Multi‐level Factor Analysis of Daily Data to Identify Subsyndromes of Behavioral and Psychological Symptoms

Abstract

AbstractBackgroundBehavioral and psychological symptoms in dementia (BPSD) are dynamic phenomena with a high amount of intraindividual variability. Many studies have used single‐level factor analysis approaches to identify subsyndromes of BPSD based upon meaningful groupings of co‐occurring symptoms yet, proposed subsyndromes have not been successfully replicated and lack stability over time. This may be due to measurement issues which do not account for dependencies in the data and create temporal misalignment. We applied an innovative multi‐level framework to overcome limitations of prior research to identify subsyndromes of BPSD.MethodsThis study used an intensive longitudinal design in which co‐residing family caregivers to persons living with dementia in the community were recruited to proxy‐report on their care recipient’s daily symptom experiences of 23 different BPSD for 8 consecutive days. A multi‐level exploratory/confirmatory factory analysis was utilized to account for nested data and separate within‐person variances from between‐level factor estimates.ResultsFamily caregivers (N = 70) were on average African American (54%) females (69%) most often caring for a parent with dementia (62%) average 74 years old requiring assistance with 4.5/7 activities of daily living. Data were provided on 79% of sampled days (n = 443 diaries). Exploratory factor analysis identified a 4‐between 3‐within factor structure based on fit statistics and clinical interpretability. The between‐level subsyndromes include wandering/aggression/care resistance (F1), affective/apathy/psychosis/vocalizations (F2), cognitive/ lability/ euphoria/ impulsivity/ compulsions (F3), and hallucinations/uncooperative (F4). Reliability estimates were high: 0.88 (F1), 0.91 (F2), 0.83 (F3), and 0.79 (F4).ConclusionThe between‐level factor structure identified through the multilevel factor analysis framework provide a more valid and reliable estimation of clinical subsyndromes within a population by separating out the variance due to within‐person daily fluctuations. At the between‐level, hallucinations was the only BPSD to cross‐load, which may indicate they are a byproduct of unmanaged depression in the affective/psychosis factor and their own unique subsyndrome. Support for the between‐level subsyndromes can be found in studies on shared neurobiological mechanisms of BPSD. The within‐level factors are similar to structures found in a recent systematic review. We argue these within‐level factors are more representative of symptom clusters, as they are driven by fluctuating contextual factors.