Geographical variation in osteoporosis in Australian women: a longitudinal analysis over 23 years

Osteoporosis affects over half of older women, whether urban compared to rural residents have different risk of osteoporosis is unclear. This 23-year longitudinal study of Australian women found lower risk of osteoporosis and fractures among women living in rural compared with urban areas, which may relate to distribution of risk factors and/or screening opportunities.

To determine whether the prevalence of osteoporosis, fractures, and dual-X-ray absorptiometry (DXA) differs between Australian women living in rural compared with urban areas.

The Australian Longitudinal Study on Women's Health, a prospective longitudinal study of Australian women, includes a cohort of women born 1946-1951, surveyed nine times from 1996-2019. Data from administrative health records were linked to survey data. Geographic area was classified as major city, inner regional, outer regional, or remote. Generalised estimating equations (GEE) explored outcomes of osteoporosis, fracture, and DXA rates. Univariable and multivariable regression were performed and bootstrapping with 100 repetitions at 95% sampling of the original dataset to ensure robust results.

A total of 13,712 women were included: 5000 (36.5%) living in major cities, 5214 (38.0%) inner regional, 2798 (20.41%) outer regional, and 700 (5.1%) remote areas. Baseline age, 47.6 (1.46) years (mean (SD)), was similar; but education, marital status, country of birth, smoking, BMI, and comorbidities varied between areas. Over 23 years, 2956 (21.6%) were diagnosed with osteoporosis, 3185 (23.2%) had an incident fracture, and 8151 (59.4%) had a DXA. On univariable analysis, women living outside major cities had lower risk of osteoporosis (inner regional OR 0.92; 95% CI 0.84, 0.99; outer regional 0.70; 0.63, 0.79; remote 0.43; 0.34, 0.53), and lower risk of fractures (inner regional area OR 0.92, 95% CI 0.84, 0.99; outer regional 0.81; 0.73, 0.90; remote 0.54; 0.42, 0.67), than women living in major cities. These differences were no longer significant on multivariable analysis, suggesting differences might be related to the distribution of risk factors. Women living in inner regional and outer regional areas were significantly less likely to have a DXA, than those living in a major city (OR 0.75, 95% CI 0.70, 0.81; OR 0.72; 95% CI 0.64, 0.77, respectively), and this remained significant on multivariable analysis.

Osteoporosis or fracture affected one-third of older Australian women, and women living outside major cities had lower rates of osteoporosis, fractures, and DXA completion.