Angiotensinogen Gene Variation and Hypertension in a Cohort Study in Japanese

Many recent case-control studies have suggested a significant relationship between M235T (the substitution of threonine for methionine at position 235 codon) polymorphism of the angiotensinogen (AGT) gene and hypertension. To investigate whether the M235T polymorphism of AGT gene affects the incidence of hypertension, a retrospective cohort study was performed among Japanese workers. The subjects were Japanese workers at an occupational site in Shimane Prefecture in Japan. The baseline data were set at the received regular health examination in 1992, and a retrospective cohort study was performed for analyzing the incidence of hypertension in 1998. The rates of M235M (MM), M235T (MT) and T235T (TT) genotypes were 4%, 32% and 64%, respectively. The relative risks of MT and TT against MM for the incidence of hypertension by single variance analysis were 1.47 [95% confidence interval (CI) 0.50 4.33] and 1.35 (95% CI 0.47 3.90), respectively. The relative risks of MT and TT against MM for the incidence of hypertension, adjusted for sex, age, body mass index, fasting glucose and cigarette smoking, drinking and exercise in 1992, were 1.49 (95% Cl 0.49 4.53) and 1.25 (95% Cl 0.42 3.74), respectively. The data from this study suggest that the M235T polymorphism of AGT gene has a weak role in the manifestation of hypertension. Further comprehensive studies are needed to resolve this issue. J Epidemiol, 2001 ; 11 : 115-119


INTRODUCTION
Essential hypertension is a multiple factorial disease that is complicated by the interaction of genetic factors and environmental factors 1).The angiotensinogen (AGT) gene variant, M235T (mutation to the threonine of the methionine in amino acid codon 235), was reported in the pathogenesis of essential hypertension by Jeunemaitre et al. 2) for the first time in 1992.
However there have been subsequently many studies investigating the hypothesis that this M235T is one of the genetic factors of essential hypertension; the hypothesis remains controversial 3-7).
Most of the reports regarding the relationship between M235T and hypertension used the method of a case -control study method that included a relatively small sample size.If the study design is not carefully conducted, a case -control study has many kinds of bias.It is very difficult to choose a suitable control -group.Thus, we carried out a retrospective cohort study with a large sample size to clarify the involvement of M235T in hypertension.

Subjects;
The research was approved by the ethics committee of the Faculty of Medicine, Tottori University and informed consent was obtained from 2042 workers in Shimane Prefecture in Japan who received their regular health examination in 1998.Excluding people who had hypertension in 1992, 1001 workers who had received the regular health examination in both 1992 and 1998 were included in this study.A retrospective cohort study for the incidence of hypertension was

RESULTS
The characteristics of subjects by sex in 1992 were shown in Table 1.There were statistically significant differences in all mean values between male and female.The proportions of the genotype found in this retrospective cohort study for MM, MT and TT were 4%, 32% and 64%, respectively (Table 2).The allele frequency of 235T was 0.801.There was no significant difference between the normotensive and hypertensive subjects in terms of the proportions of the genotypes in 1998.As shown in Table 3, MT showed a slightly lower trend of mean age, but it was not significant.No significant differences were found among the other indices.Table 4 shows the distribution of BMI, SBP, DBP and fasting glucose related to genotypes in 1998.The rates of BMI value of 26.0 or more in MM, MT and TT were 16.3% 18.9% and 16.0%, respectively.There was no distribution difference in terms of genotypes in regard to the BMI value, SBP, DBP or fasting glucose.
In regard to the relative risk (RR) for hypertension, the RRs   of MT and TT against MM by univariate analysis were 1.47 and 1.35, respectively (Table 5, Model 1).These RRs showed a slightly higher trend, but it was not significant.In both cases that sex and age were adjusted (Table 5, Model 2) and in sex, age and BMI adjusted analysis (Table 5, Model 3), RRs of MT against MM were slightly higher in comparison with Model 1, but these were not significant.When the multiple risk factors of sex, age, BMI, fasting glucose, cigarette smoking, drinking habit and exercise were adjusted, RRs of MT and TT against MM were 1.49, 1.25 respectively.Although these were not significant, a weak positive relationship was found between polymorphism and hypertension.
DISCUSSION In the population examined in this study, the frequency of the wild type MM was the lowest, while that of the mutant type TT was the highest.There were no significant differences in the BMI, SBP, DBP and fasting glucose between the genotypes.The RRs of MT and TT against MM for hypertension analyzed by correction using various factors were not significant, but there were weak correlations (1.49 and 1.25, respectively).
The allele frequency in the population was higher than that of Caucasians 2).This agreed with the results reported in many studies on investigation of the allele frequency in Japanese subjects 9-13).The allele frequency was 0.801 in the present study, and ranged from 0.604 to 0.860 in published studies.This wide range of the allele frequency is probably because there are regional differences in the genetic background in Japan.
The relationship between this genetic polymorphism and hypertension has been suggested in studies performed in Japan and other countries1-7).Kato et al. were the first to cast doubt on the relationship14).The different results for the relationship between genotypes and hypertension may be caused by the possibility of bias in choosing controls, differences between races, regional differences in the genetic background, differences in the age of subjects and multifactors of the examined disease, because most of the studies were hospital-based casecontrol studies in a small sample size 15).Recently, Kato et al.
performed a case-control study in a large sample size using the four renin-angiotensin gene including M235T, and they concluded that there was no significant evidence for disease association 16) To improve on the drawbacks of case-control studies in a small sample size, we performed a cohort study in a large sample size and obtained weak correlations between genotypes and hypertension.This study was a retrospective cohort study , in which bias could be corrected to some extent , but the adequacy of the results may be low in this study compared to a prospective cohort study.Analyses were performed using 4 models , and higher RRs of MT against MM were shown by the multivariate analyses than by the univariate analysis .It was considered to be necessary to evaluate which factors should be corrected.It was also considered necessary to investigate publication bias 14).
In conclusion, there was a weak correlation between the M235T polymorphism of the AGT gene and hypertension in a retrospective cohort study corrected for various factors in Japanese subjects performed between 1992 and 1998 .We consider it necessary to perform further investigations while evaluating the factors to be corrected and the possibility of prospective cohort studies.

ACKNOWLEDGEMENT
This research was supported in part by a Grant-in-Aid (11470096) from the Ministry of Education, Japan.
SPSS software (version 8.0, SPSS Japan Inc., Tokyo) was used for all statistical comparisons.The x2 statistic was calculated to test the distribution trend of each index by genotype.Relative risks were calculated by logistic regression analysis method.

Table 2 .
Number and distribution of subjects related to angiotensinogen codon 235 genotypes.

Table 1 .
Characteristics of subjects by sex at baseline.

Table 3 .
Characteristics of subjects for angiotensinogen codon 235 genotypes at baseline .

Table 4 .
Distribution of various indices related to angiotensinogencodon 235 genotypes in 1998 .