, as the study defined exposure by baseline use of d4T, tenofovir, PIs or NNRTIs, rather than looking at cumulative exposure more commonly adopted in most studies.

Finally, a third presentation in the same oral session reported fracture incidence in a retrospective analysis in about 2400 women (approximately 1700 HIV positive, 700 HIV negative) in the Women’s Interagency HIV Study (WIHS). Of note, this study also collected data on smoking, opiate/cocaine use and vitamin D/calcium supplementation.

Fractures occurred in 148 (9%) HIV-positive women vs 47 (7%) HIV-negative women producing incidence rates of 1.79 vs 1.41/100 PY, respectively (p=0.13 NS). Analysis by fracture site also showed no significant difference in rates by HIV status. In the multivariate analysis, HIV status remained non-significant, with only age, race (being Caucasian), HCV coinfection and serum creatinine only showing positive relationship to increased fracture
rate. The only determinants of time to fracture in HIV-positive women were age (per 10 years older HR 1.2; 95%CI 1.0–1.5, p=0.047) and previous AIDS defining illness (HR 1.9; 95%CI 1.3–2.7, p=0.0008), but not CD4 count or ARV use.

Although no impact of HIV was seen over five years of follow-up the limitations of this study was that this was broadly a pre-menopausal patient group and in a modest sample size. Also, approximately 50% women had high BMI which is associated with protecting bone density and strength.

The conclusion included recognition that fracture risk could increase in HIV-positive post-menopausal women, given that oestrogen has a protective effect and other studies have already highlighted the lower levels of bone mineral density in HIV-positive women.


Unless stated otherwise, all references are to the Programme and Abstracts of the 17th Conference on Retroviruses and Opportunistic Infections. 16-10 February 2010, San Francisco. All oral abstracts are available as webcasts.

  1. Dao C et al. Higher and increasing rates of fracture among HIV-infected persons in the HIV Outpatient Study compared to the general US population, 1994 to 2008. 17th CROI 2010. Oral abstract 128.
  2. Womack J et al. HIV-infection and fragility fracture risk among male veterans. 17th CROI 2010. Oral abstract 129.
  3. Yin M et al. Fracture rates are not increased in younger HIV+women. 17th CROI 2010. Oral abstract 130.

Links to other websites are current at date of posting but not maintained.