Oral Presentation 29th Australian and New Zealand Bone and Mineral Society Annual Scientific Meeting 2019

Observational epidemiology has been critical to understanding the epidemiology of fractures and osteoporosis, with advances particularly being possible post the advent of widely accessible DXA. This combined with increased understanding of the physiology and pathophysiology of bone from laboratory-based research has led to the development of successful pharmacological interventions to prevent minimal trauma fracture. Furthermore, research has also moved to RCTs for non-pharmacological interventions, such as exercise and falls prevention programs.  Given the vast amount of observational data already published and the need for high quality evidence to guide practice, does this mean that observational epidemiology should now be less of a priority?

The answer comes down to whether observational studies are used wisely to answer the right questions. There are key areas in which observational data remain critical to informing clinical practice and research directions.  Key examples relate to:

1. early identification of rare but important adverse events (atypical femoral fractures, ONJ)
2. identification of potential additional off-label uses of existing medications
   1. bisphosphonates and osteoarthritis;
   2. advantages of treatment choices across diseases, particularly important in context of multimorbidity e.g. antihypertensives and fracture; bisphosphonates and CVD
3. assessment of potential utility of new technologies e.g. HRpQCT
4. situations where intervention studies are particularly challenging
   1. impacts of early life factors on bone health and fractures in later life
   2. for complex interventions e.g. dietary patterns and bone health.
   3. ethically difficult studies e.g. imagine attempting a RCT of the impact of smoking on bone health!
5. potential uses of registries, clinical data from primary care and data linkage.

Observational epidemiology remains critically important and a high priority provided we avoid “so what questions” and use it wisely to support practice, policy and future research directions.