Study Population

The study population consisted of voluntary flight academy candidates aged 17 to 29 years, who were all previously screened to exclude significant medical conditions, including cardiovascular and respiratory diseases. All candidates with BMIs > 30 kg/m ² are disqualified from the flight academy and thus were excluded from the study. Candidates were evaluated by an examiner at the IAF Aero Medical Center (AMC). Medical history, physical examination, and a 12-lead resting electrocardiogram were obtained for all candidates. Candidates who had normal results on medical evaluation and completed an additional nonmedical screening process underwent transthoracic echocardiography at the AMC. Only those with normal results on echocardiography were included in the study. The study was performed retrospectively on the basis of the routine examinations performed in air force academy candidates. The results recorded were reviewed by a technician to ensure that no abnormal values were present. If abnormal values were identified, the test was reviewed, and the correct data were recorded. All data were also reviewed by a statistician, and again any values identified as abnormal were reviewed through the original medical records.

Echocardiographic Studies

We reviewed all routine echocardiographic studies of aircrew candidates performed at the AMC between January 1994 and January 2010. Studies were performed whether abnormal clinical findings (e.g., on physical examination or electrocardiography) were present or not.

All echocardiographic studies were obtained with one of three devices (HP 500 Sonos, ATL 5000HDI [Palo Alto, CA], or Philips HD 11 XE [Amsterdam, The Netherlands]). Second-generation devices (HP 500 Sonos and ATL 5000HDI) were used from 1994 to 2008. A third-generation device (Philips HD 11 XE) was used from 2008 onward. All studies were performed by one of three experienced sonographers and interpreted by one of two cardiologists specialized in echocardiography. Transthoracic echocardiography included two-dimensional, M-mode, and Doppler studies according to standard American Society for Echocardiography guidelines for obtaining images, quantification of chamber dimensions, and assessment of valvular regurgitation. All studies performed at the IAF AMC are performed in four windows. The first and second are the left and right parasternal long-axis views. The third window is the parasternal short-axis view, and the fourth is the apical four-chamber view. In cases in which elevated PASP or TR is suspected on the basis of these four windows, a subcostal window is added. All TR gradients were measured during the inspiratory phase. No saline injection was used in an attempt to enhance weak signals. Measurements were corrected for body surface area. Left ventricular mass was estimated as 1.05 × {[left ventricular end-diastolic diameter (mm) + posterior wall thickness (mm) + intraventricular septum (mm)] ³ − [left ventricular end-diastolic diameter] ³ }/1,000. Right ventricular systolic pressure is equivalent to systolic pulmonary artery pressure. The pressure gradient across the tricuspid valve was derived from tricuspid systolic regurgitant flow velocity, measured by continuous-wave Doppler, using the simplified Bernoulli equation. Right atrial pressure was estimated as 5 mm Hg. Thus, the right ventricular systolic pressure estimation equaled the tricuspid pressure gradient + 5 mm Hg. Right ventricular systolic pressure is considered to be equal to PASP.

Tests in which cardiac abnormalities were identified (with the exception of TR or elevated pulmonary artery pressure) or in the absence of measurable regurgitant tricuspid flow were excluded from the study.

The study was approved by the ethics committee of the Medical Corps of the Israel Defense Force (IDF-685-2007).

Statistical Analysis

Continuous variables are presented as mean ± SD if normally distributed or as median (interquartile range) otherwise. Spearman’s correlation analysis was used to assess the univariate correlations of PASP with age, blood pressure, and echocardiographic parameters. Multivariate linear regression analyses were performed to assess the independent contribution of age, blood pressure, and echocardiographic measures to PASP variance. The selection of variables for inclusion in the model was based on clinical judgment and univariate correlations detected. The assumption of normality was tested by evaluation of the distribution of the model residuals. Variance inflation factors were examined to detect possible collinearity between variables in the model. Variance inflation factors did not exceed 3 for all variables in each model, indicating lack of collinearity. Data were analyzed using SAS version 9.2 (SAS Institute Inc., Cary, NC).