6.3. Association between insulin resistance and GTP

In the present study, BMI, WH ratio and alcohol consumption correlated with GTP, supporting the observation of Yamada et al. (Yamada et al. 1989, Yamada et al. 1990). However, elevated GTP levels have been shown to associate even more strongly with BMI, total cholesterol and systolic blood pressure than alcohol consumption (Nilssen et al. 1990). The association between GTP levels and blood pressure has been shown to be independent of the alcohol dose (Yamada et al. 1989) and also documented in non-drinkers (Yamada et al. 1990). However, the significant relationship between GTP and blood pressure was reduced to a nonsignificant level after adjustment for plasma insulin, whereas plasma insulin remained a significant association with blood pressure (Ikai 1995). In our study, blood pressure correlated significantly with GTP in control men. The adjustment of the GTP levels for age, body mass index and alcohol consumption, but not the adjustments only for age and BMI, almost abolished the correlations. This suggests that alcohol consumption could explain a considerable part of the correlations. The same trend was also found in control females. The lack of association between GTP and blood pressure in subjects with drug-treated hypertension may be due to the effect of the antihypertensive drug therapy.

In the present study, GTP levels were associated with the components of the MS. The relationships between GTP and fasting glucose and insulin and 2-hour glucose and insulin after glucose loading are in good accordance with the previous findings (Koutis et al. 1992, Wannamethee et al. 1995). In a Cretan population with low cardiovascular mortality, hypertensives had higher GTP values even though they consumed less alcohol than normotensives. This was suggested to be related to the higher body mass index and the higher basal insulin production (estimated from C-peptide) of hypertensives than normotensive subjects (Koutis et al. 1992). High serum triglyceride level is one of the most common components of the metabolic syndrome. In our study, the GTP adjusted for age, body mass index and alcohol usage correlated with serum triglyceride values in control subjects and drug-treated women, and there was also a trend towards higher triglyceride values in the highest GTP tertile in hypertensive men. However, drug treatment may be a confounding factor. Beta blockers and diuretics are known to increase serum triglyceride values (Lithell 1993), and beta blockers were the most commonly used drugs in our male hypertensive patient cohort. Low HDL cholesterol is one component of the metabolic syndrome. In our study, however, HDL did not correlate with GTP adjusted for age, body mass index and alcohol consumption. Indeed, the relationship between GTP and HDL cholesterol was possibly confounded by the effect of alcohol consumption. This is supported by a previous study that showed an association of light-to-moderate alcohol intake with enhanced insulin-mediated glucose uptake, lower plasma glucose and insulin concentrations in response to oral glucose and also with higher HDL cholesterol (Facchini et al. 1994). The strong increasing effect of alcohol on GTP levels and the simultaneously increasing effect on HDL levels may obscure the relationship in relation to MS.

In the British Regional Heart Study (Wannamethee et al. 1995), the increased risk of ischemic heart disease mortality was more marked in those with evidence of ischemic disease at screening. The mechanism by which elevated GTP levels contribute to ischemic heart disease has not been solved. However, GTP has been previously associated with hypertension (Yamada et al. 1990), but not with insulin resistance per se. A significant relationship between non-fasting serum insulin and GTP has been observed after adjustment for age and BMI in middle-aged men (Perry et al. 1998). In the present study, hyperinsulinemia was associated with gamma-GT tertiles independently of age, BMI and alcohol usage in all study groups in both genders. This observation implies that insulin resistance might be in a central mechanism by which increased GTP levels are associated with ischemic heart disease. This is supported by the finding that high serum GTP levels have been shown to predict the development of NIDDM in men independently of serum glucose, body mass index and other predictors of NIDDM (Perry et al. 1998). Also, initial GTP levels were significantly higher in stroke patients with a history of diabetes mellitus than in non-diabetics (Peck et al. 1990).

Several mechanisms have been proposed to explain the elevation of GTP levels. One mechanism for elevated GTP could simply be the fatty infiltration of the liver. In fact, in obese non-drinking subjects, the increase in GTP could be caused by the fatty change of liver cell (Yamada et al. 1990). The elevated GTP in alcohol consumers is also related more closely to the end-organ damage of alcohol consumption than to the amount of alcohol consumed (Henningsen et al. 1983, Cusman et al. 1984, Poikolainen et al. 1985), and elevated GTP in this condition may reflect more directly the alterations in hepatic cell membrane function.

The correlation of GTP with fasting glucose and insulin and 2-hour glucose and insulin after glucose loading in the present study suggests that GTP has a close link to carbohydrate metabolism and insulin resistance. One pathophysiological mechanism resulting in the elevation of liver enzymes is an increased insulin effect on the liver causing hepatic steatosis (Nomura et al. 1986, Wanless et al. 1989, Fagerberg et al. 1993). Hyperinsulinemia has been associated with elevated liver enzymes, such as alanine and aspartate aminotransferases, independently of obesity (Fagerberg et al. 1993). Additionally, hyperinsulinemia seems to be related to increased GTP and hepatic steatosis (Ikai et al. 1995). Steatosis appears when insulin secretion is sufficient to block free fatty acid (FFA) oxidation but not sufficient to block FFA mobilization from adipose tissue (Wanless et al. 1989). FFAs are the major substrate for hepatic triglyceride synthesis, explaining the strong association between triglycerides and fasting insulin, and they are markedly cytotoxic (Wenzel & Hale 1978) which may lead to alterations in hepatic cell membrane function and to increased release of GTP.

However, insulin is degraded in the liver by the enzyme glutathione insulin transhydrogenase. GTP is involved in glutathione metabolism by mediating subtrates for the synthesis and cleavage of glutathione. Renal GTP prevents the excretion of glutathione from the body by initiating cleavage of this tripeptide (Hanigan et al. 1996). The presence of GTP in extrarenal tissues may be a means of preventing loss of glutathione and its constituent amino acids from the body. It can be hypothesized that increased concentrations of insulin in hepatic cells possibly induce GTP activity and subsequently increase GTP release from hepatic cells. This is supported by the finding that induction of GTP activity and increased intracellular glutathione concentrations correlate with increased resistance of tumours to chemotherapeutic drugs (Hanigan et al. 1996).

The results of this population-based cross-sectional study suggest that GTP may be a marker for the metabolic syndrome in the same way as for hypertriglyceridemia and high fasting insulin (Reaven 1988) or hyperuricemia (Vuorinen-Markkula & Yki-Järvinen 1994). GTP was related to many components of the metabolic syndrome even after adjustment for age, BMI and alcohol consumption. However, as the regression analyses show, BMI and triglycerides are significantly more powerful predictors of the metabolic syndrome than GTP, which explained only two percent of the variation of fasting insulin. Thus, in the absence of other evidence of disease and increased alcohol consumption, GTP could be used as a marker of the metabolic syndrome, and elevated levels of GTP should lead to a more detailed evaluation of the possibility of the metabolic syndrome.