Featuring Michael T. Yin, M.D., M.S.
Dr. Yin is well recognized for research on bone disease and other metabolic complications of HIV infection and antiretroviral therapy. The coauthor of several dozen publications on these issues, he is also a member of the Osteo Renal Exchange Program, which issued the 2015 recommendations for evaluating and managing bone disease in people with HIV. Dr. Yin is principal investigator of a NIAID-funded initiative to study the impact of menopause on bone and muscle in the Women's Interagency HIV Study (WIHS). And he is co-principal investigator of a NICHD-funded study examining bone health in antiretroviral-treated children in South Africa. Dr. Yin is also a member of Bone and Metabolic Working Groups for the WIHS, the Veterans Aging Cohort Study (VACS), and the AIDS Clinical Trials Group (ACTG). He earned his M.D. at the Columbia College of Physicians and Surgeons and added a Masters in Epidemiology at the Columbia University Mailman School of Public Health.
Focusing on Modifiable Risk Factors
Aside from age, what are the most important bone risk factors in people with HIV?
The biggest risks are prior fracture, white race, and low body weight.
You can't do anything about fracture history or race, but what about low weight? Should HIV clinicians aggressively try to get underweight people to add pounds?
I think that's a reasonable approach. You hit upon an important point. There are some modifiable risk factors and some that are not. Controlling body weight, gaining muscle mass with exercise, quitting smoking, and limiting alcohol drinking to moderate amounts are all things that are modifiable and are probably important in people who are at high risk of bone disease. For people who have really low body weight, especially if they are prone to falls, strength training and improvement in muscle mass function would be helpful.
For the majority of patients who are in a reasonable weight range, we don't know how much benefit they'll get from strength training and other exercise. But in older folks, frailer folks, those who are more prone to falls, that kind of intervention is probably going to be beneficial.
The lifestyle risk factors you just mentioned -- smoking, heavy alcohol drinking, and lack of exercise -- those are all habits that are pretty hard for people to change. Do you have any advice on how HIV clinicians can motivate people to address those factors?
Changing ingrained habits can be difficult, but many people do find the motivation to change. The United States, for example, now has more former smokers than current smokers.1 For some patients, understanding and seeing the short-term benefits of modifying their behavior are strong motivators. Clinicians can help patients find this motivation by mapping out and tracking gains from changed behavior.
Gray Zones in Interpreting DXA and FRAX
When are DXA scans and the FRAX algorithm appropriate screening tooling for low bone mineral density in people with HIV?
DXA screening is recommended for all HIV-positive people over 50.2 It's a logical extrapolation from screening guidelines in the general population, but I think it will result in lots of test results that are in the gray area. Interpreting results will require a discussion between patient and provider about whether to monitor with a follow-up DXA in 1 year after lifestyle modification, or to switch antiretrovirals and monitor, or to start bisphosphonates or other bone-specific therapy. The thresholds for each of those decisions are not clearly defined, and we hope research will provide results to help guide us in the future. FRAX -- the algorithm based on a list of clinical factors3 -- can be used in HIV patients. Guidelines from the European AIDS Clinical Society (EACS) recommend FRAX for men in their 40s and premenopausal women over 40 without fragility fracture risk.4 This is an extrapolation from screening guidelines for the general population in Europe, which differ from those followed in the United States. Available data suggest that FRAX underestimates risk in HIV-infected individuals.5 Some of that underestimation can be corrected if you consider HIV a cause of secondary osteoporosis in the FRAX calculation, but whether it is an accurate enough predictor of fracture risk in HIV to use it as a screening tool is still not clear.
You pointed out that recent HIV bone guidelines call for FRAX screening starting at age 40,2,4 but one of your studies found that young men with HIV have lower peak bone mass than young HIV-negative men.6 [See "Substance Use, Old Age, Young Age" in this issue] Should HIV clinicians be on the lookout for signs of bone problems in people in their 20s and 30s?
That's a good question. If we look, we're going to find evidence of bone deficiency in younger people with HIV, just as we did in that study. The problem with using a risk factor measure such as the FRAX -- and even bone mineral density (BMD) by DXA -- is that they are useful for predicting fracture only in older individuals. FRAX is well validated for people older than 50, although the calculator itself lets you go down to age 40. Those tools are less useful in younger people because fracture risks are generally low in younger individuals; even if a younger person has low BMD, their fracture risk is still much lower than that of an older person with the same BMD value. So the question is, in HIV-infected people who are infected at a very early age, including those infected perinatally, whether another kind of screening measure is indicated. We don't have good data to answer that question yet. The study that you pointed out and some others are just starting to assemble that data.
Your Veterans Aging Cohort Study analysis found that FRAX underestimates fracture risk in HIV-positive men 50 and older.5 Does that finding also raise concerns about the predictive power of FRAX in men in their early or later 40s?
That's an issue that we really wanted to address with that analysis. It's not a perfect analysis because we didn't have all the risk factors that are necessary to calculate FRAX. So the data we presented are illustrative but not definitive of how accurate FRAX could be in the HIV-infected population if you had all the variables. From our analysis, it appears that a modified-FRAX (calculated with all but two of the FRAX variables) has poor predictive value for fracture using commonly accepted thresholds for pharmacologic therapy (greater than 3% 10-year fracture risk at the hip). However, defining the true predictive value of FRAX in people with HIV still requires a definitive study in which all the variables are utilized. Bisphosphonates -- and before -- and beyond