Prediction of Progression Along the Transition Points within the Continuum of Cognitive Impairment from Subjective Cognitive Complaints to Dementia Using Dyadic Patterns of Subjective Reporting

Abstract

AbstractBackgroundThe validity of Subjective Cognitive Complaints (SCCs) from dyadic patterns to predict progression to dementia is not yet clear. Some studies suggest that the validity of informant report in predicting dementia increases as cognitive function and awareness of symptoms decline. Our aim was to compare validity of informant and participant reports, and its agreement, to predict progression to MCI and conversion to dementia.MethodA total of 226 participants from the CompAS study were longitudinally assessed (intervals 18–24 months). The sample consisted of SCD (198) and MCI (28) participants. SCCs from participants and informants were assessed at baseline, 1st and 2nd follow‐ups using the QAM questionnaire. Informant‐participant total score agreement (disagreement, over‐ and under‐estimation, were identically considered) were calculated. Logistic regressions were separately performed to distinguish SCD stables and participants progressing either from SCD to MCI/dementia or from MCI to dementia using SCCs reports from participants, informants and agreement as predictive variables.ResultsInformant report at baseline significantly predicted conversion to dementia from baseline to the 3rd follow‐up (β = .173; SE = .048; p<.001; OR = 1.189, CI = 1.082‐1.306), similarly to that observed for informant reports at 1st follow‐up to predict conversion from 1st to the 3rd follow‐up (β = .225; SE = .050; p = .001; OR = 1.252; CI = 1.135‐1.381), and at 2nd follow‐up to predict conversion from 2nd to the 3rd follow‐up (β = .271; SE = .083; p<.001; OR = 1.313; CI = 1.110–1.552) (see Table 1).Informant report at baseline also successfully predicted the progression to MCI from baseline to the 3rd follow‐up (β = .139; SE = .057; p = .017; OR = 1.49; CI = 1.025‐1.288), similarly to that observed for informant reports at 1st (β = .124; SE = .056; p = .026; OR = 1.132; CI = 1.015‐1.264), and the 2nd follow‐ups (β = .195; SE = .073; p = .008; OR = 1.216; CI = 1.053‐1.404) (see Table 2).Conversion to dementia (β = .113; SE = .048; p = .018; OR = 1.119; CI = 1.020‐1.229) and progression to MCI (β = .138; SE = .054; p = .011; OR = 1.148; CI = 1.032‐1.277) using Self‐report were only significantly predicted using 1st follow‐up reports.Agreement did not significantly predict the progression to MCI or dementia at any of measurement points.ConclusionsInformant report successfully predicted progression and conversion at any transition point. Self‐reports were only predictive at 1st follow‐up. Agreement did not significantly predict progression and conversion.