Secular increases in the numbers of hip fractures may occur because the proportion of elderly persons in the population is increasing and because the age-specific incidence of hip fractures may increase [1–5]. Melton et al.[6] suggest that the main reason for an increase in the number of hip fracture cases is an increase in the number of elderly in the community. Interestingly, secular decreases in the age-specific incidence of hip fractures have been reported. Melton et al.[4] reported a stabilization and decreased incidence in the last few decades in men and women (Fig. 1). Highly variable incidence rates have been reported in the Mediterranean Osteoporosis Study in 12 regions in Europe, with the female: male ratio varying from 3.8 in Seville, Spain to 0.4 in rural Turkey [5].

Ascertainment errors may be partly responsible for this heterogeneous epidemiology [7, 8]. Greater attention to the problem of osteoporosis in men may have resulted in identification of a problem that has been present for some time rather than reflecting an increase in its incidence.‘Incidence’is calculated from discharge records rather than being derived prospectively. Failure to identify all cases may result when there are several health providers. Readmission or the occurrence of a second hip fracture may inflate the figures. Small sample sizes, particularly in those over 80 years old, may result in an unstable numerator while the population size forming the denominator in the ‘incidence’calculation is often derived from dated census data. Groupings into 75+, 80+ and 85+ years are often made. Most hip fractures occur at this time and the mean age and distribution within these classifications will be changing with time. Errors in the numerator and denominator may markedly alter the incidence estimates. Thus, the rising incidence may be an artifact of better surveillance. Retrospective analyses may result in an underestimated incidence. The lower rates and varying gender ratios in