Contribution of elevation and residential proximity to the coast in explaining geographic variations in hip fracture incidence. A Norwegian Epidemiologic Osteoporosis Studies (NOREPOS) study.

Abstract

Norway is an elongated country extending north of the Arctic Circle with substantial coast-inland variation in topography and climate. Differences in hip fracture incidence between counties and a distinct seasonal variation have previously been shown. The aim of the current study was to explore these variations further by considering associations of height above sea level (elevation) and distance to the coast with hip fracture incidence. All patients with hip fractures admitted to Norwegian hospitals in the period 2009-2018 were included. Individual residential elevation and distance to the coast was calculated in Geographic Information Systems and combined with individual-level population data on all Norwegians 50years of age or older during the observation period, including hospital information on fractures. Age-standardized incidences rate and incidence rate ratios with 95% confidence intervals (IRR, 95% CI) according to elevation and coastal proximity were estimated. The associations were tested using Poisson models adjusting for sex, urban/rural location of residency, country of birth, and season of hip fracture occurrence. From 2009 to 2018, there were 85,776 first hip fractures. There was an increasing risk with higher residential elevation (above versus below mean) for women: IRR = 1.04, 95% CI: 1.02, 1.05), but not for men (IRR = 1.00, 95% CI: 0.97, 1.02). Incidence of hip fracture increased with distance from the coast. Women residing the farthest away from the coast (above versus below mean distance) had a higher age-adjusted incidence of hip fracture compared to those living closer to the coast (IRR = 1.04 (95% CI: 1.02, 1.06), whereas no association was found in men (IRR = 1.00 (95% CI: 1.00, 1.01). Combining elevation and distance to coast showed a higher incidence in women living at high elevation far from the coast compared with women living at low elevation near the coast (IRR = 1.07, 95% CI: 1.04, 1.10). A similar result was found in men but only for hip fractures occurring during March-May (IRR = 1.07, 95% CI: 1.00, 1.15). The previously shown patterns of county differences and seasonal variations were unchanged when considering geography. We found a somewhat higher incidence of hip fracture in inland residents living in areas of high elevation, as compared to those living in more coastal proximity; however, the geographic variation did not explain county and seasonal differences in fracture incidence in Norway. More in-depth analyses on temperature and climate factors may give further clues.