Study Design and Participants

The selection criteria and study design of the CARDIA study have been described in detail elsewhere. Briefly, CARDIA is a multicenter, biracial cohort study that enrolled 5,115 men and women from four US field centers (Birmingham, Alabama; Oakland, California; Chicago, Illinois; and Minneapolis, Minnesota), aged 18 to 30 years at baseline (1985–1986) with 25 years of follow-up in seven subsequent visits up until 2010–2011 (year 25 examination). For this study, we evaluated 3,239 CARDIA participants with echocardiograms obtained 20 years apart, 1990–1991 (year 5, baseline) and 2010–2011 (year 25, follow-up). One hundred seventy participants were excluded because of suboptimal images for AoD assessment. Participants with clinical cardiovascular disease events ( n = 93) and moderate or severe valvular dysfunction ( n = 43) were excluded. The final analytic cohort had 2,933 participants (91% of total year 5 and year 25 participants). All study participants gave written informed consent, and the institutional review board of each participating institution approved study.

Echocardiographic Assessments

The echocardiographic protocols for CARDIA year 5 and year 25 were designed to ensure consistency across examinations and followed existing American Society of Echocardiography guidelines at each period. Briefly, trained sonographers, following standardized protocols across all four field centers, acquired images. Experienced readers interpreted digitized images using standard software offline image analysis system transmitted to a core reading laboratory (videotape in year 5 to the University of California, Irvine, and digitally in year 25 to the Johns Hopkins University Echocardiography Reading Center in Baltimore, Maryland). AoD was measured from two-dimensionally guided parasternal M-mode tracings at the level of the sinuses of Valsalva, using the leading edge–to–leading edge approach in end-diastole. LV and left atrial dimensions were acquired using two-dimensionally guided M-mode obtained from parasternal views. LV mass (LVM) was derived according to American Society of Echocardiography guidelines. LV volumes were estimated from apical views using the Simpson method. LV ejection fraction was calculated from LV end-diastolic and end-systolic volumes. Peak early diastolic velocity (E) and peak late diastolic velocity (A) were measured from pulsed Doppler recordings of transmitral flow. Early peak diastolic mitral annular velocity (E′) was derived from tissue Doppler as the average of septal and lateral mitral annular velocities.

Speckle-Tracking Echocardiography

The speckle-tracking echocardiographic protocol and quality control procedures for CARDIA have been previously described. In a retrospective analysis of prospectively collected data, speckle-tracking echocardiography (STE) for LV myocardial strain and strain rate measures were performed offline with dedicated semiautomated two-dimensional wall motion tracking software (Toshiba Medical Systems, Tokyo, Japan). Three cardiac cycles were recorded for apical and parasternal views at an average frame rate of 46.2 frames/sec. After manual tracing of LV endocardial border, a midwall region of interest was automatically defined and adapted when necessary. Lagrangian strain was derived from the change in regional length relative to the length at end-diastole: strain = ( L [ t ] − L ₀ )/ L ₀ × 100, in which L ( t ) is the length at time t , and L ₀ is the length of the segment at the beginning of the QRS complex. Strain rate was defined as the rate of deformation, estimated from the strain temporal derivative. Strain rate parameters were presented as deformation per second. Global strain values were calculated as the average of segmental peak systolic strain. More negative strain values imply better systolic function, whereas more negative diastolic strain rate values denote worse diastolic function.

Covariates

Assessments of risk factor variables in the CARDIA study have been previously described. Briefly, the use of antihypertensive medication and smoking status (current smoker, former smoker, and never smoker) were self-reported, assessed using validated questionnaires. A physical activity score (exercise units) was obtained from the CARDIA physical activity history, a modified version of the Minnesota Leisure Time Physical Activity Questionnaire. Three seated blood pressure measurements were obtained; the mean of the second and third readings was used. Diabetes mellitus was ascertained at each examination on the basis of one or more of a combination of history of medication use (every visit), fasting glucose ≥ 126 mg/dL, 2 hours from a glucose tolerance test (glucose ≥ 200 mg/dL) or glycated hemoglobin ≥ 6.5, and the absence of pregnancy. Total cholesterol, high-density lipoprotein cholesterol, and triglycerides were determined using an enzymatic assay by Northwest Lipids Research Laboratory (Seattle, WA). Low-density lipoprotein cholesterol was derived using the Friedewald equation.

Statistical Analysis

Descriptive characteristics of study participants were summarized for men and women at year 5 and year 25. Continuous variables are presented as mean ± SD or as medians and interquartile ranges, using Student’s t test or the Wilcoxon rank sum (Mann-Whitney) test, as appropriate. Categorical variables were compared using χ ² statistics for frequencies and proportions. AoD at the year 25 examination was used for cross-sectional analyses, and 20-year change in AoD (ΔAoD), calculated as AoD in year 25 minus AoD in year 5, was used for longitudinal analyses. Multivariate linear regression models were used to assess the cross-sectional and longitudinal associations of AoD and ΔAoD with LV structure and function in both sexes. In cross-sectional analyses with LV structure and function as dependent variables, models were adjusted for age, sex, race, education, smoking status (current or former vs never), systolic blood pressure, diastolic blood pressure, resting heart rate, body mass index, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, antihypertensive medications (any vs none), and physical activity (continuous). In longitudinal analyses, the relationship between LV structure and function as dependent variables and ΔAoD was assessed with models adjusted for baseline year 5 risk factors, year 5 echocardiography parameters if available, 20-year change in risk factors, chronic risk exposure (antihypertensive use; consistent antihypertensive use at year 5 and year 25, antihypertensive use at year 5 but not year-25 and vice versa compared with never used antihypertensive medication, diabetes mellitus at year 5 and new-onset diabetes at year 25, smoking [current or former vs never]). To assess potential collinearity of selected covariates in regression models, the variance inflation factor was used. Finally, we examined the interaction of AoD with sex and race in its association with cardiac structural and functional parameters using interaction terms as well as stratified analyses. Statistical significance was set at a two-sided probability of P < .05. All statistical analyses were performed using Stata version (StataCorp, College Station, TX).

Population Characteristics

Baseline and follow-up population characteristics of the 2,933 CARDIA participants are illustrated in Table 1 . Fifty-six percent of study population was female and about 55% Caucasian. The median age at baseline and follow-up was 30 years (interquartile range: 27–33 years) and 51 years (interquartile range: 47–53 years), respectively. The prevalence of diabetes increased significantly from baseline to the 20-year follow-up examination, together with increases in mean body mass index and systolic and diastolic blood pressures in both men and women over 20 years of follow-up. Compared with women, men had higher blood pressures at each visit. Mean LVM increased from baseline to 20-year follow-up in both men and women, although men had higher mean LVM compared with women in both examinations. Conversely, LV ejection fraction was higher in women compared with men at baseline and decreased significantly in both sexes across the 20-year follow-up period. Mean absolute AoD increased from baseline to the 20-year follow-up examination and was higher in men (29.9 ± 3.6 vs 33.3 ± 3.8 mm, P < .001) compared with women (26.1 ± 3.1 vs 28.7 ± 3.4 mm, P < .001) at both the baseline and follow-up examinations. Similarly, men had a larger mean increase in AoD over 20 years compared with women (3.40 ± 3.9 vs 2.62 ± 3.4 mm, P < .001).

The results of sex-specific regression analyses adjusted for temporal changes in cardiovascular risk factors and therapies to control hypertension are shown in Table 2 . Larger AoD at CARDIA examination year 25 was associated with higher LVM (2.77 vs 2.18 g/mm in men and women, respectively), lower LV ejection fraction, and higher ratio of LVM to LV end-diastolic volume in both sexes ( P < .05 for all). Larger AoD at CARDIA examination year 25 was also associated with impaired E′ and impaired longitudinal and circumferential early diastolic strain rates by STE. In contrast, larger AoD at examination year 25 was associated with higher left atrial dimension ( P < .001) in women only.

Table 2

Coefficients for multivariable linear regression of LV structure and function with AoD and 20-year change in AoD

	AoD at year 25		Association of 20-year change in AoD
	Univariate	Multivariate	Univariate	Multivariate
	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)
Men
Structural indices
LVM	3.95 (0.4) §	2.77 (0.36) §	0.88 (0.39) †	1.68 (0.4) §
LVMVR	0.02 (0.004) §	0.01 (0.004) ‡	0.01 (0.004) ‡	0.02 (0.01) ‡
LAD	0.02 (0.003) §	0.01 (0.003) †	0.01 (0.003) ‡	0.01 (0.003) ∗
Systolic functional indices
LVEF	−0.17 (0.05) ‡	−0.15 (0.06) ‡	−0.06 (0.05)	−0.17 (0.1) ∗
Ell	0.02 (0.02)	−0.01 (0.02)	0.02 (0.02)	0.0002 (0.02)
Ecc	0.03 (0.02)	0.04 (0.02)	0.02 (0.02)	0.02 (0.02)
Diastolic functional indices
E/E′ ratio	0.02 (0.02)	−0.01 (0.02)	0.03 (0.02) ∗	−0.005 (0.02)
E′	−0.1 (0.02) §	−0.06 (0.02) §	−0.02 (0.02)	−0.03 (0.02) ∗
Ell_SRE	−0.01 (0.002) §	−0.003 (0.002) ∗	0.0005 (0.002)	−0.002 (0.002)
ECC_SRE	−0.01 (0.003) †	−0.006 (0.003) †	−0.002 (0.002)	−0.004 (0.003)
Women
Structural indices
LVM	3.53 (0.3) §	2.18 (0.27) §	0.70 (0.31) †	1.47 (0.3) §
LVMVR	0.02 (0.004) §	0.01 (0.004) §	0.01 (0.004) †	0.02 (0.01) †
LAD	0.03 (0.003) §	0.02 (0.003) §	0.02 (0.003) §	0.02 (0.004) §
Systolic functional indices
LVEF	−0.16 (0.05) ‡	−0.17 (0.05) ‡	−0.06 (0.05)	−0.19 (0.09) ∗
Ell	0.05 (0.02) ‡	−0.002 (0.02)	0.013 (0.019)	0.002 (0.02)
Ecc	0.04 (0.02) †	0.01 (0.02)	0.01 (0.02)	0.01 (0.03)
Diastolic functional indices
E/E′ ratio	0.07 (0.02) §	0.017 (0.02)	0.07 (0.02) §	0.04 (0.02) †
E′	−0.13 (0.02) §	−0.07 (0.02) §	−0.07 (0.02) §	−0.07 (0.02) §
Ell_SRE	−0.01 (0.002) §	−.008 (0.002) §	−0.004 (0.002) †	−0.089 (0.002) ‡
ECC_SRE	−0.01 (0.002) §	−0.009 (0.003) ‡	−0.005 (0.002) †	−0.01 (0.003) ‡

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. Looking Back, Looking Ahead, Revisited
2. Pediatric Appropriate Use Criteria for Outpatient Echocardiography: Practice Variations among Pediatric Cardiologists, Noncardiologist Subspecialists, and Primary Care Providers
3. Global and Regional Right Ventricular Function Assessed by Novel Three-Dimensional Speckle-Tracking Echocardiography
4. Application of Speckle-Tracking Echocardiography in an Experimental Model of Isolated Subendocardial Damage

Stay updated, free articles. Join our Telegram channel

Join