Temporal order of clinical symptoms in Alzheimer's disease.

Abstract

Background: Understanding the temporal order of clinical symptoms is critical for improving disease progression assessment in biomarker-defined Alzheimer's disease (AD). However, the sequential emergence of cognitive, neuropsychiatric, and functional manifestations remains incompletely characterized. The purpose of this study is to characterize the population-level sequence in which Alzheimer's disease symptoms appear, to help understand how the disease is progressing.

Methods: We analyzed 932 participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI), including 546 amyloid-positive individuals across the AD continuum and 386 amyloid-negative individuals in the comparison group. Longitudinal cumulative incidence was analyzed to estimate the temporal emergence of major clinical symptoms. Any symptom onset was defined using a > 50% cumulative incidence threshold. Cross-sectional relationships between the sequence of clinical symptoms and Mini-Mental State Examination (MMSE) scores were analyzed using locally estimated scatterplot smoothing (LOESS). The concordance between the clinical symptom sequence and PET-derived Braak staging was assessed using weighted kappa statistics.

Results: A reproducible temporal ordering of clinical symptoms was observed across longitudinal and cross-sectional analyses. Memory impairment emerged earliest, affecting 63.75% of participants at baseline, followed by behavioral and affective symptoms. At intermediate stages, executive dysfunction and instrumental activities of daily living (IADL) impairment affected 52.44% and 54.85% of participants, respectively. Language and visuospatial impairments appeared later, affecting 51.25% and 51.54% of participants, respectively, followed by psychotic symptoms and severe functional decline in basic activities of daily living (BADL), which suggests the disease has reached its late stage. The order in which different symptom groups appear-from memory to executive function and instrumental abilities, then language and visuospatial skills, and finally psychotic symptoms and basic abilities-shows moderate consistency respectively with CDR staging (κ = 0.596, 95%CI: 0.524-0.668), and the PET-derived Braak staging (κ = 0.517, 95%CI: 0.436-0.598).

Conclusion: Clinical symptoms in AD tend to emerge in a temporal sequence. This finding may help to better characterize disease progression clinically, serving as a symptom-staging framework for symptomatic AD and providing interpretable turning points for clinical assessment and individualized management.