If only a subset of patients will show cognitive symptoms, who will they be? Paul Schulz of the Baylor College of Medicine in Houston, Texas, presented several clues in his presentation on risk factors for cognitive impairment in ALS. Schulz and collaborators have found that cognitive symptoms correlated with ApoE genotype and diabetes.

A decade ago, few neurologists considered the cognitive side of ALS, and that attitude persists in many medical school classrooms and patient resources, Schulz said. When a colleague, Paul Massman of the University of Texas at Austin,published evidence of neuropsychological deficits in people with ALS (Massman et al., 1996), Schulz was rather embarrassed he had not noticed the symptoms in his patients; most doctors assumed their slower responses were due to the trouble they had moving or speaking.

The exact rates of cognitive impairment in ALS are uncertain because methods for assessing cognition vary, said Orla Hardiman of Trinity College in Dublin, Ireland, who did not attend the meeting. In addition, most studies have been done at centers focused on cognition, which attract more patients with such impairment, potentially skewing the results. Her (unpublished) and Schulz’s (Ringholz et al., 2005) work indicate that one-third to one-half of people with ALS are cognitively normal. The rest have some signs of cognitive or behavioral involvement, and perhaps 10-20 percent meet the criteria for frontotemporal dementia.

Understanding cognitive symptoms in ALS is important to both patients and scientists, Hardiman said. People with ALS and their families may want to make decisions about end-stage care early, before mental impairment arises. And understanding the different forms of the disease is essential for clinical trials, Hardiman said: “If we don’t stratify for cognitive impairment in clinical trials, we are actually probably introducing confounders into the study.”

The full neuropsychological battery to define cognitive skills can cost $2,000-$4,000, Schulz said, so it is worth knowing who is at risk before committing to testing. Some scientists at the meeting presented short questionnaires to screen people with ALS for cognitive symptoms. Schulz is trying to understand the risk factors that might point to a heightened chance of mental involvement. Then, basic scientists can work backwards from his clinical results to figure out how those factors influence disease.

One of the first factors Schulz tackled was a genetic basis for ALS. Approximately one-tenth of patients have a family history of disease. Schulz surmised that the familial cases would be similar to each other with respect to cognitive symptoms: “I figured it would be either everybody or nobody.” But instead, he and his collaborators found the rates of cognitive dysfunction among 37 people with familial ALS to be about the same as 392 spontaneous cases, with two-thirds of familial cases and one half of spontaneous cases exhibiting cognitive impairment(see Wheaton et al., 2007). “To be honest, I was very surprised.” It is still possible that among the familial cases, mutations in certain genes have specific effects on cognition; the researchers are currently genotyping subjects.

Schulz hypothesized that ApoE genotype, which can make a 20-year difference to onset of Alzheimer disease, might be involved in ALS as well. When he and his colleagues compared ApoE genotypes to age of onset and speed of progression of motor symptoms in 852 people with spontaneous ALS, they found no effect. “There was not a single shard of difference between the various ApoE genotypes,” Schulz said. But cognitive testing of 185 subjects in the same group gave them more exciting results: people with at least one ApoE4 allele performed worse on six different tasks. So ApoE4 may be a risk factor for cognitive impairment in ALS, although, Schulz said, the association is “not strong enough to bet the farm on.”

Schulz suspected metabolism, too, might play a role, since diabetes is a risk factor for dementia (see ARF related news story) and hypermetabolism has been linked to ALS ARF related news story). In a study of 175 diabetics and 2,196 non-diabetics, he found another surprise. Diabetics were generally diagnosed with ALS at a slightly older age than those without diabetes. “That’s the shocker here—four years later age of onset,” he said. But the subset that underwent cognitive testing—24 people with ALS and diabetes, 429 with ALS only—showed a different story. Diabetics performed worse on cognitive tests at the time of diagnosis with ALS. “There’s something funny going on in diabetes,” Schulz said, although he is not sure how to explain the results. Perhaps, he guessed, the altered blood sugar levels in diabetics protects motor neurons, but makes other neurons more vulnerable.

Metabolism and weight control are often problematic in people with ALS. Hypermetabolism can cause patients to burn too many calories, and the disease often interferes with swallowing, so patients frequently lose weight. But Schulz and colleagues have found that those who maintained their body mass index (BMI) had increased survival of up to nine months. Of course, the correlation could go in either direction: people who keep their weight up could be protected, or people with a slower disease might find it easier to maintain weight. Diabetes, Schulz surmised, might counter the effects of hypermetabolism, and thus be protective. He has not yet looked for a link between body mass and cognitive function, but hopes to test a weight-maintenance program to see if it slows disease in people with ALS.

One risk factor will probably not be enough to predict cognitive impairment in ALS patients, but a collection of factors could be significant, Schulz said. He is also looking into social cognition tasks, ethnicity, traumatic brain injury and other neurodegenerative conditions as possible tip-offs to look for cognitive impairment. “This research helps clinicians determine which ALS patients may be at greater risk for cognitive impairment, so they can better triage patients for neuropsychological testing or counseling,” wrote Susan Woolley of the Forbes Norris ALS Research Center in San Francisco, in an e-mail to ARF. Woolley attended the meeting but is not involved in Schulz’s studies.

Schulz also hopes his work will inspire basic scientists studying ALS. ApoE genotype and metabolic effects should have a place at the lab bench, he said; for example, it would be interesting to cross an ALS mouse model with a diabetes model and look for changes in motor and cognitive symptoms. Denise Figlewicz, vice president for research at the ALS Society of Canada (one of the sponsors of the June meeting), agrees. “The question of calorie consumption and overall metabolism with respect to ALS patients has been a very dynamic area in the past few years,” she wrote in an e-mail to ARF. “In addition to ‘pushing’ patients to be sure to take in enough calories, I hope that more people will get involved in studying what the data we have so far really means.”

Doctors and scientists are beginning to realize that cognition is an important part of ALS. “We are starting to go beyond the phase of wondering if there is cognitive impairment,” Schulz said. “Now we are going on to the next step, to figure out who is at risk for it, and why they are at risk.”—Amber Dance.