Abstract
INTRODUCTION
We examined midlife (1990–1992, mean age 57) and late‐life (2011–2013, mean age 75) nonalcoholic fatty liver disease (NAFLD) and aminotransferase with incident dementia risk through 2019 in the Atherosclerosis Risk in Communities (ARIC) Study.
METHODS
We characterized NAFLD using the fatty liver index and fibrosis‐4, and we categorized aminotransferase using the optimal equal‐hazard ratio (HR) approach. We estimated HRs for incident dementia ascertained from multiple data sources.
RESULTS
Adjusted for demographics, alcohol consumption, and kidney function, individuals with low, intermediate, and high liver fibrosis in midlife (HRs: 1.45, 1.40, and 2.25, respectively), but not at older age, had higher dementia risks than individuals without fatty liver. A U‐shaped association was observed for alanine aminotransferase with dementia risk, which was more pronounced in late‐life assessment.
DISCUSSION
Our findings highlight dementia burden in high‐prevalent NAFLD and the important feature of late‐life aminotransaminase as a surrogate biomarker linking liver hypometabolism to dementia.
Highlights
Although evidence of liver involvement in dementia development has been documented in animal studies, the evidence in humans is limited.
Midlife NAFLD raised dementia risk proportionate to severity.
Late‐life NAFLD was not associated with a high risk of dementia.
Low alanine aminotransferase was associated with an elevated dementia risk, especially when measured in late life.
Keywords: aminotransferase, dementia, liver fibrosis, liver‐brain axis, nonalcoholic fatty liver disease
1. BACKGROUND
Alzheimer's disease and related dementias exhibit brain perturbations in energy metabolism, lipid metabolism, and oxidative stress. 1 Research on mice with mutations in the γ‐secretase subunits, which induce amyloid pathology and dementia onset, have shown that the liver is the first peripheral organ to experience metabolic dysregulation with progression of amyloid pathology. 2 The liver plays a crucial role as the central metabolic hub, and the loss of hepatic synthesis and metabolic functionality may contribute to the pathogenesis of dementia through abnormal bidirectional transport and communication between the brain and periphery. 3 Population‐based research is needed to advance scientific understanding of the potential role of the liver in dementia development.
There is currently no consensus definition for liver integrity. Nonalcoholic fatty liver disease (NAFLD) is the most common liver disorder worldwide, characterized by the accumulation of fat in the liver, with insidious progression to liver cell damage and dysfunction. 4 Based on disease progression and severity, NAFLD is further categorized into a spectrum from simple steatosis to steatohepatitis to fibrosis. 4 Previous studies investigating the associations between NAFLD, liver fibrosis, and cognitive impairment or dementia have been inconsistent and limited by cross‐sectional study designs as well as a lack of assessment of NAFLD across the full severity spectrum and at multiple time points. 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22
Elevated alanine aminotransferase (ALT) and aspartate aminotransferase (AST) are also widely used markers in clinical practice to detect hepatocellular injury. Mild to moderate elevations are commonly associated with NAFLD, 23 , 24 whereas low levels are often considered normal. 24 However, accumulating studies consistently show that low aminotransferase levels are associated with frailty, sarcopenia, and increased risk of mortality. 25 , 26 , 27 Lower levels of ALT were also reported in patients with dementia compared to cognitively unimpaired individuals. This analysis, however, was cross‐sectional and reverse causation could not be ruled out. Furthermore, alcohol consumption, which is associated with altered aminotransferase levels, was not available as a covariate. 5 Although the specific symptoms and characteristic liver pathology of low aminotransaminase are not well understood, it may provide complementary information about liver integrity.
We quantified liver integrity on the NAFLD–fibrosis spectrum and aminotransferase levels (with specific interest in low‐normal levels) in midlife and late life in the community‐based cohort of the Atherosclerosis Risk in Communities (ARIC) Study, and prospectively assessed their associations with incident dementia over 30 years of follow‐up.
2. METHODS
2.1. Study population
ARIC is an ongoing prospective cohort of 15,792 participants aged 45 to 64 years at enrollment in 1987 to 1989, sampled from four U.S. communities (Forsyth County, North Carolina; Minneapolis, Minnesota; Washington County, Maryland; and Jackson, Mississippi). 28 In this study, midlife liver integrity was assessed in 1990 to 1992 (visit 2, mean age 57 years) and late‐life liver integrity was assessed in 2011 to 2013 (visit 5, mean age 75 years). The cohort participants were followed through December 31, 2019, for incident dementia. The following participants were excluded (Figure 1): (1) not self‐identified as Black or White (the race recruitment were mostly center specific in ARIC, and Black participants from Minneapolis and Washington County were excluded due to the small numbers); (2) had prevalent stroke or dementia at baseline visit; (3) had clinical manifestations of liver disease, identified by International Classification of Diseases, Ninth Revision (ICD‐9) codes (Table S1); (4) had significant alcohol consumption (individuals with an alcohol‐related ICD‐9 code [Table S1] and current drinkers with a weekly alcohol consumption over 140 g for women and 210 g for men); (5) had ALT or AST level ≥300 U/L (NAFLD rarely has ALT or AST >300 U/L) 24 ; and (6) had missing covariates. In analyzing the relationship between aminotransferases and dementia risk, we further restricted the analysis individuals with ALT and AST level no more than 40 U/L, which is clinically deemed the normal range and accounted for >95% of the participants in this study. 24 The institutional review boards of all ARIC study sites approved the study protocol, and all cohort participants provided written informed consent at each examination.
FIGURE 1.
Study timeline and study eligibility criteria. ALT, alanine aminotransferase; AST, aspartate aminotransferase, NAFLD, nonalcoholic fatty liver disease.
2.2. Liver integrity
2.2.1. Assessment of the NAFLD–fibrosis spectrum
The NAFLD–fibrosis spectrum was estimated by combining the fatty liver assessment model of fatty liver index (FLI) and liver fibrosis assessment model of Fibrosis‐4 (FIB‐4), 29 , 30 , 31 , 32 following the equation and algorithm shown in Figure 2.
FIGURE 2.
The algorithm of nonalcoholic fatty liver disease (NAFLD) Atherosclerosis Risk in Communities –fibrosis spectrum. GGT (γ‐glutamyl transferase) was measured in serum using a kinetic rate reaction method on the Roche Cobas 6000 chemistry analyzer; platelet count was measured using ABX Horiba Diagnostics MICROS 60‐CS. ALT, alanine aminotransferase, AST, aspartate aminotransferase; BMI, body mass index.
RESEARCH IN CONTEXT
Systematic review: We reviewed literature on the association between the liver and brain using PubMed. Previous studies on nonalcoholic fatty liver disease (NAFLD) and dementia are inconsistent and limited by cross‐sectional study design and lack of NAFLD severity assessment at multiple time points. The prospective association between aminotransferase (another indicator of liver function) and dementia is unclear. These relevant citations are cited appropriately.
Interpretation: Midlife NAFLD may induce peripheral promoters of neurodegeneration and may expose a large segment of the population to increased risk of dementia. Low alanine aminotransferase, although typically considered clinically normal, was associated with an elevated dementia risk, especially in late life, which advocates for attention to the putative role of liver hypometabolism in the pathogenesis of dementia.
Future directions: Our findings motivate research on the potential influence of hepatic function on brain health, and of related opportunities for intervention on the risk of dementia.
2.2.2. Aminotransferase measurements and classification
Serum ALT and AST , were measured on frozen specimens collected at visit 2 and visit 5 using a kinetic rate reaction method on the Roche Cobas 6000 chemistry analyzer (Roche Diagnostics, Indianapolis, IN) at the University of Minnesota in 2011 to 2013. 33 After determining a U‐shaped association between aminotransferase and incident dementia with nonparametric smoothing technique, we used the optimal equal‐hazard ratio (HR) approach to identify two optimal cut‐points that have equal log relative hazard with a minimum Akaike information criterion (AIC) value, whereby we classified aminotransferase into “low‐normal,” “normal,” and “high‐normal” levels (R package ‘CutpointsOEHR’). 34 Details are provided in the Supplementary Statistics. Selected cut‐points were 4 and 14 U/L for ALT at visit 2, and 7 and 30 U/L at visit 5. Figure S1, panel B and D shows a monotonic increase in dementia risk with increasing AST and thus there was no meaningful cut‐point identified.
2.2.3. Incident dementia
Dementia was adjudicated and ascertained through December 31, 2019, from neuropsychological assessments, annual participant or informant contact, and medical record surveillance. 35 Details are provided in the Supplementary Statistics.
2.3. Covariates
Information on age, sex, race, and education was collected at the baseline visit of ARIC (visit 1, 1987‐1989). Participants’ behavioral and clinical characteristics were assessed at the relevant time of the exposure assessment (that is, visit 2 for analysis of NAFLD and aminotransferases in midlife, and at visit 5 for analysis of NALFD and aminotransferases in late life) using standardized protocols. 36 Details are provided in the Supplementary Statistics.
2.4. Statistical evaluation
Participant characteristics in midlife (visit 2) and late life (visit 5) were summarized across NAFLD–fibrosis spectrum and aminotransferase categories. Age, sex, and race‐adjusted incidence rates of dementia were calculated for each category using Poisson regression with robust standard errors.
To quantify the prospective association of the NAFLD–fibrosis spectrum and aminotransferase categories measured in midlife and late life with incident dementia, we used multivariable Cox proportional hazards regression model and employed Efron's approximation to handle ties. Individuals classified as having no fatty liver served as the reference group for the NAFLD cohort. The normal aminotransferase level group was used as the reference group for the aminotransferase cohort.
Progressive adjusted models were fitted. Model 1 was the unadjusted model. Model 2 was adjusted for age, sex, race‐center, education, and apolipoprotein E (APOE) ɛ4 genotype. Model 3 was additionally adjusted for alcohol use and estimated glomerular filtration rate (eGFR). Model 4 further included other cardiometabolic factors, including body mass index (BMI), blood pressure, high density lipoprotein cholesterol, total cholesterol, hypertension, and diabetes. Model 3 was used as the primary model. Because the cardiometabolic factors included in model 4 are potential causal intermediates, this model was considered secondary only to quantify liver contributions to the development of dementia independent of cardiometabolic disorders.
Finally, we performed additional analyses to understand the unexpected inverse association of late‐life NAFLD with dementia: (1) NAFLD increases the risk of pre‐dementia death, which precludes the occurrence of dementia. We first fit a Cox regression model for pre‐dementia death. Then to account for the competing risk of pre‐dementia death, we performed competing risk analyses (subdistribution hazard modeling with cumulative incidence function and sub‐hazard ratios). 37 (2) As another approach, we applied inverse probability censoring weight to account for the cohort attrition due to pre‐dementia death and non‐death drop‐out. 38 For greater specificity, pooled logistic regression was used to estimate the cumulative probability of being alive and participation separately. (3) Ninety percent of Black participants from the Jackson site lacked platelet measurements at visit 2 and thus had no assessments of NAFLD–fibrosis. This resulted in only 3.6% Black participants in the NAFLD cohort at visit 2 compared with ≈25% in the ARIC cohort and in NAFLD cohort at visit 5. To confirm that the different patterns of association at visit 2 and visit 5 were not due to the different composition of Black and White participants, we repeated the analyses using FLI classified NAFLD only (no fatty liver, indeterminate fatty liver, and fatty liver) in 11,650 participants at visit 2 and 5038 participants at visit 5. (4) Stratified by the development of incident dementia, we compared weight loss since the last visit prior to the baseline visit, as unintentional weight loss is expected to cause NAFLD misclassification, which in turn would bias the association estimates.
Analyses were performed with Stata version 17.0 (StataCorp LLC, College Station, Texas) and R version 4.3.1. A p‐value < 0.05 was considered statistically significant.
3. RESULTS
3.1. Midlife and late‐life characteristics by NAFLD–fibrosis spectrum
At visit 2, the mean age of 8972 cohort participants without dementia was 57.1 (SD 5.7) years, 55.5% were female, and 3.6% were Black. There were 34.5%, 26.6%, 25.2%, 13.0%, and 0.7% of cohort participants classified as having no fatty liver, indeterminate fatty liver, low, intermediate, and high risk of liver fibrosis, respectively. Participants with severe NAFLD–fibrosis were older and more likely to be male, less educated, be former drinkers, have a lower estimated glomerular filtration rate (eGFR), higher aminotransferase, higher body mass index (BMI), higher systolic blood pressure, higher triglyceride levels, and have diabetes and hypertension (Table 1).
TABLE 1.
Characteristics of participants in the Atherosclerosis Risk in Communities (ARIC) study visit 2 (1990–1992) by categories of nonalcoholic fatty liver disease (NAFLD) –fibrosis spectrum (N = 8972).
| Total | No fatty liver | Indeterminate fatty liver | Low‐risk liver fibrosis | Intermediate‐risk liver fibrosis | High‐risk liver fibrosis | |
|---|---|---|---|---|---|---|
| N | 8972 | 3097 | 2389 | 2263 | 1165 | 58 |
| Age, years, mean (SD) | 57.1 (5.7) | 56.5 (5.7) | 57.7 (5.7) | 55.9 (5.4) | 59.9 (5.0) | 61.2 (5.3) |
| Female, n (%) | 4980 (55.5) | 2266 (73.2) | 1126 (47.1) | 1136 (50.2) | 437 (37.5) | 15 (25.9) |
| Race (center), n (%) | ||||||
| White (Forsyth County, NC, USA) | 2620 (29.2) | 1028 (33.2) | 748 (31.3) | 481 (21.3) | 348 (29.9) | 15 (25.9) |
| White (Minneapolis, MN, USA) | 3017 (33.6) | 1128 (36.4) | 783 (32.8) | 771 (34.1) | 323 (27.7) | 12 (20.7) |
| White (Washington, MD, USA) | 3007 (33.5) | 849 (27.4) | 775 (32.4) | 911 (40.3) | 445 (38.2) | 27 (46.6) |
| Black (Forsyth County, NC, USA) | 309 (3.4) | 88 (2.8) | 80 (3.3) | 91 (4.0) | 46 (3.9) | 4 (6.9) |
| Black (Jackson, MS, USA) | 19 (0.2) | 4 (0.1) | 3 (0.1) | 9 (0.4) | 3 (0.3) | 0 (0.0) |
| Education, n (%) | ||||||
| <High school | 1440 (16.0) | 367 (11.9) | 389 (16.3) | 417 (18.4) | 251 (21.5) | 16 (27.6) |
| High school or vocational school | 4129 (46.0) | 1474 (47.6) | 1070 (44.8) | 1049 (46.4) | 516 (44.3) | 20 (34.5) |
| At least some college | 3403 (37.9) | 1256 (40.6) | 930 (38.9) | 797 (35.2) | 398 (34.2) | 22 (37.9) |
| APOE ε4 status, n (%) | 2516 (28.0) | 915 (29.5) | 668 (28.0) | 613 (27.1) | 306 (26.3) | 14 (24.1) |
| Alcohol use, n (%) | ||||||
| Current | 5471 (61.0) | 1982 (64.0) | 1484 (62.1) | 1333 (58.9) | 639 (54.8) | 33 (56.9) |
| Former | 1702 (19.0) | 479 (15.5) | 457 (19.1) | 479 (21.2) | 267 (22.9) | 20 (34.5) |
| Never | 1799 (20.1) | 636 (20.5) | 448 (18.8) | 451 (19.9) | 259 (22.2) | 5 (8.6) |
| eGFR, mean (SD) | 94.4 (13.5) | 96.1 (12.5) | 93.5 (13.2) | 95.3 (13.7) | 90.0 (14.6) | 85.1 (19.8) |
| ALT, U/L, mean (SD) | 16.5 (9.5) | 12.9 (5.4) | 16.2 (8.8) | 19.6 (9.3) | 20.1 (12.9) | 25.9 (33.8) |
| AST, U/L, mean (SD) | 20.8 (7.3) | 19.3 (5.1) | 20.5 (6.7) | 20.2 (5.7) | 25.4 (10.2) | 40.0 (26.2) |
| BMI, kg/m2, mean (SD) | 27.5 (5.1) | 23.2 (2.4) | 26.6 (2.2) | 32.1 (4.8) | 31.2 (4.3) | 31.5 (4.5) |
| SBP, mmHg, mean (SD) | 119.5 (17.6) | 114.0 (16.7) | 119.1 (16.8) | 123.9 (16.7) | 126.3 (18.5) | 125.6 (18.5) |
| DBP, mmHg, mean (SD) | 71.0 (9.8) | 68.5 (9.4) | 70.9 (9.5) | 73.5 (9.7) | 72.9 (10.2) | 71.4 (10.6) |
| Triglycerides, mg/dL, mean (SD) | 140.1 (86.8) | 94.4 (40.1) | 129.0 (54.5) | 185.2 (100.8) | 193.7 (116.5) | 197.0 (179.3) |
| Total cholesterol, mg/dL, mean (SD) | 209.5 (38.4) | 201.9 (34.6) | 210.2 (38.4) | 217.3 (39.8) | 213.5 (40.7) | 206.8 (46.2) |
| HDL, mg/dL, mean (SD) | 48.4 (16.2) | 57.8 (16.7) | 47.0 (14.2) | 41.5 (12.6) | 39.6 (12.3) | 43.1 (17.0) |
| Diabetes, n (%) | 1047 (11.7) | 94 (3.0) | 188 (7.9) | 464 (20.5) | 282 (24.2) | 19 (32.8) |
| Hypertension, n (%) | 2639 (29.5) | 521 (16.9) | 619 (26.0) | 926 (41.0) | 547 (47.0) | 26 (44.8) |
Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; HDL, high‐density lipoprotein; SBP, systolic blood pressure.
Among 4938 eligible cohort participants at visit 5, the mean age was 75.5 (SD 5.1) years, 59.0% were female, and 22.9% were Black. No fatty liver, indeterminate fatty liver, low, intermediate, and high risk of liver fibrosis accounted for 25.7%, 30.1%, 13.5%, 26.4%, and 4.3% of cohort participants, respectively. Table S2 shows late‐life characteristics at visit 5 by NAFLD–fibrosis spectrum, with some patterns consistent with those in midlife; however, for age, proportion of Black, current drinkers, and diabetes, the associations followed a “J”‐ and “inverse‐J” shape.
3.2. Dementia risk by NAFLD–fibrosis spectrum in midlife and late life
During a median follow‐up of 24.5 years after midlife NAFLD–fibrosis assessment and 1789 participants developed dementia. The age, sex, and race‐adjusted incidence were 6.3 (95% confidence interval [CI], 5.6–7.0), 6.7 (6.0–7.6), 7.8 (6.9–8.9), 7.2 (6.3–8.2), and 6.3 (3.8–10.5) per 1000 person‐years, respectively, across the least to the most severe category of the NAFLD–fibrosis spectrum. The median follow‐up time in late life was 6.3 years, and 865 dementia cases were documented. The age, sex, and race‐adjusted incidences of dementia were roughly three times the estimates in midlife, at 20.8 (16.7–25.8), 17.1 (13.8–21.2), 19.9 (15.1–26.2), 18.6 (14.7–23.4), and 22.0 (15.8–30.6) per 1000 person‐years, respectively.
Using multivariable adjusted Cox regression, increasing severity of midlife NAFLD–fibrosis was associated with a graded increase in dementia risk. Compared to individuals with no fatty liver, the HR of 1.45 (1.27–1.64), 1.40 (1.21–1.63), and 2.25 (1.20–4.22) were observed for low, intermediate, and high‐risk liver fibrosis categories, respectively. After further adjustment for cardiometabolic risk factors, these associations were attenuated but remained significant (Table 2). In contrast, severe NAFLD–fibrosis in late life was not consistently associated with increased dementia risk. Instead, indeterminate fatty liver and intermediate risk liver fibrosis were associated with 25% and 17% lower dementia risk, respectively, when compared to the no fatty liver category (Table 2).
TABLE 2.
Hazard ratios (95% confidence intervals) of incident dementia by categories of nonalcoholic fatty liver disease (NAFLD)–fibrosis spectrum in midlife and late life in the Atherosclerosis Risk in Communities (ARIC) study.
| No fatty liver | Indeterminate fatty liver | Low‐risk liver fibrosis | Intermediate‐risk liver fibrosis | High‐risk liver fibrosis | |
|---|---|---|---|---|---|
| Midlife visit 2 (N = 8972) | N = 3097 | N = 2389 | N = 2263 | N = 1165 | N = 58 |
| Model 1 b | 1 (Ref.) | 1.29 (1.14, 1.45) a | 1.27 (1.12, 1.44) a | 2.03 (1.76, 2.35) a | 2.46 (1.32, 4.60) a |
| Model 2 c | 1 (Ref.) | 1.12 (0.99, 1.26) | 1.46 (1.29, 1.66) a | 1.41 (1.21, 1.63) a | 2.18 (1.16, 4.09) a |
| Model 3 d | 1 (Ref.) | 1.11 (0.98, 1.25) | 1.45 (1.27, 1.64) a | 1.40 (1.21, 1.63) a | 2.25 (1.20, 4.22) a |
| Model 4 e | 1 (Ref.) | 1.05 (0.92, 1.21) | 1.27 (1.05, 1.54) a | 1.25 (1.02, 1.53) a | 1.94 (1.01, 3.71) a |
| Late‐life visit 5 (N = 4938) | N = 1267 | N = 1487 | N = 665 | N = 1305 | N = 214 |
|---|---|---|---|---|---|
| Model 1 b | 1 (Ref.) | 0.82 (0.69, 0.97) a | 0.62 (0.49, 0.79) a | 0.76 (0.64, 0.92) a | 1.36 (1.01, 1.83) a |
| Model 2 c | 1 (Ref.) | 0.78 (0.65, 0.93) a | 0.90 (0.71, 1.16) | 0.87 (0.72, 1.04) | 1.05 (0.78, 1.42) |
| Model 3 d | 1 (Ref.) | 0.75 (0.63, 0.89) a | 0.85 (0.66, 1.09) | 0.83 (0.69, 1.00) | 1.01 (0.74, 1.37) |
| Model 4 e | 1 (Ref.) | 0.75 (0.61, 0.92)a | 0.89 (0.65, 1.23) | 0.84 (0.64, 1.11) | 1.10 (0.77, 1.58) |
Statistically significance at P < 0.05.
Model 1: Crude model.
Model 2: Age, sex, race‐center, education, and APOE ɛ4 genotype.
Model 3 (Primary model): + alcohol use, estimated glomerular filtration rate.
Model 4: + body mass index, blood pressure, high density lipoprotein cholesterol, total cholesterol, hypertension, and diabetes.
3.3. Midlife and late‐life characteristics by aminotransferase category
At visit 2, compared to normal ALT levels, individuals with low‐normal levels of ALT were more likely to be older, Black, and carriers of the APOE ɛ4 allele, less educated and current drinkers, and have less hypertension; these relations were opposite for high‐normal versus normal levels of ALT. Both low‐normal and high‐normal ALT levels were associated with more diabetes. In addition, individuals with low‐normal ALT level were more likely be classified as having no fatty liver and intermediate risk liver fibrosis (Table S3). At visit 5, the pattern of late‐life characteristics by ALT category was generally consistent with that of midlife (Table S4).
3.4. Dementia risk by aminotransferase category in midlife and late life
During a median follow‐up of 24.1 years after visit 2, a total of 2411 participants developed dementia. The age, sex, and race‐adjusted incidence were 6.3 (3.5–11.5), 7.0 (6.3–7.7), and 7.3 (6.6–8.0) per 1000 person‐years, respectively, for low‐normal, normal, and high‐normal levels of ALT, respectively. The median follow‐up time after visit 5 was 6.3 years, with 892 dementia cases documented. The age, sex, and race‐adjusted incidences of dementia were 39.9 (25.0–63.6), 18.5 (15.2–22.6), and 24.4 (17.2–34.7) per 1000 person‐years, respectively.
We observed a U‐shaped pattern in dementia risk associated with ALT. At visit 2, compared to the normal ALT, the HRs of dementia were 1.05 (0.54–2.02) for low‐normal ALT and 1.09 (1.00–1.18) for high‐normal ALT, as assessed by the primary adjustment model; the corresponding HRs at visit 5 were 2.03 (1.34, 3.06) for low‐normal ALT and 1.43 (1.03, 1.97) for high‐normal ALT. Results were robust to adjustment for cardiometabolic factors (Table 3).
TABLE 3.
Hazard ratios (95% confidence intervals) of incident dementia by categories of alanine aminotransferase (ALT) in midlife and late life in the Atherosclerosis Risk in Communities (ARIC) study.
| Midlife | ALT < 4 (Low‐normal) | ALT 4‐ < 14 (Normal) | ALT 14‐40 (High‐normal) |
|---|---|---|---|
| Visit 2 (N = 11346) | N = 48 | N = 5885 | N = 5413 |
| Model 1 b | 1.01 (0.52, 1.95) | 1 (Ref.) | 0.98 (0.91, 1.07) |
| Model 2 c | 1.05 (0.54, 2.02) | 1 (Ref.) | 1.09 (1.00, 1.19) a |
| Model 3 d | 1.05 (0.54, 2.02) | 1 (Ref.) | 1.09 (1.00, 1.18) a |
| Model 4 e | 1.03 (0.54, 1.99) | 1 (Ref.) | 1.03 (0.95, 1.12) |
| Late‐life | ALT < 7 (Low‐normal) | ALT 7‐ < 30 (Normal) | ALT 30‐40 (High‐normal) |
|---|---|---|---|
| Visit 5 (N = 4925) | N = 73 | N = 4605 | N = 247 |
| Model 1 b | 2.34 (1.56, 3.51) a | 1 (Ref.) | 0.91 (0.66, 1.25) |
| Model 2 c | 2.09 (1.38, 3.14) a | 1 (Ref.) | 1.40 (1.02, 1.93) a |
| Model 3 d | 2.03 (1.34, 3.06) a | 1 (Ref.) | 1.43 (1.03, 1.97) a |
| Model 4 e | 2.20 (1.44, 3.36) a | 1 (Ref.) | 1.41 (1.01, 1.96) a |
Statistically significance at P < 0.05.
Model 1: Crude model.
Model 2: Age, sex, race‐center, education, and APOE ɛ4 genotype.
Model 3 (Primary model): + alcohol use, estimated glomerular filtration rate.
Model 4: + body mass index, blood pressure, high density lipoprotein cholesterol, total cholesterol, hypertension, and diabetes.
3.5. Supplementary analyses for opposite association of NAFLD with dementia in midlife and late life
At visit 2, compared to no fatty liver category, the HRs of pre‐dementia death were 1.18 (1.07–1.30), 1.57 (1.43–1.73), 1.56 (1.40–1.74), and 2.66 (1.92–3.68) for indeterminate fatty liver, low, and intermediate, and high risk of liver fibrosis categories, respectively. The subhazard ratios from the competing risk model for dementia and pre‐dementia death were attenuated. Although not statistically significant, the subhazard ratio of dementia showed a discordant direction for high‐risk liver fibrosis (0.79, 0.41–1.52) compared to the estimates from traditional Cox regression model. At visit 5, only high‐risk liver fibrosis was associated with higher hazard of pre‐dementia death (1.76, 1.28–2.42), whereas for indeterminate fatty liver, an inverse association (0.78, 0.61–0.97) was observed. The use of competing risk regression did not materially change the estimates compared with the traditional Cox regression (Table S5).
Using inverse probability censoring weight to account for the cohort attrition, while we found a greater magnitude of dementia risk in individuals with midlife high‐risk liver fibrosis (3.97, 1.89–8.37), the association between late‐life NAFLD and dementia risk remained largely unchanged (Table S6). By classifying NAFLD using only the FLI, we confirmed a similar association pattern, that is, only midlife NAFLD was associated with an increased risk of dementia (Table S7). Finally, although the decline in weight‐related measures (weight, waist circumference, BMI) and triglycerides prior to visit 2 was comparable in two groups, it was more evident prior to visit 5 in individuals who developed dementia than in individuals who did not develop dementia (Table S8).
4. DISCUSSION
This study estimated the impact of liver integrity, as indicated by the NAFLD–fibrosis spectrum and aminotransferase levels, on dementia risk. Midlife NAFLD–fibrosis was associated with higher dementia risk, even after adjusting for cardiometabolic factors. Conversely, late‐life NAFLD–fibrosis was associated with lower dementia risk. The competing risk of pre‐dementia death and cohort attrition are unlikely to explain this result. In addition, there was a U‐shaped association between ALT and dementia risk, although low‐normal levels of ALT may not necessarily have the same risk profile compared to high‐normal ALT and NAFLD.
Several pathways connecting NAFLD to dementia have been proposed. Hepatic steatosis impairs insulin action, and loss of insulin signaling in turn promotes the development of NAFLD, thereby activating a vicious cycle of insulin resistance–lipotoxicity that drives NAFLD progression. 39 , 40 Hepatic and peripheral insulin resistance and metabolic dysregulation seen in NAFLD closely parallel that seen in dementia. 5 , 41 , 42 In addition, in NAFLD, accumulation of lipids in hepatocytes and infiltration of inflammatory immune cells result in hepatocyte injury and stimulate a sustained secretion of pro‐inflammatory cytokines and chemokines to the systemic circulation, leading to low‐grade systemic inflammation. 43 , 44 Inflammation increases cytokines and upregulates pro‐atherogenic transcription factors, which recruit monocytes and other immune cells to cross the blood–brain barrier and infiltrate the brain, thereby inducing neuroinflammation and disrupting brain function. 45 Along with these effects, the inflammatory state contributes to microglial activation, endothelial dysfunction, mitochondrial dysfunction, a pro‐coagulant state, and platelet activation, and exacerbates insulin resistance. Together, these factors may promote microvascular structural alterations, significant glial and neuronal cell loss, and neurotransmitter impairment and eventually leads to cerebral peripheral ischemia and neurodegeneration. 42 , 44 , 46 , 47
Many individuals with NAFLD also have excess burden of cardiometabolic risk factors. 3 However, animal studies have shown that NAFLD, even in the absence of obesity and hyperglycemia, can promote inflammation and neuronal apoptosis, cerebral hypoperfusion, and brain amyloid deposition. 43 , 48 Consistent with their evidence, in this study the graded association between midlife NAFLD–fibrosis and dementia risk persisted even after adjusting for cardiometabolic risk factors, indicating some direct liver contributions to dementia and requiring further investigations.
The inverse association of late‐life NAFLD with dementia risk merits discussion. The competing risk of pre‐dementia death and cohort attrition are unlikely to explain this result. A recent study found that neither NAFLD identified by FLI or abdominal ultrasound nor liver fibrosis assessed via transient elastography was associated with increased dementia risk in a general elderly population (mean age 70 years). Instead, NAFLD was associated with a reduced risk in the first 5 years of follow‐up. The authors noted that this phenomenon was likely the result of the reversibility of NAFLD due to weight loss. 19 In our study, we identified more weight loss and triglyceride decline before visit 5 (late life) than before visit 2 (midlife) in individuals who developed dementia than in individuals who did not. This weight loss, which may reflect poor health rather than NAFLD remission, directly affects NAFLD–fibrosis assessment and may lead to inaccuracies when using NAFLD assessment tools in older adults. 30 Our results align with previous studies that have demonstrated reduced dementia risks among individuals with a higher BMI in late life. 49
Our resutls also showed that decreased aminotransferase level was associated with increased dementia risk, which may reflect a distinct pathway by which the liver contributes to dementia compared to the connections between NAFLD and dementia, that is, reduced liver synthesis and metabolic function leading to a deficiency in energy sources and neurotransmitters. Aminotransferase, particularly ALT, is produced predominantly in the liver, 24 making it a potential surrogate biomarker for reduced liver synthesis and metabolic function, if found at low levels. Reduced liver synthesis and metabolic function contribute to or correlate with cerebral hypometabolism and occur prior to the onset of dementia. 1 , 5 Aminotransferase could also directly contribute to the dementia pathogenesis. Specifically, aminotransferase is the key enzyme to facilitate the production of glucose as a major metabolic fuel for various tissues and glutamate/glutamine acting as an excitatory neurotransmitter to maintain synapses in the brain. 5 , 23 , 39
The determinants of low aminotransaminase are not fully understood. Possible causes include advanced liver fibrosis with tissue loss 50 and primary liver hypometabolism possibly related to aging, malnutrition, or frailty. 26 , 27 In endocrine and primary care clinics, individuals with elevated aminotransferase level are considered to have high risk of NAFLD and forwarded for further evaluation. 4 If our hypothesis is confirmed by future research that low aminotransferase levels are caused by advanced liver fibrosis with substantial tissue loss, then ignoring low aminotransferase levels may result in severe cases of NAFLD being misclassified and missed opportunities for further diagnosis and timely treatment. On the other hand, if low aminotransferase levels are caused by liver hypometabolism without significant liver structural impairment, then further causal explanation of the mechanism inherent in the emergence of liver hypometabolism is necessary.
Our study has a few limitations. Histology‐based evaluation remains the gold standard for the diagnosis of liver steatosis and fibrosis, which is not indicated in healthy individuals. Misclassification is foreseeable when using noninvasive assessments that combine easily measured clinical variables. 29 , 30 , 31 , 32 Evidence show a good negative predictive value for FLI and FIB‐4 to exclude NAFLD cases, especially in community and primary care settings. 4 The positive predictive value is generally poor, however, and further biopsy examinations are recommended for those who are classified as NAFLD based on noninvasive assessments to eliminate the false‐positive cases, which unfortunately are not available in this study. Therefore, it is possible that the NAFLD group in this study includes a significant proportion of participants without true NAFLD, and no fatty liver group may include “lean NAFLD,” and thus the associations may be attenuated. Indeed, in our study, ≈40% of ARIC patients were classified as having NAFLD, which is higher than the estimates assessed in the general population (20%–35%). In addition, the NAFLD assessment models were developed and validated in patients 35 to 65 years of age, and caution should be exercised when using these models in older adults. 30 Despite those limitations, overall FLI and FIB‐4 present good performance in excluding advanced liver fibrosis and these measures represent the optimal liver assessment available in this large community‐based study and are clinically relevant. 4 Finally, we do not ascertain etiologic dementia diagnoses in all participants in ARIC. More research is needed to understand the role of the liver on different dementia types.
In conclusion, midlife NAFLD increases the risk of dementia. Although the specific signs or symptoms and the characteristic liver pathology of low aminotransaminase levels require further investigation, our findings highlight the important feature of low aminotransaminase in late life as a surrogate biomarker linking liver hypometabolism to dementia.
CONFLICT OF INTEREST STATEMENT
Dr. Mielke has served on scientific advisory boards and/or has consulted for Biogen, LabCorp, Lilly, Merck, PeerView Institute, Siemens Healthineers, and Sunbird Bio, unrelated to the current manuscript. Dr. Raffield is a consultant for the National Heart, Lung, and Blood Institute (NHLBI) Trans‐Omics for Precision Medicine Program Administrative Coordinating Center (through Westat). Dr. Hoogeveen has received research grants from Denka Seiken (to his institution) and is a consultant for Denka Seiken unrelated to the current manuscript. All other authors declare no conflicts of interest. Author disclosures are available in the Supporting information.
CONSENT STATEMENT
This study does not constitute human subjects’ research.
Supporting information
Supporting Information
Supporting Information
ACKNOWLEDGMENTS
We thank the staff and participants of the Atherosclerosis Risk in Communities (ARIC) study for their important contributions. The authors acknowledge the contributions of Dr. Gerardo Heiss, Kenan Distinguished Professor of Epidemiology at the University of North Carolina at Chapel Hill Gillings School of Global Public Health. Dr. Heiss was central to the successful development of this manuscript. He passed away on June 11, 2022. The ARIC has been funded in whole or in part with Federal funds from the National Heart, Lung, and Blood Institute, National Institutes of Health, Department of Health and Human Services, under Contract nos. 75N92022D00001, 75N92022D00002, 75N92022D00003, 75N92022D00004, 75N92022D00005. Neurocognitive data is supported by U01HL096812, U01HL096814, U01HL096899, U01HL096902, and U01HL096917 from the NIH (NHLBI, NINDS, NIA and NIDCD). Drs. Lu and Palta were in part supported by R01AG066134. Dr. Walker receives funding from the NIA Intramural Research Program. This study was funded, in part, by the NIA Intramural Research Program. Dr. Raffield is funded by R01AG075884.
Lu Y, Pike JR, Hoogeveen RC, et al. Liver integrity and the risk of Alzheimer's disease and related dementias. Alzheimer's Dement. 2024;20:1913–1922. 10.1002/alz.13601
REFERENCES
- 1. Clarke JR, Ribeiro FC, Frozza RL, De Felice FG, Lourenco MV. Metabolic dysfunction in Alzheimer's disease: from basic neurobiology to clinical approaches. J Alzheimers Dis. 2018;64:S405‐s426. [DOI] [PubMed] [Google Scholar]
- 2. Zheng H, Cai A, Shu Q, et al. Tissue‐specific metabolomics analysis identifies the liver as a major organ of metabolic disorders in amyloid precursor protein/presenilin 1 mice of Alzheimer's disease. J Proteome Res. 2019;18:1218‐1227. [DOI] [PubMed] [Google Scholar]
- 3. Bassendine MF, Taylor‐Robinson SD, Fertleman M, Khan M, Neely D. Is Alzheimer's disease a liver disease of the brain? J Alzheimers Dis. 2020;75:1‐14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Cusi K, Isaacs S, Barb D, et al. American association of clinical endocrinology clinical practice guideline for the diagnosis and management of nonalcoholic fatty liver disease in primary care and endocrinology clinical settings: co‐sponsored by the american association for the study of liver diseases (AASLD). Endocrine Practice. 2022;28:528‐562. [DOI] [PubMed] [Google Scholar]
- 5. Nho K, Kueider‐Paisley A, Ahmad S, et al. Association of altered liver enzymes with Alzheimer disease diagnosis, cognition, neuroimaging measures, and cerebrospinal fluid biomarkers. JAMA Netw Open. 2019;2:e197978. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Labenz C, Kostev K, Kaps L, Galle PR, Schattenberg JM. Incident dementia in elderly patients with nonalcoholic fatty liver disease in Germany. Dig Dis Sci. 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Weinstein G, Davis‐Plourde K, Himali JJ, Zelber‐Sagi S, Beiser AS, Seshadri S. Non‐alcoholic fatty liver disease, liver fibrosis score and cognitive function in middle‐aged adults: the Framingham Study. Liver Int. 2019;39:1713‐1721. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Solfrizzi V, Scafato E, Custodero C, et al. Liver fibrosis score, physical frailty, and the risk of dementia in older adults: the Italian Longitudinal Study on Aging. Alzheimers Dement (N Y). 2020;6:e12065‐e. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Filipović B, Marković O, Đurić V, Filipović B. Cognitive changes and brain volume reduction in patients with nonalcoholic fatty liver disease. Can J Gastroenterol Hepatol. 2018;2018:9638797. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. VanWagner LB, Terry JG, Chow LS, et al. Nonalcoholic fatty liver disease and measures of early brain health in middle‐aged adults: the CARDIA study. Obesity (Silver Spring). 2017;25:642‐651. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Seo SW, Gottesman RF, Clark JM, et al. Nonalcoholic fatty liver disease is associated with cognitive function in adults. Neurology. 2016;86:1136‐1142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Kim HM, Lee YH, Han K, et al. Impact of diabetes mellitus and chronic liver disease on the incidence of dementia and all‐cause mortality among patients with dementia. Medicine (Baltimore). 2017;96:e8753. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Shang Y, Widman L, Hagström H. Nonalcoholic fatty liver disease and risk of dementia: a population‐based cohort study. Neurology. 2022: Aug 9;99(6):e574–e582. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Shang Y, Nasr P, Ekstedt M, et al. Non‐alcoholic fatty liver disease does not increase dementia risk although histology data might improve risk prediction. JHEP Rep. 2021;3:100218. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Weinstein G, Zelber‐Sagi S, Preis SR, et al. Association of nonalcoholic fatty liver disease with lower brain volume in healthy middle‐aged adults in the Framingham Study. JAMA Neurol. 2018;75:97‐104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Celikbilek A, Celikbilek M, Bozkurt G. Cognitive assessment of patients with nonalcoholic fatty liver disease. Eur J Gastroenterol Hepatol. 2018;30:944‐950. [DOI] [PubMed] [Google Scholar]
- 17. George ES, Sood S, Daly RM, Tan SY. Is there an association between non‐alcoholic fatty liver disease and cognitive function? A systematic review. BMC Geriatr. 2022;22:47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Kim GA, Oh CH, Kim JW, et al. Association between non‐alcoholic fatty liver disease and the risk of dementia: a nationwide cohort study. Liver International. 2022;42:1027‐1036. [DOI] [PubMed] [Google Scholar]
- 19. Xiao T, van Kleef L, Ikram MK, De Knegt R, Ikram MA. Association of nonalcoholic fatty liver disease and fibrosis with incident dementia and cognition. Neurology. 2022;99:e565. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Gerber Y, VanWagner LB, Yaffe K, et al. Non‐alcoholic fatty liver disease and cognitive function in middle‐aged adults: the CARDIA study. BMC Gastroenterol. 2021;21:96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Parikh NS, Kamel H, Zhang C, et al. Association of liver fibrosis with cognitive test performance and brain imaging parameters in the UK Biobank study. Alzheimer's & Dementia. 2022. n/a. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Jeong S, Oh YH, Choi S, et al. Association of non‐alcoholic fatty liver disease with incident dementia later in life among elder adults. Clin Mol Hepatol. 2022;28:510‐521. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Sookoian S, Castaño GO, Scian R, et al. Serum aminotransferases in nonalcoholic fatty liver disease are a signature of liver metabolic perturbations at the amino acid and Krebs cycle level. Am J Clin Nutr. 2016;103:422‐434. [DOI] [PubMed] [Google Scholar]
- 24. Kwo PY, Cohen SM, Lim JK. ACG Clinical Guideline: evaluation of abnormal liver chemistries. American Journal of Gastroenterology. 2017;112. [DOI] [PubMed] [Google Scholar]
- 25. Lu Y, Pike JR, Selvin E, et al. Low liver enzymes and risk of dementia: the Atherosclerosis Risk in Communities (ARIC) Study. J Alzheimers Dis. 2021;79:1775‐1784. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Vespasiani‐Gentilucci U, De Vincentis A, Ferrucci L, Bandinelli S, Antonelli Incalzi R, Picardi A. Low alanine aminotransferase levels in the elderly population: frailty, disability, sarcopenia, and reduced survival. J Gerontol A Biol Sci Med Sci. 2018;73:925‐930. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Yamazaki H, Kamitani T, Matsui T, Yamamoto Y, Fukuhara S. Association of low alanine aminotransferase with loss of independence or death: a 5‐year population‐based cohort study. J Gastroenterol Hepatol. 2019;34:1793‐1799. [DOI] [PubMed] [Google Scholar]
- 28. Wright JD, Folsom AR, Coresh J, et al. The ARIC (Atherosclerosis Risk In Communities) Study: JACC Focus Seminar 3/8. Journal of the American College of Cardiology. 2021;77:2939‐2959. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Bedogni G, Bellentani S, Miglioli L, et al. The Fatty Liver Index: a simple and accurate predictor of hepatic steatosis in the general population. BMC gastroenterology. 2006;6:33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. 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;112:740‐751. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Koehler EM, Schouten JNL, Hansen BE, Hofman A, Stricker BH, Janssen HLA. External validation of the fatty liver index for identifying nonalcoholic fatty liver disease in a population‐based study. Clin Gastroenterol Hepatolo. 2013;11:1201‐1204. [DOI] [PubMed] [Google Scholar]
- 32. Sterling RK, Lissen E, Clumeck N, et al. Development of a simple noninvasive index to predict significant fibrosis in patients with HIV/HCV coinfection. Hepatology. 2006;43:1317‐1325. [DOI] [PubMed] [Google Scholar]
- 33. Parrinello CM, Grams ME, Couper D, et al. Recalibration of blood analytes over 25 years in the atherosclerosis risk in communities study: impact of recalibration on chronic kidney disease prevalence and incidence. Clin Chem. 2015;61:938‐947. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Chen Y, Huang J, He X, Gao Y, et al. A novel approach to determine two optimal cut‐points of a continuous predictor with a U‐shaped relationship to hazard ratio in survival data: simulation and application. BMC Med Res Methodol. 2019;19:96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Atherosclerosis Risk in Communities Neurocognitive Study. STATUS61 Derived Variable Dictionary. https://sites.cscc.unc.edu/aric/sites/default/files/public/manuals/ARIC%20STATUS61%20Derived%20Variable%20Dictionary%20190927.pdf
- 36. The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. The ARIC investigators. Am J Epidemiol. 1989;129:687‐702. [PubMed] [Google Scholar]
- 37. Austin PC, Lee DS, Fine JP. Introduction to the analysis of survival data in the presence of competing risks. Circulation. 2016;133:601‐609. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Weuve J, Tchetgen Tchetgen EJ, Glymour MM, et al. Accounting for bias due to selective attrition: the example of smoking and cognitive decline. Epidemiology. 2012;23:119‐128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Rui L. Energy metabolism in the liver. Compr Physiol. 2014;4:177‐197. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Michael MD, Kulkarni RN, Postic C, et al. Loss of insulin signaling in hepatocytes leads to severe insulin resistance and progressive hepatic dysfunction. Molecular Cell. 2000;6:87‐97. [PubMed] [Google Scholar]
- 41. Karbalaei R, Allahyari M, Rezaei‐Tavirani M, Asadzadeh‐Aghdaei H, Zali MR. Protein‐protein interaction analysis of Alzheimer`s disease and NAFLD based on systems biology methods unhide common ancestor pathways. Gastroenterol Hepatol Bed Bench. 2018;11:27‐33. [PMC free article] [PubMed] [Google Scholar]
- 42. de la Monte SM, Tong M. Brain metabolic dysfunction at the core of Alzheimer's disease. Biochem Pharmacol. 2014;88:548‐559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Kim DG, Krenz A, Toussaint LE, et al. Non‐alcoholic fatty liver disease induces signs of Alzheimer's disease (AD) in wild‐type mice and accelerates pathological signs of AD in an AD model. J Neuroinflammation. 2016;13:1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Gehrke N, Schattenberg JM. Metabolic inflammation – a role for hepatic inflammatory pathways as drivers of comorbidities in nonalcoholic fatty liver disease? Gastroenterology. 2020;158:1929‐1947. e6. [DOI] [PubMed] [Google Scholar]
- 45. Kjærgaard K, Mikkelsen ACD, Wernberg CW, et al. Cognitive dysfunction in non‐alcoholic fatty liver disease‐current knowledge, mechanisms and perspectives. J Clin Med. 2021;10:673. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Lombardi R, Fargion S, Fracanzani AL. Brain involvement in non‐alcoholic fatty liver disease (NAFLD): a systematic review. Dig Liver Dis. 2019;51:1214‐1222. [DOI] [PubMed] [Google Scholar]
- 47. Peng Z, Duggan MR, Dark HE, et al. Association of liver disease with brain volume loss, cognitive decline, and plasma neurodegenerative disease biomarkers. Neurobiol Aging. 2022;120:34‐42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Pinçon A, De Montgolfier O, Akkoyunlu N, et al. Non‐Alcoholic fatty liver disease, and the underlying altered fatty acid metabolism, reveals brain hypoperfusion and contributes to the cognitive decline in APP/PS1 mice. Metabolites. 2019;9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Hughes TF, Borenstein AR, Schofield E, Wu Y, Larson EB. Association between late‐life body mass index and dementia: the Kame Project. Neurology. 2009;72:1741‐1746. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Mofrad P, Contos MJ, Haque M, et al. Clinical and histologic spectrum of nonalcoholic fatty liver disease associated with normal ALT values. Hepatology. 2003;37:1286‐1292. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information
Supporting Information