Elsevier

Maturitas

Volume 65, Issue 2, February 2010, Pages 143-148
Maturitas

Review
Individualizing fracture risk prediction

https://doi.org/10.1016/j.maturitas.2009.12.007Get rights and content

Abstract

Low bone mineral density (BMD) and clinical factors (CRF) have been identified as factors associated with an increased relative risk of fractures. From this observation and for clinical decision making, the concept of prediction of the individual absolute risk of fractures has emerged. It refers to the individual's risk for fractures over a certain time period, e.g. the next 5 and 10 years. Two individualized fracture risk calculation tools that are increasingly used and are available on the web are the FRAX algorithm and the Garvan fracture risk calculator. These tools integrate BMD and CRFs for fracture risk calculation in the individual patient in daily practice. Although both tools include straightforward risk factors, such as age, sex, previous fractures, body weight and BMD, they differ in several aspects, such as the inclusion of other CRFs, fall risks and number of previous fractures. Both models still need to be validated in different populations before they can be generalized to other populations, since the background risk for fractures is population specific. Further studies will be needed to validate their contribution in selecting patients who will achieve fracture risk reduction with anti-osteoporosis therapy.

Introduction

The World Health Organisation (WHO) has defined osteoporosis as ‘a disease characterised by low bone mass and microarchitectural deterioration of bone tissue, leading to enhanced bone fragility and a consequent increase in fracture risk [1].’ Bone mass and bone mineral density (BMD) are most commonly measured by dual energy X-ray absorptiometry (DXA), a validated technique for measuring BMD in the lumbar spine and proximal femur. The outcome of the DXA-measurement is expressed as a T-score (T-score = (measured BMD  young adult BMD)/young adult SD). The cut-off point for osteoporosis is a T-score of 2.5 standard deviations (SDs) below the mean BMD, at the spine or at the hip, in a healthy female population aged 30 years [1], [2], [3], [4]. Approximately 15–20% of all postmenopausal women are defined as osteoporotic, using the diagnostic threshold of a T-score  −2.5 measured with DXA at the femoral neck (FN) or lumbar spine [5].

However, most patients with a fracture have no osteoporosis [6]. Many clinical risk factors (CRF) have been identified predict fracture risk, independently of each other and of BMD [7], [8]. By integrating CRFs and BMD in case-finding strategies, the risk of fracture at any BMD will also depend on the presence of CRFs. Intervention thresholds based on fracture risk calculations can therefore be different from the WHO diagnostic thresholds [9].

We review recent developments in case-finding strategies, which can be used in daily practice to calculate fracture risk in individual patients.

Section snippets

The clinical significance of fractures

The clinical significance of osteoporosis is the occurrence of fractures. Hip fractures increase morbidity and mortality and entail high socio-economic costs [3], [4], [10]. Estimates show that, in Europe alone, 890,000 persons (80% women) sustained a hip fracture (24% of total number of fractures) in the year 2000. The number of hip fractures alone could increase to 6.3 million by the year 2050 or even to 8.2 million if the assumption is made that the age-related incidence of hip fractures

Individualizing fracture risk prediction

Earlier guidelines on osteoporosis by the National Osteoporosis Foundation (NOF, in the US) and the National Osteoporosis Society (NOS, in the UK) were mainly aimed at selecting patients with pre-existing fracture or low bone mineral density for treatment [16], [17]. However, the occurrence of a fracture is a multifactorial event. Consequently, there is more than one way that an individual can attain the risk conferred by either osteoporosis or a pre-existing fracture and many CRFs have been

The fracture risk assessment tool (FRAX) [20]

The World Health Organisation (WHO) [2], [20] has developed an algorithm for individualized fracture risk prediction which is developed based on population-based cohorts from Europe, North America, Asia and Australia. The algorithm in FRAX includes the following risk factors: [4], [7], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37] age, a history of fractures, a parental history of hip fracture, low body weight or body mass index (BMI), use

Comparison of the FRAX and Garvan fracture risk calculator

Comparisons between the FRAX and Garvan tool are shown in Table 2, Table 3 and Fig. 1, Fig. 2, Fig. 3, Fig. 4. In patients without clinical risks, the fracture risk calculation for osteoporotic fractures is higher with the Garvan tool than with FRAX, since the Garvan tool predicts more fractures that FRAX (Table 2, Table 3).

The effect of fracture history on calculated fracture risk is shown in Table 2, Table 3 and Fig. 1. Calculated fracture risk increases with both tools in patients with a

Website versions

Both fracture calculations tools are freely accessible on the internet and easy to use online at the respective websites [20], [51]. The introduction of risk factors is self-explanatory and comments are available explaining in detail the definitions of CRFs.

Paper-based versions

The FRAX has simplified paper-based models in tabular format available on the website. These tables are not CRF specific, but state the risk for up to 6 CRFs for age per 5 years (range 50–90 years) and the BMD T-score per 0.5 SD (range −4.0

Conclusions

The FRAX and the Garvan fracture risk calculator are both widely available tools for individualized fracture risk prediction in daily practice. The FRAX model is implemented in the NOF, NOS and NOGG guidelines and most widely used at present. Both models still need to be validated in different populations before they can be generalized to other populations, since the background risk for fractures is population specific. Clinicians should take into account the differences between the models

Contributors

Tineke van Geel is the principle investigator and the main author of the manuscript. Piet Geusens and Geert-Jan Dinant are the supervisors of the principle investigator and responsible for the medical and scientific content of the manuscript. Joop van den Bergh has contributed substantially to the intellectual content of the manuscript. All authors have read and approved the manuscript and agree with publication of their names.

Conflict of interest

All authors report no conflict of interest.

Provenance

Commissioned and externally peer reviewed.

References (56)

  • J.A. Kanis et al.

    An update on the diagnosis and assessment of osteoporosis with densitometry. Committee of Scientific Advisors, International Osteoporosis Foundation

    Osteoporosis Int

    (2000)
  • S. Helden van et al.

    Bone and fall-related fracture risks in women and men with a recent clinical fracture

    J Bone Joint Surg Am

    (2008)
  • J.A. Kanis et al.

    Assessment of fracture risk

    Osteoporosis Int

    (2005)
  • N.D. Nguyen et al.

    Development of prognostic nomograms for individualizing 5-year and 10-year fracture risks

    Osteoporosis Int

    (2008)
  • J.A. Kanis et al.

    Approaches to the targeting of treatment for osteoporosis

    Nat Rev Rheumatol

    (2009)
  • P.D. Delmas et al.

    Strong bones in later life: luxury or necessity?

    Bull World Health Organ

    (1999)
  • W.S. Browner et al.

    Mortality following fractures in older women. The study of osteoporotic fractures

    Arch Intern Med

    (1996)
  • G.S. Keene et al.

    Mortality and morbidity after hip fractures

    BMJ

    (1993)
  • C. Cooper et al.

    Incidence of clinically diagnosed vertebral fractures: a population-based study in Rochester, Minnesota, 1985–1989

    J Bone Miner Res

    (1992)
  • W.F. Lems

    Clinical relevance of vertebral fractures

    Ann Rheum Dis

    (2007)
  • E. Seeman et al.

    7 treatment of osteoporosis: why, whom, when and how to treat. The single most important consideration is the individual's absolute risk of fracture

    Med J Aust

    (2004)
  • K.G. Saag et al.

    Progress in osteoporosis and fracture prevention: focus on postmenopausal women

    Arthritis Res Ther

    (2009)
  • N.D. Nguyen et al.

    Development of a nomogram for individualizing hip fracture risk in men and women

    Osteoporosis Int

    (2007)
  • WHO WHO. FRAX WHO fracture risk assessment tool. 07 September 2009 [cited 02 November 2009]; Web version 3.0:...
  • M. Gunnes et al.

    How well can a previous fracture indicate a new fracture? A questionnaire study of 29,802 postmenopausal women

    Acta Orthop Scand

    (1998)
  • C.M. Klotzbuecher et al.

    Patients with prior fractures have an increased risk of future fractures: a summary of the literature and statistical synthesis

    J Bone Miner Res

    (2000)
  • P.D. Ross et al.

    Prediciting vertebral fracture incidence from prevalent fractures and bone density among non-black, osteoporotic women

    Osteoporosis Int

    (1993)
  • M. Klift van der et al.

    Assessment of fracture risk: who should be treated for osteoporosis?

    Best Pract Res Clin Rheumatol

    (2005)
  • Cited by (30)

    • Adjusting FRAX® on TBS for identification of subjects at high risk of fractures

      2017, Bone
      Citation Excerpt :

      Prevalent vertebral fracture is a major risk factor of fracture. In our study, the percentage of patients with at least one VF using the VFA assessed at the time of the fracture was 37.5%, confirming the high prevalence of prevalent VFs in patients with recent non vertebral fractures [26,27]. As expected the proportion of patients at high risk of fracture (risk of MOF > 20%) is higher when the FRAX® incorporates the presence of VF (FRAX®-VFA) (28.8%), than in the FRAX® alone (24.7%) but slightly modified the value of FRAX® as 33% of patients had a previous non vertebral fracture and therefore an unchanged value.

    • A predictive mechanical model for evaluating vertebral fracture probability in lumbar spine under different osteoporotic drug therapies

      2016, Computer Methods and Programs in Biomedicine
      Citation Excerpt :

      The model uses Damage and Fracture Mechanics concepts to evaluate the fracture probability in an evolutionary algorithm and can be used in a personalized way from BMD measurements in each case. Many studies have been carried out, both in the clinical [14–18,20,21,23,76] and in the simulation fields [19,28–30,32,34,39,40,77]. But a simple and reliable model, useful as a tool for diagnosis and prevention in our daily practice, has not yet been achieved.

    • Reasons for discrepancies in hip fracture risk estimates using FRAX and Garvan calculators

      2016, Maturitas
      Citation Excerpt :

      They each use a different selection of risk factors, and clinicians using both sometimes find that hip fracture estimates are discrepant with respect to the suggested 3% treatment threshold. Most prior comparisons of FRAX and Garvan have focused on average differences in risk estimates across patient groups [9–11], or on hypothetical patients [12,13]. However, it is important for clinicians who use these calculators to understand how the differences between them apply to individual patients in a ‘real world’ context.

    • The Role of the Immune System in the Development of Osteoporosis and Fracture Risk

      2016, Osteoimmunology: Interactions of the Immune and Skeletal Systems: Second Edition
    • The utility of absolute risk prediction using FRAX<sup>®</sup> and Garvan Fracture Risk Calculator in daily practice

      2014, Maturitas
      Citation Excerpt :

      There are a number of differences between the tools. For example, FRAX® includes more health-related risk factors while the GARVAN-FRC includes number of prior fractures and recent falls [1,4,7–10]. Our objective was to investigate the utility of the FRAX® and GARVAN-FRC fracture risk prediction in a 5-year follow-up study.

    View all citing articles on Scopus
    View full text