The effect of cardiovascular risk factors on the left ventricle is well known but their effect on right ventricle has not been studied using advanced imaging techniques. The purpose of the present study was to determine the relation between the cardiovascular risk factors and right ventricular (RV) structure and function and its interaction with the left ventricle. Cardiac magnetic resonance images from 4,204 participants free of clinical cardiovascular disease in the Multi-Ethnic Study of Atherosclerosis were analyzed. Multivariate linear regression models were used to study the cross-sectional association between individual RV parameters and risk factors. All RV parameters, except for ejection fraction, decreased with age (p <0.0001). The RV mass was positively associated with systolic blood pressure (+0.4 g, p <0.0001) and high-density lipoprotein cholesterol (+0.2 g, p <0.0001). It was inversely related with diastolic blood pressure (−0.3 g, p <0.0001) and total cholesterol (−0.2 g, p <0.01). The RV end-diastolic volume was positively associated with systolic blood pressure (+1.6 ml, p <0.01) and high-density lipoprotein cholesterol (+1.8 ml, p <0.0001). It was inversely related with diastolic blood pressure (−2.2 ml, p <0.01), total cholesterol (−1.4 ml, p <0.0001), current smoking (−2.7 ml, p <0.05), and diabetes mellitus (−3.1 ml, p <0.01). The RV ejection fraction was positively associated with systolic blood pressure (+1.0%, p <0.0001), high-density lipoprotein cholesterol (+0.4%, p <0.0001) and inversely with diastolic blood pressure (−0.7%, p <0.0001). In conclusion, the mass and volumes of the right ventricle decrease with age. Cardiovascular risk factors, especially blood pressure and high-density lipoprotein cholesterol, are associated with subclinical changes in the RV mass and volume.
Magnetic resonance imaging (MRI) has emerged as an accurate tool for right ventricular (RV) evaluation and is considered a standard of reference for the evaluation of both the right and left ventricle. Previous studies of RV function using MRI in healthy volunteers have included smaller numbers of participants and did not assess the relation between cardiovascular risk factors, such as age, diabetes mellitus, systolic and diastolic blood pressure, cholesterol, and smoking, and the RV mass and volume, particularly in relation to the left ventricle. The Multi-Ethnic Study of Atherosclerosis (MESA) is an ongoing multicenter, prospective, cohort study of subjects without clinical cardiovascular disease, in whom cardiac MRI was performed to assess the cardiac structure and function. The purpose of the present study was to determine the relation between the cardiovascular risk factors and the RV structure and function and its interaction with the left ventricle.
Methods
The MESA was designed to study subclinical cardiovascular disease in subjects without previous clinical cardiovascular disease. In 2000 to 2002, MESA recruited 6,814 men and women aged 45 to 84 years from 6 United States communities. The MESA participants were non-Hispanic white, African American, Hispanic, and Asian. The exclusion criteria included clinical cardiovascular disease (physician diagnosis of heart attack, stroke, transient ischemic attack, heart failure, or angina); current atrial fibrillation; any cardiovascular procedure; pregnancy; active cancer treatment; weight >300 lb; serious medical condition precluding long-term participation; nursing home residence; cognitive inability; an inability to speak English, Spanish, Cantonese, or Mandarin; a plan to leave the community within 5 years, and chest computed tomography within the previous year. The protocols of MESA and all studies described in the present report were approved by the institutional review boards of all collaborating institutions and the National Heart, Lung, and Blood Institute (Bethesda, Maryland).
Consenting and eligible (without metal implants, devices, or fragments) participants underwent a cardiac MRI scan using 1.5 T GE (General Electric Medical Systems, Waukesha, Wisconsin) and Siemens scanners (Siemens, Erlanger, Germany) at the 6 MESA field centers. Cardiac MRI was performed using a standard protocol. The imaging was performed with a 4-element, phased-array, surface coil placed anteriorly and posteriorly, with electrocardiographic gating and brachial artery blood pressure monitoring. All images were acquired during short breath-holding (12 to 15 seconds) at rest lung volume. Imaging consisted of fast gradient echo cine images of the heart with a time resolution of <50 ms. Quantitative analysis of the right ventricle was performed for 4,204 participants with interpretable MRI scans.
The cardiac MRI examinations were transmitted to the reading center at Johns Hopkins Hospital (Baltimore, Maryland) using Digital Imaging and Communications in Medicine transfer protocol. Image analysis was done on Windows workstations using the QMASS software (research version, Medis Medical Quantification Software, Leiden, The Netherlands). The images were magnified to 250%, contrast was set to 55, image brightness was set to 55, and the window width and level were set using the autofunction in QMASS to the minimum and maximum pixel values. Image analysis was done by 2 independent analysts who had received extensive training in the MESA protocol and RV morphology. The endocardial and epicardial borders for the right ventricle were contoured manually on the short-axis cine images at the end-systolic and end-diastolic phases. The papillary muscles and trabeculae were included in the RV volume and were excluded from the RV mass on the basis of preliminary training data sets showing greater reproducibility for the analysis. The outflow tract was included in the RV volume. The contours were modified at the basal slices of the heart by carefully identifying the tricuspid valve to exclude the right atrium and avoid overestimation of the volume. The RV end-systolic volume and end-diastolic volume (EDV) were calculated using Simpson’s rule, by summation of the areas on each slice multiplied by the sum of the slice thickness and image gap. The RV mass was determined at the end-diastolic phase as the difference between the end-diastolic epicardial and endocardial volumes multiplied by the specific gravity of the heart (1.05 g/ml). The RV stroke volume was calculated by subtracting the RV end-systolic volume from the EDV. The RV ejection fraction (EF) was calculated by dividing the RV stroke volume by the RV EDV and multiplying by 100.
Reader variability in the assessment of the RV mass and volume was measured by repeat readings selected randomly. The readers were masked to the results obtained from the previous analyses. Outliers were defined as measures with a difference greater than 2 SDs from the mean relative difference. The outliers were reviewed and arbitrated by an experienced MRI physician.
Standard questionnaires were used to ascertain smoking (classified as never, former, and current). Height was measured to the nearest 0.1 cm, with the subject in stocking feet. Weight was measured to the nearest pound, with the subject in light clothing, using a balanced scale. Blood pressure at rest was measured using the Dinamap Monitor PRO 100 (Critikon, Tampa, Florida) automated oscillometric device. The serum glucose level after a 12-hour fast was measured by rate reflectance spectrophotometry using thin film adaptation of the glucose oxidase method on the Vitros analyzer (Johnson & Johnson Clinical Diagnostics, Rochester, New York). The diagnosis of diabetes mellitus was determined by the use of insulin or oral hypoglycemic medication or a fasting glucose level of ≥126 mg/dl. Impaired fasting glucose was considered present if the fasting glucose was 100 to 125 mg/dl. High-density lipoprotein and total cholesterol were assessed using standard methods.
The data are presented as the mean ± SD for continuous variables and as percentages for discrete variables. Intra- and inter-reader variability was determined using the intraclass correlation coefficient, Pearson’s correlation coefficient, and the percentage of technical error of measurement. Independent variables were age, systolic blood pressure, diastolic blood pressure, total cholesterol, high-density lipoprotein cholesterol, diabetes mellitus or impaired fasting glucose (vs normoglycemic findings), and smoking (current vs former and never). Multivariate linear regression analyses were performed to assess the independent associations of the individual risk factors with each RV parameter after adjustment for gender, ethnicity, height, and weight. In model 2, the corresponding left ventricular (LV) parameter was added. p Values <0.05 were considered significant. Multiple interaction terms were introduced to evaluate the associations of each risk factor across different strata of age, gender, and ethnicity; significance was declared at p <0.004 after applying Bonferroni’s correction for an average of 12 comparisons. To compare the association with the LV parameters, similar multivariate analysis models were performed. The β coefficients for the risk factors in each model and the percentage of variability explained by the model (adjusted r 2 ) were determined. Analyses were repeated by including the height squared or the body surface area instead of the height and weight as measures of body size. No significant change in β coefficients was detected; thus, the height- and weight-adjusted models only are presented. All analyses were done using Stata Statistical Software, release 10.0 (StataCorp, College Station, Texas).
Results
The MESA included 6,814 participants. Of these, 5,098 participants were eligible and underwent cardiac MRI examination, of which 4,204 were randomly selected for right ventricle interpretation. The data from 30 participants (<1%) were missing for ≥1 cardiovascular risk factor variables and were excluded from the study. Compared to those not included in the study (n = 2,640), the 4,174 participants were, on average, slightly younger in age (61.4 vs 63.3 years), had lower systolic blood pressure (125.5 vs 128.4 mm Hg), lower body mass index (27.9 vs 29.1 kg/m 2 ), had a greater proportion of Chinese Americans (12.4% vs 10.8%), lower proportion of African Americans (26.3% vs 30.2%), were less likely to have hypertension (42.9% vs 48.1%), less likely to be taking hypertension medication (35.8% vs 39.4%), less likely to have diabetes mellitus (11.7% vs 14.1%), and were more likely to be nonsmokers (51.6% vs 48.5%; all p <0.05). The mean age of the 4,174 participants included in the present study was 61.4 ± 10 years. Of the participants, 47.5% were men. 39.2% were white, 26.3% were African American, 22.1% were Hispanic, and 12.4% were Chinese-American ( Table 1 ).
| Parameter | Study Sample (n = 4,174) | Excluded (n = 2,640) |
|---|---|---|
| Age (years) | 61.4 ± 10.1 ⁎ | 63.3 ± 10.4 |
| Men | 47.5% | 46.6% |
| Ethnicity | ||
| White | 39.2% | 37.4% |
| Chinese American | 12.4% ⁎ | 10.8% |
| African American | 26.3% ⁎ | 30.2% |
| Hispanic | 22.1% | 21.6% |
| Height (cm) | 166.4 ± 10.0 | 166.3 ± 10.2 |
| Weight (kg) | 77.4 ± 16.2 ⁎ | 80.6 ± 18.9 |
| Systolic blood pressure (mm Hg) | 125.5 ± 21 ⁎ | 128.4 ± 22 |
| Diastolic blood pressure (mm Hg) | 71.9 ± 10.2 | 72 ± 10.4 |
| Total cholesterol (mg/dl) | 194.3 ± 35.1 | 194.0 ± 36.8 |
| High-density lipoprotein cholesterol (mg/dl) | 51.1 ± 15.0 | 50.7 ± 14.5 |
| Hypertension | 42.9% ⁎ | 48.1% |
| Treated hypertension | 35.8% ⁎ | 39.4% |
| Diabetes mellitus (treated or untreated) | 11.7% ⁎ | 14.1% |
| Impaired fasting glucose | 13.2% | 14.8% |
| Cigarette smoking | ||
| Current | 12.6% | 13.8% |
| Former | 35.8% | 37.7% |
| Never | 51.6% ⁎ | 48.5% |
| Body mass index (kg/m 2 ) | 27.9 ± 5.0 ⁎ | 29.1 ± 6.1 |
The mean ± SD for the RV parameters was 21.1 ± 4.4 g for mass, 124.2 ± 30.9 ml for EDV, 37.3 ± 14.2 ml for end-systolic volume, 86.8 ± 20.6 ml for stroke volume, and 70.4 ± 6.5% for EF. Gender-specific values for RV mass, volumes, and EF are presented in Table 2 . The mean values of RV mass and volume were significantly greater in men than in women. Gender differences remained significant for all parameters, even when indexed for body surface area (respective parameter divided by body surface area). RVEF was significantly lower in men than in women. With the sample sizes used in the present study, it was possible to detect gender differences in the RV parameters with a power >90%.
| RV Parameter | Men (n = 1,995) | Women (n = 2,209) |
|---|---|---|
| Mass (g) | 23.1 ± 4.4 | 19.2 ± 3.6 ⁎ |
| End-diastolic volume (ml) | 140.9 ± 29.6 | 109.1 ± 23.3 ⁎ |
| End-systolic volume (ml) | 45.1 ± 14.0 | 30.4 ± 10.2 ⁎ |
| Stroke volume (ml) | 95.8 ± 20.6 | 78.7 ± 16.8 ⁎ |
| Ejection fraction (%) | 68.2 ± 6.2 | 72.4 ± 6.0 ⁎ |
| Indexed for body surface area | ||
| Mass (g/m 2 ) | 11.7 ± 2.0 | 11.0 ± 1.7 ⁎ |
| End-diastolic volume (ml/m 2 ) | 71.5 ± 13.0 | 62.2 ± 10.6 ⁎ |
| End-systolic volume (ml/m 2 ) | 22.8 ± 6.5 | 17.3 ± 5.2 ⁎ |
| Stroke volume (ml/m 2 ) | 48.7 ± 9.4 | 44.9 ± 7.9 ⁎ |
The scans of 136 participants were included in the intrareader reliability analyses, and 154 were included in the inter-reader reliability analyses. The intraclass correlation coefficients, Pearson correlation coefficients, and percentage of technical error of measurement are listed in Table 3 . All parameters had intraclass correlation coefficients for both intrareader and inter-reader reliability >0.85, except for RVEF, which was slightly lower (0.75 inter-reader intraclass correlation coefficient).
| RV Parameter | Inter-Reader | Intrareader | ||||
|---|---|---|---|---|---|---|
| ICC | Pearson Correlation Coefficient | Technical Error of Measurement (%) | ICC | Pearson Correlation Coefficient | Technical Error of Measurement (%) | |
| Mass (g) | 0.87 | 0.78 | 10.9 | 0.93 | 0.87 | 7.1 |
| End-diastolic volume (ml) | 0.96 | 0.92 | 7.5 | 0.99 | 0.98 | 3.7 |
| End-systolic volume (ml) | 0.93 | 0.87 | 14.1 | 0.98 | 0.95 | 8.3 |
| Stroke volume (ml) | 0.92 | 0.85 | 10.3 | 0.97 | 0.95 | 5.3 |
| Ejection fraction (%) | 0.75 | 0.61 | 6.3 | 0.93 | 0.88 | 3.5 |
The Pearson’s correlation coefficients for RV mass, EDV, end-systolic volume, stroke volume, and EF with the corresponding LV parameter are listed in Table 4 . All RV parameters correlated positively with the LV parameters, ranging from 0.82 for the end-diastolic volume to 0.47 for EF (all p <0.0001).
| Right Ventricle | Left Ventricle |
|---|---|
| Mass | 0.62 ⁎ |
| End-diastolic volume | 0.82 ⁎ |
| End-systolic volume | 0.64 ⁎ |
| Stroke volume | 0.79 ⁎ |
| Ejection fraction | 0.47 ⁎ |
The results from the multivariate linear regression models for RV mass, EDV, and EF in relation to cardiovascular risk factors are listed in Table 5 , both before (model 1) and after (model 2) adjustment for the corresponding LV parameter. The proportion of variability explained by adding the LV to the multivariate model increased from 45% to 52% for RV mass, 52% to 74% for RV EDV, and 15% to 28% for the RVEF. Because the mean RV and LV masses were significantly different (21.1 vs 145.3 g, respectively), the percentage of change in the RV and LV parameters in relation to the individual cardiovascular risk factors are listed in Table 6 .
| RV mass (g) | RV EDV (ml) | RVEF (%) | ||||
|---|---|---|---|---|---|---|
| Proportion of Variability Explained by Model (Adjusted r 2 ) | Model 1 (45%) | Model 2 (52%) | Model 1 (52%) | Model 2 (74%) | Model 1 (15%) | Model 2 (28%) |
| Age (per 10 years) | −1.0 ⁎ (−1.2, −0.9) | −0.9 ⁎ (−1.0, −0.8) | −5.7 ⁎ (−6.5, −4.9) | −1.6 ⁎ (−2.3, −1.0) | 0.3 † (0.1, 0.5) | 0.2 (0.0, 0.4) |
| Systolic blood pressure (per 21 mm Hg) | 0.4 ⁎ (0.2, 0.5) | 0.0 (−0.2, 0.1) | 1.6 † (0.6, 2.6) | −1.6 ⁎ (−2.4, −0.8) | 1.0 ⁎ (0.7, 1.3) | 0.6 ⁎ (0.4, 0.9) |
| Diastolic blood pressure (per 10 mm Hg) | −0.3 ⁎ (−0.5, −0.2) | −0.3 ⁎ (−0.5, −0.2) | −2.2 † (−3.2, −1.2) | 0.1 (−0.9, 0.6) | −0.7 ⁎ (−1.0, −0.5) | −0.3 † (−0.6, −0.1) |
| Current smokers (vs never smokers) | −0.1 (−0.5, 0.2) | −0.5 † (−0.8, −0.2) | −2.7 ‡ (−4.8, 0.6) | −3.1 ⁎ (−4.6, −1.5) | −0.3 (−0.9, 0.3) | 0.3 (−0.2, 0.9) |
| Total cholesterol (per 35 mg/dl) | −0.2 † (−0.3, −0.1) | −0.1 ‡ (−0.3, 0.0) | −1.4 ⁎ (−2.1, −0.7) | −0.4 (−0.9, 0.1) | 0.1 (0.0, 0.3) | 0.1 (−0.1, 0.3) |
| High-density lipoprotein (per 15 mg/dl) | 0.2 ⁎ (0.1, 0.4) | 0.2 † (0.1, 0.3) | 1.8 ⁎ (1.1, 2.6) | 0.3 (−0.3, 0.8) | 0.4 ⁎ (0.2, 0.6) | 0.4 ⁎ (0.2, 0.6) |
| Impaired fasting glucose (vs normoglycemic) | −0.3 ‡ (−0.7, 0.0) | −0.4 ‡ (−0.7, −0.1) | −2.6 ‡ (−4.6, −0.6) | 0.1 (−1.4, 1.6) | 0.4 (−0.2, 0.9) | 0.4 (−0.1, 0.9) |
| Diabetes mellitus (vs normoglycemic) | 0.0 (−0.4, 0.3) | −0.2 (−0.5, 0.1) | −3.1 † (−5.2, −1.0) | −1.0 (−2.6, 0.6) | −0.2 (−0.8, 0.4) | 0.2 (−0.4, 0.7) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
