VITAMIN D IS REQUIRED FOR EFficient

Transcription

1 ORIGINAL CONTRIBUTION Occult Vitamin D Deficiency in Postmenopausal US Women With Acute Hip Meryl S. LeBoff, MD Lynn Kohlmeier, MD Shelley Hurwitz, PhD Jennifer Franklin John Wright, MD Julie Glowacki, PhD VITAMIN D IS REQUIRED FOR EFficient absorption of dietary calcium and for normal mineralization of bone. Reduction in vitamin D levels is associated with impaired calcium absorption and a compensatory increase in the level of parathyroid hormone (PTH) which, in turn, stimulates bone resorption and bone loss. According to the recent National Health and Nutrition Examination Survey Study, an estimated 26 million to 38 million US adults have osteoporosis or are at risk for osteoporosis in the hip. 1 Risk of hip fractures increases exponentially with age. Hip fractures are increasing worldwide, and are the most devastating and costly of the osteoporotic fractures. 2 Advancing age is associated with reductions in sun exposure, intake and skinactivation of vitamin D, and in vitamin D absorption, all of which may contribute to low vitamin D levels. 3 It is striking that in recent studies from England, Scotland, and South Africa, 13% to 33% of patients with hip fractures had histological signs of osteomalacia that may have been caused by prolonged vitamin D deficiency, 4-7 although not all European studies con- Context Low vitamin D levels may contribute to hip fractures in women, although limited data are available on vitamin D levels in US women admitted with acute hip fractures. Objective To determine whether postmenopausal women with hip fractures have low vitamin D and high parathyroid hormone levels compared with nonosteoporotic and osteoporotic women admitted for elective joint replacement. Design Comparative case series conducted between January 1995 and June Setting and Patients Ninety-eight postmenopausal community-dwelling women with no secondary causes of bone loss admitted for hip replacement, of whom 30 women had acute hip fractures and 68 women were admitted for elective joint replacement. Of the women admitted for elective joint replacement, 17 had osteoporosis and 51 did not. Women with comorbid conditions or who were taking medications that affect bone density and turnover were excluded. Main Outcome Measures Primary measures were levels of vitamin D and parathyroid hormone; secondary measures were body composition and markers of bone turnover. Results Women with hip fractures had lower levels of 25-hydroxyvitamin D than women without osteoporosis admitted for elective joint replacement (P =.02) and than women with osteoporosis admitted for elective joint replacement (P =.01) (medians, 32.4, 49.9, and 55.0 nmol/l, respectively; comparisons adjusted for age and estrogen intake). Parathyroid hormone levels were higher in women with fractures than women in the nonosteoporotic control group (P.001) or than elective osteoporotic women (P =.001) (medians, 5.58, 3.26, and 3.79 pmol/l, respectively; comparisons adjusted for age and estrogen intake). Fifteen patients (50.0%) with hip fractures had deficient vitamin D levels ( 30.0 nmol/l) and 11 (36.7%) had a parathyroid hormone level greater than 6.84 pmol/l. Levels of N-telopeptide, a marker of bone resorption, were greater in the women with hip fractures than in the elective nonosteoporotic controls (P =.004). Conclusions Postmenopausal community-living women who presented with hip fracture showed occult vitamin D deficiency. Repletion of vitamin D and suppression of parathyroid hormone at the time of fracture may reduce future fracture risk and facilitate hip fracture repair. Because vitamin D deficiency is preventable, heightened awareness is necessary to ensure adequate vitamin D nutrition, particularly in northern latitudes. JAMA. 1999;281: Author Affiliations: Endocrine Hypertension Division, Departments of Internal Medicine and Orthopedic Surgery, Brigham and Women s Hospital, Boston, Mass. Dr Kohlmeier is now with the Internal Medicine/Endocrinology Department, Washington Medical Group, Fremont, Calif. Financial Disclosure is listed at the end of this article. Corresponding Author and Reprints: Meryl S. Le- Boff, MD, Director of Skeletal Health and Osteoporosis Program, Endocrine-Hypertension Division, Brigham and Women s Hospital, 221 Longwood Ave, Boston, MA ( msleboff@bics.bwh.harvard.edu) American Medical Association. All rights reserved. JAMA, April 28, 1999 Vol 281, No

2 cur. 8 In the United States in 1978, Sokoloff 9 reported occult osteomalacia in 8 (25%) of 31 patients with hip fracture, and another study 10 from an elderly residential facility and hospital found 4 (12.9%) of 31 patients with osteomalacia at the time of hip fracture. Limited data are available on vitamin D status in US patients admitted for acute osteoporotic fractures. We compared such subjects with patients scheduled for elective joint replacement surgery to test the hypothesis that patients with acute hip fractures have lower vitamin D levels and higher PTH levels than do subjects without fracture with normal or even low bone mineral density. Figure 1. Selection of Postmenopausal Women Undergoing Joint Replacement Comorbid Medical Conditions (n = 190) Acute Hip (n = 262) Refused (n = 42) Postmenopausal Women (N = 805) Enrolled (n = 30) (n = 30) Comorbid Medical Conditions (n = 101) Joint Replacement (n = 543) Refused (n = 374) (n = 17) Enrolled (n = 68) Not (n = 51) Women were not included in the study if they had comorbid medical conditions or were taking medications that could affect bone, declined study participation, or had underlying hip disease other than osteoarthritis. Comorbid medical conditions included renal insufficiency, creatinine level of 177 µmol/l (0.02 mg/dl) or more, malabsorption, gastrectomy, active liver disease, acute myocardial infarction, alcoholism, and anorexia nervosa. Table 1. Patients Excluded According to Each Prespecified Exclusion Criterion No. (%) of Patients Screened (n = 389) Exclusion Criterion (n = 169) (n = 220) Excluded medication* 17 (10.1) 38 (17.3) Morbid obesity 6 (3.6) 0 (0) Rheumatic condition/underlying hip 26 (15.4) 16 (7.3) disease Dementia/unable to obtain consent 12 (7.1) 26 (11.8) Cancer within 5 y 16 (9.5) 50 (22.7) Comorbid condition 7 (4.1) 13 (5.9) Infection 0 (0) 31 (14.1) Premenopausal/perimenopausal 11 (6.5) 2 (0.9) Suppressed thyrotropin 1 (0.6) 3 (1.4) Men 1 (0.6) 0 (0) group, not osteoporotic 0 (0) 3 (1.4) Nursing home 1 (0.6) 1 (0.5) Trauma 1 (0.6) 4 (1.8) Previously enrolled 2 (1.2) 0 (0) d more than once 0 (0) 3 (1.4) Total *Glucocorticoids, androgens, anabolic steroids, anticonvulsants, calcitonin, sodium fluoride, bisphosphonates, and pharmacological doses of vitamin D ( 800 IU) were excluded because they could affect bone condition. vs fracture group, P.04. Rheumatic condition was defined as rheumatoid arthritis and ankylosing spondylitis. Underlying hip diseases other than osteoarthritis included avascular necrosis, psoriatic arthritis, and osteonecrosis. Included renal insufficiency (creatinine 1.77 µmol/l [0.02 mg/dl]), malabsorption, gastrectomy, active liver disease, acute myocardial infarction, alcoholism, and anorexia nervosa. Sex unknown prior to contact. Ineligible because residing in nursing home. METHODS Patient Recruitment A total of 805 postmenopausal women scheduled for joint replacement were identified by daily monitoring of orthopedic admissions between January 1995 and June 1998, including 262 women admitted to Brigham and Women s Hospital with acute hip fracture and 543 women scheduled for elective joint replacement either at the Brigham and Women s Hospital or the New England Baptist Hospital in Boston, Mass. Postmenopausal women who experienced either natural or surgical menopause with amenorrhea for 12 months were recruited for this study. Women receiving estrogen replacement therapy were included (46% of patients). Women were excluded if they were taking any other medications or had any disorders or abnormal admission test results that might affect bone, or had any underlying hip disease other than osteoarthritis (FIGURE 1 and TABLE 1). Of the 68 women admitted for elective joint replacement, 17 were determined to be osteoporotic (T score 2.5) according to the World Health Organization bone density criteria 11 and were analyzed separately (elective osteoporotic group) from the 51 elective patients admitted who were not osteoporotic (elective control group). There were 30 osteoporotic women with hip fractures (fracture group). Research protocols were approved by the institutional review boards of the Brigham and Women s Hospital and the New England Baptist Hospital. All patients or, if necessary, a designated family member or guardian, gave informed consent. Questionnaires Using a modification of the Nurses Health Study questionnaire 12 and dietary and detailed physical activity JAMA, April 28, 1999 Vol 281, No American Medical Association. All rights reserved.

3 questionnaires, all participants provided information regarding lifestyle, reproductive factors, dietary calcium consumption, and physical activity (TABLE 2). Spine, Proximal Femur, Total Body Bone Density, and Body Composition Bone mineral density of the spine (L1- L4), proximal femur, and total body were measured with the dual x-ray absorptiometry technique (QDR2000, Hologic Inc, Waltham, Mass). In vivo precisions (coefficient of variation percentage) in postmenopausal women for the spine, femoral neck, and trochanteric bone density measurements on different days were 1.21%, 1.74%, and 1.24%, respectively. 14 Women with vertebrae with moderately severe osteoarthritic changes, disk space narrowing, or a fracture were excluded from the analyses as these anatomic findings may elevate the spinal bone mineral density. Results were expressed as SDs and compared with bone mineral density values for age-matched healthy individuals (z score). Body composition (lean and fat mass) was determined by dual x-ray absorptiometry technique. The ratio of fat-to-lean tissue was calculated. Reproducibility of mean (SEM) for fat and lean tissue determinations in our laboratory were 1.09% (0.15%) and 0.89% (0.28%), respectively. Blood Chemistries and Assays Blood chemistries, complete blood cell counts, and tests for urinary calcium levels were performed in hospital clinical laboratories; all remaining tests were performed in the General Clinical Research Center laboratory unless otherwise specified. Blood samples were obtained preoperatively in 88% of the patients. In some instances, elective or hip fracture patients were rushed to surgery and informed consent and blood samples were obtained postoperatively. Serum intact PTH levels were measured with the sensitive Allegro immunoradiometric assay (Corning Nichols Institute, San Juan Capistrano, Calif). Serum 25-hydroxyvitamin D levels were measured using a radioimmunoassay (RIA) procedure (Incstar Corp, Stillwater, Minn) approved by the Food and Drug Administration. Levels of 1,25-dihydroxyvitamin D were measured by RIA. Urinary N-telopeptide levels, an index of bone resorption, were determined in a 24-hour urine collection by an enzyme-linked immunosorbent assay that measures cross-linked collagen peptides (Osteomark Assay, Ostex International Inc, Seattle, Wash). Bone formation markers including serum levels of osteocalcin by RIA 15 and bone-specific alkaline phosphatase (BSAP) were measured by immunoradiometric assay (Tandem-R Ostase assay, provided by Hybritech, San Diego, Calif) by Dr Gundberg of Yale University School of Medicine, New Haven, Conn. The BSAP levels were measured by immunoassay (Alphase B assay, provided, in part, by Metra Biosystems, San Diego, Calif). The interassay and intraassay coefficients of variation for all the assays ranged from 2.3% to 12.1%. Table 2. Characteristics of Enrolled Women Admitted for Surgery and Acute Hip Control (n = 51) (n = 17) (n = 30) Control vs P Value vs Variable Age, Bone, and Body Characteristics* Age, y (8.10) (10.42) (9.17) Years since menopause (9.01) (11.27) 28.0 (10.69) Bone density z score Spine (L1-L4) 1.33 (1.09) 0.33 (1.04) 0.21 (0.72) Femoral neck 1.16 (1.10) 0.32 (0.80) 0.90 (0.73) Trochanter 0.96 (1.13) 0.55 (0.74) 0.75 (1.00) Total body 1.19 (1.13) 0.87 (0.78) 0.31 (1.21) Body composition Fat g (10.04) (8.59) (10.25) % (6.28) (9.23) (10.60) Lean, g (4.02) (4.33) (6.08) Fat-lean ratio 0.95 (0.23) 0.81 (0.27) 0.56 (0.27) Intake and Activity Total calcium intake, mg/d 873 ( ) 700 ( ) 794 ( ) Alcohol, drinks per week 1 (0-14) 1 (0-10) 0 (0-16) Smoking, pack-years 8 (0-60) 0 (0-82) 8 (0-48) Physical activity, metabolic hours per week 16.4 (0-67.0) 20.2 (0-99.3) 1.5 (0-19.4) *Values measured as mean (SD) and are adjusted for estrogen intake. Bone mineral density equals SD compared with age-matched controls. Values measured as median (percentile [5th and 95th]) and adjusted for age and estrogen intake. Includes dietary and supplemental calcium. Among the elective nonosteoporotic control, elective osteoporotic, and fracture patients, 49%, 41%, and 37%, respectively, were taking a multivitamin. Metabolic hours included past and present sports, and job-related and domestic-related physical activities American Medical Association. All rights reserved. JAMA, April 28, 1999 Vol 281, No

4 Statistical Analyses Patient characteristics and markers of bone turnover were summarized as means and SDs. Analysis of variance was performed with prespecified contrasts comparing the fracture group with each of the elective groups, and having estrogen replacement therapy as a covariate. Normality tests were rejected for several calciotropic hormones, biochemical values, and baseline characteristics; therefore, medians and 5th and 95th percentiles were presented. The natural log transformed normalized intact PTH, 25- hydroxyvitamin D, N-telopeptides, BSAP (Metra Biosystems), BSAP (Hybritech), and creatinine. Analysis of variance was used for phosphate, 1,25-dihydroxyvitamin D, urinary calcium, transformed intact PTH, 25-hydroxyvitamin D, N- telopeptides, BSAP (Metra Biosystems), BSAP (Hybritech), and creatinine, with Figure 2. Women With Abnormal Parathyroid Hormone, 25-Hydroxyvitamin D, and Urinary Calcium Levels Women, % Parathyroid Hormone >6.84 pmol/l (>65 pg/ml) Group (n = 30) Group (n = 17) Control Group (n = 51) 25-Hydroxyvitamin D 30 nmol/l ( 12 ng/ml) Urinary Calcium <3 mmol/d (<120 mg/24 h) Percentage of women with elevated parathyroid hormone levels, 25-hydroxyvitamin D deficiencies and urinary calcium levels below the median. The women in the fracture group had the highest percentage of elevated parathyroid hormone levels, deficient 25- hydroxyvitamin D levels, or low-range urinary calcium concentrations. Asterisks indicate P.002 for fracture vs elective control or elective osteoporotic groups. age and estrogen replacement therapy as covariates. Cochran Mantel-Haenszel tests were used for albumin, total calcium, osteocalcin, total calcium intake, alcohol, smoking, and physical activity with estrogen replacement therapy and age above or below the median as stratification factors. Percentages and 2 tests were used to compare groups for dichotomized calciotropic hormones. Ninety-one subjects had hip surgery and 7 had knee surgery (6 from the elective control group, 1 from the elective osteoporotic group). Results were nearly identical with and without knee surgery subjects; combined results were presented. Analyses were performed using SAS software (SAS Inc, Cary, NC). RESULTS Patient Characteristics Characteristics of 98 women who were enrolled in the study are summarized in Table 2. The groups with osteoporosis (with fracture or elective) did not differ in the years since menopause, and they had comparable bone density levels (z scores) in the spine and proximal femur sites, but they differed in total body bone density. Body fat, percentage fat, and the fat-to-lean ratio showed monotonic changes among the 3 groups. Compared with the elective control group, thepostmenopausalwomenadmittedfor hip fracture had reduced exercise hours. Vitamin D and PTH Fifty percent (n = 15) of the postmenopausal women admitted with acute hip fractureshaddeficient25-hydroxyvitamin D levels ( 30.0 nmol/l [ 12 ng/ml]), 36.7% (n = 11) had elevated PTH levels ( 6.84pmol/L[ 65pg/mL]),and81.8% had urinary calcium levels below the median ( 3 mmol/d [120 mg/24 h]) (FIGURE2). Amongthewomenwithacute fracture, 8 had both deficient 25- hydroxyvitamin D levels ( 30 nmol/l) andelevatedpthlevels( 6.84pmol/L). Body Composition and Markers of Bone Turnover Circulating calcium levels were lower in the women with hip fractures than in the elective control and osteoporotic groups (TABLE 3). Compared with the women in the control group admitted for elective hip replacement, urinary N-telopeptide levels were higher in the osteoporotic elective and fracture groups (with similar z scores for the spine and hip). Levels of N- telopeptide were not different between the osteoporotic elective and fracture groups and were higher than the elective control group. Osteocalcin, BSAP (Hybritech), and BSAP (Metra Biosystems) were lower in women with fractures than in the elective osteoporotic group. The correlation between the BSAP (Hybritech) and the BSAP (Metra Biosystems) samples was 0.82 (Spearman rank correlation, P.001). Serum thyrotropin levels were comparable among the 3 groups (data not shown). There were no significant differences in calcium, albumin, phosphate, PTH, 25-hydroxyvitamin D, 1,25-dihydroxyvitamin D, urinary calcium, osteocalcin, creatinine, or N-telopeptide levels between women whose serum sample was obtained preoperatively and the 12% of women whose serum sample was obtained postoperatively. COMMENT This study shows low vitamin D and high PTH levels among elderly, community-dwelling women in Boston admitted for acute hip fracture with no secondary causes of bone loss. Half had vitamin D levels in the deficient range ( 30 nmol/l) with high-range normal PTH levels or secondary hyperparathyroidism. While there is increased awareness about the importance of calcium nutrition for skeletal health, these data suggest that deficient vitamin D levels should be of concern in women admitted with hip fractures. Circulating calcium and phosphate levels were lower in women with hip fractures compared with those admitted for elective joint replacement, consistent with vitamin D deficiency, impaired calcium and phosphate absorption, and higher PTH levels. The higher PTH levels, in turn, partially 1508 JAMA, April 28, 1999 Vol 281, No American Medical Association. All rights reserved.

5 raise circulating calcium to the lower limit of the normal range and decrease the serum phosphate concentration. Low albumin levels in women with hip fractures compared with the elective groups may reflect nutritional status and are an independent predictor of vitamin D deficiency in hospitalized subjects. 16 Although we excluded patients from nursing homes and with comorbid conditions affecting bone density, it is possible that the fracture patients were more frail, had poorer nutritional status, and were exposed to less sunlight than the control group. Data were analyzed adjusting for estrogen replacement, as indicated. Estrogen can affect bone and calciotropic hormones but not 25-hydroxyvitamin D levels. Estrogen potentially can affect PTH levels, but in women who were postmenopausal for less than 20 years (as in the case, on average, for the elective subjects), PTH levels increase with age. In women with more than 20 years since menopause, similar to the women with hip fractures, estrogen may lower PTH levels. Thus, estrogen use cannot account for higher PTH levels in women with acute hip fracture compared with the elective groups. According to Khosla et al, 17 levels of vitamin D predict PTH levels regardless of estrogen therapy or menopausal status. In our study, the increase in PTH levels in postmenopausal women with hip fractures cannot be explained by differences in creatinine level; the groups were similar and no patient had a creatinine level above 159 µmol/l (1.8 mg/ dl), a level at which intact PTH levels increase to more than 6.84 pmol/l. Levels of 1,25-dihydroxyvitamin D did not differ between the women with hip fractures and those admitted for elective joint replacement. The conversion of 25-hydroxyvitamin D to 1,25- dihydroxyvitamin D in the kidney may be augmented in patients with hip fractures by the stimulatory effects of higher PTH and lower phosphate levels on renal 1-hydroxylation of vitamin D. 18 Levels of vitamin D metabolites presented herein reflect the vitamin D status at the time of hip fracture. One large study found low levels of 1,25-dihydroxyvitamin D but not 25-hydroxyvitamin D in women with hip fractures, although the vitamin D metabolites were obtained approximately 3.9 years before the fracture. 19 Others did not find low levels of 1,25-dihydroxyvitamin D in women with hip fractures after correction for a decrease in vitamin D-binding protein. Osteoporosis results from an imbalance between bone resorption and bone formation. The development of markers of bone turnover such as the urinary N-telopeptide level, which is an index of bone resorption, and serum osteocalcin and BSAP, which are bone formation markers, make it possible to estimate bone remodeling noninvasively. In our study, both the osteoporotic women admitted for elective orthopedic surgery and those with hip fractures and comparable axial bone Table 3. Calciotropic Hormones and Biochemical Values* P Value Laboratory Test Value (Normal Range for Test) No. Control Control vs vs Calcium ( mmol/l ( ) 2.39 ( ) 2.33 ( ) [ mg/dl]) [9.70 ( )] [9.55 ( )] [9.3 ( )] Albumin (37-54 g/l) (36-45) 44 (39-51) 38 (27-48) Phosphate ( mmol/l ( ) 1.36 ( ) 1.10 ( ) [ mg/dl]) [3.8 ( )] [4.2 ( )] [3.4 ( )] Intact parathyroid hormone ( ) 3.79 ( ) 5.58 ( ) ( pmol/l [10-65 pg/ml]) 25-Hydroxyvitamin D ( ( ) 55.0 ( ) 32.4 ( ) nmol/l [9-43 ng/ml]) 1,25-Dihydroxyvitamin D ( ( ) 98.8 ( ) 80.6 ( ) pmol/l [16-65 pg/ml]) Urinary calcium ( mmol/d ( ) 3.28 ( ) 2.1 ( ) [ mg/24 h]) [139.5 ( )] [131.2 ( )] [84.0 ( )] N-Telopeptides (20-76 nmol/l bone ( ) 79.0 ( ) 64.5 ( ) collagen equivalents per mmol/l creatinine) Osteocalcin (6-16 µg/l) ( ) 11.2 ( ) 4.7 ( ) Bone specific alkaline phosphatase ( ng/ml) ( ) 18.0 ( ) 16.0 ( ) ( ng/ml) ( ) 13.8 ( ) 10.4 ( ) Creatinine (61-99 µmol/l ( ) 64.8 ( ) 68.6 ( ) [ mg/dl]) [0.90 ( )] [0.85 ( )] [0.90 ( )] *Values measured as mean (percentile [5th and 95th] and adjusted for age and estrogen intake. Calcium corrected for albumin levels. Log-transformed data analyzed. Product of Metra Biosystems, San Diego, Calif. Product of Hybritech, San Diego, Calif American Medical Association. All rights reserved. JAMA, April 28, 1999 Vol 281, No

6 density and years since menopause had higher urinary N-telopeptide levels compared with the elective women without osteoporosis. Levels of N-telopeptide, osteocalcin, and BSAP increase with age and increased bone turnover is a major determinant of osteoporosis in the elderly. 20 Osteocalcin and BSAP levels were lower in the women with fractures than women with osteoporosis in the elective group. This low osteocalcin level in the women with hip fractures compared with the osteoporotic elective group, in conjunction with higher N-telopeptide levels, suggests an uncoupling of bone formation, with increased bone resorption that may be a consequence of vitamin D deficiency. Another study showed that urinary excretion of pyridinolines was higher in elderly patients with vitamin D deficiency and elevated PTH levels than in healthy young adults. 21 Studies by Akesson et al 22 showed lower osteocalcin levels and higher urinary pyridinoline levels in women with hip fractures than in age-matched controls. Several lines of evidence indicate that subclinical vitamin D deficiency may contribute to increased fractures. First, there is seasonal variation in vitamin D levels, with a decrease in winter and spring that parallels the increased hip fracture rates during those seasons. 23 Second, the increase in PTH levels due to vitamin D deficiency leads to PTHmediated bone loss. Third, vitamin D and calcium therapy reduce PTH and lead to a decrease in markers of bone resorption, thereby reducing bone breakdown 24,25 and fractures. 24,26 Low vitamin D levels and elevated PTH levels have been reported in patients with hip fractures from Europe, Scandinavia, and Australia However, in the United States, milk is fortified with vitamin D so the intake of vitamin D is greater, and vitamin D levels are higher than in Scandinavia and Europe. 30 Peacock et al 31 determined that a level of 25-hydroxyvitamin D of less than 9.98 nmol/l can cause osteomalacia. Other studies show that a level of 25-hydroxyvitamin D below 30.0 nmol/l is a threshold for decreased bone density. 32 Results may not be comparable, however, because vitamin D assays used in earlier studies from Europe and Amsterdam had lower values than the RIA assays now approved by the Food and Drug Administration and often used in the United States. 33,34 International standards for vitamin D assays and for normality are needed for application to clinical care of patients with metabolic bone disorders. The decision to treat patients with low 25-hydroxyvitamin D levels depends on the risk of a rise in PTH levels that may have deleterious effects on the skeleton. For example, PTH levels increase well before the 25-hydroxyvitamin D level decreases to 30.0 nmol/l (or 22.5 nmol/l, the lower limit of normal for some assays). A few individuals were identified with 25-hydroxyvitamin D levels between 30.0 and 54.9 nmol/l and with elevated PTH levels. According to different models, several 25-hydroxyvitamin D threshold levels for PTH have been proposed. Ooms et al 35 reported that 25-hydroxyvitamin D levels of 25.0 nmol/l are negatively related to PTH and low bone density at a cutoff of 30 nmol/l. In elderly nursing home patients including those in Boston, PTH levels were elevated with 25-hydroxyvitamin D concentrations less than 37.4 nmol/l, thus vitamin D depletion was indicated by a 25-hydroxyvitamin D level below 37.4 nmol/l. 3,36 Peacock et al 31 defined vitamin D deficiency as 25- hydroxyvitamin D levels between 9.98 and 49.9 nmol/l, when there may be mild hyperparathyroidism and decreased bone density. 31 Chapuy et al 37 studied 1569 French subjects and found that PTH levels were stable when the 25- hydroxyvitamin D level was higher than 77.4 nmol/l, using the same RIA vitamin D assay used in our study. Haden et al 38 also showed that PTH levels start to increase at 25-hydroxyvitamin D levels of less than 62.4 nmol/l. Thus, levels of 25-hydroxyvitamin D may need to be higher than 49.9 to 73.4 nmol/l to protect against secondary hyperparathyroidism. Our study was not designed to assess seasonal changes and it did not equally span all seasons in the study interval. The study may have underestimated the frequency of low vitamin D levels in women with hip fractures by excluding 190 subjects with comorbid conditions and fractures who may have had reduced sun exposure and vitamin D intake (Table 1). A recent study showed that 57% of hospitalized patients had low vitamin D levels. 16 In conclusion, this study shows occult vitamin D deficiency in postmenopausal women with hip fractures and no other secondary causes of bone loss in the northeastern United States. Correction of vitamin D deficiency or insufficiency with vitamin D supplementation or sun exposure may lead to a reduction in the exponential rise of fractures that occur with age. The new 1997 Dietary Reference Intake guidelines from the National Institutes of Health recommend 400 IU of vitamin D supplement daily for individuals from age 51 through 70 years and 600 IU daily for those older than 70 years. Supplements of about 800 IU of vitamin D per day and calcium may be necessary to attenuate bone loss in the winter and to reduce fractures. 24,26 These findings may play an important role in the orthopedic management of acute fractures in postmenopausal women. In patients with hip fractures, vitamin D repletion at the time of fracture and resultant suppression of hyperparathyroidism may facilitate hip fracture repair, implant incorporation, and reduction of future fracture risk. Vitamin D deficiency is preventable and heightened awareness is necessary to institute public health programs to ensure adequate vitamin D nutrition in the elderly, particularly in northern latitudes. Financial Disclosure: At the time of the study, Dr LeBoff served on the Scientific Advisory Board for Incstar, which manufactures a vitamin D assay, but she is no longer a member of the board. Incstar did not provide any funding for this study, and the study is not about any technology that is owned by or contractually obligated to Incstar in any way. Funding/Support: This study was supported in part by National Institutes of Health grants R01 AG and R01 AG (Drs LeBoff and Glowacki) and NCRR General Clinical Research Center to the Brigham and Women s Hospital General Clinical Research Cen JAMA, April 28, 1999 Vol 281, No American Medical Association. All rights reserved.

7 ter, and a National Osteoporosis Foundation Fellowship Award (Dr Kohlmeier). Acknowledgment: We greatly appreciate the expert assistance of Irma Sabbag and Natalie Glass and other members of the Skeletal Health and Osteoporosis Program at Brigham and Women s Hospital, and the important contributions of Clement Sledge, MD, Thomas Thornhill, MD, and Benjamin Bierbaum, MD, and the physicians in the Department of Orthopedics at Brigham and Women s Hospital and New England Baptist Hospital. We appreciate Caren Gundberg, PhD, of Yale Medical Center for her expertise with the osteocalcin and BSAP (Hybritech) assays and Beverly Potter for overseeing quality assurance of the tests in our General Clinical Research Center Laboratory. REFERENCES 1. Looker AC, Orwoll ES, Johnston CC Jr, et al. Prevalence of low femoral bone density in older US adults from NHANES III. J Bone Miner Res. 1997;12: Ray NF, Chan JK, Thamer M, Melton LJ III. Medical expenditures for the treatment of osteoporotic fractures in the United States in 1995: report from the National Osteoporosis Foundation. J Bone Miner Res. 1997;12: Parfitt AM, Chir B, Gallagher JC, et al. Vitamin D and bone health in the elderly. Am J Clin Nutr. 1982; 36: Aaron JE, Gallagher JC, Anderson J, et al. Frequency of osteomalacia and osteoporotic fractures of the proximal femur. Lancet. 1974;1: Chalmers J, Barclay A, Davison AM, MacLeod DAD, Williams DS. Quantitative measurements of osteoid in health and disease. Clin Orthop. 1969;63: Solomon L. of the femoral neck in the elderly: bone ageing or disease? S Afr J Surg. 1973;11: Hordon LD, Peacock M. Osteomalacia and osteoporosis in femoral neck fracture. Bone Miner. 1990; 11: Lips P, Netelenbos JC, Jongen MJM, et al. Histomorphometric profile and vitamin D status in patients with femoral neck fracture. Metab Bone Dis Rel Res. 1982;4: Sokoloff L. Occult osteomalacia in American (USA) patients with fracture of the hip. Am J Surg Pathol. 1978;2: Johnston CC, Norton J, Khairi MRA, et al. Heterogeneity of fracture syndromes in postmenopausal women. J Clin Endocrinol Metab. 1985;61: Kanis JA, for the WHO Study Group. Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: synopsis of a WHO report. Osteoporos Int. 1994;4: Carey VJ, Walters EE, Colditz GA, et al, for the Nurses Health Study. Body fat distribution and risk of non insulin-dependent diabetes mellitus in women. Am J Epidemiol. 1997;145: Kriska AM, Bennett PH. An epidemiological perspective of the relationship between physical activity and NIDDM: from activity assessment to intervention. Diabetes Metab Rev. 1992;8: El-Hajj Fuleihan G, Testa MS, Angell JE, Porrino N, LeBoff MS. Reproducibility of DXA absorptiometry: a model for bone loss estimates. J Bone Miner Res. 1995;10: Gundberg CM, Grant FD, Conlin PR, et al. Acute changes in serum osteocalcin during induced hypocalcemia in humans. J Clin Endocrinol Metab. 1991; 72: Thomas MK, Lloyd-Jones DM, Thadhani RI, et al. Hypovitaminosis D in medical inpatients. N Engl J Med. 1998;338: Khosla S, Atkinson EJ, Melton LJ III, Riggs BL. Effects of age and estrogen status on serum parathyroid hormone levels and biochemical markers of bone turnover in women: a population based study. J Clin Endocrinol Metab. 1997;82: Compston JE, Silver AC, Croucher PI, Brown RC, Woodhead JS. Elevated serum intact parathyroid hormone levels in elderly patients with hip fracture. Clin Endocrinol. 1989;31: Cummings SR, Browner WS, Bauer D, et al. Endogenous hormones and the risk of hip and vertebral fractures among older women. N Engl J Med. 1998; 339: Garnero P, Sornay-Rendu E, Chapuy M, Delmas PD. Increased bone turnover in late menopausal women is a major determinant of osteoporosis. J Bone Miner Res. 1996;11: Kamel S, Brazier M, Picard C, et al. Urinary excretion of pyridinolines crosslinks measured by immunoassay and HPLC techniques in normal subjects and in elderly patients with vitamin D deficiency. Bone Miner. 1994;26: Akesson K, Vergnaud P, Gineyts E, Delmas PD, Obrant KJ. Impairment of bone turnover in elderly women with hip fracture. Calcif Tissue Int. 1993;53: Lips P, Hackeng WHL, Jongen MJM, Van Ginkel FC, Netelenbos JC. Seasonal variation in serum concentrations of parathyroid hormone in elderly people. J Clin Endocrinol Metab. 1983;57: Chapuy MC, Arlot ME, Duboeuf F, et al. Vitamin D 3 and calcium to prevent hip fractures in elderly women. N Engl J Med. 1992;327: Prestwood KM, Pannullo AM, Kenny AM, Pillbeam CC, Raisz LG. The effect of a short course of calcium and vitamin D on bone turnover in older women. Osteoporos Int. 1996;6: Dawson-Hughes B, Harris SS, Krall EA, Dallal GE. Effect of calcium and vitamin D supplementation on bone density in men and women 65 years of age or older. N Engl J Med. 1997;337: Boonen S, Aerssens J, Dequeker J. Age-related endocrine deficiencies and fractures of the proximal femur, II: implications of vitamin D deficiency in the elderly. J Endocrinol. 1996;149: Harma M, Parviainen M, Koskiinen T, Voikka V, Alhava E. Bone density, histomorphometry and biochemistry in patients with fracures of the hip or spine. Ann Clin Res. 1987;19: Baker MR, McDonald H, Peacock M, Nordin BEC. Plasma 25-hydroxyvitamin D concentrations in patients with fractures of the femoral neck. BMJ. 1979; 1: McKenna MJ. Differences in vitamin D status between countries in young adults and the elderly. Am J Med. 1992;93: Peacock M, Selby PL, Francis RM, Brown WB, Hordon L. Vitamin D deficiency, insufficiency, sufficiency, and intoxication, what do they mean? In: Schaefer NA, Grigoletti MG, Herrath DV, eds. Sixth Workshop on Vitamin D. Berlin, Germany: DeGruyter; 1985: Bouillon RA, Auwerx JH, Lissens WD, Pelemans WK. Vitamin D status in the elderly: seasonal substrate deficiency causes 1,25-dihydroxycholecalciferol deficiency. Am J Clin Nutr. 1987;45: Lips P, Chapuy MC, Dawson-Hughes B, Pols HAP. International comparison of serum 25-hydroxyvitamin D measurements. J Bone Miner Res. 1995;10 (suppl):s Jongen MJM, VanGinkel FC, van der Vijgh WJF, Kuiper S, Netelenbos JC, Lips P. An international comparison of vitamin D metabolite measurements. Clin Chem. 1984;30: Ooms ME, Lips P, Roos JC, et al. Vitamin D status and sex hormone binding globulin: determinants of bone turnover and bone mineral density in elderly women. J Bone Miner Res. 1995;10: Webb A, Pillbeam C, Hanafin, Hollick M. An evaluation of the relative contributions of exposure to sunlight and of diet to the circulating concentrations of 25-hydroxyvitamin D in an elderly nursing home population in Boston. Am J Clin Nutr. 1990;51: Chapuy MC, Preziosi P, Maamer M, et al. Prevalence of vitamin D insufficiency in an adult normal population. Osteoporos Int. 1997;7: Haden S, Fuleihan GE, Angell JE, Cotran NM, Le- Boff MS. Calcidiol and PTH levels in women attending an osteoporosis program. Calcif Tissue Int. 1999; 64: American Medical Association. All rights reserved. JAMA, April 28, 1999 Vol 281, No