Is Carer Management Style Associated with Longitudinal Functional Decline in Dementia?

INTRODUCTION

Functional disability, which refers to the inability to perform activities of daily living (ADL), is one of the key criteria for a diagnosis of dementia [1]. People living with dementia (PLwD) can over-time experience a decline in their ADL performance with progression of the syndrome, usually starting with instrumental ADLs (e.g., managing finances, driving, preparing meals), and then moving to more basic ADLs (e.g., maintaining personal hygiene, feeding themselves, going to the toilet) later on [2, 3]. Indeed, functional disability is associated with various negative consequences for the PLwD including a loss of autonomy, decreased quality of life, increased risks of being admitted to a care home and hospital, as well as death [4]. Functional disability in PLwD is also associated with increased carer burden and distress [5–7], as well as increased financial costs for health care settings [8].

Considering the impact that functional disability has for the PLwD, their families, and the community at large, studies have attempted to identify factors associated with this disability. The identified factors can broadly be categorized into intrinsic factors— seen in the PLwD and associated with their dementia pathology, and extrinsic factors— seen in the carers of the PLwD and associated with their care recipient’s dementia pathology. The main intrinsic factors identified include cognitive impairment [9], as well as neuropsychiatric symptoms such as apathy and depression [2, 10]. Although not strictly an intrinsic factor per se, the presence of medical comorbidities, which is highly prevalent among PLwD [11], has also been suggested as a factor associated with functional disability in this population [12, 13], although evidence from some studies suggest no significant association between the two [14, 15]. Meanwhile, the extrinsic factors that have been recognized include carer burden [16], depression [17], and management style (i.e., using a criticism style to manage dementia-related issues) [18]. Lastly, findings from a recent study suggested that it may be the combination of both the above intrinsic and extrinsic factors that may best explain functional disability in PLwD. Specifically, this study reported that the combination of having worse cognition, increased apathy, and carers that use a criticism style was significantly associated with greater functional disability in the PLwD [18].

Despite studies having identified both intrinsic and extrinsic factors as being associated with functional disability in dementia, very little is known about whether the combination of both types of factors are associated with longitudinal decline in ADL performance. Considering that PLwD can exhibit variation in their longitudinal patterns of functional decline [19, 20], addressing this research gap is of high importance as results would offer insight on how clinical interventions and services should be designed to help maintain function for longer. Moreover, results could help identify individuals at higher risk of functional decline over a period of time, who can subsequently be targeted and prioritized alongside their carers early on.

To this end, the main aim of this study is to build on a previous study by our team [18] and investigate whether the identified combination of intrinsic and extrinsic factors that were significantly associated with functional disability in a cohort of PLwD at baseline are also associated with longitudinal decline in their ADL performance over a 12-month period. Additionally, a secondary aim is to explore if medical comorbidities will appear to be significantly associated with functional disability at baseline, in addition to the combination of intrinsic/extrinsic factors that were associated with functional disability in our previous study [18]. We hypothesize that individuals with relatively worse cognition, increased apathy, and with carers who use more of a criticism style at baseline will be the ones who exhibit a greater longitudinal decline in their ADL performance, as these factors have been shown to be associated with increased functional disability at baseline by our previous study [18]. We also hypothesize that at baseline, in addition to the factors of cognition, apathy, and carer management style, having higher medical comorbidities will also be associated with increased functional disability, in line with previous studies [12, 13].

METHODS

This study was conducted as a part of the TASKed project, funded by the Alzheimer’s Society, sponsored by the University of East Anglia, and led by the senior author.

RESULTS

Demographics

A total of 183 dyads of PLwD and their family carers in the UK were recruited from September 2016–January 2020. A total of 141 PLwD-carer dyads passed the above-mentioned inclusion/exclusion criteria (see Methods, Participants) and were used for the present study. Of this cohort, 131 PLwD-carer dyads overlap with the sample used in our previous study [18] (which had a slightly different inclusion/exclusion criteria), for which we will be directly comparing some of our results to.

Data were collected from the 141 dyads at baseline, of which 60 were followed-up after a mean duration of 12.20 months (attrition rate = 57.4%). The relatively smaller sample size at follow-up was largely due to two factors: the COVID-19 pandemic, which prevented the study team from holding the second research visit for 39 dyads visited at baseline, as well as participant discontinuation from the study due to either severe decline or withdrawing voluntarily. Of the 60 dyads visited at follow-up, 7 had incomplete data for the variables of interest (due to their carers declining to complete the relevant measures) and were subsequently excluded from the analysis, resulting in a final sample of n = 53. A flowchart depicting the participant sample sizes included at each stage of the study is shown in Fig. 1.

OPEN IN VIEWER

Fig. 1

Flowchart illustrating participant sample sizes included at each stage of the study.

Participant demographics, both at baseline and follow-up, are shown in Table 1. Carer demographics, both at baseline and follow-up, are shown in Table 2.

OPEN IN VIEWER

Table 1

Participant Demographics

Variable	At Baseline	At Follow-Up
	Assessment	Assessment
	(n = 141)	(n = 53)
Age (Years –Mean; SD)	79.26 (7.41)	79.54 (7.40)
Gender (% M, F)	63.1, 36.9	52.8, 47.2
Education (Years –Mean; SD)	12.17 (2.66)	12.43 (2.31)
Diagnosis (AD, VaD, mixed dementia)	105, 25, 11	41, 10, 2
Illness Duration (Years From Earliest Symptoms Date to Baseline/Follow-Up Date –Mean; SD)	4.29 (3.08)	5.36 (2.99)
(0.41–20.58)	(2.00–16.58)
(Range in Years)

DISCUSSION

This was one of the first studies to investigate whether the combination of intrinsic and extrinsic factors at baseline were associated with longitudinal changes in functional disability in PLwD. Additionally, this study examined whether medical comorbidities, when combined with the intrinsic and extrinsic factors from our previous study [18] could explain the variations seen in functional disability for the PLwD at baseline. In line with our hypothesis, we found that at baseline, having a higher number of medical comorbidities was associated with increased functional disability for the PLwD. We also found that neither the level of general cognition, apathy, carer criticism/encouragement/active management, or medical comorbidities for the PLwD at baseline were significantly associated with longitudinal decline in their ADL performance over a 12-month period. This was not in line with our hypothesis that the combination of these intrinsic and extrinsic factors at baseline would be associated with longitudinal declines in ADL performance. Longitudinal declines in ADL performance were instead associated only with decline in general cognition for the PLwD.

Our results from the baseline multiple regression model suggested higher medical comorbidities as a factor that is significantly associated with functional disability in dementia, which is in support of results from other studies [13, 29]. The PLwD in our sample exhibited a range of medical comorbidities including depression, diabetes, peripheral vascular disease, and cerebrovascular accidents. The exact mechanisms underlying how medical comorbidities impact functional disability in dementia depends upon the specific type of comorbidity, however studies increasingly suggest that the presence of medical comorbidities can negatively impact one’s cognitive and physical function, which can in turn affect their functional abilities. For instance, the presence of diabetes (a disease closely associated with dementia) has been shown to be associated with lesions to white matter tissue in the brain, which can over time contribute to cognitive decline in affected individuals [30]. Indeed, cognitive decline often results in individuals being unable to engage in self-management and wellness activities (i.e., preparing healthy meals, exercise, maintain hygiene, etc.), which can subsequently have a negative impact on their functional abilities [31]. From a physiological standpoint, medical comorbidities can give rise to a number of physical issues (e.g., breathlessness, mobility issues, low energy levels, low handgrip strength, etc.), all of which can affect the PLwD’s ability to perform various functional activities [32–34]. Alternatively, having multiple diseases (especially those commonly occurring together) can often result in the diseases interacting with one another and attenuating potential compensatory pathways, resulting in greater functional disability [35, 36]. In addition to the diseases themselves, co-medication for multiple diseases can interact with one another and result in adverse drug reactions, leading to greater functional disability through means of impacting cognitive and physical function [37]. Overall, our results suggest that burden of medical comorbidities, in combination with cognitive ability, level of apathy, and carer management style of the PLwD can explain the variation in functional disability levels seen in this population at baseline.

Our results from the longitudinal multiple regression model suggests that although individuals’ level of global cognition, apathy, burden of medical comorbidities, as well as their carers’ more frequent use of criticism and encouragement management styles were associated with the level of their functional disability in a cross-sectional approach, the same combination of these variables were not associated with longitudinal decline in their ADL performance over a 12-month period. To the best of our knowledge, this is the first study to investigate if the combination of both intrinsic and extrinsic factors for functional disability are associated with longitudinal changes in PLwD’s ADL performance. Previous studies have either only identified factors associated with functional disability at a single time point or investigated intrinsic/extrinsic factors separately in their association with functional decline in PLwD [18, 38–42]. Our results are not in support of findings from previous studies, which have reported carers’ more frequent use of problem-focused coping strategies and PLwD’s higher neuropsychiatric symptoms (which includes apathy) at baseline respectively as being associated with greater longitudinal decline in ADL performance of PLwD [41, 42].

These discrepancies in our results with previous studies could be due to differences in study methodology, with the previous studies having a relatively larger sample size and assessing longitudinal change over a greater time-period (i.e., 4–6 years). More importantly, however, is that the previous studies differed from the current study in the instruments used for the outcome measure (i.e., functional disability), using the Clinical Dementia Rating scale and the Rapid Disability Rating Scale instead of the DAD used here. Although there is no current gold-standard assessment tool for functional disability in dementia [43], future studies should explore which measurement scale has the most ecological validity with regards to accurately capturing changes in functional disability overtime for PLwD. Furthermore, the previous studies used a) the Ways of Coping Checklist, which measures carers’ use of 8 different coping behaviors to look after their own wellbeing, which is different to the DMSS used here, which measures carers’ use of three management styles to practically deal with the PLwD and b) the Neuropsychiatric Inventory to measure behavioral symptom prevalence, which encompasses a range of symptoms compared to our sole focus of apathy in the CBI-R. This considered, we speculate that longitudinal functional decline in PLwD may partly be explained when looking at specific coping strategies of the carer which are not captured by the DMSS, as well as when considering the level of neuropsychiatric symptoms as a whole in the PLwD, as opposed to solely apathy levels. An interesting direction for future studies would be to explore if the development of models which involve the combination of carers’ frequency of using problem-focused coping behaviors and the PLwD’s level of neuropsychiatric symptoms at baseline can accurately predict those with greater longitudinal functional decline in the future.

Our post-hoc analysis results showed that the PLwD who had greater deterioration in their cognitive abilities over the 12-month period were the ones who had greater decline in their ADL performance, suggesting cognitive decline as the main underlying factor for functional decline in this population. With the various cognitive domains, but especially executive function and memory, and the interactions between them being suggested to underlie the ability to perform various functional tasks, it comes as no surprise that deficits in cognition contribute to functional decline [3, 9]. The nature of this relationship may be bidirectional, as decline in functional abilities have also been shown to accelerate cognitive decline in older adults [44]. Overall, this result is in support of findings from previous studies and adds to the growing body of work which show a correlation between cognitive and functional decline in dementia [40, 46]. A worthwhile direction for future studies would be to assess whether it is longitudinal decline in the PLwD’s overall cognitive abilities or specific sub-domains (i.e., executive function and memory) that is relatively more associated with longitudinal declines in their ADL performance.

The results of our study have potential clinical implications. Our findings suggesting that it may not be possible to predict longitudinal functional decline for PLwD using their baseline measurements of the various intrinsic and extrinsic factors, underscores the importance for clinicians to periodically conduct assessments of functional ability for PLwD (either in the clinic or through at home visits). Here, individuals exhibiting relatively greater changes in their functional abilities across assessments, compared to those that are more stable, could potentially be identified as being at risk for greater functional decline in the future. These individuals should ideally be targeted for the development of novel interventions that may help alleviate their functional disability. Although interventions such as cognitive training and cognitive stimulation have been suggested by previous studies to have promise in reducing functional disability in PLwD [47], in general the evidence for the effectiveness of these strategies is still low [48]. These approaches have instead been suggested to potentially be more useful for healthy older adults who may be at high risk for dementia, to alleviate their cognitive and subsequently, functional decline [48]. Importantly, our findings suggest that other extrinsic factors remain to be explored, in particular other modifiable extrinsic factors which may influence declines in ADL performance over time. This may subsequently help inform training for carers as well as contribute to the development of future prediction models and clinical interventions.

Despite our findings, there are some limitations to our study that need to be mentioned. Some participants dropped out of the study from baseline to follow up due to severe cognitive decline, resulting in a potential selection bias for the longitudinal analysis with only participants with relatively less severe decline completing the study. Furthermore, for our multiple regression models in the longitudinal analysis, our sample size was n = 53 and we had 6 independent variables. Since it is recommended to have a minimum of 10 observations per independent variable for regression models, it may be likely that the null results we obtained was due to our analysis lacking statistical power [49]. This limitation in sample size also prevented us from exploring and adding interactions between the independent variables as additional explanatory variables into the model. Another limitation is that we did not consider the factors of genotype [50], severity [51] and type of dementia [39], and age [19], which have all been suggested as influencing rates of functional decline in PLwD by previous studies. Furthermore, in the current study our outcome measure of functional disability (i.e., the DAD score) was based on information provided by the carers. With previous studies suggesting that carers can overestimate/underestimate their ratings of care recipients’ functional abilities [52], the extent to which the DAD scores of our PLwD truly reflect their functional abilities is at present unclear. It is more likely that the DAD only captures the ADL performance of the PLwD, and not be indicative of their functional abilities as a whole. To overcome these limitations, future studies should aim to replicate our analyses using larger samples and investigate if at baseline, the combination as well as interaction of factors such as genotype, dementia stage/severity, and age with various extrinsic factors are associated with longitudinal decline in functional abilities for PLwD. Future studies should also aim to assess PLwD’s functional abilities using performance-based assessments [1, 53], which can make more objective assessments of functional status and potentially detect subtle declines in functional abilities over time, which may not be possible when using carer-based assessments. Lastly, although we tried to ensure that the same researchers conducted the study visits at baseline and follow-up, this was not always possible due to changes in team members, time constrictions, and team member availability. Hence although all researchers were trained in the data collection process, inter-rater variability remains a limitation in our study.

In conclusion, we show that although the combination of various intrinsic and extrinsic factors were associated with functional disability for PLwD at baseline, these same factors were not associated with longitudinal decline in ADL performance of these individuals over a 12-month period. Our study provides a platform for future studies to investigate other extrinsic factors that may be longitudinally associated with functional disability, which can generate knowledge that can be used in future to help maintain ADL performance in PLwD.