Subcostal ecocardiography for the evaluation of left ventricular hypertrophy in older populations.
Background: Left ventricle (LV) hypertrophy is a major risk factor for mortality and morbidity. Unfortunately, the echocardiographic evaluation of LV mass from the parasternal (PS) approach is frequently limited in older, obese or emphysematous individuals. This poses a selection bias to the epidemiological assessment of LV hypertrophy. Methods: In order to assess whether the subcostal (SC) view may reliably substitute the PS view to evaluate LV mass in older adults, 614 unselected persons aged ≥65 years participating in a population-based survey on heart failure in the elderly, and 54 randomly selected healthy individuals 35-64 years old, underwent a complete echocardiographic examination including M-mode PS and SC scanning. Results: Overall, both PS and SC views were available for LV mass estimate (Devereux’s formula) in 314 subjects(age: 69.9±0.5 years, mean±SEM) with normal LV kinetics and sinus rhythm. The SC view underestimated (p<0.001) LV diastolic dimension (LVDD) and overestimated (p<0.012) free wall thickness (FWT), whereas septum thickness was similar in the two views. As a consequence, LV mass was underestimated from the SC view (76.2±1.2 g/m 2 ) as compared to the PS view (87.9±1.4 g/m 2 ; p<0.001). The SC view was 99% specific but only 25% sensitive for the diagnosis of LV hypertrophy (LV mass >116 and >104 g/m 2 in men and women, respectively), based on the PS view. When age, body weight and gender were entered in multivariate regression models with the SC measures, the first two variables significantly contributed to the prediction of both LVDD and FWT, and gender only to the prediction of LVDD. LV mass was recalculated using measures from the SC view, corrected on the basis of the regression equations. The corrected value did not differ from that obtained from the PS view and the sensitivity of the SC view for the diagnosis of LV hypertrophy increased to 51% with unmodified specificity (98%). Conclusions: The SC view significantly underestimates LV mass. However, correction for simple demographic and anthropometric variables greatly improves its reliability for estimating LV mass and hypertrophy in unselected older adults.