SCD is increasingly recognized as the earliest manifestation of Alzheimer’s Disease (AD), preceding mild cognitive impairment (MCI). However, SCD encompasses a wide range of clinical presentations and may result from various potentially treatable conditions. Identifying distinct phenotypes could enhance patient stratification and aid in predicting clinical trajectories.

We analyzed data from the AD Neuroimaging Initiative (ADNI - MERGE) database, including 353 individuals with SCD, 542 cognitively normal controls (CN), and 1113 patients with MCI. Clustering methods were applied to the SCD cohort to identify distinct cognitive subgroups. Between-cluster comparisons were conducted based on cognitive performance, cerebrospinal fluid (CSF) biomarkers (including amyloid-beta [Aβ₄₂], total tau [t-tau], and phosphorylated tau [p-tau], and neuroimaging data (structural magnetic resonance imaging [MRI] and [18F] fludeoxyglucose [FDG] positron emission tomography [PET]). Additionally, longitudinal assessments were conducted to evaluate clinical progression.

Three distinct SCD clusters were identified: (1) a “dysexecutive” cluster with poorer executive functioning, showing brain hypometabolism on FDG-PET, and levels of both p-tau and Aβ₄₂ resembling those of the MCI group; (2) a “worried-well” cluster, with individuals reporting subjective memory and attentional complaints in the absence of objective cognitive impairment; (3) an “amnesic” cluster, marked primarily by memory deficits. Longitudinal analysis showed that cluster 1 exhibited the greatest clinical decline, suggesting a higher risk for progression to neurodegenerative diseases, while clusters 2 and 3 remained relatively stable over time.