Bonnet F, Pierre-Henri D, Gastaldelli A, Laville M, Anderwald C, Konrad T, et al.; for the RISC Study Group. Liver enzymes are associated with hepatic insulin resistance, insulin secretion, and glucagon concentration in healthy men and women. Diabetes 2011;60:1660-1667. (Reprinted with permission.)
The pathophysiological mechanisms to explain the association between risk of type 2 diabetes and elevated concentrations of gamma-glutamyltransferase (GGT) and alanineaminotransferase (ALT) remain poorly characterized. We explored the association of liver enzymes with peripheral and hepatic insulin resistance, insulin secretion, insulin clearance, and glucagon concentration. We studied 1,309 nondiabetic individuals from the Relationship between Insulin Sensitivity and Cardiovascular disease (RISC) study; all had a euglycemic-hyperinsulinemic clamp and an oral glucose tolerance test (OGTT) with assessment of insulin secretion and hepatic insulin extraction. The hepatic insulin resistance index was calculated in 393 individuals. In both men and women, plasma concentrations of GGT and ALT were inversely related with insulin sensitivity (M/I) (all P < 0.01). Likewise, the hepatic insulin resistance index was positively correlated with both GGT (r = 0.37, P < 0.0001, men; r = 0.36, P < 0.0001, women) and ALT (r = 0.25, P = 0.0005, men; r – 0.18, P = 0.01, women). These associations persisted in multivariable models. Increased GGT and ALT were significantly associated with higher insulin secretion rates and with both reduced endogenous clearance of insulin and hepatic insulin extraction during the OGTT (P = 0.0005 in men; P = 0.003 in women). Plasma fasting glucagon levels increased over ALT quartiles (men, quartile 4 vs. 1 11.2 +/− 5.1 vs. 9.3 +/− 3.8 pmol/L, respectively, P = 0.0002; women, 9.0 +/− 4.3 vs. 7.6 +/− 3.1, P = 0.001). In healthy individuals, increased GGT and ALT were biomarkers of both systemic and hepatic insulin resistance with concomitant increased insulin secretion and decreased hepatic insulin clearance. The novel finding of a positive correlation between ALT and fasting glucagon level concentrations warrants confirmation in type 2 diabetes.
Obesity, type 2 diabetes, and metabolic syndrome are interrelated conditions, with increasing epidemiologic impact in the United States and the Western world. These conditions are typically associated with several complications, including coronary artery disease (CAD), cerebrovascular accidents (CVA), and nonalcoholic fatty liver disease (NAFLD). 1-3 Central to the pathophysiology of metabolic syndrome, insulin resistance is a condition in which insulin sensitivity at sites of glucose metabolism is diminished. As a result, insulin levels rise. No single test is routinely and consistently used in clinical practice to diagnose insulin resistance. Furthermore, there is no approved treatment for isolated insulin resistance in the absence of clinically overt diabetes.
Nevertheless, insulin resistance is associated with several important clinical outcomes, including cardiovascular events in patients with and without overt diabetes. 4-8 In fact, insulin resistance is an independent risk factor for cardiovascular mortality in nondiabetic patients with end-stage renal disease. 9 Furthermore, population-based data show that insulin resistance is associated with all-cause mortality in patients with a normal body mass index. 10
In addition to cardiovascular outcomes, the epidemiologic association between liver enzymes and the development of type 2 diabetes is well established. 11-15 Furthermore, both insulin resistance and type 2 diabetes are associated with an increasing prevalence of NAFLD in the United States. 2 Additionally, insulin resistance and type 2 diabetes have both been associated with adverse outcomes in patients with chronic liver disease. 1
Because of these important associations between insulin resistance and liver-related outcomes, we read with great interest the article by Bonnet et al. investigating the pathophysiologic relationship between liver enzymes and incident diabetes in healthy subjects. 16 The investigators reported many important associations between liver enzymes and various steps in glucose metabolism in a study that included patients from the Relationship between Insulin Sensitivity and Cardiovascular disease (RISC) study cohort (Table 1). The investigators concluded that liver enzymes are associated with many components of glucose metabolism in patients with diabetes, even among “healthy” individuals whose liver enzymes fell within the “normal” range.
|Measure of Glucose Metabolism||ALT Only/GGT Only/ALT and GGT||Assay Used||Multivariate Analysis Factors|
|Peripheral insulin resistance*||ALT and GGT in both men and women||Hyperinsulinemic-euglycemic clamp||Age, center, physical activity, alcohol intake, waist circumference|
|Hepatic insulin resistance*||GGT > ALT in both men and women||Endogenous glucose production|
|Insulin secretion*||GGT in both men and women; ALT in men only||Serum levels during OGTT||Age, center, physical activity, alcohol intake, waist circumference, peripheral insulin sensitivity|
|Hepatic insulin extraction and endogenous insulin clearance||GGT only in men and women||Serum levels during OGTT||Age, center|
|Glucagon concentration||ALT in both men and women; GGT in men only||Fasting serum level||Age, center, physical activity, alcohol intake, waist circumference|
|Fasting glucose||GGT > ALT in men||Fasting serum level|
|ALT > GGT in women|
|Two-hour glycemia||GGT only in both men and women||Two-hour serum level during OGTT||Age, center, physical activity, alcohol intake, waist circumference|
|Fasting glucose at 3 years||GGT in men only||Fasting serum level|
|Two-hour glycemia at 3 years||GGT only in both men and women||Two-hour serum level during OGTT|
Gold-standard techniques were used to measure insulin resistance, including the hyperinsulinemic-euglycemic clamp. The investigators used multivariate analysis to demonstrate independent associations between liver enzymes even after adjusting for potential confounders. The findings were largely applicable to both male and female participants. Finally, the investigators confirmed their findings in individuals whose liver enzymes fell within the “normal” range, emphasizing the arbitrary nature of what is considered the “normal range” for liver enzymes and the need for revision.
One potential source of bias was the inclusion of only those individuals from the RISC cohort for whom liver enzyme testing was available. It is possible that these individuals may have represented a prognostically different cohort compared to those who did not have liver enzymes drawn. This is of particular importance, because the data show that associations between liver enzymes and the pathophysiologic components of glucose metabolism persist even in patients with normal liver enzymes. Another issue is the absence of hepatic imaging to assess the presence of NAFLD. In fact, several studies demonstrate that NAFLD is an independent risk factor for cardiovascular events. Additionally, NAFLD is strongly and independently associated with insulin resistance, type 2 diabetes, and metabolic syndrome. Whether elevated liver enzymes are primarily a reflection of NAFLD (and associated insulin resistance) or whether liver enzymes elevation is associated with insulin resistance independent of NAFLD remains unclear. We suspect that NAFLD, insulin resistance, and elevated liver enzymes are tightly inter-related and part of the spectrum of the same pathophysiologic processes.
While these findings appear to be valid, it is unclear how to use this information in clinical practice. Specifically, should liver enzymes in asymptomatic “healthy” individuals be obtained to assess the risk of incident diabetes? Analogous to the measurement of C-reactive protein (CRP) levels to determine cardiovascular risk, predicting an increased risk for “incident diabetes” can have important positive and negative implications from a public health perspective. Early diabetes prediction and related preventive measures, such as lifestyle modification, could potentially result in lower rates of clinically overt diabetes as well as its complications, which could lead to a decrease in resource utilization longitudinally. Negative implications include inappropriate resource utilization and the potentially negative impact on the well-being of individuals who are now “labeled” with the possibility of an important chronic disease. Nonetheless, additional prospective data is needed for external validation of these important results. Furthermore, whether it is beneficial to institute lifestyle modification with the possible addition of metformin in a population of otherwise healthy patients with elevated liver enzymes for diabetes prevention is unclear. Multiple prospective trials have demonstrated the efficacy of lifestyle modification and metformin in preventing overt diabetes mellitus in those who are at high risk (defined as impaired glucose tolerance and elevated BMI) (17-19). For now, based on the results of this and other research, we believe that discovering abnormal liver enzymes or presence of NAFLD on an imaging study in asymptomatic individuals may predict insulin resistance and incident diabetes mellitus. This is especially important in individuals with obesity and additional risk factors for adverse cardiovascular outcomes.