Age and bone mass as predictors of fracture in a prospective study.

To study the effect of bone mass on the risk of fracture, we followed 521 Caucasian women over an average of 6.5 yr and took repeated bone mass measurements at the radius. We observed 138 nonspinal fractures in 3,388 person-yr. The person-years of follow-up and the incident fractures were cross-classified by age and bone mass. The incidence of fracture was then fitted to a log-linear model in age and bone mass. It was found that incidence of fracture increased with both increasing age and decreasing radius bone mass. When subsets of fractures were examined it was found that age was a stronger predictor of hip fractures, whereas midshaft radius bone mass was a stronger predictor of fractures at the distal forearm. We concluded that bone mass is a useful predictor of fractures but that other age-related factors associated with fractures need to be identified.

The loss of bone mineral with aging and its relationship to risk of fracture.

Longitudinal studies have shown that individuals lose bone mineral at unequal rates with aging. It has been postulated that individuals with the more rapid rates of loss constitute a separate population having an increased risk for developing fractures, i.e., osteoporosis. To examine this postulate, we made a search for a separate population of elderly women using a precise and objective measurement technique of bone mineral, photon absorptiometry. Bone mineral content (BMC) was measured in the radius of 571 Caucasian females who were age 50 or older. It was found that BMC values adjusted for width had a normal distribution in all decades and the variation in BMC values did not increase with age. Subjects with vertebral fractures (n = 108) were estimated to be losing bone mineral at the same rate as those without vertebral fractures (n= 161). Thus evidence for a separate population of rapid losers of bone mineral was not found. Reconciliation of longitudinal studies which show unequal rates of loss with the present population survey, in which evidence for unequal rates was not found, would require that (a) the rate of loss of bone mineral for an individual is not constant and/or (b) the rate of mineral loss is proportional to the amount of mineral present at maturity. The incidence of vertebral fractures was inversely proportional to BMC values. In a group of 278 women followed for 470 subject-yr, the incidence of all fractures during the study (n = 31) was also inversely proportional to BMC. These data suggest that the BMC values of osteoporotics would be at the lower end of normally distributed values for the population.

Age and activity effects on rate of bone mineral loss.

It has been postulated that the rate of mineral loss in postmenopausal women remains constant with aging and that the decreased activities of daily living associated with aging contribute to mineral loss. These hypothese were examined by measuring the bone mineral content at the midshaft of the radius with the photon absorption technique. The estimated rate of loss was calculated in a cross-sectional study as the regression coefficient of bone mineral content vs. age and in a longitudinal study as the regression coefficient of bone mineral content vs. time. In the cross-sectional study, Group A, which consisted of 264 women aged 50-72 yr, had an estimated rate of loss of -0.0114 +/- 0.0014 (SE) g/cm per yr. Group B, which consisted of 266 women aged 73-96 yr, had an estimated rate of loss of -0.0055 +/- 0.0017 g/cm per yr. In the longitudinal study, Group C consisted of 33 women aged 51-65 yr who were followed for an average of 4.5 yr with a mean number of 16 visits per subject; they were found to have a mean rate of loss of -0.00990 +/- 0.00107 g/cm per yr. Group D consisted of 38 women aged 70-91 yr who were followed for an average of 3.8 yr with a mean number of 31 visits per subjects; they were found to have a mean rate of loss of -0.00020 +/- 0.00236 g/cm per yr. The estimated and directly measured rates of loss were more rapid in the younger groups than in the older groups (A vs. B, P less than 0.001; C vs. D, P less than 0.001). These data demonstrate that the mean rate of mineral loss is not constant with aging and that in elderly subjects it is significantly slower than that of the earlier postmenopausal years. Since the elderly women were the less active, these findings suggest that factors other than decreased physical activity are more important in determining the rates of mineral loss.

Bone mineral density in relation to polymorphism at the vitamin D receptor gene locus.

Polymorphism at the vitamin D receptor gene was examined in relation to bone mineral density (BMD) at spine, femur, and forearm in 86 monozygotic (MZ) and 39 dizygotic (DZ) adult female twins. All were white, 63 pairs (44 MZ, 19 DZ) were premenopausal, and 43 pairs (31 MZ, 12 DZ) were discordant for age at menopause or use of estrogen. Each individual of the DZ pairs and one individual of MZ pairs was genotyped for ApaI, BsmI, and TaqI polymorphism at the vitamin D receptor gene locus using Southern hybridization. Intraclass correlations for BMD in MZ and DZ twin pairs indicated that heritability accounted for over 70% of BMD. There was no relationship between genotype for any of the three polymorphisms and BMD at any skeletal site in the twin population, considered either as a total population, both with and without twins discordant for age at menopause or use of estrogen, or as a premenopausal population. In DZ twin pairs discordant for alleles for the three polymorphisms, no allele was associated with higher or lower BMD. It is concluded that in this population of healthy adult females there was no relationship between these polymorphisms at the vitamin D receptor gene locus and BMD.

Sex steroids and bone mass. A study of changes about the time of menopause.

To examine the relationships between bone loss and sex steroids, 84 peri- and postmenopausal women were studied at 4-mo intervals for 3 yr. At each visit, measurements were made of bone mass at the midshaft and distal radius, of steroids, of gonadotropins, and of bone gla protein (BGP). Bone loss was approximately 1% per yr among late perimenopausal and postmenopausal groups, whereas the early perimenopausal group lost no bone. Mean serum estrogen and BGP concentrations predicted rates of bone loss. BGP was negatively correlated with the rate of bone loss (r = -0.45) and with mean estrogen concentrations (r = -0.40). Multivariate regressions showed estrogen concentrations to be strong independent predictors of the slope of bone mass over time. When BGP concentrations were added to the models, the significance of estrogen was reduced, suggesting that a portion of the estrogen effect was mediated through effects on rates of bone remodelling.

Sex steroids, bone mass, and bone loss. A prospective study of pre-, peri-, and postmenopausal women.

Although bone loss around the time of menopause is driven by estrogen deficiency, the roles of estrogens and androgens in the preservation of skeletal mass at other stages of life are less well understood. To address this issue we studied 231 women between the ages of 32 and 77 with multiple measurements of sex steroids and bone mass over a period of 2-8 yr. In all women bone mass was negatively associated with concentrations of sex-hormone binding globulin, and positively associated with weight. Bone loss occurred from all skeletal sites in peri- and postmenopausal women, but premenopausal women lost bone only from the hip (-0.3%/yr) and had positive rates of change in the radius and spine. Bone loss was significantly associated with lower androgen concentrations in premenopausal women, and with lower estrogens and androgens in peri- and postmenopausal women. Sex steroids are important for the maintenance of skeletal integrity before menopause, and for as long as 20-25 yr afterwards.

Genetic factors in determining bone mass.

This investigation was undertaken to evaluate possible genetic determinants of bone mass with the premise that inheritance of bone mass could be of etiologic importance in osteoporosis. Bone mass and width measurements were made with the photon absorption technique on the right radius of 71 juvenile and 80 adult twin paris. The variance of intrapair differences of bone mass in monozygotic (MZ) juvenile twins was 0.0013 g(2)/cm(2) compared to 0.0052 g(2)/cm(2) in the dizygotic (DZ) twins. For the adult twins the variance of intrapair differences in bone mass was 0.0069 for MZ and 0.0137 for DZ twins. Similar results were obtained for bone width. The significantly larger variation in intrapair differences in DZ twins indicates that these traits have significant genetic determinants. These intrapair differences were found to increase with age, suggesting that genetic-environmental interaction also contributes to the observed variation in bone mass. These data provide evidence that bone mass does have significant genetic factors, which alone or in conjunction with environmental factors may predispose persons to the development of osteoporosis.

Sex steroids and bone mass in older men. Positive associations with serum estrogens and negative associations with androgens.

The purpose of this study was to determine whether bone density in older men was associated with serum sex steroids or sex hormone binding globulin (SHBG). Bone density and sex steroids were measured in men over age 65 at 6-mo intervals for an average of 2.1 yr. Bone density was significantly positively associated with greater serum E2 concentrations (+0.21 < r < +0.35; 0.01 < P < 0.05) at all skeletal sites. There were weak negative correlations between serum testosterone and bone density (-0.20 < r < -0.28; 0.03 < P < 0.10) at the spine and hip. SHBG was negatively associated only with bone density in the greater trochanter (r = -0.26, P < 0.05). Greater body weight was associated with lower serum testosterone and SHBG, and greater E2. Because of these associations, regression models which adjusted for age, body weight, and serum sex steroids were constructed; these accounted for 10-30% of the variability in bone density, and showed consistent, significant positive associations between bone density and serum E2 concentrations in men, even after adjustments for weight and SHBG. These data suggest that estrogens may play an important role in the development or maintenance of the male skeleton, much as is the case for the female skeleton. These data also indicate that, within the normal range, lower serum testosterone concentrations are not associated with low bone density in men.

Long-term effects of raloxifene on bone mineral density, bone turnover, and serum lipid levels in early postmenopausal women: three-year data from 2 double-blind, randomized, placebo-controlled trials.

BACKGROUND: In postmenopausal women, raloxifene hydrochloride has favorable effects on bone and lipid metabolism and does not stimulate reproductive tissues. The studies reported herein evaluated the long-term (3-year) effects of raloxifene treatment on bone mineral density (BMD), serum lipid levels, and drug tolerability in healthy postmenopausal women. METHODS: A total of 1145 healthy European and North American postmenopausal women aged 45 through 60 years were enrolled in 2 parallel, double-blind, randomized, placebo-controlled trials of identical design and randomly assigned to receive raloxifene hydrochloride, 30, 60, or 150 mg, or placebo daily; all groups received 400 to 600 mg of elemental calcium. Assessments included measurements for BMD by dual-energy x-ray absorptiometry, markers of bone turnover, and serum lipid levels. RESULTS: Lumbar spine BMD changed from baseline to 36 months as follows: placebo (mean percentage change + SE), -1. 32% +0.22%; raloxifene, 30 mg, 0.71% +0.23%; raloxifene, 60 mg, 1. 28% +0.23%; and raloxifene, 150 mg, 1.20% +0.24%. Comparable BMD changes were observed in the hip and total body. Biochemical markers of bone turnover were suppressed by raloxifene to normal premenopausal ranges through 3 years. Serum low-density lipoprotein cholesterol was reduced 7% to 12% below baseline through 3 years. Study withdrawals due to any reason (37%) and withdrawals due to adverse events (14%) were not different among groups. The only significant adverse effect of therapy was hot flashes (25% in the 60-mg raloxifene group vs 18% in the placebo group); hot flashes were typically reported as mild and were not associated with study withdrawal (1.7% for 60-mg raloxifene vs 2.4% for placebo). CONCLUSIONS: Raloxifene preserves BMD at important skeletal sites, lowers serum low-density lipoprotein cholesterol levels, and has a tolerability profile comparable to placebo. These results indicate a favorable benefit-risk profile of raloxifene for long-term use in healthy postmenopausal women. Arch Intern Med. 2000;160:3444-3450.

Cigarette smoking, obesity, and bone mass.

This study was designed to assess the effects of smoking on bone mass and bone loss and to ascertain whether these effects are independent of effects on adiposity and hormone concentrations. A total of 84 healthy, peri- and postmenopausal women were studied prospectively over 3 1/2 years. Heavy smokers had significantly (p less than 0.05) lower radial and vertebral bone mineral content than light or nonsmokers (who did not differ from each other). In regression models, which contained measurements of obesity, pack-years smoking remained a significant predictor of bone mass. However, there were no detectable effects of smoking on rates of bone loss at any site. Smokers appear to be at greater risk of osteoporosis due to their lower bone mass. However, this reduced bone mass is already present around the time of menopause, and rates of bone loss during this period do not appear to be influenced by smoking. Furthermore, we have previously shown in this population that menopausal serum estrogen concentrations (which determine rates of bone loss) do not differ between the smokers and nonsmokers. Further studies of larger groups are required to determine whether small differences in bone loss may exist, since the power to detect such differences was not ideal in this study.

No evidence for an effect of lactase deficiency on bone mass in pre- or postmenopausal women.

The potential role for lactase deficiency in the development of low bone mass was examined in 342 adult female twins. Diminished lactase activity, defined as greater than 20 ppm increase in expired hydrogen at 2 or 2.5 h after an oral lactose load, was examined: (1) by comparing bone mass between members of twin pairs discordant for lactase activity; (2) by examining the linear association between bone mass and total expired hydrogen gas; and (3) by comparing all lactase-deficient individuals to those with persistent lactase activity. Among members of discordant (primarily DZ) pairs, the lactase-deficient member had greater bone mass 54% of the time. The correlations between the increase in expired hydrogen and bone mass at various sites were between -0.02 (femoral neck) and 0.11 (midshaft radius), suggesting no association between these variables. Finally, all lactase-deficient subjects were compared with those with normal lactase activity, regardless of twin status, and at each skeletal site the differences in bone mass were 1% or less. Thus, all primary hypotheses were not supported by these data; that is, in this large sample we could find no evidence of a detrimental effect of lactase deficiency on adult bone mass. However, baseline expired hydrogen was consistently and positively associated with bone mass at all sites, independently of age, suggesting the possibility that some aspect of intestinal function related to the activity of bacterial anaerobes may be positively associated with bone mass.

Role of physical activity in the development of skeletal mass in children.

A group of 118 children, aged 5.3-14 years, were enrolled in a prospective study of calcium supplementation and bone mass. At entry to the study, questionnaires regarding the child's usual physical activity were administered to the children and their mothers. Repeated activity assessments at 6 month intervals indicated good within-person agreement for total activity and for most individual activities. Consistent positive associations were observed between bone mineral densities (BMD) in the radius, spine, and hip and most activities. A summary measure (total hours of weight-bearing activity) was significantly related to BMD in the radius and hip, independently of age or gender effects. Self-reported sports and play activities were associated with BMD, but neither time spent watching television nor hours of physical education classes were associated either positively or negatively with skeletal mass. These data suggest that important increments in skeletal mass may result from physical activity during childhood.

Longitudinal study of bone mass in end-stage renal disease patients: effects of parathyroidectomy for renal osteodystrophy.

The effectiveness of parathyroidectomy (PTHX) for the control of secondary hyperparathyroidism was assessed in 46 adult end-stage renal disease (ESRD) patients whose bone mineral content at the midshaft and distal radius was measured using single-photon absorptiometry (SPA) every 6 months before and after the surgery. They were compared to 46 age-, race-, and sex-matched ESRD patient controls who had not undergone surgery but who had had at least five SPA studies at similar intervals. Presurgery midradius bone mass was significantly lower for PTHX patients compared to controls. Comparing changes in bone mass of PTHX patients across surgery to controls in comparable time periods showed that PTHX patients lost significantly less bone mass after surgery. Similar results were obtained when rates of change in bone mass were evaluated. When patient characteristics were examined, the effect of surgery was found to be diminished in elderly patients and in oophorectomized patients. It is concluded that PTHX can have a salutary effect on renal osteodystrophy in the appendicular skeleton, but factors other than bone mass also need to be considered in identifying those patients who will benefit from surgery.

Genetic determinants of bone mass in adult women: a reevaluation of the twin model and the potential importance of gene interaction on heritability estimates.

We estimated genetic effects on bone density in pre- and postmenopausal twins and critically considered the assumptions of the twin model. Bone mass in the radius, lumbar spine, and hip, anthropometric measurements, usual calcium and caffeine intake, tobacco and alcohol use, number of pregnancies and live births, menstrual history, usual physical activity, and medical history were measured in a volunteer sample of 171 twin pairs [124 monozygotic (MZ) and 47 dizygotic (DZ)], aged 25-80, free of diseases known to affect bone mass or mineral metabolism. At all skeletal sites, MZ intraclass correlations exceeded DZ correlations for both pre- and postmenopausal women, yielding highly significant estimates of heritability for bone mass. Adjustments for height, age, and environmental characteristics did not reduce the heritability estimates. However, many of these estimates were unrealistically high, suggesting some violation(s) of the assumptions of the twin model. Thus, the familial resemblance in bone mass is due primarily to genetic effects at all skeletal sites and at all ages, although the importance of genetic effects is diminished with aging, as evidenced by increasing within-MZ pair variability in older women. Because of failures in the assumptions of the twin model, however, particularly the greater MZ environmental similarity and the probability of gene interaction, heritability estimates are probably too high and require cautious interpretation.

Reduced rates of skeletal remodeling are associated with increased bone mineral density during the development of peak skeletal mass.

Two related studies were conducted to assess the associations between markers of skeletal modeling and remodeling in healthy children. Members of monozygotic twin pairs, aged 6-14, enrolled in a clinical trial of calcium supplementation, were studied at the end of the period of supplementation and for 3 years thereafter. Supplemented children had significantly higher rates of gain in bone mineral density (BMD) (+3% on average) during the period of supplementation accompanied by significantly lower concentrations of serum osteocalcin (OC, -15%). During postsupplement follow-up, both differences in BMD and OC disappeared. Black females, age matched to the baseline ages of the white children, had significantly lower serum concentrations of both OC and tartrate-resistant acid phosphatase (TRAP) at all ages and higher BMDs. When stratified on serum TRAP concentrations, regardless of race, children with lower concentrations had significantly higher BMDs, and no racial differences were apparent. In regression models accounting for 70-80% of the variability in BMD in children, body size and TRAP, but not race, remained significantly associated with BMD. The skeletal advantages seen with calcium supplementation and black race appear to be associated with reduced rates of skeletal turnover. Given that markers of turnover during growth reflect both skeletal modeling and remodeling, and there is no apparent advantage to reduced skeletal modeling, it seems probable that reduced remodeling is the factor that accounts for the increases in bone mass.

Suppressed bone turnover by bisphosphonates increases microdamage accumulation and reduces some biomechanical properties in dog rib.

It has been hypothesized that suppression of bone remodeling allows microdamage to accumulate, leading to increased bone fragility. This study evaluated the effects of reduced bone turnover produced by bisphosphonates on microdamage accumulation and biomechanical properties of cortical bone in the dog rib. Thirty-six female beagles, 1-2 years old, were divided into three groups. The control group (CNT) was treated daily for 12 months with saline vehicle. The remaining two groups were treated daily with risedronate (RIS) at a dose of 0.5 mg/kg per day or alendronate (ALN) at 1.0 mg/kg per day orally. After sacrifice, the right ninth rib was assigned to cortical histomorphometry or microdamage analysis. The left ninth rib was tested to failure in three-point bending. Total cross-sectional bone area was significantly increased in both RIS and ALN compared with CNT, whereas cortical area did not differ significantly among groups. One-year treatment with RIS or ALN significantly suppressed intracortical remodeling (RIS, 53%; ALN, 68%) without impairment of mineralization and significantly increased microdamage accumulation in both RIS (155%) and ALN (322%) compared with CNT. Although bone strength and stiffness were not significantly affected by the treatments, bone toughness declined significantly in ALN (20%). Regression analysis showed a significant nonlinear relationship between suppressed intracortical bone remodeling and microdamage accumulation as well as a significant linear relationship between microdamage accumulation and reduced toughness. This study showed that suppression of bone turnover by high doses of bisphosphonates is associated with microdamage accumulation and reduced some mechanical properties of bone.

Choosing between predictors of fractures.

The identification of those at highest risk of osteoporotic fractures is a clinical goal that requires appropriate statistical comparisons of potential predictors of fractures. This article provides a formal approach of comparing individual predictors (e.g., bone mass at one site vs bone mass at another), or sets of predictors (e.g., bone mass vs other risk factors), and contrasts newer methods, such as bootstrapping, to receiver-operating-characteristics (ROC) curves, which have been previously used. The advantages of the bootstrapping approach are illustrated using time-to-fracture data from a published study demonstrating the use of baseline bone mass measurements in the prediction of fractures in 521 subjects with variable lengths of follow-up, extending to 12.5 years. Bone mineral density (BMD) was shown to be significantly better than bone mineral content (BMD) in predicting fractures in free-living subjects, but not in retirement-community subjects. Bone mineral apparent density (BMAD) was also compared with BMC and BMD and shown not to improve fracture prediction in these subjects.

Do variations in hip geometry explain differences in hip fracture risk between Japanese and white Americans?

Despite lower femoral neck bone mass, Japanese women have a substantially lower incidence of hip fracture than North American whites. Reasons for this discrepancy were sought in a study of 57 Japanese and 119 white American women aged 50-79. All women were in good health. Bone mineral content (BMC) in the femoral neck, femoral neck length (NL), femoral neck angle (theta), cross-sectional moment of inertia (CSMI), safety factor (SF), and fall index (FI) were calculated using dual x-ray absorptiometry. Height and weight were greater in Americans than in Japanese (1.62 versus 1.52 m; p < 0.0001 and 66.0 versus 49.4 kg; p < 0.0001, respectively). Mean BMC in the femoral neck and CSMI were greater in Americans than in Japanese (3.91 versus 3.02 g; p < 0.0001 and 0.99 versus 0.57 cm4; p < 0.0001, respectively). NL was longer in Americans (5.6 versus 4.4 cm; p < 0.0001) and theta was larger in Americans (130 versus 128 degrees; p < 0.01), whereas SF and FI were less in Americans than in Japanese (3.41 versus 5.12; p < 0.0001 and 1.00 versus 1.40; p < 0.0001, respectively). These results indicate that despite lower bone mass, Japanese women have lower risks of structural failure in the femoral neck, attributable primarily to shorter femoral necks and, to a lesser degree, a smaller femoral neck angle. Geometric characteristics of the femoral neck in Japanese women are associated with their lower hip fracture risk, and the measurement of proximal femoral geometry, combined with bone mass, may provide further clinical information about the risk of hip fracture.

Linkage of structure at the proximal femur to chromosomes 3, 7, 8, and 19.

Risk for osteoporotic fracture is determined in part by femoral structure, which is under genetic control. We conducted a genome scan in 638 sister-pairs for structure phenotypes. Significant evidence of linkage was detected with several chromosomal regions, including confirmation of our prior linkage findings. Bone strength and resistance to fracture at the proximal femur is determined in part by structural variables. We previously reported that several structural variables, including pelvic axis length, femur axis length, femur head width, and femur midshaft width, had significant or suggestive linkage to regions of chromosomes 3, 4, 5, 7, 9, 17, and 19 in a sample of 309 white premenopausal sister pairs. We now report the results of a genome-wide linkage analysis of femoral structure variables in 437 white and 201 black healthy premenopausal sister pairs, of which 191 white pairs overlapped with our previously published sample. Multipoint quantitative linkage analysis was performed using microsatellite markers genotyped throughout the genome. In the current sample, linkage of femoral structure to chromosomes 3, 7, and 19 was confirmed in the white sister pairs, and a new linkage to chromosome 8 was identified. There was linkage at chromosome 3 to femoral head width (logarithm of the odds [LOD] = 5.0) and femur shaft width (LOD = 3.6). On chromosome 19, there was linkage to femoral neck axis length (LOD = 3.2); on chromosome 7, to femoral head width (LOD = 5.0); and on chromosome 8, to femoral head width (LOD = 6.0). The current findings emphasize the importance of increasing sample size to replicate linkage findings and identify new regions of linkage.

Universal standardization of bone density measurements: a method with optimal properties for calibration among several instruments.

The International Dual-Photon X-Ray Absorptiometry (DXA) Standardization Committee (IDSC) conducted a cross-calibration study among three models of DXA machines from three different manufacturers. In that study, 100 subjects were scanned on all three machines. A set of equations were derived to convert bone mineral density (BMD) on each machine to a "standardized BMD" (sBMD) such that sBMD from the same subject derived from different machines would be approximately the same. In a reanalysis of the cross-calibration data, we showed that the conversion method used in the IDSC study did not achieve several optimal properties desirable in such conversions. We derived new conversion equations to sBMD based on minimizing differences among sBMD from the three machines. More important is that the new conversions have no residual bias that was present in the IDSC conversions. The performance of the methods were compared on the cross-calibration data as well as an external data set. We conclude that the IDSC conversions are adequate for clinical use on other machines worldwide, but that researchers should standardize their own machines in a laboratory using the new method.