Subjects

We enrolled 134 unrelated consecutive Mexican Mestizo patients, who were referred for an HUTT test to the Electrophysiology Department of the National Institute of Cardiology in a 5-year period, with more than one unexplained syncope or presyncope episode and clinical suspicion of VVS. The inclusion criteria for participants were: adult patients, with a history of recurrent episodes of syncope or pre-syncope - defined as more than two episodes in the previous 6 months - and who agreed to sign the informed consent. Exclusion criteria were signs or symptoms of other ANS-related diseases, pregnancy, lactation, or patients who were taking medication interfering with the ANS. Subjects with a negative HUTT test (n = 46) were compared with those with a positive test (n = 88). Mexican ancestry origin was defined as individuals who were born in Mexico and traced back to the third generation. This project was designed according to the declaration of Helsinki and the Institutional Ethical and Research Committees approved the protocol. All participants were informed about the objectives and benefits of the results of this study for the Mexican population with VVS. In this case–control study, all patients signed the informed consent before enrollment.

HUTT test protocol

The HUTT test was performed between 9:00 and 12:00 h, and all the subjects had fasted the previous night. A surface three-lead electrocardiogram was continuously displayed, and blood pressure measured with a sphygmomanometer by the same researcher at intervals of 5 min and every minute when symptoms of impending syncope developed. Then, after the subjects had rested in a supine position for a minimum of 10 min to obtain stable values, a 5 min baseline recording was captured. Phase I: The participants were tilted to 70° using a motorized table with footboard support located in a silent room with a constant temperature of 21°C, for 20 min or until syncope was imminent, associated with a clear precipitous fall in blood pressure or heart rate (positive HUTT). The imminence of syncope was recognized by a subjective sensation of impending fainting, dizziness, nausea, sweating, pallor, altered mental state, and visual changes plus hypotension (systolic blood pressure < 80 mmHg). If a favorable response was not detected during Phase I of the test, the subjects were returned to the supine position, received 5 mg of sublingual isosorbide dinitrate and were tilted at the same angle for a further 12 min or less (Phase II).

DNA extraction

A 5-ml blood sample was obtained from participants, and the DNA was extracted by the method proposed by Miller et al.¹⁵. Once obtained, the DNA was measured in a Thermo Scientific™ NanoDrop™ spectrophotometer (Waltham, Massachusetts, USA) with 1–2 µL of sample, verifying the integrity of the product through an agarose gel (1%) dyed with ethidium bromide. After verification of quality, the DNA was aliquoted and stored.

Genetic analysis

The ADRA1A (rs1048101 [Company identifier: C___2696454_30], rs1383914 [C___2696575_1_], rs574584 [C___2315104_10], and rs573542 [C____903247_10]), ADRB1 (rs1801252 [C___ 8898508_10] and rs1801253 [C___8898494_10]), ADRB2 (rs1042713 [C___2084764_20] and rs1042714 [C___2084765_20]), and ADRB3 ([rs4994 [C___2215549_20]) SNPs were genotyped using the 5’ exonuclease TaqMan genotyping assays on a 7900HT Fast real-time polymerase chain reaction system following the manufacturer’s instructions (Applied Biosystems, Foster City, USA) (Table 1). The thermocycling protocol was: denaturation at 95°C for 10 min followed by 40 cycles of 15 sec at 95°C (denaturation) and 1 min at 60°C (annealing/extension). Samples previously sequenced of the different genotypes of the polymorphisms studied were included as positive controls.

Statistical analysis

For statistical analysis purposes, SPSS version 18.0 (SPSS, Chicago, IL) statistical package was used. Means or medians with standard deviation or interquartile range were calculated for baseline clinical characteristics. For comparisons between groups, we used the Chi-squared test and Student´s t-test or Mann–Whitney U-test as necessary. The Hardy-Weinberg equilibrium was evaluated by the Chi-square test. For the comparison between our Minor Allele Frequency (MAF) and those of open international databases, the Chi-squared test was used when the allele counts for each SNP was available. Logistic regression analysis was used to test for associations of polymorphisms with the HUTTT result using the following inheritance models: codominant, dominant, recessive, over-dominant, and additive. The effect size was calculated by odds ratio transformation to the Cohen’s d effect size estimator and considered as follows: small (0.2), medium (0.5), and a substantial effect (0.8). SNPstats (http://bioinfo.iconcologia.net/snpstats/start.htm) software for Windows® was used to analyze the genetic frequencies and to evaluate the linkage disequilibrium (LD, D’). Haplotype construction was made with Haploview version 3.32 (Broad Institute of Massachusetts Institute of Technology and Harvard University, Cambridge, MA, USA).

The potential role of α-ARs in VVS

Of the four polymorphisms located in the ADRA1A gene for this study, the rs1048101 was the only SNP with a nonsynonymous change; it has been associated to a greater effect of nifedipine among Chinese population¹⁷. This polymorphism was also related to the hemodynamic response in the treadmill exercise test, especially in men, being present in those with higher vasoconstriction (greater maximal systolic blood pressure)¹⁸. Hernández-Pacheco et al.⁷ studied 89 patients with VVS and compared them with 40 healthy Mexican controls. They analyzed the association of the Arg/Arg genotypes of the rs1048101 polymorphism with the occurrence of syncope (codominant model OR = 13.21, 95% CI 3.69-54.99, p < 0.001 and additive model OR = 12.68, 95% CI 3.5-53.07 p < 0.001). As reported by Sorrentino et al.¹¹, in the present work, no association was found between the same polymorphism and a positive HUTT. The main difference between these reports is that Hernández-Pacheco’s et al. group included healthy, sex- and age-matched controls, while in our study and Sorrentino’s et al., there were no healthy controls, and the reference group was the negative HUTT group. Nevertheless, when compared to the healthy population from open databases (1000 Genomes project and Exome Aggregation Consortium) there were no differences between ADRA1A receptor polymorphisms and the positivity in the HUTT test.

The other α-AR SNPs (rs1383914, rs574584, and rs573542) are in no coding regions of the gene and were included in the present study because a previous study reported the association of these SNPs with fibromyalgia in Mexican and Spanish populations. However, we did not detect any association in this study¹⁴.

The potential role of β-ARs in VVS

The rs1801252 and rs1801253 polymorphisms in ADRB1 genes were previously studied in essential hypertension in a Chinese population¹⁹. The rs1801253 polymorphism (Arg389Gly) was found to be associated with the disease. In vitro studies demonstrated that the Arg389 variant of the ADRB1 gene elicits an increased response to agonist stimulation compared with the Gly389 variant²⁰. In our study, this polymorphism was associated with a positive HUTT test.

Our study included two ADRB2 polymorphisms (rs1042713 and rs1042714) that have been previously analyzed by Diatchenko et al.²¹. This group examined 3 ADRB2 polymorphisms and reported an association of one haplotype (rs1042713, rs1042714, and rs1042717) with the presence of low blood pressure, somatization, and temporomandibular joint syndrome. In the same way, Vargas-Alarcón et al.¹⁴ found an association between β2-AR AC haplotype (rs1042713 and rs1042714) and fibromyalgia in Mexican and Spanish individuals. In our cohort, only the rs1042713 was associated with a positive HUTT test.

In the ADRB3 gene, we studied the rs4994 polymorphism, which has been associated with increased cardiovascular risk in women without evidence of obstructive coronary artery disease¹². In our study, under two different models, this polymorphism was associated with an increased risk to have a positive HUTT.

The rs4994 in the ADRB3 was overexpressed in the positive HUTT test group compared with the Mexicans in the 1000 Genomes project; this may enhance the findings of a pathologic relationship between this polymorphism and the positivity of the test.

Resolving discrepancies in the literature among genetic studies of VVS

The findings herein reported contradicted some previous studies, even when they were performed in a Mexican Mestizo population⁷-⁹, but also supported the results of others. For example, Zelazowska et al.¹⁰ and Sorrentino et al.¹¹, with limited sample sizes and in populations with the different genetic background (Italian and Polish), found a negative association for rs1801253.

At this point, in the study of a genetic basis for VVS, there is still much confusion about the validity of the results so far reported. This confusion stems from three principal hurdles: all these studies, including the one presented here, have a limited sample size; therefore, studies in many individuals are necessary to establish the real effect of these polymorphisms. The second probable source of error is a subtle population stratification in self-reported ethnicity observed within the ethnically homogeneous population, which can also occur in ethnically mixed populations and can lead to detecting significant associations at loci that have nothing to do with the disease. The gold standard to avoid this problem is the use of matched case–control patients with ancestry markers. The third point to consider is that the frequency of syncope can vary from country to country and even in their borders. Furthermore, the genetic and relevant environmental factors within populations may not be balanced between initial and subsequent studies; the variations in effect size from gene-gene and gene-environment interactions could in part explain the inability to replicate the association²².

In this study, we included five models of inheritance to be considered in future replication studies. There is no current consensus about the best model of inheritance to be chosen in association studies. However, in the last years, the additive model has been proposed as the preferred model to detect genetic association even when the real model of inheritance is different²³.

Other limitations are the increased cost and pitfalls for recruitment, which is reflected in a lack of a bigger sample size to detect a lower effect, as is usual in multifactorial disorders. However, in the results obtained here, an important effect could be detected in a few polymorphisms in this small sample because the calculated OR was high (OR=8.63, 95% CI 0.95-78.72). This can be seen in the polymorphism Ser49Gly located in ADRB1 under the recessive model with a d = 1.188, and also in the Trp64Arg located in ADRB3, under the dominant and over-dominant models with a medium effect size (d = 0.496 and 0.531) with ORs higher than 2. In the same context, the power of the study was low for the great part of polymorphisms studied but high in those polymorphisms where a high effect size was detected, and an allelic frequency was more prominent than 0.25. Another limitation of the study is the case–control approach because, due to the reduced sample size, a family approach is a better alternative to obtain best power for similar sample size and avoid the problem of subpopulation stratification²⁴.

One strategy to overcome these obstacles is to form a multiethnic consortium in the study of the genetics of VVS, to determine the participation of polymorphisms in these four genes of ADRA and ADRB in patients with VVS.