Skip Navigation


QJM Advance Access originally published online on September 23, 2006
QJM 2006 99(10):673-682; doi:10.1093/qjmed/hcl094
This Article
Right arrow Summary Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
99/10/673    most recent
hcl094v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (2)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by van Staa, T.P.
Right arrow Articles by Cooper, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by van Staa, T.P.
Right arrow Articles by Cooper, C.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press on behalf of the Association of Physicians. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

A simple clinical score for estimating the long-term risk of fracture in post-menopausal women

T.P. van Staa1,2, P. Geusens3,4, J.A. Kanis5, H.G.M. Leufkens1, S. Gehlbach2 and C. Cooper2,

From the 1Utrecht Institute for Pharmaceutical Sciences, Utrecht University, Utrecht, Netherlands, 2MRC Epidemiology Resource Centre, University of Southampton, Southampton, UK,3Department of Rheumatology, University Hospital, Maastricht, Netherlands, 4Limburg University Center, Diepenbeek, Belgium, and 5Centre for Metabolic Bone Disease (WHO Collaborating Centre), University of Sheffield Medical School, Sheffield, UK

Address correspondence to Professor C. Cooper, Professor of Rheumatology and Director, MRC Epidemiology Resource Centre, Southampton General Hospital, Southampton SO16 6YD. email: cc{at}mrc.soton.ac.uk

Received 3 March 2006 and in revised form 26 May 2006


   Summary
Top
 Summary
Introduction
 Methods
 Results
 Discussion
 Acknowledgements
 References
 
Background: Simple tools are needed to identify patients at high risk of fracture.

Aim: To develop a simple clinical tool for assessing 5-year risk of fracture.

Design: Cohort study.

Methods: The study population consisted of all women aged 50+ included in the THIN Research Database (containing computerized medical records of UK general practices). Using Cox proportional hazards models, a risk score was initially estimated from age, body mass index, and clinical risk factors. The 5-year risk of fracture (survival function) was estimated for each score.

Results: The study population included 366 104 women aged >=50 years (mean follow-up 5.8 years). Of these, 6453 suffered a hip fracture. Several characteristics independently contributed to the fracture risk score (age, body mass index, fracture and fall history, previous diagnoses and use of medication). The 5-year risks for hip fracture for patients with total scores of 10, 30 and 50 were 0.3% (95%CI 0.3–0.4%), 2.2% (95%CI 2.1–2.2%), and 13.1% (95%CI 12.5–13.7%), respectively. A woman aged 65 years with low BMI and a history of both fracture and falling would have a hip fracture risk score of 37, with a corresponding 5-year risk for a hip fracture of 4.1% (4.0–4.2%). The risk score was validated and tested in another population (from GPRD), with a good concurrence between predicted and observed risks of fracture.

Discussion: This risk score predicts the long-term risk of fracture, and could be used for targeting patients for further investigation, such as bone densitometry.


   Introduction
Top
 Summary
 Introduction
Methods
 Results
 Discussion
 Acknowledgements
 References
 
Osteoporotic fracture is a serious source of morbidity and mortality in the elderly. Prospective studies have demonstrated that bone mineral density (BMD) is a major determinant of the risk of osteoporotic fracture: the risk of hip fracture approximately doubles with a decrease in bone mineral density of 1 SD below the age-adjusted mean.1 However, because the risk of fracture is also related to bone quality and to non-skeletal risk factors, screening the BMD of whole populations is not generally recommended. Instead, case-finding strategies are widely accepted as the way to identify individuals suitable for treatment; the targeting of high-risk populations is very important for cost-effectiveness.2,3 In a case-finding strategy, individuals are identified by the presence of risk factors and subsequently undergo BMD examination, with intervention recommended when the BMD is below a given threshold. In a recent study, a case-finding strategy that combined the information of clinical risk factors and selective use of BMD better identified high-risk patient groups.4

Several studies have used information on clinical risk factors to develop scores that predict long-term risk of fractures5–8 or low BMD.9–11 For example, the FRACTURE Index was developed using data from the Study of Osteoporotic Fractures (SOF) and was based on age, fracture history, maternal hip fracture, weight, smoking status and use of arms to stand up from a chair. Each of these variables was then given a score, based on the relative increase in the risk of fracture in patients with this variable. The long-term risk of fracture was then calculated for the total score.5 However, patients in the SOF study and the other prospective studies may not be representative of the general population of older women, as they relied on volunteers, who may be healthier than women of the same age in the general population.5 The primary objective of this study was to estimate the long-term risks of fracture in a large general population of post-menopausal women.


   Methods
Top
 Summary
 Introduction
 Methods
Results
 Discussion
 Acknowledgements
 References
 
Information for the study was obtained from The Health Improvement Network (THIN) research database of computerized medical records of UK general practitioners.12 General practitioners (GP) play a key role in the UK health care system, as they are responsible for primary health care and specialist referrals. Patients are semi-permanently affiliated to a practice, which centralizes the medical information from the GPs, specialist referrals and hospitalizations. The data recorded in THIN include demographic information, prescription details, clinical events, preventive care provided, specialist referrals, hospital admissions and their major outcomes. A validation study of the General Practice Research Database (GPRD) reported a high validity with respect to fractures. Hip fractures were confirmed by the GP in 91.0% of fracture cases.13

Study population
The study population consisted of all women aged 50 years or older who were registered at one of the THIN practices. A fracture risk score and corresponding 5- and 10-year risks of fracture were calculated for each woman, using Cox proportional hazards regression models. The methodology was similar to that previously applied to an analysis of oral glucocorticoid users.14 The study population was followed from 2 year after start of computerization of the general practice to the end of THIN data collection. Only data from 1990 onwards were used. As a specific risk score has now been developed for oral glucocorticoid users,14 patients with recent use of oral glucocorticoids were excluded.

Cases were patients who had a clinical osteoporotic fracture during follow-up (i.e. fracture of the radius/ulna, humerus, rib, femur/hip, pelvis or vertebrae). The history of any type of fracture was determined. Also, the occurrence of osteoporotic fracture at any other site during follow-up was noted. In order to exclude fractures that occurred at the same time but were recorded at a different time in the medical record, any fractures that occurred in the prior 3-month period were not included.

The risk factors considered in the study included body mass index (BMI) and smoking history, where available. The analysis also evaluated the presence of diseases and use of drugs that have been associated with an increased risk of fracture in a previous GPRD study.15 These included prescriptions in the previous 6 months for central nervous system medication (anticonvulsants, hypnotics/anxiolytics, antidepressants, antipsychotics and anti-Parkinsonian drugs), recorded medical history of early menopause, and of falls in the previous 6–18 months. The presence of the following diseases was also noted: chronic obstructive pulmonary disease and asthma (ICD9 492, 493, 496), cerebrovascular accident (431, 432, 433, 434, 436), heart failure (428), rheumatoid arthritis (714), and inflammatory bowel disease (555, 556). For patients with any of these diseases, presence of a record indicating a GP visit or hospitalization for these diseases in the 6 months before was also measured. Risk factors were included as categorical variables (present vs. absent). For risk factors with missing data (smoking and BMI), indicator variables for missing values were included in the regression models. The period of follow-up period was divided into 6-month intervals. As age increases over time, and clinical risk factors may vary over time (e.g. medication use), the patient's age and the presence of clinical risk factors were assessed at the start of each 6-month interval of follow-up.

Statistical methods
Cox proportional hazards models were used to estimate the long-term risks of fracture. For each set of patient characteristics, the Cox model allows calculation of an individual's probability of fracture (i.e. survivor function). We fitted regression models with age and the risk factors. Backward regression was conducted using a significance level of p = 0.05. We also investigated possible statistical interactions between age and the risk factors (i.e. whether risk scores differed with age). The beta-coefficients of this Cox model (the exponentials of which constitute the relative rates, RR) were converted into integer risk scores. The value of each integer is the rounded sum of the predictors of the Cox model multiplied by 10. Because age and clinical risk factors varied over time, the risk score of a patient was averaged over the total follow-up period. This score represents the probability of fracturing, conditional on patient survival. The absolute risk of fracture at the time of the mean duration of follow-up was then estimated for each risk score. The average hazard rate (log of risk) over this period was then used to calculate the 5- and 10-year risks of fracture. Various methods were used to test the fitting of the Cox models.16 The proportional hazards assumption was evaluated by visual examination of the Schoenfeld residuals. We also compared the observed 5-year probability of fracture (based on the Kaplan-Meier estimate) to the probability predicted by the Cox model. This was done by dividing the study population into 10 groups based on the predicted probability of fracture. The observed and predicted probabilities for fracture were then compared. Receiver operator characteristic (ROC) curves and the areas under the ROC curve were estimated.

In this population, the vertebral fractures mostly concerned clinically symptomatic vertebral fractures, confirmed radiographically.13 Systematic morphometry of vertebral fractures was not routinely done by the GP and clinically symptomatic vertebral fractures are also under-diagnosed in UK general practice.13 In order to adjust the vertebral rates for this, we compared the rates of morphometric vertebral fracture from the European Prospective Osteoporosis Study (EPOS)17 to the rates in the THIN cohort. The vertebral fracture rates in EPOS were about 18 times higher compared to those in THIN. We used half of this ratio, a more conservative approach, in line with estimates from a recent pharmacoeconomic analysis.18 The log of this ratio was then added to the risk score (i.e. multiplying the hazard rates by this ratio). The underlying assumption was that the effects of risk factors are similar between the different populations, and between morphometric and clinically symptomatic vertebral fractures (i.e. RRs of exposure are identical).

Validation of predictive model
We obtained information from a random age-stratified sample of 50 000 women aged >=50 years from the GPRD. Patients with recent use of oral glucocorticoids were also excluded in this dataset. As some GP practices contribute data to both GPRD and THIN, patients from general practices that contribute their medical records to both THIN and GPRD were excluded. The RRs of fracture for the different risk factors were compared between the two populations. Observed fracture rates and expected rates based on risk scores as developed in THIN were also evaluated.


   Results
Top
 Summary
 Introduction
 Methods
 Results
Discussion
 Acknowledgements
 References
 
The study population consisted of 366 104 women aged >=50 years. Mean duration of follow-up was 5.8 years (median 4.7 years). There were 6453 women who suffered a hip fracture (1610 clinical vertebral and 14 011 other osteoporotic fractures).

Table 1 displays the RRs of fracture for age and risk factors. Strong risk factors for fracture included age, low BMI, fall and fracture history. Patients with one of the selected chronic diseases also had increased risks of fractures, especially in those with a recent GP visit or hospitalization. Table 2 presents the risk score for various patient characteristics. For example, for a woman aged 65 years with low BMI and fracture and fall history, the total hip fracture risk score was 37 (risk score for age 13; low BMI 6; fall history 10; fracture history 8; total 37 points). The corresponding 5-year risk for a hip fracture was 4.1% (95%CI 4.0–4.2%). The 5-year risks for hip fracture for patients with total scores of 10, 30 and 50 were 0.3% (95%CI 0.3–0.4%), 2.2% (95%CI 2.1–2.2%), and 13.1% (95%CI 12.5–13.7%), respectively (Figure 1). Table 3 shows the distribution of risk scores in the study population. The median hip fracture risk score was 38 (90% percentile: 46) for women aged 80–89 years. The area under the ROC curve was 0.84 for hip fractures, 0.69 for clinical vertebral fractures and 0.60 for other clinical osteoporotic fractures.


View this table:
[in this window]
[in a new window]
  		Table 1 Prevalence and the age-adjusted relative risk (95%CI) of fracture for age and clinical risk factors

 

View this table:
[in this window]
[in a new window]
  		Table 2 Risk score* of fracture based on age and clinical risk factors

 

Figure 1
View larger version (24K):
[in this window]
[in a new window]
  		Figure 1. Relation between our risk score and risk of fracture for 5- and 10-year periods ({circ} = 5-year; {square} = 10-year).

 

View this table:
[in this window]
[in a new window]
  		Table 3 Five-year fracture incidence at different percentiles of fracture risk scores

 
The validation dataset (GPRD) included 32 728 women followed for a mean 5.6 years. Fracture rates in the validation dataset were higher than in THIN. There were also differences within THIN in fracture rates between practices that contributed both to GPRD and THIN vs. those contributing to THIN only: the rate of hip fracture was 31% higher in those contributing both to GPRD and THIN. For clinical vertebral fractures, the difference was +64% and for other clinical osteoporotic fractures +35%. When stratifying within each dataset by deciles of fracture risk, a good concurrence was found between the predicted and observed risks of fracture (Table 4).


View this table:
[in this window]
[in a new window]
  		Table 4 Observeda and predictedb 5-year fracture risks in the dataset used for the development of the risk score (THIN) and in the dataset used for validation (GPRD), with the population divided into 10 samples based on the fracture risk score in each dataset

 

   Discussion
Top
 Summary
 Introduction
 Methods
 Results
 Discussion
Acknowledgements
 References
 
We have developed a clinical risk score that provides an easily applicable clinical method of estimating a patient's individual risk of fracture, based on routinely available clinical information. Our data suggest that that high-risk patients can be identified based on clinical risk factors.

Several other studies have evaluated the value of clinical risk factors in the prediction of fracture risk5–8 or low BMD.9–11 In addition to the FRACTURE index derived from the SOF study,5 the EPIDOS study8 developed a score for hip fracture risk based on age, neuromuscular factors (slow gait speed, reduced visual acuity and difficulty on the tandem stand test) and history of falling. The Rotterdam study used information on age, sex, height, weight, use of a walking aid and smoking to identify patients at high risk of hip fracture.6 There are also several risk assessment schemes for identifying women with osteoporosis.9–11,19,20 We found that several of these risk assessments schemes performed well in identifying women at high risk of hip fracture in our study population (Figure 2). Interestingly, there were no major differences in fracture risk between the various risk assessment schemes. A simple scheme that only included age and weight (the Osteoporosis Self-Assessment Tool10) found fracture risks similar to our more complex scheme. This may be related to the low prevalence in the general population of some of the risk factors included in our risk assessment scheme. Simple risk assessment schemes may be useful for the screening of broad populations, but may underestimate the risks in patients with risk factors not included in the scheme.


Figure 2
View larger version (18K):
[in this window]
[in a new window]
  		Figure 2. Five-year hip fracture incidence in deciles of the THIN population with different risk classification schemes (•, our risk score; {square}, Fracture Index;5 X, Rotterdam Risk Score;6 {circ}, Body weight criterion;9 #, Osteoporosis Risk Assessment Instrument;11 {Delta}, Osteoporosis Self-Assessment Tool.10 Calculation of the risk scores was as done as follows. Fracture Index: age (years): 1 if 65–69, 2 if 70–74, 3 if 75–79, 4 if 80–84, 5 if 85+; fracture history, 1; weight (kg): 1 if <57; Smoker, 1. Rotterdam Risk Score: age (years): 7 if 60–64, 14 if 65–69, 21 if 70–74, 28 if 75–79, 35 if 80–84, 42 if 85+; height (cm): 4 if 160–164, 8 if 165–169, 12 if 170–174, 16 if 175–179, 20 if >=180. Weight (kg): 2 if 90–94, 4 if 85–89, 6 if 80–84, 8 if 75–79, 10 if 70–74, 12 if 65–69, 14 if 60–64, 16 if 55–59, 18 if < 55; smoker, 9. Body weight criterion: body weight in kg. Osteoporosis Risk Assessment Instrument: age (years): 5 if 55–64, 9 if 65–74, 15 if 75+; weight (kg): 9 if <60, 3 if 60–69; oestrogen use, 2. Osteoporosis Self–Assessment Tool: 0.2x(weight in kg–age in years) truncate to yield integer.

 
Low body weight and BMI are important variables in the various risk classification schemes as published in literature and in the risk score scheme developed in this study. Several studies, including our own, have shown a significant protective effect of high body weight or BMI on hip fracture risk.21,22 This may be related to direct protective effect of adipose tissue around the hip, to the production of endogenous estrogens produced in adipose tissue, or to nutrient intake, including calcium and proteins. In addition, low body weight or BMI may be a marker of other conditions that increase the risk of osteoporosis or fracture. In our study population, there was a higher prevalence of other clinical risk factors in patients with low BMI. A review of clinical predictors of osteoporosis concluded that body weight <59 kg may be a simple and reasonably sensitive but non-specific measure for selecting women for further diagnostic testing.23 These findings suggest that elderly women with low BMI should be targeted for further diagnostic testing, including bone densitometry. Furthermore, fracture history was an important predictor for fracture risk.24

Our results confirm that clinical risk factors can be used to identify individuals at high risk of fracture. Diagnostic examination can then be targeted to individuals with higher risks. However, the magnitude of the absolute risk of fracture does not determine the type of intervention that is required; risk of fracture can be related to both skeletal-related risk factors and fall-related risk factors.25 An intervention that reduces the propensity for falling may be of limited value to individuals with a high risk of fracture due to low BMD or negative changes in bone micro-architecture. The same caveat may apply to giving a bisphosphonate to an individual with a high propensity of falling (e.g. due to sedative use). The efficacy of bisphosphonates in individuals with low BMD is well established, but they may be less efficacious in individuals with normal BMD.26

The risk estimates in this study are based on historical data. It has been argued that such data can only be an estimate for prospective prediction of fracture risk, because populations and circumstances are continuously changing.27 Also, the risk estimates in this study may not generalize to other populations, as the underlying fracture incidence and the effects of risk factors may vary. It would be more appropriate to view the risk estimates in this study as a tool to improve the prediction of fractures, rather than as definitive estimates applicable to every patient. Our findings are based on a complex mathematical model. We evaluated the key underlying assumptions and tested the predictive capacity in a different study population. However, we did not evaluate all possible interactions between the risk factors, and for certain risk factor combinations the model may therefore have over- or under-estimated fracture risks. Another limitation was that we did not have information on all risk factors for fracture (e.g. BMD, exercise, calcium intake, family history, diet), that could improve the accuracy of prediction for an individual patient. Information on BMI and smoking history was missing in about one-third of the population, as this information is not routinely measured and recorded by GPs. We can not exclude the possibility that these characteristics were only measured for selected groups of patients, but our results for the relationship to fracture are in concordance with those reported in the literature.21,22

In conclusion, a simple risk score based on age, BMI, fracture and fall history and history of other disease and concomitant drug use, can help to quantify the long-term absolute risk of fracture. This score can also be used to target preventative or investigative action to patients with higher long-term risks.


   Acknowledgements
Top
 Summary
 Introduction
 Methods
 Results
 Discussion
 Acknowledgements
References
 
The study was supported by a grant from Procter & Gamble Pharmaceuticals Limited. T.P. van Staa was previously employed by Procter & Gamble Pharmaceuticals, Egham, UK.


   References
Top
 Summary
 Introduction
 Methods
 Results
 Discussion
 Acknowledgements
 References
 
1. Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. Br Med J 1996; 312:1254–9.[Abstract/Free Full Text]

2. Compston JE, Papapoulos SE, Blanchard F. Report on osteoporosis in the European Community: current status and recommendations for the future. Working Party from European Union Member States. Osteoporos Int 1998; 8:531–4.[CrossRef][ISI][Medline]

3. Jonsson B. Targeting high-risk populations. Osteoporos Int 1998; 8:Suppl. 1, S13–16.

4. Johansson H, Oden A, Johnell O, Jonsson B, de Laet C, Oglesby A, McCloskey EV, Kayan K, Jalava T, Kanis JA. Optimization of BMD measurements to identify high risk groups for treatment–a test analysis. J Bone Miner Res 2004; 19:906–13.[CrossRef][ISI][Medline]

5. Black DM, Steinbuch M, Palermo L, Dargent-Molina P, Lindsay R, Hoseyni MS, Johnell O. An assessment tool for predicting fracture risk in postmenopausal women. Osteoporos Int 2001; 12:519–28.[CrossRef][ISI][Medline]

6. Burger H, de Laet CEDH, Weel AEAM, Hofman A, Pols HAP. Added value of bone mineral density in hip fracture risk scores. Osteoporos Int 1999; 25:369–74.

7. McGrother CW, Donaldson MM, Clayton D, Abrams KR, Clarke M. Evaluation of a hip fracture risk score for assessing elderly women: the Melton Osteoporotic Fracture (MOF) study. Osteoporos Int 2002; 13:89–96.[CrossRef][ISI][Medline]

8. Dargent-Molina P, Douchin MN, Cormier C, Meunier PJ, Breart G. EPIDOS Study Group. Use of clinical risk factors in elderly women with low bone mineral density to identify women at higher risk of hip fracture: The EPIDOS prospective study. Osteoporos Int 2002; 13:593–9.[CrossRef][ISI][Medline]

9. Michaelsson K, Bergstrom R, Mallmin H, Holmberg L, Wolk A, Ljunghall S. Screening for osteopenia and osteoporosis: selection by body composition. Osteoporos Int 1996; 6:120–6.[CrossRef][ISI][Medline]

10. Koh LK, Sedrine WB, Torralba TP, Kung A, Fujiwara S, Chan SP, Huang QR, Rajatanavin R, Tsai KS, Park HM, Reginster JY. Osteoporosis Self-Assessment Tool for Asians (OSTA) Research Group. A simple tool to identify asian women at increased risk of osteoporosis. Osteoporos Int 2001; 12:699–705.[CrossRef][ISI][Medline]

11. Cadarette SM, Jaglal SB, Kreiger N, McIsaac WJ, Darlington GA, Tu JV. Development and validation of the Osteoporosis Risk Assessment Instrument to facilitate selection of women for bone densitometry. CMAJ 2000; 162:1289–94.[Abstract/Free Full Text]

12. Walley T and Mantgani A. The UK General Practice Research Database. Lancet 1997; 350:1097–9.[CrossRef][ISI][Medline]

13. van Staa TP, Abenhaim L, Cooper C, Zhang B, Leufkens HGM. The use of a large pharmacoepidemiological database to study exposure to oral corticosteroids and risk of fractures: validation of study population and results. Pharmacoepi Drug Saf 2000; 9:359–66.[CrossRef]

14. van Staa TP, Geusens P, Pols HAP, de Laet C, Leufkens HGM, Cooper C. A simple score for estimating the long-term risk of fracture in patients using oral glucocorticoids. Q J Med 2005; 98:191–8.

15. van Staa TP, Leufkens HGM, Cooper C. Utility of medical and drug history in fracture risk prediction among men and women. Bone 2002; 31:508–14.[Medline]

16. Harrell HE, Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996; 15:361–87.[CrossRef][ISI][Medline]

17. Felsenberg D, Silman AJ, Lunt M, et al. for the European Prospective Osteoporosis Study (EPOS) group. Incidence of vertebral fracture in Europe: results from the European prospective Osteoporosis Study (EPOS). J Bone Miner Res 2002; 17:716–24.[CrossRef][ISI][Medline]

18. Kanis JA, Brazier J, Calvert N, Lloyd-Jones M, Stevenson M. Treatment of established osteoporosis. Health Technol Assess 2002; 6:1–146 Available at: [http://www.ncchta.org.execsumm/summ629.htm].[Medline]

19. Geusens P, Hochberg MC, van der Voort DJ, Pols H, van der Klift M, Siris E, Melton ME, Turpin J, Byrnes C, Ross P. Performance of risk indices for identifying low bone density in postmenopausal women. Mayo Clin Proc 2002; 77:629–37.[ISI][Medline]

20. Cadarette SM, McIsaac WJ, Hawker GA, Jaakkimainen L, Culbert A, Zarifa G, Ola E, Jaglal SB. The validity of decision rules for selecting women with primary osteoporosis for bone mineral density testing. Osteoporos Int 2004; 15:361–6.[CrossRef][ISI][Medline]

21. Cummings SR, Nevitt MC, Browner WS, Stone K, Fox KF, Ensrud KE, Cauley J, Black D, Vogt TM. Risk factors for hip fractures in white women. N Engl J Med 1995; 332:767–73.[Abstract/Free Full Text]

22. Johnell O, Gullberg B, Kanis JA, Allander E, Elffors L, Dequeker J, Dilsen G, Gennari C, Lopes Vaz A, Lyritis G, et al. Risk factors for hip fracture in European women: the MEDOS Study. Mediterranean Osteoporosis Study. J Bone Miner Res 1995; 10:1802–15.[ISI][Medline]

23. Green AD, Colon-Emeric CS, Bastian L, Drake MT, Lyles KW. Does this woman have osteoporosis? JAMA 2004; 292:2890–900.[Abstract/Free Full Text]

24. Kanis JA, Johnell O, De Laet C, Johansson H, Oden A, Delmas P, Eisman J, Fujiwara S, Garnero P, Kroger H, McCloskey EV, Mellstrom D, Melton LJ, Pols H, Reeve J, Silman A, Tenenhouse A. A meta-analysis of previous fracture and subsequent fracture risk. Bone 2004; 35:375–82.[Medline]

25. Allolio B. Risk factors for hip fracture not related to bone mass and their therapeutic implications. Osteoporos Int 1999; 9:Suppl. 2, S9–16.

26. Cummings SR, Black DM, Thompson DE, Applegate WB, Barrett-Connor E, Musliner TA, Palermo L, Prineas R, Rubin SM, Scott JC, Vogt T, Wallace R, Yates AJ, LaCroix AZ. Effect of alendronate on risk of fracture in women with low bone density but without vertebral fractures: results from the Fracture Intervention Trial. JAMA 1998; 280:2077–82.[Abstract/Free Full Text]

27. Tunstall-Pedoe H. ‘Absolute’ is inappropriate for quantitative risk estimation. Br Med J 2000; 320:723.[Free Full Text]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Summary Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
99/10/673    most recent
hcl094v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (2)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by van Staa, T.P.
Right arrow Articles by Cooper, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by van Staa, T.P.
Right arrow Articles by Cooper, C.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?