Is geographic variation in hip fracture rates related to current or former region of residence?

This study examines geographic variation in male and female age-adjusted hip fracture rates in white elderly Medicare enrollees. We assembled a cohort of more than 2 million 1992 enrollees and followed them passively through record linkage for 2 years for a hospitalization containing a diagnostic code indicating hip fracture. We simultaneously estimated rate ratios by region of residence early (inferred from the Social Security number) and late in life. Residence region early in life is associated with substantial variation in hip fracture rates. Conversely, current region has little relation with hip fracture risk.