Clinical psychologist evaluates the dangers of race-based norms in cognitive testing

A UCSF clinical psychologist has taken aim at the National Football League (NFL) for "race norming" black players diagnosed with dementia, a practice that is depriving them of the monetary awards allocated to former footballers with neurodegenerative disorders.

In her perspective that runs in JAMA Neurology on Dec. 21, 2020, Katherine Possin, Ph.D., of the UCSF Memory and Aging Center, says the denial of compensation to Black players, whose cognitive scores had been adjusted to "correct" for the race, adds fuel to the nation's "long history of structural racism and discrimination practices."

The Human Genome Project has taught us that human beings are 99.9 percent similar at the DNA level, and the remaining 0.1 percent does not vary according to socio-politically defined race categories such as Black or White."

Katherine Possin, PhD, Memory and Aging Center, University of California - San Francisco

Instead of using race norms, which typically also adjust for age, education, and gender, neuropsychologists should develop new approaches that account for "social determinants of brain health," like language, literacy, occupation, economic and financial status, she said.

Poussin's perspective follows the August 25 class-action lawsuit against the NFL filed by retired players Kevin Henry, 52, and Najeh Davenport, 41, both of whom are Black and were diagnosed with dementia after sustaining multiple concussions on the field.

Claims nixed despite dementia diagnoses

Under the landmark concussion settlement of 2013, the NFL had agreed to pay more than $1 billion to retired footballers who had suffered neurodegenerative disorders associated with chronic brain trauma. But in the cases of Henry and Davenport, their claims were denied despite testing by NFL-approved physicians that showed cognitive impairment equivalent to mild-to-moderate dementia.

In Henry's case, the NFL said a second neurological evaluation did not qualify him for compensation after his raw scores were adjusted using a "full demographic model ... which includes age, education, race/ethnicity and gender," indicating that they fell within the normal range for Blacks.

In Davenport's case, the NFL appealed his claim based on the same recalculation of cognitive scores "applying the industry-standard Heaton norms" that correct for the Black race.

Their lawsuit alleges the NFL has "repeatedly insisted that applicants' scores must be race-normed by using separate Black and White reference populations - a position that greatly reduces Black players' chances of success." It notes that an estimated 65 to 70 percent of current professional football players are Black.

The NFL is reported to have denied approximately three-quarters of the 2,000 claims for dementia submitted by all former players.

Faculty from the UCSF Memory and Aging Center had declined to join the pool of neuro specialists partnering with the NFL under the settlement agreement, after recognizing the use of race-based norms.

"Race-based norms are not scientifically sound or fair," said Bruce Miller, MD, director of the UCSF Memory and Aging Center. "Over-reliance on quantitative measures of cognition without understanding day-to-day function can lead to an incomplete picture of the illness."

Instead, Miller said, the emergence of powerful biomarkers for Alzheimer's disease and vascular dementia should be of increasing value in confirming disease diagnosis.

Alternative assessment should be used for claimants of all races

Possin, who is also affiliated with the UCSF Weill Institute for Neurosciences, described race norming as "reminiscent of a damaging, century-long history of assuming that differences on intelligence tests are primarily inherited, and then using this false assumption to legitimize the unequal distribution of resources by social class."

A more scientific alternative, which would be used for all claimants regardless of race, would evaluate those factors that influence cognitive performance, she said.

This would include accounting for the country of origin and reason for immigration; language and multilingualism; the number of years of education, quality of education and number of years of parents' education; literacy and health literacy; perceived discrimination and early life adversity; occupation, including stability, at midlife and current; income, housing, food security, and health care access at childhood, midlife and current; and residential characteristics and social cohesion at childhood, midlife and current.

Possin asserts that data and algorithms should be made publicly available "so that limitations are transparent, data can be obtained across samples and new approaches can be accelerated."

And rather than a "plug-and-play" formula of cognitive test scores, a diagnosis should be based on clinical judgment following a comprehensive evaluation of history, changes in patient functioning, neurological exam findings, lab findings, and social context.

"Changes in personality and behavior that are a very common part of the dementia syndromes that follow repeated head trauma are often not captured on standard cognitive testing," she noted, "yet can be far more disabling for patients and troubling for caregivers than deficits in cognition."