Endocrine Abstracts (2018) 57 009 | DOI: 10.1530/endoabs.57.009

The accuracy of self-reported fractures among a Belgian cohort of postmenopausal women: The FRISBEE study

F Baleanu1, P Bergmann2, V Kinnard3, L Iconaru1, SI Cappelle3, M Moreau4, M Paesmans4, R Karmali1 & JJ Body1


1Department of Endocrinology, CHU-Brugmann, Université Libre de Bruxelles, Brussels, Belgium; 2Department of Nuclear Medicine, CHU–Brugmann, Université Libre de Bruxelles, Brussels, Belgium; 3Department of Internal Medicine, CHU-Brugmann, Université Libre de Bruxelles, Brussels, Belgium; 4Data Centre, Inst. J. Bordet, Université Libre de Bruxelles, Brussels, Belgium.


In large population-based epidemiological studies of osteoporotic fractures, self-report is an important way of obtaining information. However, this method is subject to errors of recall and may result in misclassification of fracture status. Surprisingly, the accuracy of self-reported fractures has only rarely been assessed. The purpose of our study was to assess the accuracy of self-reported fractures in the FRISBEE cohort (Brussels, Belgium) of 3560 postmenopausal, aged 60–85 years. Baseline assessment parameters were collected during an interview by trained nurses. Participants were followed yearly by phone call for the occurrence of incident fragility fractures. From 967 reported fractures, 79.3% (n=767) were radiologically confirmed. Among the 20.7% (n=200) unconfirmed fractures, 56.5% (n=113) had no fracture (true false positive; the radiology report indicated that the area was investigated but no fracture was found), for 21% (n=42) no radiology report was available (no x-ray was taken or not enough information was given to find the record), 16% (n=32) reported an existing fracture (the x-ray at the time the subject reported the fracture showed an old fracture), and 6.5% (n=13) of fractures were unconfirmed because of an equivocal radiology report or wrong declared area. Based on the fracture site, among the 56.5% (n=113) of true false positive, we found a percentage of 2.7% (n=3) for hip, 9.7% (n=11) for wrist, 9.7% (n=11) for humerus, 23% (n=26) for spine, 10.6% (n=12) for ankle, 5.3% (n=6) for pelvis and 38.9% (n=44) for ‘minor’ fractures (face/skull, ribs, knee, carpal/metacarpal bones, tarsal/metatarsal bones). Further, we investigated the characteristics of individuals who gave a ‘wrong information’ by using a multivariate analysis - covariates - age, BMI, fracture site, ethnicity, education, smoking, alcohol intake, history of fracture, falls, insomnia, physical activity, calcium and vitamin D intake. We found that subjects with a higher BMI (>25), with fractures on other site than hip, a lower education level, sedentarity and subjects taking calcium supplements were more likely to report unvalidated fractures. In conclusion, the inaccuracy of self-reported fractures is far from being negligible for wrist, humerus, ankle and spine and is inacceptably high for fractures considered as minor.

Article tools

My recent searches

No recent searches.