Neuropsychiatric syndromes in Alzheimer’s disease: Systematic review and meta‐analysis of the factor structure of the neuropsychiatric inventory

Abstract

AbstractBackgroundNeuropsychiatric symptoms (NPS) are a prevalent feature in patients with Alzheimer’s disease (AD). Studies have consistently observed that certain NPS have a tendency to co‐occur with others. One major focus of NPS research has been to understand the factor structure of the Neuropsychiatric Inventory (NPI) to identify neuropsychiatric syndromes in AD. Whilst there is a large literature of studies, there exists heterogeneity in the results derived.MethodWe sought to increase our certainty about how neuropsychiatric syndromes in AD may be defined by conducting a systematic review and meta‐analysis of the literature from 1994‐2020 regarding the factor structure of the NPI. We used meta‐analytic structural equation modelling (MASEM) to pool effect sizes (correlation coefficients between NPI items) under a random effects model and fitted measurement models from published studies to identify a best‐fitting factor structure.ResultTwenty‐four studies met our eligibility criteria however effect sizes were only available for seven studies. For the NPI‐10 (7 studies, n = 5185), a five‐factor structure was found to have good fit to the data across a majority of model fit indices (CFI = 0.966; TLI = 0.940; SRMR 0.031; RMSEA 0.026 [95% CI 0.021, 0.030]). For the NPI‐12 (4 studies, n = 2397), we were unable to identify a factor structure that displayed good model fit.ConclusionOur meta‐analysis indicated that a five‐factor structure to the NPI‐10 – agitation, irritability, and aberrant motor behaviour (factor 1); delusions and hallucinations (factor 2); depression and anxiety (factor 3); euphoria and disinhibition (factor 4); and apathy (factor 5) – was the best‐fitting measurement model.