Article in brief

A large analysis of dementia rates over a 10-year period among veterans found that the incidence of dementia varied by race and ethnicity, with higher rates reported for black and Hispanic veterans. The study authors said biases, including medical decision-making bias, testing bias or survival bias, may partly explain the disparities.

A retrospective review of data on more than 1.8 million veterans found significant differences in dementia rates across racial and ethnic groups, particularly among blacks and Hispanics.

A team of researchers from the University of California, San Francisco and the San Francisco Veterans Health Care Administration (VHA) found that although there are regional variations in the incidence of dementia, black veterans and Hispanics were consistently at higher risk than others, especially whites. Compared to white veterans, Hispanic veterans had nearly twice the risk of dementia, investigators reported April 19 in JAMA.

The research team looked at dementia diagnoses made over a ten-year period. A total of 13% of VA patients, with an average age of 69 and almost all male, were diagnosed with dementia. The adjusted incidence per 1000 person-years was 14.2 for American Indian and Alaska Native patients, 12.4 for Asian patients, 19.4 for black patients, 20.7 for Hispanics and 11.5 for white patients.

“This study identifies the differences, but what we then need to know are the underlying causes of these differences. It is also important to begin to study the combined effects of biological and non-biological influences on dementia and all its subtypes. —DR. JONATHAN HAINES

Of the veterans, 88.6% were white, 9.5% were black, 1% Hispanic, 0.5% Asian, and 0.4% were Native American. Only 2.3% were female, but this matches the gender distribution of the older veteran population. All received their health care at VHA medical centers across the United States.

“American veterans are at high risk for dementia due to exposure to military-related risk factors, such as traumatic brain injury and post-traumatic stress disorder, and the high prevalence of cardiovascular risk and other non-military risk factors,” said lead investigator Kristine Yaffe, MD, professor of psychiatry, neurology, and epidemiology at the University of California, San Francisco and chief of neuropsychiatry for the VHA system. from San Francisco. “There may be genetic or resilience factors that we don’t yet know about, or maybe even a survival bias. Those old enough to have dementia may have a survival advantage. »

In a 2014 study in Lancet NeurologyDr. Yaffe and his colleagues were the first to estimate that almost 30% of the risk of dementia could be avoided.

“It’s not so much about veterans as it is about seniors in general. This is the largest, and I think the most methodologically rigorous, study of almost two million older adults to show health disparities in the incidence of dementia. It happens to be veterans, because we have national EMR data, but it’s not really veterans per se,” she said. neurology today.

“One explanation for the observed differences…is bias, including medical decision-making bias, testing bias, or survival bias,” she said. “It’s possible that clinicians are more likely to diagnose dementia in black participants or Hispanic participants due to conscious or unconscious bias.”

Additionally, non-white participants may be more likely to perform worse on cognitive tests that are often used in dementia assessments for reasons other than actual cognitive impairment, other researchers say, Dr. Yaffe noted.

The next steps will be to develop a better understanding of the individual and collective drivers of these disparities, she said. “This includes neighborhood disadvantage, quality of education, racism, etc., as well as individual aspects such as other conditions such as access to care, regional variations.”

Figure 2

“American veterans are at high risk for dementia due to exposure to military-related risk factors, such as traumatic brain injury and post-traumatic stress disorder, and the high prevalence of cardiovascular risk and other non-military risk factors There may be genetic or resilience factors that we do not yet know about, or perhaps even a survival bias Those old enough to have dementia may have a survival advantage.”—DR. KRISTINE YAFFE

“Observed differences in the diagnosis of dementia may reflect biases of the physicians making the diagnosis or biases in our cognitive tests, as well as the impact of education and medical and psychiatric variables,” added the first author. Erica Kornblith, PhD, of UCSF. Department of Psychiatry and Behavioral Sciences.

Role of clinicians

“Racial and ethnic differences underscore the need to explore the cause of these differences with the ultimate goal of improving them,” she said.

In an accompanying editorial, Gwen Yeo, PhD, principal investigator at Stanford University School of Medicine, Stanford, CA, noted that the findings are in contrast to others who have compared rates of dementia in ethnic and racial lines that have been primarily confined to local or regional health care systems or census areas.

“The fact that the veterans were all cared for in the same national health system reduces the number of variables that could affect comparisons,” she wrote.

Past comparisons of dementia rates among racial and ethnic groups revealed that education level was an important risk factor that helped explain differences between populations,” she noted.

The researchers reported that Native Americans were the least likely to live in ZIP codes where more than 25% of residents had a college education and had the highest rates of obesity (16.4%), stress disorder post-traumatic (11.2%) and alcohol-related disorder (9.1%). Hispanics had the highest rate of diabetes (36.1%) and stroke (8.2%), while black veterans had the highest rate of hypertension (73.7%).

“Clinicians should be aware of dementia risk factors among ethnic and racial populations with higher dementia rates and help control these factors, if possible,” she said, including diabetes and cardiovascular disease. .

“To the extent that dementia is preventable, particularly in underrepresented racial and ethnic populations, all of which are at increased risk, clinicians will have an important responsibility.”

Expert commentary

Kyra S. O’Brien, MD, professor of neurology and clinician at the Penn Memory and Dementia Center at the Perelman School of Medicine at the University of Pennsylvania, said, “I think it’s an interesting question whether the he increased incidence of dementia in blacks and Hispanics is due to a higher prevalence of dementia risk factors relative to inaccurate dementia diagnosis, and understanding the clinical assessment that led to the dementia diagnosis will help make light on that,” she said. neurology today.

“If the assessment were conducted without paying attention to basic education or preferred language, people could certainly be misidentified as having dementia. Clinicians may make assumptions about a person’s functional status based on inaccurate test results and possibly unconscious bias, and may waive [additional] diagnostic tests based on such false assumptions,” she said.

“There is also a need to consider the diversity of patients, including taking into account different levels of education, languages ​​and cultural beliefs, and there is a need to validate cognitive assessment tools in different languages ​​and different groups. racial. “Providers should also be trained to regularly assess a patient’s social and educational history,” Dr. O’Brien said.

“It would be helpful to delve deeply into these dementia cases. I would like to know in what contexts these diagnoses are made for each racial group. Is it in primary care, in neurology or during emergency room visits or hospitalizations? What diagnostic tests were performed in each racial group that led to the diagnosis of dementia? There is data to suggest that there are racial differences in receiving appropriate diagnostic services and prompt diagnosis, and this is something that should be investigated in this population.

picture 3

“If the assessment were conducted without paying attention to basic education or preferred language, people could certainly be misidentified as having dementia. Clinicians may make assumptions about someone’s functional status based on inaccurate test results and possibly unconscious bias and may waive [additional] diagnostic tests based on such faulty assumptions. ”—DR. KYRA S. OBRIEN

Jonathan Haines, PhD, professor and chair of the Department of Population and Quantitative Health Sciences at Case Western Reserve University School of Medicine, also commented on the findings.

“This is an interesting and well-done study that confirms the findings of other studies of differences by self-identified race/ethnicity. The authors acknowledge the various limitations in interpreting the underlying causes of these differences,” he said. neurology today. “It’s also important to note that this describes the incidence of all dementias, not just Alzheimer’s disease.”

One of Dr. Haines’ major research efforts is the Collaborative for Alzheimer’s Disease Research, an NIH-supported nationwide study of Alzheimer’s disease and other dementias, with a focus on populations that historically have not been represented in large-scale genetic research.

“As the authors point out, there are multiple potential reasons why these rates differ, including various sources of bias, multiple other unmeasured influences (e.g. social determinants of health before and after service, when and what type of service), and using billing codes (e.g., ICD-9 and ICD-10) as a proxy for the actual diagnosis,” he said.

“This study identifies the differences, but what we then need to know are the underlying causes of these differences. It is also important to begin to study the combined effects of biological and non-biological influences on dementia and all of its subtypes,” he continued.

“Potential biases need to be measured to determine which of these biases are influencing the results, but to address and understand these potential biases will require much more detailed data and it may require additional newer studies that collect this data.”

Disclosures

None of the sources cited in this article had relevant disclosures.