Résumé
Background
Subjective cognitive complaints (SCC) signal risk of objective cognitive decline. To better understand the bio‐behavioral factors that signal cognitive trajectories at‐risk for the earliest stages of Alzheimer’ disease, there is a need for well‐phenotyped large‐scale population‐based cohorts with longitudinal assessments.
Method
From the INTUITION decentralized observational brain health study (NCT0505895), using app‐collected self‐report data, we compared the demographic, clinical, and neuropsychological profiles of 2045 SCC versus 4090 age‐ and education‐matched cognitively normal (CN) participants. The SCC cohort was defined by age (> = 50 years) and presence of significant cognitive concerns based on total scores> = 4 on the 14‐item Cognitive Function Instrument (CFI‐14). Using customized assessments from Cambridge Cognition, Pattern Recognition Memory (PRM), Paired Associates Learning (PAL), Spatial Working Memory (SWM), and Match‐to‐Sample (MTS) tests were performed. SCC participants were stratified and compared by presence or absence of dementia risk factors (first‐degree family history of dementia (FH+/FH‐) or 2+ cardiovascular risk factors (CVRF2+/CVRF2‐)).
Result
Cognitive and neuropsychiatric symptom‐burden were greater among the SCC cohort compared to CN participants (e.g., E‐Cog‐12: 1.8(0.5) vs 1.2(0.3), pathological PHQ2 (Total> = 3): 14% vs 2.7%, pathological GAD‐2(Total> = 3): 11.8% vs 1.9%, respectively p<0.001). Lifestyle factors differed, with SCC participants reporting lower level of physical activity, higher substance use and impaired sleep, relative to CN. Among SCC participants the prevalence of CVRF2+ and FH+ were 53.2% and 37.8%, respectively; and 43% (CVRF2+) and 33.5% (FH+) among CN. Among SCC, individuals with CVRF2+ were more likely to be older, male, and with higher rates of anxiety/depression symptoms. Mean total CFI‐14 scores were similar between FH+/FH‐ or CVRF2+/CVRF2‐, but item‐level CFI profiles differed (e.g., word‐finding complaints in FH+/CVRF2+). SCC scored lower on key CANTAB outcomes (e.g., PAL/PRM) evaluated compared to CN. Both CVRF2+ and FH+ participants with SCC performed worse on cognitive measures of learning (SWM) and attention (MTS) compared to CVRF2‐ and FH‐ respectively.
Conclusion
In a large, decentralized cohort, SCC was associated with higher burden of modifiable and non‐modifiable risk factors. First‐degree family history of dementia and cardiovascular disease, when coupled with SCC, revealed more severe measurable objective cognitive deficits. Future work will assess individual longitudinal cognitive trajectories.