- Email: [email protected]
or Type 2 Diabetes Mellitus
Current Problems in Diagnostic RadiologyIIMB Management ReviewJournal of Cardiac FailureJournal of Exotic Pet MedicineBiology of Blood and Marrow Transplantation
Accepted Manuscript
Cognitive Performance in Individuals with Non-alcoholic Fatty Liver Disease and/or Type 2 Diabetes Mellitus Ali A. Weinstein Ph.D. , Leyla de Avila B.A. , James Paik Ph.D. , Pegah Golabi M.D. , Carey Escheik B.S. , Lynn Gerber M.D. , Zobair M. Younossi M.D., M.P.H. PII: DOI: Reference:
S0033-3182(18)30308-6 10.1016/j.psym.2018.06.001 PSYM 896
To appear in:
The End-to-end Journal
Received date: Revised date: Accepted date:
17 May 2018 31 May 2018 1 June 2018
Please cite this article as: Ali A. Weinstein Ph.D. , Leyla de Avila B.A. , James Paik Ph.D. , Pegah Golabi M.D. , Carey Escheik B.S. , Lynn Gerber M.D. , Zobair M. Younossi M.D., M.P.H. , Cognitive Performance in Individuals with Non-alcoholic Fatty Liver Disease and/or Type 2 Diabetes Mellitus, The End-to-end Journal (2018), doi: 10.1016/j.psym.2018.06.001
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Cognitive Performance in Individuals with Non-alcoholic Fatty Liver Disease and/or Type 2 Diabetes Mellitus
CR IP T
Ali A. Weinstein, Ph.D.1,2, Leyla de Avila, B.A.1, James Paik, Ph.D.1, Pegah Golabi, M.D. 1, Carey Escheik, B.S.1, Lynn Gerber, M.D.2,3, Zobair M. Younossi M.D., M.P.H. 1, 3
1
AN US
Betty and Guy Beatty Center for Integrated Research, Inova Health System, Falls Church, VA, United States. 2 Center for the Study of Chronic Illness and Disability, George Mason University, Fairfax, VA, United States. 3 Center for Liver Disease, Department of Medicine, Inova Fairfax Hospital, Falls Church, VA, United States.
M
Corresponding Author: Zobair M. Younossi, MD, MPH Betty and Guy Beatty Center for Integrated Research Claude Moore Health Education and Research Building 3300 Gallows Road, Falls Church, VA 22042 Phone: (703) 776-2540 Fax: (703) 776-4386 Email: [email protected]
Word count: 2497; Number of tables: 3; Number of figures: 1
ED
Conflict of Interest Statement: Dr. Younossi is a consultant to BMS, Gilead, AbbVie, Intercept, and GSK. All other authors have no conflict of interest to disclose.
PT
Financial Support Statement: This work was supported by the Betty and Guy Beatty Center for
CE
Integrated Research.
Authors contributions: AW, LD, CE and PG participated in the study design, helped the interpretation
AC
of the data and to draft the manuscript. JP performed the statistical analysis and helped with the interpretation of the data. LG and ZMY conceived of the study, participated substantially in its design and coordination, and helped to draft the manuscript. All authors read and approved the final manuscript.
1
ACCEPTED MANUSCRIPT
ABSTRACT Background: Individuals with Non-Alcoholic Fatty Liver Disease (NAFLD) share some common pathophysiological features with individuals with type 2 diabetes mellitus (T2DM). There is a wellestablished association between T2DM and cognitive decline, but no corollary data in people with
CR IP T
NAFLD without T2DM and whether combination of the two disorders is associated with additive deficits in cognitive performance. The purpose of this investigation is to compare measures of cognitive
performance for individuals with NAFLD, individuals with T2DM, individuals with both or neither. Methods: Using NHANES data from 2011–2014, 1,102 individuals were identified that had completed
AN US
cognitive assessments. Cognitive performance was assessed by a series of well-established tasks.
Presence of T2DM was defined as the use of diabetes medication and/or a fasting glucose measure of > 126 mg/dL. Presence of NAFLD was defined as Fatty Liver Index≥60 in the absence of other chronic liver diseases. Results: After controlling for demographics, comorbidities, and metabolic components,
M
individuals with both NAFLD and T2DM scored statistically significantly lower on a task that combines processing speed, sustained attention, and working memory (Beta=-3.44, 95% CI: -6.75 - -0.12) than
ED
individuals with neither. Individuals with T2DM without NAFLD scored statistically significantly lower on verbal fluency (Beta=-1.47, 95% CI: -2.7 - -0.23) than individuals with neither. Conclusions: Data
PT
from this study suggest that individuals with T2DM and individuals with both NAFLD and T2DM have lower cognitive performance on various tasks. These data support an approach that aims to apply
CE
preventive strategies to optimize management of T2DM in patients with NAFLD.
AC
Keywords: cognitive performance; nutrition surveys; memory; insulin resistance; steatosis; NAFLD; diabetes
2
ACCEPTED MANUSCRIPT
INTRODUCTION In the United States and globally, non-alcoholic fatty liver disease (NAFLD) is the most common cause of chronic liver disease1,2. NAFLD is associated with cardiovascular disease (CVD) and its risk factors including type 2 diabetes mellitus (T2DM), obesity, hyperlipidemia, and hypertension3,4. These
CR IP T
risk factors are known to contribute to cognitive impairment with or without the presence of CVD5,6. NAFLD is physiologically related to T2DM, as there are shared risk factors linking these diseases. For example, both individuals with NAFLD and those with T2DM are more likely to be obese7. While the global prevalence of NAFLD is 25%8, in patients with T2DM the prevalence of NAFLD is considerably
AN US
higher, occurring in up to 65-75% of these individuals9. We have experienced the existence of this comorbidity in our clinic patients. It is rare for patients seen in our clinic to be overweight, have T2DM, and NOT have evidence of NAFLD. Therefore, we wished to explore relationships between NAFLD, T2DM and cognition, since cognitive performance impairment has been well documented in people with T2DM.
M
We sought to determine the relationship between NAFLD and cognitive performance as an independent association.
ED
Along with the common comorbidity between T2DM and NAFLD, the insulin resistance that is the hallmark of T2DM has also been tightly associated with the risk of NAFLD. Insulin is delivered
PT
directly to the portal vein after secretion and the liver eliminates a large portion of portal insulin at this first pass10. In addition, NAFLD is associated with dysfunctional adipose tissue, and lipo-toxicity
CE
promotes insulin resistance and pancreatic β-cell dysfunction. Therefore, there are common pathophysiological factors in both of these diseases10.
AC
The association between T2DM and an increased risk of cognitive decline is well established11-18.
Results of neurocognitive performance tests have demonstrated cognitive impairments in those with T2DM that exceed the rate of usual aging-associated cognitive decline11, specifically in the domains of attention12–15, psychomotor speed13–15, executive functions14,16,17, and learning and memory14,18. However, there is a lack of literature on the specific relationship between NAFLD and cognitive performance. We were able to identify only one study that investigated cognitive performance in those with NAFLD19. In 3
ACCEPTED MANUSCRIPT
that investigation, statistical manipulation was used to control for the presence of T2DM in those with NAFLD. Therefore, our aim was to compare patients with NAFLD, patients with T2DM, patients with both NAFLD and T2DM, and individuals with neither (control) on measures of cognitive performance in order to determine whether there was an association and possibly an additive performance impact when
CR IP T
both conditions co-existed.
PATIENTS AND METHODS Study population
AN US
Data were obtained from two continuous cycles of the National Health and Nutrition Examination Survey (NHANES 2011–2012 and NHANES 2013–2014). These data provide a cross-sectional, population-based sampling survey of the civilian non-institutionalized population of the United States. Data obtained from these surveys include both self-reports and objective clinical measures. Surveys were
M
based on a complex, multi-stage sampling plan. Detailed information about design and sampling of the survey is available20. Cognitive performance tests were only administered to individuals aged 60 and over,
ED
therefore participants under the age of 60 were excluded from analyses. In addition, participants were excluded if they were missing data needed to calculate NAFLD or T2DM status or a history of stroke,
PT
HIV, alcoholic liver disease (ALD), hepatitis B virus (HBV), hepatitis c virus (HCV), or excessive alcohol use (EAU). Figure 1 details the cohort selection for the current analyses.
CE
The following definitions were used to define groups for these analyses. T2DM was defined as the use of diabetes medication and/or a fasting glucose measure of > 126 mg/dL. NAFLD was defined by
AC
the absence of any other causes of chronic liver disease (e.g. ALD, HBV, HCV, or iron overload) as well as a Fatty Liver Index (FLI) score ≥ 6021. ALD was defined by EAU with elevated liver enzymes. EAU was defined as ≥ 20 g per day of alcohol in men and ≥ 10 g of alcohol per day in women. Elevated liver enzymes were defined as alanine aminotransferase (ALT) ≥ 40U/L in men and ≥ 31U/L in women or aspartate aminotransferase (AST) level ≥ 37U/L in men and ≥ 31U/L in women. HBV was defined as positive Hepatitis B surface antigen and HCV was defined as positive hepatitis C virus RNA. Iron 4
ACCEPTED MANUSCRIPT
overload was defined as serum transferrin saturation ≥ 50%. In addition, the following information was used for modeling purposes: (1) poverty income ratio (PIR) (PIR<1.3=low; 1.33/5=high); (2) obesity (body-mass index ≥ 30 kg/
);
(3) hypertension (systolic blood pressure > 140 mmHG or diastolic blood pressure > 90 mmHg or history
CR IP T
of blood pressure); (4) metabolic syndrome (at least of the following present waist circumference >102 cm in men or >88 cm in women, fasting plasma glucose >110 mg/dl, BP >130/85 mm Hg, elevated triglycerides >150 mg/dl, and HDL ≤ 40 mg/dL in men or ≤ 50 mg/dL in women); and (5) insulin resistance (homeostasis of model assessment (HOMA) of > 3).
AN US
Cognitive Performance
A series of assessments in NHANES are used to examine cognitive performance. The word learning subset from the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD-WL) was used to assess immediate and delayed learning ability for new verbal information22. The animal fluency
M
test (AFT) examines categorical verbal fluency, a component of executive function23. The digit symbol substitution test (DSST), a performance module from the Wechsler Adult Intelligence Scale (WAIS III) is
Statistical Analysis
ED
a test that involves processing speed, sustained attention, and working memory24.
PT
Survey design elements (Clusters, Strata, and examination sample weights) provided by the National Center for Health Statistics (NCHS) were used to account for survey non-response, noncoverage,
CE
and sampling strategy. For national representation, original weights in our merged sample were modified using the method recommended by NCHS25. Taylor series linearization was used to calculate standard
AC
errors to take into account the complex sampling design. This study compares cognitive performance in individuals with T2DM and NAFLD to cognitive
performance in a group of individuals with neither T2DM or NAFLD from the same age-matched study cohort. The prevalence of various characteristics including demographic, clinical parameters, and outcomes across study groups was compared using a t-statistic for continuous variables and the Rao-Scott
5
ACCEPTED MANUSCRIPT
chi-square test for categorical variables. Reported estimates were not age-adjusted, and no adjustments were made for multiple comparisons. To evaluate the association between NAFLD and T2DM on cognitive performance, multivariate analysis was applied while adjusting for demographic disparity, comorbidities, and metabolic components
CR IP T
in a hierarchical fashion. Linear regressions for continuous outcome and logistic regression for binary outcome were used. All analyses were performed with SAS software, version 9.4 (SAS Institute, Cary, NC). Statistical tests were considered significant at p <0.05 (two-tailed).
AN US
RESULTS
Of the 19,931 individuals in NHANES 2011-2014 cycles, 1,102 were considered eligible for the study (mean age 69.4 years, 46.2% male, 79.7% white) (Table 1). In short, 49.6% of those eligible did not have NAFLD or T2DM, 21.7% had NAFLD only, 12.9% had T2DM only, and 15.8% had both
characteristics are detailed in Table 1.
M
NAFLD and T2DM. Statistically significant differences between these groups on demographic
ED
CERAD-WL – Immediate and Delayed Verbal Memory Immediate Verbal Memory. The weighted mean of the CERAD-WL immediate score was
PT
significantly lower in the NAFLD+T2DM group (19.20.5) than in the NAFLD only (20.50.4) group
CE
(Table 2). The T2DM only group (18.90.5) scored statistically significantly lower than NAFLD only and control groups (20.10.3) (Table 2). Multivariate regression showed that NAFLD and T2DM status were
AC
not associated with CERAD-WL immediate score. Delayed Verbal Memory. The weighted mean of the CERAD-WL delayed score was significantly
lower in the T2DM only group (5.70.3) than in the NAFLD only group (6.60.2) and the control group (6.40.1) (Table 2). Multivariate regression showed that NAFLD and T2DM status were not associated with performance on the CERAD-WL delayed. AFT – Verbal Fluency (Executive Function) 6
ACCEPTED MANUSCRIPT
The weighted mean of the AFT score was significantly lower in the NAFLD+T2DM group (16.80.5) than in the NAFLD only (18.60.3) and control groups (18.30.4)0 (Table 2). Also, the weighted mean was significantly lower in the T2DM only group (15.90.5) than the NAFLD only and control groups. Multivariate regression showed that T2DM status, age, race, and poverty ratio were
CR IP T
significantly associated with AFT scores (all p<0.05) (Table 3). After adjusting for demographic disparity, comorbidities, and metabolic components, those with T2DM scored statistically significantly lower on AFT compared to controls (Beta=-1.47, 95% CI: -2.7 - -0.23).
DSST – Processing Speed, Sustained Attention, and Working Memory
AN US
The weighted mean of the DSST score was significantly lower in the NAFLD+T2DM group than in the T2DM only and control group (47.11.7 vs 56.01.1 and 53.61.2, respectively) (Table 2). Multivariate regression showed that NAFLD and T2DM status, race, and poverty ratio were significantly associated with DSST score (all p<0.001) (Table 3). After adjusting for demographic disparity,
M
comorbidities, and metabolic components, those with NAFLD+T2DM scored statistically significantly lower on DSST compared to the control group (Beta=-3.44, 95% CI: -6.75 - -0.12). In other words,
ED
compared to controls, the coexistence of NAFLD+T2DM was significantly associated with a lower score
DISCUSSION
PT
on DSST.
CE
This study was undertaken to determine if patients with NAFLD with or without T2DM have impairment of cognitive performance as compared to individuals with T2DM alone, and individuals with
AC
neither. Although there is a rich literature examining the cognitive performance of individuals with T2DM14, there is a lack of information on the cognitive performance of individuals with NAFLD. We chose to use the NHANES survey (2011-2012 and 2013-2014) because those cycles included cognitive performance testing. These data provide detailed information on a few of the most frequently measured cognitive domains. The data, however, are only a small portion of the complexity of human cognition and
7
ACCEPTED MANUSCRIPT
only assess a small percentage of potential cognitive domains. These data provide the first assessment of cognitive performance in those with NAFLD compared to individuals with T2DM. The NHANES survey also offers a large cohort for study with standardized demographic, anthropometric, and laboratory measures that enables us to examine data from multiple domains that may
CR IP T
influence or are influenced by cognitive function. In the current sample, after controlling for potential confounding variables, there were no differences found in the verbal memory (immediate and delayed) domain. However, individuals with T2DM performed worse on verbal fluency and individuals with both NAFLD and T2DM performed worse on a task that involves processing speed, sustained attention, and
AN US
working memory than those without T2DM. This demonstrates the importance of the type of cognitive domain assessed when investigating potential cognitive performance deficits. Considering the domains assessed in this sample, individuals with NAFLD only did not demonstrate cognitive performance impairments, but those with T2DM only did.
M
From further examination of our sample, those with NAFLD only had lower rates of the metabolic syndrome (59%) compared to T2DM only (78%) and those with NAFLD+T2DM (98%). Since
ED
metabolic syndrome is a cluster of conditions (hypertension, hyperglycemia, excess visceral fat, and abnormal cholesterol/triglyceride levels), detectable cognitive performance deficits are likely dependent
PT
upon multiple pathophysiological abnormalities. These may in fact, be additive. Therefore, those with NAFLD only have a lower likelihood of impaired cognitive performance than those with T2DM only and
CE
those with NAFLD+T2DM. The mechanism for the cognitive performance decrement is, as yet unknown and may well be the combination of vascular and lipid abnormalities.
AC
In addition, NAFLD+T2DM group performed significantly worse on a task that requires a
combination of processing speed, sustained attention, and working memory. Possible explanations for this are at best speculative at this time, but the findings do suggest that NAFLD may affect brain function through region-specific processes rather than diffuse cortical dysfunction and it is related to the mechanism of cognitive performance change related to T2DM, as there is an additive effect. Additional studies are needed to examine these processes further. It is also possible that insulin resistance might play 8
ACCEPTED MANUSCRIPT
a crucial role in the relationship between NAFLD and cognitive performance, which is why the current investigation also focused on those with T2DM. Insulin resistance is a common factor in individuals with NAFLD, individuals with T2DM, and individuals with diseases that have cognitive impairment as a hallmark symptom, such as Alzheimer’s Disease3,26. In a study utilizing an animal model, it was
CR IP T
suggested that increase insulin resistance (induced by nitrosamine) lead to non-alcoholic steatohepatitis and Alzheimer’s Disease27.
The strengths of the current study include the utilization of NHANES data, which is a large and nationally representative sample of adults with standardized administration of three well-validated
AN US
cognitive performance tests and a thorough assessment of potential demographic, metabolic, and other factors that can affect cognitive performance. There are limitations to the current investigation. One of the weaknesses is the lack of histological confirmation of NAFLD. We used the FLI score as our definition for NAFLD. Although it is considered to have a good predictive value for NAFLD28,29, we may have
M
underestimated the adverse association between NAFLD and cognitive performance by utilizing this definition. Our data indicate that individuals considered to have NAFLD in our sample are comparable to
ED
previously reported characteristics of patients histologically diagnosed with NAFLD. Another limitation of the current investigation was the limited nature of the cognitive tests. There were three tests conducted
PT
and that does not allow a thorough investigation of a range of cognitive performance and domains. Future research should use a more diverse set of cognitive performance tests to determine potential deficits in
CE
other cognitive domains for individuals with NAFLD. Further, the cognitive tests were only conducted in individuals aged 60 and over. Therefore, the current results cannot be generalized to younger individuals.
AC
In summary, this analysis of a large population-based cohort suggests that individuals with
NAFLD and T2DM have lower cognitive performance. These findings are of particular importance given the rising incidence of obesity and metabolic syndrome, both risk factors for NAFLD and T2DM. The presence of NAFLD without T2DM did not appear to be associated with lower cognitive performance. These data provide support for a therapeutic approach that aims to optimally manage T2DM in patients with NAFLD, not only for hepatic and cardiovascular risks but also for their cognitive performance. 9
ACCEPTED MANUSCRIPT
Future research needs to extend the current findings, both in terms of comprehensive cognitive testing and potential pathophysiological mechanisms that help to explain the connections between these conditions and cognitive performance. A focus of future research should be on the potential factors that explain the relationship between NAFLD, T2DM, and cognitive performance, such as the cluster of risk factors
CR IP T
included in the metabolic syndrome, insulin resistance, inflammatory processes or microvascular
AC
CE
PT
ED
M
AN US
abnormalities.
10
ACCEPTED MANUSCRIPT
REFERENCES
AC
CE
PT
ED
M
AN US
CR IP T
1 Argo CK, Caldwell SH: Epidemiology and natural history of non-alcoholic steatohepatitis. Clin Liver Dis 2009; 13(4): 511–31. Doi: 10.1016/j.cld.2009.07.005. 2 Lazo M, Clark JM: The epidemiology of nonalcoholic fatty liver disease: a global perspective. Semin Liver Dis 2008; 28(4): 339–50. Doi: 10.1055/s-0028-1091978. 3 Targher G, Bertolini L, Poli F, et al: Nonalcoholic fatty liver disease and risk of future cardiovascular events among type 2 diabetic patients. Diabetes 2005; 54(12): 3541–6. 4 Stepanova M, Younossi ZM: Independent association between nonalcoholic fatty liver disease and cardiovascular disease in the US population. Clin Gastroenterol Hepatol Off Clin Pract J Am Gastroenterol Assoc 2012; 10(6): 646–50. Doi: 10.1016/j.cgh.2011.12.039. 5 Peila R, Rodriguez BL, Launer LJ, Honolulu-Asia Aging Study: Type 2 diabetes, APOE gene, and the risk for dementia and related pathologies: The Honolulu-Asia Aging Study. Diabetes 2002; 51(4): 1256–62. 6 Petrovitch H, White LR, Izmirilian G, et al: Midlife blood pressure and neuritic plaques, neurofibrillary tangles, and brain weight at death: the HAAS. Honolulu-Asia aging Study. Neurobiol Aging 2000; 21(1): 57–62. 7 Firneisz G: Non-alcoholic fatty liver disease and type 2 diabetes mellitus: the liver disease of our age? World J Gastroenterol 2014; 20(27): 9072–89. Doi: 10.3748/wjg.v20.i27.9072. 8 Younossi ZM, Koenig AB, Abdelatif D, Fazel Y, Henry L, Wymer M: Global epidemiology of nonalcoholic fatty liver disease-Meta-analytic assessment of prevalence, incidence, and outcomes. Hepatol Baltim Md 2016; 64(1): 73–84. Doi: 10.1002/hep.28431. 9 Ballestri S, Nascimbeni F, Romagnoli D, Baldelli E, Targher G, Lonardo A: Type 2 Diabetes in NonAlcoholic Fatty Liver Disease and Hepatitis C Virus Infection--Liver: The “Musketeer” in the Spotlight. Int J Mol Sci 2016; 17(3): 355. Doi: 10.3390/ijms17030355. 10 Neuschwander-Tetri BA: Non-alcoholic fatty liver disease. BMC Med 2017; 15(1): 45. Doi: 10.1186/s12916-017-0806-8. 11 Ryan CM, van Duinkerken E, Rosano C: Neurocognitive consequences of diabetes. Am Psychol 2016; 71(7): 563–76. Doi: 10.1037/a0040455. 12 Monette MCE, Baird A, Jackson DL: A meta-analysis of cognitive functioning in nondemented adults with type 2 diabetes mellitus. Can J Diabetes 2014; 38(6): 401–8. Doi: 10.1016/j.jcjd.2014.01.014. 13 Cui Y, Jiao Y, Chen H-J, et al: Aberrant functional connectivity of default-mode network in type 2 diabetes patients. Eur Radiol 2015; 25(11): 3238–46. Doi: 10.1007/s00330-015-3746-8. 14 Palta P, Schneider ALC, Biessels GJ, Touradji P, Hill-Briggs F: Magnitude of cognitive dysfunction in adults with type 2 diabetes: a meta-analysis of six cognitive domains and the most frequently reported neuropsychological tests within domains. J Int Neuropsychol Soc JINS 2014; 20(3): 278–91. Doi: 10.1017/S1355617713001483. 15 Yaffe K, Falvey C, Hamilton N, et al: Diabetes, glucose control, and 9-year cognitive decline among older adults without dementia. Arch Neurol 2012; 69(9): 1170–5. Doi: 10.1001/archneurol.2012.1117. 16 Vincent C, Hall PA: Executive Function in Adults With Type 2 Diabetes: A Meta-Analytic Review. Psychosom Med 2015; 77(6): 631–42. Doi: 10.1097/PSY.0000000000000103. 17 Rucker JL, McDowd JM, Kluding PM: Executive function and type 2 diabetes: putting the pieces together. Phys Ther 2012; 92(3): 454–62. Doi: 10.2522/ptj.20100397. 18 Chung C-C, Pimentel D, Jor’dan AJ, Hao Y, Milberg W, Novak V: Inflammation-associated declines in cerebral vasoreactivity and cognition in type 2 diabetes. Neurology 2015; 85(5): 450–8. Doi: 10.1212/WNL.0000000000001820. 19 Seo SW, Gottesman RF, Clark JM, et al: Nonalcoholic fatty liver disease is associated with cognitive function in adults. Neurology 2016; 86(12): 1136–42. Doi: 10.1212/WNL.0000000000002498.
11
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
M
AN US
CR IP T
20 Centers for Disease Control and Prevention: National Health and Nutrition Examination Survey: Analytic Guidelines, 2011-2012 2013. 21 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 Gastroenterol 2006; 6: 33. Doi: 10.1186/1471-230X6-33. 22 Morris JC, Heyman A, Mohs RC, et al: The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part I. Clinical and neuropsychological assessment of Alzheimer’s disease. Neurology 1989; 39(9): 1159–65. 23 Strauss EH, Sherman EMS, Spreen O: A Compendium of Neuropsychological Tests: Administration, Norms, And Commentary. Oxford University Press; 2006. 24 Wechsler D: WAIS-III: Administration and scoring manual : Wechsler adult intelligence scale--third edition. 3rd edition. Psychological Corporation; 1997. 25 National Center for Health Statistics: Office of Analysis and Epidemiology, Public-use Linked Mortality File. Hyattsville, MD; 2015. 26 Craft S: Insulin resistance and Alzheimer’s disease pathogenesis: potential mechanisms and implications for treatment. Curr Alzheimer Res 2007; 4(2): 147–52. 27 Tong M, Neusner A, Longato L, Lawton M, Wands JR, de la Monte SM: Nitrosamine exposure causes insulin resistance diseases: relevance to type 2 diabetes mellitus, non-alcoholic steatohepatitis, and Alzheimer’s disease. J Alzheimers Dis JAD 2009; 17(4): 827–44. 28 Angulo P: Nonalcoholic fatty liver disease. N Engl J Med 2002; 346(16): 1221–31. Doi: 10.1056/NEJMra011775. 29 Calori G, Lattuada G, Ragogna F, et al: Fatty liver index and mortality: the Cremona study in the 15th year of follow-up. Hepatol Baltim Md 2011; 54(1): 145–52. Doi: 10.1002/hep.24356.
12
ACCEPTED MANUSCRIPT
ED
M
AN US
CR IP T
Figure Legend
AC
CE
PT
Fig. 1. Cohort Selection for the Current Analyses of the NHANES 2011-2014 sample. NHANES: National Health and Nutrition Examination Survey; NAFLD: Non-Alcoholic Fatty Liver Disease; T2DM: Type 2 Diabetes Mellitus; HIV: Human Immunodeficiency Virus; ALD: Alzheimer’s Disease, HBV: Hepatitis B virus; HCV: Hepatitis C Virus. Excessive alcohol consumption was defined as ≥ 20 g per day of alcohol in men and ≥ 10 g of alcohol per day in women.
13
ACCEPTED MANUSCRIPT
Table 1. Participant Characteristics by T2DM and NAFLD Status, NHANES (2011-2014) NAFLD+ T2DM
T2DM only
NAFLD only
(n=174)
(n=142)
(n=239)
69.43 (0.30) 46.21 (1.62) 28.46 (0.29)
68.81 (0.54) 47.76 (6.14) 34.95 (0.82)
70.61 (0.94) 51.51 (5.25) 25.58 (0.19)
68.58 (0.44) 50.07 (3.33) 33.05 (0.34)
69.8 (0.41) 42.98 (2.77) 25.00 (0.20)
79.69 (2.22) 6.99 (0.93) 3.39 (0.77) 4.86 (0.81) 3.92 (0.76) 1.15 (0.44) 29.14 (2.61) 66.37 (2.01) 93.60 (1.10)
76.59 (4.10) 10.02 (1.91) 6.24 (1.99) 1.67 (0.68) 5.06 (1.32) 0.42 (0.45) 18.45 (3.79) 62.8 (5.37) 94.91 (1.86)
63.88 (5.88) 11.44 (2.80) 5.8 (1.85) 11.83 (2.65) 6.02 (1.64) 1.03 (1.00) 22.08 (4.66) 67.51 (5.03) 94.52 (2.13)
83.52 (2.47) 5.67 (1.17) 3.83 (1.03) 1.82 (0.65) 3.67 (0.98) 1.49 (0.90) 28.28 (4.08) 63.92 (3.56) 90.81 (3.04)
81.56 (2.37) 5.97 (0.89) 1.96 (0.57) 5.95 (1.14) 3.35 (0.73) 1.21 (0.57) 33.78 (3.12) 68.33 (2.65) 94.4 (1.20)
62.09 (2.95) 65.98 (4.81) 6.59 (2.11)
54.53 (5.69) 71.70 (6.02) 11.32 (3.74)
60.74 (3.16) 57.8 (4.20) 2.58 (0.89)
65.60 (3.09) 60.19 (2.69) 3.54 (1.00)
(n=1102)
Age Male 2
BMI (kg/m )
Mexican Asian Hispanic Other College Degree Married Insured Insurance Type
62.87 (2.32) 61.5 (2.13) 4.46 (0.97)
PT
Private Insurance
CE
Medicare
(n=239)
P value <.001 0.391
<.001 0.003
<.001 <.001 0.083 0.789 0.004 0.619 0.355
0.133 0.197 <.001
AC
Medicaid
M
Black
ED
White
AN US
Race
Controls
CR IP T
All
14
ACCEPTED MANUSCRIPT
Income
Medium High
17.03 (1.76) 38.28 (2.27) 44.69 (3.13)
27.8 (4.92) 40.24 (4.47) 31.95 (5.57)
26.31 (4.43) 40.36 (6.28) 33.32 (6.09)
13.54 (2.61) 40.7 (5.11) 45.76 (4.75)
14.12 (2.08) 36.22 (2.94) 49.66 (3.76)
9.31 (0.99) 36.44 (1.81) 46.21 (1.62)
7.37 (2.76) 37.14 (4.58) 47.76 (6.14)
5.83 (1.92) 32.97 (4.63) 51.51 (5.25)
9.22 (2.36) 42.06 (4.59) 50.07 (3.33)
10.52 (1.68) 34.19 (2.36) 42.98 (2.77)
41.39 (2.26) 40.44 (1.96) 46.21 (1.62)
23.09 (4.85) 45.34 (5.26) 47.76 (6.14)
28.28 (4.74) 39.77 (6.41) 51.51 (5.25)
37.09 (3.65) 48.41 (3.85) 50.07 (3.33)
51.17 (3.39) 35.21 (2.99) 42.98 (2.77)
27.11 (4.61) 55.16 (9.55) 83.37 (4.74) 86.27 (3.33) 77.87 (5.49) 34.93 (5.69) 2.35 (1.06) 19.31 (4.61) 41.82 (6.53)
92.53 (2.31) 62.12 (9.31) 76.11 (3.77) 82.34 (3.47) 58.50 (4.21) 53.57 (4.03) 8.05 (3.39) 13.80 (2.60) 26.56 (6.49)
19.43 (2.43) 17.49 (2.39) 61.99 (3.70) 74.16 (2.48) 11.09 (2.17) 43.22 (3.12) 3.27 (0.79) 11.45 (2.11) 22.75 (4.66)
Current Smoker Former Smoker Non-Smoker
AN US
Smoking Status
Excellent/Very Good Good Fair/Poor Comorbidities
Insulin Resistance
PT
Hypertension
48.84 (1.95) 40.44 (3.14) 70.97 (2.22) 78.03 (1.90) 41.30 (2.03) 47.89 (2.28) 5.94 (1.04) 15.71 (1.44) 28.28 (3.85)
ED
Obesity
High Cholesterol
CE
Metabolic Syndrome Arthritis
AC
Congestive Heart Failure Ischemic Heart Disease
Chronic Kidney Disease
M
Health Condition, self-reported
93.77 (2.8) 79.94 (6.09) 86.47 (4.05) 79.2 (4.68) 96.96 (2.58) 63.38 (4.56) 14.29 (3.11) 32.06 (4.20) 42.78 (8.24)
<.001 0.767
CR IP T
Low
0.007
0.567 0.311 0.391
<.001 0.043 0.391
<.001 <.001 <.001 0.056 <.001 <.001 <.001 <.001 0.025
Values were expressed as weighted mean/percentage (standard error)
15
ACCEPTED MANUSCRIPT
Table 2. Cognitive Performance by T2DM and NAFLD Status, NHANES (2011-2014)
Immediate verbal memory (CERAD-WL) Delayed verbal memory (CERAD-WL) Verbal fluency (AFT) Processing speed, sustained attention, working memory (DSST)
All
NAFLD+ T2DM
T2DM only
NAFLD only
Controls
19.92 (0.27) 6.34 (0.13) 17.96 (0.26) 52.50 (0.83)
19.19 (0.53) 6.11 (0.26) 16.84 (0.45) 47.12 (1.65)
18.88 (0.51) 5.67 (0.29) 15.91 (0.54) 45.08 (1.98)
20.45 (0.42) 6.63 (0.20) 18.57 (0.31) 55.99 (1.05)
20.05 (0.33) 6.38 (0.14) 18.33 (0.42) 53.60 (1.23)
CR IP T
Process Measure
AC
CE
PT
ED
M
AN US
Values were expressed as weighted mean (standard error) CERAD-WL: Consortium to Establish a Registry for Alzheimer’s Disease-Word List; AFT: Animal Fluency Test; DSST: Digit Symbol Substitution Test
16
ACCEPTED MANUSCRIPT
Table 3. Association between T2DM and NAFLD status on AFT and DSST among U.S. adults aged 60 and over, NHANES (2011-2014)
Beta (95% CI)
Weighted mean ± SE
Model 1 (n=931)
Model 2 (n=918)
Reference
Reference
AFT None
18.33 ± 0.42
NAFLD only
18.57 ± 0.31
T2DM only
15.91 ± 0.54**
-1.40 (-2.68 - -0.12)*
-1.47 (-2.70 - -0.23)*
T2DM and NAFLD
16.84 ± 0.45**
-1.14 (-2.17 - -0.10)*
-0.69 (-2.08 - 0.70)
DSST None
53.60 ± 1.23
Reference
Reference
NAFLD only
55.99 ± 1.05
1.96 (-0.60 - 4.52)
T2DM only
45.08 ± 1.98**
T2DM and NAFLD
47.12 ± 1.65**
0.53 (-0.66 - 1.73)
CR IP T
0.03 (-0.87 - 0.93)
-3.81 (-8.07 - 0.45)
AN US
-3.69 (-7.00 - -0.39)*
1.82 (-1.81 - 5.46)
-3.66 (-8.06 - 0.74)
-3.44 (-6.75 - -0.12)*
AC
CE
PT
ED
M
* p <0.05 and **p <0.01 significantly different from None. SE, standard error Model 1: adjusted for Age, male, race, and income Model 2: adjusted for Model 1 + smoking status, presence of depression, asthma, arthritis, congestive heart failure, cancer, ischemic heart disease, chronic obstructive pulmonary disease, thyroid disease, psoriasis, obesity, hypertension, hypercholesterolemia. AFT= Animal Fluency Test DSST= Digit Symbol Substitution Test
17
Accepted Manuscript
Cognitive Performance in Individuals with Non-alcoholic Fatty Liver Disease and/or Type 2 Diabetes Mellitus Ali A. Weinstein Ph.D. , Leyla de Avila B.A. , James Paik Ph.D. , Pegah Golabi M.D. , Carey Escheik B.S. , Lynn Gerber M.D. , Zobair M. Younossi M.D., M.P.H. PII: DOI: Reference:
S0033-3182(18)30308-6 10.1016/j.psym.2018.06.001 PSYM 896
To appear in:
The End-to-end Journal
Received date: Revised date: Accepted date:
17 May 2018 31 May 2018 1 June 2018
Please cite this article as: Ali A. Weinstein Ph.D. , Leyla de Avila B.A. , James Paik Ph.D. , Pegah Golabi M.D. , Carey Escheik B.S. , Lynn Gerber M.D. , Zobair M. Younossi M.D., M.P.H. , Cognitive Performance in Individuals with Non-alcoholic Fatty Liver Disease and/or Type 2 Diabetes Mellitus, The End-to-end Journal (2018), doi: 10.1016/j.psym.2018.06.001
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Cognitive Performance in Individuals with Non-alcoholic Fatty Liver Disease and/or Type 2 Diabetes Mellitus
CR IP T
Ali A. Weinstein, Ph.D.1,2, Leyla de Avila, B.A.1, James Paik, Ph.D.1, Pegah Golabi, M.D. 1, Carey Escheik, B.S.1, Lynn Gerber, M.D.2,3, Zobair M. Younossi M.D., M.P.H. 1, 3
1
AN US
Betty and Guy Beatty Center for Integrated Research, Inova Health System, Falls Church, VA, United States. 2 Center for the Study of Chronic Illness and Disability, George Mason University, Fairfax, VA, United States. 3 Center for Liver Disease, Department of Medicine, Inova Fairfax Hospital, Falls Church, VA, United States.
M
Corresponding Author: Zobair M. Younossi, MD, MPH Betty and Guy Beatty Center for Integrated Research Claude Moore Health Education and Research Building 3300 Gallows Road, Falls Church, VA 22042 Phone: (703) 776-2540 Fax: (703) 776-4386 Email: [email protected]
Word count: 2497; Number of tables: 3; Number of figures: 1
ED
Conflict of Interest Statement: Dr. Younossi is a consultant to BMS, Gilead, AbbVie, Intercept, and GSK. All other authors have no conflict of interest to disclose.
PT
Financial Support Statement: This work was supported by the Betty and Guy Beatty Center for
CE
Integrated Research.
Authors contributions: AW, LD, CE and PG participated in the study design, helped the interpretation
AC
of the data and to draft the manuscript. JP performed the statistical analysis and helped with the interpretation of the data. LG and ZMY conceived of the study, participated substantially in its design and coordination, and helped to draft the manuscript. All authors read and approved the final manuscript.
1
ACCEPTED MANUSCRIPT
ABSTRACT Background: Individuals with Non-Alcoholic Fatty Liver Disease (NAFLD) share some common pathophysiological features with individuals with type 2 diabetes mellitus (T2DM). There is a wellestablished association between T2DM and cognitive decline, but no corollary data in people with
CR IP T
NAFLD without T2DM and whether combination of the two disorders is associated with additive deficits in cognitive performance. The purpose of this investigation is to compare measures of cognitive
performance for individuals with NAFLD, individuals with T2DM, individuals with both or neither. Methods: Using NHANES data from 2011–2014, 1,102 individuals were identified that had completed
AN US
cognitive assessments. Cognitive performance was assessed by a series of well-established tasks.
Presence of T2DM was defined as the use of diabetes medication and/or a fasting glucose measure of > 126 mg/dL. Presence of NAFLD was defined as Fatty Liver Index≥60 in the absence of other chronic liver diseases. Results: After controlling for demographics, comorbidities, and metabolic components,
M
individuals with both NAFLD and T2DM scored statistically significantly lower on a task that combines processing speed, sustained attention, and working memory (Beta=-3.44, 95% CI: -6.75 - -0.12) than
ED
individuals with neither. Individuals with T2DM without NAFLD scored statistically significantly lower on verbal fluency (Beta=-1.47, 95% CI: -2.7 - -0.23) than individuals with neither. Conclusions: Data
PT
from this study suggest that individuals with T2DM and individuals with both NAFLD and T2DM have lower cognitive performance on various tasks. These data support an approach that aims to apply
CE
preventive strategies to optimize management of T2DM in patients with NAFLD.
AC
Keywords: cognitive performance; nutrition surveys; memory; insulin resistance; steatosis; NAFLD; diabetes
2
ACCEPTED MANUSCRIPT
INTRODUCTION In the United States and globally, non-alcoholic fatty liver disease (NAFLD) is the most common cause of chronic liver disease1,2. NAFLD is associated with cardiovascular disease (CVD) and its risk factors including type 2 diabetes mellitus (T2DM), obesity, hyperlipidemia, and hypertension3,4. These
CR IP T
risk factors are known to contribute to cognitive impairment with or without the presence of CVD5,6. NAFLD is physiologically related to T2DM, as there are shared risk factors linking these diseases. For example, both individuals with NAFLD and those with T2DM are more likely to be obese7. While the global prevalence of NAFLD is 25%8, in patients with T2DM the prevalence of NAFLD is considerably
AN US
higher, occurring in up to 65-75% of these individuals9. We have experienced the existence of this comorbidity in our clinic patients. It is rare for patients seen in our clinic to be overweight, have T2DM, and NOT have evidence of NAFLD. Therefore, we wished to explore relationships between NAFLD, T2DM and cognition, since cognitive performance impairment has been well documented in people with T2DM.
M
We sought to determine the relationship between NAFLD and cognitive performance as an independent association.
ED
Along with the common comorbidity between T2DM and NAFLD, the insulin resistance that is the hallmark of T2DM has also been tightly associated with the risk of NAFLD. Insulin is delivered
PT
directly to the portal vein after secretion and the liver eliminates a large portion of portal insulin at this first pass10. In addition, NAFLD is associated with dysfunctional adipose tissue, and lipo-toxicity
CE
promotes insulin resistance and pancreatic β-cell dysfunction. Therefore, there are common pathophysiological factors in both of these diseases10.
AC
The association between T2DM and an increased risk of cognitive decline is well established11-18.
Results of neurocognitive performance tests have demonstrated cognitive impairments in those with T2DM that exceed the rate of usual aging-associated cognitive decline11, specifically in the domains of attention12–15, psychomotor speed13–15, executive functions14,16,17, and learning and memory14,18. However, there is a lack of literature on the specific relationship between NAFLD and cognitive performance. We were able to identify only one study that investigated cognitive performance in those with NAFLD19. In 3
ACCEPTED MANUSCRIPT
that investigation, statistical manipulation was used to control for the presence of T2DM in those with NAFLD. Therefore, our aim was to compare patients with NAFLD, patients with T2DM, patients with both NAFLD and T2DM, and individuals with neither (control) on measures of cognitive performance in order to determine whether there was an association and possibly an additive performance impact when
CR IP T
both conditions co-existed.
PATIENTS AND METHODS Study population
AN US
Data were obtained from two continuous cycles of the National Health and Nutrition Examination Survey (NHANES 2011–2012 and NHANES 2013–2014). These data provide a cross-sectional, population-based sampling survey of the civilian non-institutionalized population of the United States. Data obtained from these surveys include both self-reports and objective clinical measures. Surveys were
M
based on a complex, multi-stage sampling plan. Detailed information about design and sampling of the survey is available20. Cognitive performance tests were only administered to individuals aged 60 and over,
ED
therefore participants under the age of 60 were excluded from analyses. In addition, participants were excluded if they were missing data needed to calculate NAFLD or T2DM status or a history of stroke,
PT
HIV, alcoholic liver disease (ALD), hepatitis B virus (HBV), hepatitis c virus (HCV), or excessive alcohol use (EAU). Figure 1 details the cohort selection for the current analyses.
CE
The following definitions were used to define groups for these analyses. T2DM was defined as the use of diabetes medication and/or a fasting glucose measure of > 126 mg/dL. NAFLD was defined by
AC
the absence of any other causes of chronic liver disease (e.g. ALD, HBV, HCV, or iron overload) as well as a Fatty Liver Index (FLI) score ≥ 6021. ALD was defined by EAU with elevated liver enzymes. EAU was defined as ≥ 20 g per day of alcohol in men and ≥ 10 g of alcohol per day in women. Elevated liver enzymes were defined as alanine aminotransferase (ALT) ≥ 40U/L in men and ≥ 31U/L in women or aspartate aminotransferase (AST) level ≥ 37U/L in men and ≥ 31U/L in women. HBV was defined as positive Hepatitis B surface antigen and HCV was defined as positive hepatitis C virus RNA. Iron 4
ACCEPTED MANUSCRIPT
overload was defined as serum transferrin saturation ≥ 50%. In addition, the following information was used for modeling purposes: (1) poverty income ratio (PIR) (PIR<1.3=low; 1.3
);
(3) hypertension (systolic blood pressure > 140 mmHG or diastolic blood pressure > 90 mmHg or history
CR IP T
of blood pressure); (4) metabolic syndrome (at least of the following present waist circumference >102 cm in men or >88 cm in women, fasting plasma glucose >110 mg/dl, BP >130/85 mm Hg, elevated triglycerides >150 mg/dl, and HDL ≤ 40 mg/dL in men or ≤ 50 mg/dL in women); and (5) insulin resistance (homeostasis of model assessment (HOMA) of > 3).
AN US
Cognitive Performance
A series of assessments in NHANES are used to examine cognitive performance. The word learning subset from the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD-WL) was used to assess immediate and delayed learning ability for new verbal information22. The animal fluency
M
test (AFT) examines categorical verbal fluency, a component of executive function23. The digit symbol substitution test (DSST), a performance module from the Wechsler Adult Intelligence Scale (WAIS III) is
Statistical Analysis
ED
a test that involves processing speed, sustained attention, and working memory24.
PT
Survey design elements (Clusters, Strata, and examination sample weights) provided by the National Center for Health Statistics (NCHS) were used to account for survey non-response, noncoverage,
CE
and sampling strategy. For national representation, original weights in our merged sample were modified using the method recommended by NCHS25. Taylor series linearization was used to calculate standard
AC
errors to take into account the complex sampling design. This study compares cognitive performance in individuals with T2DM and NAFLD to cognitive
performance in a group of individuals with neither T2DM or NAFLD from the same age-matched study cohort. The prevalence of various characteristics including demographic, clinical parameters, and outcomes across study groups was compared using a t-statistic for continuous variables and the Rao-Scott
5
ACCEPTED MANUSCRIPT
chi-square test for categorical variables. Reported estimates were not age-adjusted, and no adjustments were made for multiple comparisons. To evaluate the association between NAFLD and T2DM on cognitive performance, multivariate analysis was applied while adjusting for demographic disparity, comorbidities, and metabolic components
CR IP T
in a hierarchical fashion. Linear regressions for continuous outcome and logistic regression for binary outcome were used. All analyses were performed with SAS software, version 9.4 (SAS Institute, Cary, NC). Statistical tests were considered significant at p <0.05 (two-tailed).
AN US
RESULTS
Of the 19,931 individuals in NHANES 2011-2014 cycles, 1,102 were considered eligible for the study (mean age 69.4 years, 46.2% male, 79.7% white) (Table 1). In short, 49.6% of those eligible did not have NAFLD or T2DM, 21.7% had NAFLD only, 12.9% had T2DM only, and 15.8% had both
characteristics are detailed in Table 1.
M
NAFLD and T2DM. Statistically significant differences between these groups on demographic
ED
CERAD-WL – Immediate and Delayed Verbal Memory Immediate Verbal Memory. The weighted mean of the CERAD-WL immediate score was
PT
significantly lower in the NAFLD+T2DM group (19.20.5) than in the NAFLD only (20.50.4) group
CE
(Table 2). The T2DM only group (18.90.5) scored statistically significantly lower than NAFLD only and control groups (20.10.3) (Table 2). Multivariate regression showed that NAFLD and T2DM status were
AC
not associated with CERAD-WL immediate score. Delayed Verbal Memory. The weighted mean of the CERAD-WL delayed score was significantly
lower in the T2DM only group (5.70.3) than in the NAFLD only group (6.60.2) and the control group (6.40.1) (Table 2). Multivariate regression showed that NAFLD and T2DM status were not associated with performance on the CERAD-WL delayed. AFT – Verbal Fluency (Executive Function) 6
ACCEPTED MANUSCRIPT
The weighted mean of the AFT score was significantly lower in the NAFLD+T2DM group (16.80.5) than in the NAFLD only (18.60.3) and control groups (18.30.4)0 (Table 2). Also, the weighted mean was significantly lower in the T2DM only group (15.90.5) than the NAFLD only and control groups. Multivariate regression showed that T2DM status, age, race, and poverty ratio were
CR IP T
significantly associated with AFT scores (all p<0.05) (Table 3). After adjusting for demographic disparity, comorbidities, and metabolic components, those with T2DM scored statistically significantly lower on AFT compared to controls (Beta=-1.47, 95% CI: -2.7 - -0.23).
DSST – Processing Speed, Sustained Attention, and Working Memory
AN US
The weighted mean of the DSST score was significantly lower in the NAFLD+T2DM group than in the T2DM only and control group (47.11.7 vs 56.01.1 and 53.61.2, respectively) (Table 2). Multivariate regression showed that NAFLD and T2DM status, race, and poverty ratio were significantly associated with DSST score (all p<0.001) (Table 3). After adjusting for demographic disparity,
M
comorbidities, and metabolic components, those with NAFLD+T2DM scored statistically significantly lower on DSST compared to the control group (Beta=-3.44, 95% CI: -6.75 - -0.12). In other words,
ED
compared to controls, the coexistence of NAFLD+T2DM was significantly associated with a lower score
DISCUSSION
PT
on DSST.
CE
This study was undertaken to determine if patients with NAFLD with or without T2DM have impairment of cognitive performance as compared to individuals with T2DM alone, and individuals with
AC
neither. Although there is a rich literature examining the cognitive performance of individuals with T2DM14, there is a lack of information on the cognitive performance of individuals with NAFLD. We chose to use the NHANES survey (2011-2012 and 2013-2014) because those cycles included cognitive performance testing. These data provide detailed information on a few of the most frequently measured cognitive domains. The data, however, are only a small portion of the complexity of human cognition and
7
ACCEPTED MANUSCRIPT
only assess a small percentage of potential cognitive domains. These data provide the first assessment of cognitive performance in those with NAFLD compared to individuals with T2DM. The NHANES survey also offers a large cohort for study with standardized demographic, anthropometric, and laboratory measures that enables us to examine data from multiple domains that may
CR IP T
influence or are influenced by cognitive function. In the current sample, after controlling for potential confounding variables, there were no differences found in the verbal memory (immediate and delayed) domain. However, individuals with T2DM performed worse on verbal fluency and individuals with both NAFLD and T2DM performed worse on a task that involves processing speed, sustained attention, and
AN US
working memory than those without T2DM. This demonstrates the importance of the type of cognitive domain assessed when investigating potential cognitive performance deficits. Considering the domains assessed in this sample, individuals with NAFLD only did not demonstrate cognitive performance impairments, but those with T2DM only did.
M
From further examination of our sample, those with NAFLD only had lower rates of the metabolic syndrome (59%) compared to T2DM only (78%) and those with NAFLD+T2DM (98%). Since
ED
metabolic syndrome is a cluster of conditions (hypertension, hyperglycemia, excess visceral fat, and abnormal cholesterol/triglyceride levels), detectable cognitive performance deficits are likely dependent
PT
upon multiple pathophysiological abnormalities. These may in fact, be additive. Therefore, those with NAFLD only have a lower likelihood of impaired cognitive performance than those with T2DM only and
CE
those with NAFLD+T2DM. The mechanism for the cognitive performance decrement is, as yet unknown and may well be the combination of vascular and lipid abnormalities.
AC
In addition, NAFLD+T2DM group performed significantly worse on a task that requires a
combination of processing speed, sustained attention, and working memory. Possible explanations for this are at best speculative at this time, but the findings do suggest that NAFLD may affect brain function through region-specific processes rather than diffuse cortical dysfunction and it is related to the mechanism of cognitive performance change related to T2DM, as there is an additive effect. Additional studies are needed to examine these processes further. It is also possible that insulin resistance might play 8
ACCEPTED MANUSCRIPT
a crucial role in the relationship between NAFLD and cognitive performance, which is why the current investigation also focused on those with T2DM. Insulin resistance is a common factor in individuals with NAFLD, individuals with T2DM, and individuals with diseases that have cognitive impairment as a hallmark symptom, such as Alzheimer’s Disease3,26. In a study utilizing an animal model, it was
CR IP T
suggested that increase insulin resistance (induced by nitrosamine) lead to non-alcoholic steatohepatitis and Alzheimer’s Disease27.
The strengths of the current study include the utilization of NHANES data, which is a large and nationally representative sample of adults with standardized administration of three well-validated
AN US
cognitive performance tests and a thorough assessment of potential demographic, metabolic, and other factors that can affect cognitive performance. There are limitations to the current investigation. One of the weaknesses is the lack of histological confirmation of NAFLD. We used the FLI score as our definition for NAFLD. Although it is considered to have a good predictive value for NAFLD28,29, we may have
M
underestimated the adverse association between NAFLD and cognitive performance by utilizing this definition. Our data indicate that individuals considered to have NAFLD in our sample are comparable to
ED
previously reported characteristics of patients histologically diagnosed with NAFLD. Another limitation of the current investigation was the limited nature of the cognitive tests. There were three tests conducted
PT
and that does not allow a thorough investigation of a range of cognitive performance and domains. Future research should use a more diverse set of cognitive performance tests to determine potential deficits in
CE
other cognitive domains for individuals with NAFLD. Further, the cognitive tests were only conducted in individuals aged 60 and over. Therefore, the current results cannot be generalized to younger individuals.
AC
In summary, this analysis of a large population-based cohort suggests that individuals with
NAFLD and T2DM have lower cognitive performance. These findings are of particular importance given the rising incidence of obesity and metabolic syndrome, both risk factors for NAFLD and T2DM. The presence of NAFLD without T2DM did not appear to be associated with lower cognitive performance. These data provide support for a therapeutic approach that aims to optimally manage T2DM in patients with NAFLD, not only for hepatic and cardiovascular risks but also for their cognitive performance. 9
ACCEPTED MANUSCRIPT
Future research needs to extend the current findings, both in terms of comprehensive cognitive testing and potential pathophysiological mechanisms that help to explain the connections between these conditions and cognitive performance. A focus of future research should be on the potential factors that explain the relationship between NAFLD, T2DM, and cognitive performance, such as the cluster of risk factors
CR IP T
included in the metabolic syndrome, insulin resistance, inflammatory processes or microvascular
AC
CE
PT
ED
M
AN US
abnormalities.
10
ACCEPTED MANUSCRIPT
REFERENCES
AC
CE
PT
ED
M
AN US
CR IP T
1 Argo CK, Caldwell SH: Epidemiology and natural history of non-alcoholic steatohepatitis. Clin Liver Dis 2009; 13(4): 511–31. Doi: 10.1016/j.cld.2009.07.005. 2 Lazo M, Clark JM: The epidemiology of nonalcoholic fatty liver disease: a global perspective. Semin Liver Dis 2008; 28(4): 339–50. Doi: 10.1055/s-0028-1091978. 3 Targher G, Bertolini L, Poli F, et al: Nonalcoholic fatty liver disease and risk of future cardiovascular events among type 2 diabetic patients. Diabetes 2005; 54(12): 3541–6. 4 Stepanova M, Younossi ZM: Independent association between nonalcoholic fatty liver disease and cardiovascular disease in the US population. Clin Gastroenterol Hepatol Off Clin Pract J Am Gastroenterol Assoc 2012; 10(6): 646–50. Doi: 10.1016/j.cgh.2011.12.039. 5 Peila R, Rodriguez BL, Launer LJ, Honolulu-Asia Aging Study: Type 2 diabetes, APOE gene, and the risk for dementia and related pathologies: The Honolulu-Asia Aging Study. Diabetes 2002; 51(4): 1256–62. 6 Petrovitch H, White LR, Izmirilian G, et al: Midlife blood pressure and neuritic plaques, neurofibrillary tangles, and brain weight at death: the HAAS. Honolulu-Asia aging Study. Neurobiol Aging 2000; 21(1): 57–62. 7 Firneisz G: Non-alcoholic fatty liver disease and type 2 diabetes mellitus: the liver disease of our age? World J Gastroenterol 2014; 20(27): 9072–89. Doi: 10.3748/wjg.v20.i27.9072. 8 Younossi ZM, Koenig AB, Abdelatif D, Fazel Y, Henry L, Wymer M: Global epidemiology of nonalcoholic fatty liver disease-Meta-analytic assessment of prevalence, incidence, and outcomes. Hepatol Baltim Md 2016; 64(1): 73–84. Doi: 10.1002/hep.28431. 9 Ballestri S, Nascimbeni F, Romagnoli D, Baldelli E, Targher G, Lonardo A: Type 2 Diabetes in NonAlcoholic Fatty Liver Disease and Hepatitis C Virus Infection--Liver: The “Musketeer” in the Spotlight. Int J Mol Sci 2016; 17(3): 355. Doi: 10.3390/ijms17030355. 10 Neuschwander-Tetri BA: Non-alcoholic fatty liver disease. BMC Med 2017; 15(1): 45. Doi: 10.1186/s12916-017-0806-8. 11 Ryan CM, van Duinkerken E, Rosano C: Neurocognitive consequences of diabetes. Am Psychol 2016; 71(7): 563–76. Doi: 10.1037/a0040455. 12 Monette MCE, Baird A, Jackson DL: A meta-analysis of cognitive functioning in nondemented adults with type 2 diabetes mellitus. Can J Diabetes 2014; 38(6): 401–8. Doi: 10.1016/j.jcjd.2014.01.014. 13 Cui Y, Jiao Y, Chen H-J, et al: Aberrant functional connectivity of default-mode network in type 2 diabetes patients. Eur Radiol 2015; 25(11): 3238–46. Doi: 10.1007/s00330-015-3746-8. 14 Palta P, Schneider ALC, Biessels GJ, Touradji P, Hill-Briggs F: Magnitude of cognitive dysfunction in adults with type 2 diabetes: a meta-analysis of six cognitive domains and the most frequently reported neuropsychological tests within domains. J Int Neuropsychol Soc JINS 2014; 20(3): 278–91. Doi: 10.1017/S1355617713001483. 15 Yaffe K, Falvey C, Hamilton N, et al: Diabetes, glucose control, and 9-year cognitive decline among older adults without dementia. Arch Neurol 2012; 69(9): 1170–5. Doi: 10.1001/archneurol.2012.1117. 16 Vincent C, Hall PA: Executive Function in Adults With Type 2 Diabetes: A Meta-Analytic Review. Psychosom Med 2015; 77(6): 631–42. Doi: 10.1097/PSY.0000000000000103. 17 Rucker JL, McDowd JM, Kluding PM: Executive function and type 2 diabetes: putting the pieces together. Phys Ther 2012; 92(3): 454–62. Doi: 10.2522/ptj.20100397. 18 Chung C-C, Pimentel D, Jor’dan AJ, Hao Y, Milberg W, Novak V: Inflammation-associated declines in cerebral vasoreactivity and cognition in type 2 diabetes. Neurology 2015; 85(5): 450–8. Doi: 10.1212/WNL.0000000000001820. 19 Seo SW, Gottesman RF, Clark JM, et al: Nonalcoholic fatty liver disease is associated with cognitive function in adults. Neurology 2016; 86(12): 1136–42. Doi: 10.1212/WNL.0000000000002498.
11
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
M
AN US
CR IP T
20 Centers for Disease Control and Prevention: National Health and Nutrition Examination Survey: Analytic Guidelines, 2011-2012 2013. 21 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 Gastroenterol 2006; 6: 33. Doi: 10.1186/1471-230X6-33. 22 Morris JC, Heyman A, Mohs RC, et al: The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part I. Clinical and neuropsychological assessment of Alzheimer’s disease. Neurology 1989; 39(9): 1159–65. 23 Strauss EH, Sherman EMS, Spreen O: A Compendium of Neuropsychological Tests: Administration, Norms, And Commentary. Oxford University Press; 2006. 24 Wechsler D: WAIS-III: Administration and scoring manual : Wechsler adult intelligence scale--third edition. 3rd edition. Psychological Corporation; 1997. 25 National Center for Health Statistics: Office of Analysis and Epidemiology, Public-use Linked Mortality File. Hyattsville, MD; 2015. 26 Craft S: Insulin resistance and Alzheimer’s disease pathogenesis: potential mechanisms and implications for treatment. Curr Alzheimer Res 2007; 4(2): 147–52. 27 Tong M, Neusner A, Longato L, Lawton M, Wands JR, de la Monte SM: Nitrosamine exposure causes insulin resistance diseases: relevance to type 2 diabetes mellitus, non-alcoholic steatohepatitis, and Alzheimer’s disease. J Alzheimers Dis JAD 2009; 17(4): 827–44. 28 Angulo P: Nonalcoholic fatty liver disease. N Engl J Med 2002; 346(16): 1221–31. Doi: 10.1056/NEJMra011775. 29 Calori G, Lattuada G, Ragogna F, et al: Fatty liver index and mortality: the Cremona study in the 15th year of follow-up. Hepatol Baltim Md 2011; 54(1): 145–52. Doi: 10.1002/hep.24356.
12
ACCEPTED MANUSCRIPT
ED
M
AN US
CR IP T
Figure Legend
AC
CE
PT
Fig. 1. Cohort Selection for the Current Analyses of the NHANES 2011-2014 sample. NHANES: National Health and Nutrition Examination Survey; NAFLD: Non-Alcoholic Fatty Liver Disease; T2DM: Type 2 Diabetes Mellitus; HIV: Human Immunodeficiency Virus; ALD: Alzheimer’s Disease, HBV: Hepatitis B virus; HCV: Hepatitis C Virus. Excessive alcohol consumption was defined as ≥ 20 g per day of alcohol in men and ≥ 10 g of alcohol per day in women.
13
ACCEPTED MANUSCRIPT
Table 1. Participant Characteristics by T2DM and NAFLD Status, NHANES (2011-2014) NAFLD+ T2DM
T2DM only
NAFLD only
(n=174)
(n=142)
(n=239)
69.43 (0.30) 46.21 (1.62) 28.46 (0.29)
68.81 (0.54) 47.76 (6.14) 34.95 (0.82)
70.61 (0.94) 51.51 (5.25) 25.58 (0.19)
68.58 (0.44) 50.07 (3.33) 33.05 (0.34)
69.8 (0.41) 42.98 (2.77) 25.00 (0.20)
79.69 (2.22) 6.99 (0.93) 3.39 (0.77) 4.86 (0.81) 3.92 (0.76) 1.15 (0.44) 29.14 (2.61) 66.37 (2.01) 93.60 (1.10)
76.59 (4.10) 10.02 (1.91) 6.24 (1.99) 1.67 (0.68) 5.06 (1.32) 0.42 (0.45) 18.45 (3.79) 62.8 (5.37) 94.91 (1.86)
63.88 (5.88) 11.44 (2.80) 5.8 (1.85) 11.83 (2.65) 6.02 (1.64) 1.03 (1.00) 22.08 (4.66) 67.51 (5.03) 94.52 (2.13)
83.52 (2.47) 5.67 (1.17) 3.83 (1.03) 1.82 (0.65) 3.67 (0.98) 1.49 (0.90) 28.28 (4.08) 63.92 (3.56) 90.81 (3.04)
81.56 (2.37) 5.97 (0.89) 1.96 (0.57) 5.95 (1.14) 3.35 (0.73) 1.21 (0.57) 33.78 (3.12) 68.33 (2.65) 94.4 (1.20)
62.09 (2.95) 65.98 (4.81) 6.59 (2.11)
54.53 (5.69) 71.70 (6.02) 11.32 (3.74)
60.74 (3.16) 57.8 (4.20) 2.58 (0.89)
65.60 (3.09) 60.19 (2.69) 3.54 (1.00)
(n=1102)
Age Male 2
BMI (kg/m )
Mexican Asian Hispanic Other College Degree Married Insured Insurance Type
62.87 (2.32) 61.5 (2.13) 4.46 (0.97)
PT
Private Insurance
CE
Medicare
(n=239)
P value <.001 0.391
<.001 0.003
<.001 <.001 0.083 0.789 0.004 0.619 0.355
0.133 0.197 <.001
AC
Medicaid
M
Black
ED
White
AN US
Race
Controls
CR IP T
All
14
ACCEPTED MANUSCRIPT
Income
Medium High
17.03 (1.76) 38.28 (2.27) 44.69 (3.13)
27.8 (4.92) 40.24 (4.47) 31.95 (5.57)
26.31 (4.43) 40.36 (6.28) 33.32 (6.09)
13.54 (2.61) 40.7 (5.11) 45.76 (4.75)
14.12 (2.08) 36.22 (2.94) 49.66 (3.76)
9.31 (0.99) 36.44 (1.81) 46.21 (1.62)
7.37 (2.76) 37.14 (4.58) 47.76 (6.14)
5.83 (1.92) 32.97 (4.63) 51.51 (5.25)
9.22 (2.36) 42.06 (4.59) 50.07 (3.33)
10.52 (1.68) 34.19 (2.36) 42.98 (2.77)
41.39 (2.26) 40.44 (1.96) 46.21 (1.62)
23.09 (4.85) 45.34 (5.26) 47.76 (6.14)
28.28 (4.74) 39.77 (6.41) 51.51 (5.25)
37.09 (3.65) 48.41 (3.85) 50.07 (3.33)
51.17 (3.39) 35.21 (2.99) 42.98 (2.77)
27.11 (4.61) 55.16 (9.55) 83.37 (4.74) 86.27 (3.33) 77.87 (5.49) 34.93 (5.69) 2.35 (1.06) 19.31 (4.61) 41.82 (6.53)
92.53 (2.31) 62.12 (9.31) 76.11 (3.77) 82.34 (3.47) 58.50 (4.21) 53.57 (4.03) 8.05 (3.39) 13.80 (2.60) 26.56 (6.49)
19.43 (2.43) 17.49 (2.39) 61.99 (3.70) 74.16 (2.48) 11.09 (2.17) 43.22 (3.12) 3.27 (0.79) 11.45 (2.11) 22.75 (4.66)
Current Smoker Former Smoker Non-Smoker
AN US
Smoking Status
Excellent/Very Good Good Fair/Poor Comorbidities
Insulin Resistance
PT
Hypertension
48.84 (1.95) 40.44 (3.14) 70.97 (2.22) 78.03 (1.90) 41.30 (2.03) 47.89 (2.28) 5.94 (1.04) 15.71 (1.44) 28.28 (3.85)
ED
Obesity
High Cholesterol
CE
Metabolic Syndrome Arthritis
AC
Congestive Heart Failure Ischemic Heart Disease
Chronic Kidney Disease
M
Health Condition, self-reported
93.77 (2.8) 79.94 (6.09) 86.47 (4.05) 79.2 (4.68) 96.96 (2.58) 63.38 (4.56) 14.29 (3.11) 32.06 (4.20) 42.78 (8.24)
<.001 0.767
CR IP T
Low
0.007
0.567 0.311 0.391
<.001 0.043 0.391
<.001 <.001 <.001 0.056 <.001 <.001 <.001 <.001 0.025
Values were expressed as weighted mean/percentage (standard error)
15
ACCEPTED MANUSCRIPT
Table 2. Cognitive Performance by T2DM and NAFLD Status, NHANES (2011-2014)
Immediate verbal memory (CERAD-WL) Delayed verbal memory (CERAD-WL) Verbal fluency (AFT) Processing speed, sustained attention, working memory (DSST)
All
NAFLD+ T2DM
T2DM only
NAFLD only
Controls
19.92 (0.27) 6.34 (0.13) 17.96 (0.26) 52.50 (0.83)
19.19 (0.53) 6.11 (0.26) 16.84 (0.45) 47.12 (1.65)
18.88 (0.51) 5.67 (0.29) 15.91 (0.54) 45.08 (1.98)
20.45 (0.42) 6.63 (0.20) 18.57 (0.31) 55.99 (1.05)
20.05 (0.33) 6.38 (0.14) 18.33 (0.42) 53.60 (1.23)
CR IP T
Process Measure
AC
CE
PT
ED
M
AN US
Values were expressed as weighted mean (standard error) CERAD-WL: Consortium to Establish a Registry for Alzheimer’s Disease-Word List; AFT: Animal Fluency Test; DSST: Digit Symbol Substitution Test
16
ACCEPTED MANUSCRIPT
Table 3. Association between T2DM and NAFLD status on AFT and DSST among U.S. adults aged 60 and over, NHANES (2011-2014)
Beta (95% CI)
Weighted mean ± SE
Model 1 (n=931)
Model 2 (n=918)
Reference
Reference
AFT None
18.33 ± 0.42
NAFLD only
18.57 ± 0.31
T2DM only
15.91 ± 0.54**
-1.40 (-2.68 - -0.12)*
-1.47 (-2.70 - -0.23)*
T2DM and NAFLD
16.84 ± 0.45**
-1.14 (-2.17 - -0.10)*
-0.69 (-2.08 - 0.70)
DSST None
53.60 ± 1.23
Reference
Reference
NAFLD only
55.99 ± 1.05
1.96 (-0.60 - 4.52)
T2DM only
45.08 ± 1.98**
T2DM and NAFLD
47.12 ± 1.65**
0.53 (-0.66 - 1.73)
CR IP T
0.03 (-0.87 - 0.93)
-3.81 (-8.07 - 0.45)
AN US
-3.69 (-7.00 - -0.39)*
1.82 (-1.81 - 5.46)
-3.66 (-8.06 - 0.74)
-3.44 (-6.75 - -0.12)*
AC
CE
PT
ED
M
* p <0.05 and **p <0.01 significantly different from None. SE, standard error Model 1: adjusted for Age, male, race, and income Model 2: adjusted for Model 1 + smoking status, presence of depression, asthma, arthritis, congestive heart failure, cancer, ischemic heart disease, chronic obstructive pulmonary disease, thyroid disease, psoriasis, obesity, hypertension, hypercholesterolemia. AFT= Animal Fluency Test DSST= Digit Symbol Substitution Test
17