Elsevier

Maturitas

Volume 83, January 2016, Pages 65-71
Maturitas

Fracture experience among participants from the FROCAT study: what thresholding is appropriate using the FRAX tool?

https://doi.org/10.1016/j.maturitas.2015.10.002Get rights and content

Highlights

  • Previously, a Spanish FRAX risk thresholds of risk had been proposed from FRIDEX cohort.

  • We display the suggested algorithm in the FROCAT, a population based cohort.

  • The main fracture risk categories (low, intermediate, high) show concordance.

  • The frequencies of fragility fractures over 10-year period are similar in FRIDEX and FROCAT cohorts.

Abstract

Objective

To perform an external validation of FRAX algorithm thresholds for reporting level of risk of fracture in Spanish women (low <5%; intermediate ≥5% and <7.5%; high ≥7.5%) taken from a prospective cohort “FRIDEX”.

Methods

A retrospective study of 1090 women aged ≥40 and ≤90 years old obtained from the general population (FROCAT cohort). FRAX was calculated with data registered in 2002. All fractures were validated in 2012. Sensitivity analysis was performed.

Results

When analyzing the cohort (884) excluding current or past anti osteoporotic medication (AOM), using our nominated thresholds, among the 621 (70.2%) women at low risk of fracture, 5.2% [CI95%: 3.4–7.6] sustained a fragility fracture; among the 99 at intermediate risk, 12.1% [6.4–20.2]; and among the 164 defined as high risk, 15.9% [10.6–24.2]. Sensitivity analysis against model risk stratification FRIDEX of FRAX Spain shows no significant difference. By including 206 women with AOM, the sensitivity analysis shows no difference in the group of intermediate and high risk and minimal differences in the low risk group.

Conclusions

Our findings support and validate the use of FRIDEX thresholds of FRAX when discussing the risk of fracture and the initiation of therapy with patients.

Introduction

Osteoporosis is an asymptomatic bone disease that can lead to an increased risk of fragility fractures, commonly occurring after minor falls. It is the most common musculoskeletal disease in humans and has a growing impact on the public health systems of developed countries due to their aging populations [1], [2], [3], [4], [5], [6].

Traditionally, Bone Mineral Density (BMD), measured by a Dual-energy X-ray Absorptiometry (DXA) scan, has been the main predictor of fragility fracture [7], [8]. Despite the significant influence of BMD on the overall risk of fracture, several studies have shown that taken in isolation, it fails to deliver a cost-effective population screening test [7], [8], [9]. The current practice in most developed countries is to identify patients at high risk of fragility fractures taking into account the presence of other risk factors besides densitometric osteoporosis [9], [10], [11], [12], [13], [14].

The European Society for Clinical and Economic Evaluation of Osteoporosis and Osteoarthritis (ESCEO) [15], proposes a combined assessment of BMD and clinical risk factors for fracture to decide both diagnostic and therapeutic interventions, and the best known and most widely used is the FRAX® Tool (Fracture Risk Assessment®), which is freely available online [16]. This tool calculates the absolute risk of osteoporotic fracture over a 10-year period, considering clinical risk factors independent of bone mass in the male and female population between 40 and 90 years old, who have not received anti osteoporotic medication (AOM) [17]. FRAX is a computer-based algorithm, developed to evaluate the 10-year probability (absolute risk) of a major osteoporotic fracture (clinical spine, forearm, hip or shoulder) and the 10-year probability of hip fracture alone [http://www.shef.ac.uk/FRAX/]. This tool integrates 10 of the clinical risk factors that have shown a strong association with the incidence of fracture in previous studies according to WHO experts. It is able to recalculate the risk itself with inclusion of BMD at the femoral neck (FN) (g/cm2 or T-score). Therefore the FRAX algorithm gives the overall absolute risk for the four main fractures as well as proximal femur alone if needed [15], [16], [17], [18], [19].

The FRAX models have been developed from studying population-based cohorts from Europe, North America, Asia and Australia [20], [21]. As its developers specify, FRAX is calibrated to countries where population fracture risks and mortality rates are known. This is because the probability of fracture is calculated taking into account both the risk of fracture and the mortality rate [20], [21]. There is a consensus in approaching fracture probability based on the combined assessment of clinical risk factors, along with BMD and age, to improve sensitivity fracture prediction without specificity being adversely affected [22]. FRAX authors also specify that due to the epidemiological and economical variability across countries for medical interventions for preventing fractures, cost-effective intervention thresholds have to be country-specific as, for example, it has been made in the United Kingdom [23].

In Spain, to evaluate the fracture risk, the data used for the FRAX country specific algorithm came from different studies, most of which were retrospective hospital studies from the 1990s although a later study showed similar results [24]. There is hence an urgent need for updating fracture incidence and mortality data to provide a better approach to fracture predictions [20], [21]. In addition to what has been said, recent Spanish female population cohort studies have assessed the predictive ability of the FRAX tool locally, and analyzed the FRAX discriminative and predictive ability to predict major osteoporotic fractures [5], [25], [26], [27], [28]. The ability of the FRAX tool to discriminate between Spanish women with high or low fracture risk shows acceptable values that are similar to studies in other populations [26], [27], [28]. A more recent refinement led to the construction of a calibrated model to determine three levels of FRAX risk (low, intermediate, high) based on the analysis of the main fracture outcomes of women from the FRIDEX cohort over a 10-year period of follow up [25], [26] that better identified women at high risk of fracture (Fig. 1).

The aim of this study was to apply the same thresholds proposed by the FRIDEX study in another general population of women recruited to the FROCAT cohort.

Section snippets

Methods

The FROCAT cohort represents a Spanish cohort of men and women aged ≥40 and ≤90 years old assigned to family physicians participating in the study that were working in the Public Health Services and practices managed by the Catalan Health Institute. This institution is the main public provider of health services in Catalonia-Spain and covers around 83% of the 7.5 million population and has computerized medical records of their patients since 2001. Each family physician has in charge a group of

Results

The study population was comprised of 1090 women. Table 1 shows the baseline characteristics of the participants with the most important measurements and risk factors analyzed. A total of 154 women (14.1%) reported previous fragility fractures and 119 (10.9%) reported parental hip fracture. There were 331 (30.4%) women that suffered falls during the previous year of the end of study (2011–2012) and 206 (18.9%) women were categorized as current or past users of anti osteoporotic medication.

Discussion

The study results show that at a population level, the adjusted thresholds suggested based on the FRAX algorithm for low, intermediate or high risk of fracture in a Spanish female population perform well for prediction of incident fracture in a general population based cohort study.

Family history of hip fracture has been shown as an independent contributory factor to fracture risk in meta analyses and other studies [1], [5], [19], but no significance has been shown in this study, even though a

Funding

This study was supported in part by a research grants from the Instituto de Salud Carlos III, Ministry of Science [PI09/90507] and the Institut d'Investigació en Atenció Primària IDIAP Jordi Gol. Barcelona. Spain.

List of contributors

R. Azagra MD, PhD.

M. Zwart MD, MSc.

A. Aguyé MD.

J.C. Martín-Sánchez MSc.

E. Casado MD.

M.A. Díaz-Herrera NG.

D. Moriña PhD.

C. Cooper FMedSci.

A. Díez-Pérez MD, PhD.

E.M. Dennison PhD.

Authors’ roles

Study conduct: RA.

Data collection: FROCAT Study Group.

Data analysis: RA, JCM, DM.

Data interpretation: RA, MZ, ED.

Drafting manuscript: RA, MZ, ED.

Revising manuscript content: RA, MZ, CC, ADP, ED.

Approving final version of manuscript: RA, MZ, AA, JCM, EC, MAD, DM, CC, ADP, ED.

RA takes responsibility for the integrity of the data analysis.

Ethical approval

This study was approved by the Clinical Research Ethics Committee of the Institut d’Investigació en Atenció Primària IDIAP Jordi Gol. Barcelona. Spain.

Informed consent was obtained from all patients.

Conflict of interest

The authors declare no conflict of interest.

FROCAT study group

Vallés Occidental-Barcelona: M Carmen Yuste, CAP Badia, Badia del Vallés, Barcelona; Nuria Puchol, CAP Badia, Badia del Vallés, Barcelona, Milagros Iglesias, CAP Badia, Badia del Vallés, Barcelona; Francesc Solé, CAP Badia, Badia del Vallés, Barcelona; Francesc Julià, CAP Badia, Badia del Vallés, Barcelona; Genís Roca, CAP Sant Llatzer-Terrassa, Barcelona; Sergi Ortiz, CAP Canaletes-Cerdanyola, Barcelona. Vallés Oriental-Barcelona: Amada Aguyé, CAP Granollers, Barcelona; Mireia Rosas, CAP

Acknowledgements

We are grateful to all collaborating researchers in collecting information during fieldwork. We would also like to thank the subjects whose participation made this investigation possible.

References (30)

  • A. Pages et al.

    Burden of osteoporotic fractures in primary health care in Catalonia (Spain): a population-based study

    BMC Musculoskelet. Disord.

    (2012)
  • R. Azagra et al.

    Changing trends in the epidemiology of hip fracture in Spain

    Osteoporos. Int.

    (2014)
  • K.M.B. Huntjens et al.

    Risk of subsequent fracture and mortality within 5 years after a non-vertebral fracture

    Osteoporos. Int.

    (2010)
  • D. Marshall et al.

    Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures

    BMJ

    (1996)
  • O. Johnell et al.

    Predictive value of BMD for hip and other fractures

    J. Bone Miner. Res.

    (2005)
  • Cited by (14)

    View all citing articles on Scopus
    View full text