Introduction
Motor neuron disease (MND) [
1] is an umbrella term that subsumes several different neurodegenerative conditions, such as primary lateral sclerosis (PLS), progressive bulbar palsy (PBP), progressive muscular atrophy (PMA), with the most prevalent of those being amyotrophic lateral sclerosis (ALS) [
2,
3]. ALS and other forms of MND are progressive degenerative conditions that affects the central nervous system, impacting individuals’ physical functioning, characterised by muscle wasting of the upper and lower limbs, loss of functional abilities, including speech and movement, respiratory problems, loss of communicative abilities and autonomy [
4]. The life expectancy from diagnosis is between two to five years.
Up to 15% of people with MND (pwMND) can develop frontotemporal dementia (FTD), often typified by progressive cognitive and behavioural impairment, and a lower insight/awareness for these symptoms, in addition to physical deterioration [
5‐
7]. As MND and FTD exist on a disease spectrum [
8‐
10], individuals without dementia can develop milder cognitive and behavioural impairment, occurring in up to 50% of pwMND throughout disease stages [
11‐
13]. The most common cognitive impairments in MND are in executive functioning (particularly verbal fluency), language function and social cognition [
14,
15]. These are most often assessed through neuropsychological batteries but commonly via cognitive screens specific to MND e.g. Edinburgh Cognitive and Behavioural ALS Screen (ECAS) cognitive screen, ALS Cognitive Behavioral Screen (ALS-CBS) cognitive subscale [
16]. Previous research has shown that cognitive impairment, particularly executive dysfunction, can have a negative impact on functional decline, survival for pwMND [
17,
18], and is associated with increased caregiver burden or strain [
19].
The most common behavioural impairments are apathy (as a lack of motivation) and disinhibition [
20,
21], which have been shown to occur across disease stages [
22]. These are assessed through questionnaires, scales or semi structured behavioural interviews that are commonly specific to MND e.g. Beaumont Behavioural Inventory (BBI), ECAS behavioural interview. Previous research has shown that caregiver burden or strain is associated with behavioural impairments in MND, which can be further compounded by physical deterioration [
23]. Behavioural impairment has also been shown to have a negative impact on disease progression [
24] and ultimately on survival of pwMND [
25,
26], particularly apathy [
27].
As a progressive neurodegenerative disease, motor neuron disease can impact both the wellbeing or quality of life (QoL) of pwMND and also their caregivers or family members [
28‐
30]. Caregiver burden or strain has been shown to associate with declining physical functioning of the pwMND, with a further emphasis on the caregivers mental health, particularly depression [
31,
32]. Parallel to this, QoL for pwMND is also effected, due to associated loss of functioning as the condition progresses [
29,
33,
34]. In terms of physical aspects of QoL relating to loss function, respiratory difficulties and communication, there are interventions available for management [
35,
36]. Furthermore, psychological- or mental health-related QoL have been shown to be impacted in MND [
30,
37], particularly in relating to hopelessness and social withdrawal [
38,
39]. While there has been increasing utility of disease-specific, multidomain QoL instruments for MND [
40], many studies still utilise generic, disease-non-specific QoL measures [
41].
As such, while there has been research separately investigating cognitive or behavioural impairment and QoL in MND, there have been no systematic reviews exploring relationships between these factors. Furthermore, while use of QoL, cognitive and behavioural measures [
16,
41] have been explored in MND, it is unclear what QoL measures (generic, disease-specific) have been used in the context of cognitive and/or behavioural functioning.
Review aims
The systematic review primarily aimed to explore if there was an association between QoL and cognitive or behavioural impairment in MND. Further secondary aims were to determine what types of measures are used to explore QoL and cognitive or behavioural functioning in these studies.
The following databases were searched systematically (periods of searches and data retrieval are specified in brackets next to each database): PsychINFO (1806 to 22nd February 2022), MEDLINE (1946 to 22nd February 2022), Embase (1947 to 22nd February 2022) and AMED (1985 to 22nd February 2022) via OVID. The searches were updated on 22nd February 2023 using the same databases. Backwards citation tracking was additionally applied to all full text articles that were reviewed.
Search strategy and eligibility criteria
The searches included free text keyword terms (inclusive of spelling variations) and medical subject heading (MeSH) specific to each database. Search terms were linked by Boolean operators (i.e. AND, OR). The search terms were disease/condition of MND AND cognitive terms OR behaviour terms AND QoL terms. Exact search terms, free text keyword and MeSH search terms that were used, mapped to each database, can be found in Online Resource 1.
Eligibility criteria for inclusion of studies in the systematic review were primary and secondary quantitative research studies, with no restriction on specific quantitative study designs or on language, including individuals with an MND diagnosis, of 18 years or older. Studies required a quantitative measure of QoL or wellbeing, quantitative measure of cognitive and/or behavioural functioning. Exclusion criteria were qualitative studies (with no quantitative element), study population where individuals did not have an MND diagnosis, of 17 years old or younger. Editorials, conference abstracts, book chapters, other systematic reviews or meta-analyses were excluded.
Study selection
Following PRISMA guidelines, Stage 1 involved two reviewers screening titles and abstracts independently based on the inclusion/exclusion eligibility criteria. Articles classified as relevant by both reviewers were included in the next stage of the review and articles deemed not relevant by both reviewers were excluded. Articles where two reviewers disagreed or that were classified ambiguous, progressed to Stage 2 of the review.
In Stage 2, the two reviewers examined full articles independently, applying the inclusion/exclusion eligibility criteria. Articles classified as relevant by both reviewers were included in the review and articles deemed not relevant by both reviewers were excluded. Articles where two reviewers disagreed on inclusion/exclusion or that were classified ambiguous were forwarded to a third reviewer for adjudication.
For the final included articles, all studies were examined for risk of bias/quality and all relevant data were extracted for synthesis. See below the Risk of bias (quality) assessment and Analysis sections for details of relevant quality scoring and data extracted methodology.
Risk of bias (quality) assessment
Quality of individual studies was assessed using the National Heart, Lung, and Blood Institute (NHLBI) Study Quality Assessment Tools [
44], covering a broad range of study designs and is a well-established, frequently updated tool in clinical literature [
45]. These tools were used to determine the quality of various quantitative studies. Each item can be answered "Yes", "No", "Cannot Determine/Not Reported" and “Not Applicable”. If "No" is selected, the reviewers consider potential bias for this item. If "cannot determine" or "not reported" is selected, these may represent potential flaws in the study design, reporting and/or implementation. If “Not Applicable” is selected, this was accounted for in overall judgement of the quality/risk of bias. The quality review was completed by two reviewers independently, with studies classified as “good”, “fair” or “poor” based on consensus. The tool was adapted to assess cognitive and behavioural impairment assessment separately, with items 6 to 10 relating to exposure of interest (cognitive and behavioural impairment) being additionally subdivided.
Analysis
Narrative synthesis of quantitative data from included studies was conducted to explore if there are any patterns of associations or differences in QoL in terms of cognitive and/or behavioural impairment.
Author (year), country, study design, sample size, age, sex, measures (QoL, cognitive and/or behavioural impairment) and results were extracted from included studies. In terms of results, examples of quantitative data that were synthesised were correlations, regressions and group differences. Specifically, if available, correlations (Pearson’s r, Spearman’s rho or other relevant metrics, including p values), regression results (beta coefficients or other relevant metrics, including p values) or group differences (t, F or other relevant metrics, including p values) for cognitive or behavioural measures relative QoL measures were extracted and synthesised. Furthermore, the overall robustness of the data synthesis was also analysed and considered in relation to quality of included studies (i.e. risk of bias).
Discussion
This systematic review shows that there may be a relationship between self-perceived QoL and behavioural impairment, in the context of variable QoL, cognitive and behavioural measurements that were used. About a third of studies that measured behaviour found that there was an associated effect between this impairment and lower QoL.
These findings build on previous research suggesting that behavioural impairment is associate with strain, burden and distress for caregivers or family members [
31]. The impact of behavioural impairment may extend towards the pwMND themselves, through overall behavioural impairment negatively affecting QoL. Further to this, there is indication that certain behavioural domains such as apathy, irritability and behavioural-regulation (akin to disinhibition) may relate to lower QoL. Apathy as the most common behavioural impairment in MND [
20] may result in withdrawal from everyday life, family gatherings and social events. Caga et al. [
48] found that pwMND with apathy had lower QoL relating to community connectiveness and satisfaction in life, particularly relative to emotional elements of apathy. Emotional apathy (as emotional neutrality/indifference towards self, others and surroundings), has been shown to be more characteristic and prevalent in FTD [
60‐
62]. Emotional apathy may have overlap with diminished sympathy and empathy features that are observed in FTD and milder behavioural impairments in MND. These types of behavioural impairments may represent observable changes in the pwMND through how they interact with people in the environment and how those people reciprocally interact with them. This could result in a negative dynamic between the pwMND and the others around them, reverberating as a negative impact on QoL. As such there might be a complex interplay between self-perceived QoL, interaction and caregiver burden or strain in relation to behavioural impairment, which should be an avenue for future research. Notably studies that did detect QoL-behavioural impairment links in this systematic review predominantly utilised disease-specific measures, for example the ECAS and the ALS-CBS, whereas those that did not used disease-non-specific measures. More recently systematic reviews of measurement have propagated the consistent use of disease-specific measures of cognitive functioning and behaviour in both exploratory research but also in clinical trial research [
16,
63,
64]. This in combination with our findings emphasises further the importance of disease-specific measurement of behavioural impairment in MND.
For cognitive impairment, the studies in this systematic review overall that did not find a robust QoL association, utilised either comprehensive cognitive screens or neuropsychological assessment/tasks that were disease-specific. Associations between cognitive impairment and lower QoL were only found in one third of studies, most of which were to do with overall cognitive functioning and were assessed by disease-non-specific measures such as self-rated questionnaire measures or generic cognitive screens (e.g. SIML, ACE-R, FAB). While self-rated cognitive questionnaires give insight to cognitive functioning, they may also capture subjective perception of cognitive impairment and worry expressed by the individuals completing the measures. This may therefore induce bias of over-judgement relating to cognitive difficulties and confound any cognitive functioning-QoL relationships. The current criteria for detecting cognitive impairment recommends objective, quantifiable cognitive screening or neuropsychological assessments/tasks [
9,
10]. The findings seem to suggest that objectively assessed frontal executive or theory of mind-related impairments associates with worse QoL [
58,
59]. These types of cognitive impairments have been observed to be common in MND [
65]. Further, these cognitive domains are important for higher order processing and social interaction or understanding, as such these may impact people’s confidence in performing tasks independently and interacting with the outside world, which may result in a knock-on effect on QoL.
In terms of QoL assessment, the measures used across studies study were predominantly generic and disease-non-specific. While some of the most common measures used were the MQoL and SEIQoL-DW, the total score of the latter may not be representative of the disease-specific experience of pwMND and have variable intercorrelations with other QoL measures in MND [
66,
67]. These types of items may not quantifiably capture what is important for pwMND in the context of their condition. Disease-specific measures such as the ALSAQ-40 or the ALS-Specific QoL (ALSsQoL) instrument explore domains that are relevant to MND, for example for communication, intimacy, eating, interaction with the environment and mental health. Moreover, in studies where QoL was explored multidimensionally, this yielded meaningful findings about the experiences of pwMND in relation to QoL associated with mental health and social connectedness. As such, this is supportive of disease-specific measures being used for assessment of QoL in MND.
In assessing included studies, there was apparent reporting and methodological limitations, either in relation to sample size, eligibility, study population or statistical reporting. While different types of MND share clinicopathological similarities, they can have varying rates of progressions and regions effected. As such this could variably impact self-perceived QoL dependent on the type of MND, and may have a complex interaction with cognitive or behavioural profiles. Of further note is that more objective measures of cognitive functioning could have been used. All these factors impact the quality of included studies, increase the chance of bias resulting in cautious interpretation of the findings of this systematic review.
Recommendations
The core recommendation for future research centre around study population, more standardised reporting of methods and statistics, consistent design and identification of cognitive and behavioural impairment (both longitudinally and cross-sectionally) as well as more targeted application of MND-specific measures, across QoL, cognition and behaviour. In particular, future research should use disease specific, multidimensional cognitive and behavioural screen (i.e. ECAS) to determine impairments. Ideally, individuals should undergo assessment using a comprehensive neuropsychological battery exploring-specific cognitive domains and thorough behavioural as well as neuropsychiatric assessment to reliably determine impairments for pwMND. Notably few studies explicitly applied the criteria for classification of cognitive and behavioural impairment in MND [
9,
10], which may be further confound relative to consistency of findings and would be an avenue for future research.
Strengths and limitations
A strength of this systematic review is that it was able to extract and meaningfully interpret findings from a relatively heterogeneous collection of included studies that utilised variable measurements methods of cognitive or behavioural impairment and QoL. However, this systematic review does have its own limitations. Due to the different study designs (longitudinal, cross-sectional) in combination with the differential reporting of and types of statistical analysis used in the included studies, it was not possible to perform a more in-depth synthesis of results (i.e. meta-analysis). As such, as the field develops further, it might provide opportunity for more in-depth analysis of the relationship between QoL-cognitive and behavioural impairment in MND.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.