Are amyotrophic lateral sclerosis patients cognitively normal?

Methods.

One hundred patients with probable or definite ALS seen in the ALS Center at the University of California at San Francisco (UCSF) within the last 2 years underwent a brief cognitive screen and clinical interview as part of their initial visit. All subjects tested met criteria for the diagnosis of clinically definite or probable ALS based on the World Federation of Neurology Criteria, and none of the subjects carried a premorbid diagnosis of a cognitive disorder.13 These 100 patients (62 men, 38 women) had a mean age of 65 years (range 34 to 90 years). We queried the patients’ families for the presence of language, executive, or behavioral deficits. Patients were tested with the Mini-Mental State Examination (MMSE).14 In addition to the MMSE, patients were asked to rapidly name words that started with “D” for 1 minute and then as many animals as they could for 1 minute. Those patients unable to speak wrote or typed out the words on a communication device.

Word generation tests have been well validated both for spoken and written output and for phonemic and category fluency with good reliability and validity, with many studies correlating verbal fluency with left frontal lesions and bilateral frontal lobe lesions.15 Verbal fluency tests have been shown to be particularly useful in identifying cognitive deficits in ALS patients.16 The dysarthria and motor deficits common to ALS patients require alternations to be made when administering fluency tests. The majority of patients had intact spoken output that did not limit speed or intelligibility of output. Any patients who could write faster than speak were encouraged to write their answers, and all patients tested by speaking spoke at least 1 word per 2 seconds. As for most individuals, the output of words slowed significantly over the course of the 60 seconds, with >75% of words being generated in the first 15 seconds and 90% being generated in the first 30 seconds, indicating that generation is limited primarily by thinking of words as opposed to motor speed.15 Despite this, we allowed patients with even mild dysarthria additional time to try to generate more words. None of the patients testing abnormally benefited from the extra time; in fact, they all thought of their words in the first 15 or 30 seconds. It is noteworthy that the nondemented dysarthric subjects, while scoring in the normal range after 1 minute, did improve their score further with the extra minute. Patients with poor motor speed were given a second timed task to recopy their list of words. No patient took the entire 60 seconds to recopy the words, indicating that they were limited by thinking of the words as opposed to writing them.

The criterion for being deficient in word generation was either a score of 8 on D words or a score of 13 on animals. This was a conservative estimate based on literature controls and the lower limits of normal from our own age-matched control group, which were set at 2 standard deviations below the mean of the log-transformed data.15 More normal subjects score in the higher ranges than in the lower ranges, so the data are not normally distributed, requiring transformation of the data to determine a mean and standard deviation. We did not adjust the results for age, sex, or education, as we designed this to be a quick, simple screening tool to identify those for further testing rather than using it as a diagnostic test by itself.

All 100 subjects were asked to undergo a detailed language assessment by a speech pathologist and behavioral and neuropsychological evaluation by a neuropsychologist without knowledge of their test results with word generation. Forty-four subjects agreed to undergo further testing. Reasons for test refusal included worry about a cognitive abnormality being found, death prior to returning for their scheduled appointment, and inability to return to tolerate lengthy testing given the degree of ALS disability.

All of the 44 patients agreeing to further testing were administered a battery of neuropsychological tests and a self-report clinical interview about mental status functioning in either the UCSF ALS Center or the Memory and Aging Center. A semistructured interview conducted with an informant provided information about patient neuropsychiatric status and general level of functioning. Those patients with abnormal test results were then formally conferenced in the Memory and Aging Center to determine their diagnosis using the Neary criteria for FTLD.2 Patients were grouped into three broad categories: nondemented, possible FTLD, and probable or definite FTLD. Patients diagnosed with FTLD were subclassified into the frontal variant (FTD), PPA, or SD.

Neuropsychological functioning was assessed within four broad domains: executive functioning, memory, visuospatial skills, and language. Executive functioning measures evaluated conceptual ability, response inhibition, generation, and set shifting. They included the Wisconsin Card Sort, California Card Sort, California Trail Making Test, Stroop Interference Test, Verbal Fluency, and Design Fluency.17,18⇓ Episodic memory measures included the California Verbal Learning Test–Short Form and the Visual Reproduction subtest of the Wechsler Memory Scale III.19,20⇓ Working memory was measured with the Digit Span and Spatial Span subtests from the Wechsler Memory Scale III.20 Language function was measured by the Boston Naming Test, and visuospatial skill was measured with the Wechsler Adult Intelligence Scale III Block Design.21,22⇓

Neuropsychiatric functioning was assessed with the Neuropsychiatric Inventory (NPI), a caregiver report measure designed to assess 12 disturbances common in dementia: delusions, hallucinations, agitation, dysphoria, anxiety, apathy, irritability, euphoria, disinhibition, aberrant motor behavior, nighttime behavior disturbance, appetite, and eating abnormalities.23 Practical daily functioning was measured with the Clinical Dementia Rating Scale (CDR), a semistructured interview conducted with an informant. Areas of assessment included memory, orientation, judgment, problem solving, community affairs, home and hobbies, and personal care.24 Functional status was also assessed using the Blessed–Roth Dementia Scale. The NPI and CDR data were used to aid with the diagnosis of dementia.

Each neuropsychological test was taken from one of three sources: the Wechsler Memory Scale III, the Wechsler Adult Intelligence Scale III, and the Dellis–Kaplan Executive Function Scale. These three batteries are among the best standardized and most commonly accepted tools available, with strong reliability and validity and adequate sensitivity and specificity to detect organic brain damage.15

Tests were selected that did not depend on speed so patients would not be limited based on slow speech and slow handwriting. After written informed consent was given, a subset of subjects scoring within normal ranges on screening word generation tests underwent detailed neuropsychological testing for research purposes under a protocol approved by our institutional review board. Any patient testing abnormally on these tests was also formally conferenced in the Memory and Aging Center to confirm a diagnosis of FTLD based on Neary research criteria.2

Results.

Prior to our formal assessment of cognition, none of the subjects had been diagnosed with FTLD or any other cognitive disorder. Of the 100 patients studied, approximately 10% of patients or their families reported trouble with executive function, personality changes, and/or language difficulty. Nearly 20% of all patients were observed by staff to be disinhibited, with inappropriate disrobing, touching of staff, and/or inappropriate jokes. None of the subjects tested below the lower limits for age on the MMSE, yet 31% tested in the abnormal range on word generation tests. Seventeen patients were given a clinical referral to the Memory and Aging Center based on symptoms detected in the ALS clinic, and the remaining 83 patients were offered further neuropsychological testing as a part of an institutional review board–approved research protocol. The testing overlapped completely between the Memory and Aging Center and our research protocol.

Of the 31 patients in whom abnormal word generation was demonstrated in the initial brief cognitive screening, 17 went on to have more detailed testing to determine if they qualified for a diagnosis of FTLD. Twelve of these 17 patients were diagnosed with probable or definite FTLD (8 FTD, 3 PPA, and 1 SD), 3 were diagnosed with possible FTLD (1 PPA and 2 FTD), and 2 of the patients tested in the low normal range but have undergone cognitive decline since testing and are returning for follow-up testing. Deficits on the Stroop Interference Test, Verbal Fluency Test, California Card Sorting Test, and the Wisconsin Card Sorting test best identified FTD variant patients. Performance on the Boston Naming Test best identified patients with PPA and SD. Depression was excluded as a contributing factor to performance with the Beck Depression Scale in all patients. Consistent with Neary criteria for FTLD, patients had higher levels of apathy, disinhibition, and poor social monitoring on the NPI and largely intact memory and visuospatial skills on the CDR.

Another 27 patients who tested in the normal range on word generation tests went on to have a detailed neuropsychological assessment. Nineteen were considered normal, whereas six patients had profound behavioral abnormalities and met research criteria for FTLD (four FTD, one SD, and one PPA), one patient had a diagnosis of possible PPA, and one patient had a diagnosis of possible FTD.

As shown in the figure, comparing the groups scoring normally versus abnormally on word generation tests revealed an increased number of bulbar-onset patients with abnormalities in word generation tests. Subjects were typically tested early in their disease course when their forced vital capacity (FVC) was >50%, helping to eliminate hypoxia or hypercapnia as a potential cause of altered mental status.

• Download figure
• Open in new tab
• Download powerpoint

Figure. Word generation test results for 100 consecutive patients with ALS. The numbers of ALS patients testing normally on the 2-minute word generation tests are shown in white. The numbers of ALS patients testing abnormally are shown in black. The patients are divided into two groups based on their site of onset to show that half of bulbar-onset patients test abnormally on word generation tests compared with fewer limb-onset patients.

In summary, of the 44 ALS patients who underwent detailed neuropsychological testing, 18 were confirmed as having probable or definite FTLD (12 FTD, 4 PPA, 2SD); 5 were confirmed as having possible FTLD, meeting most but not all criteria for a definite diagnosis (2 PPA, 3 FTD); 2 tested in the low normal range and are returning for repeat testing; and 19 subjects were normal. Their clinical characteristics are summarized in table 2. Patients with dementia were more likely to be older (p < 0.002), have a lower FVC (p < 0.003), and have a family history of dementia (p < 0.01) when compared with the nondemented group; however, they were at a similar stage of their disease based on their ALS functional rating score.

Discussion.

This study evaluates the incidence of dementia in a large group of consecutive patients with ALS using detailed language, neuropsychological, and behavioral testing. The presence of frontal deficits was very high, as has been reported in smaller series. As shown in table 1, approximately 60% of ALS patients from multiple smaller studies may have demonstrable neuropsychological deficits, supporting the generalizability of our results. Testing revealed 52% of the 44 ALS subjects met recently developed research criteria for possible or probable FTLD. All the major subtypes of FTLD (FTD, SD, and PPA) were present in proportion to what is seen with sporadic FTLD, whereas bulbar presentations of ALS seemed overrepresented in the patients who also had FTLD. This strong association between ALS and FTLD offers new clues into the biological underpinnings of both disorders and also brings new clinical opportunities and challenges to the diagnosis and management of the ALS patient.

The clinical approach to ALS has been strongly influenced by the premise that the coassociation of dementia is low. The findings in this study contradict this premise as half of our ALS cohort had significant cognitive abnormalities. In a neurologic condition where swallowing, movement, and activities of daily living are steadily disintegrating, it is easy to overlook the appearance of a frontal lobe dementia whose major features are apathy, euphoria, or deficits in planning and executive control. Compounding the difficulty, owing to the presence of frontal lobe dysfunction, the patients lose insight into their illness and rarely complain of cognitive, emotional, or behavioral changes.

The unexpectedly high frequency of dementia in ALS underscores the need for neuropsychological testing and caregiver/family interviews to document changes in functioning. The simple 2-minute word generation test was an excellent screening tool with high sensitivity for identifying patients in whom more detailed neuropsychological evaluations were needed and likely to show further frontal/executive deficits. Neuropsychological testing in ALS proved difficult, and tasks needed to be chosen that did not depend on motor speech. Risk factors for developing dementia may include older age, lower FVC, and family history of dementia (see table 2).

Knowing whether or not an ALS patient has cognitive impairment brings up important considerations regarding routine management. Dementia patients often need simpler tools for communication and more involvement of the caretaker. Nutritional maintenance may require closer supervision of eating as gorging is a prominent abnormality seen in FTLD but can threaten the life of the ALS patient who already is experiencing impaired swallowing. The presence of FTLD impacts a patient’s ability to ambulate safely, given the loss of impulse control. Furthermore, mental competency will have to be carefully scrutinized because ALS patients and their families often face decisions regarding a patient’s disability and autonomy, the legal validity of new wills, and even the patient’s right to participate in assisted suicide. It even calls into question the idea of ALS patients having a “nice” personality (Wilbourn and Mitsumoto, unpublished results, 1998). Finally, in FTLD, selective serotonin reuptake inhibitors can be particularly valuable for the treatment of depression as well as the alleviation of impulsive and compulsive behaviors. These compounds should now be considered in the treatment of these behaviors in ALS.

Continued study of this link between ALS and FTLD is needed at the clinical, imaging, and genetic levels. The vulnerability of frontal and motor neurons suggests a significant overlap between these two diseases and supports the idea of a potential common etiology. The presence of tau and ubiquitin inclusions in FTLD and FTLD/ALS syndromes offers new molecular probes for the study of ALS. What initially appears as confound can ultimately offer new weapons in the battle against ALS and FTLD.