Article Text


Cognitive dysfunction in lower motor neuron disease: executive and memory deficits in progressive muscular atrophy
  1. Joost Raaphorst1,2,
  2. Marianne de Visser1,
  3. Marie-José van Tol3,
  4. Wim H J P Linssen2,
  5. Anneke J van der Kooi1,
  6. Rob J de Haan4,
  7. Leonard H van den Berg5,
  8. Ben Schmand1,6
  1. 1Departments of Neurology, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands
  2. 2Department of Neurology, Sint Lucas Andreas Hospital, Amsterdam, The Netherlands
  3. 3Department of Psychiatry, Leiden University Medical Centre, Leiden, The Netherlands
  4. 4Clinical Epidemiology and Biostatistics, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands
  5. 5Department of Neurology, Rudolf Magnus Institute of Neuroscience, University Medical Centre Utrecht, Utrecht, The Netherlands
  6. 6Department of Psychonomics, University of Amsterdam, Amsterdam, The Netherlands
  1. Correspondence to J Raaphorst, Department of Neurology (H2-235), Academic Medical Centre, PO Box 22660, 1100 DD Amsterdam, The Netherlands; j.raaphorst{at}


Aim In contrast with findings in amyotrophic lateral sclerosis (ALS), cognitive impairments have as yet not been shown in the lower motor neuron variant of motor neuron disease, progressive spinal muscular atrophy (PMA). The objective of this study was to investigate cognitive function in PMA and to compare the cognitive profile with that of ALS. In addition, visuospatial functions were assessed comprehensively; these tests are underrepresented in earlier neuropsychological investigations in ALS.

Methods 23 PMA and 30 ALS patients (vital capacity >70% of predicted value) underwent a neuropsychological assessment adapted to motor impairments: global cognitive and executive functioning, psychomotor speed, memory, language, attention and visuospatial skills. The results were compared with age, education and sex matched controls and with normative data.

Results Compared with controls, PMA patients performed worse on attention/working memory (digit span backward), category fluency and the Mini-Mental State Examination. Compared with normative data, PMA patients most frequently showed impairment on three measures: letter–number sequencing, and immediate and delayed story recall. 17% of PMA patients showed cognitive impairment, defined as performance below 2 SDs from the mean of normative data on at least three neuropsychological tests. In ALS, similar but more extensive cognitive deficits were found. Visuospatial dysfunction was not found in PMA and ALS.

Conclusions 17% of PMA patients have executive and memory impairments. PMA with cognitive impairment adds a formerly unknown phenotype to the existing classification of motor neuron diseases.


Statistics from


Cognitive impairments have been found in patients with motor neuron disease with upper motor neuron (UMN) involvement: 30% of patients with amyotrophic lateral sclerosis (ALS) and a fair proportion of patients with primary lateral sclerosis have executive and memory deficits.1–4 Whether this holds true for patients with only lower motor neuron (LMN) signs is unclear.5 Progressive spinal muscular atrophy (PMA) is an adult onset progressive LMN disorder. There is still debate on whether PMA is a distinct disease entity or whether it represents one end of the spectrum of motor neuron diseases. In favour of the latter claim are the clinical, genetic and pathological features that PMA shares with ALS.6–9 We therefore hypothesised that PMA patients may exhibit cognitive impairments in similar domains compared to ALS. PMA seldom starts in bulbar neurons. As cognitive impairments have been found to be related to bulbar onset ALS in some studies, cognitive impairment may not be encountered as frequently in PMA compared with ALS.10–12

The main objective of this study was to explore the occurrence of cognitive dysfunction and its clinical correlates in PMA in comparison with that of ALS. In addition, we aimed to further define the cognitive profile of ALS by measuring visuospatial functions. Visuospatial tests are underrepresented in earlier neuropsychological investigations in ALS and thus the presence of visuospatial deficits is unclear.12 13 Visuospatial dysfunction may be suspected in ALS as patients with other disorders affecting the motor system (eg, Parkinson's disease or dystonia) have shown visuospatial deficits.14 15



Patients were recruited between January 2007 and January 2009 from the outpatient clinics of the Academic Medical Centre, Amsterdam and the University Medical Centre Utrecht. Spouses and friends of patients were asked to participate as controls. We chose this control group to ensure that it would be matched for education and age, which are important variables that may influence cognitive measures.

The medical ethical committees of the hospitals approved the study. Written informed consent was obtained from all participants. Included were PMA patients who fulfilled the criteria as described previously6: (1) a disease duration of less than 5 years from the time of diagnosis; (2) clinical and electrophysiological evidence of LMN involvement in two or more of four regions (bulbar, cervical, thoracic and lumbosacral); (3) no conduction blocks on nerve conduction studies; and (4) no clinical UMN signs and symptoms, including forced yawning, crying and laughing, clonus of masseter reflex, (sub)clonic myotatic reflexes, Hoffmann–Trömner sign, extensor plantar response or spasticity. All patients with ALS included in the study were classified as probable or definite according to the revised El Escorial criteria.16 Patients and controls were excluded if they had dementia according to consensus criteria (Diagnostic and Statistical Manual of Mental Disorders, American Psychiatric Association 1994 and Lund–Manchester criteria17), a history of another neurological disorder associated with cognitive impairment, a vital capacity lower than 70% of the predicted value (to prevent bias of respiratory compromise on cognitive measures),18 severe dysarthria or anarthria, or if they were unable to push a button with the index or middle finger of their dominant hand. Patients had to speak Dutch fluently and they had to be free of psychoactive medications.

Clinical assessment

The Amyotrophic Lateral Sclerosis Functional Rating Scale (ALSFRS-R) was used to evaluate the functional status of the patients.19 To assess UMN involvement, an unvalidated scale was used that summated myotatic and pathological UMN reflexes and pseudobulbar affect. The scale is a modification of the scale used by Ellis et al (table 1).20

Table 1

Upper motor neuron scale

In addition, the following clinical variables were assessed: site of onset (bulbar of limb); degree of bulbar involvement (defined as a score below 4 on one or more of the bulbar items of the ALSFRS-R); disease duration (defined as the time between the first symptom and the neuropsychological assessment); vital capacity (percentage of predicted value: predVC); and years of formal education.

Neuropsychological assessment

Neuropsychological tests were administered in a fixed order with rest periods if needed. To further avoid a negative effect of fatigue, testing was done in an outpatient clinic in the vicinity of the patients' homes.

Premorbid intellectual ability (National Adult Reading Test, Dutch version, DART), global cognitive functioning (Mini-Mental State Examination (MMSE)) and six cognitive domains were tested: attention and working memory, executive functions, psychomotor speed, memory, language and visuospatial abilities (for neuropsychological tests and references, see appendix A, available online). Five tests relied on speed: Stroop Test Parts A (word naming), B (colour naming) and C (word interference on colour naming), category and letter fluency. The Stroop Test Part A was used to control for a possible negative effect of dysarthria on the fluency tests by generating the measure fluency–motor speed corrected, according to the following formula (for letter fluency):Letter fluencymotor speed corrected=(180(time per word for Stroop part A)fluency score)fluency scoreWhere fluency score=total amount of produced words with letters K, O and M in 180 s (60 per letter). The time per word for Stroop Part A is the mean time (in seconds) it took the subject to pronounce a colour word. A similar formula for category fluency was used.

The Stroop Test Part B (colour naming) and C (word interference on colour naming) were not accommodated for motor impairment. These tests were used to calculate the Stroop interference time (Stroop C–Stroop B, not motor speed dependent). The mental rotation task is described in detail in appendix B (available online). The Hospital Anxiety and Depression Scale (HADS) was used to examine symptoms of depression or anxiety.21

Statistical analyses

Differences in demographic and clinical characteristics between the PMA and ALS patients and the control group were analysed using a two group t test. When data and their log or inverse transformations were not normally distributed, non-parametric Mann–Whitney U tests were used to analyse differences between groups. The χ2 test was used to analyse nominal variables. Differences between the scores for neuropsychological measures of PMA and ALS patients and the control group were expressed in Hedge's g effect size. By convention, effect sizes of 0.2, 0.5 and 0.8 are considered small, moderate and large, respectively.22 A second set of analyses was performed to examine the cognitive deficits in a manner similar to clinical practice. Standard scores, either scaled scores (mean 10; SD 3) or T scores (mean 50; SD 10) were derived from available normative data in test manuals. Normative scores were taken from the following sources: Stroop Colour Word Test and RBMT,,23 COWAT and category fluency,24 BNT25 and MWCST.26

Frequency of cognitive dysfunction

A test score was considered impaired if it was more than 2 SDs below the mean score of the normative sample, after correction for age, gender and, if possible, education. Cognitive dysfunction was considered to be present if performance on more than two neuropsychological tests was impaired (measures without adjustment for motor impairment (Stroop Test Parts A, B and C) were excluded from this analysis). This criterion ensured that less than 5% of the control subjects would be impaired. In addition, the criterion minimises the possibility that impaired performance reflects a chance finding due to the large number of measures employed.

Demographic and clinical variables associated with cognitive impairment

To identify variables associated with cognitive impairment, ALS and PMA patients with and without cognitive impairment were compared using the following variables: age; education; HADS total score and subscores; disease duration; ALSFRS-R (including a bulbar sum score); site of onset; and predVC. A possible relation of UMN signs with cognitive impairment was explored in three ways: firstly, performance of the PMA and ALS patient groups on the neuropsychological tests were compared; secondly, in the PMA and ALS patients together (MND patients), the UMN score was compared between MND patients with and without cognitive impairment; and thirdly, within MND patients, correlations between the UMN score and performance on the neuropsychological tests were analysed.


Demographic and clinical characteristics

Twenty-three PMA patients, 30 ALS patients and 24 age, sex and education matched controls were included. Age, sex distribution, years of education, DART-IQ estimate, HADS scores and the ALSFRS-R were not significantly different in patients and controls (PMA vs controls, ALS vs controls and PMA vs ALS) (table 2). The median sum score of the three ALSFRS-R bulbar items (normal score=12) was 10 (range 6–11) in ALS patients and 12 (range 9–12) in PMA patients (p<0.05). The UMN score was higher in ALS compared with PMA patients (p<0.001).

Table 2

Demographic and clinical characteristics of patients and controls

Neuropsychological performance of PMA and ALS patients and control subjects

Compared with controls, PMA patients performed worse on an attention/working memory test (digit span backward), category fluency and the MMSE. Compared with controls, ALS patients showed impaired performance on attention/working memory (letter–number sequencing), category and letter fluency, Stroop Test Part B (not corrected for motor impairment), visual memory and naming (table 3). Effect sizes ranged from negligible to moderate (table 3). There were no statistical differences in performance on the neuropsychological tests between the PMA and ALS patient groups. Not all patients completed the neuropsychological battery: two visuospatial tests were not performed by some patients due to fatigue, technical problems (one ALS patient, one control) and disease progression (not being able to push a button) between the moment the patient decided to participate and the neuropsychological investigation (two ALS patients, one PMA patient).

Table 3

Neuropsychological test results in patients and controls: raw scores, p values and effect sizes

Frequency of cognitive dysfunction

Individual performance compared with normative data showed three measures that were most frequently impaired in PMA and ALS patients: letter–number sequencing, and immediate and delayed story recall. In the remaining tests, the frequency of impairment was 10% or lower (figure 1).

Figure 1

Percentages of patients and controls with abnormal scores: lower than 2 SDs below the population mean. Tests with abnormal scores in less than 5% of progressive spinal muscular atrophy (PMA) and amyotrophic lateral sclerosis (ALS) patients are omitted. HC, healthy controls; MWCST, Modified Wisconsin Card Sorting Test.

Seventeen per cent of PMA patients, 27% of ALS patients and 4% of controls displayed cognitive impairment, defined as a score >2 SD below the mean of normative data on at least three neuropsychological tests (table 4). Except for the Stroop interference condition and the Rey Auditory Verbal Learning Test recognition, impairments on all tests were observed in one or more of the eight cognitively impaired ALS patients. Tests that showed impairments in at least 50% of the cognitively impaired PMA and ALS patients were letter–number sequencing, and immediate and delayed story recall.

Table 4

Number of impaired tests and percentages of patients and controls demonstrating impairments

Clinical variables associated with cognitive impairment

The cognitively impaired ALS patients had more often bulbar onset compared with the ALS patients without cognitive impairment (p<0.05). None of the PMA patients had bulbar onset. One of the four cognitively impaired PMA patients and six of the eight cognitively impaired ALS patients had bulbar involvement (NS). Cognitively impaired ALS and PMA patients did not differ from patients without cognitive dysfunction with respect to age, education, HADS total score and subscores, disease duration, ALSFRS-R, predVC or UMN score. When PMA and ALS patients were analysed together (n=53), the UMN score did not differ between patients with and without cognitive impairment, and nor did the UMN score correlate with any neuropsychological test score.


Cognitive dysfunction in PMA

This study is the first to demonstrate cognitive dysfunction in a relatively large cohort of PMA patients taking into account the potential negative influence of motor impairment and respiratory compromise on the neuropsychological assessment. Group differences between PMA and controls were found for attention/working memory (digit span) and category fluency. Seventeen per cent of PMA patients showed cognitive impairment, with attention/working memory (letter–number sequencing) and story recall being most frequently abnormal. Some tests showed group effects while they were not the most frequently abnormal tests. This inconsistency may result from a significant proportion of patients that have subclinical levels of impairment (eg, in category fluency) yielding group differences, but no individual impairments.

The cognitive impairments in PMA patients, together with similar findings in ALS and PLS patients, suggest that extramotor cerebral involvement is present in MND regardless of the presence or absence of UMN signs.1–3 Indeed, when PMA and ALS patients were analysed together, no correlation between the degree of upper motor neuron signs and the presence of cognitive impairment could be demonstrated in this study. One earlier study also examined cognitive functions in PMA with adaptation of tests to motor impairment and exclusion of patients with respiratory weakness. This study did not show cognitive deficits. However, it may have been underpowered with 12 PMA patients undergoing eight neuropsychological tests.5

In the present study, PMA patients were found to have executive dysfunction with impairments on tests of category fluency and digit span backward. The profile of executive dysfunction is supported by the results of the letter–number sequencing test, a working memory task that was the most frequent abnormal test compared with norm scores in our PMA patients (22%). This is in agreement with impairments of attention and working memory in patients with ALS in our and other studies.13 27–30

As executive dysfunction, including fluency deficits, has been demonstrated in patients with depression,31 it is worth mentioning that our PMA patients did not show signs of depression, and no differences in depression or anxiety scores were shown between PMA patients with and without cognitive impairment.

Cognitive dysfunction in ALS: memory

It has been suggested that in ALS retrieval is predominantly affected and recognition/encoding is relatively intact. Retrieval requires integrity of the prefrontal cortex for its executive component. Indeed, in ALS, both prefrontal dysfunction and free recall problems have been shown.1 32 33 However, our findings in ALS patients indicate that in addition to retrieval (story recall), recognition is impaired, as the Doors B test showed group effects in ALS patients compared with controls. The Doors B test does not rely on active retrieval but on passive recognition and its results are more specifically related to the process of encoding. Therefore, impairments on the Doors B test in ALS patients suggest that encoding deficits may also underlie memory impairments in ALS.1 10 12 13 Whether encoding deficits in ALS are related to hippocampus dysfunction or attention deficits, or both, needs to be studied further.

Cognitive dysfunction in ALS: visuospatial functions

in the present study, no visuospatial abnormalities were found in ALS (or in PMA) patients compared with controls. We used more sensitive tests (eg, space relations) compared with earlier studies that employed solely the judgment of line orientation or the Visual Space and Object Perception battery.1 27 34 35 Therefore, the findings of the present study demonstrate the ‘frontotemporal’ cognitive profile in non-demented ALS, as memory, language and executive dysfunction were observed while visuospatial dysfunction was absent.2 13 27 33 35–37

Comparison of the cognitive profiles of PMA and ALS

In both PMA and ALS patients, category fluency deficits were found. Also, in both PMA and ALS patients, diverging results regarding executive tests were found: impaired fluency (letter and or category fluency) and normal results on the modified Wisconsin Card Sorting Test and Stroop interference tests. This pattern is in agreement with other neuropsychological studies in ALS33 36 and it emphasises the different pathways in the executive system which may be selectively involved in MND.28 In addition to executive deficits, in ALS, but not in PMA, the cognitive profile includes recognition and language deficits.27 38

The MMSE was significantly lower in PMA patients but not in ALS patients compared with controls. However, differences between the MMSE scores were negligible and have thus very little clinical significance. In addition, the MMSE has a strong ceiling effect in controls (ie, cognitively normal people perform (almost) perfectly, resulting in small variance of scores in controls).

Bulbar involvement

The finding of cognitive impairment in our PMA patients, in whom bulbar symptoms were absent or if they occurred during the course were very mild, corroborates earlier findings that cognitive impairment does occur in MND patients with isolated limb involvement.39 The potentially negative effects of dysarthria on neuropsychological measures are therefore not applicable to the great majority of PMA patients in this study.1 In our ALS patients, cognitive impairment was related to bulbar onset but not to bulbar involvement.11 These results show that in MND patients bulbar onset predisposes to cognitive impairment but is not a conditio sine qua non.

Upper motor neuron involvement

This study shows that MND patients have cognitive impairments that are not per se related to the presence of UMN signs. One must bear in mind that the absence of (bulbar or spinal) UMN signs does not exclude UMN pathology in MND: pyramidal tract pathology in the spinal cord may go undetected in PMA patients and is only identified at autopsy.9 Indeed, it is interesting to examine whether cognitive deficits predict the development of UMN signs (ie, ALS). However, this requires a longitudinal study which is beyond the scope of this paper.

Strengths and limitations


Firstly, bias due to severe motor impairment, dysarthria and respiratory dysfunction were minimised in our study design. Secondly, a wide range of cognitive domains were measured, including three tests assessing different visuospatial abilities. Thirdly, two analytical procedures were applied to assess cognitive function (ie, comparison with normative data and comparison of neuropsychological performance with a matched control group).


Firstly, some normative datasets are of better quality than other datasets (eg, correction for education is not possible for every test). Secondly, our criterion for respiratory failure was based on vital capacity only (VC <70% of predicted value), which is a widely used and validated measure. However, this may not have entirely excluded those patients with incipient respiratory failure. Currently, other methods are known to have a higher sensitivity to asses respiratory failure in MND (eg, SNIFF nasal inspiratory pressure). Thirdly, patients with MND may complain of fatigue which may influence neuropsychological scores negatively. We tried to exclude fatigue as much as possible by appropriate timing and setting of the neuropsychological examination. However, as we did not quantify fatigue, we cannot completely exclude such an influence on our data.

Finally, as we took data from a clinic based cohort, recruitment bias may be suspected. However, in The Netherlands, the great majority of patients with (suspected) MND, and not just ‘special’ cases, are referred to specialised clinics (eg, Amsterdam and Utrecht) to verify the diagnosis. We estimate that the proportion of patients who refused to cooperate with the study was about 20–30%.

In conclusion, in this study, executive dysfunction and verbal recall deficits were demonstrated in PMA. In ALS, similar but more extensive cognitive deficits were found. The cognitive impairments in different MND phenotypes may contribute to understanding the extramotor involvement and the heterogeneity within the MND spectrum.


The Mental Rotation Test was developed and kindly provided by Sjoerd J de Vries and Theo Mulder (Centre for Human Movement Sciences, University Medical Centre Groningen). Theo Mulder is now Director of Research, Royal Netherlands Academy of Arts and Sciences. We gratefully acknowledge the assistance of Hanneke Börger, Hepke Grupstra (Academic Medical Centre, Amsterdam) and Inge van Beilen (University Medical Centre Utrecht) with the recruitment of the patients. We thank the patients and their partners for their participation and the staff of many neurology and rehabilitation outpatient clinics in The Netherlands for their hospitality.


View Abstract

Supplementary materials


  • Funding The study was supported by a grant from The Netherlands ALS Foundation (JR).

  • Competing interests None.

  • Ethics approval The study was conducted with the approval of the medical ethics committees of the hospitals.

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

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.