Background
The aim of this study was to determine the association between cardiovascular health (CVH) in young adulthood and left ventricular (LV) structure and function later in life.
Methods
Participants from the Coronary Artery Risk Development in Young Adults study, which recruited black and white participants aged 18 to 30 years at baseline, were included; echocardiography was performed at year 25. CVH at year 0 was defined on the basis of blood pressure, total cholesterol, fasting glucose, body mass index, smoking status, diet, and physical activity. Two, 1, or 0 points were assigned to each component for ideal, intermediate, and poor levels of each component. Participants were stratified into CVH groups on the basis of point score: ≤8 (poor), 9 to 11 (intermediate), and 12 to 14 (ideal).
Results
The distribution of CVH at year 0 was as follows: poor, n = 264 (9%); intermediate, n = 1,315 (47%); and ideal, n = 1,224 (44%). Individuals with ideal and intermediate CVH at year 0 had significantly lower LV end-diastolic volume and lower LV mass index at year 25. In participants with ideal and intermediate CVH, the multivariate-adjusted odds ratios for diastolic dysfunction at year 25 was 0.52 (95% CI, 0.37–0.73) and 0.63 (95% CI, 0.46–0.83), respectively, compared with participants with poor CVH. Participants with ideal and intermediate CVH had significantly lower odds for LV hypertrophy; the LV mass index was 5.3 to 8.7 g/m 2.7 lower ( P < .001 for both) than in participants with poor CVH.
Conclusion
Greater levels of CVH in young adulthood are associated with lower LV mass and lower risk for diastolic dysfunction 25 years later.
The American Heart Association (AHA) published a set of priorities to promote cardiovascular health (CVH) in the general population. The AHA’s 2020 strategic impact goals define ideal, intermediate, and poor levels of three health factors (blood pressure, total cholesterol, and fasting blood glucose) and four health behaviors (nonsmoking status, body mass index (BMI), physical activity, and nutrition) that form the basis of defining levels of CVH. Previous data from epidemiologic cohorts have reported that the CVH profile is associated with overall survival, cardiovascular mortality, fatal and nonfatal cardiovascular disease (CVD) events, reduced burden of subclinical atherosclerosis, and cancer incidence. However, data are sparse on the longitudinal association of CVH in young adulthood with left ventricular (LV) structure and function later in life.
Subclinical LV systolic and diastolic dysfunction is an important precursor of incident heart failure and is associated with markedly increased risk for all-cause mortality. Indeed, an observational study from Rochester, Minnesota, reported that incident heart failure is nearly always preceded by preclinical systolic or diastolic dysfunction. Even in young adults, the presence of LV diastolic dysfunction is associated with increased risk for clinical CVD and all-cause mortality. The objective of our study was to determine if CVH factors and behaviors in young adulthood are associated with LV structure and systolic and diastolic function later in life. An association between optimal CVH factors and behaviors with favorable LV structure and function would provide important insight into prevention strategies for heart failure.
Methods
Cohort
The Coronary Artery Risk Development in Young Adults (CARDIA) study is an ongoing community-based study funded by the National Heart, Blood, and Lung Institute. Details of the study design and objectives have been published elsewhere. Briefly, the initial cohort included 5,115 individuals recruited in 1985 and 1986 from four urban communities: Birmingham, Alabama; Chicago, Illinois; Minneapolis, Minnesota; and Oakland, California. The sample consisted of approximately equal proportions of black and white men and women, aged 18 to 30 years at the time of enrollment and aged 43 to 55 years at the time of echocardiography. A total of 3,499 participants were examined at year 25, representing 72% of the surviving cohort. We excluded participants who experienced myocardial infarctions, strokes, or coronary revascularization during the follow-up period ( n = 103 [2% of the overall cohort]). We also excluded two transgender participants and 200 participants with incomplete CVH information. Among the remaining 3,194 participants, 60 did not undergo echocardiography, and 331 had incomplete echocardiographic measurements. Thus, 2,803 participants constituted the sample for this analysis.
The CARDIA protocol for the measurement of covariates has been described in detail. Following a 5-min rest, an Omron sphygmomanometer (Omron, Kyoto, Japan) was used to measure systolic blood pressure. Total and high-density lipoprotein cholesterol was measured enzymatically. Fasting blood glucose was measured using the hexokinase ultraviolet assay. Height and weight were measured with participants wearing no shoes and only light clothing, and BMI was calculated as weight in kilograms divided by the square of height in meters. Smoking status was determined by self-report and validated in a prior study using serum cotinine levels. Dietary intake was measured using an interviewer-administered food frequency questionnaire. Intake of potassium, calcium, fiber, and saturated fat was calculated from the University of Minnesota Nutrition Coordinating Center Food Composition Database, version 10. The CARDIA physical activity history was used to derive the physical activity score and has been described in detail elsewhere. We used the education measure from the year 25 examination.
Definition of CVH
The seven components of the CVH profile are blood pressure, total cholesterol, fasting blood glucose, BMI, smoking status, physical activity, and diet. The classification for each health behavior and health factor is shown in Table 1 and is identical to definitions used in a previous CARDIA analysis. We used partition values for physical activity and diet that were also similar to a previous analysis. For the physical activity score, the highest 40% corresponds to 480.0 and 307.3 exercise units for men and women, respectively. Approximately 300 exercise units corresponds to the recommendation by the American College of Sports Medicine for the amount of exercise needed to support weight loss. The dietary score calculation has also been described in detail previously. The highest 40% is intended to correspond with the Dietary Approaches to Stop Hypertension eating pattern.
| Component | Optimal (2 points) | Intermediate (1 point) | Poor (0 points) |
|---|---|---|---|
| Smoking | Never or quit >12 months | Former ≤12 months | Current |
| BMI (kg/m 2 ) | <25 | 25–29.9 | ≥30 |
| PA | Top 40% on PA questionnaire | Second 40% on PA questionnaire | Lowest 20% on PA questionnaire |
| Diet | Highest 40% diet score | Second 40% on diet score | Lowest 20% on diet score |
| Blood pressure (mm Hg) | <120/<80 without any medications | SBP 120–139 or DBP 80–89 or treated to <120/<80 | SBP ≥ 140 or DBP ≥ 90 |
| Total cholesterol (mg/dL) | <200 without medication | 200–239 or treated to <200 | ≥240 |
| Fasting glucose (mg/dL) | < 100 without medication | 100–125 or treated to <100 | ≥126 |
We defined a composite CVH score for each participant at the year 0 examination, as has been described previously. Each individual component was assigned a score of 0 (poor), 1, (intermediate), or 2 (optimal). Thus, a participant with optimal levels of all components would have a score of 14. We then classified participants with point scores ≤ 8 as having poor overall CVH, those with scores of 9 to 11 as having intermediate overall CVH, and those with scores ≥ 12 as having ideal overall CVH.
Echocardiography
The design, acquisition, and storage of echocardiographic data were similar to a protocol that has been previously described and are similar to recommendations from the American Society of Echocardiography. Participants underwent two-dimensional, M-mode, Doppler, and tissue Doppler scanning at the year 25 examination using an Artida cardiac ultrasound scanner (Toshiba Medical Systems, Otawara, Japan). The complete scanning protocol, cardiologists and technologists, and members of the CARDIA Echo Committee are available online at the CARDIA study web site ( http://cardia2.dopm.uab.edu/ ).
We used the biplane method of disks in the apical four-chamber view to calculate LV end-diastolic volume (LVEDV) and LV end-systolic volume (LVESV). Stroke volume was calculated as LVEDV − LVESV, and LV ejection fraction (LVEF) was calculated as (LVEDV − LVESV)/LVEDV. LV volumes were indexed to height. LV mass was derived from diastolic measurements of LV internal diameter in diastole (LVIDd), LV posterior wall thickness (LVPWT), and interventricular septal thickness (IVST). LV mass was then calculated according to the formula developed by Devereux and indexed to height 2.7 , as previously done in CARDIA : LV mass (g) = 0.80{1.04 [(LVIDd + IVST + LVPWT) 3 − LVIDd 3 ]} + 0.6. Parasternal long-axis two-dimensional views were used to guide the assessment of M-mode anteroposterior diameter of the left atrium. Left atrial volumes were acquired from a two-dimensional four-chamber view, measured at the point of maximum atrial volume, and calculated using the biplane method of disks.
Pulsed-wave Doppler was performed in the apical four-chamber view to assess mitral inflow velocities. Peak early filling (E wave) and peak late filling (A wave) were measured. Pulsed-wave Doppler tissue imaging was used to acquire early and late mitral annular velocities in diastole. The sample volume was placed within 1 cm of the mitral valve septal and lateral insertion sites; E′ denotes the early mitral annular velocity. The tissue Doppler velocity at the lateral mitral annulus in systole was also acquired (denoted as S′). We estimated LV filling pressures by dividing E velocity by lateral E′ velocity.
For the categorical classification of diastolic dysfunction, we used an age-based partition value for lateral E′ velocity as our primary definition of diastolic dysfunction: lateral E′ velocity < 10 cm/sec was considered diastolic dysfunction. This value was derived in a population-based study, was used in a large clinical trial, and is also recommended as a partition value in guidelines from the American Society of Echocardiography. In a secondary analysis, we defined diastolic dysfunction as (1) increased left atrial volume index (>34 mL/m 2 ) and (2) either abnormal lateral E′ (<10 cm/sec) or septal E′ (<8 cm/sec). We elected not to assign grades of diastolic dysfunction in favor of this method, which is more feasible and more easily reproduced. In separate analyses, we also examined measures of diastolic function, including tissue Doppler velocities and estimated filling pressures, as continuous variables. LV hypertrophy (LVH) was defined on the basis of sex-specific partition values published by the American Society of Echocardiography. LV mass index ≥ 45 g/m 2.7 was used as the partition value to define LVH in women, and ≥49 g/m 2.7 was used as the partition value in men. LVEF < 50% was considered systolic dysfunction.
Statistical Analysis
We used χ 2 tests for categorical variables and analysis of variance for continuous variables, as appropriate, to compare baseline characteristics across CVH groups. We then compared year 25 echocardiographic parameters across CVH groups, using χ 2 tests for categorical parameters and analysis of variance for continuous parameters. We used multivariate-adjusted linear regression analyses to determine the association between CVH group at year 0 and continuous echocardiographic parameters at year 25. Multivariate models were adjusted for age, race, sex, and year 25 education level. Logistic regression modeling was used to determine the association between CVH group and LVH, diastolic dysfunction, and systolic dysfunction at year 25. We focused the multivariate analyses on the dependent variables of LV mass and LVH, systolic function (as represented by ejection fraction), and diastolic function because these measures represent the most clinically important intermediate phenotypes of CVD. The poor CVH group was used as the referent in all models. We performed sensitivity analyses excluding obesity from the CVH score. For the sensitivity analyses, a score of 10 to 12 was considered ideal, 6 to 9 was considered intermediate, and ≤5 was considered poor.
We also analyzed the association between the change in CVH over time with diastolic dysfunction and LVH 25 years later, using a method previously described in CARDIA. As above, each individual component was assigned a score of 0, 1, or 2 for poor, intermediate, or ideal CVH status, respectively. We assessed the total CVH score for each participant at year 0 and year 20, which ranged from 0 to 14. The change in CVH was defined as the difference of CVH score between year 0 and year 20. We used year 20 because the diet score was not available at year 25.
We used spline analysis for logistic regression to investigate the relationship between LV diastolic dysfunction (primary definition) and LV mass with the continuous CVH point score. To fit the spline, we attempted to fit linear, quadratic, and cubic curves with one interior knot placed on the median, two interior knots placed on tertiles, and three interiors knots placed on quartiles, respectively. The curves appeared similar, and we chose to present the quartile splines with one interior knot. This model was chosen on the basis of the optimal model fit as determined by the Akaike information criterion. The multivariate-adjusted odds ratios (ORs) for diastolic dysfunction and LVH were plotted against the CVH point score, using a point score of 7 as the referent value. We also calculated the attributable risk for LVH and diastolic dysfunction, which represents the excess risk for these outcomes in participants with poor or intermediate CVH. SAS version 9.3 (SAS Institute, Inc, Cary, North Carolina) was used for all analyses. Two-sided P values < .05 were considered to indicate statistical significance.
Results
Baseline Characteristics
Table 2 shows the year 0 characteristics of the study population stratified by CVH group. The prevalence of ideal, intermediate, and poor CVH at year 0 was 44% ( n = 1,224), 47% ( n = 1,315), and 9% ( n = 264), respectively. As expected, participants with poor CVH had more adverse levels of cardiovascular risk factors compared with those with intermediate and ideal CVH. Black participants constituted a majority of those with poor CVH (78%) and only 28% of those with ideal CVH ( P for trend < .001). Participants with ideal CVH had more years of education than those with lower levels of CVH ( P for trend < .001). Table 3 shows the year 0 characteristics of the entire cohort, the participants examined at year 25, and the participants included in the study. Table 3 also shows the year 25 characteristics of those available for analysis and those ultimately included and demonstrates that the study cohort consisted of fewer black participants, fewer smokers, and a lower mean BMI than the overall cohort and excluded participants.
| Covariate | CVH point score stratum | P | ||
|---|---|---|---|---|
| Poor CVH ( n = 264) | Intermediate CVH ( n = 1,315) | Ideal CVH ( n = 1,224) | ||
| Age (y) | 25.8 ± 3.5 | 24.8 ± 3.6 | 25.1 ± 3.6 | <.001 |
| Black race | 205 (78%) | 750 (57%) | 341 (28%) | <.001 |
| Men | 98 (37%) | 578 (44%) | 524 (43%) | .13 |
| Education (y) | 14 ± 2 | 15 ± 2 | 16 ± 2 | <.001 |
| Systolic blood pressure (mm Hg) | 115 ± 11 | 111 ± 11 | 107 ± 9 | <.001 |
| Diastolic blood pressure (mm Hg) | 71 ± 11 | 68 ± 10 | 67 ± 8 | <.001 |
| Creatinine (mg/dL) | 1.1 ± 0.9 | 1.0 ± 0.2 | 1.0 ± 0.2 | .32 |
| BMI (kg/m 2 ) | 28.8 ± 6.6 | 24.6 ± 4.6 | 22.5 ± 2.5 | <.001 |
| Total cholesterol (mg/dL) | 198 ± 41 | 180 ± 34 | 168 ± 26 | <.001 |
| HDL cholesterol (mg/dL) | 50 ± 15 | 53 ± 13 | 55 ± 12 | <.001 |
| LDL cholesterol (mg/dL) | 130 ± 37 | 113 ± 32 | 101 ± 24 | <.001 |
| Triglycerides (mg/dL) | 93 ± 77 | 71 ± 40 | 62 ± 29 | <.001 |
| Fasting blood glucose (mg/dL) | 86 ± 20 | 82 ± 10 | 81 ± 7 | <.001 |
| Current smoker | 177 (67%) | 461 (35%) | 57 (5%) | <.001 |
| Antihypertensive medication | 26 (10%) | 25 (2%) | 2 (1%) | <.001 |
| Diabetes medication | 12 (5%) | 5 (1%) | 0 (0%) | <.001 |
| Covariate | Entire cohort ( n = 5,113) | All year 25 participants ( n = 3,482) | Study cohort ( n = 2,803) |
|---|---|---|---|
| Year 0 baseline characteristics | |||
| Age (y) | 24.8 ± 3.7 | 25.1 ± 3.6 | 25.0 ± 3.6 |
| Sex (% female) | 54.5 ( n = 2,785) | 56.6% ( n = 1,980) | 57.3% ( n = 1,579) |
| Race (% black) | 48.4% ( n = 2,637) | 46.9% ( n = 1,640) | 46.0% ( n = 1,267) |
| Education (y) | 13 ± 2 | 14 ± 2 | 14 ± 2 |
| BMI (kg/m 2 ) | 24.5 ± 5.0 | 24.5 ± 5.0 | 24.0 ± 4.4 |
| SBP (mm Hg) | 110 ± 11 | 110 ± 11 | 110 ± 11 |
| DBP (mm Hg) | 69 ± 10 | 68 ± 9 | 68 ± 9 |
| Total cholesterol (mg/dL) | 177 ± 33 | 177 ± 33 | 177 ± 33 |
| HDL cholesterol (mg/dL) | 53 ± 13 | 53 ± 13 | 53 ± 13 |
| Fasting glucose (mg/dL) | 83 ± 16 | 82 ± 10 | 82 ± 10 |
| Current smoker | 30% ( n = 1,544) | 26.7% ( n = 927) | 24.7% ( n = 679) |
| Year 25 characteristics | |||
| Education (y) | — | 15 ± 3 | 15 ± 3 |
| BMI (kg/m 2 ) | — | 30 ± 7 | 29 ± 7 |
| SBP (mm Hg) | — | 120 ± 16 | 119 ± 16 |
| DBP (mm Hg) | — | 75 ± 11 | 74 ± 11 |
| Total cholesterol (mg/dL) | — | 192 ± 37 | 192 ± 36 |
| HDL cholesterol (mg/dL) | — | 58 ± 18 | 59 ± 18 |
| Fasting glucose (mg/dL) | — | 99 ± 29 | 98 ± 27 |
| Current smoker | — | 17% ( n = 589) | 16% ( n = 429) |
Echocardiographic Parameters
Table 4 shows echocardiographic characteristics at year 25 stratified by CVH group at year 0. Participants with ideal and intermediate CVH had lower indexed LVEDV and LVESV than those with poor CVH. Participants with ideal and intermediate CVH also had smaller LV internal diastolic diameter. Participants with ideal CVH had significantly lower LV mass index and left atrial volume compared with participants who had poor or intermediate CVH. There was no difference in the mean LVEF across the CVH groups; however, with participants with ideal and intermediate CVH had higher lateral S′ velocity than those with poor CVH. Given the differences observed in mean heart rate and stroke volume, there was a significant difference observed in cardiac index across CVH groups as well.
| Covariate | CVH point score stratum | P -value for trend | ||
|---|---|---|---|---|
| Poor ( n = 264) | Intermediate ( n = 1,315) | Ideal ( n = 1,224) | ||
| LVEDV index (mL/m) | 41 ± 9 | 38 ± 9 | 37 ± 9 | <.001 |
| LVESV (mL/m) | 16 ± 6 | 15 ± 6 | 14 ± 6 | <.001 |
| LVIDd (cm) | 5.2 ± 0.5 | 5.1 ± 0.5 | 5.1 ± 0.5 | <.001 |
| Heart rate (beats/min) | 68 ± 11 | 66 ± 10 | 64 ± 10 | <.001 |
| LV stroke volume (mL) | 93 ± 21 | 88 ± 21 | 86 ± 21 | <.001 |
| Left atrial volume index (mL/m 2 ) | 28 ± 9 | 27 ± 8 | 27 ± 8 | .008 |
| Left atrial diameter (cm) | 3.9 ± 0.5 | 3.7 ± 0.5 | 3.6 ± 0.5 | <.001 |
| LV mass index (g/m 2.7 ) | 47 ± 15 | 41 ± 12 | 37 ± 10 | <.001 |
| LVH (N,%) | 111 (44%) | 338 (26%) | 169 (14%) | <.001 |
| Cardiac index (L/min/m 2 ) | 3.3 ± 0.9 | 3.2 ± 0.8 | 3.1 ± 0.8 | <.001 |
| LVEF (%) | 62 ± 8 | 62 ± 7 | 62 ± 7 | .97 |
| Systolic lateral S′ (cm/sec) | 8.6 ± 2.0 | 9.2 ± 2.3 | 9.3 ± 2.1 | <.001 |
| Lateral E′ velocity (cm/sec) | 10.9 ± 2.9 | 11.6 ± 2.7 | 12.3 ± 2.8 | <.001 |
| Septal E′ velocity (cm/sec) | 8.6 ± 2.5 | 9.2 ± 2.3 | 9.8 ± 2.4 | <.001 |
| E/E′ ratio | 8.0 ± 3.0 | 7.3 ± 2.4 | 6.6 ± 2.1 | <.001 |
| Mitral E/A ratio | 1.2 ± 0.4 | 1.4 ± 0.4 | 1.3 ± 0.3 | <.001 |
| Mitral deceleration time (msec) | 179 ± 41 | 178 ± 38 | 178 ± 39 | .89 |
| Diastolic dysfunction (primary) | 101 (41%) | 338 (26%) | 220 (18%) | <.001 |
| Diastolic dysfunction (secondary) | 36 (14%) | 100 (8%) | 71 (6%) | <.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
