Validation of ten-year fracture risk prediction: a clinical cohort study from the Manitoba Bone Density Program

INTRODUCTION: Absolute 10-year fracture risk is the preferred method for fracture risk assessment. The validity of applying published fracture rates from one population to another population is uncertain. METHODS: 20,579 women age 47.5 years or older at the time of baseline femoral neck bone mineral density (BMD) were identified in a database containing all clinical DXA results for the Province of Manitoba, Canada. Individual 10-year fracture risk was predicted from age-only and age plus femoral neck T-score using published 10-year fracture risk for Swedish women. Health service records were assessed for the presence of non-trauma 'osteoporotic' fracture codes (hip, clinical spine, wrist, humerus) subsequent to BMD testing (86,447 person-y follow up, 1173 patients with osteoporotic fractures). Fracture rates were derived for subgroups stratified by age (5-year strata) and estimated risk (5% strata). 10-year fracture rates were computed directly by the Kaplan-Meier method (10-year continuous data) and by the actuarial method (two 5-year periods with adjustments for aging, death and expected BMD loss). RESULTS: Direct and actuarial methods gave nearly identical point estimates, but the latter were more precise. There was a strong linear correlation between predicted and observed fracture rates based upon age-only (r = 0.95) and age plus BMD (r = 0.99). For age strata 50 to 75, and for estimated risk strata from 0-5% to 20-25%, the confidence intervals overlapped the line of identity. For women age >77.5 or estimated risk >25%, observed exceeded estimated fracture rates. This is explained by healthy selection bias whereby elderly women referred for BMD testing have lower mortality than expected, hence more years at risk for fracture. Corrected for survival bias, women age >77.5 had observed fracture rates no different than predicted. CONCLUSION: Swedish 10-year fracture risk data are generally applicable to the Canadian female population referred for clinical BMD testing, though fracture rates were underestimated in the oldest and highest risk subgroups due to healthy selection bias.     

Bone three-dimensional microstructural features of the common osteoporotic fracture sites

Osteoporosis is a common metabolic skeletal disorder characterized by decreased bone mass and deteriorated bone structure, leading to increased susceptibility to fractures. With aging population, osteoporotic fractures are of global health and socioeconomic importance. The three-dimensional microstructural information of the common osteoporosis-related fracture sites, including vertebra, femoral neck and distal radius, is a key for fully understanding osteoporosis pathogenesis and predicting the fracture risk. Low vertebral bone mineral density (BMD) is correlated with increased fracture of the spine. Vertebral BMD decreases from cervical to lumbar spine, with the lowest BMD at the third lumbar vertebra. Trabecular bone mass of the vertebrae is much lower than that of the peripheral bone. Cancellous bone of the vertebral body has a complex heterogeneous three-dimensional microstructure, with lower bone volume in the central and anterior superior regions. Trabecular bone quality is a key element to maintain the vertebral strength. The increased fragility of osteoporotic femoral neck is attributed to low cancellous bone volume and high compact porosity. Compared with age-matched controls, increased cortical porosity is observed at the femoral neck in osteoporotic fracture patients. Distal radius demonstrates spatial inhomogeneous characteristic in cortical microstructure. The medial region of the distal radius displays the highest cortical porosity compared with the lateral, anterior and posterior regions. Bone strength of the distal radius is mainly determined by cortical porosity, which deteriorates with advancing age.     

Performance of risk assessment instruments for predicting osteoporotic fracture risk: a systematic review

Summary

We systematically reviewed the literature on the performance of osteoporosis absolute fracture risk assessment instruments. Relatively few studies have evaluated the calibration of instruments in populations separate from their development cohorts, and findings are mixed. Many studies had methodological limitations making susceptibility to bias a concern.

Introduction

The aim of this study was to systematically review the literature on the performance of osteoporosis clinical fracture risk assessment instruments for predicting absolute fracture risk, or calibration, in populations other than their derivation cohorts.

Methods

We performed a systematic review, and MEDLINE, Embase, Cochrane Library, and multiple other literature sources were searched. Inclusion and exclusion criteria were applied and data extracted, including information about study participants, study design, potential sources of bias, and predicted and observed fracture probabilities.

Results

A total of 19,949 unique records were identified for review. Fourteen studies met inclusion criteria. There was substantial heterogeneity among included studies. Six studies assessed the WHO’s Fracture Risk Assessment (FRAX) instrument in five separate cohorts, and a variety of risk assessment instruments were evaluated in the remainder of the studies. Approximately half found good instrument calibration, with observed fracture probabilities being close to predicted probabilities for different risk categories. Studies that assessed the calibration of FRAX found mixed performance in different populations. A similar proportion of studies that evaluated simple risk assessment instruments (≤5 variables) found good calibration when compared with studies that assessed complex instruments (>5 variables). Many studies had methodological features making them susceptible to bias.

Conclusions

Few studies have evaluated the performance or calibration of osteoporosis fracture risk assessment instruments in populations separate from their development cohorts. Findings are mixed, and many studies had methodological limitations making susceptibility to bias a possibility, raising concerns about use of these tools outside of the original derivation cohorts. Further studies are needed to assess the calibration of instruments in different populations prior to widespread use.     

Hospitalized osteoporotic vertebral fracture increases the risk of stroke: A population‐based cohort study

The association between osteoporosis and cardiovascular diseases has been demonstrated. Higher cardiovascular risk has also been correlated with vertebral fractures. However, the association between osteoporotic vertebral fracture and the possibly higher risk of stroke remains uncertain. This study aimed to evaluate the incidence, risk, and type of stroke in patients with osteoporotic vertebral fracture. Patients with osteoporotic vertebral fracture were identified (n = 380) and 10 age‐ and sex‐matched controls per case (comparison group, n = 3795) were chosen from a nationwide representative cohort of 999,997 people from 1998 to 2005. Both groups were followed‐up for stroke events for 3 years, matched by propensity scores with adjustments for covariates such as comorbidities (ie, hypertension, diabetes, arrhythmia, or coronary heart diseases) and exposure to medications (ie, aspirin, lipid lowering drug, or nitrates), and assessed by Kaplan‐Meier and Cox regression analyses. The incidence rate of stroke in the osteoporotic vertebral fracture group (37.5 per 1000 person‐years; 95% confidence interval [CI], 27.5–51.2) was significantly higher than in the comparison group (14.0 per 1000 person‐years; 95% CI, 12.0–16.4, p < 0.001). Stroke was more likely to occur in the osteoporotic vertebral fracture patients than in the normal controls (crude hazard ratio [HR] 2.68, 95% CI 1.89–3.79, p < 0.001; adjusted HR 2.71, 95% CI 1.90–3.86, p < 0.001). In conclusion, patients with osteoporotic vertebral fracture have a higher risk of stroke (ie, both ischemic and hemorrhagic) and require stroke prevention strategies. © 2013 American Society for Bone and Mineral Research.     

Treatment Implications for Men When Switching From Male to Female Bone Mineral Density Reference Data: The Manitoba Bone Density Program

Since 2001, the International Society for Clinical Densitometry Official Position has been to use a young female normative database in women and a young male normative in men for T-scores. Several prospective studies have shown that men and women with identical hip bone mineral density (BMD) have the same fracture rates, and so there has been reconsideration of whether female reference data should be adopted for men. We studied 4691 men age 50 yr and older with baseline dual-energy X-ray absorptiometry assessments to explore how a change in BMD reference data from male to female would affect the number of men meeting National Osteoporosis Foundation (NOF) intervention criteria. We found that use of male vs female BMD reference data for T-score calculation did affect individual eligibility criteria for treatment under the NOF guidelines, but that overall differences in treatment rates were small when eligibility for treatment considered any of the NOF intervention criteria. Specifically, the majority of men who no longer qualified for treatment based upon T-scores calculated from female as opposed to male reference data would still qualify for treatment based upon other NOF intervention criteria. In addition, men only eligible for treatment under NOF criteria when using male reference data were at low fracture risk. We conclude that choice of male or female reference data for T-score calculation in men has little effect on overall treatment eligibility rates under NOF guidelines.     

Bone microarchitecture assessed by TBS predicts osteoporotic fractures independent of bone density: the Manitoba study

The measurement of BMD by dual‐energy X‐ray absorptiometry (DXA) is the “gold standard” for diagnosing osteoporosis but does not directly reflect deterioration in bone microarchitecture. The trabecular bone score (TBS), a novel gray‐level texture measurement that can be extracted from DXA images, correlates with 3D parameters of bone microarchitecture. Our aim was to evaluate the ability of lumbar spine TBS to predict future clinical osteoporotic fractures. A total of 29,407 women 50 years of age or older at the time of baseline hip and spine DXA were identified from a database containing all clinical results for the Province of Manitoba, Canada. Health service records were assessed for the incidence of nontraumatic osteoporotic fracture codes subsequent to BMD testing (mean follow‐up 4.7 years). Lumbar spine TBS was derived for each spine DXA examination blinded to clinical parameters and outcomes. Osteoporotic fractures were identified in 1668 (5.7%) women, including 439 (1.5%) spine and 293 (1.0%) hip fractures. Significantly lower spine TBS and BMD were identified in women with major osteoporotic, spine, and hip fractures (all p < 0.0001). Spine TBS and BMD predicted fractures equally well, and the combination was superior to either measurement alone (p < 0.001). Spine TBS predicts osteoporotic fractures and provides information that is independent of spine and hip BMD. Combining the TBS trabecular texture index with BMD incrementally improves fracture prediction in postmenopausal women. © 2011 American Society for Bone and Mineral Research     

Nulliparity and fracture risk in older women: the study of osteoporotic fractures.

Whether nulliparity increases fracture risk is unclear from prior studies, which are limited by small samples or lack of measured bone mineral density. No study has evaluated whether the effect of parity differs by skeletal site. We prospectively analyzed the relationship of parity to the risk of incident nontraumatic hip, spine, and wrist fractures in 9704 women aged 65 years or older participating in the Study of Osteoporotic Fractures to determine if parity reduces postmenopausal fracture risk, and if so, if this risk reduction is (1) greater at weight-bearing skeletal sites and (2) independent of bone mineral density. Parity was ascertained by self-report. Incident hip and wrist fractures were determined by physician adjudication of radiology reports (mean follow-up, 9.8 years) and spine fractures by morphometric criteria on serial radiographs. The relationship of parity to hip and wrist fracture was assessed by proportional hazards models. Spine fracture risk was evaluated by logistic regression. Compared with parous women, nulliparous women (n = 1835, 19%) had an increased risk of hip and spine, but not wrist, fractures. In multivariate models, parity remained a significant predictor only for hip fracture. Nulliparous women had a 44% increased risk of hip fractures independent of hip bone mineral density (hazards ratio, 1.44; 95% CI, 1.17-1.78). Among parous women, each additional birth reduced hip fracture risk by 9% (p = 0.03). Additionally, there were no differences in mean total hip, spine, or radial bone mineral density values between nulliparous and parous women after multivariate adjustment. In conclusion, childbearing reduces hip fracture risk by means that may be independent of hip bone mineral density.     

Prediction of fracture risk in men: a cohort study

FRAX is a tool that identifies individuals with high fracture risk who will benefit from pharmacological treatment of osteoporosis. However, a majority of fractures among elderly occur in people without osteoporosis and most occur after a fall. Our aim was to accurately identify men with a high future risk of fracture, independent of cause. In the population‐based Uppsala Longitudinal Study of Adult Men (ULSAM) and using survival analysis we studied different models' prognostic values (R²) for any fracture and hip fracture within 10 years from age 50 (n = 2322), 60 (n = 1852), 71 (n = 1221), and 82 (n = 526) years. During the total follow‐up period from age 50 years, 897 fractures occurred in 585 individuals. Of these, 281 were hip fractures occurring in 189 individuals. The rates of any fracture were 5.7/1000 person‐years at risk from age 50 years and 25.9/1000 person‐years at risk from age 82 years. Corresponding hip fractures rates were 2.9 and 11.7/1000 person‐years at risk. The FRAX model included all variables in FRAX except bone mineral density. The full model combining FRAX variables, comorbidity, medications, and behavioral factors explained 25% to 45% of all fractures and 80% to 92% of hip fractures, depending on age. The corresponding prognostic values of the FRAX model were 7% to 17% for all fractures and 41% to 60% for hip fractures. Net reclassification improvement (NRI) comparing the full model with the FRAX model ranged between 40% and 53% for any fracture and between 40% and 87% for hip fracture. Within the highest quintile of predicted fracture risk with the full model, one‐third of the men will have a fracture within 10 years after age 71 years and two‐thirds after age 82 years. We conclude that the addition of comorbidity, medication, and behavioral factors to the clinical components of FRAX can substantially improve the ability to identify men at high risk of fracture, especially hip fracture. © 2012 American Society for Bone and Mineral Research.     

Clinical risk factors for osteoporotic fracture: A population‐based prospective cohort study in Korea

Clinical risk factors (CRFs), either alone or in combination with bone mineral density, are used to determine the fracture risk for clinical assessment and to determine intervention thresholds. Because fracture risk is strongly affected by ethnicity and population‐specific differences, we sought to identify Korean‐specific CRFs for fracture, in combination with quantitative ultrasound (qUS) measurements of the radius and tibia. A total of 9351 subjects (4732 men and 4619 women) aged 40 to 69 years were followed for a mean of 46.3 ± 2.2 months. We obtained CRF information using a standardized questionnaire and measured anthropometric variables. Speed of sound at the radius (SoSR) and tibia (SoST) were measured by qUS. Fracture events were recorded using a questionnaire, and a height‐loss threshold was used as an indicator of vertebral fracture. Relative risks were calculated by Cox regression analysis. A total of 195 subjects (61 men and 134 women) suffered low‐trauma fractures. Older age, lower body mass index (BMI), and previous fracture history were positively associated with fracture risk in both sexes. Decreased hip circumference, lack of regular exercise, higher alcohol intake, menopause, and osteoarthritis history were further independent CRFs for fracture in women. However, neither SoSR nor SoST was independently associated with fracture risk. In this study, we identified the major Korean‐specific CRFs for fracture and found that smaller hip circumference was a novel risk factor. This information will allow optimal risk‐assessment targeting Koreans for whom treatment would provide the greatest benefit. © 2010 American Society for Bone and Mineral Research     

Construction and validation of a simplified fracture risk assessment tool for Canadian women and men: results from the CaMos and Manitoba cohorts

Summary

A procedure for creating a simplified version of fracture risk assessment tool (FRAX?) is described. Calibration, fracture prediction, and concordance were compared with the full FRAX tool using two large, complementary Canadian datasets.

Introduction

The Canadian Association of Radiologists and Osteoporosis Canada (CAROC) system for fracture risk assessment is based upon sex, age, bone mineral density (BMD), prior fragility fracture, and glucocorticoid use. CAROC does not require computer or web access, and categorizes 10-year major osteoporotic fracture risk as low (<10%), moderate (10?C20%), or high (>20%).

Methods

Basal CAROC fracture risk tables (by age, sex, and femoral neck BMD) were constructed from Canadian FRAX probabilities for major osteoporotic fractures (adjusted for prevalent clinical risk factors). We assessed categorization and fracture prediction with the updated CAROC system in the CaMos and Manitoba BMD cohorts.

Results

The new CAROC system demonstrated high concordance with the Canadian FRAX tool for risk category in both the CaMos and Manitoba cohorts (89% and 88%). Ten-year fracture outcomes in CaMos and Manitoba BMD cohorts showed good discrimination and calibration for both CAROC (6.1?C6.5% in low-risk, 13.5?C14.6% in moderate-risk, and 22.3?C29.1% in high-risk individuals) and FRAX (6.1?C6.6% in low-risk, 14.4?C16.1% in moderate-risk, and 23.4?C31.0% in high-risk individuals). Reclassification from the CAROC risk category to a different risk category under FRAX occurred in <5% for low-risk, 20?C24% for moderate-risk, and 27?C30% for high-risk individuals. Reclassified individuals had 10-year fracture outcomes that were still within or close to the original nominal-risk range..

Conclusion

The new CAROC system is well calibrated to the Canadian population and shows a high degree of concordance with the Canadian FRAX tool. The CAROC system provides s a simple alternative when it is not feasible to use the full Canadian FRAX tool.     

Bone turnover biomarkers and risk of osteoporotic hip fracture in an Asian population

While epidemiologic studies suggest that bone turnover biomarkers may predict hip fracture risk, findings are inconsistent and Asian data are lacking. We conducted a matched case–control (1:1) study nested in the Singapore Chinese Health Study, a population-based prospective cohort of Chinese men and women (45–74 years) recruited from 1993 to 1998 in Singapore. One hundred cases with incident hip fracture and 100 individually matched controls were randomly selected from 63,257 participants. Serum bone turnover biomarkers, namely bone alkaline phosphatase (bone ALP), osteocalcin (OC), procollagen type I N propeptide (PINP), N-terminal and C-terminal crosslinking telopeptide of type I collagen (NTX-I and CTX-I) were measured using immunoassays. Hip fracture cases had significantly higher serum levels of OC, PINP, CTX-I and NTX-I than controls (p < 0.05). There was a dose-dependent positive relationship between OC, PINP, CTX-I and NTX-I and risk of hip fracture (all Ps for trend ≤ 0.006), where the risk was significantly increased by 4.32–8.23 folds for the respective BTM [Quartile (Q) 4 vs. Q1]. The odds ratio [OR (95% CI)] at the highest quartile (Q4) was 6.63 (2.02–21.18) for PINP and 4.92 (1.67–14.51) for CTX-I. The joint effect of PINP and CTX-I showed a 7-fold increase in risk (OR: 7.36; 95% CI: 2.53–21.41) comparing participants with higher levels of PINP (Q4) and CTX-I (Q3–Q4) to those with low levels of PINP (Q1–Q3) and CTX-I (Q1–Q2). Our data demonstrated that higher serum levels of bone turnover biomarkers were associated with increased risk of hip fracture in an Asian population.     

FRAX在国内骨质疏松性骨折风险评估中的应用进展

骨质疏松性骨折是骨质疏松症最常见和最严重的并发症。骨质疏松患者一旦发生骨折,严重影响其生活质量,可致残或致死,而由此产生的治疗和护理费用给家庭及社会带来巨大的负担,因此进行骨折风险评估、预防骨质疏松性骨折显得尤为重要。WHO推荐FRAX作为骨折风险的预测工具,国外对此已有诸多报道,而国内使用FRAX评估骨折风险的研究尚处于起步阶段,相关报道仍较少。本文就FRAX在国内各类疾病患者、不同地区的应用现状作一综述,旨在探讨FRAX对国人的适用性,预测未来骨折风险,降低骨折发生率、提高生活质量。     

Disease-specific perception of fracture risk and incident fracture rates: GLOW cohort study

Summary

Accurate patient risk perception of adverse health events promotes greater autonomy over, and motivation towards, health-related lifestyles.

Introduction

We compared self-perceived fracture risk and 3-year incident fracture rates in postmenopausal women with a range of morbidities in the Global Longitudinal study of Osteoporosis in Women (GLOW).

Methods

GLOW is an international cohort study involving 723 physician practices across ten countries (Europe, North America, Australasia); 60,393 women aged ≥55 years completed baseline questionnaires detailing medical history and self-perceived fracture risk. Annual follow-up determined self-reported incident fractures.

Results

In total 2,945/43,832 (6.8 %) sustained an incident fracture over 3 years. All morbidities were associated with increased fracture rates, particularly Parkinson's disease (hazard ratio [HR]; 95 % confidence interval [CI], 3.89; 2.78–5.44), multiple sclerosis (2.70; 1.90–3.83), cerebrovascular events (2.02; 1.67–2.46), and rheumatoid arthritis (2.15; 1.53–3.04) (all p?<?0.001). Most individuals perceived their fracture risk as similar to (46 %) or lower than (36 %) women of the same age. While increased self-perceived fracture risk was strongly associated with incident fracture rates, only 29 % experiencing a fracture perceived their risk as increased. Under-appreciation of fracture risk occurred for all morbidities, including neurological disease, where women with low self-perceived fracture risk had a fracture HR 2.39 (CI 1.74–3.29) compared with women without morbidities.

Conclusions

Postmenopausal women with morbidities tend to under-appreciate their risk, including in the context of neurological diseases, where fracture rates were highest in this cohort. This has important implications for health education, particularly among women with Parkinson's disease, multiple sclerosis, or cerebrovascular disease.     

Bone mineral density and bone loss measured at the radius to predict the risk of nonspinal osteoporotic fracture.

Low bone mineral density (BMD) and, probably, the rate of bone loss (RBL) are associated with the risk of osteoporotic fractures. To estimate the risk of nonspinal fracture in osteoporotic women, we measured BMD and RBL in a prospective study (average follow-up, 5.38 years) in 656 postmenopausal women. The women were considered in three groups: group A (whole population), group B (women under the age of 65 years) and group C (women over the age of 65 years). At the beginning of the study, BMD was measured at the distal radius (DR) and at the proximal radius (PR) using a single-energy densitometer. BMD measurements made 2 years previously in the same patients were used to calculate RBL. Then patients were checked annually for nonspine fracture due to minor trauma. During follow-up, 121 nonspinal fractures were detected. Women with fractures were older and had lower BMD. With the Cox regression, age-corrected BMD at both DR and PR predicts fracture risk in groups A and B but not in group C. After correction for potential confounders, DR still predicts fractures in groups A and B whereas PR predicts fractures only in group B. In group C, only the RBL at the PR was predictive of the fracture risk as well as in the other two groups. Specific types of fractures are predictable in the whole population at the wrist. In conclusion, radial BMD predicts the risk of nonspine fractures except in women over the age of 65 years. The RBL at the PR is an effective predictor of fracture risk also in women over the age of 65 years.     

Bone remodeling markers: Assessment of fracture risk and fracture risk reduction

There is an increasing repertoire of laboratory tests available for assessing the bone remodeling process. Biochemical markers of bone remodeling can be measured in serum or urine, and have a number of potential roles in the management of fracture risk. Differences in remodeling between individuals might be related to fracture risk and could be used to target therapy. Change in remodeling with therapy could be related to fracture risk reduction and the choice of therapy could be influenced by knowledge of bone remodeling. Biochemical therapeutic monitoring may improve patient understanding and therapeutic adherence. The rate of bone remodeling is weakly predictive of fracture risk in individuals who are not receiving therapy; however, it is not clear whether this is independent of other risk factors. There is increasing evidence that change in bone mineral density does not explain fracture risk reduction with antiresorptive therapies, and that therapeutic benefit might be explained by change in bone turnover. Additional studies and information are required to allow these scientific advances to be translated into cost-effective and validated clinical protocols. Optimizing the precision and accuracy of bone turnover assessment remains an important priority.     

Bone reinnervation after fracture: a study in the rat.

Reinnervation after tibial fracture in the rat was studied by analyzing the occurrence of growth-associated protein 43 (GAP-43), a marker for regenerating nerve fibers, and protein gene product 9.5 (PGP-9.5), a marker for mature nerve fibers, by immunohistochemistry. At 3 days postfracture, GAP-43--immunoreactive nerve fibers were first observed in the fracture hematoma and periosteum. At 7 days postfracture, abundant sprouting of GAP-43--positive fibers was seen in the callus, hyperplastic periosteum, and edge of fibrocartilage. In the latter region, the nerve fibers were nonvascular, showing dense ramifications and terminal sprouting close to chondroid cells. At 14 days and 21 days postfracture, many GAP-43--positive fibers were still sprouting into the fibrocartilage and new woven bone. Fine varicose GAP-43--positive fibers also were present in the bone marrow. In contrast to GAP-43, PGP-9.5-positive nerve fibers were observed only occasionally at 3 days postfracture but gradually increased in number from day 14 to 21. Our study shows that intense nerve regeneration occurs in early fracture healing partly unrelated to neovascularization. Considering that neuronal mediators have been shown to participate in local bone formation and resorption, the nerve regeneration observed may prove to be essential for delivery of neuronal mediators required for normal callus formation and/or neovascularization.     

Coffee, tea and caffeine consumption in relation to osteoporotic fracture risk in a cohort of Swedish women

Introduction Consumption of coffee and tea, and total intake of caffeine has been claimed to be associated with osteoporotic fracture risk. However, results of earlier studies lack consistency. Methods We examined this relation in a cohort of 31,527 Swedish women aged 40-76 years at baseline in 1988. The consumption of coffee, caffeinated tea and the intake of caffeine were estimated from a self-administered food frequency questionnaire (FFQ). Multivariate-adjusted hazards ratios (HRs) of fractures with 95% confidence intervals (95% CIs) were estimated by Cox proportional hazards models. Results During a mean follow-up of 10.3 years, we observed 3,279 cases with osteoporotic fractures. The highest (>330 mg/day) compared with the lowest (<200 mg/day) quintile of caffeine intake was associated with a modestly increased risk of fracture: HR 1.20 (95% CI: 1.07–1.35). A high coffee consumption significantly increased the risk of fracture (p for trend 0.002), whereas tea drinking was not associated with risk. The increased risk of fracture with both a high caffeine intake and coffee consumption was confined to women with a low calcium intake (<700 mg/day): HR 1.33 (95% CI: 1.07–1.65) with ≥4 cups (600 ml)/day of coffee compared to <1 cup (150 ml)/day. The same comparison but risk estimated for women with a high propensity for fractures (≥2 fracture types) revealed a HR of 1.88 (95% CI: 1.17–3.00). Conclusions In conclusion, our results indicate that a daily intake of 330 mg of caffeine, equivalent to 4 cups (600 ml) of coffee, or more may be associated with a modestly increased risk of osteoporotic fractures, especially in women with a low intake of calcium.