Fibrillin-1 genotype and risk of prevalent hypertension: A study in two independent populations

Abstract

Objective. Mutations in the fibrillin-1 gene are the cause of Marfan syndrome. We wanted to investigate the relationship between a mutation in this gene and risk of prevalent hypertension. Methods. In a cross-sectional study, the effect of a G–A substitution in intron 27 in the fibrillin-1 gene (rs11856553) on risk of prevalent hypertension was studied in two large population-based studies: the Health 2006 study, consisting of 3193 women and men, age 18–69 years, and the MONICA10 study, consisting of 2408 women and men, age 41–72 years. In 1646 MONICA10 participants, blood pressure (BP) was also measured by 24-h ambulatory recordings. Results. Among the 3193 Health 2006 participants 23 had the G–A variant, and among the 2408 MONICA10 participants 18 had the G–A variant. In Health 2006, the odds ratio estimate (95% confidence intervals) for the G–A variant for risk of hypertension, defined as systolic (S) BP ≥ 140 mmHg or diastolic (D) BP ≥ 90 mmHg or on antihypertensive medicine, was 2.67 (1.14–6.18), p = 0.022. The corresponding figure for moderate to severe hypertension, defined as SBP ≥ 160 mmHg or DBP ≥ 100 mmHg, was 9.68 (4.24–22.12), p < 0.0001. In MONICA10, the odds ratio estimate (95% confidence intervals) for the G–A variant for risk of moderate to severe ambulatory hypertension, defined as 24-h mean SBP ≥ 150 mmHg or 24-h mean DBP ≥ 90 mmHg, was 5.73 (1.96–16.7), p = 0.0014. Conclusion. The G–A substitution in the fibrillin-1 gene (rs11856553) is a rare genetic variant that is associated with an increased risk of prevalent hypertension, particularly of moderate to severe prevalent hypertension.

Key Words::

• Ambulatory blood pressure
• cardiovascular disease
• epidemiology
• fibrillin-1 gene
• general population
• hypertension

Introduction

Fibrillin-1 is a major structural component in elastic fibers (1,2). The importance of fibrillin-1 in normal vascular formation was originally underscored by the finding that mutations in the fibrillin-1 gene are the cause of Marfan syndrome (1,2). The role of mutations in this gene in more common cardiovascular diseases (CVD), including hypertension, is still not clear, although a study by Medley et al. indicated that the single-nucleotide polymorphism (SNP) 5 in intron 27 in the fibrillin-1 gene was associated with aortic stiffness and coronary artery disease severity in patients with ischemic heart disease (3). So far, no studies have been published relating mutations in the fibrillin-1 gene to risk of hypertension and CVD in the general population.

In this study, we primarily investigated the effect of a G to A substitution at SNP 5 in intron 27 in the fibrillin-1 gene (rs11856553) on risk of prevalent hypertension in two large population-based, cross-sectional studies. In one of these studies, blood pressure (BP) was determined both in the office and by 24-h ambulatory BP monitoring (4), the latter method being the gold standard for the assessment of the total BP burden on the arterial wall (5), and the gold standard to make the best diagnosis of hypertension (6). In addition, we also investigated the effect of the G to A substitution at SNP 5 in intron 27 in the fibrillin-1 gene on risk of incident CVD in a prospective substudy to complement our main BP investigation.

Methods

Study populations

The two studies, the MONICA10 study and the Health 2006 study, have been described in detail elsewhere (4,7,8). The first study consisted of 2656 re-invited participants who originated from the Danish part of the MONItoring of trends and determinants in CArdiovascular disease (MONICA) health survey (9), which was performed from 1982 to 1984. The MONICA participants were randomly selected from 11 municipalities within Copenhagen County. Furthermore, for the MONICA protocol, participants were selected to represent equal numbers of women and men aged 30, 40, 50 and 60 years. The MONICA10 study was done from 1993 to 1994 in the Research Centre for Prevention and Health in Glostrup, Denmark. All MONICA participants completed a questionnaire about current and previous diseases, intake of medication, and lifestyle risk factors. The MONICA study was conducted in agreement with the Second Helsinki Declaration and was approved by the local Ethics Committee system in Denmark. Written informed consent was obtained from all subjects. The present MONICA investigation of genetic determinants of cardiovascular phenotypes was approved as a separate protocol (KA-04130). In total, 248 MONICA10 participants were excluded from the present investigation, owing to (i) a previous diagnosis of myocardial infarction or stroke or use of digoxin or nitrates (n = 163), or (ii) no material available for genotyping (n = 85), leaving 2408 women and men to be included in the analyses. In addition, in a subgroup, consisting of 1646 participants, BP was also determined by 24-h ambulatory BP monitoring, as described in detail elsewhere (4). To be included in the ambulatory BP subgroup, all 1646 participants had to have at least 14 readings of systolic (SBP) and diastolic BP (DBP) during the day and at least seven readings of SBP and DBP during the night, complying with international recommendations (4).

The Health 2006 study is a population-based study investigating the prevalence and risk factors of chronic diseases such as asthma, allergies, CVD and diabetes (7). The study took place between June 2006 and June 2008. A random sample of the general population aged 18–69 years living in the South-Western part of Copenhagen County was drawn from the Civil Registration System. A total of 7931 persons were invited to a health examination and 3471 (43.8%) participated. Participation rate was significantly associated with female gender and increasing age (p < 0.001) (7). The health examination included a self-administered questionnaire, as described above, and a health examination. The Health 2006 study was approved by the ethics committee of Copenhagen County (KA-20060011) and the Danish Data Protection Agency. Informed written consent was obtained from all participants. In total, 278 Health 2006 participants were excluded from the present investigation owing to (i) a diagnosis of myocardial infarction (n = 45) or other major cardiovascular diseases (n = 120), (ii) no material available for genotyping (n = 89), or (iii) participation in the MONICA10 study (n = 24), leaving 3193 women and men to be included in the analyses.

Genotyping

Genomic DNA was isolated from blood samples, which had been stored in EDTA tubes at − 20°C, using a commercial available kit. The A:A, G:A and G:G variants at SNP 5 in intron 27 in the fibrillin-1 gene (rs11856553) were genotyped by KBioscience (http://www.kbioscience.co.uk). The in-house assay KASPer or the TaqMan system (Applied Biosystems) were used. The assays were validated prior to use and by 5% duplicates in the population. Using the test for heterozygote excess in Genepop v.4.0.10 on the web: http://genepop.curtin.edu.au/, we found no significant departure from Hardy–Weinberg equilibrium in either MONICA10 (p = 0.97) or Health 2006 (p = 0.99).

Lipid and glucose analyses

Fasting concentrations of lipids were measured using commercially available enzymatic methods, and plasma glucose was analyzed using a hexokinase reagent kit, as described elsewhere (8).

Clinical and demographic variables

A physical examination was performed by trained staff and included anthropometric measurements and recording of BP. Office BP was measured in the sitting position after 5 min of rest using a random zero mercury sphygmomanometer (4,7,8). The mean of two measurements was reported. In a subset of the MONICA10 population (n = 1646), 24-h ambulatory BP monitoring was performed. The ambulatory BP measurement was initiated immediately after recording of office BP using a validated Takeda TM-2421 (A&D, Tokyo, Japan) device (4). Oscillometric recordings were used in this study (4). BP recordings were made every 15 min between 07:00 and 23:00 h, and every 30 min between 23:00 and 07:00 h (4). Means of ambulatory BP were computed with weights according to the time interval between successive readings (4).

Definition of clinical conditions

For the present study, we defined hypertension by conventional criteria as office SBP ≥ 140 mmHg or office DBP ≥ 90 mmHg or use of BP-lowering drugs. In addition, we defined uncontrolled hypertension as office SBP ≥ 140 mmHg or office DBP ≥ 90 mmHg, irrespective of use of antihypertensive medicine. Furthermore, irrespective of use of antihypertensive medicine, we defined moderate to severe office hypertension grade 1 as office SBP ≥ 150 mmHg or office DBP ≥ 90 mmHg, moderate to severe office hypertension grade 2 as office SBP ≥ 150 mmHg or office DBP ≥ 100 mmHg and moderate to severe office hypertension grade 3 as office SBP ≥ 160 mmHg or office DBP ≥ 100 mmHg.

With respect to cut-off points for ambulatory hypertension, we first applied the criteria proposed by the IDACO (International Database on Ambulatory blood pressure monitoring in relation to Cardiovascular Outcomes) investigators (10): mean 24-h SBP ≥ 130 mmHg or mean 24-h DBP ≥ 80 mmHg, irrespective of use of antihypertensive medicine. In addition, we made the following cut-off points to define prevalent moderate to severe ambulatory hypertension grade 1 and 2, respectively: mean 24-h SBP ≥ 140 mmHg or mean 24-h DBP ≥ 85 mmHg, and mean 24-h SBP ≥ 150 mmHg or mean 24-h DBP ≥ 90 mmHg.

Finally, we defined present of diabetes as use of antidiabetic drugs and/or a fasting plasma glucose ≥ 7.0 mmol/l.

Events for prospective substudy

Complete follow-up with respect to mortality was obtained from the Civil Registration System. Information on CVD mortality was obtained from the National Death Register. Information on hospitalizations was obtained from the National Health Register. The pre-specified primary event was the combination of CVD mortality (ICD-8 codes 390–448, or ICD-10 codes I00–I79), ischemic heart disease (ICD-8 codes 410–414, or ICD-10 codes I20–I25) and stroke (ICD-8 codes 431, 433 or 434, or ICD-10 codes I61 or I63). The pre-specified secondary events were ischemic heart disease, CVD mortality, acute myocardial infarction (ICD-8 code 410, or ICD-10 codes I21–I22) and stroke. The diagnoses of acute myocardial infarction and stroke have been validated in the National Patient Registry (11,12), as well as the causes of death in the National Death Register (13).

Statistical analyses

Analyses were performed with the Statistical Analysis System (SAS), version 9.1 (SAS Institute, Cary, North Carolina, USA). Baseline characteristics, presented as mean ± standard deviation (SD) or proportions in percent, were compared with t-tests for continuous variables and Fisher’s exact test for categorical variables. Multiple logistic regression models were constructed to compute odds ratio estimates, adjusted for age and sex and other variables of interest, for risk of prevalent hypertension based on office BP readings and 24-h ambulatory BP recordings. Cumulative incidence of the various CVD events was compared by Fisher’s exact test. In addition, survival was analyzed by the Cox proportional-hazard method, adjusted for age and sex. In our outcome analyses, when participants experienced multiple events, we considered only the first. All statistical tests were two-sided and significance was concluded for values of p < 0.05.

Results

Among the 3193 participants in Health2006 one had the A:A variant and 22 had the G:A variant. In the analyses of the Health 2006 participants presented below, we treated heterozygote and homozygote A variants as one group. Table I summarizes the general characteristics of the Health 2006 participants by fibrillin-1 genotypes. Compared with participants with the G:G variant, participants with the A:A/G:A variants were older, had a higher mean SBP and DBP, had a higher prevalence of hypertension and diabetes, and a larger proportion of them received medical treatment for hypertension.

Table I. General characteristics of fibrillin-1 genotypes in the Health 2006 population.

Variable	A:A/G:A genotypes (n = 23)	G:G genotype (n = 3170)	p
Age, years	55.5 ± 10.9	48.8 ± 13.0	0.014
Female sex, %	52.2	56.1	0.83
Total cholesterol, mmol/l	5.19 ± 1.20	5.10 ± 1.02	0.66
HDL cholesterol, mmol/l	1.49 ± 0.35	1.54 ± 0.43	0.56
Office systolic blood pressure, mmHg	141 ± 22	130 ± 18	0.0024
Office diastolic blood pressure, mmHg	89 ± 13	82 ± 11	0.0007
Pulse pressure, mmHg	52 ± 16	48 ± 12	0.15
Hypertension, %	60.9	36.8	0.028
Treatment of hypertension, %	26.1	13.0	0.11
Body mass index, kg/m^2	27.0 ± 5.1	25.8 ± 4.6	0.24
Diabetes, %	21.7	4.0	0.0021
Current smoking, %	17.4	22.9	0.63

Data are mean ± standard deviation or proportions in percent. HDL, high-density lipoprotein; hypertension, office systolic blood pressure ≥ 140 mmHg or office diastolic blood pressure ≥ 90 mmHg or use of blood pressure lowering drugs; diabetes, use of antidiabetic drugs or a fasting plasma glucose ≥ 7.0 mmol/l.

Among the 2408 participants in MONICA10 none had the A:A variant but18 had the G:A variant. Table II summarizes the general characteristics of the MONICA10 participants by fibrillin-1 genotypes. Although participants with the G:A variant had a 5-mmHg higher mean SBP than participants with the G:G variant, the difference between the two genotypes did not reach statistical significance (p = 0.24). Furthermore, there were no significant differences in any of the other variables listed between the two groups (p ≥ 0.24).

Table II. General characteristics of fibrillin-1 genotypes in the MONICA10 population.

Variable	G:A genotype (n = 18)	G:G genotype (n = 2390)	p
Age, years	56.6 ± 8.6	54.0 ± 10.6	0.29
Female sex, %	61.1	50.7	0.38
Total cholesterol, mmol/l	6.02 ± 0.85	6.16 ± 1.11	0.60
HDL cholesterol, mmol/l	1.52 ± 0.42	1.45 ± 0.42	0.47
Office systolic blood pressure, mmHg	134 ± 22	129 ± 19	0.24
Office diastolic blood pressure, mmHg	84 ± 11	82 ± 11	0.51
Pulse pressure, mmHg	50 ± 15	47 ± 13	0.26
Hypertension, %	44.4	36.2	0.47
Treatment of hypertension, %	11.1	8.2	0.66
Body mass index, kg/m^2	25.1 ± 4.1	25.9 ± 4.2	0.40
Diabetes, %	5.5	3.5	0.64
Current smoking, %	44.4	45.9	0.90

Data are mean ± standard deviation or proportions in percent. HDL, high-density lipoprotein; hypertension, office systolic blood pressure ≥ 140 mmHg or office diastolic blood pressure ≥ 90 mmHg or use of blood pressure lowering drugs; diabetes, use of antidiabetic drugs or a fasting plasma glucose ≥ 7.0 mmol/l.

The results of our logistic regression analyses are summarized in Tables III and IV. Table III shows the odds ratio estimates for risk of the various definitions of hypertension for the fibrillin-1 gene variants in Health 2006; first with no adjustments and then with successive adjustments for sex and age, and sex, age and diabetes, as the prevalence of diabetes varied significantly between the two gene variants in Health 2006. The odds ratio estimates for risk of hypertension for the A:A/G:A variants increased substantially, more than twofold, with increasing severity of hypertension definition. Furthermore, the odds ratio estimates were diminished by up to 30% by adjustments for sex and age, whereas further adjustment for diabetes had only minor effect, less than 9%. It is also seen that the risk of moderate to severe hypertension grade 2 and 3 for the A:A/G:A variants was highly statistically significant, p < 0.0001, even after adjustments. Finally, adjustment for body mass index instead of diabetes did not change the results (data not shown).

Table III. Odds ratio estimates (95% confidence intervals) for risk of hypertension for fibrillin-1 gene variants in the Health2006 study.

Variables	Proportion of fibrillin-1 genotypes: A:A/G:A vs G:G	Odds ratio (95% CI) unadjusted: A:A/G:A vs G:G	Odds ratio (95% CI) adjusted for sex and age: A:A/G:A vs G:G	Odds ratio (95% CI) adjusted for sex, age and diabetes: A:A/G:A vs G:G
Hypertension	60.9% (14/23) vs 36.8% (1168/3170)	2.67 (1.14–6.18)*	1.87 (0.73–4.79)	1.70 (0.65–4.47)
Uncontrolled hypertension	60.9% (14/23) vs 32.6% (1035/3170)	3.20 (1.38–7.42)^†	2.41 (0.95–6.07)	2.33 (0.92–5.91)
Moderate to severe hypertension, grade 1	60.9% (14/23) vs 26.5% (841/3170)	4.31 (1.86–9.99)^‡	3.47 (1.41–8.58)^†	3.51 (1.41–8.67)^†
Moderate to severe hypertension, grade 2	60.9% (14/23) vs 15.7% (498/3170)	8.35 (3.59–19.39)^§	7.31 (2.88–18.57)^§	7.22 (2.83–18.40)^§
Moderate to severe hypertension, grade 3	52.2% (12/23) vs 10.1% (321/3170)	9.68 (4.24–22.12)^§	8.12 (3.37–19.94)^§	8.01 (3.27–19.58)^§

CI, confidence interval; hypertension, office systolic blood pressure (SBP) ≥ 140 mmHg or office diastolic blood pressure (DBP) ≥ 90 mmHg or use of blood pressure lowering drugs; uncontrolled hypertension, office SBP ≥ 140 mmHg or office DBP ≥ 90 mmHg; moderate to severe hypertension grade 1, office SBP ≥ 150 mmHg or office DBP ≥ 90 mmHg; moderate to severe hypertension grade 2, office SBP ≥ 150 mmHg or office DBP ≥ 100 mmHg; moderate to severe hypertension grade 3, office SBP ≥ 160 mmHg or office DBP ≥ 100 mmHg; diabetes, use of antidiabetic drugs or a fasting plasma glucose ≥ 7.0 mmol/l. *p < 0.05; ^†p < 0.01; ^‡p < 0.001; ^§p < 0.0001.

Table IV shows the odds ratio estimates for risk of the various definitions of ambulatory hypertension for the fibrillin-1 gene variants in MONICA10; first with no adjustments and then with adjustments for sex and age. With respect to hypertension based on office BP recordings, none of the various definitions of hypertension were significantly different between the two gene variants (p ≥ 0.078), although the same trend was observed in MONICA10 as in Health 2006 (data not shown). Among the 1646 MONICA10 participants with 24-h ambulatory BP measurements, 16 had the G:A variant. The odds ratio estimates for risk of ambulatory hypertension for the G:A variant increased substantially, more than twofold, with increasing severity of ambulatory hypertension definition. It is also seen that adjustments for age and sex did not decrease the size of the odds ratio estimates, and that the risk of moderate to severe ambulatory hypertension grade 2 for the G:A variant was robustly statistically significant, p < 0.01.

Table IV. Odds ratio estimates (95% confidence intervals) for risk of ambulatory hypertension for fibrillin-1 gene variants in the MONICA10 study.

Variables	Proportion of fibrillin-1 genotypes: A:A/G:A vs G:G	Odds ratio (95% CI) unadjusted: A:A/G:A vs G:G	Odds ratio (95% CI) adjusted for sex and age: A:A/G:A vs G:G
Ambulatory hypertension	56.3% (9/16) vs 38.0% (620/1630)	2.09 (0.78–5.65)	2.28 (0.81–6.40)
Moderate to severe hypertension, grade 1	43.8% (7/16) vs 17.7% (289/1630)	3.61 (1.33–9.77)*	3.93 (1.36–11.4)*
Moderate to severe hypertension, grade 2	31.3% (5/16) vs 7.4% (120/1630)	5.73 (1.96–16.7)^†	6.54 (2.12–20.2)^†

CI, confidence interval; ambulatory hypertension, 24-h mean systolic blood pressure (SBP) ≥ 130 mmHg or 24-h mean diastolic blood pressure (DBP) ≥ 80 mmHg; moderate to severe ambulatory hypertension grade 1, 24-h mean SBP ≥ 140 mmHg or 24-h mean DBP ≥ 85 mmHg; moderate to severe ambulatory hypertension grade 2, 24-h mean SBP ≥ 150 mmHg or 24-h mean DBP ≥ 90 mmHg. *p < 0.02; ^†p < 0.01.

With respect to the prospective substudy to complement our main investigation of hypertension, we used data from the MONICA10 study. The median study duration, from baseline until follow-up on 31 December 2006, was 12.6 years (5th and 95th percentiles: 4.4–13.4 years). Cumulative incidences of the various events in the main MONICA10 study population, consisting of 2408 participants, are summarized in Table V. Participants with the G:A genotype had a significantly higher (p = 0.011) incidence of the combined CV event and of ischemic heart disease and CV mortality. Of note, none of the 2408 participants suffered from aortic dissection or rupture or was diagnosed with Marfan syndrome during follow-up.

Table V. Cumulative incidence of cardiovascular disease events by fibrillin-1 genotypes in the entire MONICA10 study population.

Event	G:A genotype	G:G genotype	p
Cardiovascular disease	33.3% (6/18)	13.0% (311/2390)	0.011
Ischemic heart disease	22.2% (4/18)	8.2% (197/2390)	0.033
Cardiovascular mortality	16.7% (3/18)	5.5% (131/2390)	0.039
Acute myocardial infarction	11.1% (2/18)	4.2% (100/2390)	0.15
Stroke	5.6% (1/18)	2.9% (69/2390)	0.50

The results of our Cox proportional hazard models, performed in the main MONICA10 study population, consisting of 2408 participants, are shown in Table VI. Adjusted for age and sex, participants with the G:A genotype had a significantly higher risk (p = 0.014) of the combined CV event and of ischemic heart disease and CV mortality. Further adjustments for baseline office SBP and DBP did not really change the relationships between the G:A variant and the various events (Table VI). Finally, if we restricted our Cox proportional hazards models to the 1646 MONICA10 participants with 24-h ambulatory BP measurements, the age and sex adjusted hazard ratio (95% confidence interval) for the combined CV event was 2.67 (1.09–6.53), p = 0.032, a relationship that was no longer significant when further adjustments were made for 24-h mean SBP and DBP (2.01 (0.81–4.96), p = 0.13).

Table VI. Hazard ratio estimates (95% confidence interval) for the fibrillin-1 G:A genotype for risk of cardiovascular disease events in the entire MONICA10 study population.

Events	Hazard ratio (95% CI)* G:A vs G:G genotype	p	Hazard ratio (95% CI)^† G:A vs G:G genotype	p
Cardiovascular disease	2.76 (1.23–6.19)	0.014	2.55 (1.131–5.77)	0.024
Cardiovascular mortality	3.53 (1.12–11.1)	0.031	3.38 (1.06–10.8)	0.039
Ischemic heart disease	2.81 (1.04–7.56)	0.041	2.45 (0.90–6.64)	0.079
Acute myocardial infarction	2.82 (0.70–11.5)	0.15	2.61 (0.6310.7)	0.18
Stroke	2.36 (0.33–17.0)	0.40	2.17 (0.30–15.9)	0.44

*Adjusted for age and sex. CI, confidence interval. ^†Adjusted for age, sex and baseline office systolic and diastolic blood pressure.

Additional analyses

We also investigated the differences in BP handled as a continuous variable between the fibrillin-1 gene variants. In all situations, participants with the A:A/G:A variants had higher BP levels. In Health 2006, adjusted for sex, age, diabetes and medical treatment for hypertension, the difference in mean SBP (95% confidence interval) was 6 mmHg (− 1 to 12), p = 0.072, and the difference in mean DBP was 6 (2–10) mmHg, p = 0.0057. In MONICA10, adjusted for sex, age and medical treatment for hypertension, the difference in 24-h mean SBP was 5 mmHg (− 1 to 11), p = 0.093, and the difference in 24-h mean DBP was 4 mmHg (0–8), p = 0.042.

Discussion

The principal finding of this investigation was that individuals with the G to A substitution at SNP 5 in intron 27 in the fibrillin-1 gene (rs11856553) had a remarkably increased risk of moderate to severe hypertension, based on office BP measurements in the Health 2006 study and based on 24-h ambulatory BP measurements in the MONICA10 study. Another important finding was that MONICA10 participants with the G:A variant also had an increased risk of CVD. We speculate that the cause of the increased CVD risk in participants with the G:A variant in MONICA10 may be moderate to severe hypertension.

This study is the first to explore the relationships between a fibrillin-1 gene mutation and risk of prevalent hypertension and risk of incident of CVD in a large population-based sample, including replication with respect to risk of prevalent hypertension. We initiated this study with the rs11856553 SNP in the fibrillin-1 gene, because mutations in the fibrillin-1 gene are believed to be responsible for the vascular damage, which in the end is the cause of Marfan syndrome (1,2), because Medley et al. (3) reported that the SNP 5 in intron 27 in the fibrillin-1 gene was associated with aortic stiffness and coronary artery disease severity in patients with ischemic heart disease, and because the role of mutations in the fibrillin-1 gene in more common CVD, including hypertension, is basically unknown in the general population. Initially, we thought that variation in the fibrillin-1 gene might affect BP though a change in vascular stiffness (3), as fibrillin-1 is a major component of the elastic fibers in the arterial wall (1,2).

The G to A substitution in intron 27 at SNP 5 in the fibrillin-1 gene is a rare genetic variant. In both health 2006 and in MONICA10 the prevalence of the A variant was ∼0.4%. Nevertheless, both in Health 2006 and MONICA10 it was possible to identify highly statistically significant and remarkably strong relationships between the G:A variant and a diagnosis of moderate to severe hypertension. Thus, even after adjustments for relevant confounders, odds ratios greater than 6 for risk of moderate to severe hypertension were found in both study populations. Therefore, the G to A substitution at SNP 5 in intron 27 in the fibrillin-1 gene appears to be a relevant yet rare marker of risk or may be a rare cause of moderate to severe hypertension in the general population. In this context, it is of note that rs11856553 was not identified as a susceptibility locus for hypertension by the previous genome-wide association studies (14); possibly, the rarity of the G:A variant (minor allele frequency < 0.4%), lack of accurate phenotyping by 24-h ambulatory BP recordings, and little attention to more severe degrees of hypertension in the various genome-wide association studies (14) may explain the discrepancy between the previous genome-wide association studies and our study.

So, if the G to A substitution at SNP 5 in intron 27 in the fibrillin-1 gene is a cause of moderate to severe hypertension, which mechanisms may be involved? In this situation, it is important to remember that the G to A substitution in question here is intronic, and to date no functional effects of this variant have been reported. So, whether the G:A genotype is indeed directly involved in the pathogenesis of hypertension and CVD, or the G:A genotype is a marker of the presence of another disease causing gene remains to be clarified. However, it is important to remember here that in a recently published cholesterol study it was demonstrated that common non-coding DNA variants identified by genome-wide association studies can directly contribute to clinical phenotypes (15). Furthermore, with respect to mechanisms we initially thought that variation in the fibrillin-1 gene might affect BP though a change in vascular stiffness (3). In this perspective, several studies in humans have examined the relationships between fibrillin-1 genotypes and estimates of large artery stiffness, such as pulse pressure, aortic impedance and aortic pulse wave velocity, but the results have overall been inconsistent with different results for the same fibrillin-1 gene mutations (3,16–18). In the present investigation, an increased pulse pressure was not a characteristic feature of participants with the G to A substitution. In the MONICA10 study population, we also have measurements of aortic pulse wave velocity (19), an index of aortic stiffness, but we found no relevant difference in this hemodynamic measurement between the G:A and G:G genotypes, either (data not shown), so other mechanisms than increased large artery stiffness may be involved in the possible BP-raising effect of the G:A substitution at SNP 5 in intron 27 in the fibrillin-1 gene found in this investigation. Finally, with respect to pathophysiology and large artery stiffness, it is noteworthy that the G:A variant appeared to be more strongly associated with DBP than with SBP, a finding also pointing towards mechanisms beyond large artery stiffness.

Limitations and strengths

With respect to limitations, we have to acknowledge that although our two study populations consisted of more than 3000 and 1600 participants, respectively, an even larger sample size would probably have been helpful, as the G:A variant is indeed a rare genetic variant.

With respect to strengths, it is a major strength of this investigation that it was possible to identify highly statistically significant and remarkably strong relationships between the G:A substitution at SNP 5 in intron 27 in the fibrillin-1 gene and risk of moderate to severe hypertension in two independent population-based studies, which were carried out more than 10 years apart. In this respect, we were able to comply with the recommended replication standards in genetic research. It is also a major strength of this investigation that we measured BP by 24-h BP monitoring in one of the two population-based studies, so our BP findings were clearly not the result of “white coat” hypertension. Finally, the fact that we could identify the G:A variant as a risk marker of CVD in our prospective substudy serves to inspire even more confidence in our findings overall that the G:A substitution at SNP 5 in intron 27 in the fibrillin-1 gene has something to do with hypertension, the intermediate CVD phenotype studied here.

Perspective

In this investigation, we found for the first time that a rare genetic variant in the fibrillin-1 gene had a remarkably strong effect on risk of moderate to severe hypertension. In general, primary or essential hypertension is believed to be the result of the actions of many different genes, probably all with relatively minor effects on BP levels, although rare genes with pronounced effects on BP and risk of hypertension, such as the monogenic or Mendelian forms of hypertension, have been identified (20). Our investigation raises the possibility that essential hypertension more often than expected may be the result of rare genetic variants, which have not yet been identified, with major effects on BP levels and risk of hypertension. Thus, overall our results support the common diseases multiple rare variant hypothesis (21,22). Consequently, as a supplement to genome wide association studies (14), it may be a good idea to select patients with moderate to severe hypertension and perform detailed phenotyping and then sequence relevant loci believed to contain functional and structural genes that are believed to be involve in BP regulation, the way it was done with the tyrosine hydroxylase gene, which encodes the rate-limiting enzyme in catecholamine biosynthesis (23).

Acknowledgements

This study was funded by grants from the Danish Heart Foundation, the Danish Medical Association Research Fund/Volten, the Danish Pharmaceutical Association, the Lundbeck Foundation and the Novo Nordisk Foundation.

Conflicts of interests and disclosures: None of the authors has any conflicts of interests or disclosures with respect to the content of this manuscript.