Essential hypertension is a common, polygenic, complex disorder resulting from interaction of several genes with each other and with environmental factors such as obesity, dietary salt intake, and alcohol consumption. Since the underlying genetic pathways remain elusive1, currently most studies focus on the genes coding for proteins that regulate blood pressure as their physiological role makes them prime suspects.

The Renin-Angiotensin System (RAS) has a central role in regulating blood pressure and sodium homeaostasis. Genes encoding components of RAS, including angiotensinogen (AGT), angiotensin-converting enzyme (ACE), angiotensinogen II type-1 receptor (AGTR1), and renin, have been extensively investigated as genetic determinants of essential hypertension 2. Polymorphisms of RAS3 genes seem also to play a role in the development of diseases that cause secondary hypertension45. Subjects carrying the ACE D allele have unanimously been shown to have increased ACE serum activity36 while the T235 AGT variant has been associated with elevated angiotensinogen levels 7. However, so far there are no consistent findings.

In 1992, the M235T AGT TT polymorphism was first reported to be associated with hypertension 8. This finding has not been confirmed by all investigators910. Although no relationship between the ACE gene and hypertension was observed in one early linkage study 11 and most recent studies including one meta-analysis 121314, several studies have suggested a role: hypertensive individuals have a high prevalence of the D allele or DD genotype 31516.

The inconsistent results might be explained in part by the genetic and environmental heterogeneity among different ethnic groups 13. On the other hand, one recent study 17 reported that the MM, AA, CC, DD/ID genotype combination was associated with a substantially higher prevalence of hypertension in the participants to the Olivetti Heart Study, even though no individual effect of each isolated genotype was detected.

The present study investigates the relationship between variants of the I/D ACE gene and M235T AGT gene, and the presence and severity of essential hypertension in a large homogeneous German population. The effect of a combination of ACE and AGT gene polymorphisms on hypertension was also examined.

Demographic data are summarized in Table 1. Hypertensives were older (58.80 ± 13.22 years) than controls (41.24 ± 12.66 years, p < .0001), and more often male (53.2% vs. 44.4%, p = .001).

AGT M/T genotyping was successful in 637 hypertensives and 720 normotensives, and ACE I/D genotype was analyzed in 636 hypertensive and 719 control subjects. The roles of age and gender in the association between hypertension and ACE and AGT gene polymorphisms were examined by comparing the effects of ACE and AGT genotypes for hypertension in men and women, young and elderly subjects respectively. "Young" was defined as "age < 50 years old", and "elderly" was characterized as "age ≥ 50 years old".

Although some recent studies 15164445 suggested a unique sex-specific role of ACE in essential hypertension, no significant association of essential hypertension with the ACE gene I/D polymorphism was observed in this German population of 1,358 for either gender. This finding confirms earlier observations in another German population 31, in other Caucasian populations 11121446, and in one meta-analysis 13. The distribution of the ACE genotypes was in Hardy-Weinberg equilibrium in this German population while that was not the case in Pereira et al's study 1. One possible explanation is the ethnic difference. Pereira et al. 1 showed that there were statistically different ACE I/D polymorphism genotypic frequencies in different ethnic groups.

Surprisingly, the cross-sectional study presented here shows a higher prevalence of the T/T M235T AGT gene in the control group compared to the hypertensive group. AGT TT genotype was associated with a decrease in the risk for hypertension (odds ratio-TT vs. MM: 0.52; 95% CI: 0.28 to 0.96) and a more significant association was found in women (odds ratio-TT vs. MM: 0.28; 95% CI: 0.1 to 0.78), compared to men. This is in stark contrast to findings from previous studies, including two German datasets 3132, and three meta-analyses 102247, which reported that the AGT 235 T-allele and/or TT genotype significantly increased the risk for essential hypertension in Caucasians: odds ratio T vs. M was 1.20 (95% CI: 1.11 to 1.29) 10, odds ratio TT vs. MM was 1.31 47, and odds ratio TT vs. MM was 1.19 (95% CI: 1.10 to 1.30)22. In agreement with the previous meta-analyses, the meta-analysis presented here showed increased odds for hypertension (odds ratio: TT vs. MM: 1.21) in Caucasians conferred by TT, and the odds ratio rose by 2.5% when the present study was excluded. Of the 25 studies included in the present meta-analysis, the present study was the only one in which the AGT T235T genotype decreased odds for hypertension. Nevertheless, the quality of meta-analysis results depends on the quality of the individual studies included, and unusual sample sizes might bias the finding. For example, one single study 43 included in the previously largest meta-analysis 22 was exceptionally large, giving it enormous weight. The highly variable study quality implies that all interpretations must be made with great caution, as was explicitly pointed out by Kunz et al. 10

Although no significant difference was observed in AGT T allele frequency distribution between hypertensives and controls with respect to age and gender, the frequency of the AGT T allele among normotensives was higher than that among hypertensives (0.36 vs. 0.33). This was inconsistent with one previous meta-analysis in Caucasians, which showed that among controls, the mean allele frequency for the AGT T allele was 0.41 (95% CI: 0.34 to 0.48), and among cases, increased to 0.45 (95% CI: 0.38 to 0.52) 10. In the present study, the frequency of the AGT T allele among hypertensives (0.33) was outside the lower border of the 95% confidence interval (0.38 to 0.52) reported in 10. This may reveal the specific genetic background of this particular German population.

The AGT genotype distribution is not in Hardy-Weinberg equilibrium (P < 0.001) while no deviations from Hardy-Weinberg equilibrium are observed on the ACE genotype distribution in the same population. This may be explained by a shift toward a higher frequency of MT individuals (62.4%) instead of TT individuals (3.7%) in this specific population.

The study population presented here contains a large proportion (65%) of patients with renal disease. While selection of participants based on patient records excluded those patients that had symptoms suggesting the diagnosis of secondary hypertension at first contact, the possibility remains that at least a part of the study population suffers from renal rather than essential hypertension. It should be noted, however, that the majority of studies included in the presented meta-analysis does not give specific information regarding renal function of hypertensives, and the largest study 43 is population based and does not name any specific, kidney related exclusion criteria.

Another possible explanation is that the AGT T allele frequency may decrease with age, which was reported from the United Arab Emirates 48. The prevalence of hypertension increases with advancing age. According to the National Health and Nutrition Examination Survey III (NHANES III) prevalence estimates for the years 1988–1994, American Whites aged 55 to 64 years have a more than threefold higher prevalence of hypertension (42.1%) than those aged 35 to 44 years (11.3%) 49. Frossard et al detected that AGT T allele frequency decreased with age in the United Arab Emirates48. The ACE DD genotype was found associated with human longevity 50. In the present study, normotensives are significantly younger than hypertensives (41 ± 12 yrs vs. 59 ± 13 yrs). It is conceivable that many of the young individuals are at hypertension risk because of their ACE or AGT genotype, but have not yet shown hypertension at the time of genotyping, and may develop hypertension in their older age. This might lead to some misclassifications and hence reduce the power of this study. On the other hand, two studies of German populations 3251 reported that the AGT T allele was a risk factor for hypertension in individuals younger than 50 years of age. In the present study, young hypertensives had a higher frequency of the AGT T allele than elderly hypertensives (0.36 vs. 0.32), but the difference didn't reach statistical significance (p = .22). It is possible that the small percentage (24.9%) of the studied hypertensives under 50 years of age biased the finding.

A number of studies 51525354 examined the relationship between RAS genotype and the severity of hypertension, but their results were contradictory. In accordance with two 5254 of them including one German dataset, the present study fails to find an effect of the AGT or ACE genotype on the severity of hypertension. Nevertheless, in the present study, hypertensives that carried at least one copy of the AGT T allele (TT or MT: n = 397) were less likely to take two or more antihypertensive medications than those with MM genotype (n = 225) (odds ratio -TT or MT vs. MM: 0.797; 95% CI: 0.57 to 1.12; P = .185), and their average number of antihypertensive drugs was lower (2.09 vs. 2.20; P = .276). Despite not reaching statistical significance, this observation was in contrast to Schunkert et al's study 51 on another German population (subjects initially participated in the MONICA Augsburg cohort baseline survey), which found that the carriers of the AGT T allele (n = 418) had a 2.1-fold higher probability of taking two or more antihypertensive drugs than individuals with the MM genotype (n = 216). It is worth pointing out that in the present study, the number of subjects taking antihypertensive medications is much larger than in Schunkert et al's study (622 vs. 143). While the data is far from statistical significance, the trend is in line with those findings that associate rather the M allele with hypertension in the present study.

The effect of a combination of RAAS genes' polymorphisms on blood pressure has been investigated in the participants to the Olivetti Heart Study 17, in which the MM, AA, CC, DD/ID genotype was detected to be associated with a substantially higher prevalence of hypertension in the absence of detectable effects of each individual polymorphism at any single locus. The present study showed very similar findings in women: MM, DD/ID genotype significantly increased the odds for hypertension (odds ratio: 1.45; 95% CI: 1.04 to 2.02), while TT, DD/ID and TT, II/ID were significantly associated with lower prevalence of hypertension (odds ratio: 0.33, 95% CI: 0.11 to 0.99; odds ratio: 0.31, 95% CI: 0.10 to 0.92). The risk of hypertension in the women with TT, DD/ID and TT, II/ID, however, didn't change much compared with those with TT alone (odds ratio: 0.31, 95% CI: 0.12 to 0.83). This suggests that there is only a slight synergistic effect between the AGT and ACE genes.

Although in most surveys the prevalence of hypertension appears to be equal in women and men 55, sex-specific effects of ACE or AGT genes on hypertension have been reported recently 1643. For instance, Sethi et al. 43 found the AGT TT associated with an increase in risk for hypertension in women but not in men from the Copenhagen City Heart Study with a population of 9100 subjects, and an association of the ACE DD genotype with increased diastolic blood pressure was detected in men, but not in women from the Framingham Heart Study 16. In the present study, the AGT TT genotype was negatively correlated to hypertension in women only while no sex-specific effect of the ACE gene was shown. It is possible that the fact is covered by the different gender distributions: in this population, there are more women in normotensives than in hypertensives (55.6% vs. 46.8%, P = .001).

In addition to the factor declared above, the result may be influenced by the study design and the composition of the sample population. The study design itself may influence the results. As mentioned above, this study was a cross-sectional study where subjects were assessed at a single time, and at that time most of the controls were younger than 50 years old. Animal studies have shown that hypertension genes may be activated for only certain periods during the life history of an organism 2. Hence, some of them might develop hypertension at an older age, resulting from the activated hypertension genes. Longitudinal studies are needed to further examine the relationship between hypertension and genes at different ages.

The population can be described as static population with a mixed Germanic-Slavonic background. Due to the location (a provincial town on the German-Polish border) and political and historical setting (little population fluctuation during the over 140 years of imperial, fascist and socialist rule, no influx due to lack of economic attractivity after reunification), the population may be assumed as homogeneous. Several recent studies 565758596061 have reported significant differences in prevalence of hypertension between Germany and Poland, which, however, are assumed to be largely dependent on life-style differences, mostly salt intake. These differences, however, are unlikely to play a role in the study population due to its homogeneity with regards to life style preferences following successful assimilation over many generations. In the case of ACE polymorphism with all allele frequencies greater than 15%, there is no need to examine population stratification 62. This becomes more urgent for AGT where there is no Hardy- Weinberg equilibrium. As the sample information, however, does not include data on the ethnic background of the probands, additional haplotyping carried out on the samples would not have allowed to rectify for the historical ethnic background (Germanic versus Slavonic).6263

In the meantime, it should be noted that the present study was carried out on an unusually large population (second only to one in 63 studies included in Sethi's meta-analysis22). Given the large homogeneous population-based sample, the findings cannot be attributed to simple selection bias. Therefore, the finding that the AGT TT genotype associated with a decreased risk for essential hypertension is likely to be true for this particular German population.

Despite the limitations mentioned, this cross-sectional study does not support the notion that the ACE I/D polymorphism contributes to the prevalence and severity of essential hypertension. However, the M235T TT genotype of AGT gene was detected to confer a significantly decreased risk for the prevalence of hypertension in women from this particular population. Despite the large sample size, the present study fails to revise the odds ratio in a meta-analysis of a total of 25 studies on the association between the AGT M235T polymorphisms and hypertension in caucasians. This observation may reflect a very specific local inheritance pattern of the AGT genotypes. If this holds true, studies aiming at drug development based on genomic traits must be scrutinized rigorously as therapeutic recommendations may be valid for selected subpopulations only64.

The author(s) declare that they have no competing interests.

M.N. designed and initiated the study, analyzed the samples and wrote part of the manuscript. A-L.Z. and M.L. did the statistical analysis. A.M. devised the concept for meta-analysis and wrote the manuscript. L.P. created the figures. M.N. and A.M. should be considered as joint co- authors.