Association between Gamma-Glutamyl Transferase Levels and Pancreatobiliary Cancer Risk in Patients with Diabetes: Evidence from the National Health Insurance Cooperation Health Checkup 2009 to 2012

Ji Hye Heo; Jun Goo Kang; Kyungdo Han; Kyong Joo Lee

doi:10.5009/gnl240322

Online first

Split Viewer

Online first

Association between Gamma-Glutamyl Transferase Levels and Pancreatobiliary Cancer Risk in Patients with Diabetes: Evidence from the National Health Insurance Cooperation Health Checkup 2009 to 2012

Ji Hye Heo^1,2 , Jun Goo Kang¹ , Kyungdo Han³ , Kyong Joo Lee¹

Author Info ⁺

¹Department of Internal Medicine, Hallym University College of Medicine, Chuncheon, Korea; ²Division of Endocrinology, Gerontology and Metabolism, Stanford University School of Medicine, Stanford, CA, USA; ³Department of Statistics and Actuarial Science, Soongsil University, Seoul, Korea

Correspondence to: Kyong Joo Lee
ORCID https://orcid.org/0000-0002-9175-9164
E-mail smild123@hallym.or.kr

Kyungdo Han
ORCID https://orcid.org/0000-0002-6096-1263
E-mail hkd917@naver.com

Received: July 13, 2024; Revised: September 24, 2024; Accepted: October 4, 2024

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Gut Liver.

Published online December 6, 2024

Abstract

Go to

Abstract
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ACKNOWLEDGEMENTS
CONFLICTS OF INTEREST
AUTHOR CONTRIBUTIONS
References

Background/Aims: Elevated gamma-glutamyl transferase (GGT) levels indicate hepatic dysfunction and have been linked to an increased risk of pancreatobiliary cancers. However, this association, particularly in individuals with diabetes mellitus (DM), requires elucidation. We aimed to examine the association between elevated serum GGT levels and pancreatobiliary cancer risk in patients with diabetes.
Methods: Our study included data from the National Health Insurance Service (NHIS) database for 2,459,966 adults aged >20 years diagnosed with DM between 2009 and 2012. We examined the association between serum GGT levels and pancreatobiliary cancer risk, considering DM-related factors. Serum GGT levels were categorized into quartiles, and Cox proportional hazards analysis was performed to evaluate the association between serum GGT levels and pancreatobiliary cancer risk.
Results: Over a median follow-up period of 7.2 years, 21,795 patients (0.89%) were newly diagnosed with pancreatobiliary cancer. The adjusted hazard ratio for pancreatobiliary cancer in quartiles 2–4 compared with that in quartile 1 was 1.091, 1.223, and 1.554, respectively, demonstrating a significant upward trend (p<0.001). This association remained consistent across all cancer types and was independent of the DM duration or treatment regimen.
Conclusions: Elevated serum GGT levels were independently associated with an increased risk of pancreatobiliary cancer, regardless of the duration of DM or the use of oral hypoglycemic agents and insulin. While these findings suggest the potential utility of serum GGT as a biomarker for identifying individuals at higher risk of pancreatobiliary cancer within the diabetic population, further research is needed to validate its clinical applicability.

Keywords: Pancreatobiliary cancer, Gamma-glutamyltransferase, Diabetes mellitus, Risk, Biomarkers

INTRODUCTION

Go to

Abstract
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ACKNOWLEDGEMENTS
CONFLICTS OF INTEREST
AUTHOR CONTRIBUTIONS
References

Serum gamma-glutamyl transferase (GGT) is a crucial indicator of hepatic dysfunction and surrogate marker of excessive alcohol consumption.^1,2 GGT is essential in glutathione metabolism and facilitates the transfer of amino acids across cell membranes, contributing to numerous cellular processes.³ Elevated GGT levels are linked to various health conditions, including metabolic syndrome,⁴ hypertension,⁵ diabetes mellitus (DM),⁶ cardiovascular disease,⁷ and cancer.⁸ Numerous studies have established GGT as a well-known marker of oxidative stress, which is a critical factor in carcinogenesis.^9-11 For instance, elevated GGT levels have been associated with an increased risk of various non-pancreatobiliary cancers, such as liver cancer,¹² colorectal cancer,¹³ and breast cancer,¹⁴ suggesting its potential role as a biomarker for cancer risk across different malignancies. Elevated GGT levels were positively correlated with the risk of gastrointestinal cancers, including pancreatobiliary cancer, which is a significant global health concern owing to its lethality.¹³ However, the consistency of these findings across studies remains variable, necessitating further investigations using a larger population cohort.⁸

DM, characterized by chronic hyperglycemia due to insulin resistance, is often accompanied by elevated GGT levels.^2,6,15 DM is associated with an increased susceptibility to gastrointestinal cancers, including pancreatobiliary cancer.^16-20 However, the direct relationship between serum GGT level and pancreatobiliary cancer risk, particularly in individuals with DM, remain unknown.²⁰

Understanding the complex relationship between serum GGT levels, glycemic status, and the risk of pancreatobiliary cancer is essential for uncovering new insights into the disease and improving strategies for risk assessment and intervention. This study aimed to examine the link between elevated serum GGT level and risk of pancreatobiliary cancer in diabetic patients, utilizing a large-scale dataset from the National Health Insurance Service (NHIS). Furthermore, we aimed to assess the impact of glycemic status on this relationship and investigate the combined effects of serum GGT levels and glycemic status on pancreatobiliary cancer incidence.

MATERIALS AND METHODS

Go to

Abstract
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ACKNOWLEDGEMENTS
CONFLICTS OF INTEREST
AUTHOR CONTRIBUTIONS
References

1. NHIS database and NHIS health checkup program

We used data from the NHIS, a government program initiative in 2000 that encompasses approximately 98% of the Korean population. The NHIS conducts a National Health Screening Program every 2 years for all insured individuals, which includes a self-reported survey on lifestyle factors, anthropometric data, and laboratory tests.²¹ Additionally, the NHIS database holds information on healthcare services covered for reimbursement, such as treatments, procedures, and diagnoses, categorized under the International Classification of Diseases, 10th revision (ICD-10). Certified hospitals authorized by the NHIS are responsible for conducting these health checkups and trained examiners regularly undergo qualification assessments administered by the NHIS.

2. Study population

We included 2,746,079 Korean adults with diabetes from the NHIS database who participated in health examinations between 2009 and 2012. Among them, we excluded individuals <20 years old (n=441); who had been previously diagnosed with cancer according to ICD-10, regardless of the type (n=84,796); with missing variables (n=177,502); and who received a diagnosis within 1 year following their enrolment (n=23,373) by applying a 1-year lag period to minimize reverse causality. Ultimately, 2,459,966 individuals with diabetes were included in the analysis (Fig. 1). The participants were monitored starting 1 year after the baseline examination date, up until either the diagnosis of pancreatobiliary cancer or December 31, 2018, whichever occurred first.

Figure 1.Flowchart of the selection process of study participants. NHIS, National Health Insurance Service.

3. Data collection and definition of comorbidities

Comprehensive demographic and lifestyle data were collected through standardized self-reported questionnaires. Income level was divided into two groups, with the lower 25% being one category. Smoking habits were classified as never smoked, ex-smoker, or current smoker. Alcohol intake was categorized as non (0 g/day), mild (<30 g/day), or heavy (≥30 g/day). Regular exercise was defined as participating in vigorous physical activity at least three times per week or moderate physical activity at least five times per week.

Anthropometric data, including height, weight, and waist circumference, were recorded, and body mass index (BMI) was calculated by dividing weight (in kilograms) by height (in meters) squared. Blood samples were taken after an overnight fast, and enzymatic methods were used to measure serum levels of glucose, total cholesterol, triglycerides, high-density lipoprotein cholesterol, aspartate aminotransferase, alanine aminotransferase, GGT, and creatinine. New-onset diabetes was characterized by fasting blood glucose ≥126 mg/dL at the index date, with no prior claims for diabetes (ICD-10 code E11-14) or prescriptions for antidiabetic medication before this index.²² The duration of diabetes was calculated from the time of first assignment of ICD-10 codes (E11.x–E14.x) alongside antidiabetic prescriptions up to the index date.²³ Hypertension was defined as blood pressure ≥140/90 mm Hg or at least one annual claim for antihypertensive prescriptions under ICD-10 codes I10-I15. Dyslipidemia was identified by either a total cholesterol level ≥240 mg/dL or at least one claim annually for lipid-lowering medications under the ICD-10 code E78.²² Chronic kidney disease was defined as an estimated glomerular filtration rate <60 mL/min/1.73 m², determining through a combination of ICD-10 codes N18-19, Z49, Z94.0, and Z99.2. Obesity was categorized by a BMI ≥25 kg/m², based on the Asia-Pacific guidelines from the World Health Organization.

This study was approved by the Institutional Review Board of Hallym University Sacred Heart Hospital (IRB No. HALLYM 2020-05-001). All the data were fully anonymized before access; therefore, the requirement for informed consent was waived.

4. Outcomes

Due to potential variations in GGT levels attributed to different measurement devices and the absence of a reported threshold linked to heightened pancreatobiliary cancer risk, individuals in the cohort were segmented into four groups based on their GGT quartile levels. The GGT levels were divided into sex-specific quartiles.^24,25 The cutoffs for each quartile for males were as follows: Q1 (<27 IU/L), Q2 (<43 IU/L), Q3 (<77 IU/L), and Q4 (≥77 IU/L). Similarly, that for females were Q1 (<17 IU/L), Q2 (<24 IU/L), Q3 (<36 IU/L), and Q4 (≥36 IU/L). The primary outcome was the incidence of pancreatobiliary cancer in each group, categorized by the GGT quartile level. The secondary objective was to evaluate the potential interaction between elevated GGT levels and the risk of pancreatobiliary cancer, stratified by factors such as the duration of DM (new-onset DM, type 2 DM of <5 years, and type 2 DM of ≥5 years), and use and number of oral hypoglycemic agents (OHA; classified as non, <3, and ≥3 agents), and insulin use. Pancreatobiliary cancers include pancreatic, common bile duct (CBD), and gallbladder (GB) cancers. Newly diagnosed pancreatobiliary cancer during the follow-up period were identified using the ICD-10 codes from the NHIS database (pancreatic cancer: C25, CBD cancer: C23-C24, GB cancer: C23), with the index date being the first recorded date of diagnosis.

5. Statistical analyses

Baseline characteristics of the participants are presented as means±standard deviations or medians (interquartile ranges) for continuous variables, and as numbers (percentages) for categorical variables. Group comparisons were made using the independent t-test for continuous variables and the chi-square test for categorical variables. The incidence of pancreatobiliary cancer was calculated by dividing the number of cases by 1,000 person-years. Cox proportional hazards regression was conducted to assess the association between GGT levels and pancreatobiliary cancer, with hazard ratios (HRs) and 95% confidence intervals (CIs) calculated. The multivariate-adjusted proportional hazards model was applied in several steps: model 1 was adjusted for age and sex, while model 2 additionally adjusted for smoking status, alcohol intake, regular exercise, BMI, hypertension, dyslipidemia, fasting glucose levels, diabetes duration, OHA use, and insulin use. The Kaplan-Meier method and log-rank tests were used to compare the cumulative incidence of pancreatobiliary cancer across the GGT quartiles. Statistical significance was defined as p<0.05. All analyses were conducted using SAS (version 9.4; SAS Institute Inc., Cary, NC, USA) and R 3.1.0 (R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Go to

Abstract
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ACKNOWLEDGEMENTS
CONFLICTS OF INTEREST
AUTHOR CONTRIBUTIONS
References

1. Baseline characteristics of the study population

The analysis included 2,459,966 individuals; 21,795 individuals (0.89%) received a new diagnosis of pancreatobiliary cancer over a median follow-up period of 7.2 years (interquartile range, 6.0 to 8.1 years). The mean age of the participants was 57.2±12.3 years, with 60.1% being male. The baseline characteristics of the participants stratified into four quartiles based on GGT levels are shown in Table 1. The higher GGT quartile groups (Q3–Q4) had a higher proportion of individuals who consumed alcohol or smoked than the lower quartile groups (Q1–Q2). The prevalence of hypertension, dyslipidemia, and chronic kidney disease progressively increased with increasing GGT quartiles. The prevalence of new-onset DM exhibited a rising trend with increasing GGT quartiles, whereas the proportion of type 2 diabetes of ≥5 years decreased with increasing GGT quartiles. Furthermore, OHA and insulin use decreased with increasing GGT quartiles.

Table 1. Baseline Characteristics

Variable	Total (n=2,459,966)	GGT				p-value
Variable	Total (n=2,459,966)	Q1 (n=618,846)	Q2 (n=621,445)	Q3 (n=610,091)	Q4 (n=609,584)	p-value
Age, yr	57.2±12.3	59.3±13.2	58.2±12.2	56.3±11.9	54.8±11.6	<0.001
Male sex	1,479,229 (60.1)	370,888 (59.9)	368,091 (59.2)	372,527 (61.1)	367,723 (60.3)	<0.001
Income, low 25%	518,388 (21.1)	129,529 (21.0)	130,121 (21.0)	126,404 (20.7)	132,334 (21.7)	<0.001
BMI, kg/m²	25.1±3.4	23.8±3.0	24.9±3.2	25.7±3.3	25.9±3.6	<0.001
Smoking						<0.001
Never	1,362,183 (55.4)	376,567 (60.9)	357,743 (57.6)	323,418 (53.0)	304,455 (50.0)
Ex	449,365 (18.3)	118,625 (19.2)	116,711 (18.8)	112,659 (18.5)	101,370 (16.6)
Current	648,418 (26.3)	123,654 (19.9)	146,991 (23.6)	174,014 (28.5)	203,759 (33.4)
Drinking						<0.001
Non (0 g/day)	1,394,234 (56.7)	435,419 (70.4)	383,809 (61.8)	317,317 (52.0)	257,689 (42.3)
Mild (<30 g/day)	816,777 (33.2)	164,202 (26.5)	199,708 (32.1)	225,062 (36.9)	227,805 (37.4)
Heavy (≥30 g/day)	248,955 (10.1)	19,225 (3.1)	37,928 (6.1)	67,712 (11.1)	124,090 (20.3)
Regular exercise	504,585 (20.5)	143,035 (23.1)	133,003 (21.4)	119,783 (19.6)	108,764 (17.8)	<0.001
Hypertension	966,238 (39.3)	207,611 (33.6)	237,363 (38.2)	250,728 (41.1)	270,536 (44.4)	<0.001
Dyslipidemia	685,228 (27.9)	125,565 (20.3)	164,452 (26.5)	185,813 (30.5)	209,398 (34.4)	<0.001
CKD	280,753 (11.4)	83,948 (13.6)	75,861 (12.2)	65,055 (10.7)	55,889 (9.2)	<0.001
DM duration						<0.001
New onset	981,237 (39.9)	209,294 (33.8)	228,665 (36.8)	254,622 (41.7)	288,656 (47.4)
<5 yr	769,850 (31.3)	171,995 (27.8)	195,891 (31.5)	201,167 (33.0)	200,797 (32.9)
≥5 yr	708,879 (28.8)	237,557 (38.4)	196,889 (31.7)	154,302 (25.3)	120,131 (19.7)
OHA						<0.001
Non	1,017,164 (41.4)	223,013 (36.0)	237,946 (38.3)	261,346 (42.8)	294,859 (48.4)
<3	1,090,638 (44.3)	289,488 (46.3)	289,530 (46.6)	267,883 (43.9)	243,737 (40.0)
≥3	352,164 (14.3)	106,345 (17.2)	93,969 (15.1)	80,862 (13.3)	70,988 (11.6)
Waist circumference, cm	85.46±8.89	82.45±8.63	85.10±8.58	86.86±8.60	87.50±8.87	<0.001
Fasting glucose, mg/dL	145.02±47.03	137.93±45.36	143.01±46.56	147.26±46.94	152.01±48.08	<0.001
Total cholesterol, mg/dL	197.07±45.92	184.73±42.89	195.1±42.85	201.73±45.07	206.92±49.61	<0.001
HDL-C, mg/dL	52.26±28.69	51.77±28.49	51.66±28.94	51.86±28.56	53.76±28.72	<0.001
LDL-C, mg/dL	112.92±84.52	109.46±78.23	114.33±76.19	115.3±75.62	112.6±104.65	<0.001
AST, IU/L	26.15 (26.14–26.16)	20.98 (20.96–20.99)	23.10 (23.08–23.12)	26.46 (26.44–26.48)	36.68 (36.63–36.73)	<0.001
ALT, IU/L	26.31 (26.29–26.33)	18.36 (18.34–18.38)	22.90 (22.88–22.93)	28.60 (28.56–28.63)	40.16 (40.10–40.22)	<0.001
Estimated GFR, mL/min/1.73 m²	85.06±36.20	83.68±36.61	83.97±35.78	85.27±35.83	87.36±36.46	<0.001
Triglyceride, mg/dL	146.56 (146.45–146.66)	113.15 (113.00–113.29)	138.03 (137.85–138.21)	160.29 (160.07–160.50)	185.22 (184.95–185.49)	<0.001
GGT, IU/L	36.92 (36.89–36.96)	16.12 (16.1–16.13)	27.02 (27.00–27.04)	42.98 (42.95–43.02)	101.18 (101.00–101.35)	<0.001

Data are presented as the mean±SD, number (%), median (interquartile range).

GGT, gamma-glutamyl transferase; BMI, body mass index; CKD, chronic kidney disease; DM, diabetes mellitus; OHA, oral hypoglycemic agent; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; AST, aspartate aminotransferase; ALT, alanine aminotransferase; GFR, glomerular filtration rate.

2. Association between serum GGT level and pancreatobiliary cancer risk

During the follow-up period, a total of 21,795 new cases of pancreatobiliary cancer were identified, including 11,776 cases of pancreatic cancer, 3,168 cases of CBD cancer, and 8,807 cases of GB cancer. Table 2 presents a comprehensive breakdown of theses incident cases, incidence rates, and adjusted HRs based on serum GGT quartile levels for each type of pancreatobiliary cancer. A significant association between serum GGT levels and pancreatobiliary cancer risk was observed after adjusting for various covariates, including age, sex, BMI, smoking, drinking, regular exercise, hypertension, dyslipidemia, fasting glucose, DM duration, OHA use, and insulin use. For pancreatobiliary cancer, the adjusted HRs for Q2, Q3, and Q4 compared to Q1 were 1.09 (95% CI, 1.05 to 1.13), 1.22 (1.18 to 1.27), and 1.55 (1.49 to 1.62), respectively (p for trend <0.001). Notably, the risk of pancreatobiliary cancer consistently increased with rising serum GGT levels across all cancer types. For pancreatic cancer, the adjusted HRs for Q2, Q3, and Q4 compared to Q1 were 1.06 (95% CI, 1.01 to 1.11), 1.09 (95% CI, 1.03 to 1.15), and 1.29 (95% CI, 1.22 to 1.36), respectively (p for trend <0.001). For CBD cancer, the adjusted HRs for Q2, Q3, and Q4 compared to Q1 were 1.07 (95% CI, 0.97 to 1.18), 1.29 (95% CI, 1.16 to 1.42), and 1.52 (95% CI, 1.37 to 1.69), respectively (p for trend <0.001). For GB cancer, the adjusted HRs for Q2, Q3, and Q4 compared to Q1 were 1.15 (95% CI, 1.08 to 1.22), 1.41 (95% CI, 1.33 to 1.50), and 1.99 (95% CI, 1.87 to 2.12), respectively (p for trend <0.001). Participants in the higher GGT quartiles exhibited a significantly greater cumulative incidence of pancreatobiliary cancer than those in the lowest quartiles (Fig. 2). The Kaplan-Meier analysis indicated that individuals in the highest GGT quartile had a markedly higher probability of developing pancreatobiliary cancer compared to those in the lowest quartile. Moreover, the increased risk of pancreatobiliary cancer in the upper GGT quartiles observed in the earlier years persisted through a median follow-up of 7.2 years (p<0.001, log-rank test).

Figure 2.Association between serum gamma-glutamyl transferase (GGT) quartile level and pancreatobiliary cancer risk. (A) Pancreatobiliary cancer, (B) pancreatic cancer, (C) common bile duct cancer, and (D) gallbladder cancer. Q1-Q4, GGT quartile level. All p for trend <0.001.

Table 2. Association between Serum GGT Level and Pancreatobiliary Cancer Risk

GGT	No.	Event	Incidence rate per 1,000 PY	Model 1, HR (95% CI)^*	Model 2, HR (95% CI)^†
Pancreatobiliary cancer
Q1	618,846	5,639	1.33	1 (reference)	1 (reference)
Q2	621,445	5,426	1.26	1.08 (1.04–1.12)	1.09 (1.05–1.13)
Q3	610,091	5,133	1.21	1.20 (1.16–1.25)	1.22 (1.18–1.27)
Q4	609,584	5,597	1.35	1.53 (1.47–1.59)	1.55 (1.49–1.62)
Pancreatic cancer
Q1	618,846	3,243	0.76	1 (reference)	1 (reference)
Q2	621,445	3,064	0.71	1.04 (0.99–1.10)	1.06 (1.01–1.11)
Q3	610,091	2,696	0.64	1.07 (1.02–1.13)	1.09 (1.03–1.15)
Q4	609,584	2,773	0.67	1.27 (1.21–1.34)	1.29 (1.22–1.36)
Common bile duct cancer
Q1	618,846	817	0.19	1 (reference)	1 (reference)
Q2	621,445	780	0.18	1.06 (0.97–1.18)	1.07 (0.97–1.18)
Q3	610,091	786	0.19	1.27 (1.15–1.41)	1.29 (1.16–1.42)
Q4	609,584	785	0.19	1.50 (1.36–1.66)	1.52 (1.37–1.69)
Gallbladder cancer
Q1	618,846	2,097	0.49	1 (reference)	1 (reference)
Q2	621,445	2,070	0.48	1.13 (1.06–1.20)	1.15 (1.08–1.22)
Q3	610,091	2,120	0.50	1.38 (1.30–1.47)	1.41 (1.33–1.50)
Q4	609,584	2,520	0.61	1.95 (1.84–2.07)	1.99 (1.87–2.12)

GGT, gamma-glutamyl transferase; PY, person-year; HR, hazard ratio; CI, confidence interval.

*Adjusted for age, sex; ^†Adjusted for age, sex, smoking, drinking, regular exercise, body mass index, hypertension, dyslipidemia, fasting glucose, diabetes mellitus duration, oral hypoglycemic agent, insulin.

3. Effect of serum GGT level on pancreatobiliary cancer risk stratified by comorbidities

We performed stratified analyses to examine the association between GGT levels and pancreatobiliary cancer risk, stratified by DM duration, OHA use, and insulin use. In the subgroup analyses based on DM duration, and OHA and insulin use, we observed a consistent increase in pancreatobiliary cancer risk with escalating serum GGT levels across all subgroups (Table 3). Subgroup analyses by age (20–64 years vs ≥65 years), sex (male vs female), BMI (<25 kg/m² vs ≥25 kg/m²), smoking status (never smoker vs ex-smoker vs current smoker), and alcohol consumption (non drinker vs mild drinker vs heavy drinker) showed an increased risk of pancreatobiliary cancer with increasing GGT levels across all subgroups. Furthermore, the increased risk of pancreatobiliary cancer associated with increased GGT levels was significantly more pronounced in older adults, non-obese individuals, and ex-smokers than that in adults younger than 65 years, obese individuals, and never smokers or current smokers, respectively (p values for interaction <0.05).

Table 3. Subgroup Analysis of the Effect of Serum GGT Level on Pancreatobiliary Cancer Risk Stratified by Comorbidities

Subgroup		GGT	No.	Pancreatobiliary cancer event	Incidence rate per 1,000 PY	Model 1, HR (95% CI)^*	Model 2, HR (95% CI)^†
Age	20–64 yr	Q1	382,977	1,836	0.68	1 (reference)	1 (reference)
		Q2	420,250	2,202	0.74	1.09 (1.03–1.16)	1.11 (1.04–1.18)
		Q3	452,369	2,385	0.75	1.11 (1.04–1.17)	1.19 (1.12–1.26)
		Q4	483,130	2,934	0.88	1.30 (1.23–1.38)	1.49 (1.40–1.58)
	≥65 yr	Q1	235,869	3,803	2.46	1 (reference)	1 (reference)
		Q2	201,195	3,224	2.39	0.97 (0.93–1.02)	1.08 (1.03–1.13)
		Q3	157,722	2,748	2.59	1.05 (1.01–1.11)	1.24 (1.18–1.31)
		Q4	126,454	2,663	3.23	1.32 (1.25–1.38)	1.61 (1.53–1.69)
	p for interaction					0.004	0.039
Sex	Male	Q1	370,888	3,729	1.47	1 (reference)	1 (reference)
		Q2	368,091	3,387	1.33	0.90 (0.86–0.95)	1.11 (1.06–1.16)
		Q3	372,527	3,088	1.20	0.81 (0.78–0.85)	1.22 (1.17–1.29)
		Q4	367,723	3,413	1.37	0.93 (0.89–0.98)	1.60 (1.52–1.69)
	Female	Q1	247,958	1,910	1.11	1 (reference)	1 (reference)
		Q2	253,354	2,039	1.15	1.03 (0.97–1.10)	1.06 (1.00–1.13)
		Q3	237,564	2,045	1.23	1.11 (1.04–1.18)	1.22 (1.14–1.30)
		Q4	241,861	2,184	1.31	1.18 (1.11–1.26)	1.48 (1.39–1.58)
	p for interaction					<0.001	0.169
BMI	<25 kg/m²	Q1	417,367	3,794	1.33	1 (reference)	1 (reference)
		Q2	328,446	2,947	1.30	0.98 (0.93–1.02)	1.08 (1.03–1.14)
		Q3	262,804	2,354	1.30	0.97 (0.93–1.03)	1.20 (1.14–1.26)
		Q4	246,503	2,664	1.62	1.22 (1.16–1.28)	1.61 (1.52–1.70)
	≥25 kg/m²	Q1	201,479	1,845	1.32	1 (reference)	1 (reference)
		Q2	292,999	2,479	1.21	0.92 (0.86–0.98)	1.09 (1.03–1.16)
		Q3	347,287	2,779	1.14	0.87 (0.82–0.92)	1.23 (1.16–1.30)
		Q4	363,081	2,933	1.17	0.89 (0.84–0.95)	1.49 (1.40–1.58)
	p for interaction					<0.001	0.038
Smoking	Never	Q1	376,567	3,265	1.26	1 (reference)	1 (reference)
		Q2	357,743	3,102	1.24	0.99 (0.94–1.04)	1.10 (1.04–1.15)
		Q3	323,418	2,879	1.27	1.01 (0.96–1.06)	1.26 (1.20–1.33)
		Q4	304,455	2,927	1.40	1.11 (1.06–1.17)	1.57 (1.49–1.65)
	Ex	Q1	118,625	1,220	1.50	1 (reference)	1 (reference)
		Q2	116,711	1,058	1.31	0.87 (0.80–0.95)	1.08 (0.99–1.17)
		Q3	112,659	984	1.26	0.84 (0.77–0.92)	1.26 (1.16–1.37)
		Q4	101,370	1,002	1.45	0.97 (0.89–1.06)	1.68 (1.54–1.82)
	Current	Q1	123,654	1,154	1.38	1 (reference)	1 (reference)
		Q2	146,991	1,266	1.26	0.91 (0.84–0.99)	1.08 (0.99–1.17)
		Q3	174,014	1,270	1.06	0.77 (0.71–0.84)	1.10 (1.01–1.19)
		Q4	203,759	1,668	1.22	0.89 (0.82–0.96)	1.42 (1.32–1.54)
	p for interaction					<0.001	0.008
Drinking	Non (0 g/day)	Q1	435,419	4,232	1.42	1 (reference)	1 (reference)
		Q2	383,809	3,618	1.36	0.96 (0.91–1.00)	1.09 (1.04–1.14)
		Q3	317,317	2,996	1.36	0.95 (0.91–1.00)	1.23 (1.17–1.29)
		Q4	257,689	2,671	1.52	1.07 (1.02–1.12)	1.55 (1.48–1.64)
	Mild (<30 g/day)	Q1	164,202	1,261	1.10	1 (reference)	1 (reference)
		Q2	199,708	1,503	1.08	0.98 (0.91–1.05)	1.09 (1.01–1.18)
		Q3	225,062	1,575	1.01	0.91 (0.85–0.98)	1.18 (1.10–1.27)
		Q4	227,805	1,802	1.16	1.06 (0.98–1.14)	1.55 (1.44–1.66)
	Heavy (≥30 g/day)	Q1	19,225	146	1.10	1 (reference)	1 (reference)
		Q2	37,928	305	1.16	1.06 (0.87–1.29)	1.16 (0.95–1.41)
		Q3	67,712	562	1.20	1.09 (0.91–1.31)	1.37 (1.14–1.64)
		Q4	124,090	1,124	1.34	1.27 (1.03–1.46)	1.66 (1.40–1.97)
	p for interaction					0.471	0.796
DM duration	New onset	Q1	209,294	1,172	0.83	1 (reference)	1 (reference)
		Q2	228,665	1,379	0.88	1.07 (0.99–1.15)	1.15 (1.07–1.25)
		Q3	254,622	1,431	0.82	0.99 (0.92–1.07)	1.22 (1.13–1.32)
		Q4	288,656	1,869	0.96	1.16 (1.08–1.25)	1.56 (1.45–1.69)
	<5 yr	Q1	171,995	1,676	1.40	1 (reference)	1 (reference)
		Q2	195,891	1,833	1.33	0.95 (0.89–1.01)	1.07 (1.00–1.15)
		Q3	201,167	1,872	1.32	0.94 (0.88–1.01)	1.20 (1.13–1.29)
		Q4	200,797	2,057	1.48	1.06 (0.99–1.13)	1.51 (1.41–1.61)
	≥5 yr	Q1	237,557	2,791	1.71	1 (reference)	1 (reference)
		Q2	196,889	2,214	1.61	0.94 (0.89–1.00)	1.07 (1.01–1.13)
		Q3	154,302	1,830	1.70	0.99 (0.94–1.05)	1.24 (1.17–1.32)
		Q4	120,131	1,671	2.05	1.20 (1.13–1.27)	1.60 (1.50–1.70)
	p for interaction					0.005	0.435
OHA	Non	Q1	223,013	1,300	0.86	1 (reference)	1 (reference)
		Q2	237,946	1,472	0.91	1.05 (0.98–1.13)	1.15 (1.07–1.24)
		Q3	261,346	1,504	0.84	0.98 (0.91–1.05)	1.23 (1.14–1.33)
		Q4	294,859	1,966	0.99	1.15 (1.07–1.23)	1.59 (1.48–1.71)
	<3	Q1	289,488	3,073	1.53	1 (reference)	1 (reference)
		Q2	289,530	2,927	1.44	0.94 (0.89–0.99)	1.08 (1.02–1.13)
		Q3	267,883	2,734	1.45	0.95 (0.90–10.00)	1.23 (1.17–1.30)
		Q4	243,737	2,692	1.60	1.05 (0.99–1.10)	1.52 (1.44–1.60)
	≥3	Q1	106,345	1,266	1.71	1 (reference)	1 (reference)
		Q2	93,969	1,027	1.55	0.90 (0.83–0.98)	1.06 (0.97–1.15)
		Q3	80,862	895	1.57	0.92 (0.84–1.00)	1.20 (1.10–1.31)
		Q4	70,988	939	1.92	1.12 (1.03–1.22)	1.61 (1.48–1.75)
	p for interaction					0.029	0.421
Insulin	No	Q1	550,937	4,840	1.27	1 (reference)	1 (reference)
		Q2	567,896	4,812	1.22	0.95 (0.92–0.99)	1.1 (1.05–1.14)
		Q3	566,712	4,580	1.16	0.91 (0.87–0.95)	1.22 (1.17–1.27)
		Q4	567,500	4,979	1.28	1.01 (0.97–1.05)	1.54 (1.48–1.61)
	Yes	Q1	67,909	799	1.79	1 (reference)	1 (reference)
		Q2	53,549	614	1.72	0.96 (0.87–1.07)	1.04 (0.93–1.15)
		Q3	43,379	553	1.92	1.07 (0.96–1.19)	1.25 (1.12–1.39)
		Q4	42,084	618	2.29	1.29 (1.16–1.43)	1.64 (1.48–1.83)
	p for interaction					<0.001	0.249

GGT, gamma-glutamyl transferase; PY, person-year; HR, hazard ratio; CI, confidence interval; BMI, body mass index; DM, diabetes mellitus; OHA, oral hypoglycemic agent.

*Adjusted for age, sex; ^†Adjusted for age, sex, smoking, drinking, regular exercise, BMI, hypertension, dyslipidemia, fasting glucose, DM duration, OHA, and insulin.

DISCUSSION

Go to

Abstract
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ACKNOWLEDGEMENTS
CONFLICTS OF INTEREST
AUTHOR CONTRIBUTIONS
References

In this large cohort study using NHIS data, we observed that individuals in the highest GGT quartile had an approximately 55% higher risk of pancreatobiliary cancer, across all subtypes, than those in the lowest quartile among the Korean population with diabetes over a median 7.2-year follow-up period. This elevated risk persisted regardless of DM duration or OHA and insulin use, underscoring the potential of GGT level as a valuable biomarker for identifying individuals with type 2 DM at an increased risk of pancreatobiliary cancer, irrespective of their DM status or treatment type.

Our study demonstrates that increased GGT levels in patients with DM are associated with an increased risk of pancreatobiliary cancer, including pancreatic, CBD, and GB cancers. Interestingly, even within normal GGT ranges (Q1, Q2, Q3 for females; Q1, Q2 for males), diabetic individuals may still face an elevated risk, suggesting GGT could serve as a clinically significant biomarker, even within traditionally normal limits. Previous population-based studies have shown inconsistent results regarding GGT and pancreatobiliary cancer risk, with focusing specifically on diabetic patients. In the Ohsaki Cohort involving 15,031 Japanese adults, individuals in the highest GGT quartile exhibited an 89% higher risk of pancreatic cancer than those in the lowest quartile. However, this trend was primarily observed in current drinkers rather than never drinkers, largely reflecting alcohol consumption patterns,²⁶ as GGT is a marker for heavy alcohol use.²⁷ The U.K. Biobank Cohort, comprising 421,032 adults, also supported the association between elevated serum GGT levels and an increased risk of pancreatic cancer,²⁸ although alcohol consumption was not considered beyond the broad spectrum. Our large-scale NHIS cohort study of 2,459,966 participants showed a 29% increased risk of pancreatic cancer in diabetics in the highest GGT quartile, even after adjusting for alcohol consumption. Yin et al.²⁹ also found elevated GGT to be a marker of aggressive tumor behavior in intrahepatic cholangiocarcinoma, though their sample size was limited. In our study, those with the highest GGT levels had a 52% and 99% increased risk of CBD and GB cancer, respectively, regardless of alcohol consumption (non, mild, or heavy drinkers). Establishing an association between GGT levels and pancreatobiliary cancer in individuals with diabetes is crucial because DM is an established risk factor for pancreatobiliary cancer.^30,31 However, to our knowledge, only a few studies have investigated the relationship between serum GGT levels and the development of pancreatobiliary cancer in people with diabetes, considering DM status, duration, and treatment regimen. Therefore, we stratified the participants according to DM duration, number of OHAs used, and insulin use and found that an increase in GGT level was associated with a higher risk of pancreatobiliary cancer across all diabetes status subgroups. While the risk did not notably increase with longer diabetes duration, and increased OHA or insulin use compared with the controls, pancreatobiliary cancer risk increased consistently with increasing GGT levels, regardless of the various conditions in individuals with DM. These results further support the link between GGT levels and pancreatobiliary cancer risk in patients with diabetes. A recent Korean study demonstrated that higher GGT levels were associated with a higher risk of gastrointestinal cancers, including esophageal, stomach, and colorectal cancers, which was more pronounced in individuals with DM than in those without DM.¹³

The association between high GGT levels, diabetes, and pancreatobiliary cancer is complex, involving multiple interrelated mechanisms. GGT, an enzyme primarily found in the liver, plays a crucial role in various metabolic pathways, including glucose metabolism and oxidative stress regulation.⁹ In patients with diabetes, chronic hyperglycemia may contribute to elevated GGT levels, possibly through pathways involving oxidative stress-induced hepatic injury.³² Elevated GGT levels may impair red blood cell membranes, resulting in the release of harmful transition metals, which can initiate a cascade of pro-oxidant reactions.⁹ Consequently, elevated GGT levels have been implicated in the pathogenesis of various cancers, including pancreatobiliary cancer, through mechanisms, such as oxidative stress-induced DNA damage, chronic inflammation, and dysregulation of cellular signaling pathways.³³ In the context of intrahepatic cholangiocarcinoma, elevated GGT levels are more likely a consequence of cholestasis caused by tumor progression, reflecting advanced disease and poor prognosis.²⁹ This mechanism differs from the GGT elevation seen in diabetic patients or those with chronic alcohol use, where the increase is often linked to underlying metabolic dysfunction or liver injury, rather than direct tumor-related obstruction. Recognizing the distinct mechanisms of GGT elevation in different patient populations is crucial for understanding its role as a biomarker. The complex relationship between metabolic dysregulation, oxidative stress, and carcinogenesis highlights the need for further research to clarify these pathways and explore potential therapeutic targets for preventing or treating pancreatobiliary cancer in patients with diabetes and elevated GGT levels.

Serum GGT measurements are quick and cost-effective tests routinely performed in clinical practice. Additionally, elevated GGT levels have been associated with various health conditions, including diabetes and cancer.³⁴ Thus, integrating GGT measurements into routine clinical assessments of patients with DM may potentially contribute to early detection and risk stratification of pancreatobiliary cancer, facilitating timely interventions and improved patient outcomes.

The strengths of our study include its large-scale investigation using data from the NHIS database, which allowed a comprehensive analysis of the association between serum GGT levels and pancreatobiliary cancer risk. Additionally, our study accounted for potential confounders in the analyses and the various stratified analyses based on the characteristics of the study participants enhanced the reliability of our findings. Nonetheless, this study had some limitations. First, the observational nature of the study limited the ability to establish a causal relationship between elevated GGT levels and pancreatobiliary cancer risk. Although we observed a significant association between these two factors, further longitudinal or interventional studies are required to confirm the causality. Second, the use of administrative data, although comprehensive, may introduce bias or misclassification of variables due to factors, such as coding errors or incomplete records. Additionally, the exclusion of intrahepatic cholangiocarcinoma and perihilar cholangiocarcinoma from our analysis may limit the generalizability of our findings, as these subtypes can affect the overall risk assessment of pancreatobiliary cancer. Finally, our study focused on the Korean population, which may limit the generalizability of our findings to other populations.

In conclusion, our study supports the association between elevated serum GGT levels and pancreatobiliary cancer risk in patients with diabetes, independent of diabetes status and treatment regimen. These findings highlight the potential utility of GGT as a biomarker for an increased risk of pancreatobiliary cancer and underscores the importance of further research in this area. However, further prospective studies are required to elucidate the underlying mechanisms linking GGT levels to pancreatobiliary cancer risk and validate the GGT as a biomarker for the early detection and risk stratification of pancreatobiliary cancer.

ACKNOWLEDGEMENTS

Go to

Abstract
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ACKNOWLEDGEMENTS
CONFLICTS OF INTEREST
AUTHOR CONTRIBUTIONS
References

This study was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (No. RS-2023-00243402).

CONFLICTS OF INTEREST

Go to

Abstract
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ACKNOWLEDGEMENTS
CONFLICTS OF INTEREST
AUTHOR CONTRIBUTIONS
References

No potential conflict of interest relevant to this article was reported.

AUTHOR CONTRIBUTIONS

Go to

Abstract
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ACKNOWLEDGEMENTS
CONFLICTS OF INTEREST
AUTHOR CONTRIBUTIONS
References

Study concept: J.H.H., K.H., J.G.K., K.J.L. Study design: J.H.H., K.H., J.G.K. Data acquisition: J.H.H., K.H. Data analysis and interpretation: J.H.H., K.H., J.G.K., K.J.L. Statistical analysis: K.H., J.H.H., K.J.L. Drafting of the manuscript: J.H.H., K.H., J.G.K., K.J.L. Critical revision of the manuscript for important intellectual content: J.H.H., K.H., J.G.K., K.J.L.

References

Go to

Abstract
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ACKNOWLEDGEMENTS
CONFLICTS OF INTEREST
AUTHOR CONTRIBUTIONS
References

Teschke R, Rauen J, Neuefeind M, Petrides AS, Strohmeyer G, assignee. Alcoholic liver disease associated with increased gamma-glutamyltransferase activities in serum and liver. Adv Exp Med Biol 1980;132:647-654.
Ishizaka N, Ishizaka Y, Toda E, Yamakado M, Koike K, Nagai R, assignee. Association between gamma-glutamyltransferase levels and insulin resistance according to alcohol consumption and number of cigarettes smoked. J Atheroscler Thromb 2010;17:476-485.
Pompella A, Corti A, Paolicchi A, Giommarelli C, Zunino F, assignee. Gamma-glutamyltransferase, redox regulation and cancer drug resistance. Curr Opin Pharmacol 2007;7:360-366.
Lee DS, Evans JC, Robins SJ, et al, assignee. Gamma glutamyl transferase and metabolic syndrome, cardiovascular disease, and mortality risk: the Framingham Heart Study. Arterioscler Thromb Vasc Biol 2007;27:127-133.
Lee DH, Ha MH, Kim JH, et al, assignee. Gamma-glutamyltransferase and diabetes: a 4 year follow-up study. Diabetologia 2003;46:359-364.
Lee DH, Silventoinen K, Jacobs DR, Jousilahti P, Tuomileto J, assignee. 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.
Emdin M, Passino C, Pompella A, Paolicchi A, assignee. Gamma-glutamyltransferase as a cardiovascular risk factor. Eur Heart J 2006;27:2145-2146.
Mok Y, Son DK, Yun YD, Jee SH, Samet JM, assignee. γ-Glutamyltransferase and cancer risk: the Korean cancer prevention study. Int J Cancer 2016;138:311-319.
Koenig G, Seneff S, assignee. Gamma-glutamyltransferase: a predictive biomarker of cellular antioxidant inadequacy and disease risk. Dis Markers 2015;2015:818570.
Strasak AM, Pfeiffer RM, Klenk J, et al, assignee. Prospective study of the association of gamma-glutamyltransferase with cancer incidence in women. Int J Cancer 2008;123:1902-1906.
Strasak AM, Rapp K, Brant LJ, et al, assignee. Association of gamma-glutamyltransferase and risk of cancer incidence in men: a prospective study. Cancer Res 2008;68:3970-3977.
Kazemi-Shirazi L, Endler G, Winkler S, Schickbauer T, Wagner O, Marsik C, assignee. Gamma glutamyltransferase and long-term survival: is it just the liver?. Clin Chem 2007;53:940-946.
Hong SW, Lee HJ, Han K, et al, assignee. Risk of gastrointestinal cancer in patients with an elevated level of gamma-glutamyltransferase: a nationwide population-based study. PLoS One 2021;16:e0245052.
Staudigl C, Concin N, Grimm C, et al, assignee. Prognostic relevance of pretherapeutic gamma-glutamyltransferase in patients with primary metastatic breast cancer. PLoS One 2015;10:e0125317.
Nano J, Muka T, Ligthart S, et al, assignee. Gamma-glutamyltransferase levels, prediabetes and type 2 diabetes: a Mendelian randomization study. Int J Epidemiol 2017;46:1400-1409.
Ben Q, Xu M, Ning X, et al, assignee. Diabetes mellitus and risk of pancreatic cancer: a meta-analysis of cohort studies. Eur J Cancer 2011;47:1928-1937.
Jing W, Jin G, Zhou X, et al, assignee. Diabetes mellitus and increased risk of cholangiocarcinoma: a meta-analysis. Eur J Cancer Prev 2012;21:24-31.
Ma Y, Yang W, Song M, et al, assignee. Type 2 diabetes and risk of colorectal cancer in two large U.S. prospective cohorts. Br J Cancer 2018;119:1436-1442.
Valent F, assignee. Diabetes mellitus and cancer of the digestive organs: an Italian population-based cohort study. J Diabetes Complications 2015;29:1056-1061.
Van Hemelrijck M, Jassem W, Walldius G, et al, assignee. Gamma-glutamyltransferase and risk of cancer in a cohort of 545,460 persons: the Swedish AMORIS study. Eur J Cancer 2011;47:2033-2041.
Lee J, Lee JS, Park SH, Shin SA, Kim K, assignee. Cohort profile: the National Health Insurance Service-National Sample Cohort (NHIS-NSC), South Korea. Int J Epidemiol 2017;46:e15.
Kim YH, Kang JG, Lee SJ, et al, assignee. Underweight increases the risk of end-stage renal diseases for type 2 diabetes in Korean population: data from the National Health Insurance Service Health Checkups 2009-2017. Diabetes Care 2020;43:1118-1125.
Kim WJ, Lee SJ, Lee E, Lee EY, Han K, assignee. Risk of incident dementia according to glycemic status and comorbidities of hyperglycemia: a nationwide population-based cohort study. Diabetes Care 2022;45:134-141.
Wu X, Liang D, Sun J, Lin Y, Wu S, assignee. Association between sex-specific serum gamma-glutamyltransferase and incidence of hypertension in a Chinese population without metabolic syndrome: a prospective observational study. Front Cardiovasc Med 2021;8:644044.
Li DD, Xu T, Cheng XQ, et al, assignee. Serum gamma-glutamyltransferase levels are associated with cardiovascular risk factors in China: a nationwide population-based study. Sci Rep 2018;8:16533.
Tsuboya T, Kuriyama S, Nagai M, et al, assignee. Gamma-glutamyltransferase and cancer incidence: the Ohsaki cohort study. J Epidemiol 2012;22:144-150.
van Beek JH, de Moor MH, Geels LM, et al, assignee. The association of alcohol intake with γ-glutamyl transferase (GGT) levels: evidence for correlated genetic effects. Drug Alcohol Depend 2014;134:99-105.
Liao W, Yang Y, Yang H, Qu Y, Song H, Li Q, assignee. Circulating gamma-glutamyl transpeptidase and risk of pancreatic cancer: a prospective cohort study in the UK Biobank. Cancer Med 2023;12:7877-7887.
Yin X, Zheng SS, Zhang BH, et al, assignee. Elevation of serum γ-glutamyltransferase as a predictor of aggressive tumor behaviors and unfavorable prognosis in patients with intrahepatic cholangiocarcinoma: analysis of a large monocenter study. Eur J Gastroenterol Hepatol 2013;25:1408-1414.
Huxley R, Ansary-Moghaddam A, Berrington de González A, Barzi F, Woodward M, assignee. Type-II diabetes and pancreatic cancer: a meta-analysis of 36 studies. Br J Cancer 2005;92:2076-2083.
Ren HB, Yu T, Liu C, Li YQ, assignee. Diabetes mellitus and increased risk of biliary tract cancer: systematic review and meta-analysis. Cancer Causes Control 2011;22:837-847.
Gohel MG, Chacko AN, assignee. Serum GGT activity and hsCRP level in patients with type 2 diabetes mellitus with good and poor glycemic control: an evidence linking oxidative stress, inflammation and glycemic control. J Diabetes Metab Disord 2013;12:56.
Traverso N, Ricciarelli R, Nitti M, et al, assignee. Role of glutathione in cancer progression and chemoresistance. Oxid Med Cell Longev 2013;2013:972913.
Lee CH, Han K, Kim DH, Kwak MS, assignee. Repeatedly elevated γ-glutamyltransferase levels are associated with an increased incidence of digestive cancers: a population-based cohort study. World J Gastroenterol 2021;27:176-188.

Article

ahead

Gut and Liver

Published online December 6, 2024