Introduction

Essential hypertension (EH), which accounts for ∼95% of hypertensive cases, is an increasingly serious health problem in the developed countries. In China, hypertension is one of the fastest growing diseases of the past 30 years. According to a 2002 survey, the prevalence rate of hypertension among Chinese adults was ∼18.8%, with a total of 170 million people suffering from hypertension.1 EH is generally regarded as a paradigmatic multi-factorial disease that is determined by a combination of genetic factors, environmental stimuli and their interaction.2 It is estimated that ∼20–60% of the inter-individual variation of blood pressure (BP) is genetically controlled.3 Accordingly, the discovery of many potential hypertension-susceptibility genes has allowed for a better understanding of disease etiology.

The rennin–angiotensin–aldosterone system is an important regulator of BP.4 Angiotensinogen (AGT) is a liver protein that interacts with renin to produce angiotensin I, the prohormone of angiotensin II, which is the major effector molecule of rennin–angiotensin–aldosterone system. AGT gene variants can modify the plasma AGT concentration, which has been directly linked with arterial BP.5 Among these variants, the AGT T174M polymorphism (rs4762), a C to T conversion at nucleotide position 521 in exon 2, results in the replacement of threonine by methionine at codon 174. It will be included in the meta-analysis.

In 1992, the AGT gene T174M polymorphism was first reported to be related to EH prevalence by Jeunemaitre et al.5 Since then, there has been a great effort to further elucidate their association. For the Chinese population, some studies6, 7, 8, 9, 10, 11, 12, 13 have implied that the T174M polymorphism is associated with EH or BP, whereas other studies14, 15, 16, 17, 18, 19, 20, 21 were unable to replicate these findings. In fact, many studies focusing on the Chinese population provided equivocal or largely negative evidence for the association between the polymorphism and hypertension. To fully elucidate the effect of the T174M polymorphism on EH risk in the Chinese population, we conducted a carefully designed meta-analysis that included all the eligible case–control studies published to date.

Methods

Identification and eligibility of relevant studies

To search for all the studies that examined the association of the T174M polymorphism with EH risk in the Chinese population, we conducted a computerized literature search of the PubMed, EMBASE, CBM (China Biological Medicine Database), CNKI (China Nation Knowledge Infrastructure Platform), Wanfang and VIP databases, using the following keywords and subject terms: ‘AGT’, ‘polymorphism’, ‘hypertension’ and ‘Chinese or China or Taiwanese or Taiwan’. The present meta-analysis eligibility deadline was February 2011. Eligible publications had to be written in either Chinese or English. The references of all retrieved articles were also screened. To prevent data duplication, when a report overlapped with another study, only the most detailed study was included. If an article reported results on different ethnic sub-populations, each sub-population was treated as a separate study in our meta-analysis. Studies included in the meta-analysis had to meet all the following criteria: (i) the presentation of an investigation of the relationship between the AGT T174M polymorphism and EH in the Chinese population, (ii) the use of an unrelated case–control design (family-based study design with linkage considerations was excluded), (iii) the available genotype frequency, (iv) the genotype distribution of the control population had to be in Hardy–Weinberg equilibrium (HWE) and (v) the study had to define hypertension as systolic (SBP) ⩾140 mm Hg and/or diastolic (DBP) ⩾90 mm Hg22, 23 and/or treatment with anti-hypertensive medication. If the genotype frequency was not reported, we contacted the original authors by e-mail to obtain the missing data.

Data extraction

To minimize the selection bias, two authors independently extracted the information from each study. Disagreements were resolved by discussion between the authors. The following information was gathered from each study: first author, year of publication, racial background and resident region of study population, genotype detection method of each study, diagnostic standard, matching in sex and age, number of cases and controls, and distribution of genotypes and alleles in both the case and control groups.

Statistical analysis

As case–control studies were used, odds ratios (ORs) corresponding to a 95% confidence interval (CI) were applied to assess the strength of the association between the T174M polymorphism and hypertension, and the OR was calculated according to the method described by Woolf.24 We tested only the dominant genetic model (MM+MT vs. TT) because (i) the low frequency of homozygosity for high-risk alleles would yield a considerable number of studies with zero cell counts, generating unreliable OR estimates and (ii) the combination of MM and MT genotypes into one group was utilized in most primary studies included in our meta-analysis. The current strategy for analysis was consistent with that used in some previous studies.25, 26, 27, 28, 29

In our study, two models of meta-analysis were applied for dichotomous outcomes in Review-Manager 5.0.25 software (The Cochrane Collaboration, Oxford, UK): the fixed-effects model and the random-effects model. The fixed-effects model, using the Mantel–Haenszel method, assumes that studies are sampled from populations with the same effect size, making an adjustment to the study weights according to the in-study variance. The random-effects model, using DerSimonian and Laird's method, assumes that studies are taken from populations with varying effect sizes, and calculates the study weights both from in-study and between-study variances, considering the extent of variation or heterogeneity. We performed a χ2-based Q-statistic test to assess the between-study heterogeneity.30 Heterogeneity was considered significant for P<0.10 because of the low power of the statistic. The inconsistency index I2 was also calculated to evaluate the variation caused by heterogeneity rather than by chance. Higher values of the index indicate the existence of heterogeneity.31 The fixed-effects model (if P>0.10) or the random-effects model (if P<0.10) were used to pool the results.32 The significance of the pooled ORs was determined by the Z-test, and a P<0.05 was considered significant.

Subgroup analysis according to racial descent was carried out for the Han Chinese and non-Han Chinese minority populations to estimate ethnic-specific OR. In addition, subgroup analysis according to different diagnostic standards for hypertension was also performed.

Sensitivity analyses were conducted by sequential removal of single studies in an attempt to identify the potential influence of the individual data set on the pooled ORs. Publication bias was investigated by funnel plot, in which the standard error of the log of each study was plotted against its corresponding OR. An asymmetric plot suggested possible publication bias. Funnel-plot asymmetry was assessed by Egger's linear regression test.33 We performed a t-test to determine the significance of the intercept, and a P<0.05 was considered significant. HWE was tested with a χ2 test for goodness of fit based as applied by a web program (http://ihg.gsf.de/cgi-bin/hw/hwa1.pl). All other statistical analyses were performed using ReviewManager 5.0.25 and the Stata version 10.0 software (Stata Corporation, College Station, Texas, USA). All P-values were two-sided.

Results

Studies included in the meta-analysis

After the literature search and selection applying our inclusion criteria, 22 relevant articles on the relationship between the T174M polymorphism and EH in the Chinese population were identified. Among the 22 eligible articles, a study by Chiang et al.34 was replaced with a later report21 that included a larger population. Moreover, Fang et al.35 and Niu et al.36, 37, 38 were excluded as they were family-based studies. In addition, the genotype data of the control population provided in He et al.39 was excluded, as it deviated from HWE (PHWE <0.0001). After exclusion, 16 studies, comprising 3828 hypertensive patients and 3251 controls, were collected as considered appropriate for the meta-analysis.6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 An included study by Niu et al.11 was an unpublished thesis that was acquired from a medical doctorate dissertation database. This database is a common sub-database shared by the CNKI and Wanfang databases. Yuan et al.13 provided data on two Chinese minority populations: the Hani and the Yi. The two minorities were treated as separate studies. The characteristics of the selected studies are summarized in Table 1. The populations among these studies were as follows: thirteen studies involved Han Chinese subjects (3051 cases and 2620 controls), and three studies involved non-Han Chinese minority populations (777 cases and 631 controls) including Kazakh, Yi, Hani and Aims populations. Of the 16 studies, 69% (11/16) stated that the age and sex status were well matched between the study case and control population; 75% (12/16) were age-matched and 94% (15/16) were gender-matched. All the studies used a blood sample for genotyping.

Table 1 Detailed characteristics of eligible studies considered in the meta-analysis
Full size table

Main meta-results and subgroup analysis

The distribution of genotypes and alleles in the individual studies is listed in Table 2. The significance level for HWE testing for controls is also shown in Table 2. We observed a wide variation of the 174M allele frequencies in cases and controls ranging from 0.0463 to 0.234 and 0.0375 to 0.1993, respectively, across different studies. Accordingly, the pooled overall frequency of the 174M allele in the Chinese population was 12.64% in hypertensive cases and 10.78% in normotensive controls. The main results of this meta-analysis and the heterogeneity test are listed in Table 3. For all subjects, the random-effects model was used to pool the results, as the between-study heterogeneity was significant. There was no significant association between the T174M polymorphism and hypertension in the total of the 16 studies using the dominant genetic model (MM+MT vs. TT: P=0.24, OR=1.14, 95% CI 0.92–1.41, Pheterogeneity=0.002, I2=57%; Figure 1).

Table 2 Sample size, the distribution of genotypes and allele frequencies of cases and controls, and P-values of HWE in controls
Full size table
Table 3 OR (95% CI) of the association of the T174 M polymorphism and hypertension in different subgroups under the dominant genetic contrast
Full size table
Figure 1
figure 1

Meta-analysis examining the overall association between the T174M polymorphism and hypertension under the dominant genetic model (MM+MT vs. TT). ‘Events’ indicates the total count of individuals with the MM+MT genotypes. ‘Total’ indicates the total number of individuals. A full color version of this figure is available at the Hypertension Research journal online.

Full size image

In the subgroup analysis by ethnicity, all studies were categorized into two groups: Han Chinese and non-Han Chinese minorities. For the latter, there was only one study that considered Kazakh, Yi, Hani and Aims populations. The 174M allele was more common in Han Chinese cases and controls (13.09% and 11.24%, respectively) than in non-Han Chinese minorities (10.88% and 8.87%, respectively). For Han Chinese, no evidence of association between the T174M polymorphism and hypertension in the dominant genetic model could be found (MM+MT vs. TT: P=0.45, OR=1.11, 95% CI 0.84–1.47, Pheterogeneity=0.0005, I2=66%, random-effects model; Table 3). For non-Han Chinese minorities, no significant between-study heterogeneity existed (Pheterogeneity=0.57), and the fixed-effects model was used to pool the results. As with the Han Chinese, no association was found between the T174M polymorphism and hypertension in the minority populations (Table 3).

For the subgroup analysis based on different diagnostic standards, the data from three studies9, 17, 18 were combined to form a hypertensive group (SBP ⩾160 mm Hg and/or DBP ⩾95 mm Hg, 1978 WHO criteria.40) In this subgroup, a significant association was found in the dominant genetic model and the between-study heterogeneity was insignificant (MM+MT vs. TT: P=0.03, OR=1.71, 95% CI 1.07–2.74, Pheterogeneity=0.27. I2=24%, fixed-effects model; Figure 2).

Figure 2
figure 2

Meta-analysis examining the association between the T174M polymorphism and systolic (SBP) ⩾160 mm Hg and/or diastolic (DBP) ⩾95 mm Hg hypertension under the dominant genetic model (MM+MT vs. TT). ‘Events’ indicates the total number of the MM+MT genotype individuals and ‘Total’ indicates the total number of the MM+MT genotype plus the TT genotype individuals. A full color version of this figure is available at the Hypertension Research journal online.

Full size image

Sensitivity analysis

To investigate the impact of individual data sets on the pooled OR, we sequentially deleted data from single studies involved in the meta-analysis. No individual study had an undue influence on the pooled ORs, and the between-study heterogeneity still existed for all subjects and the subgroup analysis of Han Chinese when any single study was excluded.

Publication bias

A Begg's funnel plot and Egger's test were performed to assess the publication bias in the literature. As shown in Figure 3, the shape of the funnel plot did not reveal any evidence of obvious asymmetry in the dominant model, and the Egger's test suggested an absence of publication bias among all studies (t=0.26, P=0.800 for MM+MT vs. TT).

Figure 3
figure 3

Begg's funnel plot analysis was used to detect publication bias for the dominant model (MM+MT vs. TT). No asymmetry was found as indicated by the P-value of the Egger's test.

Full size image

Discussion

The literature examining the relationship between the T174M polymorphism and EH risk in the Chinese population was replete with small studies with conflicting findings. No clear consensus has been reached. Therefore, we performed the present meta-analysis on the Chinese population that included 16 studies from 12 provinces with 3828 cases and 3251 controls. Only the dominant genetic model was selected to reduce the chance of false-positive findings. Unfortunately, we were unable to identify a significant association between the T174M polymorphism and hypertension in all subjects. However, several potential explanations may explain the lack of association between them in the population.

First, it should be noted that hypertension is a complex polygenic disease. A single polymorphism or gene likely has weak effects on the individual's phenotype, as complex traits presumably arise from multiple interacting polymorphisms or genes. A study by Hegele et al.41 showed that the AGT T174M polymorphism only accounted for 3.1% of the total variation in SBP in men. Some previous studies revealed that the T174M variant was in tight linkage disequilibrium with other variants such as coding region variants, promoter variants and other yet unknown functional AGT polymorphisms.5, 42, 43, 44 Therefore, it is necessary to evaluate the combined effect of T174M with other relevant polymorphisms in the AGT gene or different genes. The results from Yuan et al.13 suggested there was no evidence of association between T174M polymorphism by itself with hypertension in the Hani minority in China. However, when the T174M polymorphism was analyzed in combination with the M235T variant in the coding region, a significantly elevated risk of hypertension (OR=1.62, 95% CI 1.02–2.59; P=0.043) could be found. A meta-analysis by Ji et al.45 reported that the M235T variant increases the risk of hypertension in the Chinese population (OR=1.54, 95% CI 1.16–2.03, P=0.002), whereas the current meta-analysis has indicated no significant effect of the T174M variant on the incidence of hypertension. Owing to the low frequency of T174M, the statistical power of the research might be limited to detecting differences in OR estimates. Consequently, given that our study focused on the effect of a single polymorphism to determine the genetic determinants of EH, negative findings were not surprising. With a tight linkage disequilibrium between the M235T and T174M polymorphisms, the relationship between M235T and EH might partially be attribute to the effect of T174M. Additional well-designed studies with a larger population, especially studies investigating the combined effect of T174M and other polymorphisms are needed to fully elucidate the relationships between the polymorphisms and EH in the future.

Second, hypertension is an acknowledged multi-factorial disease. Beside genetic background, environmental factors and individual biological characteristics may also influence the occurrence and development of hypertension. The former includes salt intake, smoking and alcohol consumption, and the latter includes race, age, gender, body mass index and BP. For example, in at least one study, age was a determinant attribute of the penetrance of genetic variants, and the age-dependent genetic effects could not be ignored.46 Hiroyasu et al.47 reported that there was a higher prevalence of the T174M variant among persons with hypertension onset <55 years compared with those with later onset. Without comprehensively considering these factors, any analysis may fail in exploring the independent role of suspected polymorphism in hypertension.

China is a very large multi-ethnic country with 56 identified ethnic groups. Among these groups, Han Chinese are the largest ethnic group, making up over 93% of the total population.48 In the subgroup analysis, we divided all studies into two subgroups: Han Chinese and non-Han Chinese minorities. No significant association between EH and the T174M polymorphism was observed among studies considering Han Chinese. These results were in accordance with the results for the overall population. For non-Han Chinese minorities, a negative result was also obtained. In this subgroup, the genetic background was quite complex and four ethnic minorities were included: Kazakh, Yi, Hani and Aims. Moreover, both the studies and population of non-Han Chinese minorities were limited. Interestingly, a statistically significant association of the T174M polymorphism with the hypertensive group (SBP ⩾160 mm Hg and/or DBP ⩾95 mm Hg) was detected when different diagnostic standards were used (P=0.03, OR=1.71, 95% CI 1.07–2.74). The result might be worth deliberating because the pooled sample size of three studies was relatively small (216 cases and 279 controls). A sampling bias might exist, which would lead to the increase in the probability of a false-positive (type I error). Thus, after considering the possible bias, the positive results must be considered with caution.

In the present meta-analysis, the frequency of the 174M allele varied widely across the populations of Chinese and Han Chinese, which suggested that there was considerable heterogeneity in these populations. The genetic background of the Chinese population was intricate owing to the large number of ethnicities included. Furthermore, some studies49, 50 indicate that the genetic profile of the Han Chinese is mixed, and heterogeneity could exist. Sensitivity analysis was performed to investigate the heterogeneity source in both the overall and the subgroup analyses. The analysis indicated that the corresponding pooled ORs were not materially altered, thus indicating our results were statistically robust. Heterogeneity, however, still existed in the sensitivity analysis. It was possible that differences in diagnostic standards across the studies were a source of heterogeneity. In addition, heterogeneity might also result from specific clinical characteristics.

Our findings on the T174M polymorphism were different from the results from a meta-analysis by Pereira et al.,27 in which a similar population, East Asians, were studied. In that meta-analysis, there was a significant association between the T174M polymorphism and hypertension in the East Asian population (10 studies were included, only 4 of which considered ethnic Chinese subjects). The discrepancy between the Pereira results and the findings of our meta-analysis might be due to the potential publication bias observed in the previous meta-study. In addition, many Chinese studies in local journals were not included in that meta-analysis, potentially leading to a language bias.

To summarize, our meta-analysis failed to provide evidence for the genetic association of the AGT gene T174M polymorphism with hypertension risk in the Chinese population. Similar results were found in the subgroup analyses of Han Chinese and non-Han Chinese minorities. However, the T174M polymorphism did present a significant association with the hypertensive group (SBP ⩾160 and/or DBP ⩾95 mm Hg). Additional studies and large case–control studies, especially studies stratified for different stages of hypertension, should be performed to clarify the association between the T174M polymorphism and EH in the Chinese population. Further studies investigating gene–gene, gene–environment and their mutual interactions are certainly important, as well.