Non-alcoholic fatty liver disease and risk of type 2 diabetes

Non-alcoholic fatty liver disease (NAFLD) covers a spectrum of liver disease from simple steatosis to non-alcoholic steatohepatitis (NASH) and cirrhosis. NAFLD is commonly associated with features of the metabolic/insulin resistance syndrome (‘Metabolic/Obese NAFLD’) and may therefore predict type 2 diabetes (T2DM). For this review, we searched for prospective studies examining whether NAFLD predicts T2DM, and if so, whether this occurs independently of factors such as age and obesity. These studies included NAFLD diagnosed by ultrasonography (n = 6) or liver enzymes (n = 14). All ultrasonography studies found NAFLD to predict the risk of T2DM independently of age, and in 4 out of 6 studies NAFLD was also a predictor independently of BMI. NAFLD was a predictor of T2DM in all 14 studies where NAFLD was diagnosed by liver enzymes. In 12 of these studies, ALT or AST or GGT were significant predictors of T2DM risk, independently of age and BMI. NAFLD, however, is heterogeneous and may also be caused by common genetic variants. The I148M variant in PNPLA3 and the E167K variant in TM6SF2 are both associated with increased liver fat content, but not features of the metabolic/insulin resistance syndrome. These genetic forms of NAFLD predict NASH and cirrhosis but not T2DM. Taken together these data imply that ‘Metabolic/Obese NAFLD’ predicts T2DM independently of age and obesity and support the role of hepatic insulin resistance in the pathogenesis of this disease.

Introduction

Non-alcoholic fatty liver disease (NAFLD) is defined as hepatic steatosis not caused by excess use of alcohol (>20 g/day in women, >30 g/day in men), viruses such as hepatitis B or C, autoimmune hepatitis, use of hepatotoxic drugs or other compounds, or rare genetic forms [1]. It covers a range of conditions from simple steatosis to non-alcoholic steatohepatitis (NASH) and cirrhosis. NAFLD is currently the most common liver disorder with an estimated worldwide prevalence of 25% [2]. Depending on the method of diagnosis, 65–87% of patients with type 2 diabetes (T2DM) have NAFLD ∗[3], [4]. NAFLD is the second most common cause of being on a waiting list for a liver transplant in the US [5] and the most common cause of hepatocellular carcinoma (HCC) in both US [6] and UK [7].

The metabolic/insulin resistance syndrome is a well-established predictor of T2DM, although overt hyperglycemia only develops in those whose beta-cells fail to sustain hyperinsulinemia in the face of insulin resistance [8]. The liver is the site of production of glucose and very low-density lipoprotein (VLDL) -triglycerides. In subjects with ‘metabolic/obese NAFLD’, the liver is insulin resistant leading to overproduction of both glucose and VLDL [8]. Glucose in turn stimulates insulin secretion thereby inducing hyperinsulinemia. The increase in VLDL leads to lowering of the concentration of high-density lipoprotein (HDL) cholesterol. These changes are often observed in obese subjects, but are also observed independently of obesity [9]. NAFLD is thus closely linked to the pathogenesis of the metabolic syndrome raising the possibility that NAFLD predicts T2DM, even independently of obesity.

In addition to the association of NAFLD with the metabolic/insulin resistance syndrome, two common genetic variants increase the risk of NAFLD. A variant in the patatin-like phospholipase domain-containing 3 (PNPLA3) (rs738409[G], encoding I148M) confers to NAFLD susceptibility by increasing liver fat content, risk of inflammation, and fibrosis (‘PNPLA3 NAFLD’) ∗[10], ∗[11]. Genetic variation in the transmembrane 6 superfamily member 2 (TM6SF2) (rs58542926[T], encoding E167K) is also associated with liver fat accumulation and increased risk of NASH (‘TM6SF2 NAFLD’) ∗[12], [13]. Insulin resistance is not a characteristic of these two conditions [14], although genetic and metabolic causes of NAFLD may both exist in the same person [15].

The ensuing discussion is focused on reviewing studies which have examined whether NAFLD, diagnosed either by liver enzymes, ultrasonography, other imaging techniques, or by liver biopsy, predicts T2DM, and if so, whether this is observed independently of obesity and other established predictors of T2DM. We will also briefly comment on whether and why NAFLD should be screened for in the diabetes clinic.