Cognitive impairment in some US veterans may be due to treatable hepatic encephalopathy (HE) rather than dementia, new research suggested.
From 5%-10% of veterans diagnosed with dementia had possible undiagnosed cirrhosis, implicating HE as a contributor to cognitive impairment, found the study by Jasmohan S. Bajaj, MD, of Virginia Commonwealth University and Richmond VA Medical Center, Virginia, and colleagues.
The research was prompted, in part, by an earlier case study by Bajaj and colleagues that showed that two older men diagnosed with dementia and Parkinson's disease actually had HE, meaning their symptoms were due to advanced but treatable liver disease.
"Once they were properly diagnosed, whatever had been considered dementia was gone," Bajaj told Medscape Medical News. "The spouse of one man said, 'My husband is a different person now.' It's not that clinicians don't know how to treat HE; the problem was that they did not suspect it."
Among veterans with cirrhosis, concomitant dementia is common and is difficult to distinguish from HE, but the extent to which patients with dementia also have undiagnosed cirrhosis and HE is unknown, the authors of the current study wrote. "Undiagnosed cirrhosis among veterans with dementia could raise the possibility that part of their cognitive impairment may be due to reversible HE," they added.
To investigate, the researchers examined the prevalence and risk factors of undiagnosed cirrhosis — and therefore, possible HE — among US veterans.
The study was published online in JAMA Network Open.
Dementia or Cirrhosis?
Using the VHA Corporate Data Warehouse, researchers analyzed medical records of 177,422 US veterans diagnosed with dementia but not cirrhosis between 2009 and 2019 and with sufficient laboratory test results to calculate their Fibrosis-4 (FIB-4) scores. The mean age was 78.35 years, 97.1% were men, and 80.7% were White individuals.
The FIB-4 score for each patient was calculated using the most recent alanine aminotransferase (ALT) or aspartate aminotransferase (AST) levels or values and platelet values that were closest to the index date during the two years after the index dementia date.
Age is in the numerator of the FIB-4 score calculation; hence, higher age could lead to an erroneously high FIB-4 score, the authors noted. Therefore, for patients older than 65 years, the researchers entered 65 years as an input variable, rather than the actual age.
A FIB-4 score > 2.67 was suggestive of advanced fibrosis, whereas a score > 3.25 was suggestive of cirrhosis.
A total of 18,390 (10.3%) veterans had a FIB-4 score > 2.67, and 9373 (5.3%) had a FIB-4 score > 3.25.
In multivariable logistic regression models, a FIB-4 score > 3.25 was associated with older age (odds ratio [OR], 1.07), male sex (OR, 1.43), congestive heart failure (OR, 1.48), viral hepatitis (OR, 1.79), an Alcohol Use Disorders Identification Test score showing problem drinking (OR, 1.56), and chronic kidney disease (OR, 1.11).
In contrast, a FIB-4 score > 3.25 was inversely associated with the White race (OR, 0.79), diabetes (OR, 0.78), hyperlipidemia (OR, 0.84), stroke (OR, 0.85), tobacco use disorder (OR, 0.78), and rural residence (OR, 0.92).
Similar findings were associated with the FIB-4 greater than 2.67 threshold.
In a follow-up validation study among 89 veterans diagnosed with dementia at a single center, the researchers found similar results: 4.4%-11.2% of participants had high FIB-4 scores, suggestive of HE.
After investigating further, they concluded that 5% of patients in that cohort had reasons other than cirrhosis for their high FIB-4 scores. The remaining patients (95%) had evidence of cirrhosis, had risk factors, and/or had no other explanation for their high FIB-4 scores.
"The combination of high FIB-4 scores and other risk factors for liver disease in patients with dementia raises the possibility that reversible HE could be a factor associated with cognitive impairment," the authors wrote. "These findings highlight the potential to enhance cognitive function and quality of life by increasing awareness of risk factors and diagnostic indicators of advanced liver disease that may be associated with HE as a factor or as a differential diagnosis of dementia among clinicians other than liver specialists."
FIB-4 Screening Advised
"An elderly patient with cirrhosis used to be an oxymoron, because we never used to have people who lived this long or were diagnosed this late with cirrhosis," Bajaj told Medscape Medical News. "It's a good problem to have because people are now living longer, but it also means that we need to have every single person who is taking care of patients with what is deemed to be dementia know that the patient could also have an element of encephalopathy."
Increased awareness is important because, unlike dementia, encephalopathy is very easily treated, Bajaj said. "The biggest, easiest, correctable cause is to figure out if they have severe liver disease, and if that's the case, your friendly neighborhood gastroenterologist is waiting for you," he added.
The finding that cirrhosis was present in 95% of patients in the validation cohort is "very impressive, as they had excluded from the consideration all those with obvious cirrhosis before the FIB-4 was done," said William Carey, MD, acting hepatology section head in the Department of Gastroenterology, Hepatology and Nutrition at Cleveland Clinic's Digestive Disease Institute in Ohio. "This validates FIB-4 as a powerful tool for cirrhosis case-finding."
Ordering a FIB-4 "is within the skill set of every healthcare provider," Carey, who was not involved in the study, told Medscape Medical News. "Patients with altered mental status, including suspected or proven dementia, should be screened for possible cirrhosis, as future management will change. Those with elevated FIB-4 results should also be tested for possible HE and treated if it is present."
The study was partly funded by VA Merit Review grants to Bajaj. Bajaj reported receiving grants from Bausch, Grifols, Sequana, and Mallinckrodt outside the submitted work. Carey reported no relevant disclosures.