Study Population

Subjects were derived from the Asklepios Study, an extensively phenotyped population-representative random sample of 2,524 male and female volunteers, aged 35 to 55 years, from the Belgian communities of Erpe-Mere and Nieuwerkerken, free of clinically overt CVD at baseline. An in-depth description of the Asklepios Study protocol has been published.

Exclusion criteria were (1) clinical presence of atherosclerosis or atherothrombosis (symptomatic >50% stenosis, prior atherothrombotic event, revascularization), (2) major concomitant illness, (3) type 1 diabetes mellitus and type 2 diabetes mellitus if proven macrovasculopathy or significant renal impairment, (4) conditions precluding accurate hemodynamic assessment (atrial fibrillation, pregnancy), and (5) inability to provide informed consent. The study complied with the Declaration of Helsinki, the protocol was approved by the ethics committee of Ghent University Hospital, and all subjects gave written informed consent.

Participant Examination: Overview

After obtaining written informed consent, review of questionnaire data, and a 15-min rest, measurements included basic clinical data, blood sampling, echocardiographic examination, carotid and femoral artery ultrasound, and arterial tonometry. All measurements were performed by a single observer. Blood pressure (BP) was recorded using cuff-patient matched bilateral triplicate measurements on a sitting subject using a validated oscillometric device (Omron HEM-907; Omron, Kyoto, Japan). Physical activity was derived from questionnaire data (intensity and frequency averaged throughout the year). Subjects performing leisure time physical activity of ≥3.5 metabolic equivalents at least once per 2 weeks were classified as “active.” “Ever smokers” were current or ex-smokers. “Current smokers” were subjects smoking ≥1 cigarette per week. “Ex-smokers” were subjects who had stopped smoking (>7 days), with a cumulative exposure of ≥100 cigarettes.

Cardiac Imaging

All subjects underwent resting electrocardiographically gated transthoracic echocardiography (Vivid 7; GE Vingmed Ultrasound AS, Horten, Norway). LVIDd and LV internal dimension at end-systole were measured. These measurements were done in the minor and major axes. The minor axis was measured from a two-dimensional parasternal long-axis view at the level of the mitral valve tips. The major axis was measured from the apical four-chamber view from the visual apex to the middle of the line connecting the mitral valve insertions. Sphericity was defined as LVIDd _(minor) divided by LVIDd _(major) , expressed as a percentage. Further measurements included LV wall thickness (interventricular septum and PW), LV mass, ejection fraction, systolic (s′) and early diastolic (e′) mitral annular pulsed-wave tissue Doppler velocities, pulsed-wave Doppler early (E) and late (A) diastolic transmitral flow velocities, and the propagation velocity of the E wave. The latter was measured as the slope of the isovelocity line from the mitral tips to a position 4 cm distally into the left ventricle, using the first aliasing velocity. All analyses were performed offline by a single blinded, measurement-dedicated reader.

Carotid and brachial artery BP waveforms were obtained by applanation tonometry using a pen-type transducer (SPT 301; Millar Instruments, Houston, TX) and dedicated hardware and acquisition software. The brachial artery tonometric waveform, calibrated to the oscillometric systolic BP (SBP) and diastolic BP, was used to obtain a true mean BP, instead of an empirical estimate. Carotid artery tonometry readings were calibrated, assuming that the difference between mean BP and diastolic BP is constant along the arterial tree.

Biochemical Analyses

All subjects were fasting, had refrained from smoking for ≥6 hours, and were screened for intercurrent infection or inflammation before blood sampling (in which case blood sampling was postponed). Serum parameters were measured on a Modular P automated system (Roche Diagnostics GmbH, Mannheim, Germany), in an ISO-9002-certified reference laboratory. Impaired fasting glycemia denotes a fasting glucose level ≥100 and <126 mg/dL (diabetes). The coefficients of variation of all tests described above were <3.0%.

FH of CVD

FH of CVD was determined using the Asklepios Family History Questionnaire, designed specifically for this study. It provides data on the occurrence of CVD spanning 4 generations of a respondent’s family (parents, grandparents, siblings, and offspring; only genetic family). Because the participants had several days to complete the questionnaire, they could obtain additional information from family members. The study nurse reviewed the questionnaire together with the subject during the visit at the study center. The FH questionnaire was completed by 2,491 of 2,524 Asklepios subjects. For 2,151 of these 2,491 subjects (86.4%), all necessary information was available to evaluate FH. We excluded 340 subjects (13.6%) who could not be correctly classified because of insufficiently accurate knowledge of FH. For this study, we used two definitions of FH of CVD. FH by conventional guidelines-based definition (cFH) was the occurrence of premature CVD (<55 years for men and <65 years for women) in a first-degree relative. In addition, to obtain potentially higher sensitivity, we explored the effect of using an extended FH definition (eFH). The eFH definition also incorporates the number of affected relatives, late CVD (≥55 years for men, ≥65 years for women) and second-degree relatives (grandparents). In particular, we further divided the cFH subjects with negative FH into eFH low-risk (≤1 second-degree relative with late-onset CVD or no known CVD in any first-degree or second-degree relative) and eFH moderate-risk classes (≥1 first-degree relative with late-onset CVD or ≥2 second-degree relatives with late-onset CVD or 1 second-degree relative with premature CVD). The cFH-positive FH group was further complemented with subjects with ≥2 second-degree relatives with premature CVD to obtain the eFH high-risk group.

In sensitivity analyses, we also used FH of hypertension, which was first defined as having one first-degree relative (parents or siblings) with hypertension and in a second analysis as having two first-degree relatives with hypertension.

To further explore the intertwined relationship among FH, cardiovascular risk factors, and echocardiographic findings, we also reversed the question described above and sought to determine whether echocardiographic findings provided incremental information toward a FH that identifiable risk factors did not provide.

Finally, in the light of maternal-fetal-placental studies, because we had access to birth weights in 1,802 subjects, we tested whether there was an effect of low birth weight (<2,500 g ) on LV function or structure.

Statistical Analyses

Statistical analysis was performed using SPSS version 19.0 (SPSS, Inc., Chicago, IL). First, we assessed the associations (unadjusted) between the different FH categories and LV functional and structural parameters. Second, we used a general linear model including confounders (which could be both related to FH and are known to be related to echocardiographic findings)—age, sex, height, weight, central SBP, drug-treated hypertension, heart rate, physical activity, and glycemia—to evaluate the association between FH and LV function and structure over and above these confounders.

Given that LV mass is strongly related to body height and weight, which are also heritable, we assessed the association between FH of CVD and LV mass index, normalized for height using appropriate allometric powers derived from this population as previously described.

To determine whether echocardiographic findings provide incremental information toward the presence of a FH that identifiable risk factors do not provide, we performed an exploratory analysis using FH as the outcome variable. First, we did a binary logistic regression analysis (entry P = .10, removal P = .15, forward conditional). The first block included traditional cardiovascular risk factors (age, sex, height, weight, central SBP, drug-treated hypertension, heart rate, physical activity, and glycemia). In the second block, we added the echocardiographic parameters. In a second (backward) binary logistic regression analysis, we included all risk factors and all echocardiographic parameters in one block and repeatedly removed the strongest predictive parameter. In a third binary logistic regression analysis, we included all risk factors and only one echocardiographic parameter in one block and again repeatedly removed the strongest predictive parameter. Finally, we performed a general linear model analysis, first by only including echocardiographic parameters in the model and then by adding classical risk factors.

All parameters are reported as mean ± SD (all parameters were checked for normality). P values < .05 were considered to indicate statistical significance.