Nonalcoholic Fatty-Liver Disease Doubles Death Rates in Diabetes

Liam Davenport

March 13, 2017

MANCHESTER, UK — Type 2 diabetes patients who have been admitted to the hospital with nonalcoholic fatty-liver disease (NAFLD) face a substantially increased risk of death from a range of causes, as well cardiovascular-disease events, indicating that NAFLD is an independent risk factor for morbidity and mortality in this patient population, a new large-scale UK study indicates.

The trial of more than 130,000 type 2 diabetes patients, presented last week at the Diabetes UK Professional Conference 2017, showed that hospital admission for NAFLD was associated with a doubling of mortality from all causes, including raised death risk from cardiovascular disease, hepatocellular cancer, and other noncancer causes, particularly for those with more severe NAFLD.

Moreover, NAFLD admission was also linked to an increased risk of cardiovascular disease, with all associations independent of major risk factors such as age, sex, deprivation, cholesterol, smoking, and blood pressure.

"What does this mean?" asked study presenter Sarah Wild, MD, PhD, Usher Institute of Population Health Sciences and Informatics, University of Edinburgh, Scotland.

"One of the key things is that the early stages of NAFLD, particularly the steatosis, respond to the sort of lifestyle changes that are recommended for people with type 2 diabetes in terms of weight loss and increased physical activity, with, of course, the additional suggestion that it's beneficial to restrict alcohol intake as part of a healthy lifestyle."

"Although…prevention is preferable — particularly because there are no licensed drugs available for nonalcoholic fatty-liver disease — we all know that these lifestyle changes are very difficult to achieve, which is why we need more research to find effective treatments," for NAFLD, she added.

"Remarkable" Increase in Risk of Death from Liver Cancer, Cardiovascular Disease

During the question-and-answer session, Dr Wild was asked whether — given that standard cardiovascular risk factors were fairly well controlled in the current study — healthcare professionals should be even "more aggressive in treating those risk factors, or should we be looking for new ones?"

She said the specific population studied here was a little younger (mean age 60 to 63 years) than a standard type 2 diabetes cohort, which may have contributed to their risk factors "looking a bit better." It would therefore be useful to "age-standardize the comparison of risk factors," she explained.

"We'd expect the risk factors to be slightly worse if we looked at it properly, so I think there is potential for better control of risk factors."

Another audience member described the findings as extraordinary, noting that, "Despite all the competing causes of increased mortality in people with diabetes, we could still see an independent effect of NAFLD coming through to increase a whole range of mortality outcomes, from cardiovascular disease through to a remarkable increase in the hazard ratio for hepatocellular carcinoma."

He continued: "I think what this is telling us, given, as you say, that a lot of the traditional risk factors are very similar in both of the groups, is that there are liver-specific molecules, cytokines for example, that are released…in this liver condition, and are potentially having an influence outside the liver."

Another possibility is that the findings are due to "a change in gut microbiota that's contributing to the NAFLD," he added, noting, "It's a case of 'watch this space' for the future."

First Large-scale Look at how NAFLD Affects Outcomes in Diabetes

Opening her presentation, Dr Wild said that NALFD has "only recently been recognized to be strongly associated with both type 2 diabetes and obesity." NAFLD represents a spectrum of disease from simple steatosis to fibrosis and is thought to be present in up to 70% of people with type 2 diabetes.

She highlighted two previous small cohort studies, with one indicating that in people with diabetes, NAFLD is associated with an increase in all-cause mortality (Am J Gastroenterol. 2015;105:1567-1573), and the other revealing an increase in incident cardiovascular disease among NALFD patients (Diabetes Care. 2007;30:1212-1218 ).

However, there have been no reports from large populations, and "one of the reasons for this is that it's really quite difficult to get a measure of NAFLD in population-based studies."

She and her colleagues therefore retrospectively analyzed data extracted in 2014 from Scottish Care Information–Diabetes, a national, population-based register linked to national mortality and hospital-discharge records, on individuals aged 40 to 89 years diagnosed with type 2 diabetes between 2004 and 2013 who had at least one hospital admission.

This yielded 173,716 people, of whom 18,421 were not included due to having no record of a hospital admission, 3175 of whom were excluded due to having one or more of the following: alcoholic liver disease, viral hepatitis, autoimmune hepatitis, or hemochromatosis, and 18,808 of whom were excluded due to missing data.

The final cohort therefore consisted of 133,312 individuals, of whom 1998 (1.5%) had a history of hospital admission with NAFLD. These included 1283 (64.0%) with less severe fatty liver or nonalcoholic steatohepatitis (NASH) and 715 (36.0%) with one or more of fibrosis, sclerosis, cirrhosis, or portal hypertension (PH).

Due to the large sample size, the majority of baseline characteristics were significantly different between type 2 diabetes patients with and without NAFLD, although the absolute differences were typically small.

The More Severe the NAFLD, the Higher the Mortality Risk

Analysis revealed that, over a median follow-up of 4.7 years, type 2 diabetes patients with NAFLD had overall a significantly increased risk of an incident or recurrent cardiovascular-disease event compared with those without NAFLD, at a hazard ratio (HR) of 1.62.

Hazard Ratios for Outcomes Among Various Categories of NAFLD in Type 2 Diabetes Patientsa

Outcome Whole NAFLD group (n=1998) Fatty-liver/NASHb subgroupc (n=1283) Cirrhosis/ fibrosis/sclerosis/PH subgroupd (n=715)
Incident or recurrent CVD event 1.62 1.66 1.57
All-cause mortality 2.11 1.29 3.20
CVD mortality 1.39 1.10 1.78
Hepatocellular cancer (HCC) 41.9 2.42 90.8
Cancer mortality (excluding HCC) 1.15 0.81 1.60
Other causes of death 3.16 1.89 4.82
a. Compared with those with type 2 diabetes, history of one or more hospital admissions, complete data, and no history of chronic liver disease
b. Nonalcoholic steatohepatitis
c. Subgroup based on people with specific mention of fatty liver or NASH but no mention of cirrhosis, fibrosis, sclerosis, or PH
d. Subgroup with mention of cirrhosis, fibrosis, sclerosis or portal hypertension (PH) in hospital records

Furthermore, those with NALFD also had more than double the risk of all-cause mortality compared with those with type 2 diabetes but no record of NAFLD, including significantly increased death risk from cardiovascular disease, hepatocellular carcinoma and other cancers, and increased risk of other cases of death.

And death risk was increased markedly in more severe NAFLD patients compared with those with less severe forms of the disease, across all causes examined.

Referring to the previously mentioned studies, which were "smaller but with better characterization of the NAFLD," Dr Wild said, "Interestingly, despite these different approaches, our results were consistent with the findings of these two smaller studies."

"What we've been able to do is extend this work because of our larger sample size and cause-specific mortality."

She also pointed out that, in this study, she and her colleagues "used the only feasible measure of NAFLD in population-based studies at the moment: diagnosis from a record of a hospital admission."

Consequently, "the event rates that we recorded are likely to reflect people with more severe liver disease," she observed, because those with undiagnosed NAFLD or NAFLD diagnosed in primary care or outpatients would have been included in the comparison group, thereby meaning the risks reported here are likely an "an underestimate" of the true effects of NAFLD, she concluded.

This work is presented on behalf of Scottish and Southampton Diabetes and Liver Disease Group and the Scottish Diabetes Research Network Epidemiology Group. It was supported by funding from the Scottish government through the Scottish Diabetes Group. The authors report no relevant financial relationships.

For more diabetes and endocrinology news, follow us on Twitter and on Facebook.

Diabetes UK Professional Conference 2017. March 9, 2017; Manchester, UK. Abstract A54


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.