Evidence of distinct subjective cognitive decline profiles in a community sample

Abstract

AbstractBackgroundSubjective Cognitive Decline (SCD) refers to a self‐perceived decline in cognitive ability despite normal performance on objective tests. SCD has been conceptualized as a marker of preclinical Alzheimer’s disease (AD) but research has uncovered considerable heterogeneity in symptoms and outcomes. Therefore, it is likely that SCD encompasses distinct subtypes, similarly to Mild Cognitive Impairment (MCI). Here we sought to identify and characterize subgroups based on patterns of subjective cognitive symptoms.MethodsCommunity‐dwelling participants aged ≥50 years (n = 483), located within the United States, were recruited via a crowdsourcing website, and completed a survey consisting of the following: SCD Questionnaire‐MyCog, PROMIS depression scale, and demographic questions. The SCD‐Q includes 24 items assessing self‐perceived recent decline on memory, language, and executive functioning tasks with a yes or no format. Two‐step cluster analysis using AIC was performed on all SCD‐Q items.ResultsThe analysis yielded a 3‐cluster solution with fair cohesion and separation (silhouette measure = 0.46). Cluster 1 (No Subjective Decline) included 36% of participants reporting no decline except for sporadic endorsement (≤2%) on a few memory tasks. Cluster 2 (Mild Subjective Decline) consisted of 42% of participants who frequently (≤50%) endorsed decline on select memory and language tasks, and less frequently on executive functioning tasks (≤25%). Cluster 3 (Moderate Subjective Decline) included 22% of participants who almost unanimously (≥80%) reported decline on a wide range of memory and language tasks, and frequently (≥60%) on executive functioning tasks. Clusters differed on total SCD‐Q scores (1<2<3, M = 0.1, 3.5, 13.7), percent below the SCD cutoff of 7/24 (1 and 2<3, 0%, 8.4%, 100%), depression score (1<2<3, M = 10.9, 13.0, 16.7), gender (percent women 1<2 and 3, 58.7%, 71.0%, 704%) and AD family history (percent with history 1<2 and 3, 20%, 29%, 34.9%) but not on age.ConclusionWe identified three distinct subgroups in our community sample, defined by escalating subjective symptom burden, growing involvement of language and executive functioning domains and increasing prevalence of AD risk factors (female gender/AD family history). All Moderate‐Subjective‐Decline subgroup members but none/very few of the others, scored above the published SCD cut‐off. Longitudinal outcomes are needed to establish subgroup validity.