γ-Glutamyltranspeptidase and Incident Stroke Among Japanese Men and Women
The Circulatory Risk in Communities Study (CIRCS)
Background and Purpose—Although serum γ-glutamyltranspeptidase (GGT) levels have been associated with cardiovascular disease incidence, few studies have taken into account the effect of alcohol intake on GGT levels. In this study, we examined the relationship between GGT and stroke incidence according to drinking status.
Methods—We conducted a prospective cohort study of Japanese women (N=6281) and men (N=3471) aged 40 to 69 years living in communities under systematic surveillance for stroke incidence.
Results—During the 18-year follow-up, 202 (3.2%) women and 230 (6.6%) men had strokes. Serum GGT levels were positively associated with risk of total stroke for women but not men. The multivariable hazard ratios of total stroke for the highest quartile of GGT compared with the lowest quartile were 1.56 (95% CI, 1.01 to 2.39) for women and 1.37 (95% CI, 0.89 to 2.11) for men. Moreover, GGT was associated with total and ischemic stroke risks for never-drinking women.
Conclusions—Serum GGT is associated with risk of total and ischemic strokes for Japanese women, especially never-drinkers.
Previous studies have indicated that higher levels of serum γ-glutamyl transpeptidase (GGT) are associated with incidence of stroke.1 GGT is a well-known enzyme marker for alcohol consumption, liver disease,2 and a potential marker for oxidative stress,3 but few studies have taken into account the effect of alcohol intake on GGT levels. We tried to determine whether drinking status affects the association of GGT with stroke incidence by using long-term follow-up data for middle-aged Japanese men and women.
Subjects and Methods
The populations surveyed included 7352 women and 4764 men aged 40 to 69 years who participated in cardiovascular risk surveys between 1986 and 1993. The communities were the rural communities of Ikawa, Noichi, and Kyowa and the suburb of Yao. Persons with missing serum data (652 women and 554 men), a history of cardiovascular disease (48 women and 113 men), or liver damage (serum aspartate aminotransferase [AST] >50 IU/L and/or serum alanine aminotransferase [ALT] >50 IU/L; 369 women and 625 men) were excluded. The remaining 9752 participants (6281 women and 3471 men) were followed up until the end of 2007. The median follow-up time was 18.1 years. This study was approved by the Ethics Committee of the Osaka Medical Center for Health Science and Promotion.
Details of the risk factor survey have been described elsewhere.4 GGT was measured with the γ-Glu-PNA substrate method. All measurements were performed at the laboratory of the Osaka Medical Center for Health Science and Promotion.
End Point Determination
Stroke incidence was determined from national health insurance claims, reports by local physicians, ambulance records, death certificates, reports by public health nurses and health volunteers, and cardiovascular risk surveys. The incidence of strokes was determined by a panel of 3 physicians. The determination of stroke subtypes was conducted primarily by using CT/MRI findings, which were available for 93% of the stroke cases. Strokes that were diagnosed clinically but showed no lesion on CT/MRI films were classified according to the clinical criteria.
We calculated differences among quartiles of serum GGT levels in terms of age- and community-adjusted mean values or prevalence of potential cardiovascular risk factors at baseline with the general linear models procedure. Hazard ratios (HRs) and 95% CIs of stroke were calculated by using Cox proportional hazards regression models. Possible confounding variables were smoking status (nonsmoker, current smoker), alcohol consumption (nondrinker, former drinker, and current drinker [<23 g/week, 23 to 46 g/week, >46 g/week]), serum total cholesterol (mmol/L), systolic blood pressure (mm Hg), antihypertensive medication use (no, yes), diabetes (no, yes), body mass index (kg/m2), AST (IU/L), ALT (IU/L), and serum albumin (g/dL).
All statistical analyses were performed with the SAS System (Version 9.1; SAS Inc.). All probability values for statistical tests were 2-tailed, and values of <0.05 were regarded as statistically significant.
Table 1 shows the baseline characteristics according to serum GGT levels for both women and men.
Among the 6281 women and 3471 men, 432 had incident strokes during the 18-year follow-up period: 202 (3.2%) strokes, 106 (1.7%) ischemic strokes, 78 (1.2%) hemorrhagic stroke, and 18 (0.3%) unclassified strokes for women; and the corresponding values for men were 230 (6.6%), 144 (4.1%), 63 (1.8%), and 23 (0.7%). The proportions of current drinkers and never-drinkers were 10.9% and 89.8% for women and 73.2% and 20.9% for men.
Age- and community-adjusted risks of stroke showed GGT-level dependent associations for both genders (Table 2). After adjustment for cardiovascular risk factors and liver function variables, the associations were still significant for women, but not for men, and remained unchanged after leaving out AST and ALT from the model (not shown in the table). When we limited the subjects to never-drinkers, the associations became slightly stronger for women.
The positive associations for never-drinking women were observed primarily for ischemic stroke, but not hemorrhagic stroke (Table 3). No significant associations were observed among never-drinking men.
A major finding of the present study was that serum GGT levels appear to be associated with risk of total and ischemic strokes for Japanese women, especially never-drinkers. A Finnish study found that serum GGT was positively associated with risk of total stroke for both genders1 but did not examine the effect of drinking status. Another Japanese study showed a positive association between GGT and cardiovascular disease mortality among never-drinking women but no association for either current or never-drinkers among men.5 The gender difference in the associations of GGT with stroke was probably due to the much higher prevalence of drinkers. This indicated the need for further study using a sufficient number of never-drinking men.
The mechanisms by which higher serum GGT levels increase the risk of stroke have not been fully elucidated. GGT is known to be elevated in persons with low high-density lipoprotein cholesterol, high total cholesterol, and high levels of triglycerides, glucose, AST, ALT,6 and serum albumin.2 However, we found significant associations between GGT and risk of stroke among women even after adjustment for these covariates. GGT could serve as a marker of inflammation and oxidative stress,3 which may lead to development of athero- and arteriosclerosis. Moreover, catalytic active GGT has been detected in atherosclerotic plaques of coronary and carotid arteries and its hydrolysis glutathione to derive cysteinyl glycine, which may trigger the production of reactive oxygen species and the oxidation of low-density protein.7
Potential limitations of this study warrant mentioning. The statistical power for detecting any significant associations for never-drinking men was limited. Second, we analyzed the associations between GGT and stroke incidence by using a single assessment of GGT at baseline, which may have led to misclassification of GGT categories among individuals. Finally, although the associations of GGT with stroke incidence were found to be independent of the traditional risk factors, other residual confounders such as physical activity and high-density lipoprotein cholesterol may affect the associations.
In conclusion, our findings indicate that serum GGT level is a predictor for incidence of total and ischemic strokes among Japanese never-drinking women. Exploration of the biological mechanisms for that association may thus prove to be useful for stroke prevention.
Sources of Funding
This study was supported by Grants-in-Aid for Science Research from the Ministry of Education, Science, Sports and Culture of Japan (Nos. 14207019 and 19390174).
- Received September 28, 2009.
- Accepted October 6, 2009.
Jousilahti P, Rastenyte D, Tuomilehto J. Serum γ-glutamyltransferase, self-reported alcohol drinking, and the risk of stroke. Stroke. 2000; 31: 1851–1855.
Yamada J, Tomiyama H, Yambe M, Koji Y, Motobe K, Shiina K, Yamamoto Y, Yamashita A. Elevated serum levels of alanine aminotransferase and gamma glutamyltransferase are marker of inflammation and oxidative stress independent of the metabolic syndrome. Atherosclerosis. 2006; 189: 198–205.
Simamoto T, Komachi Y, Inada H, Doi M, Iso H, Sato S, Kitamura A, Iida M, Konishi M, Nakanishi N. Trends for coronary heart disease and stroke and their risk factors in Japan. Circulation. 1989 79: 503–515.
Aldo P, Michele E, Erri G, Eugenio C, Claudio P, George P, Alfonso P. Human atherosclerotic plaques contain gamma-glutamyltranspeptidase enzyme activity. Circulation. 2004; 109: 1440.