Risk Factors and Population-Attributable Fractions for Incident Hip Osteoarthritis

Abstract

Background: Despite the huge burden of hip osteoarthritis (OA) and the lack of effective treatment, research into the primary prevention of hip OA is in its infancy. Purpose: We sought to evaluate risk factors for incident clinical and incident radiographic hip OA among middle-aged and older adults, to evaluate the importance of risk factors from a preventive perspective, and to estimate the percentage of new cases attributable to these risk factors. Methods: We retrospectively reviewed data from the Rotterdam study, an open-population cohort study of individuals aged 55 years or older. Data including baseline age, sex, body mass index, smoking status, education level, diagnosis of diabetes, C-reactive protein (CRP), cam morphology, acetabular dysplasia, radiographic thumb OA, radiographic hip OA, and hip pain were assessed for their association with incident clinical hip OA and incident radiographic hip OA separately, after 11 years of follow-up. The population-attributable fractions (PAFs) of statistically significant modifiable risk factors were calculated, as well. Results: New onset of clinical hip OA was seen in 19.9% (544 of 2729) and incident radiographic hip OA in 9.9% (329 of 3309). Female sex, education level below average (PAF 21.4%), and radiographic hip OA (PAF 3.4%) were statistically significantly associated with incident clinical hip OA. Female sex, age, overweight (PAF 20.0%), cam morphology (PAF 7.9%), acetabular dysplasia (PAF 3.6%), and radiographic thumb OA (PAF 4.7%) were statistically significantly associated with radiographic hip OA. Conclusions: Our retrospective analysis suggests that, from a primary prevention perspective, the most important modifiable risk factors among middle-aged and older individuals may be low educational level for incident clinical hip OA and overweight for incident radiographic hip OA. Further study is warranted.