Explaining Variability in the Prevalence of Achilles Tendon Abnormalities: A Systematic Review With Meta-analysis of Imaging Studies in Asymptomatic Individuals
Sean Docking, Harvi F. Hart, Ebonie Rio, Madeline Hannington, Jill Cook, Adam G Culvenor
Abstract
Objective To estimate the prevalence of, and factors associated with, Achilles tendon abnormalities observed on imaging in asymptomatic individuals. Design Systematic review with stratified meta-analysis and meta-regression. Literature Search Embase, Scopus, MEDLINE, CINAHL, SPORTDiscus, and Web of Science were searched from 1980 to August 2020. Study Selection Criteria We included studies that reported the prevalence of Achilles tendon abnormalities, observed with any imaging modality, in an asymptomatic population. We excluded studies if participant mean age was younger than 12 years or if participants had current/previous lower-limb tendon injuries/symptoms or other systemic conditions. Data Synthesis Random-effects proportion meta-analysis was used to estimate prevalence. We used meta-regression for continuous variables (mean age and body mass index [BMI], sample size, proportion of female participants) and stratified categorical variables (imaging modality and participation in physical activity) to explain between-study heterogeneity. Results We included 91 studies (10 156 limbs, 5841 participants). The prevalence of Achilles tendon abnormalities on imaging ranged from 0% to 80% per participant. Between-study heterogeneity was high (I 2 >90%, P<.001), precluding data pooling. Between-study heterogeneity was partly explained by participant mean BMI (slope, 2.8% per 1-unit increase in BMI; 95% confidence interval: 0.57%, 5.03%; P = .015) and participation in physical activity per limb, and mean age of 40 years old or older (P = .022) per participant. Conclusion There was substantial variability in the prevalence of Achilles tendon abnormalities on imaging in asymptomatic individuals. Higher prevalence of abnormalities was associated with older age (40 years old or older), higher BMI, and participation in physical activity. A large proportion of heterogeneity remains unaccounted for, likely due to variations in abnormality definitions and study design. J Orthop Sports Phys Ther 2021;51(5):232–252. Epub 28 Mar 2021. doi:10.2519/jospt.2021.9970