Risk categorisation through standard deviations – the challenge of bone density measurements: A focus group study among women attending the Nord-Trøndelag Health Study (HUNT)

Abstract

Bone density measurements play an important part in the categorisation of osteoporosis as a risk factor in modern medicine. According to the World Health Organization, people are categorised as having osteoporosis when their bone mineral density (BMD) value is 2.5 standard deviation (SD) below the young adult mean value, and as having osteopenia when the value lies between one and 2.5 SDs below the young adult mean value. The categorisation according to SDs makes osteoporosis a rather unique case within the medical world of risk definitions. We invited women who had their bone mass scanned in the Nord-Trøndelag Health Study (HUNT) to participate in focus groups and share their scanning experiences. Nine groups of women met three times for a total of 27 focus group discussions. Our findings illustrate that having their BMD measured contributed to a substantial confusion, related in particular to feedback given as SDs, the choice of reference populations and the body sites chosen for BMD. Some of the women had had their BMD measured at different health institutions using different reference groups, which resulted in women being told that they were osteoporotic at one institution, but not at the other. As the different institutions also scanned various body sites, including the wrist, hip and spine, the women were also confused about what site(s) provides the best information about their bone status. Overall this study shows that osteoporosis presents us with a particularly challenging example of risk categorisation.

Keywords:

• risk
• osteoporosis
• bone mass density
• screening

Acknowledgements

This study has been financially supported by the Norwegian Women's Public Health Association and the Research Council of Norway. We thank the HUNT Research Centre for their assistance in the recruitment of the participants. Mrs Liv Åldstedt made a valuable contribution by transcribing the majority of the focus group discussions. We are also grateful to two anonymous reviewers for their valuable comments on an earlier version of this manuscript.

Notes

1. This argument is about which choice that serves best to identify the individuals that will suffer a fracture. Applying the T-score can be criticised for identifying too many as osteoporotic, whereas the Z-score can be seen as identifying too few. In both cases the problem is related to the translation of population data to individual cases.

2. However, the absence of risk, i.e. an acceptably low probability of experiencing an unwanted event, cannot be soundly inferred from a single identified incident of non-occurrence, or vice versa.