Fig. 3From: Recruitment and retention of participant and study partner dyads in two multinational Alzheimer’s disease registration trialsForest plots of the estimated relative risk of dropping out comparing adult child dyads to spousal dyads (top panel) and other dyads to spousal dyads (middle panel). The bottom panel shows the estimated relative risk of dropping out comparing to subpopulations of participants differing in age by 10 years. Each of the three plots shows three estimates: an unadjusted model, a model adjusting only for study partner (SP) type or participant age, and a fully adjusted model containing all potential confounders. Notice that the relationship between study partner type and dropout is attenuated after adjusting for participant age, but the relationship between participant age and dropout remains after adjusting for study partner type and all potential confoundersBack to article page