Background

Left ventricular (LV) volume and mass have prognostic relevance. Overall size of the left ventricle as it appears in noncontrast CT is a composite of the ventricular volume and myocardial mass. We describe a method to estimate the LV size using a single cross-section in noncontrast CT and determined normal ranges on the basis of a large population cohort.

Methods

The Multi-Ethnic Study of Atherosclerosis with 6814 participants from 4 ethnicities who were free of known cardiovascular disease and enrolled between 2000 and 2002 form the basis of our analysis. LV size was calculated from a single cross-sectional slice obtained by either nonenhanced electron beam or multidetector CT. LV size was adjusted to body surface area to obtain the LV size index, which was adjusted for age, sex, race or ethnicity, hypertension, hyperlipidemia, and diabetes.

Results

There were significant differences in LV size index by race which were further influenced by age and sex. Higher values were noted in men in all ethnic groups across all age groups. Similarly, LV size index uniformly decreased with age across all ethnic and sex categories. Caucasians had the lowest and African Americans had the highest LV size index across all age and sex categories. In multivariate regression analyses adjusted for age, sex, race or ethnicity, hypertension, hyperlipidemia, smoking, and diabetes mellitus, the significant differences were noted between male vs female (median difference, 17.5 cc/m(2); P < .001), ethnic groups (Caucasian, reference group; Asian, 3.7 cc/m(2); African American, 8.3 cc/m(2); and Hispanic, 5.6 cc/m(2); P < .001), and age groups (45-54 years, reference group; 55-64 years, -5.2 cc/m(2); 65-74 years, -11.4 cc/m(2); and 74-84 years, -12.5 cc/m(2)).

Conclusions

This study provides normative values for LV size as determined from a single, nonenhanced CT cross-section and indexed to body surface area, and it demonstrates that the LV size index varies by age, sex, and ethnic background.

Reference limits for echocardiographic quantification of cardiac chambers in Hispanics are not well studied. We examined the reference values of left atrium and left ventricle (LV) structure in a large ethnically diverse Hispanic cohort. Two-dimensional transthoracic echocardiography was performed in 1818 participants of the Echocardiographic Study of Latinos (ECHO-SOL). Individuals with body mass index ≥30 kg/m(2), hypertension, diabetes mellitus, coronary artery disease, and atrial fibrillation were excluded leaving 525 participants defined as healthy reference cohort. We estimated 95th weighted percentiles of LV end systolic volume, LV end diastolic volume, relative wall and septal thickness, LV mass, and left atrial volume. We then used upper reference limits of the 2005 and 2015 American Society of Echocardiography (ASE) and 95th percentile of reference cohort to classify the Hispanic Community Health Study/Study of Latinos (HCHS/SOL) target population into abnormal and normal. Reference limits were also calculated for each of 6 Hispanic origins. Using ASE 2015 defined reference values, we categorized 7%, 21%, 57%, and 17% of men and 18%, 29%, 60%, and 26% of women as having abnormal LV mass index, relative, septal, and posterior wall thickness, respectively. Conversely, 10% and 11% of men and 4% and 2% of women were classified as having abnormal end-diastolic volume and internal diameter by ASE 2015 cutoffs, respectively. Similar differences were found when we used 2005 ASE cutoffs. Several differences were noted in distribution of cardiac structure and volumes among various Hispanic/Latino origins. Cubans had highest values of echocardiographic measures, and Central Americans had the lowest. This is the first large study that provides normal reference values for cardiac structure. It further demonstrates that a considerable segment of Hispanic/Latinos residing in the United States may be classified as having abnormal measures of cardiac chambers when 2015 and 2005 ASE reference cutoffs are used.