HIV Independently, but Modestly, Increases Risk of Fragility Fractures, Study of U.S. Male Veterans Finds

A Discussion With Julie Womack, Ph.D.

Below is the transcript of a press conference held at CROI 2010 on Feb. 18, in which Julie Womack, Ph.D., of the VA Connecticut Health Care System, discusses a study she presented regarding the risk of fragility fractures among HIV-infected male U.S. veterans.

Julie Womack: Our research explored HIV infection and fragility fractures among male veterans.1 Decreased bone mineral density is more common among HIV-infected than among uninfected individuals. This becomes a clinical problem when it leads to fragility fractures. But we don't know whether fragility fractures, or fractures that occur with little to no trauma, are more common among HIV-infected than uninfected individuals.

Julie Womack, Ph.D.

Julie Womack, Ph.D.

The purpose of our study was twofold. First: to determine how common fragility fractures are among those with HIV infection. And secondly: to compare the rate of fracture over time among HIV-infected and uninfected male veterans.

Over 119,000 male veterans were included in this study, 34% of whom were HIV infected. Our subjects came from the Veterans Aging Cohort Study [VACS] Virtual Cohort, which is a prospective observational cohort of all individuals diagnosed with HIV infection, who are receiving care in the Veterans Health Administration, and a two-to-one sample of age, race and site-matched uninfected veterans.

Approximately 2% of our subjects had any of the three fractures of interest: wrist, hip or vertebral. These proportions were modestly different by HIV status, 2.3% among HIV-infected and 1.7% among uninfected men. Wrist fractures were less common among older men, and thus were likely related to trauma rather than to bone fragility. We thus excluded wrist fractures from the rest of our analyses.

We then compared the rates of hip or vertebral fractures over time. Because of the size of our cohort and the richness and variety of the data available, we were also able to control for many of the established risk factors for fragility fractures, in addition to HIV status. The risk for having a fracture that was associated with HIV infection was 1.53 in the unadjusted and 1.38 in the adjusted models. Very loosely, this means that when we looked at HIV alone as a risk factor, it increased the risk of fracture by approximately 50%.

However, when we looked at HIV in the context of other important risk factors, it increased the risk of fracture by approximately one third. This suggests that HIV infection adds a modest increased risk for fragility fracture, even when accounting for other established risk factors, including low body weight, stroke, white race, alcohol abuse or dependence, and older age. Of note, each of these factors was associated with a greater increase in the risk of fracture than was HIV status.

To summarize, fragility fracture remains a relatively uncommon event. HIV is a modest independent risk factor for fragility fractures. And other conditions including low body weight, stroke, white race, alcohol abuse or dependence, and older age, appear to confer greater risk for fragility fracture than HIV infection.

If someone is diagnosed with osteoporosis by bone scan or has already suffered a fragility fracture, treatment with one of the bisphosphonates has been shown to reduce fracture risk. But most people who go on to experience fragility fractures do not have osteoporosis based on bone scans. For these individuals, primary prevention is essential, including weight-bearing exercise, adequate vitamin D and calcium, reducing alcohol intake, smoking cessation and fall prevention. Thorough screening for established risk factors for fragility fractures is also important.

Reporter #1: You said that most people who go on to fragility fractures do not in fact have osteoporosis diagnosable by scan before, right? So the preventative measures that you were talking about, weight bearing, exercising and the like, who would that be aimed at? People with identifiable risk factors?

Julie Womack: Certainly anyone with an identifiable risk factor, one would target whatever that risk factor was, but we would generally recommend that anyone. Certainly, our study focused on men, so looking at men, probably in their 40s and 50s, beginning to focus on preventive measures amongst those men, everyone.

Reporter #2: I was wondering if Dr. McComsey [who also presented the findings of her study in this press conference] and Dr. Womack could speak about each other's bone mineral density results. Outright for clinicians, you both said it's important, but then Dr. Womack said that the rates were not that high -- of course this is male veterans, which is a different risk group than, for example, perimenopausal women. I just wondered if you could comment about the results of each other and the significance of bone mineral.

Grace McComsey: Sure. I did comment on the fractures in general being more common in HIV. There has been at least one study looking at that, so not specifically, fragility fracture, just any fractures were more common in HIV patients versus not HIV. And as you mentioned, my study, for example, 85% were males and the median age was 38 years. So these were young people. The fact that we don't see fractures yet on them, it doesn't mean that it won't happen. Fractures are like hypertension and stroke: If you have a decrease in bone density and you don't quite see fractures during the study, it doesn't mean it's not clinically relevant, because they will develop later on. It just -- all that my study means is bone mineral density went down. It puts them at higher risk of fractures.

Now we did look, if you were at my presentation, we did look at fractures during the study. It's a two- to four-year study, but we still saw about 5% of patients develop fractures during the study. There was no significant difference between the arms. The fractures were distributed equally among all the arms of the study.

So I do think, even though the increased risk that was reported of fragility fracture is modest -- it's not huge -- it is increased, even in this population that shouldn't be at high risk of a bone mineral density issue. I didn't hear about smoking. Probably all veterans smoke, so you can't really tease out the smoking, but that would be another important risk.

Julie Womack: We actually did not look at bone mineral density. We just looked at fragility fractures based on ICD-9 [International Statistical Classification of Diseases and Related Health Problems] codes, so that may be one of the explanations for the difference in the outcomes. We didn't look at smoking. The reason for that is that we don't have that information in the data set. However, in a subset of the patients for whom we do have the data, smoking history didn't really differ much by HIV status.

Reporter #3: Dr. Womack, did you not let the vets in that had osteoporosis? And did you see something by regimen that would have indicated something? Because you didn't mention that at all.

Julie Womack: Good question. We actually didn't look at diagnostic codes for osteoporosis. And as I said, we did not look at bone mineral density specifically. We defined our fractures, wrist, hip and vertebral, by ICD-9 codes. Among the HIV-infected men, we did specifically look at CD4 count, tenofovir [TDF, Viread] use, stavudine [d4T, Zerit] use, NNRTI [non-nucleoside reverse transcriptase inhibitor] and protease inhibitor use. And the only one of those covariates that was significantly associated with fractures was CD4 count. And there, the higher CD4 count, the lower the risk of fracture, but it was very modest.

This transcript has been lightly edited for clarity.


  1. Womack J, Goulet J, Gibert C, et al. HIV-infection and fragility fracture risk among male veterans. In: Program and abstracts of the 17th Conference on Retroviruses and Opportunistic Infections; February 16-19, 2010; San Francisco, Calif. Abstract 129.

Download the audio file