Die Prävalenz der radiologischen Gonarthrose in Bezug zu Lebensalter, Geschlecht, Jahrgangskohorte und ethnischer Zugehörigkeit. Eine Metaanalyse

AIM: This metaanalysis was performed to evaluate the prevalence of the radiological assessed knee osteoarthritis in the whole community. MATERIALS AND METHODS: Medical databases (Medline, EMBASE, Cochrane) were searched for the strategy: ["Osteoarthritis" and "Knee" and "Prevalence"]. The deadline for the search was 31.12.2009. Two investigators (first and senior author) independently made the selection from 17 studies (from a total of 1428) according to the inclusion criteria: a cross-sectional study of the whole community, radiological investigation and definition of knee ROA by an established radiological score. Only studies in English or German language were evaluated. Effect sizes (event rate, odds ratio [OR] and confidence interval [CI]) were calculated by the software "Comprehensive Metaanalysis V2". Study heterogeneity (I2) was determined accordingly to Higgins. RESULTS: The kappa index for interobserver validity was k = 0.948. All studies judged the grade of osteoarthritis according to the Kellgren-Lawrence (KL) score. For calculation of knee ROA KL grades 2+ were estimated only. The total prevalence of knee ROA was 24.3 % (CI 23.4-25.2 %). The whole prevalence in male patients was 24.3 % (CI 23.4-25.2 %); I2 = 59.4 (p = 0.002) and in female patients 32.6 % (CI 31.8-33.4 %); I2 = 49,1 (p < 0.001). Younger male patients (age 50-) had a prevalence of 5.6 (CI 4.5-6.8). In older patients (80+) the male prevalence was 44.5 % (CI 39.6-49.5 %). In this age group female patients had a prevalence of 71.6 % (CI 67.6-75.3 %). The higher prevalence of knee ROA in female patients was significant (OR = 1.8 [1.7-1.9]; I2 = 46.0 [p < 0.001]). The prevalence of knee ROA was higher in male Asians compared with male Caucasians (OR = 1.1, CI 0.9-1.2; p = 0.080) in tendency. This difference was significant in female patients (OR = 2.2; CI 2.0-2.4; p < 0.001). Furthermore another trend was evaluated. Female patients (70-79 years) from the birth-year cohort 1920- had a prevalence of 37.8 % (CI 35.9-39.7)%. In contrast female patients from the birth-year cohort 1920 had a prevalence of 62.8 % (CI 60.8-64.8 %) at 70-79 years. This difference was significant (OR = 2.8; CI 2.5-3.1; p < 0.001). CONCLUSIONS: This investigation confirms the high prevalence of knee ROA. The evaluated data may serve as a reference for medical or scientific investigations in the future.