The heterogeneity of subjective cognitive decline: A data‐driven approach on a population‐based sample

Abstract

AbstractBackgroundSubjective Cognitive Decline (SCD) has been largely studied as a risk condition for cognitive decline, mainly in memory clinics or research settings. SCD samples are highly heterogeneous, including both individuals with early Alzheimer’s disease and others with psychological vulnerabilities and/or physical comorbidity. Aims: identify distinct SCD subgroups in a population‐based sample of cognitively normal (CN) older adults; explore cross‐sectional differences among SCD subgroups and controls (noSCD); investigate the effect of SCD status on 8‐year cognitive changes.MethodThe InveCe.Ab population‐based study (NCT01345110) enrolled people aged 70‐74 years residing in Abbiategrasso (Milan). Participants underwent multidimensional assessment (blood sampling, lifestyle interview, medical and neuropsychological assessment) at baseline (Nov 2009‐Jan 2011) and after 2, 4 and 8 years. CN participants at baseline completing at least 60% of neuropsychological assessment were included in the present investigation. Those reporting any cognitive complaints to the geriatrician were classified as SCD. Hierarchical cluster analysis was conducted in SCD including 9 relevant variables: education, depressive symptoms, cardiovascular risk, comorbidity, subjective cognitive complaints (numbers) and 4 neuropsychological scores measuring memory, executive functions, visuospatial abilities and language respectively. Cross‐sectional differences among the SCD subgroups identified and noSCD were performed using Kruskal‐Wallis test. Longitudinal cognitive changes according to baseline SCD status (noSCD, SCD subgroups) were estimated using linear mixed models, adjusted for age and education.ResultTwo clusters were identified in SCD. Table 1 shows cross‐sectional results. SCD1 (N=141), compared to SCD2 and CN, were less educated, have higher comorbidities, depressive symptoms and SCD complaints. SCD1 showed lower performances on executive and visuospatial functions, while SCD2, despite being healthier and more educated, showed reduced episodic memory. At longitudinal analysis, all cognitive functions declined due to aging (Time; p<.001). SCD2 displayed a steeper worsening on executive function (Group*Time; p<.001).ConclusionThe presence of SCD at the population level could be informative of future cognitive decline when reported by individuals with higher education. Future studies are needed to clarify the meaning of this condition in the aged population, in order to identify early those at increased risk of actual cognitive decline.