Longitudinal study of care needs and behavioural changes in people living with dementia using in-home assessment data

Abstract

Background: People living with dementia often experience changes in independence and daily living, affecting their well-being and quality of life. Behavioural changes correlate with cognitive decline, functional impairment, caregiver distress, and care availability.

Methods: We use data from a 3-year prospective observational study of 141 people with dementia at home, using the Bristol Activities of Daily Living Scale, Neuropsychiatric Inventory and cognitive assessments, alongside self-reported and healthcare-related data.

Results: Here we show, psychiatric behavioural symptoms and difficulties in activities of daily living, fluctuate alongside cognitive decline. 677 activities of daily living and 632 psychiatric behaviour questionnaires are available at intervals of 3 months. Clustering shows three severity-based groups. Mild cognitive decline associates with higher caregiver anxiety, while the most severe group interacts more with community services, but less with hospitals.

Conclusions: We characterise behavioural symptoms and difficulties in activities of daily living in dementia, offering clinically relevant insights not commonly considered in current practice. We provide a holistic overview of participants' health during their progression of dementia.

Fig. 1. Prevalence of difficulties in activities of daily living and psychiatric behaviours of the cohort relative to cognitive scores and how they relate to proxy-rater well-being.

a Activities of daily living assessment (BADL), (b) psychiatric behaviour assessment (NPI) and (c) proxy-rater well-being assessment (HADS) according to the participants' cognitive score in the SMMSE (n = 119). a In the activities of daily living assessment, a score of ≥1 indicated difficulty with that component/activity of daily living, in the last 2 weeks. b For the psychiatric behaviour assessment, a score of ≥1 indicated presence of each neuropsychiatric symptom, within the last month. Frequency and severity of the symptom is not illustrated here. c From the proxy-rater well-being assessment, we show the percentage of either normal, borderline or abnormal anxiety and depression of proxy-raters, that correspond to participants with either normal, mild, moderate, moderately severe or severe SMMSE scores. NPI refers to the psychiatric behaviour assessment, BADL refers to the activities of daily living assessment and HADS refers to well-being of proxy-rater. SMMSE Standardised Mini-Mental State Examination, NPI Neuropsychiatric Inventory, BADL Bristol Activities of Daily Living, HADS Hospital Anxiety and Depression Scale.

Fig. 2. In-home assessment scores per cluster group at baseline for activities of daily living, psychiatric behaviours and proxy-rater well-being.

Baseline total activities of daily living and psychiatric behaviour assessment scores per cluster from K-Means clustering model (Calinski-Harabasz Score: 18.52). Each cluster is described on the x-axis, Severe, Moderate and Mild. Model features included, but were not limited to, baseline (a) activities of daily living assessment (n = 87), (b) total behaviour (n = 87) and (c) total behaviour proxy-rater distress scores (n = 87). Kruskal–Wallis test was conducted to compare cluster groups and post-hoc Dunn’s test was conducted to investigate group differences. Significance threshold was set at p value < 0.05; exact p values can be found in Supplementary Table 4. (* <0.05; ** <0.01; *** <0.001). NPI refers to the psychiatric behaviour assessment, NPI distress is proxy-rater distress associated with participants' psychiatric behaviour and BADL refers to the activities of daily living assessment. NPI Neuropsychiatric Inventory, BADL Bristol Activities of Daily Living.

Fig. 3. Proportion of standardised mini-mental state examination cognitive severity scores per cluster group at baseline.

We illustrate the proportion of participants in each cluster that scored either normal, mild, moderate, moderately severe or severe at baseline in their SMMSE (n = 87). Cluster grouping (Severe, Moderate and Mild) was based on previously described method using baseline activities of daily living and psychiatric behaviour assessment scores only. SMMSE Standardised Mini-Mental State Examination.

Fig. 4. Average in-home assessment scores per cluster group across study time points.

Average raw scores per assessment over the assessment time-points, including the exit timepoint, per cluster (n = 87). a Average cognitive SMMSE scores per cluster. b Average cognitive ADAS-Cog scores per cluster. c Average activities of daily living assessment scores per cluster. d Average psychiatric behaviour assessment scores per cluster. The average score per scale for the whole cohort is indicated by the grey dashed line in each subplot. Each of the solid lines represent a cluster group. NPI refers to the psychiatric behaviour assessment and BADL refers to the activities of daily living assessment. SMMSE Standardised Mini-Mental State Examination, ADAS-Cog Alzheimer’s Disease Assessment Scale Cognitive Subscale, BADL Bristol Activities of Daily Living, NPI Neuropsychiatric Inventory.

Fig. 5. Average cumulative sum of healthcare-related data per cluster group.

For each cluster group, Severe, Moderate and Mild, we calculated the average cumulative sum of: (a) comorbidities before baseline, (b) comorbidities after baseline, (c) healthcare events and encounters, and (d) behavioural observations, per cluster (n = 87). Distinct patterns of behaviour, comorbidities and healthcare events are illustrated for individual clusters.