Study Design and Participants

The Coronary Artery Risk Development in Young Adults (CARDIA) study design and population characteristics have been previously described. Briefly, CARDIA is a prospective cohort study designed to evaluate development and progression of coronary disease risk factors in young adults. Initially, 5,115 healthy black and white men and women aged 18 to 30 years (1985–1986) were enrolled and examined at four field centers in Birmingham, Alabama; Chicago, Illinois; Minneapolis, Minnesota; and Oakland, California. Of 3,498 participants attending the year 25 (2010–2011) examination, 3,474 underwent echocardiography, including 2D STE. This study involved a retrospective review of prospectively collected images in a total of 557 CARDIA participants who remained healthy at the year 25 evaluation. The institutional review board at each site approved the study protocol. Written informed consent was obtained from all participants. The work was conducted in accordance with the Declaration of Helsinki.

Echocardiographic Acquisition

Transthoracic echocardiography was performed using a commercially available ultrasound system (Artida; Toshiba Medical Systems, Tokyo, Japan) with a 1.8- to 4.2-MHz phased-array transducer. Images were acquired by experienced sonographers, trained centrally and certified for standardized protocols across all four field centers. Two-dimensional images were obtained from apical four-chamber and two-chamber views, as well as from parasternal long-axis and short-axis views. Image acquisition was optimized by adjusting gain, compression, and transducer frequency. Sector width and image depth were adjusted to ensure that the entire left ventricle could be visualized throughout a complete cardiac cycle. Focus depth was positioned at the LV midcavity. At frame rates ranging from 40 to 90 frames/sec, at least two cardiac cycles were acquired from each view. M-mode and Doppler-derived methods were performed for complete a conventional echocardiographic evaluation. The examinations were recorded in Digital Imaging and Communications in Medicine standard, as well as in vendor-specific format, and transmitted to the year 25 echocardiographic reading center at Johns Hopkins University.

Echocardiographic Analysis

All analyses were performed offline by four experienced readers, trained centrally and certified according to predefined analysis protocol. Conventional echocardiographic measures were evaluated according to previously published guidelines, using commercially available DigiView version 3.7.7 (Digisonics, Houston, TX). LV ejection fraction was estimated using the Simpson biplane method.

LV 2D STE was performed using the validated Advanced Cardiology Package 2D Wall Motion Tracking version 3.0 (Toshiba Medical Systems). Longitudinal deformation was assessed from apical four-chamber and two-chamber views; circumferential deformation and radial deformation were assessed from the parasternal short-axis view at the midventricular level ( Figure 1 ). A region of interest was automatically defined after manual selection of landmarks in the endocardial border. If necessary, further manual adjustments of region-of-interest thickness and endocardial and epicardial borders were made. Finally, myocardial tracking was obtained automatically. If needed, borders were edited or retraced. End-diastole was considered as the beginning of the QRS segment, while end-systole was defined as the lowest LV volume.

Two-dimensional STE. Longitudinal strain analyzed in apical four-chamber view (A) and apical two-chamber view (B) and circumferential strain evaluated in parasternal short-axis view (C) .

Strain was calculated as the peak systolic change in the segment length relative to the length at end-diastole and represented as a percentage. Strain rate is defined as the deformation ratio, estimated as the temporal derivative of strain. Strain rate parameters were analyzed at systolic peak, early diastolic peak, and late diastolic peak, and represented as deformation per second. Both strain and strain rate are presented as global and segmental measurements. Each echocardiographic view was divided into a six-segment model; global values were defined as the average of segmental peaks. Peak systolic strain was also analyzed at endocardial, midwall, and epicardial layers.

Longitudinal and circumferential strain and strain rate parameters reflect shortening, so during systole, more negative values represent enhanced deformation, and during diastole, enhanced deformation rate is represented by more positive values. Radial strain and strain rate parameters reflect thickening, so during systole, more positive values represent enhanced deformation, and during diastole, enhanced deformation rate is represented by more negative values ( Figure 2 ).

Global LV strain and strain rate curves. Strain was measured as the maximum systolic deformation ( yellow points ); strain rate was assessed at systolic peak ( orange points ), early-diastolic peak ( blue points ), and late-diastolic peak ( purple points ).

Echocardiographic Quality Control and Reproducibility

Quality control and reproducibility of echocardiographic measurements in CARDIA year 25 examinations have been published. Briefly, segments unsuitable for myocardial tracking were excluded. Echocardiographic views with more than three segments excluded were not considered for global measurements. Moreover, 2D images were classified as optimal if scored as good or excellent or suboptimal if scored as fair or poor, according to the 2D spatial resolution. Acquisition and reading procedures were highly reproducible. Inter- and intrareader reproducibility of speckle-tracking-derived variables were assessed for all four analysts in subsets of 40 and 160 images, respectively. Intraclass correlation coefficient and residual coefficient of variation on the basis of a linear mixed model were used to assess variability. The intraclass correlation coefficients and coefficients of variation of longitudinal apical four- and two-chamber, circumferential, and radial strains were 0.55 (10.4%), 0.71 (10.7%), 0.67 (12.9%), and 0.84 (15%), respectively, for interreader reproducibility and 0.79 (6.6%), 0.87 (5.5%), 0.81 (6.8%), and 0.89 (12.1%) for intrareader reproducibility.

Healthy Subgroup

A healthy subgroup was established from CARDIA participants who underwent year 25 follow-up. For this purpose, individuals with histories of cardiovascular disease or those with hypertension, dyslipidemia, diabetes, obesity, former and current smoking, subclinical atherosclerosis, cancer, hyperthyroidism, or human immunodeficiency virus infection identified in CARDIA year 25 examinations were excluded. Cardiovascular disease was defined as the presence of one or more of the following: history of myocardial infarction, coronary revascularization, cardiomyopathy, congestive heart failure, any aortic stenosis or other cardiac valve disease greater than mild, or pulmonary hypertension. Hypertension was defined as blood pressure ≥ 140/90 mm Hg or use of antihypertensive medications; diabetes mellitus was defined as a combination of one or more of the following: fasting plasma glucose ≥ 126 mg/dL, 2-hour oral glucose tolerance test ≥ 200 mg/dL, glycated hemoglobin ≥ 6.5%, or use of antidiabetes medications; dyslipidemia was defined as triglycerides ≥ 150 mg/dL, high-density lipoprotein cholesterol < 40 mg/dL for men and high-density lipoprotein cholesterol < 50 mg/dL for women; and obesity was defined as body mass index (BMI) ≥ 30 m ² /kg. Subclinical atherosclerosis was considered as coronary artery calcium score > 0 or mean common carotid intima-media thickness > 1.0 mm. Calcium score and intima-media thickness were measured in 85% and 99% of the healthy population, respectively. Only nonpregnant women were included in the healthy subgroup. Individuals excluded from the apparently healthy subgroup were included in the subgroup with known cardiovascular risk factors and other diseases. The most frequent exclusion criterion was obesity (43.5%), followed by smoking history (38.3%), hypertension (33.0%), subclinical atherosclerosis (28.0%), dyslipidemia (16.9%), diabetes mellitus (10.4%), cancer (7.5%), cardiovascular disease (5.8%), hyperthyroidism (1.7%), human immunodeficiency virus infection (0.9%), and pregnancy (0.1%). After application of exclusion criteria, the entire healthy population was composed of 557 individuals.

Statistical Analysis

Baseline characteristic comparisons were evaluated using Student’s t test, the Wilcoxon rank sum test, and the χ ² test as appropriate. Reference ranges for deformation parameters are described as mean and standard deviation and 95% prediction intervals. Normality of the data was verified using the Shapiro-Wilk test and histograms. Because of skewed distribution, radial early diastolic strain rate and all late diastolic strain rate measurements were log-transformed to calculate prediction intervals. One-way analysis of variance was used to verify differences among race and sex categories, with the Tukey test for post hoc multiple comparisons. Multivariate linear regression was used to assess the association between strain and strain rate (as a dependent variable) in the healthy subgroup, and demographics (age, sex, and race), educational years, anthropometric and hemodynamic variables (BMI, heart rate, and systolic blood pressure), and technical parameters (frame rate and spatial resolution of 2D images). The variance inflation factor was used to evaluate collinearity, with a mean value of 1.15 (ranging from 1.01 to 1.29) reflecting absence of significant multicollinearity. All tests were two tailed, and differences were considered statistically significant at P < .05. All statistical analyses were done using Stata version 12.1 (StataCorp, College Station, TX).

Demographics and Speckle-Tracking Feasibility

Baseline characteristics of the participants are listed in Table 1 . The healthy subgroup included 557 individuals with a mean age of 49.6 ± 3.6 years, 61.6% women, and 69.5% whites. In a multiple comparison, white men and black men were taller (178.4 ± 6.4 and 177.0 ± 7.2 cm, respectively) than white women and black women (165.5 ± 6.3 and 165.4 ± 6.8 cm, respectively; P < .001). Overweight (BMI ≥ 25 kg/m ² and <30 kg/m ² ) was identified in 254 individuals (45.6%). White women showed lower systolic blood pressure compared with black men ( P < .05). The prevalence of exclusion criteria is detailed in the Appendix ( Supplemental Table 1 , available at www.onlinejase.com ). Speckle-tracking feasibility in the healthy subgroup was high; global parameters for apical four-chamber, apical two-chamber, and parasternal short-axis views were defined respectively in 91.4% ( n = 509), 70.2% ( n = 391), and 93.5% ( n = 521) of the echocardiograms. Unsuitable tracking was more frequent in the anterior and anterolateral segments, especially at the mid and apical levels ( Figure 3 ). The mean frame rates for apical four-chamber, apical two-chamber, and parasternal short-axis views were, respectively, 46.3 ± 1.6, 46.5 ± 2.3, and 46.8 ± 3.4 frames/sec. In the healthy population, strain rate was highly feasible given frame rates < 50 frames/sec in the apical four-chamber view (99.6%), apical two-chamber view (99.4%), and short-axis view (100%).

Segmental tracking feasibility. Number of segments analyzed. AA , Apical anterior; AI , apical inferior; AL , apical anterolateral; Ant , anterior; AS , apical inferoseptal; BA , basal anterior; BI , basal inferior; BL , basal anterolateral; BS , basal anteroseptal; Inf , inferior; Lat , lateral; MA , mid anterior; MI , mid inferior; ML , mid anterolateral; MS , mid inferoseptal; Pos , posterior; SA , anteroseptal; SI , inferoseptal.

Reference Ranges and Regional Patterns of LV Deformation

Global strain and strain rate reference ranges for the left ventricle are presented in Table 2 . Mean values for longitudinal strain in the apical four- and two-chamber views were −15.9% and −16.6%, whereas circumferential and radial peak systolic strain values were −15.7% and 36.6%, respectively.

Circumferential peak systolic shortening significantly increased from the epicardium (−11.7± 2.2%) toward the endocardium (−23.2 ± 3.6%), and a similar pattern was found for radial thickening, increasing from the epicardium (31.1 ± 9.7%) toward the endocardium (43.8 ± 14.3%; P < .001 for all). For longitudinal shortening in the apical four- and two-chamber views, the increasing gradient from epicardium to endocardium was significant in the apex (−13.8 ± 3.7% to −19.5 ± 4.8% in the four-chamber view; −13.7 ± 3.6% to −19.5 ± 5.2% in the two-chamber view; P < .001 for all) but not significant in the mid and basal LV regions, as illustrated in Figure 4 .

Analysis of strain according to myocardial layers. Outer, mid, and inner segments correspond respectively to epicardial, midventricular, and endocardial layers. A , Anterior; AL , anterolateral; AS , anteroseptal; I , inferior; IL , inferolateral; IS , inferoseptal.

Longitudinal shortening increased significantly from the base toward the apex in inferior (−16.0 ± 4.4% to −20.0 ± 5.3%) and inferoseptal (−12.3 ± 2.9% to −20.3 ± 4.9%) regions ( P < .001 for all) and, in contrast, decreased from the base toward the apex in anterior (−16.4 ± 4.6% to −10.3 ± 4.1%) and anterolateral (−15.5 ± 5.1% to −11.7 ± 4.2%) walls ( P < .001 for all).

Reference values for regional strain in the healthy are presented in the appendix ( Supplemental Table 2 , available at www.onlinejase.com ).

Race and Sex Differences

Race and sex differences are reported in Table 3 . White women showed significantly better systolic deformation than both white and black men, as demonstrated by longitudinal and circumferential systolic strain and strain rate ( P < .05 for all). The lowest values of systolic deformation were found in black men, especially for longitudinal systolic strain, with lower values than all the other race and sex categories ( P < .05). Regarding diastolic deformation, race and sex differences were statistically significant for early diastolic strain rate parameters, but not for late diastolic strain rate. Women showed significant better early diastolic deformation than men, and black men exhibited the lowest values of early diastolic deformation, as evaluated by longitudinal early diastolic strain rate ( P < .05). For systolic and diastolic deformation, black women and white men showed intermediate values, usually reflecting lower deformation than white women and higher deformation than black men.

Correlates of Strain and Strain Rate in the Healthy Subgroup

Sex demonstrated the highest correlation with longitudinal strain according to standardized β coefficient (female sex was associated with better systolic deformation than male sex, P < .001; Table 4 ). Race was also independently related to longitudinal strain (white race was associated with better deformation than black race, P < .05). Advancing age was related to more negative longitudinal early diastolic strain rate, indicative of lower diastolic function. BMI within the nonobese range was associated with more negative longitudinal and circumferential early diastolic strain rate and thus lower diastolic function. Higher heart rate was related to more positive longitudinal strain and more negative radial strain, indicating lower systolic function. Higher systolic blood pressure, within the normal and prehypertensive ranges, was associated with more positive longitudinal strain as well as more negative radial strain, reflecting lower longitudinal deformation and higher thickening.

Table 4

Correlates of strain and early diastolic strain rate in the healthy population

R 2	Peak systolic strain
	Longitudinal four-chamber				Longitudinal two-chamber				Circumferential				Radial
	0.13				0.19				0.05				0.04
Constant term ∗	−22				−22				−16				18
	β	SE	S-β		β	SE	S-β		β	SE	S-β		β	SE	S-β
Frame rate	−0.002	0.060	−0.001		−0.034	0.049	−0.032		0.001	0.034	0.001		−0.281	0.140	−0.088	†
Suboptimal image (vs optimal)	0.624	0.222	0.121	†	0.669	0.285	0.113	†	0.597	0.286	0.091	†	−1.042	1.192	−0.038
Age	0.028	0.027	0.046		0.025	0.033	0.039		−0.015	0.033	−0.021		0.056	0.138	0.019
Female (vs male)	−0.948	0.207	−0.209	‡	−1.200	0.245	−0.250	‡	−0.812	0.252	−0.150	†	0.274	1.049	0.012
White (vs black)	−0.593	0.233	−0.122	†	−0.978	0.276	−0.188	‡	−0.261	0.283	−0.045		−1.377	1.180	−0.058
Educational years	0.002	0.041	0.002		−0.012	0.049	−0.013		−0.003	0.050	−0.002		0.363	0.209	0.082
BMI	0.021	0.036	0.028		0.054	0.045	0.063		0.007	0.043	0.008		0.056	0.180	0.015
Heart rate	0.037	0.010	0.153	‡	0.058	0.012	0.226	‡	0.009	0.013	0.033		−0.129	0.052	−0.109	†
Systolic blood pressure	0.023	0.009	0.120	†	0.015	0.010	0.073		0.014	0.011	0.060		0.107	0.046	0.112	†

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Ascending Aortic Dimension Measurements: Apples, Oranges, and Lemons
2. Diagnostic Accuracy of Transthoracic Echocardiography for Infective Endocarditis Findings Using Transesophageal Echocardiography as the Reference Standard: A Meta-Analysis
3. The Effect of the Superior Cavopulmonary Anastomosis on Ventricular Remodeling in Infants with Single Ventricle
4. Evaluating Appropriate Use of Pediatric Echocardiograms for Chest Pain in Outpatient Clinics

Stay updated, free articles. Join our Telegram channel

Join