Empirical classification of neuropsychiatric symptoms and association of classes with diagnostic progression and cognitive decline in MCI and AD populations.

Abstract

Background: To identify classes of cognitively impaired older individuals based on their neuropsychiatric symptoms(NPS) and to investigate the contribution of NPS class to cognitive decline and Alzheimer's disease(AD) risk in mild cognitive impairment(MCI).

Methods: Our study included 1,472 participants(age range 55-91) from the Alzheimer's Disease Neuroimaging Initiative(ADNI) who were diagnosed with MCI or mild AD and had a complete neuropsychiatric Inventory at their baseline visit. We employed latent class analysis to categorize groups by NPS patterns. Linear mixed models of repeated measures(LMMRMs) were used to compare changes in cognitive performance across 5years as a function of NPS class. Subsequently, the Cox proportional hazards model was employed in individuals with MCI to assess whether rate of conversion to AD differed across the NPS groups.

Results: We identified three latent classes of NPS: No NPS (n=799, 51.7%), Apathy/Affective (n=572, 39.8%), Complex (n=108, 8.5%) NPS. In longitudinal analyses we observed interactions between class and time, indicating accelerated cognitive decline in memory and executive function in the Apathy/Affective class. In MCI, hazard ratios for conversion to AD were 1.39(95% CI: 1.10-1.76) for the Apathy/Affective class and 2.03(95% CI: 1.33-3.10) for the Complex class compared to the No NPS group after adjusting for age, sex, education, global cognition, and ApoE4 positivity.

Conclusions: Among cognitively impaired elderly, empirically derived clusters of NPS profiles were associated with cognitive decline and risk of conversion from MCI to AD. Such NPS classes may reflect specific neurobiological mechanisms within or related to AD-related neurodegeneration. Further studies with biological markers are needed to clarify these neurobiological mechanisms.