Abstract
Background: Recent studies suggest that high homocysteine levels are associated with an increased risk of fractures. Homocysteine levels are known to be influenced by vitamin B and folate supply or status, and poor renal function can result in higher levels independent of nutritional adequacy.
Objective: The aim of the study was to determine the associations between fasting homocysteine levels and incident hip fractures, and the effects of other factors on hip fracture risk.
Design: We conducted a case-control study in the Women’s Health Initiative Observational Study, a study of postmenopausal women (n = 93,676) recruited in the United States. We selected 400 incident cases of hip fracture and 400 controls matched on age, ethnicity, and blood draw date among women not on osteoporosis therapies. Outcome measures included physician-adjudicated, incident hip fractures. Baseline lifestyle and nutritional questionnaires were performed.
Results: The risk of hip fracture increased 1.38-fold [95% confidence interval (CI), 1.14, 1.66] for each sd increase in serum homocysteine level after adjustment for fracture risk factors. This association was not affected by adjustment for dietary folate, B6, or B12 intake, but it diminished after adjustment for cystatin-C level (odds ratio, 1.08; 95% CI, 0.66–1.79), a measure of renal function not affected by muscle mass. Among women in the highest quartile of homocysteine and cystatin-C compared to those without elevations in either biomarker, the risk of hip fracture was substantially elevated (odds ratio, 2.8; 95% CI, 1.61–4.87).
Conclusions: This study indicates that high homocysteine levels are associated with an increased risk of hip fracture, which could be accounted for by poor renal function.
High homocysteine levels are associated with an increased risk of hip fractures, which could be accounted for by poor renal function.
Osteoporosis is a major public health problem in the United States, especially in women. Nutritional, lifestyle, hormonal, and genetic factors contribute to the risk of fractures. The risk of osteoporotic fractures rises exponentially with age, resulting in 2 million cases of fractures annually in the United States, including 300,000 hip fractures, the most serious fracture (1). The health consequences of hip fractures are enormous and include progressive loss of function, inability to ambulate independently in more than half of those affected, and mortality in up to 24% of women (2). Although 44 million Americans currently have osteoporosis and/or low bone mass, the recent Surgeon General’s Report on Osteoporosis projects that the number of Americans affected will rise to 52.4 million by the year 2010 (3).
High homocysteine levels have been associated with an increased risk of fractures, although the factors that contribute to this fracture risk are not fully elucidated. Homocysteine is an intermediary amino acid formed by the conversion of methionine to cysteine. Homocystinuria is a rare autosomal recessive disease characterized by increased circulating homocysteine levels, with clinical manifestations involving the eyes, vasculature and nervous system, and skeleton including the development of early osteoporotic fractures (4,5,6). Adults without homocystinuria who have high homocysteine levels are also at risk for fractures (7,8,9). Nutritional factors such as vitamin B12, B6, and folate are cofactors in homocysteine metabolism, and vitamin intakes may inversely affect plasma homocysteine levels (10,11,12). Gene polymorphisms related to homocysteine metabolism also may result in high homocysteine levels. Thus, nutritional factors such as B12/folate deficiency and genetic factors may affect homocysteine levels and contribute to fracture risk.
Chronic kidney disease affects an estimated 19 million Americans (13). Recent data show that glomerular filtration rate as measured by cystatin-C is an important determinant of plasma homocysteine. Lewerin et al. (14) conducted a population-based study among 209 community-dwelling subjects with mean age of 76 yr. They found that cystatin-C levels were correlated with plasma homocysteine concentrations (r = 0.45; P < 0.001) independent of vitamin B (12) and folate status (14). Additionally, Fried et al. (15) recently reported that kidney dysfunction, as assessed by cystatin-C, is associated with an increased risk of hip fracture.
To advance our understanding of the mechanisms that lead to osteoporotic hip fractures, we used the Women’s Health Initiative Observational Study (WHI-OS) to determine the relationship between homocysteine levels and the risk of incident hip fractures in the largest study of postmenopausal women with hip fractures. We then examined both dietary intakes of vitamin B6, B12, and folate and renal function as measured by cystatin-C levels, as potential explanatory factors for any observed association.
Subjects and Methods
The WHI-OS is a multicenter; multiethnic study designed to address the common causes of death, disability, and impaired quality of life among postmenopausal women in the United States. The study includes 93,676 women aged 50 to 79 yr at entry between 1993 and 1998 from 40 clinical centers in the United States. Details of the scientific rationale, eligibility requirements, and other aspects of the study design of the WHI-OS have been published previously (16). Women completed screening and enrollment questionnaires and underwent a physical examination, and a fasting blood specimen was collected at baseline. Blood was processed for long-term storage at −70 C. The study was reviewed and approved by human subjects review committees at each participating institution, and written, informed consent was obtained from all women enrolled in the study.
For this current analysis, we excluded participants with a history of hip fracture and those who were on certain medications including estrogen therapy, androgens, and selective estrogen receptor modulators, bisphosphonates, calcitonins, and PTH at baseline. In addition, we excluded women with confirmed or possible pathological cause for hip fractures occurring during the study, hip fractures with only local adjudication, unknown ethnicity, and inadequate available plasma.
Assessment of outcome
Data on the incidence of total fractures or radiologically confirmed fractures of the hip, spine, forearm, or wrist were collected at baseline and annually to ascertain the occurrence of a new fracture. Participants were asked annually whether they had a fracture since they last completed the medical history questionnaire and, if so, to identify the location (vertebral, shoulder, upper arm, lower arm, wrist, hip, upper leg, lower leg, or foot). All hip fractures were centrally adjudicated. A validation study of self-reported fractures was conducted on the WHI-OS and found good agreement between self-report and adjudicated fracture for hip (78%) (17).
Nested case-control design
Using this baseline population, we conducted a nested case-control study within the WHI-OS. The study population included 400 cases of incident hip fracture and 400 controls. For each case we selected one control using the risk set sampling technique described by Prentice and Breslow (18); i.e. for each incident hip fracture case, a control was selected at random from a list of all women at risk of the outcome. In addition, controls were matched on race/ethnicity, age (within 1 yr), and date of blood draw (within 4 months).
Data collection
Laboratory measurements
Total homocysteine levels (fasting samples) were measured at baseline in serum using a HPLC assay (Medical Research Laboratories International, Highland Heights, KY). The coefficient of variation was 7.3–7.6%, with a range of 5–15 μmol/liter. The laboratory personnel were blinded to case-control status, and samples were analyzed in random order.
Serum cystatin-C was measured using the Dade Behring BN-II nephelometer with a particle-enhanced immunonepholometric assay (GMI, Inc., Ramsey, MN) (interassay coefficient of variation, 5.7%; sensitivity, 0.02 mg/liter).
Hip fracture risk factors and dietary variables
Cohort members completed detailed self-administered questionnaires on demographic information, medical history, reproductive history, nutrition, smoking and alcohol, family history, personal habits, and recreational physical activity. Ethnicity was self-identified as white not of Hispanic origin, African-American, Hispanic, American Indian, Alaskan Native, Asian or Pacific Islander, or unknown. Smoking status was determined from lifetime smoking of at least 100 cigarettes, current daily cigarette smoking, and self-report of smoking cessation. Alcohol consumption was computed from the personal habits questionnaire. Total dietary and supplemental intake of folate and vitamins B6 and B12 was determined from the food frequency questionnaires and the supplement use questionnaires; nutrient intake analysis from the food frequency questionnaire data was adjusted with folic acid fortification of the food supply, which was initiated in 1996 (19).
Anthropometric measurements were obtained at baseline by trained and certified Clinical Center staff. Weight was measured to the nearest 0.1 kg, and height was recorded to the nearest 0.1 cm. Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters. Physical activity was assessed by questions on the frequency and duration of several types of recreational activity, and metabolic equivalent (MET) scores were computed as the product of days per week, minutes per day, and the MET value for each activity (20).
Statistical analysis
All statistical analyses were performed with SAS version 9.1 (SAS Institute, Inc., Cary, NC). For all analyses, we categorized homocysteine into quartiles based on the distribution in the control group. The homocysteine levels greater than 50 μmol/liter were determined to be biologically implausible for postmenopausal women and were set to missing during the analysis.
Conditional logistic regression model was used to examine the association between the quartiles of homocysteine and the risk of hip fracture. Multivariate models were used to account for the potential confounders including BMI, parental history of hip fracture, treated diabetes, alcohol use, smoking, and history of stroke.
We estimated the odds ratios (OR) and the 95% confidence intervals (CI) using the lowest quartile as the reference category. Tests for trend were conducted based on continuous scale of the log-transformed plasma homocysteine value and also by the ordinal value of each quartile. All P values reported are two-tailed, and values less than 0.05 were considered statistically significant. We also examined the association between hip fracture, homocysteine, and other possible mediators by the additional adjustment for folate, vitamins B6 and B12, and cystatin-C one at a time and jointly. The relationship between homocysteine and cystatin-C was further explored using models including both biomarkers as continuous variables and then as quartiles.
Results
Table 1 shows the comparison of baseline characteristics in cases and controls. As expected, controls were slightly heavier but were more physically active compared with cases. There were no differences in baseline characteristics between cases and controls and treated diabetes or a history of myocardial infarction, which are also associated with high homocysteine levels (Table 1). Cases had lower calcium intake, smoked more, and tended to have an increased stroke history at baseline compared with controls in the study. The mean baseline homocysteine concentrations were higher in hip fracture cases as compared with controls (11.7 vs. 10.7 μmol/liter; P = 0.001).
Table 1.
Comparison of baseline characteristics among cases of hip fractures and controls
Characteristics | Control | Case | Pvalue |
---|---|---|---|
Age at screening (yr) | 70.8 ± 6.2 | 70.8 ± 6.2 | 0.986 |
BMI (kg/m2) | 27.4 ± 5.1 | 26.0 ± 5.2 | 0.000 |
Total energy expended from physical activity (MET) | 13.9 ± 15.3 | 10.7 ± 12.7 | 0.002 |
Years since menopause | 22.0 ± 9.0 | 22.3 ± 8.8 | 0.644 |
Total calcium intake (mg) | 1167.0 ± 683.9 | 1072.5 ± 694.2 | 0.053 |
Total vitamin D intake (μg) | 9.3 ± 6.9 | 9.6 ± 9.9 | 0.707 |
Total folate intake (μg) | 443.9 ± 267.6 | 450.4 ± 320.9 | 0.757 |
Total B12 intake (μg) | 22.8 ± 80.9 | 24.6 ± 76.7 | 0.753 |
Homocysteine (μmol/liter) | 10.7 ± 3.2 | 11.7 ± 4.7 | 0.001 |
Cystatin-C (mg/liter ) | 1.06 ± 0.24 | 1.10 ± 0.30 | 0.023 |
Alcohol intake (>1 drink per day) (%) | 10.6 | 9.8 | 0.608 |
Current smokers (%) | 2.5 | 9.1 | 0.000 |
History of treated diabetes (%) | 4.8 | 6.0 | 0.433 |
History of myocardial infarction (%) | 3.5 | 5.5 | 0.170 |
History of stroke (%) | 2.0 | 4.5 | 0.046 |
Parental history of hip fracture (%) | 16.0 | 20.0 | 0.141 |
Data are expressed as mean ± sd unless indicated otherwise.
Table 2 shows the OR and 95% CI for the risk of incident hip fracture by quartiles of homocysteine levels. The OR for a hip fracture for each increase in sd of plasma homocysteine was 1.36 (95% CI, 1.13, 1.63). In the conditional logistic regression model accounting for the matched design and additionally adjusted for BMI, the OR between extreme quartiles of homocysteine was 1.49 (95% CI, 0.98, 2.29). Further adjustment of several potential confounders including parental history of hip fracture, treated diabetes, alcohol use, smoking, history of stroke, and total calcium intake resulted in an OR of 1.37 (95% CI, 0.86, 2.19). The test of linearity based on log-transformed continuous scale and across quartiles produced somewhat different P for linear trend (Table 2). Therefore, the results from both methods are included in the table. The CIs for the quartiles show the precision of the quartiles. The significance of our results was determined according the P value for trend.
Table 2.
OR (95% CI) for the risk of hip fracture (cases/controls) by quartiles of plasma homocysteine and the relative risk for each quartile compared to the lowest quartile
Homocysteinea | Unadjusted | BMI-adjusted | MV-adjustedb |
---|---|---|---|
Per sd (4.06 μmol/liter) increase | 1.31 (1.12, 1.53) | 1.36 (1.16, 1.60) | 1.36 (1.13, 1.63) |
P1 for linear trend (log-transformed) | P = 0.002 | P < 0.001 | P = 0.002 |
Per μmol/liter increase | 1.07 (1.03, 1.11) | 1.08 (1.04, 1.12) | 1.07 (1.02, 1.13) |
No. of missing pairs (total = 400) | 7 | 11 | 24 |
Quartiles (according to control group) | |||
P2 for linear trend across quartiles | P = 0.078 | P = 0.033 | P = 0.102 |
1st quartile (<8.4 μmol/liter) (reference) | 1 | 1 | 1 |
2nd quartile (8.4 to <10.0 μmol/liter) | 0.88 (0.58, 1.32) | 0.91 (0.60, 1.39) | 0.85 (0.54, 1.35) |
3rd quartile (10.0 to <12.5 μmol/liter) | 1.10 (0.75, 1.63) | 1.12 (0.75, 1.68) | 1.10 (0.71, 1.69) |
4th quartile (≥12.5 μmol/liter) | 1.34 (0.89, 2.01) | 1.49 (0.98, 2.29) | 1.37 (0.86, 2.19) |
P1, Test of linear trend based on continuous scale and log-transformed plasma homocysteine value; P2, test of linear trend based on ordinal values of the homocysteine quartile; MV, multivariate.
Hip fracture case and control selection matched on age, ethnicity, and blood draw date.
Multivariate adjustment includes BMI, parental history of hip fracture, treated diabetes, alcohol use, smoking, history of stroke, and total calcium intake.
Table 3 shows the ORs for risk of hip fracture by quartiles of homocysteine with additional adjustment for levels of cystatin-C and intakes of folate and vitamins B6 and B12. The addition of folate and vitamins B6 and B12 intake in the multivariate model did not change the OR for homocysteine substantially. Adjustment for cystatin-C attenuated the association of homocysteine and risk of hip fractures. The table also presents the correlations between homocysteine and folate, vitamins B6 and B12, and cystatin-C. As anticipated, there was a negative linear correlation between homocysteine levels and folate intake (r = −0.25; P < 0.0001) and B6 and B12 intakes. There was also a positive correlation between homocysteine and cystatin-C (r = 0.45; P < 0.0001).
Table 3.
Adjusted ORs for hip fractures, according to baseline homocysteine quartiles (μmol/liter)
(Total pairs = 400) | rb | 1st (<8.4) | 2nd (8.4 to <10.0) | 3rd (10.0 to <12.5) | 4th (≥12.5) | P-trend1 | P-trend2 |
---|---|---|---|---|---|---|---|
Base analysisc | |||||||
OR (95% CI) (n = 376)a | 1.0 | 0.85 (0.54–1.35) | 1.10 (0.71–1.69) | 1.37 (0.86–2.19) | 0.002 | 0.102 | |
P | 0.500 | 0.670 | 0.189 | ||||
Base analysisc + adjusted for total folate intaked | |||||||
OR (95% CI) (n = 376)a | −0.248 | 1.0 | 0.86 (0.54–1.36) | 1.12 (0.73–1.73) | 1.40 (0.87–2.25) | 0.002 | 0.085 |
P | <0.0001 | 0.513 | 0.608 | 0.164 | |||
Base analysisc + adjusted for total vitamin B6 intaked | |||||||
OR (95% CI) (n = 376)a | −0.041 | 1.0 | 0.87 (0.55–1.37) | 1.10 (0.72–1.70) | 1.38 (0.87–2.21) | 0.002 | 0.098 |
P | 0.254 | 0.540 | 0.651 | 0.175 | |||
Base analysisc + adjusted for total vitamin B12 intaked | |||||||
OR (95% CI) (n = 376)a | −0.057 | 1.0 | 0.86 (0.54–1.36) | 1.11 (0.72–1.71) | 1.38 (0.86–2.20) | 0.003 | 0.096 |
P | 0.111 | 0.515 | 0.636 | 0.181 | |||
Base analysisc + adjusted for plasma cystatin-C | |||||||
OR (95% CI) (n = 372)a | 0.452 | 1.0 | 0.79 (0.50–1.26) | 0.96 (0.62–1.49) | 1.08 (0.66–1.79) | 0.070 | 0.580 |
P | <0.0001 | 0.326 | 0.849 | 0.756 | |||
Base analysisc + adjusted for total folate,d B6,d B12,d and cystatin-C | |||||||
OR (95% CI) (n = 372)a | 1.0 | 0.81 (0.51–1.28) | 0.98 (0.62–1.53) | 1.11 (0.67–1.85) | 0.050 | 0.516 | |
P | 0.367 | 0.923 | 0.678 | ||||
Base analysisc + adjusted for RAND 36 physical functioning >90 | |||||||
OR (95% CI) (n = 352)a | −0.161 | 1.0 | 0.91 (0.57–1.46) | 1.12 (0.72–1.75) | 1.44 (0.88–2.35) | 0.003 | 0.088 |
P | <0.0001 | 0.691 | 0.622 | 0.144 |
P-trend1, Test of linear trend based on continuous scale and log-transformed plasma homocysteine value; P-trend2, test of linear trend based on ordinal values of the homocysteine quartile.
n indicates number of case-control pairs included in the analysis.
Pearson correlation coefficients with homocysteine.
Matched on age, ethnicity, blood draw date; controlled for BMI, parental history of hip fracture, treated diabetes, smoking, alcohol use, history of stroke, and total calcium intake.
Total intake from diet and supplement.
We also conducted a secondary analysis that shows the ORs for risk of hip fracture by quartiles of homocysteine and cystatin-C (Table 4). The odds ratio for women in the highest quartiles of both homocysteine and cystatin-C is 2.8 (1.61, 4.87; P for interaction = 0.026) compared with women in the lower three quartiles of both homocysteine and cystatin-C after adjusting for BMI, parental history of hip fracture, treated diabetes, alcohol use, smoking, history of stroke, total calcium intake, and total intake of folate and vitamins B6 and B12. The interaction of homocysteine and cystatin-C levels was not statistically significant when the biomarkers were studied as continuous variables or in quartile form.
Table 4.
Adjusted ORs for hip fractures, according to quartiles of homocysteine (μmol/liter) and cystatin-C (mg/liter)
Cystatin-C quartiles | Homocysteine quartilesc
|
|||
---|---|---|---|---|
1st (<8.4) | 2nd (8.4 to <10.0) | 3rd (10.0 to <12.5) | 4th (≥12.5) | |
4th (≥1.15) | 1.10 (0.68, 1.77)a | 2.80 (1.61, 4.87)a,b | ||
3rd (1.01 to <1.15) | 0.86 (0.53, 1.39)a | |||
2nd (0.91 to <1.01) | 1.00 (ref) | |||
1st (<0.91) |
Matched on age, ethnicity, blood draw date, and adjusted for BMI, parental history of hip fracture, treated diabetes, smoking, alcohol use, history of stroke, total calcium intake, folate, vitamin B6, and vitamin B12.
P < 0.001.
P for interaction between 4th quartile vs. not of cystatin-C and 4th quartile vs. not of homocysteine = 0.026.
Discussion
The present study supports a modest association between high homocysteine levels and increased hip fracture risk (7,8,9). In this study of postmenopausal women enrolled in the WHI-OS, every sd increase in serum homocysteine levels was associated with a significant 36% elevated risk of hip fracture. The quartile analysis revealed that most of this association was concentrated in the upper quartile. Our results remained unchanged after additional adjustment for total intake of folate and vitamins B6 and B12. Adjustment for cystatin-C, however, diminished the homocysteine association; women in the highest quartile of both homocysteine and cystatin-C had a much larger, 2.8-fold risk of suffering hip fracture compared with women without elevations in either biomarker. These data suggest that high homocysteine levels are associated with an increased risk of hip fracture, especially in women with poor renal function as assessed by cystatin-C levels rather than dietary intakes of vitamins B6 and B12 or folate.
Moderately elevated homocysteine levels increase with age and occur in about 5–7% of the population (21,22). Data from two different cohort studies show that high homocysteine levels are associated with risk for fractures (7,8,9). In the Framingham Study, hip fractures doubled in women (n = 146 hip fractures) who had homocysteine levels in the highest quartile (8). Of note, absolute homocysteine levels in the highest quartile of the women in that study were higher than those from our study. Nonfasting plasma samples for the Framingham study were collected between 1979 and 1982, and women were followed for incident hip fractures through 1998. In the WHI-OS, blood samples were collected in a fasting state between 1993 and 1998 (34% of blood drawn before 1996). Because foods were fortified with folic acid in the United States starting in 1996 (to reduce neural tube defects), the high homocysteine levels in the Framingham Study may be attributable, in part, to the different intakes of folic acid; fortification of foods may result in a reduction in high fasting homocysteine levels from 18.7 to 9.8% (10,11). In the Rotterdam and Longitudinal Aging Study Amsterdam studies, fractures doubled in women with homocysteine levels in the highest quartile (7). However, a recent population-based study in Sweden was conducted to examine the association between homocysteine and bone turnover, bone mineral density, and fracture risk. The authors reported that high homocysteine levels were associated with higher bone turnover and poor physical performance, but not total fracture risk (23).
In addition, the mechanisms through which high homocysteine levels affect bone have not been fully elucidated. Studies in vivo in animal models have shown that hyperhomocysteinemia leads to a disruption of collagen cross-links that may adversely affect bone formation and skeletal strength (24,25,26). Other studies indicate that high homocysteine levels are associated with increased markers of bone resorption, with a net increase in bone breakdown relative to formation (23,27). End-stage renal disease is associated with an increased risk of fractures in association with several metabolic bone disorders (28). Although bone density testing does not distinguish the type of renal osteodystrophy, the National Health and Nutrition Examination Survey III analyses indicate that 85% of women with glomerular rates less than 60 cc/min have osteoporosis according to bone density criteria (29). New data also show that mild to moderate kidney dysfunction, as assessed by a novel marker, cystatin-C, is associated with an increased risk for hip fracture (10). Recently, Lewerin et al. (14) reported that renal function as assessed by cystatin-C is also a determinant of plasma homocysteine levels. Our new data indicate the involvement of renal function in the relationship between homocysteine and hip fracture risk. It is important to note that previous studies reporting the relationship between homocysteine and fracture risk did not include cystatin-C in their analyses. Recent data from the WHI-OS that specifically examined the association of cystatin-C concentrations, levels of renal impairment, and hip fracture showed that hip fracture risk was elevated in women whose estimated glomerular filtration rates were less than 60 mg/ml per 1.73 m2 (30,31).
Cystatin-C is a low molecular weight protein produced by nucleated cells. It is filtered by the kidney and undergoes metabolism after tubular resorption. Unlike creatinine, cystatin-C is not affected by muscle mass and age. Thus, serum creatinine may not serve as the appropriate biomarker to precisely measure mild to moderate renal insufficiency (14,32,33). Our study confirms an association between homocysteine levels and hip fracture risk, but it indicates that the risk for hip fracture is no longer significant when these data are adjusted for cystatin-C level, an index of renal function. Hence, our findings indicate that homocysteine is associated with hip fracture only when renal function is impaired, and that homocysteine may be a mediator of the association between renal insufficiency and hip fracture or share a causal pathway. However, the present study was not designed to determine whether the two biomarkers share a common causal pathway or how each biomarker relates to the pathophysiology of osteoporosis and hip fracture.
As expected, our data show a negative linear correlation between homocysteine levels and intake of folate (r = −0.25); intakes of folate and B vitamins were similar in hip fracture cases and controls. Previous studies by Selhub et al. (11) reported an inverse relation between homocysteine and plasma folate. In another study by Ubbink et al. (34), there were significantly lower plasma concentrations of vitamin B6, B12, and folate in men with hyperhomocysteinemia.
Vitamin B12 status has also been linked to reduced bone mass in elderly women. Investigators for the Hordaland study recently reported that plasma folate, but not vitamin B12, was an independent predictor of hip fracture among women (31). Although the association between B vitamins/folate and homocysteine has been reported, there currently are not conclusive data to prove the relationship between folate and B vitamins and risk of fractures (35). One study reported that low serum folate and not homocysteine is responsible for the association between homocysteine and risk of osteoporotic fracture in elderly persons (36). Among 111 women and men with hip fractures, low levels of B vitamins were a risk factor for hip fractures, but the association of homocysteine and hip fractures was independent of these levels. More recently, Sato et al. (37) showed that in patients with strokes, supplementation with folate and B vitamins reduced hip fracture risk. Randomized, controlled studies are being conducted to understand better the effects of B vitamins and folate and reductions in homocysteine levels on fracture risk.
Our study includes a population of postmenopausal women from clinical centers all over the United States, thus enhancing the generalizability of the results. Strengths of our study are the inclusion of 1) the large number of adjudicated hip fractures (n = 400) and carefully matched controls (n = 400); and 2) dietary information and lifestyle variables, which enabled us to evaluate for intakes of folate and B vitamins and other confounding factors. A limitation of our study is that we did not measure serum levels of folate and B vitamins. Although there are advantages of using cystatin-C as a marker of renal function, a limitation is that we did not also measure creatinine levels. Another limitation is that there is a possibility of residual confounding by the variables that we cannot account for (38).
High homocysteine levels are associated with an increased risk of hip fracture and could be accounted for by poor renal function. These novel biomarkers may lead to a new understanding about the pathophysiology of hip fracture and how it can be prevented. Intervention trials targeting homocysteine levels should carefully evaluate the role of renal function when assessing the efficacy of nutritional supplementation.
Supplementary Material
Acknowledgments
The authors appreciate the contribution of the WHI-OS participants. M.S.L. thanks Mary Pettinger for her thoughtful contributions to the statistical analyses for this manuscript.
Footnotes
Funding support for this research was provided by the National Institutes of Health (NIH)-WHI Vanguard Center Grant (WHI-9219E) and NIH ROI Hip Fracture Umbrella Study. The WHI program is funded by the National Heart, Lung, and Blood Institute, U.S. Department of Health and Human Services. Support was also provided by NIH Grant Ancillary Study 90: 5 RO1 AR048919-04 and WHI Federal Contract NO1-WH-2-2110.
Trial Registration: The name of trial registry, registration number, and URL of the registry will be forthcoming.
These data were selected as an oral presentation at the Annual Meeting of The Endocrine Society in 2006 and at the World Congress on Hyperhomocysteinemia meeting in Saarbruecken, Germany, in June 2007.
M.S.L., A.L., and L.W. had full access to all the data in this study and take responsibility for the integrity of the data and, to the best of their knowledge, the accuracy of the data and the data analysis.
Disclosure Summary: R.N., A.L., L.W., R.J., J.L., D.C.B., C.K., C.L., A.M.T., and S.C. have nothing to declare. J.C. currently is working on projects involving research grants from Merck, Eli Lilly, Pfizer Pharmaceuticals, and Novartis Pharmaceuticals. M.S.L. works on active research projects that are supported by the National Institutes of Health. J.C. has received honorariums from Merck, Eli Lilly, and Novartis Pharmaceuticals. M.S.L. has an unrestricted Center of Excellence education grant (2005–2006) from Abbott. At the time of this study, J.C. was on the Speaker’s Bureau for Merck.
First Published Online January 27, 2009
Abbreviations: BMI, Body mass index; CI, confidence interval; MET, metabolic equivalent; OR, odds ratio.
References
- Burge R, Dawson-Hughes B, Solomon DH, Wong JB, King A, Tosteson A 2007 Incidence and economic burden of osteoporosis-related fractures in the United States, 2005–2025. J Bone Miner Res 22:465–475 [DOI] [PubMed] [Google Scholar]
- LeBoff MS, Hawkes WG, Glowacki J, Yu-Yahiro J, Hurwitz S, Magaziner J 2008 Vitamin D-deficiency and post-fracture changes in lower extremity function and falls in women with hip fractures. Osteoporos Int 19:1283–1290 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kelly EW, Jonson SR, Cohen ME, Shaffer R 2000 Stress fractures of the pelvis in female navy recruits: an analysis of possible mechanisms of injury. Mil Med 165:142–146 [PubMed] [Google Scholar]
- Grieco AJ 1977 Homocystinuria: pathogenetic mechanisms. Am J Med Sci 273:120–132 [PubMed] [Google Scholar]
- Morreels Jr CL, Fletcher BD, Weilbaecher RG, Dorst JP 1968 The roentgenographic features of homocystinuria. Radiology 90:1150–1158 [DOI] [PubMed] [Google Scholar]
- Mudd SH, Skovby F, Levy HL, Pettigrew KD, Wilcken B, Pyeritz RE, Andria G, Boers GH, Bromberg IL, Cerone R, Fowler B, Grobe H, Schmidt H, Schweitzer L 1985 The natural history of homocystinuria due to cystathionine beta-synthase deficiency. Am J Hum Genet 37:1–31 [PMC free article] [PubMed] [Google Scholar]
- van Meurs JB, Dhonukshe-Rutten RA, Pluijm SM, van der Klift M, de Jonge R, Lindemans J, de Groot LC, Hofman A, Witteman JC, van Leeuwen JP, Breteler MM, Lips P, Pols HA, Uitterlinden AG 2004 Homocysteine levels and the risk of osteoporotic fracture. N Engl J Med 350:2033–2041 [DOI] [PubMed] [Google Scholar]
- McLean RR, Jacques PF, Selhub J, Tucker KL, Samelson EJ, Broe KE, Hannan MT, Cupples LA, Kiel DP 2004 Homocysteine as a predictive factor for hip fracture in older persons. N Engl J Med 350:2042–2049 [DOI] [PubMed] [Google Scholar]
- Raisz LG 2004 Homocysteine and osteoporotic fractures–culprit or bystander? N Engl J Med 350:2089–2090 [DOI] [PubMed] [Google Scholar]
- Jacques PF, Selhub J, Bostom AG, Wilson PW, Rosenberg IH 1999 The effect of folic acid fortification on plasma folate and total homocysteine concentrations. N Engl J Med 340:1449–1454 [DOI] [PubMed] [Google Scholar]
- Selhub J, Jacques PF, Wilson PW, Rush D, Rosenberg IH 1993 Vitamin status and intake as primary determinants of homocysteinemia in an elderly population. JAMA 270:2693–2698 [DOI] [PubMed] [Google Scholar]
- Lee JS, Lacroix AZ, Wu L, Cauley JA, Jackson RD, Kooperberg C, Leboff MS, Robbins J, Lewis CE, Bauer DC, Cummings SR 2008 Associations of serum sex hormone-binding globulin and sex hormone concentrations with hip fracture risk in postmenopausal women. J Clin Endocrinol Metab 93:1796–1803 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stevens LA, Coresh J, Greene T, Levey AS 2006 Assessing kidney function–measured and estimated glomerular filtration rate. N Engl J Med 354:2473–2483 [DOI] [PubMed] [Google Scholar]
- Lewerin C, Ljungman S, Nilsson-Ehle H 2007 Glomerular filtration rate as measured by serum cystatin C is an important determinant of plasma homocysteine and serum methylmalonic acid in the elderly. J Intern Med 261:65–73 [DOI] [PubMed] [Google Scholar]
- Fried LF, Biggs ML, Shlipak MG, Seliger S, Kestenbaum B, Stehman-Breen C, Sarnak M, Siscovick D, Harris T, Cauley J, Newman AB, Robbins J 2007 Association of kidney function with incident hip fracture in older adults. J Am Soc Nephrol 18:282–286 [DOI] [PubMed] [Google Scholar]
- 1998 Design of the Women’s Health Initiative clinical trial and observational study. The Women’s Health Initiative Study Group. Control Clin Trials 19:61–109 [DOI] [PubMed] [Google Scholar]
- Chen Z, Kooperberg C, Pettinger MB, Bassford T, Cauley JA, LaCroix AZ, Lewis CE, Kipersztok S, Borne C, Jackson RD 2004 Validity of self-report for fractures among a multiethnic cohort of postmenopausal women: results from the Women’s Health Initiative observational study and clinical trials. Menopause 11:264–274 [DOI] [PubMed] [Google Scholar]
- Prentice R, Breslow N 1978 Retrospective studies and failure time models. Biometrika 65:153–158 [Google Scholar]
- Patterson RE, Kristal AR, Tinker LF, Carter RA, Bolton MP, Agurs-Collins T 1999 Measurement characteristics of the Women’s Health Initiative food frequency questionnaire. Ann Epidemiol 9:178–187 [DOI] [PubMed] [Google Scholar]
- Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM, Strath SJ, O'Brien WL, Bassett Jr DR, Schmitz KH, Emplaincourt PO, Jacobs Jr DR, Leon AS 2000 Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc 32(9 Suppl):S498–S504 [DOI] [PubMed] [Google Scholar]
- Ueland PM, Refsum H 1989 Plasma homocysteine, a risk factor for vascular disease: plasma levels in health, disease, and drug therapy. J Lab Clin Med 114:473–501 [PubMed] [Google Scholar]
- McCully KS 1996 Homocysteine and vascular disease. Nat Med 2:386–389 [DOI] [PubMed] [Google Scholar]
- Gerdhem P, Ivaska KK, Isaksson A, Pettersson K, Vaananen HK, Obrant KJ, Akesson K 2007 Associations between homocysteine, bone turnover, BMD, mortality, and fracture risk in elderly women. J Bone Miner Res 22:127–134 [DOI] [PubMed] [Google Scholar]
- Masse PG, Boskey AL, Ziv I, Hauschka P, Donovan SM, Howell DS, Cole DE 2003 Chemical and biomechanical characterization of hyperhomocysteinemic bone disease in an animal model. BMC Musculoskelet Disord 4:2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Masse PG, Rimnac CM, Yamauchi M, Coburn SP, Rucker RB, Howell DS, Boskey AL 1996 Pyridoxine deficiency affects biomechanical properties of chick tibial bone. Bone 18:567–574 [DOI] [PubMed] [Google Scholar]
- McKusick VA 1966 Heritable disorders of connective tissue. St. Louis, MO: C. V. Mosby [Google Scholar]
- Herrmann M, Kraenzlin M, Pape G, Sand-Hill M, Herrmann W 2005 Relation between homocysteine and biochemical bone turnover markers and bone mineral density in peri- and post-menopausal women. Clin Chem Lab Med 43:1118–1123 [DOI] [PubMed] [Google Scholar]
- Hodsman AB, Bauer DC, Dempster DW, Dian L, Hanley DA, Harris ST, Kendler DL, McClung MR, Miller PD, Olszynski WP, Orwoll E, Yuen CK 2005 Parathyroid hormone and teriparatide for the treatment of osteoporosis: a review of the evidence and suggested guidelines for its use. Endocr Rev 26:688–703 [DOI] [PubMed] [Google Scholar]
- Klawansky S, Komaroff E, Cavanaugh Jr PF, Mitchell DY, Gordon MJ, Connelly JE, Ross SD 2003 Relationship between age, renal function and bone mineral density in the US population. Osteoporos Int 14:570–576 [DOI] [PubMed] [Google Scholar]
- LaCroix AZ, Lee JS, Wu L, Cauley JA, Shlipak MG, Ott SM, Robbins J, Curb JD, Leboff M, Bauer DC, Jackson RD, Kooperberg CL, Cummings SR 2008 Cystatin-C, renal function, and incidence of hip fracture in postmenopausal women. J Am Geriatr Soc 56:1434–1441 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gjesdal CG, Vollset SE, Ueland PM, Refsum H, Drevon CA, Gjessing HK, Tell GS 2006 Plasma total homocysteine level and bone mineral density: the Hordaland Homocysteine Study. Arch Intern Med 166:88–94 [DOI] [PubMed] [Google Scholar]
- Dharnidharka VR, Kwon C, Stevens G 2002 Serum cystatin C is superior to serum creatinine as a marker of kidney function: a meta-analysis. Am J Kidney Dis 40:221–226 [DOI] [PubMed] [Google Scholar]
- Kyhse-Andersen J, Schmidt C, Nordin G, Andersson B, Nilsson-Ehle P, Lindstrom V, Grubb A 1994 Serum cystatin C, determined by a rapid, automated particle-enhanced turbidimetric method, is a better marker than serum creatinine for glomerular filtration rate. Clin Chem 40:1921–1926 [PubMed] [Google Scholar]
- Ubbink JB, Vermaak WJ, van der Merwe A, Becker PJ 1993 Vitamin B-12, vitamin B-6, and folate nutritional status in men with hyperhomocysteinemia. Am J Clin Nutr 57:47–53 [DOI] [PubMed] [Google Scholar]
- Herrmann M, Stanger O, Paulweber B, Hufnagl C, Herrmann W 2006 Folate supplementation does not affect biochemical markers of bone turnover. Clin Lab 52:131–136 [PubMed] [Google Scholar]
- Ravaglia G, Forti P, Maioli F, Servadei L, Martelli M, Brunetti N, Bastagli L, Cucinotta D, Mariani E 2005 Folate, but not homocysteine, predicts the risk of fracture in elderly persons. J Gerontol A Biol Sci Med Sci 60:1458–1462 [DOI] [PubMed] [Google Scholar]
- Sato Y, Honda Y, Iwamoto J, Kanoko T, Satoh K 2005 Effect of folate and mecobalamin on hip fractures in patients with stroke: a randomized controlled trial. JAMA 293:1082–1088 [DOI] [PubMed] [Google Scholar]
- McLean RR, Jacques PF, Selhub J, Fredman L, Tucker KL, Samelson EJ, Kiel DP, Cupples LA, Hannan MT 2008 Plasma B vitamins, homocysteine, and their relation with bone loss and hip fracture in elderly men and women. J Clin Endocrinol Metab 93:2206–2212 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.