Maintain Your Brain trial: Early findings and lessons learned from adherence and compliance data

Abstract

AbstractBackgroundTechnology and web‐based approaches potentially provide scalable population‐based interventions to reduce modifiable risk factors for dementia. Key issues in online interventions are recruitment and retention. To devise strategies to improve population reach We investigated which factors influence recruiting and maintaining participants in such an intervention, the in‐progress Maintain Your Brain trial.MethodInvitations were sent to people aged 55‐77 years from the 45 and Up study, a population‐based cohort study of one in ten people aged 45 years and older in New South Wales, Australia (n = 267,000). For MYB, participants were required to be eligible for at least one of four modules to be enrolled (physical activity, nutrition, brain training and mental wellbeing). All participants received modules based on their risks and were randomly allocated to either personalised coaching (intervention) or static information (control). Associations between participant characteristics (listed Table 1) and likelihood of completing set assessment tasks was assessed at two key stages – end of baseline and end of 12‐month follow‐up using stepwise (forward) regression.ResultsOf 96,418 people invited, 12,281 (13%) participants started baseline and completed a mean of 6.2 (SD 4.3) of ten assessments. Of these, 6,236 (6%) were enrolled in the trial. At 12‐months participants completed a mean of 5.0 of 8 assessments (SD = 3.8). Completion rate of the primary outcome (two tasks) was 62% (3,869). In the final regression model for baseline (Table1), overall associations were weak even though statistically significant, with only years of education not entered in the final model. The follow‐up model included retirement status, gender, baseline dementia risk and baseline wellbeing. However, this model (df1 = 1, df2 =6231; R2 = .01) accounted for even less variation than baseline model (R2 = .04).ConclusionOverall, regression models of participant characteristics accounted for a low amount of variation in task completion rates at both baseline and follow‐up. Participants were less likely to complete baseline tasks if they were older, male, not living with a spouse or alone and not retired or had lower dementia risk score and more psychological distress.