Non-linear relationship between gamma-glutamyl transferase and type 2 diabetes mellitus risk: secondary analysis of a prospective cohort study

Abstract

Objective

To investigate the association between gamma-glutamyl transferase (GGT) and type 2 diabetes mellitus (T2DM) risk.

Methods

This was a secondary analysis based on a publicly available DRYAD dataset that included 15 444 study participants that received medical examinations at a single centre in Japan between 2004 and 2015. Crude, minimally-adjusted and fully-adjusted regression models were used to evaluate the relationship between GGT levels and T2DM risk.

Results

The study participants (mean ± SD age of 43.72 ± 8.90 years; 8415 of 15 444 [54.49%] were male) were followed-up for a median of 1968 days (5.39 years). After adjusting for potential covariates, a non-linear relationship between the baseline GGT level and T2DM incidence was observed. The inflection point for T2DM risk was 10 IU/l GGT; below this point, the T2DM incidence increased by 1.18-fold per unit change in GGT. Above this point, the association between GGT and the incidence rate of T2DM became nonsignificant.

Conclusion

Baseline GGT exhibited a non-linear association with T2DM incidence. Elevated GGT levels should be incorporated into routine screening for individuals at high risk of T2DM, allowing for early intervention targeting GGT to potentially reduce T2DM-related morbidity.

Keywords

Gamma-glutamyl transferase type 2 diabetes mellitus non-linearity

Introduction

Type 2 diabetes mellitus (T2DM) is a major public health problem and the incidence has been continually rising in recent years.¹ According to the Centers for Disease Control and Prevention in the US, approximately one out of every 10 adults currently has T2DM.¹ Advances in medical technology have led to the development of new therapeutic agents, but mortality associated with the complications of T2DM remains high.² It is estimated that about five million people died from T2DM in 2015 worldwide, which accounted for 12.8% of all-cause mortality in adults.³ Therefore, identification of factors associated with the pathogenesis of T2DM is of great importance for improving screening of the high-risk population and facilitating early prevention strategies.

Gamma-glutamyl transferase (GGT) is a liver enzyme involved in the gamma glutamyl cycle, which is related to glutathione (GSH) synthesis and degradation.⁴ Previous studies have demonstrated a close relationship between GGT, oxidative stress and insulin resistance, which are two critical pathophysiological processes contributing to T2DM.^5–8 Despite extensive research focused on the potential relationship between GGT and the risk of T2DM, no consensus has been reached to date.^9–11 Therefore, this current study used publicly available data from a longitudinal study conducted in Japan to further evaluate the role of GGT in T2DM.

Patients and methods

Data source and study population

This current study undertook a secondary analysis of open-access data from the DRYAD database (https://datadryad.org/). The raw data can be freely obtained from this website and reanalysed with rationalcitation (Dryad data package: Okamura et al. (2019) Data from: Ectopic fat obesity presents the greatest risk for incident type 2 diabetes: a population-based longitudinal study, Dryad, Dataset, https://doi.org/10.5061/dryad.8q0p192).¹² This secondary analysis was undertaken in the Department of Cardiology, Luoyang Central Hospital Affiliated to Zhengzhou University, Luoyang, Henan Province, China.

The original study involved a population-based prospective cohort recruited from a medical examination programme at Murakami Memorial Hospital, Gifu, Japan.¹² Data for 15 464 participants between 2004 and 2015 were extracted to explore the relationship between various obesity-related phenotypes and the risk of T2DM. Exclusion criteria were under-medication, alcoholic fatty liver disease, viral hepatitis, T2DM and fasting plasma glucose (FPG) ≥6.1 mmol/l. The original study was approved by the ethics committee of Murakami Memorial Hospital and written informed consent was obtained from all participants.

The database file downloaded from DRYAD contained information for the following parameters: age, sex, body mass index (BMI), waist circumference (WC), ethanol consumption, smoking status, exercise habit, fatty liver, and baseline levels of systolic blood pressure (SBP), diastolic blood pressure (DBP), alanine aminotransferase (ALT), aspartate aminotransferase (AST), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), FPG, glycosylated haemoglobin (HbA_1c) and GGT. Regular exercise was defined as playing any type of sport more than once a week. T2DM was defined as HbA_1c ≥6.5%, FPG ≥7 mmol/l or self-reported.¹² In the current study, the baseline GGT level was evaluated as an independent variable and the dependent variable was the risk of T2DM. The median follow-up time was 1968 days (5.39 years).

Statistical analyses

Statistical analyses were performed using EmpowerStats (http://www.empowerstats.com, X&Y Solutions Inc., Boston, MA, USA) and R packages (http://www.R-project.org). Continuous variables are shown as mean ± SD or median (interquartile range) according to data distribution. Categorical variables are presented as frequencies and percentages. Differences among groups (GGT tertiles) were detected using one-way analysis of variance (for normally distributed data), Kruskal–Wallis H tests (for skewed distribution), χ²-test analysis (for categorical variables) or Fisher’s exact test. Univariate and multivariate Cox regression analyses were conducted to evaluate the relationship between GGT levels and T2DM risk. According to the STROBE recommendation, results for all three adjusted models are presented.¹³ In the crude model, there was no adjustment for additional variables; the minimally-adjusted model only adjusted for sex and age; and in the fully-adjusted model, confounding covariates that could alter the original effect size (hazard ratio) by >10% or with a P-value for the regression coefficient <0.1 were added to the model as confounders.¹⁴ For sensitivity analysis, GGT was also converted into a categorical variable according to tertiles (low, median, high). A trend test was conducted by substituting the median GGT values of the three groups into the Cox proportional hazards regression equation as a continuous variable. The P-value for the regression coefficient of this continuous variable is the P-value for the trend. A potential non-linear relationship of GGT with the risk of T2DM was also evaluated using the Cox proportional hazards regression model with cubic spline functions.^15,16 When a remarkable segmented trend was revealed, the inflection point was detected using the recursive algorithm, and a two-piecewise Cox proportional hazard model was applied to both sides of the inflection point. The log likelihood ratio test was employed to determine the most suitable model for describing the association between GGT and T2DM risk.^15,16 A P-value < 0.05 (two-sided) was considered to be statistically significant.

Results

A total of 15 444 participants were included in the final analysis after excluding 20 participants with abnormal liver function, which was defined as ALT or AST > 120 U/l (i.e. three-times higher than the normal range). The mean ± SD age of the participants was 43.72 ± 8.90 years and more than half were male (8415 of 15 444; 54.49%). The baseline demographic and clinical characteristics of the study participants are summarized according to the tertiles of baseline GGT level (Table 1) in order to show the distribution of other confounding factors and, at the same time, to explore which variables might be associated with GGT levels. Compared with the other two tertile groups, participants in the high tertile for GGT were more likely to be male, older, ex/current smokers, alcohol drinkers and to have more risk factors for T2DM at baseline, including a higher rate of fatty liver and higher levels of BMI, WC, BP, TC, HDL-C, TG, HbA_1c, as well as FPG. Accordingly, they were more susceptible to the development of T2DM later in life (250 of 5400; 4.63%); while, the T2DM incidence rates were only 0.82% (33 of 4009) and 1.46% (88 of 6035) in the low GGT and median GGT groups, respectively.

Table 1.

Baseline demographic and clinical characteristics of the study participants (n = 15 444) that were included in the final analysis of the relationship between gamma-glutamyl transferase (GGT) and type 2 diabetes mellitus (T2DM) risk.

	GGT tertile			Statistical analysis^a	Statistical analysis^b
Characteristic	Low3.00–11.00 IU/l	Median12.00–18.00 IU/l	High19.00–399.00 IU/l	Statistical analysis^a	Statistical analysis^b
Number of participants	4009	6035	5400
Age, years	42.29 ± 8.54	43.49 ± 9.07	45.02 ± 8.77	P < 0.001	P < 0.001
Sex				P < 0.001	–
Female	3294 (82.17%)	2906 (48.15%)	829 (15.35%)
Male	715 (17.83%)	3129 (51.85%)	4571 (84.65%)
BMI, kg/m²	20.76 ± 2.47	21.68 ± 2.88	23.60 ± 3.20	P < 0.001	P < 0.001
WC, cm	71.10 ± 7.39	75.40 ± 8.13	81.61 ± 8.53	P < 0.001	P < 0.001
Ethanol consumption, g/week	1.00 (0.00–5.13)	1.00 (0.00–44.00)	36.00 (1.00–132.00)	P < 0.001	P < 0.001
Ex/current smoker				P < 0.001	–
No	3185 (79.45%)	3668 (60.78%)	2167 (40.13%)
Yes	824 (20.55%)	2367 (39.22%)	3233 (59.87%)
Habit of exercise				P = 0.005	–
No	3344 (83.41%)	4902 (81.23%)	4491 (83.17%)
Yes	665 (16.59%)	1133 (18.77%)	909 (16.83%)
Fatty liver				P < 0.001	–
No	3869 (96.51%)	5307 (87.94%)	3542 (65.59%)
Yes	140 (3.49%)	728 (12.06%)	1858 (34.41%)
Baseline SBP, mmHg	107.66 ± 13.09	113.54 ± 14.13	120.62 ± 14.73	P < 0.001	P < 0.001
Baseline DBP, mmHg	66.63 ± 9.04	70.81 ± 9.86	76.10 ± 10.33	P < 0.001	P < 0.001
Baseline ALT, IU/l	12.00 (10.00–15.00)	16.00 (13.00–20.00)	24.00 (18.00–32.00)	P < 0.001	P < 0.001
Baseline AST, IU/l	15.00 (12.00–17.00)	17.00 (14.00–20.00)	20.00 (17.00–25.00)	P < 0.001	P < 0.001
Baseline TC, mmol/l	4.91 ± 0.82	5.07 ± 0.84	5.34 ± 0.87	P < 0.001	P < 0.001
Baseline HDL-C, mmol/l	1.58 ± 0.37	1.49 ± 0.41	1.34 ± 0.38	P < 0.001	P < 0.001
Baseline TG, mmol/l	0.62 ± 0.33	0.80 ± 0.51	1.24 ± 0.82	P < 0.001	P < 0.001
Baseline FPG, mmol/l	4.97 ± 0.39	5.13 ± 0.39	5.34 ± 0.38	P < 0.001	P < 0.001
Baseline HbA_1c, mmol/mol	32.47 ± 3.47	33.06 ± 3.39	33.40 ± 3.64	P < 0.001	P < 0.001
T2DM				P < 0.001	–
No	3976 (99.18%)	5947 (98.54%)	5150 (95.37%)
Yes	33 (0.82%)	88 (1.46%)	250 (4.63%)
Follow-up time, days	2485.00 (1077.00–3809.00)	1837.00 (938.50–3115.50)	1941.50 (901.75–3352.75)	P < 0.001	P < 0.001

Data presented as mean ± SD, median (interquartile range) or n of participants (%).

^aP-value calculated by one-way analysis of variance for normally distributed continuous variables, χ²-test analysis for categorical variables or Fisher’s exact test for categorical variables with expected frequency < 10.

^bP-value calculated by Kruskal–Wallis H Test (skewed distribution).

BMI, body mass index; WC, waist circumference; SBP, systolic blood pressure; DBP, diastolic blood pressure; ALT, alanine aminotransferase; AST, aspartate aminotransferase; TC, total cholesterol; HDL-C, high-density lipoprotein cholesterol; TG, triglycerides; FPG, fasting plasma glucose; HbA_1c, glycosylated haemoglobin.

The baseline demographic and clinical characteristics between participants with and without T2DM are compared in Table 2. It could be seen that participants with male sex, aging, smoking and drinking, insufficient exercise and elevated levels of all metabolic parameters were at high risk of T2DM.

Table 2.

Comparison of baseline demographic and clinical characteristics between of the study participants (n = 15 444) that were included in the final analysis of the relationship between gamma-glutamyl transferase (GGT) and type 2 diabetes mellitus (T2DM) risk stratified according to whether or not they had T2DM.

Characteristic	T2DM group	Non-T2DM group	Statistical analysis^a	Statistical analysis^b
Number of participants	371	15073
Age, years	47.11 ± 8.49	43.63 ± 8.89	P < 0.001	P < 0.001
Sex			P < 0.001	–
Female	87 (23.45%)	6942 (46.06%)
Male	284 (76.55%)	8131 (53.94%)
BMI, kg/m²	25.02 ± 3.82	22.04 ± 3.07	P < 0.001	P < 0.001
WC, cm	85.04 ± 10.19	76.25 ± 8.96	P < 0.001	P < 0.001
Ethanol consumption, g/week	4.20 (0.00–90.00)	1.00 (0.00–60.00)	P < 0.001	P = 0.005
Ex/current smoker			P < 0.001	–
No	145 (39.08%)	8875 (58.88%)
Yes	226 (60.92%)	6198 (41.12%)
Habit of exercise			NS	–
No	320 (86.25%)	12417 (82.38%)
Yes	51 (13.75%)	2656 (17.62%)
Fatty liver			P < 0.001	–
No	150 (40.43%)	12568 (83.38%)
Yes	221 (59.57%)	2505 (16.62%)
Baseline SBP, mmHg	121.99 ± 15.62	114.30 ± 14.90	P < 0.001	P < 0.001
Baseline DBP, mmHg	77.18 ± 10.24	71.44 ± 10.47	P < 0.001	P < 0.001
Baseline ALT, IU/l	24.00 (18.00–38.00)	17.00 (13.00–23.00)	P < 0.001	P < 0.001
Baseline AST, IU/l	20.00 (16.00–25.50)	17.00 (14.00–21.00)	P < 0.001	P < 0.001
Baseline GGT, IU/l	24.00 (16.50–36.00)	15.00 (11.00–22.00)	P < 0.001	P < 0.001
Baseline TC, mmol/l	5.42 ± 0.89	5.12 ± 0.86	P < 0.001	P < 0.001
Baseline HDL-C, mmol/l	1.19 ± 0.33	1.47 ± 0.40	P < 0.001	P < 0.001
Baseline TG, mmol/l	1.49 ± 0.97	0.90 ± 0.64	P < 0.001	P < 0.001
Baseline FPG, mmol/l	5.61 ± 0.36	5.15 ± 0.41	P < 0.001	P < 0.001
Baseline HbA_1c, mmol/mol	36.92 ± 3.99	32.93 ± 3.45	P < 0.001	P < 0.001
Follow-up time, days	2164.00 (1085.00–3256.00)	1960.00 (984.00–3445.00)	NS	NS

Data presented as mean ± SD, median (interquartile range) or n of participants (%).

^aP-value here was calculated by T-test for normally distributed continuous variables, χ²-test analysis for categorical variables or Fisher’s exact test for categorical variables with expected frequency < 10.

^bP-value calculated by Kruskal–Wallis H Test (skewed distribution).

Three multivariate analysis models were constructed to detect the independent association of GGT with T2DM risk. As shown in Table 3, the GGT level, when calculated as a continuous variable, was positively associated with the incidence of T2DM in both the crude unadjusted model and the minimally-adjusted model (P < 0.0001 for both models), of which a one unit (IU/l) change would lead to a 1% increase in T2DM risk. When being converted as a categorical variable, the influence of GGT on T2DM risk became more pronounced. With the low tertile group as the reference, the T2DM risk increased to 2.24-fold and 6.53-fold in median and high tertile groups, respectively, in the crude unadjusted model (P < 0.0001 for trend); and were 1.97-fold and 5.23-fold after minimal adjustment for age and sex (P < 0.0001 for trend). However, this association became insignificant after the addition of other confounding covariates, including sex, age, ethanol consumption, fatty liver, BMI, baseline ALT, baseline AST, baseline TG, baseline HbA_1c, baseline FPG and baseline SBP (median tertile versus low tertile: hazard ratio [HR] 1.08, 95% confidence interval [CI] 0.71, 1.66; high tertile versus low tertile: HR 1.40, 95% CI 0.90, 2.20; P = 0.0563 for trend).

Table 3.

Cox regression analyses of the association between glutamyl transferase (GGT) level and the risk of type 2 diabetes mellitus (T2DM) using three different models.

Exposure	Crude unadjusted modelHR (95% CI) P-value	Minimally-adjusted modelHR (95% CI) P-value	Fully-adjusted modelHR (95% CI) P-value
GGT	1.01 (1.01, 1.01) P < 0.0001	1.01 (1.01, 1.01) P < 0.0001	1.00 (1.00, 1.01) P = 0.1309
GGT tertile
Low	Ref	Ref	Ref
Median	2.24 (1.50, 3.35) P < 0.0001	1.97 (1.30, 2.98) P = 0.0013	1.08 (0.71, 1.66) P = 0.7142
High	6.53 (4.54, 9.39) P < 0.0001	5.23 (3.49, 7.84) P < 0.0001	1.40 (0.90, 2.20) P = 0.1380
P-value for trend	P < 0.0001	P < 0.0001	P = 0.0563

Multivariate models: crude model unadjusted model did not adjust for other covariates; minimally-adjusted model adjusted for age and sex; fully-adjusted model adjusted for age, sex, ethanol consumption, smoking status, fatty liver, body mass index, baseline alanine aminotransferase, baseline aspartate aminotransferase, baseline total cholesterol, baseline high-density lipoprotein cholesterol, baseline triglycerides, baseline glycosylated haemoglobin, baseline fasting plasma glucose and baseline systolic blood pressure.

HR, hazard ratio; CI, confidence interval; Ref, reference.

As shown in Figure 1, a non-linear relation between GGT and T2DM was identified after adjusting for age, sex, ethanol consumption, smoking status, fatty liver, BMI, ALT, AST, TC, HDL-C, TG, HbA_1c, FPG and SBP (log likelihood ratio test, P < 0.001). The inflection point was determined to be 10 IU/l GGT. As shown in Table 4, every unit increase in GGT level would lead to a 1.18-fold increased risk of T2DM (95% CI 1.21, 3.94, P = 0.0095) when GGT was <10 IU/l. Saturation was reached at 10 IU/l (P = 0.1768).

Figure 1.

Non-linear relationship between gamma-glutamyl transferase (GGT) and type 2 diabetes mellitus risk. A nonlinear relationship was detected after adjusting for age, sex, ethanol consumption, smoking status, fatty liver, body mass index, alanine aminotransferase, aspartate aminotransferase, total cholesterol, high-density lipoprotein cholesterol, triglycerides, glycosylated haemoglobin, fasting plasma glucose and systolic blood pressure.

Table 4.

The results of two-piecewise Cox proportional hazard model of the relationship between gamma-glutamyl transferase (GGT) and type 2 diabetes mellitus (T2DM) risk.

Inflection point of GGT	Effect size (HR)	95% CI	P-value
<10 IU/l	2.18	1.21, 3.94	P = 0.0095
≥10 IU/l	1.00	1.00, 1.01	P = 0.1768

Effect: T2DM risk.

Cause: GGT level (per one IU/l change).

Adjusted for age, sex, ethanol consumption, smoking status, fatty liver, body mass index, baseline alanine aminotransferase, baseline aspartate aminotransferase, baseline total cholesterol, baseline high-density lipoprotein cholesterol, baseline triglycerides, baseline glycosylated haemoglobin, baseline fasting plasma glucose and baseline systolic blood pressure.

HR, hazard ratio; CI, confidence interval.

As the GGT level was affected by various physiological and demographic factors, which could be evidenced by the wide range of GGT levels in Table 1 and other similar studies, many studies have only focused on the association between GGT and T2DM within the normal range of GGT concentrations. Therefore, the current study also evaluated whether there was a non-linear relationship between GGT and T2DM among participants with normal GGT (set to 51 IU/l and 33 IU/l in males and females, respectively).¹⁷ However, the results were consistent with the main analysis including all participants, confirming the non-linear relationship (Figure 2 and Table 5).

Figure 2.

Non-linear relationship between gamma-glutamyl transferase (GGT) and type 2 diabetes mellitus risk among study participants exhibiting GGT levels within the normal range (set to 51 IU/l and 33 IU/l in males and females, respectively).¹⁷ A similar nonlinear relationship was detected as observed for the entire study population after adjusting for age, sex, ethanol consumption, smoking status, fatty liver, body mass index, waist circumference, alanine aminotransferase, aspartate aminotransferase, total cholesterol, high-density lipoprotein cholesterol, triglycerides, glycosylated haemoglobin, fasting plasma glucose and systolic blood pressure.

Table 5.

The results of two-piecewise Cox proportional hazard model of the relationship between gamma-glutamyl transferase (GGT) and type 2 diabetes mellitus (T2DM) risk among study participants exhibiting GGT levels within the normal range (set to 51 IU/l and 33 IU/l in males and females, respectively).¹⁷

Inflection point of GGT	Effect size (HR)	95% CI	P-value
<10 IU/l	2.05	1.14, 3.69	P = 0.0168
≥10 IU/l	1.01	0.99, 1.02	P = 0.3264

Effect: T2DM risk.

Cause: GGT level (per one IU/l change).

Adjusted for age, sex, ethanol consumption, smoking status, fatty liver, body mass index, waist circumference, baseline alanine aminotransferase, baseline aspartate aminotransferase, baseline total cholesterol, baseline high-density lipoprotein cholesterol, baseline triglycerides, baseline glycosylated haemoglobin, baseline fasting plasma glucose and baseline systolic blood pressure.

HR, hazard ratio; CI, confidence interval.

The influence of other factors, including sex, age, smoking and drinking habits, BMI, TC, TG, HDL-C, as well as FPG, on the relationship between GGT and DM risk were also explored, but no significant interactions were observed (supplemental data: Table S1).

Discussion

This current secondary analysis demonstrated a non-linear relationship between GGT level and T2DM risk. The level of GGT was positively associated with T2DM risk when GGT was less than 10 IU/l. The relationship became insignificant when GGT values exceeded this threshold.

Gamma-glutamyl transferase is a highly glycosylated membrane protein, which is located on the surface of the plasma membrane of almost all cells except erythrocytes.¹⁸ It is encoded by a gene located on chromosome 22 and a total of 13 homologues have been identified on the human GGT gene.¹⁹ Transcription of the various mRNA subtypes depends upon the cell type and stimulating agent.²⁰ Little is known about the complex and precise regulation of GGT gene expression. The highest level of activity of GGT is in the kidney, being about 10-times of that of the liver, but serum GGT is mainly derived from hepatocyte microsomes and the epithelial cells of the bile capillaries.²¹ GGT consists of two subunits: a 50–62 kDa large chain and a 22–30 kDa smaller chain.^22,23 The large subunit comprises of an intracellular N-terminal sequence, a transmembrane hydrophobic domain and an extracellular domain.^22,23 It is mainly responsible for the anchorage of GGT in the plasma membrane.^22,23 The active catalytic centre is located on the small subunit.²⁴ The primary physiological function of GGT is the catalysis of GSH hydrolysis, transcellular transport and absorption of amino acids, as well as the degradation of toxicants.²⁵ Elevated serum GGT level have been shown to be associated with hepatobiliary diseases (inflammatory, damage and obstruction), metabolic diseases (acute myocardial infarction, hypertension, diabetes mellitus and stroke) and even tumours (ovarian cancer, gastric cancer and colorectal cancer).^26–34

The current secondary analysis demonstrated a significant association between GGT level and T2DM risk, independent of age, sex, ethanol consumption, smoking status, fatty liver, BMI, ALT, AST, TC, HDL-C, TG, HbA_1c, FPG and SBP, which might be explained by the following mechanisms. First, in our opinion, GGT is not only a sensitive indicator of an enhanced oxidative stress state, but it is also a direct contributor to oxidative stress. The gamma glutamyl cycle, mediated by GGT, is the most important endogenous antioxidant system.³⁵ Oxidative stress caused by alcohol abuse, high glucose exposure or some other stimuli could activate multiple signal cascades, including Ras, P38 mitogen-activated protein kinases, protein kinase C ζ, extracellular signal-regulated protein kinases 1 and 2, phosphatidylinositol 3-kinase, Akt, protein kinase C δ and protein kinase A, which promotes the expression of GGT.¹⁹ In turn, the cysteinyl-glycine dipeptide produced by GGT would reduce iron to ferrous iron; and the latter would induce to the development of hydrogen peroxide and endogenous superoxide anion radicals, further aggravating oxidative stress injury.³⁶ Secondly, in our opinion, increased GGT levels are also a sign of insufficient antioxidant capacity. GGT is essential in the maintenance of GSH homeostasis.³⁷ In contrast, of the six enzymes involved in the GSH cycle, only GGT is located outside the cell.³⁸ Therefore, it is the rate limiting enzyme in the whole process.³⁹ In contrast, the cysteine required in the generation of GSH mainly depends on the extracellular GGT catalytic production.⁴⁰ Glutathione depletion could lead to an upregulation of GGT as a compensatory response.⁴¹ Thirdly, GGT levels also reflect the inflammatory state as they have been reported to be positively correlated with a variety of inflammatory substances, including C-reactive protein, tumour necrosis factor-α and interleukin-6.^42–44 In addition, GGT itself is also proinflammatory by mediating the transformation of leukotriene C4 containing glutathione into leukotriene D4.⁴⁵ Both oxidative stress and inflammation have been shown to be involved in the development of T2DM.⁴⁶ GGT level was also reported to be positively associated with many T2DM-related risk factors, including red meat intake, visceral fat accumulation and hyperlipidaemia.^47,48

The non-linear relationship between GGT and T2DM risk found in this current study was also previously reported by a meta-analysis,¹⁰ which evaluated 24 eligible studies and found that the GGT level contributes to an increased risk of T2DM in a non-linear dose-dependent manner. The previous meta-analysis identified an inflection point of 35 IU/l in which one unit change in GGT led to a 34% increase in the risk of T2DM.¹⁰ Although this inflection point was higher than that identified in this current study, both analyses indicated that a saturation effect was reached.¹⁰ The non-linear relationship identified between GGT and T2DM not only has a very important clinical value, but also improves the understanding of the role of GGT in T2DM. The findings of this current study suggest that GGT level was closely associated with T2DM risk even when the level was under 10 IU/l. This effect might be reversible during this period. The timely intervention in terms of GGT level and its related T2DM risk factors may provide some clinical significance for T2DM prevention in the future. Meanwhile, as the regulation of translation, existence form in serum and the catalytic specificity of GGT remain to be elucidated fully, the discovery of this non-linear relationship between GGT and T2DM risk, which is more complicated than a simple linear association, would support the further exploration of GGT. For example, though GGT does not depend on the transport of specific lipoproteins in serum, four interchangeable serum GGT fractions have been identified according to their molecular weight: big-GGT (b-GGT; 2000 kDa), medium-GGT (1000 kDa), small-GGT (250 kDa) and free-GGT (70 kDa); among which, there is limited evidence that b-GGT might be closely associated with metabolic syndrome.⁴⁹ However, the routine method used to measure serum GGT is the initial rate method, with a normal range of 0–50 IU/l, which is sensible, but non-specific because it cannot discriminate between the four existing GGT fractions. To a certain extent, this limitation of the GGT detection method hinders further interpretation of the relationship between GGT and T2DM, which makes it difficult to determine whether the complex non-linear relationship is caused by the different specificity of the GGT fractions.

All limitations of the original study are also applicable to the current study.¹² First, as all of the participants were of Japanese descent, further analyses are needed to establish the generalizability to populations with other genetic backgrounds. Secondly, the insulin concentration was not measured in the original study, so the interaction between insulin and GGT was not considered. Thirdly, the incidence of T2DM might be underestimated due to the lack of oral glucose tolerance testing. Lastly, as participants with under-medication, alcoholic fatty liver disease, viral hepatitis or abnormal liver function at baseline were excluded from the final analyses, the conclusion reached in this current study might not be applicable for the general population. Because of limited data availability, other potential covariates were not included in the adjusted model.

In conclusion, this current secondary analysis of a large-scale, long-term dataset in a Japanese population suggests that an increase in the circulating level of GGT might be a risk factor in the development of T2DM. This suggests the potential clinical value in considering the non-linear impact of GGT for routine screening of high-risk individuals to improve early intervention for T2DM prevention.

Supplemental Material

sj-pdf-1-imr-10.1177_0300060520937911 - Supplemental material for Non-linear relationship between gamma-glutamyl transferase and type 2 diabetes mellitus risk: secondary analysis of a prospective cohort study

Supplemental material, sj-pdf-1-imr-10.1177_0300060520937911 for Non-linear relationship between gamma-glutamyl transferase and type 2 diabetes mellitus risk: secondary analysis of a prospective cohort study by Hao Wang, Lixia Li and Shouyan Zhang in Journal of International Medical Research

Supplemental Material

sj-pdf-2-imr-10.1177_0300060520937911 - Supplemental material for Non-linear relationship between gamma-glutamyl transferase and type 2 diabetes mellitus risk: secondary analysis of a prospective cohort study

Supplemental material, sj-pdf-2-imr-10.1177_0300060520937911 for Non-linear relationship between gamma-glutamyl transferase and type 2 diabetes mellitus risk: secondary analysis of a prospective cohort study by Hao Wang, Lixia Li and Shouyan Zhang in Journal of International Medical Research

Footnotes

Acknowledgement

The authors thank all the staff members in their institutions. The authors specially thank the data providers of this study. They completed the whole study and shared their data selflessly and kindly.¹² They are (the rankings and institutions of these researchers are ranked according to the original reference12): Professor T Okamura, Professor Y Hashimoto, Professor M Hamaguchi, Professor A Obora, Professor T Kojima and Professor M Fukui (corresponding author) (Department of Diabetology, Kameoka Municipal Hospital, Kameoka, Japan. mhama@koto.kpu-m.ac.jp.). The authors also specially thank Professor Chi Chen (Guizhou Medical University; Yi-er College) for providing statistical analysis guidance for the current paper. The authors thank Changzhong Chen, Xinglin Chen, Jie Liu, Wenjing Liu, Wenquan Liang and Haibo Li of Yi-er College. We would like to thank Editage () for English language editing.

Author contributions

Hao Wang and Lixia Li contributed to data acquisition, analysis and manuscript drafting. Shouyan Zhang contributed to data interpretation, manuscript conception design and revision. All the authors read and approved the final manuscript.

Declaration of conflicting interests

The authors declare that there are no conflicts of interest.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

ORCID iDs

Lixia Li

Shouyan Zhang

Supplemental material

Supplemental material for this article is available online.

References

Centers for Disease Control and Prevention. National Diabetes Statistics Report, 2020. Atlanta, GA: Centers for Disease Control and Prevention, US Department of Health and Human Services, https://www.cdc.gov/diabetes/library/features/diabetes-stat-report.html (2020, accessed 24 June 2020).

Banerjee

Talukdar

Banerjee

, et al. Type II diabetes mellitus and obesity: common links, existing therapeutics and future developments. J Biosci 2019; 44: 150.

Ogurtsova

da Rocha Fernandes

Huang

, et al. IDF Diabetes Atlas: global estimates for the prevalence of diabetes for 2015 and 2040. Diabetes Res Clin Pract 2017; 128: 40–50.

Speisky

Shackel

Varghese

, et al. Role of hepatic gamma-glutamyltransferase in the degradation of circulating glutathione: studies in the intact guinea pig perfused liver. Hepatology 1990; 11: 843–849.

Henriksen

Diamond-Stanic

Marchionne

EM.

Oxidative stress and the etiology of insulin resistance and type 2 diabetes.

Free Radic Biol Med 2011; 51: 993–999.

Shin

Chang

Shin

, et al. Elevated serum gamma-glutamyltransferase levels are independently associated with insulin resistance in non-diabetic subjects. Diabetes Res Clin Pract 2009; 84: 152–157.

Sreeram

Suryakar

Dani

NH.

Is gamma-glutamyl transpeptidase a biomarker for oxidative stress in periodontitis?

J Indian Soc Periodontol 2015; 19: 150–154.

Tangvarasittichai

Oxidative stress, insulin resistance, dyslipidemia and type 2 diabetes mellitus.

World J Diabetes 2015; 6: 456–480.

Ahn

Shin

Nam

, et al. The association between liver enzymes and risk of type 2 diabetes: the Namwon study. Diabetol Metab Syndr 2014; 6: 14.

10.

Kunutsor

Abbasi

Adler

AI.

Gamma-glutamyl transferase and risk of type II diabetes: an updated systematic review and dose-response meta-analysis. Ann Epidemiol 2014; 24: 809–816.

11.

Lee

Cho

Burgess

, et al. Serum gamma-glutamyl transferase and risk of type 2 diabetes in the general Korean population: a Mendelian randomization study. Hum Mol Genet 2016; 25: 3877–3886.

12.

Okamura

Hashimoto

Hamaguchi

, et al. Ectopic fat obesity presents the greatest risk for incident type 2 diabetes: a population-based longitudinal study. Int J Obes (Lond) 2019; 43: 139–148.

13.

von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int J Surg 2014; 12: 1495–1499.

14.

Kernan

Viscoli

Brass

, et al. Phenylpropanolamine and the risk of hemorrhagic stroke. N Engl J Med 2000; 343: 1826–1832.

15.

Cleveland

WS.

Robust Locally Weighted Regression and Smoothing Scatterplots. J Am Stat Assoc 1979; 74: 829–836.

16.

Lin

Chen

XD.

The analysis of threshold effect using Empower Stats software. Zhonghua Liu Xing Bing Xue Za Zhi 2013; 34: 1139–1141 [Article in Chinese, English abstract].

17.

Hutcheon

Lisonkova

Joseph

KS.

Epidemiology of pre-eclampsia and the other hypertensive disorders of pregnancy.

Best Pract Res Clin Obstet Gynaecol 2011; 25: 391–403.

18.

Srivastava

Awasthi

Miller

, et al. Studies on gamma-glutamyl transpeptidase in human and rabbit erythrocytes. Blood 1976; 47: 645–650.

19.

Zhang

Forman

HJ.

Redox regulation of gamma-glutamyl transpeptidase.

Am J Respir Cell Mol Biol 2009; 41: 509–515.

20.

Heisterkamp

Groffen

Warburton

, et al. The human gamma-glutamyltransferase gene family. Hum Genet 2008; 123: 321–332.

21.

Shaw

London

Petersen

LE.

Isolation of gamma-glutamyltransferase from human liver, and comparison with the enzyme from human kidney. Clin Chem 1978; 24: 905–915.

22.

Hughey

Coyle

and Curthoys

NP.

Comparison of the association and orientation of gamma-glutamyltranspeptidase in lecithin vesicles and in native membranes. J Biol Chem 1979; 254: 1124–1128.

23.

Tsuji

Matsuda

and Katunuma

Studies on the structure of gamma-glutamyltranspeptidase. II. Location of the segment anchoring gamma-glutamyltranspeptidase to the membrane. J Biochem 1980; 87:1567–1571.

24.

Balakrishna

Prabhune

AA.

Gamma-glutamyl transferases: a structural, mechanistic and physiological perspective. Front Biol 2014; 9: 51–65.

25.

Whitfield

JB.

Gamma glutamyl transferase. Crit Rev Clin Lab Sci 2001; 38: 263–355.

26.

Eminler

Irak

Ayyildiz

, et al. The relation between liver histopathology and GGT levels in viral hepatitis: more important in hepatitis B. Turk J Gastroenterol 2014; 25: 411–415.

27.

Leonard

Neptun

Popp

JA.

Serum gamma glutamyl transferase as a specific indicator of bile duct lesions in the rat liver. Am J Pathol 1984; 116: 262–269.

28.

Lee

Evans

Robins

, et al. Gamma glutamyl transferase and metabolic syndrome, cardiovascular disease, and mortality risk: the Framingham Heart Study. Arterioscler Thromb Vasc Biol 2007; 27: 127–133.

29.

Bozduman

Yildirim

Cicek

Biomarkers of nondipper hypertension in prehypertensive and hypertensive patients. Biomark Med 2019; 13: 371–378.

30.

Lee

Silventoinen

Jacobs

, et al. Gamma-glutamyltransferase, obesity, and the risk of type 2 diabetes: observational cohort study among 20,158 middle-aged men and women. J Clin Endocrinol Metab 2004; 89: 5410–5414.

31.

Yang

Kim

, et al. Gamma-glutamyl transferase predicts future stroke: A Korean nationwide study. Ann Neurol 2018; 83: 375–386.

32.

Wang

, et al. γ-Glutamyl cyclotransferase contributes to tumor progression in high grade serous ovarian cancer by regulating epithelial-mesenchymal transition via activating PI3K/AKT/mTOR pathway. Gynecol Oncol 2018; 149: 163–172.

33.

Yang

Liu

, et al. Prognostic Significance of Serum Uric Acid and Gamma-Glutamyltransferase in Patients with Advanced Gastric Cancer. Dis Markers 2019; 2019: 1415421.

34.

Kubo

Hanaoka

Kuriki

, et al. Rapid detection of metastatic lymph nodes of colorectal cancer with a gamma-glutamyl transpeptidase-activatable fluorescence probe. Sci Rep 2018; 8: 17781.

35.

Meister

Glutathione-ascorbic acid antioxidant system in animals. J Biol Chem 1994; 269: 9397–9400.

36.

Lonardo

Romagnoli

Gamma glutamyl transferase: a novel cardiovascular outfit for an old liver test.

Indian J Med Res 2016; 143: 4–7.

37.

Lieberman

Wiseman

Shi

, et al. Growth retardation and cysteine deficiency in gamma-glutamyl transpeptidase-deficient mice. Proc Natl Acad Sci U S A 1996; 93: 7923–7926.

38.

Meister

On the enzymology of amino acid transport.

Science 1973; 180: 33–39.

39.

Nemesánszky

Lott

JA.

Gamma-glutamyltransferase and its isoenzymes: progress and problems.

Clin Chem 1985; 31: 797–803.

40.

Bachhawat

Kaur

Glutathione Degradation. Antioxid Redox Signal 2017; 27: 1200–1216.

41.

Chinta

Kumar

Zhang

, et al. Up-regulation of gamma-glutamyl transpeptidase activity following glutathione depletion has a compensatory rather than an inhibitory effect on mitochondrial complex I activity: implications for Parkinson’s disease. Free Radic Biol Med 2006; 40: 1557–1563.

42.

Lee

Jacobs

DR.

Association between serum gamma-glutamyltransferase and C-reactive protein.

Atherosclerosis 2005; 178: 327–330.

43.

Reuter

Schnekenburger

Cristofanon

, et al. Tumor necrosis factor alpha induces gamma-glutamyltransferase expression via nuclear factor-kappaB in cooperation with Sp1. Biochem Pharmacol 2009; 77: 397–411.

44.

Bradley

Fitzpatrick

Jacobs

, et al. Associations between γ-glutamyltransferase (GGT) and biomarkers of atherosclerosis: the Multi-ethnic Study of Atherosclerosis (MESA). Atherosclerosis 2014; 233: 387–393.

45.

Anderson

Allison

Meister

Interconversion of leukotrienes catalyzed by purified gamma-glutamyl transpeptidase: concomitant formation of leukotriene D4 and gamma-glutamyl amino acids.

Proc Natl Acad Sci U S A 1982; 79: 1088–1091.

46.

Wright

Scism-Bacon

Glass

LC.

Oxidative stress in type 2 diabetes: the role of fasting and postprandial glycaemia.

Int J Clin Pract 2006; 60: 308–314.

47.

Montonen

Boeing

Fritsche

, et al. Consumption of red meat and whole-grain bread in relation to biomarkers of obesity, inflammation, glucose metabolism and oxidative stress. Eur J Nutr 2013; 52: 337–345.

48.

Stranges

Dorn

Muti

, et al. Body fat distribution, relative weight, and liver enzyme levels: a population-based study. Hepatology 2004; 39: 754–763.

49.

Fornaciari

Fierabracci

Corti

, et al. Gamma-glutamyltransferase fractions in human plasma and bile: characteristic and biogenesis. PLoS One 2014; 9: e88532.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.09 MB

0.08 MB