Are annual fluctuations in hip fracture incidence dependent upon the underlying mortality rate?

There is substantial variation in hip fracture incidence rates from one year to the next. To determine whether this is related to varying mortality rates in the underlying population, we carried out an ecological analysis in Rochester and Olmsted County, Minnesota. In a given year, mortality in the general population was not related to hip fracture incidence nor did the fracture incidence rates influence the overall death rate. There was a negative correlation between hip fracture incidence and mortality in the preceding year among women, but not men; the trend was statistically significant among women less than 80 years of age. Both cervical and intertrochanteric hip fractures in women had a similar relationship to mortality in the population the year before. There was no consistent relationship between hip fracture incidence and mortality rates two or three years previously. Although many hip fractures occur in frail individuals near the end of life, patterns of mortality in the underlying population cannot account for the annual variation in hip fracture incidence.