Article Text

Original research
Disentangling the relationship between social cognition, executive functions and behaviour changes in amyotrophic lateral sclerosis
Free
  1. Francesca Palumbo1,
  2. Barbara Iazzolino1,
  3. Stefano Callegaro1,
  4. Antonio Canosa1,2,
  5. Umberto Manera1,2,
  6. Rosario Vasta1,
  7. Maurizio Grassano1,
  8. Enrico Matteoni1,
  9. Sara Cabras1,
  10. Giorgio Pellegrino1,
  11. Paolina Salamone1,
  12. Laura Peotta1,
  13. Federico Casale1,
  14. Giuseppe Fuda1,
  15. Cristina Moglia1,2,
  16. Adriano Chio1,2,3,
  17. Andrea Calvo1,2
  1. 1 "Rita Levi Montalcini" Department of Neuroscience, ALS Centre, University of Turin, Turin, Italy
  2. 2 SC Neurologia 1U, Azienda Ospedaliero-Universitaria Città della Salute e della Scienza di Torino, Turin, Italy
  3. 3 Institute of Cognitive Science and Technologies, National Research Council, Rome, Italy
  1. Correspondence to Dr Francesca Palumbo; francesca.palumbo{at}unito.it

Abstract

Background Social cognition (SC) deficits are included in the amyotrophic lateral sclerosis-frontotemporal spectrum disorder (ALS-FTDS) revised diagnostic criteria. However, the impact of SC assessment on cognitive classification and the cognitive–behavioural correlates of SC remain unclear. This cross-sectional study aimed to assess the impact of SC assessment on ALS-FTDS categorisation and explore the relationship of SC with executive functions (EF) and behaviour changes in a cohort of ALS patients.

Methods 121 patients and 56 healthy controls from the Turin ALS Centre underwent cognitive/behavioural testing, including the SC subdomains of facial emotion recognition, and cognitive and affective theory of mind (ToM).

Results Patients performed significantly worse than controls in all SC explored domains, and 45% of patients exhibited a deficit in at least one SC test, dissociated from the presence of EF deficits. In 13% of cases, the SC deficit was isolated and subclinical. SC assessment contributed to the attribution of cognitive impairment in 10% of patients. Through a statistical clustering approach, we found that ToM only partially overlaps with EF while behaviour changes are associated with emotional disorders (anxiety and depression).

Conclusions SC is overall independent of EF in ALS, with ToM only partially associated with specific EF measures, and behaviour changes associated with emotional disorders. The influence of SC on cognitive categorisation and the frequent identification of a subclinical SC impairment have implications in a clinical setting, considering the substantial impact of cognitive impairment on disease burden and therapeutic choices.

  • COGNITIVE NEUROPSYCHOLOGY
  • MOTOR NEURON DISEASE
  • COGNITION

Data availability statement

Data are available on reasonable request. All data relevant to the study are included in the article or uploaded as online supplemental information. Data will be available on reasonable request by interested researchers.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • Social cognition (SC) deficits are part of the amyotrophic lateral sclerosis-frontotemporal spectrum disorder (ALS-FTDS) revised diagnostic criteria. However, the impact of SC assessment on the cognitive–behavioural categorisation and the relationship between SC, executive functions (EFs) and behaviour changes in ALS remain unclear.

WHAT THIS STUDY ADDS

  • An SC impairment may be present in up to 45% of ALS patients, and SC is overall independent of EF in ALS, with theory of mind only partially associated with specific EF subcomponents. In 13% of cases, the SC deficit was isolated and subclinical. SC assessment contributed to attributing cognitive impairment in 10% of patients. Behaviour changes (apathy and disinhibition) tend to associate with emotional disorders (anxiety and depression) and are overall independent of EF and SC.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • The influence of SC on cognitive categorisation and the frequent identification of a subclinical SC impairment have clinical implications, considering the significant impact of cognitive and behavioural impairment on disease burden and therapeutic choices. In the research setting, these findings may help to disentangle the cognitive and behavioural manifestation of the ALS-FTD spectrum and encourage exploration of SC impairment as an early marker of cognitive or behavioural dysfunction.

Introduction

In the last decades, social cognition (SC) has been investigated in neurodegenerative disorders, including amyotrophic lateral sclerosis (ALS).1 The spectrum of cognitive and behavioural dysfunction in ALS patients is highly heterogeneous, and impairment of facial emotion recognition (FER) and theory of mind (ToM) has been reported in ALS patients since the early stages of the disease.2 Consequently, SC deficits have been incorporated into the Revised Diagnostic Criteria for ALS-frontotemporal spectrum disorder (ALS-FTDS).3 However, the impact of SC assessment on the ALS-FTDS categorisation remains largely unknown, and the cognitive and behavioural determinants of SC, particularly its relationship with executive functions (EFs) and behaviour changes (BCs), are subject to ongoing debate.4 Furthermore, there are inconsistent results regarding the accuracy of the Edinburgh Cognitive and Behavioural ALS Screen (ECAS)5 in detecting SC and EF impairments in ALS, particularly concerning its sensitivity.6 However, there is general agreement on its high specificity.7

This study aims to explore the relationship between SC abilities, EF and BC in ALS patients, evaluate the accuracy of ECAS in detecting SC deficits and assess the impact of SC evaluation on patients’ cognitive classification according to ALS-FTDS Revised Diagnostic Criteria.3

Methods

Case ascertainment

We enrolled 121 consecutive patients at the Turin ALS Centre between February 2019 and May 2023, meeting inclusion and exclusion criteria: diagnosis of probable, probable laboratory-supported or definite ALS,8 no neurological comorbidities, no medications potentially affecting cognitive performance (GABAergic, cholinergic, adrenergic or serotoninergic systems), no major depression (The Diagnostic and Statistical Manual of Mental Disorders-Fifth Edition criteria)9 and no history of addiction. Demographic and clinical data were collected, including age, sex, education, site of onset, ALSFRS-R, forced vital capacity, onset delay (time between symptom onset and neuropsychological assessment), diagnostic delay (time between symptom onset and diagnosis), use of ventilatory support (invasive or non-invasive ventilation) and use of nutritional support. 56 sex-matched and age-matched healthy controls were recruited, also meeting the same exclusion criteria, among patients’ caregivers and non-health professional volunteers in the hospital.

Neuropsychological evaluation and cognitive–behavioural categorisation

All patients and controls underwent an extensive neuropsychological battery assessing EFs, memory, visuospatial function, SC and language.3 Patients underwent cognitive evaluation as part of the diagnostic workup. EFs were assessed through Letter Fluency test (FAS), Category Fluency test (CAT),10 Trail Making Test B-A (TMT B-A),11 Frontal Assessment Battery (FAB)12 and Nelson’s modified version of Wisconsin Card Sorting Test (WCST).13 FER was assessed with the Ekman 60-Faces test (EK-60F),14 Affective ToM with the Reading the Mind in the Eyes Test-36 faces full version (RMET-36)15 and the Story-Based Empathy Task-Emotion Attribution (SET-EA).16 Cognitive ToM was evaluated through Story-Based Empathy Task-Intention Attribution (SET-IA). Details of the complete neuropsychological battery are provided in online supplemental material 1 online supplemental file 1, 17. Neurobehavioural dysfunction was assessed through direct observation by the neuropsychologist, patient history18 and the Frontal Systems Behaviour Scale (FrSBe). We used the Family version of FrSBe, assessed by a close relative, to account for potential loss of insight in patients. Higher FrSBe scores indicate more severe behavioural impairment, categorised as normal (≤59), borderline (60–64) and pathological (≥65), with an increase of ≥10 points from the premorbid condition.19 Emotional disorders (anxiety and depression) were assessed using the Hospital Anxiety and Depression Scale (HADS). We addressed item 8 (I feel slow down) with the patient to exclude the potential impact of the physical disability on the response. Scores of ≥8 on the HADS anxiety and depression scales were considered thresholds indicating symptoms.20 Raw test scores, including SC tests, were age-adjusted and education-adjusted according to the Italian normative. Neuropsychological test deficit, including SC tests, was defined as a score <2 SDs from the Italian normative. For EK-60F, deficits in individual emotion recognition (happiness, sadness, fear, anger, disgust and surprise) were identified using cut-off scores from Italian normative data.21 A global FER deficit was defined for an EK-60F total score below the cut-off while a selective FER deficit was defined for a score below the cut-off for at least one emotion recognition. Affective ToM deficit was defined as a deficit in SET-EA or RMET-36,15 16 and cognitive ToM deficit was defined as a deficit in SET-IA.16 An SC deficit was defined as a deficit in at least one of the following: selective FER, global FER, affective ToM or cognitive ToM. An EF deficit was defined as a deficit in at least one EF test. Mild SC or EF impairment indicates borderline scores (between 1 SD and 2 SDs compared with the mean of Italian normative) in at least one SC or EF test. BC was defined as the presence of apathy, disinhibition or both. Mild BC denotes borderline scores in one of FrSBe ‘after’ subscores. Emotional disorders were defined as the presence of anxiety, depression or both. Deficit in ECAS was determined using single-task cut-off values from Italian normative.22

Statistical methods

Descriptive statistics (mean±SD, range for dimensional data and proportions for dichotomous data) characterised the sample. Shapiro-Wilk test assessed distribution normality. In the case of non-normal distribution, non-parametric tests (Mann-Whitney U test with Bonferroni correction for multiple comparisons) were used to compare means. All p values are two tailed, and significance was set at p<0.003 (0.05/12). Data were analysed using the Statistical Package for the Social Sciences (SPSS V.25.0 for Windows).

Multiple linear regressions and random forest analysis

To explore the SC-EF relationship, we conducted a multiple linear regression (MLR) with Bonferroni correction for multiple comparisons in the entire population. Independent variables included sex, age, education, onset site, ALS-related pathogenic variants (SOD1, TARDBP43, FUS and c9orf72), HADS-Anxiety, HADS-Depression and corrected scores from EF tests (FAS, CAT, FAB, TMT B-A and WCST); dependent variables were corrected scores from SC tests (EK-60F, RMET-36, SET-GS, SET-IA and SET-EA). To explore the impact of cognitive profile on the SC-EF relationship, we conducted a second MLR in the group of ALS-CN patients and a third MLR in the group of ALSci-bi-cbi-FTD patients, using the same variables. Another MLR assessed the SC–BC relationship with independent variables, including sex, age, education, onset site, the presence of pathogenic variants, SC tests and dependent variables FrSBe ‘after’ scores (apathy, disinhibition and total score). To evaluate if performances in EF tests could predict a deficit in SC tests, we used random forest (RF) Analysis, a machine learning method that involves creating a random ensemble of decision trees to make accurate predictions for classification tasks.23 We used the same independent variables of MLR analysis. Due to the limited data sample, we used a custom binary output variable for the SC test scores. We assigned 0 if at least one score was <2 DS from the mean and one otherwise. This resulted in assigning 0–26 patients and 1–95 patients. We used stratified k-fold cross-validation, where k represents the number of groups, we divided the dataset into for training the model (in our case, k=3).24 25 Precision, which measures the percentage of items predicted as negative that are actually negative, and recall, which captures the percentage of actual negative items correctly predicted as negative, were assessed using an independent cohort.

Cluster analysis

To evaluate how closely EF, SC, behavioural symptoms (BCs) and emotional disorders were related, we performed a cluster analysis using as input values corrected scores of FAS, CAT, FAB, TMT B-A, WCST, RMET-36, EK-60F, SET-IA, SET-EA, SET-GS, FrSBe ‘after’ (apathy, disinhibition, dysexecutive, total score), HADS-Anxiety and HADS-Depression. To ensure uniformity, we standardised the data using a min-max scaler, rescaling in the range (0–1).26 We then performed a principal component analysis (PCA) using a k-means cluster algorithm. We used Mann-Whitney U to asses p values for pairwise cluster mean comparison, with Bonferroni correction for multiple comparisons. Significance was set at p<0.003 (0.05/16). The data for RF and cluster analysis were analysed using Python V3.6.

ECAS assessment of EFs and SC

We assessed the accuracy of ECAS in identifying deficits in EF and SC, comparing it against neuropsychological evaluations for EF and SC. Additionally, since verbal fluencies are considered a measure of EF, we also evaluated the accuracy of ECAS in identifying deficits in verbal fluency. Sensitivity was calculated as true positives/(true positives+false negatives), and specificity was calculated as true negatives/(true negatives+false positives). We investigated the relationship between ECAS EF and Fluency with the ECAS SC subscores using MLR analysis. We considered as independent variables sex, age, education, onset site, presence of pathogenic variants, Fluency ‘S’ and ‘T’ scores, along with and EF subscores (backward/reverse digit span, alternation, sentence completion). The dependent variable was the ECAS SC subscore.

Impact of SC assessment on strong classification

According to revised ALS-FTDS diagnostic criteria, patients were categorised into cognitively normal ALS patients (ALS-CN), ALS patients with cognitive impairment (ALSci), behavioural impairment (ALSbi), cognitive and behavioural impairment (ALScbi) and FTD (ALS-FTD). To assess the impact of SC evaluation on cognitive categorisation, patients were initially classified without SC assessment, then reclassified taking SC into account. Cognitive impairment leading to a categorial shift (from CN to ALSci or from ALSbi to ALScbi) was defined as a deficit in global FER, affective ToM or cognitive ToM, coexisting with a deficit in a non-overlapping EF measure.3

Results

In the 2019–2022 period, 206 patients with ALS attending the Turin ALS Centre were tested for cognitive and behavioural functions. Exclusions comprised 56 not meeting recruitment criteria, patients who sought a second opinion with confirmed diagnosis without further investigations and 29 lacking FrSBe scores due to caregiver unavailability during the neuropsychological assessments. The study then includes 121 ALS patients (figure 1). Demographic and clinical features are detailed in table 1.

Figure 1

Flow chart reporting the patients’ selection process. ALS, amyotrophic lateral sclerosis; CG, caregiver; FrSBe, Frontal System Behaviour Scale; SC, social cognition.

Table 1

Demographic and clinical features of patients and controls

Prevalence of SC, EF deficits and behaviour changes

55 (45.4%) patients exhibited an SC deficit. Specifically, 10 (8.2%) showed a deficit in global FER, 29 (23.9%) in selective FER, 11 (9.0%) in affective ToM and 12 (9.9%) in cognitive ToM. Fear recognition was the most impaired (23, 19.0 %), followed by sadness (11, 9.0%), anger (10, 8.2%), surprise (7, 5.7%), disgust (6, 4.9%) and happiness (4, 3.3%). BCs were observed in 18 (14.8%) patients. Among them, 12 (9.9%) exhibited only apathy, 5 (4.1%) showed both apathy and disinhibition, and 1 (0.8%) exhibited only disinhibition. Among healthy controls, 2 (3.6%) showed a selective FER deficit in fear; no other cognitive impairment was detected (online supplemental material 2).

Patterns of SC deficits related to EF deficits and behaviour changes

Nine patients (7.4%) showed isolated selective FER deficits while 3 (2.4%) had an isolated global FER deficit. Three (1.6%) displayed an isolated ToM deficit, 15 (12.3%) had an isolated EF deficit and 7 (5.7%) had isolated BC. Figure 2 illustrates the pattern of SC impairment related to EF and BC. Patients with isolated FER or ToM impairment were predominantly male (13 males, 4 females), with spinal onset (16 spinal onset, 1 bulbar onset), normal cognitive profile (13 CN, 1 ALSbi and 3 ALSci), mean age 62.53 (SD 8.71) years and mean education 11.65 (SD 4.12) years. Among them, two (10.5%) exhibited BC (one apathy, one apathy with disinhibition) and one carried a TARDBP43 pathogenic mutation. When comparing SC-impaired patients to those without impairment, we observed poorer performance in TMT B-A and ECAS (p<0.001), with fewer years of education in the former group (p<0.001). In the comparison of cognitively normal (CN) patients with and without SC impairment, no significant differences were found (table 2). Among the 66 patients without SC deficit, 9 (13.6%) showed BC (7 apathy and 2 apathy with disinhibition). In the group of 55 patients with SC deficit, 9 (18.0%) also showed BC (5 apathy, 1 disinhibition and 3 apathy with disinhibition) (χ2=0.67).

Figure 2

Figure showing the distribution of FER, ToM, EF deficits and behaviour changes in the studied population. We included in FER group patients with a selective or global deficit in FER, in the ToM group patients with a deficit in affective ToM, cognitive ToM or both. We included in EF group patients with deficit in at least one EF test, excluded SC tests. We included in BC group patients with apathy, disinhibition or both. ALS, amyotrophic lateral sclerosis; BC, behaviour changes; EF, executive functions; FER, Facial Emotion Recognition; ToM, theory of mind.

Table 2

Comparison between patients with versus without social cognition deficit on clinical features and cognitive–behavioural tests

MLR and RF analysis results

In MLR analysis, SC tests showed no significant relationship with EF tests, even after adjusting for key variables influencing cognitive performances (table 3). This result was confirmed when stratifying patients by cognitive status (online supplemental material 3). Furthermore, in MLR analysis, no significant relationship between SC and BC was observed, even after adjusting for key variables with potential impact on BC (online supplemental material 4).

Table 3

Social cognition and executive functions relationship: multiple linear regression analysis results

In RF analysis for class 0 (patients with SC deficit, defined as a score <2 DS from the mean in at least one SC test), the model exhibited a recall of 27% and precision of 78% (online supplemental material 5). The low recall value, indicating the percentage of actual negative items correctly predicted by the model, supports the absence of an association between EF and SC deficit.

Cluster analysis results

PCA analysis reduced input variables from 16 to 10, retaining 92% of the information. The analysis used k=3 clusters. The first cluster (n=17) comprised patients with emotional disorders (anxiety and depression) and mild apathy but with normal EF and SC. The second cluster (n=72) included patients with normal or mild EF impairment (WCST borderline scores), normal SC and no emotional disorders or BC. The third cluster (n=32) included patients with EF deficit (WCST and FAB deficit) and mild ToM impairment (SET borderline scores) without emotional disorders or BC. The mean ECAS scores were 99.75 (13.82), 105.84 (13.63) and 74.30 (21.98) for the first, second and third cluster, respectively (p=0.122). Figure 3 and online supplemental table 4 provide details on the clusters and significant differences in clinical features, cognitive, behavioural and emotional profiles.

Supplemental material

Figure 3

Figure showing the three clusters. The tests that significantly differentiate the three clusters (p<0.003) are listed next to the connecting arrows. CAT, Category Fluency test Test; EF, executive functions; EK-60F, Ekman 60-Faces test; FAB, Frontal Assessment Battery; FAS, Letter Fluency Test; FER, Facial Emotion Recognition; HADS, Hospital Anxiety and Depression Scale; RMET-36, Reading the Mind in the Eyes Test-36 faces full version; SET-EA, Story Based Empathy Task-Emotion Attribution; SET-IA, Story Based Empathy Task-Intention Attribution; SET-GS, Story-Based Empathy Task-Global Score; TMT B-A, Trail Making Test B-A; ToM, theory of mind; WCST, Wisconsin Card Sorting Test.

ECAS sensitivity and specificity for SC, fluency and EF deficit

ECAS detected EF deficits with 63.63% sensitivity and 94.31% specificity, and fluency deficits with 70.00% sensitivity and 80.18% specificity. It showed 32.2% sensitivity in detecting SC deficits with 96.9% specificity (online supplemental materials 6–8). In the MLR analysis, the ECAS SC subscore showed no significant relationship with the ECAS EF and Fluency subscores (online supplemental material 9).

Impact of SC assessment on strong classification

Excluding the SC assessment, 6 (4.9%) patients were diagnosed as ALS-FTD, 10 (8.2%) as ALScbi, 8 (6.6%) as ALSbi, 26 (21.4%) as ALSci and 71 (58.6%) were classified as cognitively normal (ALS-CN). Taking into account the SC assessment, 6 (8.4%) out of 71 ALS-CN patients were reclassified as ALSci, and 2 (25.0%) out of the 8 ALSbi patients were reclassified as ALScbi. Reclassification was due to deficits in WCST and EK-60F GS in two cases, in SET-EA and WCTS in two cases, in FAB and SET-GS in three cases, in SET-IA and SET-EA in one case. Reclassified patients were predominantly male (six males, two females), with spinal onset (five spinal, three bulbar), mean age 68.63 (SD 9.46) years and mean education 7.88 (SD 2.47) years. One of them carried a pathogenic variant in the c9orf72 gene (figure 4).

Figure 4

Figure showing the changes in ALS-FTDS classification with social cognition assessment. Six patients were reclassified from ALS-CN to ALSci and two patients were reclassified from ALSbi to ALScbi. ALSbi, amyotrophic lateral sclerosis-behavioural impairment; ALScbi, amyotrophic lateral sclerosis-cognitive and behavioural impairment; ALSci,amyotrophic lateral sclerosis-cognitive impairment; ALS-CN, amyotrophic lateral sclerosis-cognitively normal; FTD, frontotemporal dementia.

Discussion

This study aimed to assess the relationship between SC abilities, EF and BC in ALS patients, and estimated, for the first time, the impact of SC assessment on the cognitive–behavioural classification according to the ALS-FTD Consensus Criteria.3 In our cohort, we observed SC deficit in up to 45% of cases. In 13% of cases, this deficit was isolated, with ~10% of patients exhibiting an isolated deficit in FER and ~3% in affective or cognitive ToM assessed by SET; conversely, RMET-36 did not identify patients with isolated affective ToM deficit.

When comparing patients with and without SC deficit in the entire cohort, the former group showed poorer performance in an EF test, consistent with findings from a previous study.2 However, we found no significant differences in EF performances when comparing CN patients with and without SC deficit. In addition, the MLR and the supervised learning approach (RF) revealed no association between EF and SC deficits. Consistently, using a clustering approach, we identified one cluster characterised by normal EF or mild EF impairment dissociated from SC impairment and another characterised by both EF deficit and mild cognitive and affective ToM impairment. Indeed, FER did not exhibit any clustering pattern with EF. These findings indicate that EF impairment could only partially relate to a ToM impairment while does not relate to an FER impairment. Our results are consistent with previous studies conducted in bvFTD27 and ALS28 and further demonstrated that the substantial dissociation between ToM and EF also remained unchanged when differentiating patients according to cognitive status. This supports the hypothesis that a slight impairment in EF is insufficient to determine a ToM deficit, however, ToM performances may be affected when the EF impairment is more severe, in line with the current theoretical framework. Current evidences suggest that, while specific ToM abilities rely on autonomous neurobiological mechanisms, a common neural basis shared with EF, represented by affected prefrontal cortical regions in ALS, underlies general ToM abilities. These general abilities include shifting perspective or manipulating mental representation to adapt behaviour.29

The co-occurrence of EF and SC deficits in the early stages of the disease may have practical implications in the clinical setting. According to the revised ALS-FTDS criteria, the diagnosis of cognitive impairment (ALSci) is established when deficits are present in two non-overlapping measures of EF, including SC.3 Analysing the results from SC tests, approximately 8.5% of CN patients were reclassified as ALSci, and 25% of ALSbi were reclassified as ALScbi. Overall, FER and ToM assessments revealed cognitive impairment in about 10% of patients in our cohort. These results reinforce the findings of a previous study by our group, which demonstrated the increased sensitivity of the revised ALS-FTD Consensus Criteria compared with the previous version in detecting early cognitive signs of the disease.30

Furthermore, this study highlighted the presence of a subgroup of ALS patients with isolated and subclinical SC impairment. Patients with subclinical SC impairment were predominantly male, had a spinal onset and showed a more pronounced impairment in FER than ToM. Only a small percentage (~11%) of them exhibited behavioural impairment. These results hold practical implications, emphasising the importance of SC evaluation, FER included, particularly in patients with these phenotypic traits. In addition, they suggest the possibility of sex differences in SC abilities, with males more likely to exhibit isolated SC impairment since the early stages of the disease. This warrants further investigations.

The importance of evaluating at least two subdomains within SC is further emphasised by the ECAS’s moderate-to-low sensitivity for SC deficits, which is consistent with previous research.6 These results are overall expected, considering that ECAS primarily explores ToM through the Yoni Task while our SC assessment encompassed both FER and ToM abilities. Furthermore, including two distinct tests assessing ToM (RMET-36 and SET) enhanced the sensitivity of our evaluation. Conversely, our results confirmed ECAS’s high specificity in identifying both SC and EF deficits. Further studies exploring additional SC subdomains, such as decision-making and moral judgement, are necessary to ascertain the role SC could have in attributing cognitive impairment independently from EF.

Notably, cluster analysis revealed that emotional disorders (anxiety and depression) and apathy tend to cluster together and are not associated with cognitive dysfunctions, including SC. Our results align with previous studies conducted on depression and cognitive impairment in ALS,31 32 highlighting that anxiety also follows a similar pattern of dissociation with cognitive impairment. The dissociation between apathy and cognitive dysfunctions highlights the high heterogeneity within the ALS-FTD spectrum. It is possible that this evidence partially depends on the difficulty of detecting subtle behavioural impairments or on the limited ecological validity of many SC tests. Social functioning, reliant on interpreting interpersonal cues,33 is often elusive towards a classical psychometric approach. Thus, a contextualised social stimulus may be easier to decipher than an FER or ToM task, with the subject still displaying appropriate behaviour. We cannot exclude that a subtle early impairment in SC abilities may foreshadow the onset of BCs in later disease stages.

Further studies, with more in-depth cluster analysis based on cognitive and behavioural assessment, are needed to comprehend the nature of dysexecutive syndrome in ALS and neurodegenerative disorders.

This study is not without limitations. First, we did not consider some domains of SC, such as decision-making and moral judgement, which, to date, have rarely been explored in ALS. Second, due to the limited data sample, we were unable to conduct a cluster analysis stratified according to cognitive profile.

This study demonstrated that SC and EF are largely independent in ALS patients, with ToM only partially associated with specific EF measures. It also highlighted the significant recurrence of a subclinical SC impairment involving FER and ToM, primarily affecting males with spinal onset. Moreover, it revealed that FER and ToM assessments significantly enhance the detection of early cognitive signs of the disease. Indeed, specific patterns of associations between cognitive impairment and behaviour changes were not identified, with apathy being associated with emotional disorders. Our findings hold practical implications in a clinical setting, considering the substantial impact of cognitive and behavioural impairment on prognosis, disease burden and therapeutic choices. Furthermore, they prompt further exploration of the high heterogeneity within the ALS-FTD spectrum and the potential role of SC impairment as an early marker of cognitive and behavioural dysfunction in ALS patients.

Data availability statement

Data are available on reasonable request. All data relevant to the study are included in the article or uploaded as online supplemental information. Data will be available on reasonable request by interested researchers.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and the study was approved by the Ethics Committee Comitato Etico Azienda Ospedaliero Universitaria Città della Salute e della Scienza (protocol no. 314/2021, 26 July 2021). Participants gave informed consent to participate in the study before taking part.

Acknowledgments

The authors would like to thank all participants involved in this study.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • FP and BI contributed equally.

  • CM, AC and AC contributed equally.

  • Contributors Study concept and design: FP, BI, ACalvo, AChiò and CM. Acquisition, analysis or interpretation of data: BI, FP, StC, LP, ACanosa, UM, RV, MG, ACalvo, AChiò and CM. Drafting of the manuscript: FP, BI, ACalvo, AChiò, CM and StC. Critical revision of the manuscript for important intellectual content: ACalvo, AChiò, CM, StC, ACanosa, UM, RV, MG, EM, SaC, GP, PS, LP, FC and GF. Obtained funding: ACalvo, AChiò and CM. Administrative, technical or material support: StC, MG, EM, SaC, GP, PS, FC and GF. Study supervision: ACalvo, AChiò andCM. AChiò is responsible for the overall content as guarantor.

  • Funding This work was supported by the Italian Ministry of Health (Ministero della Salute, Ricerca Sanitaria Finalizzata, grant RF-2016-02362405); the Progetti di Rilevante Interesse Nazionale programme of the Ministry of Education, University and Research (grant 2017SNW5MB); Horizon 2020 (grant RF H2020-SC1-DTH-2020-1, grant agreement ID: 101017598). This study was performed under the Department of Excellence grant of the Italian Ministry of Education, University and Research to the 'Rita Levi Montalcini' Department of Neuroscience, University of Torino, Italy.

  • Disclaimer The funders had no role in data collection or analysis and did not participate in writing or approving the manuscript.

  • Competing interests ACalvo has received a research grant from Cytokinetics. AChio serves on scientific advisory boards for Mitsubishi Tanabe, Biogen, Roche, Denali Pharma, Cytokinetics, Lilly and Amylyx and has received a research grant from Biogen. None of the other authors has any conflict of interest to disclose.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Linked Articles

  • Editorial commentary
    Gail Robinson