ORs were estimated from attrition-weighted generalised estimating equations of falls occurring after randomisation that were reported at each follow-up year. The model adjusted for the number of falls in the year before the baseline assessment, pair status for randomisation, and baseline age, sex, race, recruitment source, field site, education, diabetes, hypertension, history of stroke, smoking status, 11-item Center for Epidemiologic Studies Depression Scale score, Short Physical Performance Battery balance score, hearing loss severity, Hearing Handicap Inventory for the Elderly Screening Version score, and global cognition factor score. Interactions between randomisation and time, and between time and each covariate, were specified. The x-axis shows the adjusted OR for the intervention versus control with values greater than 1 (favouring the control) to the left and values less than 1 (favouring the intervention) to the right of the vertical line; the x-axis scale is not linear with values above 1 compressed for presentation purposes. OR=odds ratio.
00088-x%2Ffig3_a.webp&w=1920&q=75 "Learn How to Make Forest Plot for Your Research Paper")