Investigation into the Relationship between Autoimmune Disease and Non-Alcoholic Fatty Liver Disease

Author(s): Timnit Sisay

Mentor(s): Lynn Gerber, Health Administration and Policy

Abstract

Introduction

Nonalcoholic fatty liver disease(NAFLD) is a common form of chronic liver disease. NAFLD is associated with obesity, type 2 diabetes, hyperlipidemia, hypertension, hyperuricemia, and metabolic syndrome(MetS). Autoimmune disorders (AI) are characterized by dysregulation of the immune response. NAFLD and MetS are often associated with an increase in inflammation. We asked if there is a significant difference between the prevalence of NAFLD (by MetS diagnosis), in people with a diagnosis of AI compared with those without AI.

Methods

A retrospective, chart review study was conducted using the electronic health record program, EPIC, for patients admitted to five Inova hospitals from 2017 – 2018. There were 1,547 patients with an AI diagnosis and 306,380 patients without. The exclusion criteria for all were: age< 18, liver disease, alcohol use disorder, pregnancy, cancer. 

Results:

Data generated for the non-AI cohort: [mean age= (51.51 years ± 19.43), 61.2% female, 52.7% white, 45.1% hypertension , 21.0% obese, 8.17% diabetes (type 1 or type 2) ]. Data generated for the AI cohort: [mean age= (62.48 years ± 18.09), 73.6% female, 69.9% white, 51.6% metabolic syndrome, 62.0% hypertension , 35.1% obese, 41.2% dyslipidemia, 22.5% diabetes (type 1 or type 2), 22.6% type 2 diabetes ]. Furthermore, in a binomial, one sample nonparametric test, patients in the AI groups had a significantly higher prevalence of hypertension (p<0.001), obesity (p<0.001), and diabetes (type 1 or type 2) (p<0.001).

Conclusion:

People with AI population are older, more likely to be female, significantly higher prevalence of obesity, hypertension, and diabetes. This association may be related to other factors and additional analyses are needed to identify which factors are most highly correlated. Since AI is associated with decreased life expectancy, the additional finding of a high prevalence of MetS, a treatable condition, should be an important treatment target for this group.

Video Transcript

Hello everyone, my name is Timnit Sisay, and my mentor for this project is Dr. Lynn Gerber. And for this project I investigated the relationship between autoimmune disease and non-alcoholic fatty liver disease. So for some background on this topic. Non-alcoholic fatty liver disease is also known as NAFLD and it’s very common and affects about 25% of people globally. NAFLD is also associated with metabolic syndrome and the risk factors for metabolic syndrome are obesity, abdominal obesity, type 2 diabetes, hyperlipidemia or dyslipidemia, and hypertension. Also, NAFLD can be diagnosed with the presence of metabolic syndrome, in other words, if someone has metabolic syndrome, then they are also considered to have NAFLD. So the reason why I am interested in looking at the relationship between autoimmune disease and NAFLD is because autoimmune disease patients might have a higher risk for NAFLD because of the inflammation caused by autoimmune disease or the medications they are taking to treat autoimmune disease. Also, autoimmune disease is inflammatory, and this suggests that there may be immune system abnormalities which may increase the likelihood of metabolic syndrome or the risk factors of metabolic syndrome. So, I hypothesized that diagnosis with an inflammatory disorder would increase the risk of NAFLD by increasing the risk of components of metabolic syndrome. 

For my research question I asked: “Is there a significant difference between the prevalence of risk factors for metabolic syndrome (which are obesity, hypertension, and diabetes) in the hospitalized Inova patient population with a diagnosis of autoimmune disease compared to those without autoimmune disease? 

For the methodology of this project, I conducted a retrospective chart review with the use of patient data from 2017 to 2018. 1,548 patients with an autoimmune disease diagnosis were included and 306,380 patients with no autoimmune disease were included. Also, the exclusion criteria for all patients were children, liver disease, alcohol use disorder, pregnancy, cancer. And the statistical test I used was the binomial test. 

For the results. When comparing the autoimmune disease group to the non-autoimmune disease group we found that the autoimmune group was older and more likely to be female and white. And for more information on the autoimmune group. For this study, the criteria for metabolic syndrome diagnosis was having 2 or more of the risk factors for metabolic syndrome, and those who have metabolic syndrome are considered to have NAFLD. By this definition, about 51.6% of autoimmune patients had NAFLD which is very high. These patients also had a high prevalence of obesity, hypertension, type 2 diabetes, and dyslipidemia which are also risk factors of metabolic syndrome. When comparing the autoimmune disease group to the non-autoimmune disease group. We found that the autoimmune group had a significantly higher prevalence of obesity, hypertension, and diabetes than the non-autoimmune group. And this also had a p-value of less than 0.001. 

To conclude, we found that the autoimmune group had a higher prevalence of the risk factors of metabolic syndrome than the non-autoimmune group. This suggests that these patients may be more likely to have NAFLD than the non-autoimmune group. And so, additional analysis of overall NAFLD prevalence in the non-autoimmune group would be beneficial. Also I found that autoimmune disease patients tend to be older, more likely to be female, and have a higher prevalence of obesity, hypertension and diabetes. There are also other factors which may contribute to the association between autoimmune disease and NAFLD. Some of these factors are inflammation, medication, and differences in activity level. So additional analyses are needed to identify which factors are most highly correlated. And this concludes my presentation, thank you so much for listening!

4 replies on “Investigation into the Relationship between Autoimmune Disease and Non-Alcoholic Fatty Liver Disease”

Nice presentation, Timnit. I am really interested in auto-immunity. The correlation between autoimmune disorders and MetS is interesting. Is it possible that there is an autoimmune component to MetS or do you think it’s more likely that generalized inflammation is the culprit? Is there any way to figure that out?

Hello Dr. Lee,

This is an interesting question. It is possible that generalized inflammation (non autoimmune disease related) could be the culprit of MetS. Also, autoimmune disease is inflammatory which shows that generalized inflammation could be the culprit of MetS. Overall, many factors could play a role such as inflammation, corticosteroid use, limited mobility due to autoimmune disease, older age, etc… More analyses which examine all the factors which may play a role in MetS’ association with autoimmunity would help us determine which factor is most correlated.

Also, I am not quite sure of how we would be able to study if MetS is caused by generalized inflammation or autoimmune disease (an inflammatory disease) because both groups experience inflammation. Possibly, this would require studying the relationship between MetS and inflammation while excluding individuals with autoimmune disease.

Thank you for watching the presentation and commenting!

Great visuals to represent your data! I think it’s really interesting you were able to identify differences across such a wide umbrella of autoimmune diseases. Do you think you will focus on any certain diseases in future analysis?

Hello Natalie,

In future analysis, if I were to focus on a specific disease, I would be interested in focusing on rheumatoid arthritis because it was the most prevalent form of autoimmune disease in the population. Also, it would be interesting to see the prevalence of MetS and prevalence of the risk factors for MetS across all forms of autoimmune disease in this population.

Thank you for watching the presentation and commenting!

Leave a Reply