Table 2.

Sensitivity analyses for the primary outcome

Reason sensitivity analysis	How is the sensitivity analysis performed?
Check the influence of inappropriate inclusion, potentially confounding co-morbid conditions and trial medication use	Participants will be excluded from the analysis if: • major inclusion or exclusion criteria are violated • incorrect trial medication is taken • major neurological or psychiatric disease was present at baseline
Check the influence of classifying participants who did not understand the instructions at follow-up as having accelerated cognitive decline	The last observation carried forward method will be used for patients with missing MMSE and A&E RBI-scores at follow-up if the reason for missing is the inability of the patient to understand the instructions (instead of classifying them as having accelerated cognitive decline)
Check for bias by differential lost to follow-up (worst case scenario)	All patients with missing MMSE and A&E RBI-scores at follow-up will be considered to have accelerated cognitive decline
Investigate the impact of further baseline variables on the RBI score result, Check for confounding by depression symptoms	Age, gender, years of formal education, race, ethnicity and language and CES-D (score < 16, ≥16) are included as covariates in the logistic regression analysis