A multicentre cross-sectional study has revealed that patients living with pure motor amyotrophic lateral sclerosis experience subtle yet statistically significant difficulties when discriminating negative facial emotions, even when standard cognitive testing and comprehensive psychological screenings detect no overt cognitive or mood disturbances. Diagnostic frameworks such as the revised Strong criteria group these multi-system clinical presentations under the overarching classification of amyotrophic lateral sclerosis–frontotemporal spectrum disorder. Under this clinical taxonomy, approximately half of all diagnosed individuals exhibit measurable cognitive or behavioural alterations, and roughly ten to fifteen percent meet diagnostic thresholds for frontotemporal dementia. Patients whose clinical evaluations reveal no demonstrable cognitive deficits or behavioural changes are categorized as exhibiting pure motor amyotrophic lateral sclerosis. The present findings suggest that standard clinical screening tools may routinely underestimate subtle socio-cognitive changes occurring in individuals classified within this ostensibly purely motor subset.

Social cognition represents a multifaceted cognitive domain encompassing the perceptual decoding of affective expressions alongside theory of mind, which reflects the higher-order capacity to infer the distinct thoughts, intentions, and emotional states of others. This omission persists despite growing clinical recognition that interpersonal emotional friction and altered social interaction can place profound stress on domestic caregivers and diminish overall family quality of life. Compounding the diagnostic challenge, basic emotion recognition is known to be vulnerable to affective state changes, such as comorbid depressive symptoms, which affect approximately one third of all patients diagnosed with the condition. Although existing screening batteries like the Edinburgh Cognitive and Behavioural ALS Screen include an adapted theory of mind assessment embedded within executive measures, they omit dedicated facial emotion recognition tasks.

To rigorously examine whether facial emotion discrimination is impaired in individuals strictly verified as having pure motor presentation, researchers from neuromuscular referral centres in Basel, St. Gallen, and Ulm conducted a harmonized cross-sectional study between November 2018 and December 2024. The participant cohort comprised thirty-five patients diagnosed with probable or definite disease according to the revised El Escorial criteria. Every enrolled patient underwent stringent multi-tiered screening to ensure the complete absence of secondary cognitive or behavioural compromise. Participants demonstrated completely normal baseline scores across all individual cognitive subdomains of the German or Swiss German versions of the Edinburgh Cognitive and Behavioural ALS Screen, deliberately including non-specific domains such as memory and visuospatial performance. In addition, carer interviews utilizing the dedicated Behaviour Screen verified that no patient exhibited apathy or frontotemporal behavioural changes. Potential mood confounders were controlled by enforcing a pre-specified cutoff score below twenty-nine on the twelve-item ALS-Depression-Inventory, an assessment designed to prevent motor weakness from artificially inflating clinical depression scores. Diagnostic stability was subsequently validated across all thirty-five cases via a mandatory one-year clinical follow-up.

For rigorous baseline comparison, the investigative team recruited thirty-five psychiatrically and neurologically healthy control participants from the Memory Clinic research pool in Basel. The control cohort underwent individual-level one-to-one matching with the clinical patient cohort, maintaining exact matching on biological sex while aligning as closely as possible on age and total years of formal education. Healthy participants were thoroughly screened to rule out covert cognitive decline or affective disturbance. Their cognitive normality was objectively verified using the Montreal Cognitive Assessment, requiring performance at or above the tenth normative percentile, alongside an informant-rated score of 3.29 or lower on the Informant Questionnaire on Cognitive Decline in the Elderly. Controls also completed age-stratified depressive screening, requiring a score of nine or lower on the Beck Depression Inventory for participants younger than sixty-five years, or a score of five or lower on the Geriatric Depression Scale for those aged sixty-five and older. Neither group had a history of substance abuse, major systemic medical disorders, recent invasive ventilation, or general anaesthesia administration within the preceding three months.

Experimental evaluation was conducted in a quiet setting using standardized computer hardware and software running an identical German-language experimental script under untimed conditions. Participants completed two sequential computerized facial processing tasks designed with morphed photographic stimuli derived from the established Facial Expressions of Emotion: Stimuli and Tests battery. To confirm that any observed perceptual deficit was specific to emotional decoding rather than generalized visual facial processing, participants first completed the Facial Sex Matching Task. This preliminary control measure presented side-by-side photographic pairs of morphed neutral human faces and required participants to judge whether both individuals shared the same biological sex. Across six consecutive untimed test trials, the task generated a total performance score ranging from zero to six. If advanced motor impairment prevented an affected patient from operating the computer input mouse, a trained clinical examiner entered the explicitly indicated verbal response without providing evaluative performance feedback.

Following baseline non-emotional screening, participants completed a shortened five-emotion version of the Facial Emotion Intensity Rating Test – Congruent and Incongruent. While conventional assessments utilize static facial photographs displayed at peak expression intensity that frequently elicit confounding ceiling effects among healthy cohorts, this test utilizes morphed human faces calibrated at subtle, threshold-level emotional intensity. The task evaluates five fundamental emotions: anger, disgust, fear, happiness, and sadness across male and female faces. Each unique face was displayed five separate times, accompanied by a single verbal label presented above the portrait in a fixed sequential sequence: anger, sadness, fear, happiness, and disgust. Over fifty randomized trials, participants rated how intensely the displayed face portrayed the designated label using an unnumbered seven-point scale ranging from no emotion to maximum emotion. This specific paradigm demanded both congruent ratings, where the label matches the portrayed facial affect, and incongruent ratings, where the label diverges from the underlying facial expression.

Primary statistical analyses focused on the contrast rating score for negative emotions, designated as the NET-Contrast metric, which isolates discrimination accuracy by subtracting the mean of twenty-four incongruent negative ratings from the mean of eight congruent negative ratings across anger, fear, sadness, and disgust. Evaluated using a two-sided Welch’s t-test assuming independent samples, the analysis revealed that patients with pure motor disease achieved significantly lower NET-Contrast scores than matched healthy controls (t(68.0) = -2.62, p = 0.011), with a mean score difference of -0.64 (95% confidence interval [-1.13, -0.15]). This group divergence represented a moderate statistical effect size, documented by a Cohen’s d of 0.63. A trial-level linear mixed-effects model incorporating random participant slopes and stimulus intercepts fully reproduced this finding (beta = -0.64, standard error = 0.24, p = 0.011). When congruent and incongruent response scores were examined in isolation, neither reached independent statistical significance (p = 0.137 and p = 0.128, respectively), indicating that the impairment emerges primarily from the relative differentiation between matching and non-matching emotional categories.

Exploratory post-hoc evaluations examined individual emotional categories and valence boundaries to establish whether the observed perceptual deficit targeted specific feelings. When contrast metrics were calculated for isolated negative emotions, Holm-adjusted comparisons revealed statistically significant decrements in pure motor patients for anger (adjusted p = 0.026) and fear (adjusted p = 0.039), whereas differences in sadness and disgust failed to meet statistical significance. However, a two-by-four mixed analysis of variance examining the group-by-emotion interaction was not significant (F(3, 204) = 1.65, p = 0.178, partial eta squared = 0.024), demonstrating no statistical evidence that the impairment varied significantly across the four negative emotions. Similarly, assessing cross-valence discrimination between negative emotions and happiness yielded a small-to-medium effect that did not achieve significance on its own (t(65.6) = -1.80, p = 0.077, Cohen’s d = 0.43). A two-by-two mixed analysis of variance revealed no significant group-by-valence interaction (F(1, 68) = 0.56, p = 0.459), indicating that the clinical deficit was not proven to be exclusively confined to negative emotions over positive valence.

Crucially, performance on the non-emotional Facial Sex Matching Task demonstrated that basic structural face processing remained preserved across both clinical cohorts. The Wilcoxon rank-sum test indicated no significant difference between patients and healthy controls (W = 585, p = 0.725), with both groups performing near task ceiling as indicated by maximum scores achieved by 54 percent of patients and 57 percent of controls. Sensitivity analyses further verified that clinical findings were not driven by multi-site recruitment differences across centers. Although patients evaluated in Basel exhibited higher baseline contrast scores, restricting comparative evaluations exclusively to the larger Ulm cohort of twenty-two patients maintained statistical significance (p = 0.014, Cohen’s d = 0.69), as did excluding Basel participants entirely from the overall clinical pool (n = 31, p = 0.001, Cohen’s d = 0.83). Individual patient data distributions overlapped considerably with controls: while 74 percent of patients scored below the median of the healthy control group, only 14 percent scored below the lowest individual score observed among controls.

These experimental observations align with the overarching concept of an amyotrophic lateral sclerosis–frontotemporal spectrum disorder, wherein pathological neurodegeneration spans a continuous gradient from predominantly motor phenotypes to extensive frontotemporal degeneration. Nonetheless, several methodological parameters contextualize the findings. Cognitive stratification utilized the validated Edinburgh multi-domain screen rather than an exhaustive full-battery neuropsychological assessment, leaving open the possibility of subtle, unmeasured cognitive variance. Furthermore, the cross-sectional structure of the investigation prevents determination of whether subtle emotion recognition deficits precede broader cognitive deterioration or progress along an autonomous trajectory.

Subject of Research: Facial emotion discrimination performance in pure motor amyotrophic lateral sclerosis

Article Title: Impaired discrimination of negative facial emotions in pure motor ALS

Article References: Meyer, A., Semenkova, A., Wezel, S., Weber, M., Schweikert, K., Schläger, R., Wendebourg, J., Piguet, O., Berres, M., Braun, N., Lulé, D., & Sollberger, M. (2026). Impaired discrimination of negative facial emotions in pure motor ALS. Journal of Neurology, 273(11), Article 709. https://doi.org/10.1007/s00415-026-14241-x

Image Credits: AI Generated

DOI: 10.1007/s00415-026-14241-x

Keywords: Amyotrophic Lateral Sclerosis, Pure Motor ALS, Social Cognition, Emotion Recognition, Facial Emotion Discrimination, Negative Facial Emotions, FEIRT-CIC, Edinburgh Cognitive and Behavioural ALS Screen, ALS-Frontotemporal Spectrum Disorder, Theory of Mind, Facial Sex Matching Task, Neurodegenerative Disease, Cognitive Screening, Depression Screening, Frontotemporal Dementia, Affective Processing, Clinical Neuropsychology