Clinical Chemistry 53:1
71–77 (2007)
␥-Glutamyltransferase as a Predictor of Chronic
Kidney Disease in Nonhypertensive and
Nondiabetic Korean Men
Seungho Ryu,1* Yoosoo Chang,2 Dong-Il Kim,1 Won Sool Kim,1 and Byung-Seong Suh1
Background: Little research has been done to examine
whether ␥-glutamyltransferase (GGT) is prospectively
associated with the development of chronic kidney
disease (CKD). We performed a prospective study to
examine the association between GGT and the risk for
the development of CKD.
Methods: The study cohort included a total of 10 337
healthy males with normal baseline kidney functions
and no proteinuria. Participants were workers in a
semiconductor manufacturing company and its 13 affil-
iates. CKD was defined as either the presence of pro-
teinuria or a glomerular filtration rate (GFR) of <60
mL 䡠 minⴚ1 䡠 (1.732)ⴚ1. Cox proportional hazards models
were used to calculate the adjusted hazard ratios in
separate models for CKD.
Results: During a follow-up period of 25 774.4 person-
years, 366 men developed CKD. After adjustments were
made for age, baseline GFR, triglyceride, and HDL-C,
the risk for CKD increased with an increasing quartile
of serum GGT (p for trend <0.001). The top one fourth
of serum GGT vs the bottom one fourth of relative risks
for CKD was 1.90 (95% confidence interval, 1.37–2.63).
These associations were also apparent in participants
who consumed <20 g/day of alcohol and those with
normal weight, with values of alanine aminotransfer-
ase within reference intervals, or with C-reactive pro-
tein <3.0 mg/L, and participants without metabolic
syndrome.
Conclusions: Our findings, which were obtained from a
large work-site cohort and excluded individuals with
diabetes and hypertension, indicated that serum GGT
1
Department of Occupational Medicine and 2 Health Screening Center,
Kangbuk Samsung Hospital, Sungkyunkwan University, School of Medicine.
Seoul, Korea.
*Address correspondence to this author at: Kangbuk Samsung Hospital,
108 Pyung dong, Jongro-Gu, Seoul, Korea 110-746. Fax 82-2-2001-2626; e-mail
sh703.yoo@samsung.com.
Received August 24, 2006; accepted November 1, 2006.
Previously published online at DOI: 10.1373/clinchem.2006.078980
71
Lipids, Lipoproteins,
and Cardiovascular
Risk Factors
may be an early predictor for the development of CKD,
independent of baseline confounding factors.
© 2007 American Association for Clinical Chemistry
The number of patients with end-stage renal disease
(ESRD)3 is increasing worldwide (1, 2 ). In 1999, the US
Renal Data System documented a persistent increase in
kidney failure that required dialysis therapy or transplan-
tation in 340 000 patients and a projected increase to
651 000 patients by 2010 (3 ). In Korea, there has been a
recent dramatic increase in the prevalence of ESRD pa-
tients requiring renal replacement therapy, from 303.6 per
million population in 1994 to 854.0 per million population
in 2004 (4 ).
Chronic kidney disease (CKD) often progresses to
ESRD with its attendant complications; and the treatment
of the earlier stages of CKD is effective in slowing the
progression toward ESRD (5, 6 ). Thus, the identification
of the precursors of CKD is very important. Few prospec-
tive studies, however, have provided data on the risk
factors for the development of CKD among the popula-
tion in Asia.
Serum ␥-glutamyltransferase (GGT) has been used
widely as an index of alcohol intake or liver dysfunction
(7–9 ). Serum GGT was proposed as a sensitive marker of
oxidative stress (10 ) because it has dose–response associ-
ations with many cardiovascular disease risk factors, as
well as future risk of diabetes; however, little research has
been done to examine whether GGT is associated with the
prospective development of CKD.
The goal of this study was to evaluate the association
between GGT and the risk for CKD in nonhypertensive
and nondiabetic male Korean workers.
3
Nonstandard abbreviations: ESRD, end-stage renal disease; CKD, chronic
kidney disease; GGT, ␥-glutamyltransferase; GFR, glomerular filtration rate;
ALT, alanine aminotransferase; AST, aspartate aminotransferase; HDL-C,
HDL-cholesterol; HOMA-IR, Homeostasis Model Assessment of insulin resis-
tance; CRP, C reactive protein; SBP, systolic blood pressure; RR, relative risk;
CI, confidence interval; BMI, body mass index; ROS, reactive oxygen species.
72 Ryu et al.: ␥-Glutamyltransferase and Chronic Kidney Disease
Materials and Methods
study population
A prospective cohort study was conducted to examine the
association between serum GGT and risk for CKD in
nonhypertensive and nondiabetic male Korean workers
who were employed at one of the biggest semiconductor
manufacturing companies in Korea and its 13 affiliates.
All employees participate in either annual or biennial
health examination, as is required by Korea’s Industrial
Safety and Health law. The study population included
male workers ⱖ40 years of age who underwent an annual
comprehensive health examination and workers 30 to
39 years of age who underwent a biennial comprehensive
health examination. A total of 15 347 workers, 30 to 59
years of age, participated in the comprehensive health
examinations at a university hospital in Seoul, Korea,
between January and December of 2002. Among the
15 347 workers, 3532 (23.0%) were excluded for various
reasons: 27 (0.7%) had a history of malignancy; 9 (0.3%)
had a history of liver cirrhosis; 16 (0.5%) had a history of
cardiovascular disease; 125 (3.5%) were receiving medical
treatment for dyslipidemia; 11 (0.3%) were receiving
medical treatment for hepatitis; 7 (0.2%) were receiving
medical treatment for current kidney disease; 246 (7.0%)
did not possess information about their past medical
histories; 91 (2.6%) did not have a urinalysis performed;
279 (7.9%) were taking medication for diabetes or had
fasting glucose concentrations ⱖ126 mg/dL; 2688 (76.1%)
were taking medication for hypertension or had
BPs ⱖ140/90 mmHg at their initial examinations; 260
(7.4%) had prevalent proteinuria; and 261 (7.4%) had
glomerular filtration rates (GFRs) of ⬍60 mL 䡠 min⫺1 䡠
(1.732)⫺1 at the time of their initial examinations. Because
some individuals had more than 1 exclusion criterion, the
total number eligible for the study was 11 815. The study
participants were reexamined at the same hospital annu-
ally, over a period of ⬃3.5 successive years, until May
2006. We excluded an additional 1478 individuals from
our cohort who did not participate in consecutive annual
or biennial health examinations during the follow-up
period. Ultimately, 10 337 male workers were enrolled in
the analysis and were observed for the development of
CKD, and their mean (SD) follow-up periods were 2.49
(0.86) years. This study was approved by the Institutional
Review Board at Kangbuk Samsung Hospital.
measurements
The initial health examinations that were performed in
2002 included a medical history, a physical examination, a
questionnaire about health-related behavior, anthropo-
metric measurements, and biochemical measurements.
The medical history and history of prescription drug use
were assessed by the examining physicians. All the par-
ticipants were asked to respond to a questionnaire on
health-related behavior. Questions about alcohol intake
included the frequency of alcohol consumption on a
weekly basis and the usual amount that was consumed on
a daily basis. We considered persons reporting that they
smoked to be current smokers. In addition, the partici-
pants were asked about their weekly frequency of phys-
ical activity, such as jogging, bicycling, and swimming
that lasted long enough to produce perspiration.
Fasting blood samples were drawn from an ante-
cubital vein from participants after they had fasted ⬎12 h.
The fasting serum glucose, total cholesterol, triglycerides,
LDL cholesterol, HDL-cholesterol (HDL-C), uric acid,
blood urea nitrogen, creatinine, ␥-glutamyltransferase
(GGT), alanine aminotransferase (ALT), and aspartate
aminotransferase (AST) concentrations were measured
enzymatically with an automatic analyzer (Advia 1650
AutoAnalyzer, Bayer Diagnostics). The fasting serum
glucose was measured with the hexokinase method.
Total cholesterol and serum triglycerides were measured
with enzymatic colorimetric tests, low-density lipopro-
tein-cholesterol was measured with the homogeneous
enzymatic colorimetric test, and HDL-C was measured
with the selective inhibition method (Bayer Diagnostics).
Insulin concentrations were measured with immunora-
diometric assays (Biosource), with intra- and interassay
CVs of 4.7% to 12.2%. We performed homeostasis model
assessments of insulin resistance (HOMA-IR). High sen-
sitivity-C reactive protein (CRP) was analyzed by per-
forming particle-enhanced immunonephelomety with the
BNTM System (Dade Behring). The result was presented as
milligrams per liter, and the minimum detectable CRP
concentration was 0.175 mg/L after performing 1:20 sam-
ple dilution. The serum creatinine was measured with the
alkaline picrate (Jaffe) method. The coefficients of varia-
tion for the creatinine determinations were ⱖ3% from
2002 to 2005. The Korean Association of Quality Assur-
ance for Clinical Laboratories assessed the quality control
of the laboratory, both internally and externally, on a
regular basis. The urine protein concentration was deter-
mined at each examination from the results of a single
urine dipstick semiquantitative analysis (URiSCAN®
Urine strip, YD Diagnostics). Dipstick urinalysis was
performed on fresh, midstream urine samples that were
collected in the morning. The results of the urine test were
based on a scale that quantified proteinuria as absent,
trace, 1⫹, 2⫹, 3⫹, and 4⫹. The dipstick results of 1⫹, 2⫹,
3⫹, and 4⫹ corresponded to protein concentrations of
30, 100, 300, and 1000 mg/dL, respectively. Kidney func-
tion was estimated by the GFR, which was calculated
with the simplified Modification of Diet in Renal Disease
Study equation that is defined as GFR ⫽ 186.3 ⫻ (serum
creatinine)⫺1.154 ⫻ age⫺0.203 (17, 18 ). Proteinuria was de-
fined as a finding of 1⫹ or greater. CKD was defined as
either proteinuria or a GFR ⬍60 mL 䡠 min⫺1 䡠 (1.732)⫺1.
Trained nurses measured sitting blood pressure with a
standard mercury sphygmomanometer. The 1st and 5th
Korotkoff sounds were used to estimate the systolic blood
pressure (SBP) and the diastolic blood pressure. Height
and weight were measured after an overnight fast with
study participants wearing a lightweight hospital gown
 Clinical Chemistry 53, No. 1, 2007 73

and no shoes. The body mass index (BMI) was calculated
as the patient’s weight (in kilograms) divided by the
square of the patient’s height (in meters).
statistical analysis
The ␹2-test and the one-way ANOVA were used to
analyze the statistical differences among the characteris-
tics of the study participants at the time of enrollment in
relation to the serum GGT concentrations. Categories of
serum GGT comprised the following quartiles: ⬍18, 19 –
25, 26 –39, and ⱖ40. The incidence density was expressed
as the number of cases divided by the person-years from
the baseline until the development of CKD, with the
assumption of a date of diagnosis in the middle of the
follow-up period or until the final physical examination.
The incidence densities were compared by calculating
the incidence density ratios with the 95% confidence
interval (CI). We used the Cox proportional hazards
model to calculate the adjusted hazard ratios in the model
for CKD. The data were first adjusted for age alone, then
for the multiple covariates. In the multivariate models,
we included variables, which include age, baseline GFR,
BMI, fasting glucose, SBP, total cholesterol, uric acid,
HOMA-IR, CRP, smoking, alcohol consumption, incident
diabetes, and incident hypertension, that might confound
the relationship between the serum GGT and CKD. For
the linear trends of risk, the number of quartiles was used
as a continuous variable and tested on each model. The
data were analyzed and the statistical analysis for the data
was performed with SPSS version 12.0 software (SPSS
Inc.). All the reported P values were 2-tailed, and those
⬍0.05 were considered to be statistically significant.
Results
At baseline, the 10 337 study participants had a mean (SD)
age of 36.9 (4.8) years, 46.9% were current smokers, and
24.4% exercised regularly ⱖ1 time per week. The overall
prevalence of overweight or obesity (BMI ⱖ25 kg/m2)
was 34.0%. The participants who did not have follow-up
measurements were, on average, 0.5 years older (P ⫽
0.003) and had a fasting serum glucose ⬃0.03 mmol/L
lower (P ⫽ 0.033) than those who were included in the
analytic cohort; but there were no differences in the
prevalence of metabolic syndrome and obesity, lipid
profiles, uric acid concentrations, SBP, or baseline GFR.
The characteristics of the study participants in relation
to the serum GGT concentrations are illustrated in
Table 1. For all of the listed variables there were clear
dose–response relationships with serum GGT concentra-
tions. Age, BMI, fasting serum glucose, systolic and
diastolic blood pressures, total cholesterol, triglycerides,
LDL-C, uric acid, creatinine, HOMA-IR, CRP, current
smoking, and current alcohol drinking were associated
positively, whereas HDL-C, GFR, and exercise were asso-
ciated inversely. The percentage of those who had meta-
bolic syndrome also increased in correlation with an
increase in the serum GGT.
During 25 774.4 person-years of follow-up, 366 new
incident cases of CKD developed (Table 2). After adjust-
ments were made for age, baseline GFR, triglyceride, and

Table 1. Baseline characteristics of the study participants by GGT concentration.

GGT (U/L)

I (<18) II (19–25) III (26–39) IV (>40) P for trend

Number 2934 2389 2526 2488
Age, years 36.1 (4.7) 36.7 (4.8) 37.2 (4.9) 37.8 (4.8) ⬍0.001
BMI, kg/m2 22.4 (2.4) 23.5 (2.5) 24.5 (2.5) 25.4 (2.6) ⬍0.001
Fasting serum glucose, mmol/L 4.93 (0.46) 5.00 (0.48) 5.06 (0.51) 5.19 (0.53) ⬍0.001
Systolic BP, mmHg 110.6 (9.5) 111.8 (9.4) 112.7 (9.2) 113.8 (9.2) ⬍0.001
Diastolic BP, mmHg 71.0 (7.4) 72.0 (7.3) 72.6 (7.3) 73.6 (7.2) ⬍0.001
Total cholesterol, mmol/L 4.83 (0.80) 5.14 (0.82) 5.35 (0.85) 5.57 (0.94) ⬍0.001
HDL-C, mmol/L 1.38 (0.30) 1.35 (0.30) 1.32 (0.29) 1.31 (0.29) ⬍0.001
LDL-C, mmol/L 2.88 (0.69) 3.10 (0.71) 3.20 (0.75) 3.29 (0.82) ⬍0.001
Triglyceride, mmol/L 1.08 (0.85–1.46) 1.34 (0.99–1.78) 1.58 (1.16–2.17) 1.92 (1.41–2.67) ⬍0.001
Creatinine, ␮mol/L 98.8 (9.3) 99.5 (9.5) 99.7 (9.5) 100.3 (9.7) ⬍0.001
Uric acid, ␮mol/L 337.8 (61.9) 355.1 (61.9) 366.4 (68.4) 378.9 (72.6) ⬍0.001
GFR, mL 䡠 min⫺1 䡠 (1.732)⫺1 80.1 (73.7–85.2) 78.7 (72.8–83.2) 78.5 (72.2–83.1) 77.9 (71.5–82.7) ⬍0.001
Insulin, pmol/L 42.3 (34.2–53.3) 47.2 (37.2–63.3) 53.0 (40.8–67.4) 60.2 (45.2–72.4) ⬍0.001
HOMA-IR 1.33 (1.06–1.73) 1.52 (1.17–2.01) 1.70 (1.29–2.22) 1.99 (1.47–2.46) ⬍0.001
CRP, mg/L 0.3 (0.2–0.7) 0.5 (0.2–0.9) 0.6 (0.3–1.1) 0.7 (0.4–1.4) ⬍0.001
Current smoker, % 39.5 44.3 49.3 55.5 ⬍0.001
Current drinker, %a 28.8 37.5 46.5 58.4 ⬍0.001
Regular exerciser, %a 26.3 25.8 24.8 20.5 ⬍0.001
Metabolic syndrome, % 1.9 3.8 9.6 15.2 ⬍0.001
Data are mean (SD), median (interquartile range), or percent.
a
ⱖ1 time per week
74 Ryu et al.: ␥-Glutamyltransferase and Chronic Kidney Disease
HDL-C, GGT was significantly associated with the risk for
CKD [adjusted relative risk (RR) 1.13 (95% CI, 1.06 –1.20)
per 1-SD increase in the natural log of GGT]. In the
categorical analyses, the risk for CKD increased with
increasing quartiles of serum GGT (P for trend ⬍0.001),
but the associations across quartiles of GGT seemed to be
nonlinear. For study participants in the 4th quartile, CKD
risk was significantly increased [adjusted RR, 1.90 (95%
CI, 1.37–2.63)]. These results did not change after further
adjustments for obesity, baseline HOMA-IR, or CRP.
To explore whether the risk for CKD with a serum
GGT was mediated by the subsequent development of
hypertension or diabetes, we fit additional models by
adjusting for incident hypertension or diabetes at the time
of follow-up. During 3 years of follow-up, 145 (1.4% of the
cohort) developed incident diabetes and 1464 (14.2%)
developed hypertension. After the inclusion of incident
hypertension or diabetes, the risk for CKD still increased
with increasing quartiles of serum GGT (for trend
⬍0.001).
These associations were apparent in study participants
who were alcohol drinkers of ⱕ20 g/day or were non-
overweight (BMI ⬍25 kg/m2) or even lean (BMI ⬍23
kg/m2), in participants with ALT concentrations within
the reference interval or CRP ⬍3.0 mg/L, and in those
without the metabolic syndrome (Table 3).
In the unadjusted analyses, the quartiles of both ALT
and AST were associated with a significantly increased
risk for CKD (P for trend ⬍0.001 and ⬍0.001, respec-
tively). After adjustments were made for potential con-
founders, however, neither ALT nor AST was related
significantly to the incidence of CKD.
Discussion
The results of this study demonstrated that the risk for
CKD increased in correlation with an increase in the
serum GGT in nonhypertensive and nondiabetic men.
Other liver enzymes, such as ALT and AST, were not
related significantly to the incidence of CKD. To the best
of our knowledge, little research has been done to exam-
ine whether serum GGT is associated with the prospective
development of CKD, so we were unable to compare our
results with those of other studies.
The mechanisms by which the serum GGT increases
the risk for CKD are not fully understood. In clinical
practice, an increased concentration of serum GGT is
conventionally interpreted as a marker of alcohol abuse or
liver disease (7–9 ). In this study, the serum GGT pre-
dicted the incidence of CKD independently of the amount
of alcohol consumed, as well as in drinkers of ⱕ20 g/day.
The dose–response relationship between the serum GGT
and the incidence of CKD was observed among individ-
uals with ALT concentrations within the reference inter-
val. Furthermore, neither ALT nor AST was related sig-
nificantly to the incidence of CKD. Therefore, excessive
alcohol consumption or liver diseases do not explain our
results. Obesity could play a role in the association of
serum GGT with the incidence of CKD, because obese
adults have an increased risk for developing CKD (19 ). In
this study, however, the serum GGT was predictive of
CKD even among lean men with a BMI ⬍23 kg/m2.
Insulin resistance syndrome may explain the serum GGT–
incident CKD relationship, because these conditions have
been associated strongly with serum GGT (20 ), and
several studies reported that insulin resistance was asso-
ciated with an increased risk for CKD (21, 22 ). In this
study, however, the risk for CKD increased with increas-
ing quartiles of serum GGT independently of HOMA-IR,
as well as in men without the metabolic syndrome.
Systemic low-grade inflammation as assessed by CRP is
another possible mechanism linking GGT with CKD de-
velopment. A previous study showed that low-grade
inflammation may be a predictor for a change in kidney
function in the elderly (23 ), but our results, derived from
apparently healthy persons, showed that CRP was not
independently associated with incident CKD.
Recently, serum GGT has been proposed as a sensitive
and reliable marker of oxidative stress (10 ). In the Coro-
nary Artery Risk Development in Young Adults study,
circulating concentrations of serum and dietary antioxi-
dant vitamins, such as ␤-carotene, ␣-carotene, ␤-cryptox-
anthin, zeaxanthin/lutein, and ␣-tocopherol, were related
inversely in a dose–response manner to the serum GGT
concentration within its reference interval (24 ). The same
study also showed that serum GGT concentration pre-
dicted fibrinogen and C-reactive protein, which are mark-
ers of inflammation (25 ). These associations suggested
that serum GGT may be a biological marker of oxidative
stress. In addition, prospective cohort studies have shown
that serum GGT predicted the development and out-
comes of many diseases (10 –16 ). On the basis of the
findings of experimental and human studies, we attribute
the association of serum GGT with CKD incidence in our
study to a mechanism related to oxidative stress. Several
experimental studies have demonstrated that O2⫺1 caused
direct vasoconstriction of the renal cortical and medullary
vessels and increased the intracellular calcium in the
vascular smooth muscle and endothelial cells (26 –28 ).
These studies suggested that renal reactive oxygen species
(ROS) caused renal vasoconstriction and sodium reten-
tion, and renal damage. Other studies reported that rats
receiving a high sodium diet that caused increased arte-
riolar and venular O2⫺1 production exhibited severe renal
damage, decreased GFR and renal plasma flow, and high
renal superoxide production (29 –30 ). Furthermore, treat-
ment with vitamin C and E decreased the production of
renal superoxide and renal damage, and prevented the
decrease in renal hemodynamics (31 ). Recent human
studies indicated that individuals with essential hyperten-
sion have decreased antioxidant capacity and produce
excessive amounts of ROS, and that more than half of
these hypertensive patients were salt sensitive and had
progressive renal damage (32–34 ). These experimental
and human studies provided evidence to support the
 Table 2. Adjusted RR of incidence of CKD with explanatory factors.
Multivariate-adjusted RR (95% CI)

Age-adjusted RR (95% CI) Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

Age per 1-year increment 1.08 (1.06–1.10) 1.08 (1.06–1.10) 1.07 (1.05–1.09) 1.07 (1.05–1.09) 1.07 (1.05–1.09) 1.06 (1.04–1.09)
GGT quartiles
I (⬍18) 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference)
II (19–25) 1.42 (1.00–2.02) 1.20 (0.84–1.72) 1.17 (0.82–1.68) 1.16 (0.81–1.66) 1.15 (0.80–1.64) 1.27 (0.87–1.87)
III (26–39) 1.65 (1.18–2.30) 1.28 (0.91–1.80) 1.22 (0.87–1.73) 1.20 (0.84–1.69) 1.19 (0.84–1.69) 1.34 (0.93–1.95)
IV (ⱖ40) 2.66 (1.95–3.63) 1.90 (1.37–2.63) 1.77 (1.27–2.45) 1.71 (1.22–2.39) 1.71 (1.22–2.39) 1.87 (1.31–2.67)
P for trend ⬍0.001 ⬍0.001 ⬍0.001 0.001 0.001 ⬍0.001
GGT per 1 SD increment 1.21 (1.14–1.28) 1.13 (1.06–1.20)
GFR per 1 SD increment 0.40 (0.35–0.47) 0.42 (0.37–0.49) 0.42 (0.37–0.49) 0.42 (0.36–0.48) 0.42 (0.37–0.49) 0.41 (0.35–0.47) 0.42 (0.36–0.49)
Obesityb 1.86 (1.51–2.29) 1.11 (0.88–1.38)
Impaired fasting glucosec 1.46 (0.84–2.55)
Prehypertensiond 1.15 (0.81–1.64)
Low HDLe 2.16 (1.69–2.75) 1.86 (1.45–2.39) 1.84 (1.44–2.37) 1.91 (1.49–2.45) 1.89 (1.47–2.43) 1.85 (1.44–2.39) 1.93 (1.48–2.52)
High TGf 2.56 (2.07–3.16) 1.64 (1.30–2.07) 1.77 (1.41–2.22) 1.61 (1.27–2.03) 1.59 (1.26–2.01) 1.57 (1.23–1.99) 1.51 (1.18–1.93)
TC per 1 SD increment 1.21 (1.10–1.33)
LDL-C per 1 SD increment 0.97 (0.88–1.08)
HOMA-IR per 1 SD increment 1.26 (1.17–1.36) 1.05 (0.95–1.16)
Clinical Chemistry 53, No. 1, 2007

CRP ⱕ3.0 mg/L 1.38 (0.95–1.99) 1.23 (0.85–1.78)

Uric acid, 1 SD 1.37 (1.25–1.51)
Current smoker 1.03 (0.83–1.26)
Ethanol, 10–20 g per day 0.97 (0.75–1.26)
Ethanol, ⬎20 g per day 1.26 (0.97–1.66)
Incident hypertension 2.03 (1.62–2.55) 1.85 (1.47–2.33) 1.83 (1.45–2.31) 1.81 (1.44–2.28) 1.93 (1.52–2.46)
Incident diabetes 1.40 (0.76–2.56)
Multivariate models are adjusted for all other variables listed for the model.
a
TC, total cholesterol; LDL-C, low-density lipoprotein cholesterol; TG, triglyceride; b BMI ⱖ25 kg/m2; c fasting serum glucose ⱖ6.1 mmol/L; d blood pressure ⱖ130/85 mmHg; e HDL-C ⬍1.036 mmol/L; f triglyceride
ⱖ1.695 mmol/L.
75
76 Ryu et al.: ␥-Glutamyltransferase and Chronic Kidney Disease
Table 3. Adjusted RR of incidence of CKD by quartile of GGT concentrations according to subgroups of alcohol
consumption <20 g ethanol per day, BMI <25 kg/m2 or <23 kg/m2, ALT <35 U/L, CRP <3.0 mg/L,
and nonmetabolic syndrome
I (<18)
Ethanol per day ⬍20 g (n ⫽ 8742) 1.00
BMI ⬍25 kg/m2 (n ⫽ 6818) 1.00
BMI ⬍23 kg/m2 (n ⫽ 3776) 1.00
ALT ⬍35 U/L (n ⫽ 7299) 1.00
CRP ⬍3.0 mg/L (n ⫽ 9121) 1.00
Nonmetabolic syndrome (n ⫽ 9572) 1.00
Adjustment for age, baseline GFR, incident hypertension, loge triglyceride, and HDL-C.
hypothesis that oxidative stress plays an important role in
renal damage, but further interventional studies that
target serum GGT are necessary to clarify the mechanisms
of renal damage.
Our study had several limitations. First, we used an
estimated GFR instead of a directly measured GFR to
define CKD. A recent review article (35 ) reported that
current GFR estimates had greater inaccuracy in popula-
tions without known CKD than in those with the disease.
Nonetheless, current GFR estimates facilitate the detec-
tion, evaluation, and management of CKD, and many
organizations recommend the use of equations that esti-
mate GFR for the evaluation of renal function in epide-
miologic studies and in clinical practice (35 ). Second, the
dipstick urinalysis has imperfect sensitivity and specific-
ity, but the number of false-positive results can be higher
due to menstruation or comorbid illnesses in women and
older persons, circumstances that did not affect our study
population of healthy men, ages 30 to 59 years. In addi-
tion, the National Kidney Foundation Kidney Disease
Outcome Quality Initiation Advisory Board recom-
mended that under most circumstances spot urine sam-
ples rather than timed urine collection (overnight or 24-h)
can be used for detection and monitoring proteinuria in
adults, (18 ). Third, our findings may be biased by the
healthy worker effect. However, although those partici-
pants who did not undergo follow-up measurements
were, on average, 0.5 years older than those who were
included in the analytic cohort, there were no differences
between the 2 groups in the prevalence of metabolic
syndrome or in obesity, lipid profiles, uric acid, SBP, or
GFR concentrations at the baseline. Thus, the exclusion of
participants without follow-up measurements would not
be expected to bias our results. Finally, ethnic factors that
are characteristic for the Asian population are not well
established with respect to using equations that estimate
GFR. Therefore, these equations need to be validated in
additional studies with large Asian cohorts.
In conclusion, our findings, which were obtained from a
large work-site cohort that excluded individuals with
diabetes and hypertension, indicated that the serum GGT
Multivariate-adjusted RR (95% CI), GGT (U/L)
II (19–25) III (26–39) IV (>40) P for trend
1.02 (0.70–1.49) 1.08 (0.75–1.55) 1.45 (1.02–2.06) 0.025
1.20 (0.78–1.86) 1.28 (0.82–1.99) 1.76 (1.14–2.73) 0.012
1.28 (0.69–2.36) 1.28 (0.65–2.54) 2.17 (1.12–4.19) 0.033
1.04 (0.71–1.53) 1.13 (0.76–1.67) 1.75 (1.18–2.60) 0.007
1.15 (0.77–1.71) 1.15 (0.78–1.69) 1.62 (1.12–2.35) 0.006
1.15 (0.79–1.67) 1.08 (0.74–1.57) 1.63 (1.14–2.32) 0.006
may be an early predictor for the development of CKD,
independent of baseline confounding factors and the
subsequent development of hypertension. In addition,
our results supported previous studies that reported
oxidative stress played an important role in the produc-
tion of renal damage.
References
1. US Renal Data System. USRDS 2003 Annual Data Report. The
National Institutes of Health, National Institutes of Diabetes and
Digestive and Kidney Diseases. Bethesda, MD; 2003.
2. Stengel B, Billon S, Van Dijk PC, Jager KJ, Dekker FW, Simpson K,
et al. Trends in the incidence of renal replacement therapy for
end-stage renal disease in Europe, 1990 –1999. Nephrol Dial
Transplant 2003;18:1824 –33.
3. Xue JL, Ma JZ, Louis TA, Collins AJ. Forecast of the number of
patients with end-stage renal disease in the United States to the
year 2010. J Am Soc Nephrol 2001;12:2753– 8.
4. Korean Society of Nephrology, Registry Committee. Renal replace-
ment therapy in Korea: Insan Memorial Dialysis Registry 2004.
Korean J Nephrol 2004;24:S367–S83 (Korean).
5. Iseki K, Ikemiya Y, Fukiyama K. Risk factors of end-stage renal
disease and serum creatinine in a community-based mass screen-
ing. Kidney Int 1997;51:850 – 4.
6. Hsu CY, Chertow GM. Chronic renal confusion: insufficiency,
failure, dysfunction, or disease. Am J Kidney Dis 2000;36:415– 8.
7. Skinner HA, Holt S, Schuller R, Roy J, Israel Y. Identification of
alcohol abuse using laboratory tests and a history of trauma. Ann
Intern Med 1984;101:847–51.
8. Conigrave KM, Saunders JB, Reznik RB, Whitfield JB. Prediction of
alcohol-related harm by laboratory test results. Clin Chem 1993;
39:2266 –70.
9. Conigrave KM, Degenhardt LJ, Whitfield JB, Saunders JB, He-
lander A, Tabakoff B; WHO/ISBRA Study Group. CDT, GGT, and
AST as markers of alcohol use: the WHO/ISBRA collaborative
project. Alcohol Clin Exp Res 2002;26:332–9.
10. Lee DH, Blomhoff R, Jacobs DR Jr. Is serum gamma glutamyl-
transferase a marker of oxidative stress? Free Radic Res 2004;
38:535–9.
11. Wannamethee G, Ebrahim S, Shaper AG. Gamma-glutamyltrans-
ferase: determinants and association with mortality from ischemic
heart disease and all causes. Am J Epidemiol 1995;142:699 –
708.
12. Lee DH, Ha MH, Kim JH, Christiani DC, Gross MD, Steffes M, et al.
Gamma-glutamyltransferase and diabetes: a 4-year follow-up
study. Diabetologia 2003;46:359 – 64.
 Clinical Chemistry 53, No. 1, 2007
13. Brenner H, Rothenbacher D, Arndt V, Schuberth S, Fraisse E,
Fliedner TM. Distribution, determinants, and prognostic value of
gamma-glutamyltransferase for all-cause mortality in a cohort of
construction workers from southern Germany. Prev Med 1997;26:
305–10.
14. Miura K, Nakagawa H, Nakamura H, Tabata M, Nagase H, Yoshida
M, et al. Serum gamma-glutamyl transferase level in predicting
hypertension among male drinkers. J Hum Hypertens 1994;8:
445–9.
15. Jousilahti P, Rastenyte D, Tuomilehto J. Serum gamma-glutamyl
transferase, self-reported alcohol drinking, and the risk of stroke.
Stroke 2000;31:1851–5.
16. Nakanishi N, Suzuki K, Tatara K. Serum gamma-glutamyltrans-
ferase and risk of metabolic syndrome and type 2 diabetes in
middle-aged Japanese men. Diabetes Care 2004;27:1427–32.
17. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more
accurate method to estimate glomerular filtration rate from serum
creatinine: a new prediction equation. Modification of Diet in
Renal Disease Study Group. Ann Intern Med 1999;130:461–70.
18. National Kidney Foundation. K/DOQI clinical practice guidelines
for chronic kidney disease: evaluation, classification, and stratifi-
cation. Am J Kidney Dis 2002;39:S1–266.
19. Gelber RP, Kurth T, Kausz AT, Manson JE, Buring JE, Levey AS, et
al. Association between body mass index and CKD in apparently
healthy men. Am J Kidney Dis 2005;46:871– 80.
20. Lonardo A, Lombardini S, Scaglioni F, Carulli L, Ricchi M, Ganazzi
D, et al. Hepatic steatosis and insulin resistance: does etiology
make a difference? J Hepatol 2006;44:190 – 6.
21. Chen J, Muntner P, Hamm LL, Fonseca V, Batuman V, Whelton PK,
et al. Insulin resistance and risk of chronic kidney disease in
nondiabetic US adults. J Am Soc Nephrol 2003;14:469 –77.
22. Chen J, Muntner P, Hamm LL, Jones DW, Batuman V, Fonseca V,
et al. The metabolic syndrome and chronic kidney disease in U.S.
adults. Ann Intern Med 2004;140:167–74.
23. Fried L, Solomon C, Shlipak M, Seliger S, Stehman-Breen C,
Bleyer AJ, et al. Inflammatory and prothrombotic markers and the
progression of renal disease in elderly individuals. J Am Soc
Nephrol 2004;15:3184 –91.
77
24. Lee DH, Gross MD, Jacobs DR Jr. Cardiovascular Risk Develop-
ment in Young Adults Study. Association of serum carotenoids
and tocopherols with gamma-glutamyltransferase: the Cardiovas-
cular Risk Development in Young Adults (CARDIA) Study. Clin
Chem 2004;50:582– 8.
25. Lee DH, Jacobs DR Jr, Gross M, Kiefe CI, Roseman J, Lewis CE, et
al. Gamma-glutamyltransferase is a predictor of incident diabetes
and hypertension: the Coronary Artery Risk Development in Young
Adults (CARDIA) Study. Clin Chem 2003;49:1358 – 66.
26. Lenda DM, Sauls BA, Boegehold MA. TP receptor-mediated vaso-
constriction in microperfused afferent arterioles: roles of O(2)(⫺)
and NO. Am J Physiol Renal Physiol 2000;279:F302– 8.
27. Zou AP, Li N, Cowley AW Jr. Production and actions of superoxide
in the renal medulla. Hypertension 2001;37:547–53.
28. Lounsbury KM, Hu Q, Ziegelstein RC. Calcium signaling and
oxidant stress in the vasculature. Free Radic Biol Med 2000;28:
1362–9.
29. Lenda DM, Sauls BA, Boegehold MA. Reactive oxygen species
may contribute to reduced endothelium-dependent dilation in rats
fed high salt. Am J Physiol Heart Circ Physiol 2000;279:H7–H14.
30. Meng S, Roberts LJ 2nd, Cason GW, Curry TS, Manning RD Jr.
Superoxide dismutase and oxidative stress in Dahl salt-sensitive
and -resistant rats. Am J Physiol Regul Integr Comp Physiol
2002;283:R732– 8.
31. Tian N, Thrasher KD, Gundy PD, Hughson MD, Manning RD Jr.
Antioxidant treatment prevents renal damage and dysfunction and
reduces arterial pressure in salt-sensitive hypertension. Hyperten-
sion 2005;45:934 –9.
32. Russo C, Olivieri O, Girelli D, Faccini G, Zenari ML, Lombardi S, et
al. Anti-oxidant status and lipid peroxidation in patients with
essential hypertension. J Hypertens 1998;16:1267–71.
33. Lacy F, O’Connor DT, Schmid-Schonbein GW. Plasma hydrogen
peroxide production in hypertensives and normotensive subjects
at genetic risk of hypertension. J Hypertens 1998;16:291–303.
34. Lea JP, Nicholas SB. Diabetes mellitus and hypertension: key risk
factors for kidney disease. J Natl Med Assoc 2002;94:7S–15S.
35. Stevens LA, Coresh J, Greene T, Levey AS. Assessing kidney
function: measured and estimated glomerular filtration rate.
N Engl J Med 2006;354:2473– 83.