| |
Reduced bone density in HIV-infected women
|
| |
| |
AIDS: Volume 18(3) 20 February 2004
Dolan, Sara Ea; Huang, Jeannie Sb; Killilea, Kathleen Ma; Sullivan, Meghan Pa; Aliabadi, Negara; Grinspoon, Stevena
From the aNeuroendocrine Unit and Program in Nutritional Metabolism,
Massachusetts General Hospital, Boston, Massachusetts, USA and the bProgram in Pediatric Gastroenterology and Nutrition, University of California San Diego, San Diego, California, USA.
......In logistic regression analysis among all subjects (n = 147), HIV-infected subjects were significantly more likely to have osteopenia (odds ratio 2.5, 95% confidence interval 1.1-5.8, P = 0.03) controlling for age (P = 0.005), BMI (P = 0.03), race (P = 0.67), and menstrual function (P = 0.75)......
........We demonstrate significantly lower bone density at the lumbar spine and hip among HIV-infected women as compared to female control subjects of similar age, weight and racial background. This is the first report of reduced bone density among otherwise healthy, ambulatory HIV-infected women in the era of highly active antiretroviral therapy (HAART). Although the absolute difference in bone density between the groups was relatively modest, HIV-infected subjects were 2.5 times more likely to have osteopenia as compared to healthy control subjects. Importantly, we controlled for age, race, BMI and menstrual function in our final model. Lower bone density may predispose HIV-infected subjects to increased morbidity and further bone loss with aging.......
....... In our cohort, total body fat and percentage fat mass were significantly lower in the HIV-infected subjects, which may reflect subcutaneous fat atrophy associated with highly active antiretroviral use. Alternatively, fat loss may reflect a prior history of wasting and malnutrition in HIV-infected women. Although the subjects in this study were normal weight, we also assessed historical low weight as a marker for potential prior weight loss. Reduced fat and prior history of low weight were significantly correlated with low bone density, suggesting that prior malnutrition and changes in fat distribution may contribute to reduced bone density in this population. These data are similar to recent data by Nolan et al. and Mondy et al. demonstrating significant associations between weight, body composition and bone density in HIV-infected men......
...........Among premenopausal women with anorexia nervosa, prolonged periods of low weight and suboptimal calcium intake, have been associated with osteopenia. In non HIV-infected women, numerous studies suggest the importance of weight in achieving and maintaining normal bone density. Although current weight was similar between the subjects, lowest adult weight tended to be lower, and was highly significant as a predictor of bone density in regression modeling among the HIV-infected patients........
SUMMARY
Although bone density has been previously investigated in HIV-infected men, little is known regarding bone density in HIV-infected women.
Bone density was measured by dual-energy X-ray absorptiometry in 84 ambulatory, HIV-infected females and 63 healthy female control subjects similar in age (41 ± 1 versus 41 ± 1 years, P = 0.83), body mass index (26.0 ± 0.6 versus 27.0 ± 0.5 kg/m2, P = 0.44) and racial background (% non-Caucasian, 61 versus 51%; P = 0.24, HIV-infected versus control).
Lumbar spine (1.02 ± 0.02 versus 1.07 ± 0.02 g/cm2, P = 0.03) and total hip (0.93 ± 0.01 versus 0.99 ± 0.01 g/cm2, P = 0.004) bone density were reduced in HIV-infected compared with control subjects.
Osteopenia was demonstrated in 54 versus 30% (P = 0.004) of HIV-infected versus control subjects and was 2.5 times more likely in a multivariate model accounting for age, race, menstrual function and body mass index.
Urinary N-telopeptides of type 1 collagen (NTx) (39.6 ± 3.5 versus 29.9 ± 2.0 nM/mM urine creatinine, P = 0.03) and osteoprotegerin (4.76 ± 0.23 versus 3.39 ± 0.17 pmol/l, P <= 0.0001) were increased in HIV-infected compared with control subjects.
Among the HIV-infected women, bone density correlated with weight (r = 0.41, P < 0.001) and inversely with urinary NTx (r = -0.28, P = 0.01). Bone density did not differ by current or past protease inhibitor, nucleoside reverse trancriptase inhibitor, or non-nucleoside reverse transcriptase inhibitor exposure.
HIV-infected women demonstrate reduced bone density. Altered nutritional status, hormonal function and body composition may contribute to lower bone density in HIV-infected women. Consideration should be given to testing bone density in HIV-infected women with risk factors for osteopenia.
Introduction
Low bone density has recently been reported among HIV-infected men receiving HAART, but the mechanism is unknown. Although women comprise an increasing proportion of HIV and AIDS cases, little is known regarding bone density in the growing population of HIV-infected women. In a prior study, we have shown reduced bone density among HIV-infected women with the AIDS wasting syndrome. We now investigate bone density in normal weight, ambulatory HIV-infected women. Our data demonstrate reduced bone density in this population in association with nutritional and hormonal factors. Consideration should be given to assessing bone density in HIV-infected women with significant risk factors for osteopenia.
Subjects
Eighty-four HIV-infected women and 63 HIV-negative female controls were recruited through community advertisement and primary care provider referral between March 2000 and April 2003. Subjects who had used megace, ketoconazole, antidiabetic agents, hormone replacement therapy, bisphosphonates, steroids, growth hormone, oral contraception pills, Depo Provera, Progestasert IUD, testosterone or any other anabolic agents within the prior 3 months were excluded. Subjects who engaged in substance abuse, were pregnant or breastfeeding in the past year, who had a history of oophorectomy, or were diagnosed with an illness affecting bone were also excluded from participation. Inclusion criteria for HIV-infected participants included age between 18 and 60 years, previously diagnosed HIV infection, without a change in or initiation of an antiretroviral regimen within 6 weeks of enrollment, and a body mass index (BMI) of 20-35 kg/m2. Duration of HIV and antiretroviral medication history was obtained via patient interview at the screen visit as was data regarding menstrual history and age of menarche. Women were characterized as eumenorrheic (normal menstrual function) or oligomenorrheic (less than three menstrual periods in the 3 months prior to study). Female control subjects met the same entrance criteria and enzyme-linked immunosorbent assay (ELISA) testing verified HIV-negative status.
Eligible subjects were seen at the General Clinical Research Centers at the Massachusetts General Hospital and the Massachusetts Institute of Technology. Menstruating subjects were seen in the early follicular phase (within 10 days of initiation of menses).
All subjects gave written informed consent and the study was approved by the Human Research Committee at the Massachusetts General Hospital and the Committee on Use of Human Subjects at Massachusetts Institute of Technology.
Bone density assessment
Bone mineral density of the lumbar spine, hip (total hip and femoral neck), and total body were measured by dual X-ray absorptiometry (DXA) using a Hologic 4500 densitometer (Hologic Inc., Waltham, Massachusetts, USA). The in vivo precision for the measurement of bone density using the DXA technique is 0.5- 1.5% at the lumbar spine, and the standard deviation of the lumbar spine bone density is 0.01 g/cm2. Osteopenia and osteoporosis were defined according to World Health Organization criteria (osteopenia: T score < -1.0 SD and >= -2.5 SD; osteoporosis: T score < -2.5 SD). Ethnicity-specific T-scores provided by the manufacturer were used to determine osteopenia and osteoporosis (Hologic, Inc.).
Body composition measurement
Total fat and lean body mass were measured by DXA. The DXA technique has a precision error of 3.0% for total body fat mass, and 1.5% for total lean body mass.
Nutrition evaluation
Participants completed a 4-day food record prior to their baseline visit. These records were reviewed with each patient by a registered dietician and analyzed using a computerized nutrition software product (NDS Version 4.01 and 4.02-NDS-R;, Regents of the University of Minnesota, Minneapolis, Minnesota, USA) to quantify total caloric, protein, and fat intake, as well as total calcium and vitamin D intake (including intake from supplements). Historical low adult body weight was determined by patient interview.
Laboratory methods
Blood sampling was performed after a 12-h overnight fast. Sampling occurred at the same time of year in both patient groups (data not shown). Serum osteocalcin was measured using a two-site immunoradiometric assay (Nichols Institute Diagnostics, San Juan Capistrano, California, USA). The intra-assay coefficient of variation (CV) was 3.2-5.2%. Serum 25-hydroxyvitamin D was measured by radio-immunoassay (RIA) kit (DiaSorin Inc., Stillwater, Minnesota, USA). The intra-assay CV was 8.6-12.5%. 1,25-dihydroxyvitamin D was measured using a competitive ELISA with an intra-assay CV of 6.6% (American Laboratory Products Company, Windham, New Hampshire, USA). Serum parathyroid hormone (PTH) was measured using a two-sided immunoradiometric assay (Nicholas Institute Diagnostics), the intra-assay CV was 1.8-3.4%. Osteoprotegerin (OPG) (standard range 0-30 pmol/l) was measured by enzyme immunoassay (American Laboratory Products Company).
Serum estradiol was measured by RIA kit (Diagnostic Systems Laboratories, Inc., Webster, Texas, USA) with an intra-assay CV of 6.5-8.9%. Serum luteinizing hormone (LH) was measured using a solid-phase immunoradiometric assay (Diagnostic Products Corporation, Los Angeles, California, USA) with an intra-assay CV of 1.0-1.6%. Serum follicular stimulating hormone (FSH) was measured using a solid-phase immunoradiometric assay (Diagnostic Products Corporation) with an intra-assay CV of 2.2-3.8%. Serum calcium, phosphorous, albumin and creatinine were measured using standard laboratory techniques.
Twenty-four hour urine collections were performed for urine N-telopeptides of type 1 collagen (NTx), measured using a competitive-inhibition ELISA with an intra-assay CV of < 20% (Ostex International Inc., Seattle, Washington, USA). Urine NTx levels were corrected for urinary creatinine. Lactic acid levels were measured using an enzymatic, colorimetric method with lactic oxidase and 4-aminoantipyrine (Roche North America, Indianapolis, Indiana, USA). The intra-assay CV was 0.62-0.92%.
CD4 cell counts were measured by flow cytometry (FACS scan analyzer, Becton Dickinson and Co., San Jose, California, USA). HIV-RNA was quantified using a sandwich nucleic acid hybridization procedure, the Quantiplex HIV-RNA Assay (Chiron Corporation, Emeryville, California, USA). The lower limit of detection was 50 copies/ml.
Results
Demographic data
A total of 110 HIV-infected women and 80 HIV-negative patients were screened for the study. Nine HIV-positive subjects and 12 control subjects were found ineligible at screening. Seventeen HIV-positive and 5 control subjects met eligibility and did not continue participation after screening. Eighty-four HIV-infected subjects and 63 control subjects completed the study.
The HIV-infected and control subjects recruited for this study were similar in age (41 ± 1 versus 41 ± 1 years, P = 0.83), BMI (26.0 ± 0.6 versus 27.0 ± 0.5 kg/m2, P = 0.44) and race (61 versus 51% non-Caucasian, P = 0.24, HIV-infected versus control) (Table 1). Age of menarche was similar among both groups (13 ± 0 versus 13 ± 0 years, P = 0.44). The percentage of subjects with oligomenorrhea was not different between the groups (29 versus 19%, HIV-infected versus control, P = 0.23). The percentage of women smoking was larger in the HIV-infected group, (44 versus 26%, HIV-infected versus control, P = 0.03). In addition, the percentage of patients with a history of past substance abuse (40 versus 3%, P < 0.001) and past intravenous drug use (36 versus 0%, P < 0.001) was larger in the HIV-infected than control population.
The mean duration of HIV disease for HIV-positive study participants was approximately 8 years. A majority of subjects (93%) reported prior history of antiretroviral exposure, with 68% reporting prior protease inhibitor (PI) exposure, 93% reporting prior exposure to nucleoside reverse transcriptase inhibitors (NRTIs), and 48% reporting prior exposure to non-nucleoside reverse transcriptase inhibitors (NNRTIs). Forty-two percent of HIV-infected subjects were receiving PIs, 80% were receiving NRTIs, and 27% were receiving NNRTIs at the time of study. Mean CD4 cell count was 385 and viral load was undetectable in 42% of HIV-infected subjects.
Bone density
Bone density was lower at the lumbar spine (1.02 ± 0.02 versus 1.07 ± 0.02 g/cm2, P = 0.03), total hip (0.93 ± 0.01 versus 0.99 ± 0.01 g/cm2, P = 0.004) and femoral neck (0.82 ± 0.01 versus 0.87 ± 0.01 g/cm2, P = 0.01) in HIV-infected compared with control subjects. Significant differences remained controlling for race in all analyses (P = 0.02 for the lumbar spine, P = 0.008 for the femoral neck and P = 0.002 for the total hip) as well as for smoking status. Osteopenia at either the hip or spine was demonstrated in 54 versus 30% of HIV-infected subjects versus controls, P = 0.004 and osteoporosis was demonstrated in 10 versus 5% HIV-infected versus control, P = 0.27.
In logistic regression analysis among all subjects (n = 147), HIV-infected subjects were significantly more likely to have osteopenia (odds ratio 2.5, 95% confidence interval 1.1-5.8, P = 0.03) controlling for age (P = 0.005), BMI (P = 0.03), race (P = 0.67), and menstrual function (P = 0.75).
Endocrine parameters
Serum calcium, phosphorous and creatinine PTH, osteocalcin and 25-hydroxyvitamin D levels were similar between the two groups. The percentage of subjects with 25 hydroxyvitamin D below 25 nmol/l (28 versus 20%, did not differ between treatment groups, P = 0.31). 1,25-dihydroxyvitamin D was significantly reduced in the HIV-infected subjects compared with the control group [61 ± 6 versus 81 + 6 pmol/l, P = 0.01 (25.4 ± 2.3 versus 33.8 ± 2.5 pg/ml)]. Estradiol, LH and FSH were not different between the groups (Table 2). 18 versus 13% of HIV-infected versus control subjects were menopausal based on increased FSH, P = 0.49.
Bone markers
Urinary NTx (39.6 ± 3.5 versus 29.9 ± 2.0 nM/mM urine creatinine, P = 0.03), and serum OPG (4.76 ± 0.23 versus 3.39 ± 0.17 pmol/l, P <= 0.0001) were significantly increased in the HIV-infected group compared with the control subjects. Urinary creatinine excretion did not differ between the two groups. Osteocalcin was not different between the groups.
Body composition
Total body fat (22.0 ± 1.0 versus 26.0 ± 1.0 kg, P = 0.02) and percentage body fat (30.9 ± 0.7 versus 33.9 ± 0.7%, P = 0.003) were significantly lower in the HIV-infected subjects compared with control subjects. Total lean body mass was similar between the HIV-infected subjects and control subjects.
Dietary measures and metabolic markers
Total caloric, protein, fat, calcium and vitamin D intake were not different between the two groups of women. Lactic acid levels were increased among the HIV-infected subjects (1.01 ± 0.07 versus 0.69 ± 0.03 mmol/l, HIV-infected subjects versus control subjects, P < 0.001).
Correlates of low bone density in HIV-infected subjects
Current BMI, historical low adult weight, fat and lean body mass were significantly associated with hip and spine bone density among the HIV-infected women (Table 4). Urine NTx was inversely associated with bone density at both the hip and lumbar spine. Lactic acid levels were not associated with bone density at either the total hip or lumbar spine (Table 4). In a stepwise regression analysis, lowest adult weight (P < 0.0001) and urine NTx (P = 0.047) entered the model and were significant predictors of hip bone density among the HIV-infected group, overall R 2 = 0.39 for the model.
Although menstrual function (percentage oligomenorrheic), FSH and estradiol did not differ between the HIV and control groups, significant differences in bone density, menstrual function and gonadotropin levels were seen within the HIV group. Bone density was reduced (lumbar spine 0.96 ± 0.02 versus 1.06 ± 0.02 g/cm2, P = 0.002) and NTx increased (53.3 ± 12.1 versus 33.2 ± 2.8 nM/mM urine creatinine, P = 0.03) in oligomenorrheic versus eumenorrheic HIV-infected women. Similarly, bone density was reduced at all sites in the HIV-infected women with FSH > 15 IU/l versus those with FSH < 15 IU/l (0.92 ± 0.03 versus 1.05 ± 0.02 g/cm2, P < 0.001 for the lumbar spine; 0.74 ± 0.02 versus 0.84 ± 0.02 g/cm2, P = 0.002 for the femoral neck; 0.86 ± 0.02 versus 0.96 ± 0.03 g/cm2, P = 0.03 for the total hip). Bone markers did not correlate with PTH, 25-hydroxyvitamin D or 1,25-dihydroxyvitamin D levels (data not shown). Bone density was not lower among smokers than non-smokers and did not correlate significantly with pack years of smoking (r = -0.04, P = 0.75 for the lumbar spine; r = -0.09, P = 0.43 for the femoral neck; and r = -0.12, P = 0.26 for the total hip).
Bone density was not different at any site by current or prior PI, NRTI or NNRTI exposure and did not correlate with total duration of PI, NRTI or NNRTI use. Similarly, no effect of PIs, NRTIs or NNRTIs was seen on bone density in multivariate modeling including simultaneous terms for total duration PI use, NRTI use and NNRTI use.
DISCUSSION by authors
We demonstrate significantly lower bone density at the lumbar spine and hip among HIV-infected women as compared to female control subjects of similar age, weight and racial background. This is the first report of reduced bone density among otherwise healthy, ambulatory HIV-infected women in the era of highly active antiretroviral therapy (HAART). Although the absolute difference in bone density between the groups was relatively modest, HIV-infected subjects were 2.5 times more likely to have osteopenia as compared to healthy control subjects. Importantly, we controlled for age, race, BMI and menstrual function in our final model. Lower bone density may predispose HIV-infected subjects to increased morbidity and further bone loss with aging.
Hormonal factors are an important determinant of bone density in women. In this study, neither age of menarche or percentage of subjects with oligomenorrhea were different in the HIV groups in comparison with the control groups, as confirmed by the nearly identical estradiol, FSH and LH levels between the groups. Nonetheless, 29% of HIV-subjects were oligomenorrheic and bone density was reduced in the oligomenorrheic compared to eumenorrheic HIV-infected subjects. These data suggest that abnormal menstrual function may contribute to low bone density in a subpopulation of HIV-infected women. Disturbances in menstrual function have been shown in other studies of HIV-infected women, and may result from HIV, associated illness or malnutrition, or medications. Subjects in the current study were normal weight and ambulatory, without concurrent opportunistic infection. Gonadotropin and estradiol levels were not different between the groups and women on estrogen or other therapies known to affect bone were excluded from the study. Although menopausal status was similar between the groups, bone density was lower, as anticipated, among the HIV-infected patients with increased FSH as compared with the HIV-infected patients with normal FSH, suggesting that such patients are likely to experience even more bone loss with the onset of menopause.
Other factors in addition to menstrual dysfunction may also contribute to abnormal bone turnover and reduced bone density in HIV-infected women. HIV status was a significant predictor of osteopenia in regression analysis controlling for menstrual function, age, weight and race. Furthermore, the pattern of bone turnover demonstrated among the HIV-infected patients in this study is not typical of simple estrogen deficiency. In estrogen deficiency, overall bone turnover is increased, including markers of formation and resorption. In contrast, indices of bone formation were not increased in association with markers of resorption among our cohort of HIV-infected women, suggesting a potential uncoupling of bone formation and resorption. Aukrust et al. have previously demonstrated increased bone resorption and decreased bone formation among a mostly male population of HIV-infected patients. In addition, prior studies of low weight non HIV-infected women suggest a similar pattern of uncoupling and increased resorptive markers in the context of low or normal formation markers.
Osteoprotegerin (OPG), the decoy receptor of RANK ligand, neutralizes the stimulatory effect of RANK ligand on the differentiation and proliferation of osteoclasts. In our HIV-infected cohort, serum concentrations of OPG were significantly increased. Increased levels of OPG may reflect a compensatory mechanism to down regulate increased bone resorption, as seen in prior studies of menopausal women.
Among premenopausal women with anorexia nervosa, prolonged periods of low weight and suboptimal calcium intake, have been associated with osteopenia. In non HIV-infected women, numerous studies suggest the importance of weight in achieving and maintaining normal bone density. Although current weight was similar between the subjects, lowest adult weight tended to be lower, and was highly significant as a predictor of bone density in regression modeling among the HIV-infected patients. In our study population, total calcium and vitamin D intake and renal function were not different between the groups, but 1,25-dihydroxyvitamin D levels were significantly reduced in the HIV-infected subjects. Haug et al. have previously demonstrated a defect in 1-alpha vitamin D hydroxylation in HIV-infected subjects that may be associated with a PI effect on 1-alpha hydroxylation. However, 25-hydroxyvitamin D, PTH and serum calcium levels were not different, suggesting that overall calcium homeostasis was not significantly affected. Nevertheless, 1,25-dihydroxyvitamin D may have direct effects on bone and further studies of the biology of 1,25-dihydroxyvitamin D are needed in the HIV population.
In our cohort, total body fat and percentage fat mass were significantly lower in the HIV-infected subjects, which may reflect subcutaneous fat atrophy associated with highly active antiretroviral use. Alternatively, fat loss may reflect a prior history of wasting and malnutrition in HIV-infected women. Although the subjects in this study were normal weight, we also assessed historical low weight as a marker for potential prior weight loss. Reduced fat and prior history of low weight were significantly correlated with low bone density, suggesting that prior malnutrition and changes in fat distribution may contribute to reduced bone density in this population. These data are similar to recent data by Nolan et al. and Mondy et al. demonstrating significant associations between weight, body composition and bone density in HIV-infected men.
No differences in bone density were observed in the HIV-infected group by current and/or prior exposure to PI, NRTI, or NNRTI arguing against a direct effect of antiretroviral medications on bone density in this population. In addition, lactic acidemia may occur in HIV-infected patients as a result of NRTI therapy, and prior reports have implicated lactic acidemia as an important factor in HIV-associated osteopenia among men. Although HIV-infected women had significantly elevated lactate levels as compared to HIV-negative controls, there was no demonstrable relationship between lactate levels and bone density.
This study has a number of advantages, including exclusive enrollment of women. Smaller studies enrolling primarily men and limited numbers of HIV-infected women also demonstrate reduced bone density, but have not recruited simultaneous controls of similar age, weight and racial background. The use of T score to demonstrate an increased prevalence of osteopenia in our subjects is justified as the lowest age among the patients studied was 24. Further longitudinal studies will be interesting to see if bone density changes over time in this population.
In conclusion, HIV-infected women demonstrate lower bone density compared to healthy control subjects similar in age, weight and racial composition. This is the first report of osteopenia in ambulatory, normal weight, HIV-infected women. Altered nutritional status, hormonal function and body composition may contribute to lower bone density in this population. Further studies are needed to determine potential contributions of HIV per se to alterations in bone metabolism in HIV-infected females. HIV-infected women with low prior weight, increased bone resorption, and significant fat loss may be at particular risk for low bone density. Consideration should be given to screening HIV-infected women at high risk for osteopenia, particularly among those at or entering the menopausal transition, given the likelihood of further bone loss with aging. Among osteopenic HIV-infected women, bisphosphonates may be useful, especially given that bone resorption is increased. Further study of the safety and efficacy of bisphosphonates in osteopenic HIV-infected women is necessary.
|
|
| |
| |
|
|
|
