Vital statistics data may significantly underestimate mortality associated with dementia, a longitudinal analysis suggested.
The percentage of deaths attributable to dementia was 13.6% -- 2.7 times the percentage recorded on their death certificates -- in a study of older adults followed for up to 10 years, reported Andrew Stokes, PhD, of Boston University School of Public Health, and co-authors in .
"There are unique reasons why national mortality statistics might underestimate the mortality burden of Alzheimer's and related dementias," Stokes said.
"First, there's no routine screening in primary care, so high rates of undiagnosed dementia are high," he told 鶹ý. There's also stigma attached to dementia, which can affect coding on the death certificate, he noted.
In addition, many older adults have multiple comorbidities, Stokes pointed out. "It can be easy to attribute death to a more proximate cause, which could be cardiovascular disease or pneumonia or sepsis or other causes," he said.
To conduct their analysis, Stokes and co-researchers followed a sample of participants from the (HRS), a nationally representative longitudinal survey of U.S. older adults, from baseline cognitive assessment in 2000 through 2009, linking HRS data with National Death Index records.
Of 7,342 people in the sample, 60% were women. At baseline, 64% were 70 to 79 years old, 31% were 80 to 89, and 5% were 90 to 99. The researchers used adjusted hazard ratios relating dementia and all-cause mortality to calculate population-attributable fractions and compared results with cause-of-death information on death certificates.
An estimated 13.6% (95% CI 12.2%-15.0%) of deaths were attributable to dementia. In contrast, the proportion of deaths with dementia as an underlying cause on death certificates was 5.0% (95% CI 4.3%-5.8%). Incorporating deaths attributable to cognitive impairment without dementia (CIND) led to an even greater gap.
The mortality burden of dementia was higher among Blacks (24.7%, 95% CI 17.3-31.4) and Hispanics (20.7%, 95% CI 12.0%-28.5%) than white participants (12.2%, 95% CI 10.7-13.6), as well as in people with less than a high school education (16.2%, 95% CI 13.2%-19.0%) compared with the college-educated (9.8%, 95% CI 7.0%-12.5%).
Dementia-related deaths were underestimated more among Blacks (7.1 times) and Hispanics (4.1 times) than white people (2.3 times). Underestimation was greater in men than women, and in people with less than a high school education than in more educated people.
"The overarching findings are not surprising," observed Matthew Baumgart, vice president of health policy for the Alzheimer's Association in Chicago, who wasn't involved with the study.
"We know that Alzheimer's and dementia causes more deaths -- and creates a greater burden -- than what is officially reported," Baumgart told 鶹ý.
"While the system has gotten better at reporting deaths due to Alzheimer's disease, this study shows that additional improvements are needed so we can properly address the full extent of the burden Alzheimer's and other dementias are having on our country," he added. "We need to expand access to better quality care for people living with dementia, and to do that, we need to recognize the true burden it causes."
Understanding what people die of is important for resource allocation, Stokes said. "The prominence of dementia as a cause of death on the American mortality landscape is potentially significantly underestimated," he noted. "If it's a more prominent cause of death, that means we need to align our resources accordingly."
"Likewise, patterns across education, racial, and ethnic groups are quite different: Black and Hispanic communities appear to be much more affected by the burden of dementia when you use HRS data than when you use mortality statistics on their own," he added.
The researchers noted several limitations to their study. They could not determine mortality associations for dementia subtypes like Alzheimer's disease or vascular dementia. Although they used validated criteria for assessing dementia, classification may be subject to measurement error. And while they adjusted for sociodemographic, health, and geographic covariates that could be associated with dementia and mortality risk, unmeasured confounding may have occurred.
Disclosures
This work was supported by a National Institute on Aging grant.
Stokes reported receiving grants from Ethicon Inc outside the submitted work. No other disclosures were reported.
Primary Source
JAMA Neurology
Stokes A, et al "Estimates of the Association of Dementia With US Mortality Levels Using Linked Survey and Mortality Records" JAMA Neurol 2020; DOI: 10.1001/jamaneurol.2020.2831.