Key Points
Question
Could patients with dementia have undiagnosed cirrhosis and possibly missed treatable cognitive impairment?
Findings
In this cohort study including 177 422 veterans with a diagnosis of dementia, a high Fibrosis-4 score suggestive of advanced liver fibrosis to cirrhosis was observed in 5% to 10% of the patients who previously did not have a diagnosis of cirrhosis. These findings were supported by 2 validation cohorts within the Richmond VA Medical Center.
Meaning
The findings of this study suggest that clinicians treating patients with dementia should investigate a reversible factor associated with cognitive decline, including missed or undiagnosed cirrhosis with hepatic encephalopathy.
Abstract
Importance
Dementia and hepatic encephalopathy (HE) are challenging to distinguish clinically. Undiagnosed cirrhosis in a patient with dementia can lead to missed opportunities to treat HE.
Objective
To examine the prevalence and risk factors of undiagnosed cirrhosis and therefore possible HE in veterans with dementia.
Design, Setting, and Participants
A retrospective cohort study was conducted between 2009 and 2019 using data from the Veterans Health Administration (VHA) and 2 separate validation cohorts from the Richmond Veterans Affairs Medical Center. Data analysis was conducted from May 20 to October 15, 2023. Participants included 177 422 US veterans with a diagnosis of dementia at 2 or more clinic visits, no prior diagnosis of cirrhosis, and with sufficient laboratory test results to calculate the Fibrosis-4 (FIB-4) score.
Exposures
Demographic and clinical characteristics.
Main Outcomes and Measures
An FIB-4 score (>2.67 suggestive of advanced fibrosis and >3.25 suggestive of cirrhosis), capped at age 65 years even for those above this cutoff who were included in the analysis.
Results
Among 177 422 veterans (97.1% men; 80.7% White; mean (SD) age, 78.35 [10.97] years) 5.3% (n = 9373) had an FIB-4 score greater than 3.25 and 10.3% (n = 18 390) had an FIB-4 score greater than 2.67. In multivariable logistic regression models, FIB-4 greater than 3.25 was associated with older age (odds ratio [OR], 1.07; 95% CI, 1.06-1.09), male gender (OR, 1.43; 95% CI, 1.26-1.61), congestive heart failure (OR, 1.48; 95% CI, 1.43-1.54), viral hepatitis (OR, 1.79; 95% CI, 1.66-1.91), Alcohol Use Disorders Identification Test score (OR, 1.56; 95% CI, 1.44-1.68), and chronic kidney disease (OR, 1.11; 95% CI, 1.04-1.17), and inversely associated with White race (OR, 0.79; 95% CI, 0.73-0.85), diabetes (OR, 0.78; 95% CI, 0.73-0.84), hyperlipidemia (OR, 0.84; 95% CI, 0.79-0.89), stroke (OR, 0.85; 95% CI, 0.79-0.91), tobacco use disorder (OR, 0.78; 95% CI, 0.70-0.87), and rural residence (OR, 0.92; 95% CI, 0.87-0.97). Similar findings were associated with the FIB-4 greater than 2.67 threshold. These codes were associated with cirrhosis on local validation. A local validation cohort of patients with dementia showed a similar percentage of high FIB-4 scores (4.4%-11.2%).
Conclusions and Relevance
The findings of this cohort study suggest that clinicians encountering patients with dementia should be encouraged to screen for cirrhosis using the FIB-4 score to uncover reversible factors associated with cognitive impairment, such as HE, to enhance outcomes.
This cohort study examines the presence of undiagnosed cirrhosis and hepatic encephalopathy in US veterans with dementia.
Introduction
Dementia poses a major clinical, psychosocial, and financial burden worldwide,1,2 with a suboptimal therapeutic landscape. Metabolic encephalopathies, such as hepatic encephalopathy (HE), are reversible with therapy, although their diagnosis requires a high index of suspicion.3 Major changes in the epidemiologic factors of liver disease are attributed to widespread advances for viral hepatitis care with increasing rates of metabolic syndrome and alcohol use disorder (AUD).4 However, a substantial proportion of advanced liver disease remains undiagnosed, especially among US veterans, who have major risk factors for both cirrhosis and dementia.5,6
Many veterans receiving Veterans Administration care are aging and often have comorbid conditions that affect cognitive function, such as AUD, posttraumatic stress disorder (PTSD), and head injuries, in addition to risk factors for cirrhosis.7,8 Among veterans with cirrhosis, concomitant dementia is common and is challenging to distinguish from HE, but the extent to which patients with diagnoses of dementia also have undiagnosed cirrhosis and HE is unknown.9 Undiagnosed cirrhosis among veterans with dementia could raise the possibility that part of their cognitive impairment may be due to reversible HE.10 Therefore, the determinants of undiagnosed cirrhosis among veterans with dementia is important to identify those eligible for screening and subsequent HE therapy.
Our aims were to examine in a national cohort of US veterans with dementia without previously diagnosed cirrhosis the prevalence of Fibrosis-4 (FIB-4) score greater than 3.25 suggestive of undiagnosed cirrhosis and FIB-4 greater than 2.67 suggestive of advanced liver fibrosis, identify characteristics related to this diagnosis, and validate these findings in 2 cohorts from the Richmond Veterans Affairs Medical Center.
Methods
Data Collection
Using the VHA Corporate Data Warehouse (CDW), we extracted information on outpatient or inpatient diagnoses of dementia from 2009 to 2019. We limited our study analysis up to 2019 to minimize the confounding effect of the COVID-19 pandemic. This cohort study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline and was approved by the Richmond VAMC IRB. A waiver of informed consent was granted because this was deidentified database research.
We used validated International Classification of Diseases, 9th Revision (ICD-9) and International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD-10) codes to identify patients with 2 or more dementia diagnoses at different dates.9 We then collected the index date, considered to be the second date of diagnosis, and removed patients with evidence of an outpatient or inpatient cirrhosis diagnosis or complication from cirrhosis using validated ICD-9 and ICD-10 codes, and/or lactulose or rifaximin prescriptions during the study period (eAppendix in Supplement 1).11 Patients without alanine aminotransferase (ALT) or aspartate aminotransferase (AST) levels or values over 250 U/L (to convert to microkatals per liter, multiply by 0.0167), as well as patients without platelet counts or with abnormally low platelet counts (≤50 × 109/μL; to convert to 109 per liter, multiply by 1), were excluded.
We calculated the FIB-4 score for each patient using the most recent AST, ALT, and platelet values that were closest to the index date during the 2 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.12 Due to this limitation of the FIB-4 score for any patients older than 65 years, we entered 65 years as an input variable (rather than their actual age) to calculate the FIB-4 score. Therefore, we did not exclude any patient solely due to age. We collected demographic data, including age, US region (reference level, North Atlantic) of VHA, residence status (urban vs rural), sex, body mass index, race and ethnicity, and comorbidities (details and codes the eAppendix in Supplement 1). Along with other parameters, data on race and ethnicity could be helpful in focusing which patient groups are likely to have a higher proportion of undiagnosed cirrhosis. Alcohol Use Disorders Identification Test (AUDIT-C) scores were obtained, and problem drinking was defined by a score of greater than 4 in men or greater than 3 in women per VHA guidance.13 All comorbidities were collected up to 2 years before the index diagnosis date of dementia; FIB-4 scores greater than 3.25 and greater than 2.67 were calculated, and cohorts above or below these cutoffs were compared.
For validation, we extracted data on Richmond VAMC patients included in this cohort to determine the reasons for high FIB-4 scores. Two clinicians (J.S.B. and G.K.) reviewed each medical record in detail. A second cohort of patients with dementia seen by the geriatrics clinic at the Richmond VAMC between January 1 and June 30, 2023, who were still alive underwent a detailed medical records review to examine the presence of high FIB-4 scores and potential causes.
Statistical Analysis
Data analysis was performed from May 20 to October 15, 2023. Continuous variables are expressed as the mean (SD). The Charlson Comorbidity Index (CCI) score is expressed as the median (IQR), due to both the discrete nature and skewed distribution. Categorical variables are expressed as counts and percentages of the total. We used 2-sample t tests, Pearson χ2 tests, or the Wilcoxon rank sum test as appropriate to compare characteristics between the groups.
Statistical significance was defined as P < .05 for all hypothesis tests conducted, and all tests were 2-sided. RStudio, version 4.1.2 (R Foundation for Statistical Computing) was used for all statistical analysis.
All characteristics that were significantly different between the high FIB-4 and low FIB-4 groups (≥/<3.25 and≥/<2.67) in univariable models were included in multivariable logistic regression models. Age was capped at 65 years within these analyses, although individuals older than 65 years were included.
Results
CDW Cohort Description
We identified 287 257 veterans with dementia, of whom 38% were ineligible, resulting in 177 422 patients in the cohort (Figure). The major reasons for exclusion were lack of laboratory test values (AST, ALT, and platelet counts; n = 75 368) for FIB-4 calculation, diagnosed cirrhosis (n = 33 251), and abnormal laboratory test values (platelet count ≤50 × 103/μL [n = 383], AST≥250 U/L [n = 403], and ALT≥250 U/L [n = 430]). Their mean (SD) age was 78.35 (10.97) years, 97.1% were male, 2.9% were female, 80.7% were White, 5.3% (n = 9373) had an FIB-4 score greater than 3.25, and 10.3% (n = 18 390) had an FIB-4 score greater than 2.67. Patients with an FIB-4 score greater than 3.25 were more likely to be older, male, Hispanic, and reside in the Southeast; patients with FIB-4 greater than 3.25 were less likely to be White, live in rural areas, and less likely to be from the Midwest (Table 1). Patients with a high FIB-4 score had a higher CCI score for comorbid conditions, the most common of which were chronic kidney disease (CKD) and cardiovascular conditions. Veterans with FIB-4 scores greater than 3.25 also had lower percentages of tobacco use disorder, sleep apnea, PTSD, head injury, and major depressive disorder. While lower in proportion than these comorbid conditions, patients with high FIB-4 scores had higher AUDIT-C scores, rates of prior hepatitis B and C, and HIV infection. Homelessness was not different between groups, while patients with high FIB-4 scores had a significantly lower mean body mass index than the rest. Similar results were seen using the greater than FIB-4 2.67 cutoff (Table 2).
Figure. Flowchart of Patients With Dementia Diagnosis and Fibrosis-4 (FIB-4) Scores Across the Corporate Data Warehouse (CDW) and 2 Validation Cohorts.
aUsing FIB-4 >3.25 threshold, 9373 (5.3%) had high FIB-4.
Table 1. Characteristics of Patients With FIB-4 Level Greater Than 3.25 vs Patients With FIB-4 Level 3.25 or Less.
| Variable | No. (%) | P value | |
|---|---|---|---|
| FIB-4 > 3.25 | FIB-4 ≤ 3.25 | ||
| Patients | 9373 (5.3) | 168 049 (94.7) | NA |
| Age, mean (SD), y | 80.78 (9.65) | 78.21 (11.02) | <.001 |
| Sex | |||
| Male | 9211 (98.3) | 163 148 (97.1) | <.001 |
| Female | 162 (1.7) | 4901 (2.9) | |
| Race (White) | 6726 (75.2) | 129 266 (80.4) | <.001 |
| Ethnicity (Hispanic or Latino) | 654 (7.3) | 10 872 (6.7) | .04 |
| BMI, mean (SD) | 25.78 (5.28) | 26.73 (5.40) | <.001 |
| Charlson Comorbidity Index, median (IQR) | 3.00 (2.00-5.00) | 3.00 (2.00-5.00) | <.001 |
| Diabetes | 3471 (37.0) | 65 296 (38.9) | <.001 |
| Hypertension | 7511 (80.1) | 128 989 (76.8) | <.001 |
| Chronic kidney disease | 2584 (27.6) | 34 529 (20.5) | <.001 |
| Hyperlipidemia | 6007 (64.1) | 111 923 (66.6) | <.001 |
| Posttraumatic stress disorder | 1285 (13.7) | 26 220 (15.6) | <.001 |
| Major depressive disorder | 2237 (23.9) | 44 618 (26.6) | <.001 |
| HIV | 66 (0.7) | 683 (0.4) | <.001 |
| Hepatitis B or C | 377 (4.0) | 3353 (2.0) | <.001 |
| Head injury | 256 (2.7) | 6027 (3.6) | <.001 |
| Problem drinking using AUDIT-C | 358 (4.1) | 4690 (3.0) | <.001 |
| Tobacco use disorder | 827 (8.8) | 15 884 (9.5) | .04 |
| Cerebrovascular disease or stroke | 2142 (22.9) | 36 923 (22.0) | .04 |
| Peripheral vascular disease | 1963 (20.9) | 29 094 (17.3) | <.001 |
| Congestive heart failure | 3278 (35.0) | 42 119 (25.1) | <.001 |
| Sleep apnea | 1221 (13.0) | 24 688 (14.7) | <.001 |
| Rural residencea | 2601 (28.4) | 53 387 (32.4) | <.001 |
| Homelessness | 585 (6.2) | 10 491 (6.2) | >.99 |
| Region | |||
| Midwest | 1779 (19.0) | 37 975 (22.6) | <.001 |
| Continental | 1533 (16.4) | 28 443 (16.9) | |
| North Atlantic | 2156 (23.0) | 37 809 (22.6) | |
| Southeast | 2392 (25.5) | 39 236 (23.3) | |
| Pacific | 1513 (16.1) | 24 586 (14.6) | |
Abbreviations: AUDIT-C, Alcohol Use Disorders Identification Test; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); FIB-4, Fibrosis-4; NA, not applicable.
Contained missing values. Only complete cases considered.
Table 2. Characteristics of Patients With FIB-4 Level Greater Than 2.67 vs Patients With FIB-4 Level 2.67 or Less.
| Variable | No. (%) | P value | |
|---|---|---|---|
| FIB-4 > 2.67 | FIB-4 ≤ 2.67 | ||
| Patients | 18 390 (10.4) | 159 032 (89.6) | |
| Age, mean (SD), y | 81.04 (9.49) | 78.04 (11.09) | <.001 |
| Sex | |||
| Male | 18 085 (98.3) | 154 274 (97.0) | <.001 |
| Female | 305 (1.7) | 4758 (3.0) | |
| Race (White)a | 13 401 (76.3) | 122 591 (80.6) | <.001 |
| Ethnicity (Hispanic or Latino)a | 1278 (7.2) | 10 248 (6.7) | .007 |
| BMI, mean (SD) | 25.87 (5.34) | 26.78 (5.40) | <.001 |
| Charlson Comorbidity Index, median (IQR) | 3.00 (2.00-5.00) | 3.00 (2.00-5.00) | <.001 |
| Diabetes | 6605 (35.9) | 62 162 (39.1) | <.001 |
| Hypertension | 14 519 (79.0) | 121 981 (76.7) | <.001 |
| Chronic kidney disease | 4796 (26.1) | 32 317 (20.3) | <.001 |
| Hyperlipidemia | 11 915 (64.8) | 106 015 (66.6) | <.001 |
| Posttraumatic stress disorder | 2487 (13.5) | 25 018 (15.7) | <.001 |
| Major depressive disorder | 4350 (23.7) | 42 505 (26.7) | <.001 |
| HIV | 118 (0.6) | 631 (0.4) | <.001 |
| Hepatitis B or C | 644 (3.5) | 3086 (1.9) | <.001 |
| Head injury | 475 (2.6) | 5808 (3.7) | <.001 |
| Problem drinking using AUDIT-C | 620 (3.7) | 4428 (3.0) | <.001 |
| Tobacco use disorder | 1520 (8.3) | 15191 (9.6) | <.001 |
| Cerebrovascular disease or stroke | 4157 (22.6) | 34 908 (22.0) | .04 |
| Peripheral vascular disease | 3721 (20.2) | 27 336 (17.2) | <.001 |
| Congestive heart failure | 6244 (34.0) | 39 153 (24.6) | <.001 |
| Sleep apnea | 2377 (12.9) | 23 532 (14.8) | <.001 |
| Rural residenceb | 5210 (29.0) | 50 778 (32.5) | <.001 |
| Homelessness | 1093 (5.9) | 9983 (6.3) | .08 |
| Region | |||
| Midwest | 3395 (18.5) | 36 359 (22.9) | <.001 |
| Continental | 3079 (16.7) | 26 897 (16.9) | |
| North Atlantic | 4264 (23.2) | 35 701 (22.4) | |
| Southeast | 4719 (25.7) | 36 909 (23.2) | |
| Pacific | 2933 (15.9) | 23 166 (14.6) | |
Abbreviations: AUDIT-C, Alcohol Use Disorders Identification Test; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); FIB-4, Fibrosis-4; NA, not applicable.
Race and ethnicity categories reported as in the Corporate Data Warehouse.
Contained missing values. Only complete cases considered.
Multivariable Analysis: Risk Factors for High FIB-4 Scores
In the FIB-4 greater than 3.25 model, all variables remained significant except for head injury, major depression, PTSD, ethnicity, HIV, peripheral vascular disease, and sleep apnea (Table 3). Among the categorical variables, history of hepatitis B and C was the risk factor with the highest odds for a high FIB-4 score (OR, 1.79; 95% CI, 1.66-1.91), score followed by AUDIT-C (OR, 1.56; 95% CI, 1.44-1.68), congestive heart failure (OR, 1.48; 95% CI, 1.43-1.54), male gender (OR, 1.43; 95% CI, 1.26-1.61), and chronic kidney disease (OR, 1.11; 95% CI, 1.04-1.17). Diabetes was the categorical variable with the lowest odds (OR, 0.78; 95% CI, 0.73-0.84) with a high FIB-4 score, followed by tobacco use disorder (OR, 0.78; 95% CI, 0.70-0.87), White race (OR, 0.79; 95% CI, 0.73-0.85), living in the Midwestern vs the North Atlantic region (OR, 0.84; 95% CI, 0.77-0.91), hyperlipidemia (OR, 0.84; 95% CI, 0.79-0.89), stroke (OR, 0.85; 95% CI, 0.79-0.91), and rural residence (OR, 0.92; 95% CI, 0.87-0.97). Among continuous variables, an FIB-4 score greater than 3.25; for a 1-unit increase in age and in CCI, the odds of a high FIB-4 score both independently increased by 1.07, while a 1-unit increase in body mass index decreased the odds of FIB-4 greater than 3.25 by 0.97. Using a high FIB-4 cutoff score of 2.67, findings were similar for the variables. When using this cutoff value, hypertension was no longer a significant risk factor within the multivariable model, and peripheral vascular disease and PTSD became significant.
Table 3. Logistic Regression for the Overall Cohort With Dementia.
| Variable | FIB-4 score >3.25 | FIB-4 score >2.67 | ||
|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | |
| Age (capped at 65 y) | 1.07 (1.06-1.09) | <.001 | 1.09 (1.08-1.10) | <.001 |
| Race, White (vs all others) | 0.79 (0.73-0.85) | <.001 | 0.80 (0.76-0.85) | <.001 |
| Ethnicity – Hispanic (vs all others) | 1.08 (0.99-1.17) | .12 | 1.04 (0.97-1.11) | .28 |
| Rural residence (vs urban) | 0.92 (0.87-0.97) | .002 | 0.93 (0.90-0.97) | .001 |
| Gender, male | 1.43 (1.26-1.61) | <.001 | 1.51 (1.38-1.64) | <.001 |
| Body mass index | 0.97 (0.97-0.98) | <.001 | 0.98 (0.97-0.98) | <.001 |
| Co-morbidities | ||||
| Charlson Comorbidity Index | 1.07 (1.06-1.09) | <.001 | 1.04 (1.03-1.05) | <.001 |
| Chronic kidney disease | 1.11 (1.04-1.17) | .002 | 1.12 (1.07-1.17) | <.001 |
| Hypertension | 1.10 (1.04-1.17) | .003 | 1.03 (0.98-1.08) | .22 |
| Diabetes | 0.78 (0.73-0.84) | <.001 | 0.77 (0.73-0.81) | <.001 |
| Hyperlipidemia | 0.84 (0.79-0.89) | <.001 | 0.89 (0.85-0.93) | <.001 |
| Stroke | 0.85 (0.79-0.91) | <.001 | 0.88 (0.83-0.92) | <.001 |
| Peripheral vascular disease | 1.04 (0.98-1.10) | .19 | 1.07 (1.07-1.12) | .004 |
| Sleep apnea | 0.99 (0.92-1.06) | .83 | 0.98 (0.92-1.03) | .42 |
| Head injury | 0.95 (0.80-1.09) | .44 | 0.92 (0.81-1.03) | .13 |
| Congestive heart failure | 1.48 (1.43-1.54) | <.001 | 1.52 (1.48-1.56) | <.001 |
| Hepatitis B or C | 1.79 (1.66-1.91) | <.001 | 1.79 (1.69-1.89) | <.001 |
| Problem drinking, AUDIT-C | 1.56 (1.44-1.68) | <.001 | 1.41 (1.32-1.50) | <.001 |
| Tobacco use disorder | 0.78 (0.70-0.87) | <.001 | 0.77 (0.71-0.84) | <.001 |
| HIV | 0.87 (0.56-1.18) | .39 | 1.11 (0.87-1.35) | .39 |
| Major depressive disorder | 0.96 (0.91-1.02) | .20 | 0.96 (0.92-1.00) | .06 |
| Posttraumatic stress disorder | 0.93 (0.86-1.00) | .05 | 0.92 (0.86-0.97) | .001 |
| Regions (vs Northeast) | ||||
| Continental | 0.99 (0.92-1.07) | .81 | 1.00 (0.95-1.06) | .92 |
| Midwest | 0.84 (0.77-0.91) | <.001 | 0.79 (0.74-0.84) | <.001 |
| Pacific | 1.07 (0.99-1.15) | .09 | 1.04 (1.99-1.10) | .14 |
| Southeast | 1.07 (1.00-1.14) | .05 | 1.09 (1.04-1.14) | .001 |
Abbreviations: AUDIT-C, Alcohol Use Disorders Identification Test; FIB-4, Fibrosis-4.
Medical Records Review of All Richmond VAMC Patients From 2009 to 2019
Eighty-nine patients of the cohort were from the Richmond VAMC. The mean (SD) FIB-4 score was 5.12 (2.41), with a platelet count of 124.3 (54.5) × 103/μL; ALT level, 33.0 (24.6) U/L; AST level, 61.5 (41.9) U/L; and true age, 81.2 (10.0) years. Obesity was found in 20 patients (22.4%), diabetes in 27 patients (30.3%), and AUD in 13 patients (14.6%). Hepatitis C status was checked in only 59 patients (66.3%); of these, a positive hepatitis C virus antibody was found in 2 patients (3.3%); 4 of 42 patients (9.5%) with hepatitis B testing were positive.
Thrombocytopenia was found in 76 patients (85.4%) (mean [SD] platelet count, 117.4 [71.3]), 2 of whom had myelodysplastic syndrome. In the remaining 74 patients, an AST to ALT ratio greater than 1 was also seen in 57 patients (64.0%), and hypoalbuminemia was observed in 24 patients (26.9%), suggesting potential advanced liver disease. Liver stiffness was identified in 5 additional patients (3 with ultrasonography and 2 with magnetic resonance imaging), which showed greater than F3 stiffness in all. Imaging showed cirrhosis in 10 patients (5 with nodular liver, 4 with ascites, and 1 with varices). Three additional patients had potential HE but were not treated for it and none underwent an esophagogastroduodenoscopy. Two patients without thrombocytopenia had a liver biopsy that showed no or mild fibrosis. Of the remaining 17 patients with a high FIB-4 score, medical records review no evidence to rule out cirrhosis; all had potential cirrhosis risk factors: 8 patients with AUD, obesity, and type 2 diabetes; 7 with obesity and diabetes; and 2 with AUD only.
Ultimately, we concluded that the definite reasons other than cirrhosis in patients with high FIB-4 scores were found in only 4 patients (2 with liver biopsy and 2 with myelodysplastic syndrome: 5%). The remaining patients (95%) either had evidence of cirrhosis, had risk factors, and/or had no other explanation for their high FIB-4 score.
Review of Patients With Dementia in Richmond VAMC
Between January 1 and June 30, 2023, 80 veterans were seen in the geriatric clinics for evaluation of dementia. Using the age-adjusted FIB-4 score (for patients aged >65 years, age was entered as 65 years to calculate the score), 9 patients (11.2%) had an FIB-4 score greater than 2.67 with a mean (SD) of 3.78 (0.67). These patients had a mean platelet count of 129.9 (47.8) × 103/μL; ALT level, 26.7 (13.2) U/L; AST level, 33.3 (11.8) U/L; and age, 79.0 (7.7) years. A local validation cohort of patients with dementia showed a similar percentage of high FIB-4 scores (4.4%-11.2%). Of these 9 patients, 4 had diabetes, 3 had prior AUD, none were obese, 5 were checked for hepatitis C virus antibodies (1 was positive), and 1 already had been diagnosed with cirrhosis and was monitored by the hepatology department. The remaining 8 patients were further evaluated; 1 more had an alternative explanation (myelodysplastic syndrome) for thrombocytopenia. Potential cirrhosis therefore was possibly present in 7 patients (8.7%) (2 patients only had 1 visit with the high FIB-4 score, 5 had high FIB-4 scores on earlier visits along with thrombocytopenia. The geriatricians were encouraged to seek gastroenterology/hepatology investigation.
Discussion
In a national cohort study of veterans diagnosed with dementia who were not already diagnosed with cirrhosis, we found high FIB-4 scores in 10.3% of patients. The prevalence of potentially undiagnosed cirrhosis, represented by a high FIB-4 score, was higher in non-White and Hispanic vs White patients, urban vs rural veterans, patients with more comorbidities, and patients with AUD and viral hepatitis. 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. 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.
Cirrhosis, HE, and dementia are prevalent in the rapidly increasing older veteran population, creating the potential for overlap and diagnostic confusion.9,10 In older patients with risk factors for both cirrhosis and dementia, the lack of clinically applicable biomarkers and diagnostic tests complicates accurate diagnoses and treatment.10,14 Therefore, to diagnose or exclude HE as a potential factor associated with cognitive impairment, a strong suspicion for underlying cirrhosis is needed. There is a growing awareness of cirrhosis among primary care physicians, where FIB-4 is being promoted as a tool for screening.15,16 The validity of these tools has been established in earlier studies, but a higher prevalence of traditional liver-related risk factors, such as AUD and hepatitis B and C, in the high FIB-4 score groups, and the relative similarity of comparison of patient characteristics between the 2 FIB-4 score thresholds increases confidence in the results. While AUD and viral hepatitis are well known to be associated with advanced liver disease,17 these findings should encourage health care professionals taking care of patients with dementia, including primary care, urgent care clinicians, geriatricians, and neurologists, to seek these out and diagnose cirrhosis as a potential factor associated with cognitive dysfunction.
The factors associated with high FIB-4 scores in this population differed in important ways from previous studies of the general population. While some traditional risk factors, such as AUD and viral hepatitis, were associated with high FIB-4 scores, others (eg, diabetes and hyperlipidemia) were not. Although the reasons are unclear, diabetes and hyperlipidemia are associated with vascular dementia, independent of liver disease.18,19,20 Depression and PTSD, which are known to be occur more often among people with cirrhosis than the general population, were not associated with high FIB-4 scores in this analysis. This apparent lack of association may have been due to controlling for AUD, which may mediate the associations between PTSD, depression, and cirrhosis.21,22,23 Alternatively, dementia may be so often associated with mental health symptoms that the association between mental health symptoms and potential liver disease was not identified in this enriched cohort. While hepatitis C virus and AUD diagnoses were expectedly associated with high FIB-4 scores in this population, metabolic and other comorbidities that have been associated with cirrhosis in other populations24 were not associated with high FIB-4, likely because our population was restricted to older patients with dementia whose baseline burden of these diseases differs from that of the general population. Because the factors associated with high FIB-4 scores in the general population may differ in the context of dementia, a higher index of suspicion may be needed to ensure proper diagnosis of liver disease in this complex population. Additionally, in a large sample size, associations with small or clinically irrelevant absolute differences between the groups may still be statistically significant. However, to our knowledge, this study is among the first of its kind to address these potential risk factors; studies are needed to evaluate these risk factors further.
We used the FIB-4 index as a marker of potential liver disease in this study. The FIB-4 index is a scalable, inexpensive, and useful approach to identify patients with possible advanced liver fibrosis or cirrhosis.12 We used several validated methods to exclude previously diagnosed cirrhosis in this population.9 However, the large number of people with diagnosed dementia and diagnosed cirrhosis represent another population of interest; these have challenges in comanagement because HE and dementia overlap.9 The use of FIB-4 scores necessitated exclusion of patients who did not have liver enzyme levels available, and few were excluded for extremes of laboratory test values to avoid confounding. Despite these exclusions, we found that 5% to 10% of veterans with dementia had potential undiagnosed cirrhosis. We found similar patterns with FIB-scores 4 greater than 2.67 or greater than 3.25 when comorbid conditions, as well as sociodemographic characteristics, were compared between patients with and without low or high FIB-4 scores. While the FIB-4 is certainly not a perfect index, it is widely adopted to exclude the presence of advanced liver disease in the general population due to its ease of calculation and interpretation.25,26 The FIB-4 score also adds specificity to the alternative of using AST or ALT levels alone, since patients with even advanced liver disease often have laboratory values within the reference ranges.27
A particular challenge when studying FIB-4 scores in patients with dementia is the outsized role of older age, which can falsely increase this score. McPherson et al28 proposed using actual age with a cutoff level greater than 2.0 in metabolic dysfunction-associated steatotic liver disease fibrosis; a subsequent publication used this modification for people older than 65 years.29 However, the ages of patients included in these studies spanned both sides of 65 years, unlike our study where the mean age was 78.35 years. Thus, if these rules were adapted in our cohort of primarily older veterans, a potential overinflation of the FIB-4 score leading to numerous false-positive results may occur. To combat the potential confounding factor of age, we capped patient age at 65 years in the FIB-4 calculation, even though all patients were included. Combined with our high threshold choices of 2.67 and 3.25 for probable advanced fibrosis, this ensured that scores would be conservative and decreased the risk of false-positive results or biased to the null. The main drawback would be neglecting the actual effect of age in patients older than 65 years. In general, research and validation of the FIB-4 index in an older population is limited, and further studies are needed to address this gap. Regardless, the presence of cirrhosis among patients with high FIB-4 scores in this cohort was largely corroborated by medical records review in 2 validation cohorts.
The disparity in potential undiagnosed cirrhosis in veterans with dementia who lived in urban areas, were of Hispanic ethnicity, and were not White, is an important issue. These racial and geographic disparities echo a prior study.30 Dementia disproportionately affects Black and Hispanic veterans and is diagnosed later in the disease course in these populations, which has been attributed to lack of high-quality health care access for minoritized populations.31,32,33 A lack of access to health care services could also explain potentially underdiagnosed cirrhosis, which reiterates the need for focus on these subpopulations to accurately diagnose cirrhosis and potentially HE.34,35,36
We focused on veterans with dementia with a high FIB-4 score due to the important therapeutic implications of a missed cirrhosis diagnosis and, therefore, a missed HE diagnosis. An earlier granular study showed that patients with combined amnestic (related to dementia) and anamnestic (related to cirrhosis) cognitive dysfunction in cirrhosis have a worse quality of life, which improves after HE therapy.14 Moreover, there is a major overlap between HE and dementia in veterans with diagnosed cirrhosis.9 Even if the ultimate diagnosis of HE is not considered to be a factor associated with the cognitive dysfunction in those diagnosed with cirrhosis, hepatocellular cancer and other cirrhosis complications also need to be treated or monitored.37
Strengths and Limitations
The advantages of the CDW are the national scope, potential to study geographic variations, and the diverse population of veterans whose primary source of health care is the VHA system. The limitations are the use of coding and relative lack of granularity in the cirrhosis and dementia diagnoses. Also, the relative proportion of HE in those with undiagnosed cirrhosis, and the potential HE treatment needs to be addressed, which can only be done by an individual patient assessment. Therefore, we further analyzed this with a detailed medical records review of earlier (CDW) and current dementia cohorts. This study had the unintended positive consequence of increasing the awareness of collaborating geriatricians, who were advised to refer the patients to hepatology services. However, more work is needed to increase recognition of cirrhosis and optimize care for people with a known overlap between cirrhosis and dementia.
Conclusions
The findings of this cohort study suggest that, in a national veterans cohort of patients with dementia, 5% to 10% of the patients have laboratory values suggestive of possible undiagnosed cirrhosis that could implicate HE as a contributor to overall cognitive impairment. Patients in this study with possibly undiagnosed cirrhosis were more likely to be of races other than White, Hispanic, and urban-dwelling, and more likely to have AUD and viral hepatitis. These percentages of affected patients were corroborated with medical records review of 2 separate validation cohorts. In conclusion, FIB-4 should be considered as a screening tool to detect cirrhosis and potential HE in older veterans with dementia. Those with high scores (eg, >2.67) should be considered for further evaluation and treatment.
eAppendix. ICD Codes Used
Data Sharing Statement
References
- 1.Collaborators GBDDF; GBD 2019 Dementia Forecasting Collaborators . Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: an analysis for the Global Burden of Disease Study 2019. Lancet Public Health. 2022;7(2):e105-e125. doi: 10.1016/S2468-2667(21)00249-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.World Health Organization. Dementia. March 15, 2023. Accessed December 14, 2023. https://www.who.int/news-room/fact-sheets/detail/dementia
- 3.Vilstrup H, Amodio P, Bajaj J, et al. Hepatic encephalopathy in chronic liver disease: 2014 practice guideline by the American Association for the Study of Liver Diseases and the European Association for the Study of the Liver. Hepatology. 2014;60(2):715-735. doi: 10.1002/hep.27210 [DOI] [PubMed] [Google Scholar]
- 4.Kim IH, Kisseleva T, Brenner DA. Aging and liver disease. Curr Opin Gastroenterol. 2015;31(3):184-191. doi: 10.1097/MOG.0000000000000176 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Walker M, El-Serag HB, Sada Y, et al. Cirrhosis is under-recognised in patients subsequently diagnosed with hepatocellular cancer. Aliment Pharmacol Ther. 2016;43(5):621-630. doi: 10.1111/apt.13505 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Serper M, Kaplan DE, Shults J, et al. Quality measures, all-cause mortality, and health care use in a national cohort of veterans with cirrhosis. Hepatology. 2019;70(6):2062-2074. doi: 10.1002/hep.30779 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.O’Malley KA, Vinson L, Pless Kaiser A, Sager Z, Hinrichs K. Mental health and aging veterans: how the Veterans Health Administration meets the needs of aging veterans. Public Policy Aging Rep. 2020;30(1):19-23. doi: 10.1093/ppar/prz027 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Karel MJ, Wray LO, Adler G, et al. Mental health needs of aging veterans: recent evidence and clinical recommendations. Clin Gerontol. 2022;45(2):252-271. doi: 10.1080/07317115.2020.1716910 [DOI] [PubMed] [Google Scholar]
- 9.Adejumo A, Noll A, Rogal SS, et al. Dementia frequently coexists with hepatic encephalopathy but not other cirrhosis complications in US veterans. Am J Gastroenterol. 2023;118(3):475-480. doi: 10.14309/ajg.0000000000002189 [DOI] [PubMed] [Google Scholar]
- 10.Bajaj JS, Gentili A, Wade JB, Godschalk M. Specific challenges in geriatric cirrhosis and hepatic encephalopathy. Clin Gastroenterol Hepatol. 2022;20(8S):S20-S29. doi: 10.1016/j.cgh.2022.04.035 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Beste LA, Leipertz SL, Green PK, Dominitz JA, Ross D, Ioannou GN. Trends in burden of cirrhosis and hepatocellular carcinoma by underlying liver disease in US veterans, 2001-2013. Gastroenterology. 2015;149(6):1471-1482.e5. doi: 10.1053/j.gastro.2015.07.056 [DOI] [PubMed] [Google Scholar]
- 12.Sterling RK, Lissen E, Clumeck N, et al. ; APRICOT Clinical Investigators . Development of a simple noninvasive index to predict significant fibrosis in patients with HIV/HCV coinfection. Hepatology. 2006;43(6):1317-1325. doi: 10.1002/hep.21178 [DOI] [PubMed] [Google Scholar]
- 13.US Dept of Veterans Affairs. Viral hepatitis and liver disease: alcohol use disorders identification test. September 11, 2019. Accessed December 14, 2023. https://www.hepatitis.va.gov/alcohol/treatment/audit-c.asp
- 14.Bajaj JS, Duarte-Rojo A, Xie JJ, et al. Minimal Hepatic encephalopathy and mild cognitive impairment worsen quality of life in elderly patients with cirrhosis. Clin Gastroenterol Hepatol. 2020;18(13):3008-3016.e2. doi: 10.1016/j.cgh.2020.03.033 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Halfon P, Ansaldi C, Penaranda G, et al. Prospective screening of liver fibrosis in a primary care cohort using systematic calculation of FIB-4 in routine results. PLoS One. 2021;16(7):e0254939. doi: 10.1371/journal.pone.0254939 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Davyduke T, Tandon P, Al-Karaghouli M, Abraldes JG, Ma MM. Impact of Implementing a “FIB-4 first” strategy on a pathway for patients with NAFLD referred from primary care. Hepatol Commun. 2019;3(10):1322-1333. doi: 10.1002/hep4.1411 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Huang DQ, Terrault NA, Tacke F, et al. Global epidemiology of cirrhosis - aetiology, trends and predictions. Nat Rev Gastroenterol Hepatol. 2023;20(6):388-398. doi: 10.1038/s41575-023-00759-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Huang H, Liu Z, Xie J, Xu C. NAFLD does not increase the risk of incident dementia: a prospective study and meta-analysis. J Psychiatr Res. 2023;161:435-440. doi: 10.1016/j.jpsychires.2023.03.041 [DOI] [PubMed] [Google Scholar]
- 19.Wang L, Sang B, Zheng Z. Risk of dementia or cognitive impairment in non-alcoholic fatty liver disease: a systematic review and meta-analysis. Front Aging Neurosci. 2022;14:985109. doi: 10.3389/fnagi.2022.985109 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Thrift AP, Nguyen TH, Pham C, et al. The prevalence and determinants of NAFLD and MAFLD and their severity in the VA primary care setting. Clin Gastroenterol Hepatol. 2023;21(5):1252-1260.e5. doi: 10.1016/j.cgh.2022.05.046 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Dworkin ER, Bergman HE, Walton TO, Walker DD, Kaysen DL. Co-occurring post-traumatic stress disorder and alcohol use disorder in US military and veteran populations. Alcohol Res. 2018;39(2):161-169. [PMC free article] [PubMed] [Google Scholar]
- 22.Teeters JB, Lancaster CL, Brown DG, Back SE. Substance use disorders in military veterans: prevalence and treatment challenges. Subst Abuse Rehabil. 2017;8:69-77. doi: 10.2147/SAR.S116720 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Hasin DS, Grant BF. Major depression in 6050 former drinkers: association with past alcohol dependence. Arch Gen Psychiatry. 2002;59(9):794-800. doi: 10.1001/archpsyc.59.9.794 [DOI] [PubMed] [Google Scholar]
- 24.Lazarus JV, Mark HE, Anstee QM, et al. ; NAFLD Consensus Consortium . Advancing the global public health agenda for NAFLD: a consensus statement. Nat Rev Gastroenterol Hepatol. 2022;19(1):60-78. doi: 10.1038/s41575-021-00523-4 [DOI] [PubMed] [Google Scholar]
- 25.Yamamichi N, Shimamoto T, Okushin K, et al. Fibrosis-4 Index efficiently predicts chronic hepatitis and liver cirrhosis development based on a large-scale data of general population in Japan. Sci Rep. 2022;12(1):20357. doi: 10.1038/s41598-022-24910-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Park H, Yoon EL, Kim M, et al. Reappraisal of Fibrosis-4 Index and non-alcoholic fatty liver disease fibrosis score for advanced fibrosis in average-risk population. Front Med (Lausanne). 2022;9:1024836. doi: 10.3389/fmed.2022.1024836 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Hunt CM, Lee TH, Morgan TR, Campbell S. One ALT is not like the other. Gastroenterology. 2023;165(2):320-323. doi: 10.1053/j.gastro.2023.04.009 [DOI] [PubMed] [Google Scholar]
- 28.McPherson S, Hardy T, Dufour JF, et al. Age as a confounding factor for the accurate non-invasive diagnosis of advanced NAFLD fibrosis. Am J Gastroenterol. 2017;112(5):740-751. doi: 10.1038/ajg.2016.453 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Park H, Yoon EL, Ito T, et al. Diagnostic performance of the Fibrosis-4 Index and nonalcoholic fatty liver disease fibrosis score in lean adults with nonalcoholic fatty liver disease. JAMA Netw Open. 2023;6(8):e2329568. doi: 10.1001/jamanetworkopen.2023.29568 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Rooks RN, Simonsick EM, Klesges LM, Newman AB, Ayonayon HN, Harris TB. Racial disparities in health care access and cardiovascular disease indicators in Black and White older adults in the Health ABC Study. J Aging Health. 2008;20(6):599-614. doi: 10.1177/0898264308321023 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Kornblith E, Bahorik A, Boscardin WJ, Xia F, Barnes DE, Yaffe K. Association of race and ethnicity with incidence of dementia among older adults. JAMA. 2022;327(15):1488-1495. doi: 10.1001/jama.2022.3550 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Moon H, Badana ANS, Hwang SY, Sears JS, Haley WE. Dementia prevalence in older adults: variation by race/ethnicity and immigrant status. Am J Geriatr Psychiatry. 2019;27(3):241-250. doi: 10.1016/j.jagp.2018.11.003 [DOI] [PubMed] [Google Scholar]
- 33.Mehta KM, Yeo GW. Systematic review of dementia prevalence and incidence in United States race/ethnic populations. Alzheimers Dement. 2017;13(1):72-83. doi: 10.1016/j.jalz.2016.06.2360 [DOI] [PubMed] [Google Scholar]
- 34.El-Serag HB, Kramer J, Duan Z, Kanwal F. Racial differences in the progression to cirrhosis and hepatocellular carcinoma in HCV-infected veterans. Am J Gastroenterol. 2014;109(9):1427-1435. doi: 10.1038/ajg.2014.214 [DOI] [PubMed] [Google Scholar]
- 35.Spiewak T, Taefi A, Patel S, Li CS, Chak E. Racial disparities of Black Americans hospitalized for decompensated liver cirrhosis. BMC Gastroenterol. 2020;20(1):245. doi: 10.1186/s12876-020-01392-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Jones PD, Lai JC, Bajaj JS, Kanwal F. Actionable solutions to achieve health equity in chronic liver disease. Clin Gastroenterol Hepatol. 2023;21(8):1992-2000. doi: 10.1016/j.cgh.2023.03.043 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Tapper EB, Parikh ND. Diagnosis and management of cirrhosis and its complications: a review. JAMA. 2023;329(18):1589-1602. doi: 10.1001/jama.2023.5997 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eAppendix. ICD Codes Used
Data Sharing Statement