| |
HIV Infection and Bone Mineral Density in Middle-Aged Women
|
| |
| |
Clinical Infectious Diseases April 1 2006;42:1014-1020
Julia H. Arnsten,1,2,3 Ruth Freeman,1,4 Andrea A. Howard,1,2 Michelle Floris-Moore,1,2 Nanette Santoro,4 and Ellie E. Schoenbaum1,2,4
Departments of 1Medicine, 2Epidemiology and Population Health, 3Psychiatry and Behavioral Sciences, and 4Obstetrics, Gynecology, and Women's Health, Albert Einstein College of Medicine, Montefiore Medical Center, Bronx, New York
"Women with HIV infection had significantly lower BMD at both the femoral neck and lumbar spine, significantly lower T scores at both sites, and significantly lower Z scores at the femoral neck than did women without HIV infection......The independent association between HIV infection and reduced BMD suggests that HIV infection may exert a direct effect on bone metabolism....Our observation that low BMD is associated with HIV infection, even in the presence of significant excess body mass, supports findings by other researchers that body composition factors may play a paradoxical role in affecting BMD.....middle-aged women with HIV infection demonstrate low BMD, compared with HIV-negative women with similar behavioral risk factors. Non-black, HIV-positive women with lower BMI may be at particular risk for osteopenia or osteoporosis....Excess body weight should, therefore, not be considered protective against loss of BMD among high-risk women. High-risk women, particularly those entering the menopause transition, should be screened for osteopenia, and calcium and vitamin D therapy should be considered for middle-aged women with HIV infection and opiate dependence."
ABSTRACT
Background. Osteopenia is prevalent in persons with human immunodeficiency virus (HIV) infection and is part of a normal sequence of aging in women. Most studies of bone mineral density (BMD) and HIV infection have focused on men or have lacked a comparison group of individuals without HIV infection with similar behavioral risk factors.
Methods. We analyzed BMD in 495 middle-aged women (defined as women 40 years of age); 263 women had HIV infection, and 232 women were HIV-negative with behavioral risk factors similar to those of the HIV-positive group.
Results. The median age of the women in the study was 44 years, 54% were black, and 92% had used illicit drugs. Femoral neck BMD and lumbar spine BMD were reduced in women with HIV infection, compared with women without HIV infection (femoral neck BMD, 1.01 +/- 0.13 g/cm2 vs. 1.05 +/- 0.13g/cm2; ; lumbar spine BMD, 1.21 +/- 0.17 g/cm2 vs. 1.24 +/- 0.17 g/cm2; p=.04). In addition to HIV infection, other factors independently associated with lower BMD in both sites were being older, not being black, and having a low body weight. In race-stratified multivariate analyses, HIV infection was associated with BMD only in non-black women. Among HIV-positive women, 84% had taken antiretrovirals, and 62% had taken protease inhibitors, but their use was not associated with BMD. Methadone treatment was also independently associated with reduced lumbar spine BMD.
Conclusion. Middle-aged women with HIV infection have reduced BMD, compared with women at similar risk for HIV infection, independent of antiretroviral use. Among HIV-positive women, those who are not black, who are underweight, and who use opiates may be at particular risk. Although the prevalence of reduced BMD in this cohort was higher among women with HIV infection than among those without (27% vs. 19%), the overall prevalence of reduced BMD was low, compared with national estimates and with previous studies involving HIV-positive women and men.
Widespread use of HAART has dramatically decreased HIV-associated mortality. Because HIV-related mortality is declining while HIV incidence remains stable, more older adults are living with HIV infection. These epidemiologic trends suggest that more women with HIV infection will experience menopause and its sequelae, including osteoporosis. Because osteoporotic bony fractures are associated with significant morbidity, any further risk of osteoporosis associated with HIV infection may have a dramatic impact on the health of older women with HIV infection.
Although bone mass is largely genetically determined, certain lifestyle and hormonal factors that are highly prevalent among women with HIV infection are associated with low peak bone mass and disordered bone metabolism [1-4]. These include physical inactivity [5], decreased intake of calcium and vitamin D [6], cigarette smoking [7-9], alcohol use, depression [10-12], and opiate use [13]. Hypogonadism has been documented in up to 60% of HIV-positive men and may also be prevalent in premenopausal, HIV-positive women [14, 15]. Increased levels of proinflammatory cytokines associated with HIV infection, including IL-1, IL-6, and TNF, may also contribute directly to accelerated bone loss [16].
Several studies have reported an increased prevalence of reduced bone mineral density (BMD) among persons with HIV infection [16-24]. Although some early studies found an association between reduced BMD and protease inhibitor therapy [20, 23, 25, 26], other studies have not confirmed this association [19, 21, 22, 27, 28] or have suggested that other antiretrovirals may be important etiologic agents [29]. Other risk factors for osteoporosis, such as low body weight, physical inactivity, cigarette smoking, and opiate use, may be particularly prevalent in HIV-positive persons [30]. To date, most studies of bone density in persons with HIV infection have focused on men or have lacked a comparison group of persons without HIV infection with behavioral risk factors (i.e., drug use) similar to those of HIV-positive persons. The independent effect of HIV infection on BMD in middle-aged women who have a history of drug use is not currently known. The objective of this study was to determine the association between reduced BMD and HIV infection, HAART, and other lifestyle factors in a cohort of middle-aged women either with or at risk for HIV infection.
Discussion
In this large, cross-sectional study, we found lower BMD among middle-aged women with HIV infection than among women with similar behavioral risk factors without HIV infection, but neither nucleoside reverse-transcriptase inhibitor nor protease inhibitor therapy was associated with low BMD. Other factors independently associated with reduced BMD were older age, non-black race, lower body weight, past or current estrogen use, and past bony fractures. Although most of these are well-known risk factors for osteopenia, HIV infection has not been previously associated with reduced BMD in middle-aged women who use drugs after controlling for other risk factors. In race-stratified analyses, the association between HIV infection and reduced BMD was limited to non-black women and was independent of body weight.
Although the prevalence of low BMD in this cohort was higher among HIV-positive women than it was among women without HIV infection (27% vs. 19%), the overall prevalence of reduced BMD was low, compared with national estimates among similarly aged white women and with previous studies of HIV-positive women and men. Among white women aged >50 years in the United States, the prevalence of reduced BMD is approximately twice as high as the prevalence we observed among the non-black women in this cohort; 51%-70% of white women aged 50 years in the United States have reduced BMD (34%-50% with osteopenia, and 17%-20% with osteoporosis) [33], compared with the observed prevalence among non-black women in our study of 30% (25% with osteopenia, and 5% with osteoporosis). Among male and female patients with HIV disease, previous studies have estimated that the prevalence of reduced BMD is 39%-76% (33%-65% with osteopenia, and 6%-16% with osteoporosis) [18, 19, 21, 24, 34]. In contrast, only 34% of HIV-positive, non-black women and 22% of HIV-positive, black women in our cohort had reduced BMD.
Probable explanations for the comparatively low rate of reduced BMD that we observed are the high number of black women and the extremely high prevalence of excess weight and obesity in this cohort. Previous studies have documented lower rates of peak bone mass among white women, compared with black women [35, 36], but similar studies focusing on black and white women with HIV infection are lacking. Regarding excess weight, 68% of women with HIV infection and 84% of women without HIV infection in this cohort were overweight or obese. Previous studies documenting high rates of reduced BMD in HIV-positive persons have primarily included subjects with lean or normal body mass [17, 19, 21, 24].
Our findings illustrate the protective effect that excess body weight exerts against bone loss, but they also highlight the important and under-recognized problem of obesity among women with HIV infection who use drugs. Although HIV infection in the pre-HAART era was characterized by wasting, advances in HIV treatment may now be contributing to the rise of obesity among HIV-positive persons in this country. Excess weight may be considered desirable among some women with HIV infection, because it may suggest a healthy immune system. Similarly, obesity among women who use drugs may mask opiate or cocaine addiction, which are frequently associated with weight loss. Weight loss may not be a priority among obese, HIV-positive women who use drugs or among their health care providers.
The independent association between HIV infection and reduced BMD suggests that HIV infection may exert a direct effect on bone metabolism. Previous studies have documented that hormone levels, cytokines, and body composition factors contribute to reduced BMD in HIV-positive women and men [17-19, 21, 24]. Hypogonadally mediated alterations of bone metabolism may lead to increased bone resorption during HIV infection, and proinflammatory cytokines, such as IL-1, IL-6, and TNF, act to increase bone resorption and decrease bone formation. Because we did not assess hormone levels, biochemical markers of bone turnover, or cytokines, we cannot explain the mechanism by which HIV infection adversely affects BMD. Our observation that low BMD is associated with HIV infection, even in the presence of significant excess body mass, supports findings by other researchers that body composition factors may play a paradoxical role in affecting BMD. Previous, smaller, cross-sectional studies have noted an association between abdominal visceral adiposity and reduced BMD independent of BMI [17, 28], but prospective studies are needed to better describe the relationship between changes in body composition associated with HIV infection and BMD.
Our finding that methadone therapy is independently associated with reduced BMD of the lumbar spine has not been previously reported. Opiate use has been associated with central hypogonadism, which can lead to low BMD through reduced levels of circulating luteinizing hormone, estrogen, and testosterone [13]. However, previous studies demonstrating this association have been small and have not controlled for confounding variables, such as body weight and comorbid illness. Our results suggest that HIV-positive women in methadone maintenance programs may be at particular risk for osteopenia and osteoporosis and should be counseled to take measures to prevent bone loss.
Our study has several advantages, including the large and ethnically diverse cohort, exclusive enrollment of middle-aged women, and inclusion of a comparison group of HIV-negative individuals with similar behavioral risk factors as the enrolled HIV-positive women. Smaller studies that enrolled primarily men or limited numbers of HIV-positive women also demonstrate reduced BMD but did not recruit appropriate, simultaneous control subjects with similar drug use behavior. In 2 prior studies that focused exclusively on HIV-positive women, the mean age was several years younger than it was in our cohort, the effect of opiate use on BMD was not considered, and the comparison groups were comprised of healthy women [24, 34]. Because our study included both HIV-positive women and HIV-negative women with similar behavioral risk factors, we were able to identify HIV infection as a risk factor for low BMD after controlling for lifestyle variables.
Despite these strengths, our study has limitations. Because of the cross-sectional design, we were unable to evaluate causal relationships between HIV infection, opiate use, and BMD. In addition, lack of assessment of hormone, biochemical, or cytokine levels precludes us from drawing conclusions about the underlying mechanisms of the observed associations. Finally, we did not measure abdominal visceral fat, so we were unable to determine its relationship to BMD.
In conclusion, middle-aged women with HIV infection demonstrate low BMD, compared with HIV-negative women with similar behavioral risk factors. Non-black, HIV-positive women with lower BMI may be at particular risk for osteopenia or osteoporosis. Although a large proportion of the women in this cohort were overweight or obese, HIV infection was associated with reduced BMD independent of body weight. Excess body weight should, therefore, not be considered protective against loss of BMD among high-risk women. High-risk women, particularly those entering the menopause transition, should be screened for osteopenia, and calcium and vitamin D therapy should be considered for middle-aged women with HIV infection and opiate dependence.
Results
Study participants. Participant characteristics are listed in table 1. Of the 495 participants, 263 (53%) had HIV infection. Compared with women without HIV infection, HIV-positive women were significantly more likely to be black and unemployed and significantly less likely to be current smokers, depressed, or overweight or obese. HIV-positive women were also less likely to have used heroin or have been treated with methadone. Among HIV-positive women, 44% had been diagnosed with HIV infection >10 years ago, and the vast majority (78%) were antiretroviral treatment-experienced.
Risk factors for low BMD were common among both HIV-positive and HIV-negative women: 91% were current or former cigarette smokers, 58% had symptoms of depression, many were opiate users (48% had never been in a methadone maintenance program, and 27% had used heroin in the past 5 years), and most were physically inactive (42% exercised <1 time per week, and 53% watched TV >4 h per day). Low body weight, however, was uncommon; nearly three-quarters of participants had body mass indexes (BMIs) in the overweight or obese range (BMI, 25). Even among HIV-positive women, only 32% were in the lean-to-normal range (BMI, <25). Women with HIV infection had significantly lower BMD at both the femoral neck and lumbar spine, significantly lower T scores at both sites, and significantly lower Z scores at the femoral neck than did women without HIV infection.
Factors associated with low BMD.
In univariate analyses, factors associated with low BMD (p<.05 for both sites) were older age, non-black race, lower body weight, HIV infection, past or current estrogen use, and past bony fracture. Methadone maintenance treatment was associated with reduced BMD at the lumbar spine only (p<.05). After adjusting for these and other factors that met the criterion of p<.2 in univariate analysis, independent associations with low BMD at both sites included older age, non-black race, lower body weight, HIV infection, past or current estrogen use, and past bony fracture (table 2). Past prednisone use was independently associated with low BMD at the femoral neck only, and methadone maintenance treatment was associated with low BMD at the lumbar spine only. Active cocaine use was included in the lumbar spine model because it was associated with low BMD in univariate analysis (p<.1).
Among HIV-positive women, low BMD was not associated with longer duration of HIV infection, lower CD4 cell counts, class of antiretroviral (nucleoside reverse-transcriptase inhibitors or protease inhibitor), or duration of antiretroviral therapy (table 3). Substituting past or present use of nucleoside reverse-transcriptase inhibitor or protease inhibitor therapy for duration of therapy did not significantly alter these results.
Prevalence of low BMD.
Among 495 participants, 114 (23%) had low BMD (osteopenia or osteoporosis) of the femoral neck or spine, including 101 (20%) with osteopenia and 13 (3%) with osteoporosis. Low BMD was significantly more prevalent among HIV-positive women than among women without HIV infection (both femoral neck and spine, 27% vs. 19%, ; femoral neck only, 12% vs. 7%, p=.06; spine only, 21% vs. 15%, p=.06).
Race-stratified analysis.
Because race was strongly associated with BMD and there were unequal proportions of black women in the HIV-positive and HIV-negative groups, we calculated the prevalence of low BMD in black and non-black women separately and constructed additional linear regression models stratified by race (black vs. non-black). Unique models were constructed for the femoral neck and lumbar spine for both race strata that included the same variables as in the combined models.
Regarding the race-specific prevalence of low BMD, defined as a T score 1.0 SD below the average peak bone mass in young adult women, 68 non-black women (30%) and 46 black women (18%) had low BMD of the femoral neck or lumbar spine. In univariate analysis, low BMD at either site was more prevalent among HIV-positive women than among HIV-negative women in both race strata (34% vs. 26% among non-black women, p=.2; 22% vs. 11% among black women, p=.03).
Among non-black women, factors independently associated with low femoral neck BMD were age (beta= -0.008 per year; p<.0001), weight (beta= 0.003 per kg; p<.0001), and HIV infection (beta= -0.008 ; p<.01), and factors independently associated with low lumbar spine BMD were age (beta = -0.008 per year; p<.001), weight (beta= 0.001 per kg; p=.03), HIV infection (beta= -0.05 ; p=.03), and methadone maintenance treatment (beta= -0.04 ; p=.05). It is of note that, despite a higher prevalence of underweight women in the HIV-positive group, weight and HIV infection were each independently associated with low BMD.
In contrast, in the models including only black women, HIV infection was not associated with low BMD at either site. Among black women, factors independently associated with low femoral neck BMD were age (beta= -0.08 per year; p<.01), weight (beta= 0.002 per kg; p<.0001), and past prednisone use (beta= -0.08 ; p<.01), and factors independently associated with low lumbar spine BMD were weight (beta= 0.001 per kg; p=.05), current or past estrogen use (beta= -0.07 ; p<.01), and methadone maintenance treatment (beta= -0.06 ; p=.03).
Methods
Study participants. We studied BMD in 495 participants from the Menopause Study, an ongoing longitudinal study of menopause and its sequelae in women either with or at risk for HIV infection. Participant recruitment and study design have been described elsewhere [31, 32]. In brief, between September 2001 and July 2003, community-based women were enrolled to form a cohort in which about 50% of subjects had HIV infection and about 50% did not have HIV infection. In both HIV strata, 50% of participants reported using illicit drugs in the past 5 years, and 50% reported other high-risk behavior, such as sexual contact with an injection drug user. Menopause Study participants attend semiannual research visits, which include a standardized interview, blood collection for HIV serological tests and T lymphocyte studies, and weight and height measurements. The Menopause Study was approved by the institutional review boards of Montefiore Medical Center and Albert Einstein College of Medicine, and all participants provided written, informed consent.
Interview data. Interview data included sociodemographic characteristics, medical history (including amenorrhea), use of antiretrovirals and other medications, exercise habits, past and current drug use (e.g., use of tobacco, heroin and cocaine), alcohol use, and substance abuse treatment. Regular exercise was defined as moderate or strenuous exercise at least once per week. Symptoms of depression were assessed using the Center for Epidemiologic Studies Depression scale, which has been validated in several populations, including drug users and HIV-positive persons.
BMD. BMD of the hip (femoral neck) and lumbar spine (L2-L4) were measured by dual x-ray absorptiometry using a Prodigy densitometer with GE Lunar software, version 6.8 (GE Healthcare). Dual x-ray absorptiometry scans were performed a median of 16 days (range, 0-152 days) after the baseline interview. Control data used by this machine are derived from the National Health and Nutrition Study, which uses data from white women to determine average peak bone mass. Osteopenia is defined as a T score >1.0 SD and <2.5 SDs below the average peak bone mass in young adult women, and osteoporosis is defined as a T score >2.50 SDs below the mean peak bone mass in young adult women. Because more than one-half of the women in the cohort were black, and these normative values are derived from white women, our primary analyses used BMD (g/cm2) rather than T or Z scores.
Data analysis. Associations of BMD at the femoral neck and lumbar spine with sociodemographic characteristics, drug and alcohol use, and clinical variables, including HIV serostatus and HAART use, were determined using the Student's t and Spearman's correlation coefficient tests for non-normally distributed data. P values were corrected for multiple comparisons using the Bonferroni method. Factors independently associated with BMD were assessed with linear regression analysis; all factors meeting the criterion of on univariate analysis were included in the model.
|
|
| |
| |
|
|
|
