A linear regression model can estimate hip fracture incidence

Abstract

Hip fractures are a real health problem in the elderly [1]. Before launching a program for hip fracture prevention, knowing the incidence of hip fractures in a determined region is mandatory. In 2006 we published in Archives of Osteoporosis a manuscript on the incidence of osteoporotic hip fractures in our area (Salamanca, Spain) over the period 1994–2002, establishing a mathematical regression model to estimate the number of fractures up to 2015 [2]. Following the same methods used in this previous work [2], we collected data on hip fractures during the period 2004–2009 in order to ascertain whether the proposed model proved to be exact or not. These data are presented in Tables 1 and 2, and they show that the incidence of osteoporotic hip fractures in 2004–2009 concurs with the estimation of our mathematical model. Although incidence of fractures is increasing with time, it seems to reach a plateau or even showing some decrease. Some authors have pointed out the fact that the incidence of hip fractures is declining in the last few years [3, 4]. This change in the trend towards a decrease in hip fracture incidence is not homogeneous and may vary among