FTO Is Associated with Aortic Valve Stenosis in a Gender Specific Manner of Heterozygote Advantage: A Population-Based Case-Control Study

Background Single nucleotide polymorphisms (SNPs) within the Fat mass and obesity associated (FTO) gene have been linked with increased body weight. However, the data on an association of FTO with cardiovascular diseases remains conflicting. Therefore, we ascertained whether FTO is associated with aortic valve stenosis (AVS), one of the most frequent cardiovascular diseases in the Western world. Methods and Findings In this population-based case-control study the FTO SNP rs9939609 was analyzed in 300 German patients with AVS and 429 German controls of the KORA survey S4, representing a random population. Blood samples were collected prior to aortic valve replacement in AVS cases and FTO rs9939609 was genotyped via ARMS-PCR. Genotype frequencies differed significantly between AVS cases and KORA controls (p = 0.004). Separate gender-analyses uncovered an association of FTO with AVS exclusively in males; homozygote carriers for the risk-allele (A) had a higher risk to develop AVS (p = 0.017, odds ratio (OR) 1.727; 95% confidence interval (CI) 1.087–2.747, recessive model), whereas heterozygote carriers for the risk-allele showed a lower risk (p = 0.002, OR 0.565, 95% CI 0.384–0.828, overdominant model). After adjustment for multiple co-variables, the odds ratios of heterozygotes remained significant for an association with AVS (p = 0.008, OR 0.565, 95% CI 0.369–0.861). Conclusions This study revealed an association of FTO rs9939609 with AVS. Furthermore, this association was restricted to men, with heterozygotes having a significantly lower chance to develop AVS. Lastly, the association between FTO and AVS was independent of BMI and other variables such as diabetes mellitus.


Pilot Study
In a pilot study, 71 pilot AVS cases and 442 pilot controls of a previously published control group [1] were analyzed. An association of FTO rs1121980 (in linkage with rs9939609) with AVS of heterozygote advantage was apparent applying the co-dominant model. Based on these data the required sample size for a study power of 80% was calculated with χ2 test by comparing genotype frequencies of AVS pilot cases with pilot controls under specification of number of participants (G*Power, www.gpower.hhu.de). A sample size of 255 AVS cases was calculated.
As recent publications focused on FTO rs9939609, we also performed the initial study for rs9939609. No further analysis were performed in the pilot study. Data of the pilot study (pilot AVS cases and pilot controls) have not been used in the current study.

Screening and identification of patients with AVS
All patients admitted to the Department of Cardiovascular Surgery at the University Hospital Düsseldorf for operative therapy of aortic valve disease were screened for inclusion in the current study. All patients included in this study were enrolled in a prospective manner. Correct inclusion of patients was founded on three independent screening columns: 1) All patients underwent a thorough evaluation by transthoracic echocardiography (TTE) with or without additional transesophageal echocardiography (TEE) to ensure the diagnosis of AVS prior to admission to the hospital for cardiac surgery. In addition to the echocardiography and as a routine procedure, all patients over 50 years old underwent left heart catheterization and selective coronary angiography to look for -3 -relevant coronary artery disease. In a prospective manner medical records and charts of patients admitted for cardiac surgery were screened by P.A. and A.L. for diagnosed AVS. Documented history of rheumatic fever as derived from the medical records was used to diagnose rheumatic AVS.
2) As a second independent diagnostic step, all patients undergoing any procedure for any type of valvular heart disease received an intraoperative TEE, which a) confirms the primary pathology, b) rules out any other undetected abnormality, and c) confirms the procedural success at the end of the operation. The intraoperative TEE was performed and evaluated by a senior attending anesthesiologist, primarily involved in the section of cardiac anesthesia and routinely performing TEEs with specific focus on assessment of valve function. Patients with intraoperative TEE finding diverting from the preoperatively described lesions were excluded from this study.
3) As a third column for the diagnosis of AVS, the macroscopic aspect of the aortic valve was evaluated by the senior surgeon. The latter step was furthermore used to diagnose the presence of a bicuspid aortic valve.
The assessment of the senior surgeon involved in the operation was used to further confirm the diagnosis of AVS.

Acquisition of clinical variables
KORA data acquisition has been previously described in detail [2,3]. For the current study, a data request has been submitted to KORA-gen for the variables: residence in Germany, age at survey, sex, body weight, body length, total cholesterol, HDL, LDL, triglycerides, diabetes mellitus, diastolic and systolic blood pressure, peripheral artery disease and smoking status.
For AVS cases, data points were prospectively collected during the entire study period and in the same or similar fashion as described for KORA controls. In brief, the variables: country of birth, age and sex were registered at the time of hospital admission and rely on personal identification documents. Gender was

Genotyping
Genotyping SNPs via tetra-primer ARMS-PCR allows identification of two alleles in one PCR reaction.