Page 104 - JCTR-11-5

Journal of Clinical and
Translational Research Uric acid, CTGF genotype, and prostate cancer

Alcohol intake was calculated based on self-reported 2.5. Statistical analysis
usual monthly consumption of beer, wine (including Baseline characteristics were compared between subjects
Japanese saké [15% alcohol] and fortified wines [17–20% with and without hyperuricemia, and with different CTGF
alcohol]), and spirits (including whiskey, gin, brandy, or genotypes: common allele (C) homozygotes (genotype CC:
other liquor) among current drinkers. The factors used termed CTGF–CC) and minor allele (T) carriers (genotypes
to obtain estimates of alcohol content in all beverages CT or TT: termed CTGF–T), the latter having been found
consumed were 3.7% for beer, 10% for wine, and 38% for to be associated with longevity in our previous study.
27
spirits. Continuous variables were analyzed using Student’s t-test,
Hypertension was defined as systolic blood pressure while categorical variables were compared using the χ test.
2
(BP) ≥140 mm Hg, diastolic BP ≥90 mm Hg, or the self- Cox proportional hazard models were used to assess
reported use of antihypertensive medications. Normal BP the association of hyperuricemia and CTGF genotype
(normotension) was defined as systolic BP <140 mm Hg, with prostate cancer. The main effects, hazard ratio (HR),
diastolic BP <90 mm Hg, and not taking antihypertensive and 95% confidence intervals (95% CI) of hyperuricemia
medication. The percentage of calories from animal protein and CTGF genotype on prostate cancer incidence were
was calculated using a 24-h dietary recall by dividing the estimated using a multivariate Cox proportional hazard
calories from animal protein by the total calories. More model that included hyperuricemia and/or CTGF
detailed information can be found elsewhere. 26 genotype while adjusting for confounders. This model

2.3. Genotyping was referred to as the main effect model. The interaction
effect of hyperuricemia and CTGF genotype was tested
Among the 12 tagging single-nucleotide polymorphisms using a “full model,” which extended the main effect model
(SNPs) in CTGF that we tested in a previous case–control by including an interaction term between hyperuricemia
study for association with longevity, carriers of the minor and CTGF genotype. The goodness of fit between the full
(T) allele of rs9399005 had a significantly longer lifespan. model and the main effect model was compared using
27
Therefore, rs9399005 was chosen as the SNP of interest for likelihood ratio tests. Stratified analyses were conducted to
the present study. assess the association between hyperuricemia and prostate
Genotyping was performed using DNA extracted from cancer within each CTGF genotype. The Cox proportional
the buffy coat of blood samples collected at Kuakini-HHP hazard assumption was tested for the stratified Cox models.
Examination 4 (1991–1993), and the samples were kept at All tests were two-sided, and a p<0.05 was considered
–70°C. For participants who did not attend the Kuakini- statistically significant. All statistical analyses were
HHP Examination 4, we used serum samples available from performed using the Statistical Analysis System version 9.4
Examination 3. For the latter, DNA was amplified using a (Cary, NC, USA).
combination of QIAmp cell-free DNA isolation followed 3. Results
by REPLI-g Single-Cell WGA and WTA amplification
(QIAGEN Sciences, Germantown, MD, USA). Genotyping 3.1. Baseline characteristics
was performed using TaqMan on an Applied Biosystems From the 8,006 middle-aged men who participated in
QuantStudio 12K Flex system (ThermoFisher Scientific, Kuakini-HHP Examination 1, we excluded 82 men with
Waltham, MA, USA). any type of cancer at baseline, 35 men without uric acid
2.4. Ascertainment of prostate cancer measurements, 1,399 men without CTGF rs399005
genotype data, and 231 men who self-reported having a
All incident cancer cases diagnosed between 1965 and 1999 history of gout at baseline. As a result, our analytical sample
were captured by the Kuakini-JHCS surveillance program. included 6,259 subjects. Over a median follow-up period of
For cancer incidence among subjects who died before 29.7 years, 285 prostate cancer cases were identified from
or did not participate in the Kuakini-JHCS examination baseline to December 1999. Table 1 presents the baseline
(1971–1974), ascertainment was conducted retrospectively characteristics of subjects by hyperuricemia status and
according to the criteria of the Kuakini-JHCS surveillance CTGF rs399005 genotype. Subjects with hyperuricemia
program when the cancer surveillance program began. were younger, less physically active, had a higher BMI,
22
Prostate cancer incidence was determined by a physician had a higher prevalence of hypertension, smoked more
consensus group using hospital records, tumor registry cigarettes, drank more alcohol, and had a higher dietary
data, and confirmation through histological evidence. percentage of calories from animal protein intake.
The Kuakini-JHCS surveillance program concluded on However, no baseline variables were associated with the
December 31, 1999. CTGF genotype.

Volume 11 Issue 5 (2025) 98 doi: 10.36922/JCTR025260029