Background:​ The present study evaluated the association between blood pressure variability and the risk of hip/vertebral fractures in middle aged and elderly patients.

Methods:​ This was a retrospective observational study of patients attending family medicine outpatient clinics, recruited from 1st January 2000 to 31st December 2003 and were followed up until 31st December 2019. Standard deviation (SD), root mean square (RMS), coefficient of variation (CV) and a variability score (defined as the number of changes in blood pressure (diastolic and systolic) of 5 mm Hg or more) were used as measures of blood pressure variability. The primary outcome was a composite of new onset hip or vertebral fractures.

Results:​ A total of 57,810 patients were included. Over a median follow-up of 5894 days (interquartile range:​ 3505–6487), 3285 patients (5.68%) developed new onset hip/vertebral fractures. The crude incidence rates were 4.95%, 5.31%, and 7.2% for diastolic blood pressure-CV and 5.0%, 5.28%, and 7.08% for systolic blood pressure-CV in the first, second, and third tertiles, respectively. Survival analysis demonstrated differences in hip/vertebral fracture amongst the tertiles of systolic and diastolic blood pressure variability (P < 0.0001).

Conclusions:​ Measures of blood pressure variability were significantly associated with incident hip/vertebral fractures. They can be incorporated into existing clinical scores to improve risk stratification.