Abstract
Background.
Higher arterial stiffness may contribute to future alterations in left ventricular systolic and diastolic function. We tested this hypothesis in individuals with youth-onset type 2 diabetes from the Treatment Options for type 2 Diabetes in Adolescents and Youth (TODAY) study.
Methods.
Arterial stiffness [(pulse wave velocity (carotid-femoral, femoral-foot and carotid-radial), augmentation index, brachial distensibility] was measured in 388 participants with type 2 diabetes (mean age 21 years; diabetes duration 7.7 ± 1.5). To reflect overall (composite) vascular stiffness, the five arterial stiffness measures were aggregated. An echocardiogram was performed in the same cohort 2 years later. Linear regression models assessed whether composite arterial stiffness was associated with left ventricular mass index, systolic and diastolic function, independent of age, sex, race-ethnicity, current cigarette smoking, and long-term exposure (time weighted mean values over 9.1 years) of hemoglobin A1c, blood pressure, and body mass index. Interactions between arterial stiffness and time weighted mean hemoglobin A1c, blood pressure, and body mass were also examined.
Results.
After adjustment, arterial stiffness remained significantly associated with left ventricular mass index and diastolic function measured by mitral valve E/Em despite attenuation by time weighted mean body mass index. A significant interaction revealed greater adverse effect of composite arterial stiffness on mitral valve E/Em among participants with higher levels of blood pressure over time. Arterial stiffness was unrelated to left ventricular systolic function.
Conclusions.
The association of higher arterial stiffness with future left ventricular diastolic dysfunction suggests the path to future heart failure may begin early in life in this setting of youth-onset type 2 diabetes.
Trial Registration:
Keywords: arterial stiffness, cardiac structure, diastolic function, type 2 diabetes
Introduction
Obesity, elevated blood pressure (BP), and diabetes are associated with cardiovascular target organ injury in adolescence and young adulthood, in particular, carotid femoral pulse wave velocity (PWV) and left ventricular (LV) structure and function (1–3). It can be hypothesized that higher arterial stiffness, as measured by higher PWV, may independently impact LV structure and function, by increasing LV afterload, stimulating LV hypertrophy and myocardial fibrosis (4, 5). Over time, this could result in systolic and diastolic dysfunction with subsequent heart failure, either related to reduced LV ejection fraction or heart failure with preserved ejection fraction (6).
We and others have demonstrated worsening of obesity, hypertension and glycemic control as individuals with youth-onset type 2 diabetes (T2D) progress into adulthood, and also worsening of arterial stiffness, LV structure and function (6–9). However, the relationship between arterial stiffness to subsequent alterations in LV structure and function has not been examined. Studies in youth and adults are limited by either cross-sectional design (5), or, in longitudinal studies (6, 7), by the absence of data on both PWV and cardiac imaging, obtained with a sufficient interval between arterial stiffness assessment and cardiac imaging, to determine the impact of arterial stiffness on cardiac structure and function, independent of cardiovascular risk factors.
The Treatment Options for T2D in Adolescents and Youth (TODAY) study, a randomized clinical trial (2004–2011) of treatments for recent onset of T2D in adolescents, with observational follow up (TODAY2) of the cohort for an additional 9 years (2011–2020), obtained sequential assessments of arterial stiffness and echocardiography, two years apart during the observational follow up phase. This allowed testing of the hypothesis that arterial stiffness would adversely impact LV structure and function later, independent of obesity, hypertension, and glycemic control. Additionally, we hypothesized that participants with more arterial stiffness or adverse echocardiographic measures would have higher BP, higher body mass index (BMI), and worse glycemic control.
Material and methods
TODAY Study Design
Detailed descriptions of the TODAY clinical trial protocol (www.clinicaltrials.gov: NCT00081328) and the primary outcome results have been published (8, 10). In brief, the TODAY study was a multicenter randomized clinical trial designed to assess the effect of metformin monotherapy, metformin plus rosiglitazone, or metformin plus an intensive lifestyle intervention on time to treatment failure defined as loss of glycemic control (hemoglobin A1c [HbA1c] ≥8% for 6 months or inability to wean from temporary insulin after acute metabolic decompensation) in adolescents with youth-onset T2D (n=699 randomized, aged 11–17 years old, duration of diabetes <2 years, and BMI >85 percentile for age and sex). At the end of the TODAY clinical trial, 572 of the original 699 TODAY participants enrolled in the TODAY observational follow-up study (TODAY2) designed to assess the long-term complications of diabetes.
The analyses reported include 388 participants from 15 clinical sites who had arterial stiffness assessments conducted 7.0±1.3 years after randomization in the TODAY study and a subsequent echocardiogram approximately 2 years later (9.1±1.3 years after randomization). All participants in TODAY2 were eligible but some did not complete both assessments. Comparison of the 388 included in the analysis sample to the other 311 TODAY participants (out of 699 randomized) shows that those included in this sample were slightly younger at baseline (mean ± SD 13.8±2.0 vs. 14.2±2.0 years old) but not different in regard to sex, race-ethnicity, baseline BMI and HbA1c, or baseline duration of diabetes. Participants found to have monogenic diabetes mutations after randomization (n=22) were excluded from these analyses.
Risk Factor Assessment
Age, sex, medication usage and race-ethnicity were self-reported in the study, the latter using U.S. Census–based questions. Height, weight, BP and laboratory data were measured using a study-wide protocol (10). BMI was calculated as kg/m2 and mean arterial BP calculated as [(2*diastolic BP) + systolic BP]/3.
Self-reported cigarette smoking was collected within one month of the echocardiogram and categorized as either “yes” (used within the past month) or “no” (never used/not used within the past month). HbA1c was measured at the TODAY Central Biochemistry Laboratory (CBL, Seattle, WA) using a dedicated high-performance liquid chromatography method (TOSOH Biosciences Inc., South San Francisco, CA). HbA1c and fasting labs (lipids) were obtained according to standardized procedures and shipped to the CBL (10). As HbA1c, mean arterial BP, and BMI were obtained multiple times throughout the TODAY study, time weighted means of all follow-up values (approximately 30 values for each variable) were used to represent the cumulative or chronic exposure since randomization in the TODAY clinical trial.
Outcome Measures
The five arterial stiffness measurements included 1) carotid femoral pulse wave velocity (PWV, primary outcome), 2) carotid radial PWV, 3) femoral foot PWV, 4) augmentation index (AIx), and 5) brachial distensibility (BrachD) as previously described (2, 11). PWV and AIx measurements were obtained using the SphygmoCor CPV system (AtCor Medical, Lisle, IL), and BrachD was measured using the DynaPulse 2000 (PulseMetric, San Diego, CA). Coefficients of variability from our lab are <7% for PWV measures and <9% for BrachD. Intraclass correlation coefficients for AIx are between 0.7 and 0.9. All measurements were conducted fasting and after the participant rested for 10 minutes. All prescriptions and over-the-counter medications were held on the day of testing until all tests were complete. TODAY staff conducting the arterial stiffness measurements were certified for performance of all procedures by a central Vascular Reading Center located in Cincinnati, Ohio.
A composite score to represent the arterial stiffness burden was constructed by calculating a separate z-score ((x-μ)/σ) for 3 assessments of PWV (carotid femoral, carotid radial, and femoral to foot), AIx, and BrachD using the mean and standard deviation from the TODAY cohort. This allowed scores to be combined to obtain a measure of composite (global) arterial stiffness. Z-scores were then summed and divided by the total number of measures assessed. The BrachD z-score value was reversed prior to summing since lower BrachD values are indicative of worsening. No other adjustments for directionality were needed since higher values for PWV and AIx are indicative of worse stiffness. Secondary analyses included calculating a composite arterial stiffness z-score based on the mean and standard deviation from published lean controls (12) using the same method as above.
Two-dimensional transthoracic echocardiograms were performed with the participant lying in a left lateral decubitus position to maximize image quality. Parasternal short axis, long axis, and apical views were obtained as previously described (1, 6). This allowed measurement of LV size and structure, tissue and pulsed Doppler imaging of right and LV inflow tracts, and for two dimensional images to allow later retrieval to obtain measurements of LV strain. Images were read offline by a single technician using commercially available software (Digisonics, Houston, TX). Measurements were made and abnormal thresholds were chosen according to the American Society of Echocardiography standards (13). The pulsed Doppler sample volume was placed at the tips of the mitral valve leaflets to measure LV inflow peak early flow velocity (E) simultaneously with each tissue Doppler measurement. The pulsed Doppler sample volume was placed at the junction of the mitral annulus and the septal left ventricular wall to measure septal peak Em. The pulsed Doppler sample volume was placed at the junction of the mitral annulus and the lateral left ventricular wall to measure lateral peak Em. Left atrial (LA) diameter, rather than LA area, was reported, as previous quality control studies in this cohort showed poorer reproducibility of LA area (unpublished data). Echocardiogram measurements of LV structure included LV mass, relative wall thickness, and LA size. Measurements of cardiac systolic function included LV ejection fraction, circumferential strain and LV peak longitudinal 2 and 4 chamber strain. Right ventricular (RV) function was assessed by tricuspid annular plane systolic excursion (TAPSE). Measurements of diastolic function were made using the average of the E/Em ratio. All speckle tracking and strain measurements were analyzed by a single technician with TomTec (Unterschleissheim, Germany) at a frame rate per second of 50 for global, radial, and regional myocardial deformation and strain, volumes, mass, and ejection fraction. For strain measurements, the intra-observer variability was <10%.
To test the secondary hypothesis that participants with the most adverse arterial stiffness or echocardiographic measures would have higher BP, BMI, and HbA1c, all participants were categorized into four risk groups based on values in the upper quintile of distribution of the vascular outcomes (composite score generated from the TODAY cohort), cardiac outcomes (combined LV mass and E/Em mean septal-lateral mitral valve) or both. The four groups were as follows, 1) vascular and echo risk: all values ≥80th percentile; 2) vascular risk: composite arterial stiffness score ≥80th percentile but not the echo values; 3) echo risk: echo values ≥80th percentile but not the composite score; and 4) no vascular or echo risk: no value ≥80th percentile.
Statistical Analyses
Descriptive statistics are presented as mean ± SD, median [25th–75th percentile] or percent. Multivariable linear regression models were used to evaluate the association between arterial stiffness (composite score as the independent variable) and cardiac structure (LV mass index, relative wall thickness and LA area), systolic function (LV ejection fraction, LV peak longitudinal 2 and 4 chamber strain, circumferential strain), RV function (TAPSE) and diastolic function (E/Em mean septal-lateral), the dependent variables assessed about 2 years later. Secondary analyses examined similar relationships as above using carotid femoral PWV. Unstandardized beta coefficients with corresponding p-values were derived from the linear regression models and the squared partial correlation coefficient (R2) was used to quantify the strength of the linear relationship between the variables.
All models were evaluated before and after adjustment for potential confounders; however, all included an adjustment for heart rate by default. Covariates included a mix of continuous (age and time weighted mean HbA1c, mean arterial BP, BMI), binary (sex, cigarette smoking) or categorical (race-ethnicity) variables. These risk factors were included in the models given cross-sectional associations with arterial stiffness and echocardiogram measurements in other studies (2, 11). Interaction terms between arterial stiffness and time-weighted mean HbA1c, mean arterial BP, and BMI were also added to the models in order to examine the moderating effects of these variables. Original TODAY intervention (metformin, metformin +lifestyle, metformin +rosiglitazone) was evaluated in initial analysis, but not retained in the models due to lack of effect.
Finally, to compare the means or proportions of covariates described above (HbA1c, sex, etc.) across the four groups of adverse risk phenotypes, overall tests of difference as well as pairwise comparisons were carried out using analysis of variance, and logistic or multinomial regression, as appropriate. Statistical significance was defined as p<0.05. The TODAY2 study was not powered to test the hypotheses proposed in these post-hoc secondary analyses. As a result, no adjustments were made for multiple testing.
Results
Demographic and metabolic variables at study randomization (enrollment), the time of the arterial stiffness measurements, and at the follow-up echocardiogram are presented in Table 1. The cohort was 63.9% female and 34% were self-reported Non-Hispanic Black, 40.2% Hispanic, 19.9% Non-Hispanic white and 5.9% reported other race- ethnicity group. The mean arterial stiffness measurements and 2-year follow-up echocardiogram measurements including cardiac structure, systolic and diastolic function are presented in Table 2.
Table 1.
Demographic and metabolic characteristics of the TODAY participants, n=388*
| Variable | At TODAY study randomization (n=388) | At Arterial Stiffness Assessment (n=388) | At Echocardiogram Assessment (n=388) |
|---|---|---|---|
| Duration in study (years) | n/a | 7.0 ± 1.3 | 9.1 ± 1.3 |
| Type 2 diabetes duration (years) | 0.6 ± 0.5 | 7.7 ± 1.5 | 9.7 ± 1.5 |
| Age (years) | 13.8 ± 2.0 | 20.8 ± 2.5 | 22.8 ± 2.5 |
| Smoking in the past month (%) | 3.1% | 21.0% | 22.9% |
| BMI (kg/m2) | 35.0 ± 7.8 | 36.8 ± 8.2 | 36.2 ± 8.3 |
| Systolic BP (mm Hg) | 112.4 ± 10.7 | 118.7± 12.1 | 120.4 ± 13.1 |
| Diastolic BP (mm Hg) | 66.3 ± 8.0 | 73.0 ± 9.8 | 75.4 ± 10.6 |
| Mean arterial BP (mm Hg) | 81.7 ± 8.1 | 88.2 ± 10.0 | 90.4 ± 10.9 |
| Heart rate (beats per minute) | n/a | 78.9 ± 12.1 | 75.8 ± 13.0 |
| On anti-hypertensive medications (%) | 5.7% | 43.3% | 33.3% |
| Total cholesterol (mg/dL) | 143.6 ± 28.2 | 166.4 ± 36.8 | 177.0 ± 48.3 |
| LDL cholesterol (mg/dL) | 83.2 ± 24.1 | 97.8 ± 31.3 | 98.0 ± 32.6 |
| HDL cholesterol (mg/dL) | 38.7 ± 8.9 | 42.5 ± 12.3 | 45.2 ± 12.6 |
| Triglycerides (mg/dL) | 91 [66 – 131] | 105 [69 – 155] | 114 [75 – 183] |
| Triglycerides to HDL-cholesterol ratio | 3.1 ± 2.3 | 3.9 ± 5.7 | 6.8 ± 37.7 |
| HbA1c (%) | 6.0 ± 0.8 | 8.8 ± 2.9 | 9.6 ± 3.1 |
Mean ± SD, median [25th – 75th percentile] or percent. n/a means not applicable.
Table 2.
Arterial stiffness and 2-year later echocardiogram measures in the TODAY participants, n=388
| Variable | Mean ± SD (n=388) |
|---|---|
| Arterial Stiffness Measures | |
| carotid femoral PWV (m/s) | 6.43 ± 1.57 |
| carotid radial PWV (m/s) | 7.74 ± 1.31 |
| femoral foot PWV (m/s) | 8.45 ± 2.60 |
| AIx (%) | 8.95 ± 10.90 |
| BrachD (mm/mmHg) | 6.02 ± 1.17 |
| Composite arterial stiffness z-score* | 0.003 ± 0.598 |
| Echocardiogram Measures | |
| Cardiac Structure | |
| LV mass/height2.7 (g/m2.7) | 38.80 ± 9.71 |
| LV relative wall thickness (cm) | 0.34 ± 0.06 |
| LA internal dimension (cm) | 3.68 ± 0.47 |
| LA area (cm2) | 18.22 ± 3.47 |
| Systolic Function | |
| LV ejection fraction (%) | 66.20 ± 6.21 |
| Longitudinal 4-chamber strain (%) | −19.47 ± 3.44 |
| Longitudinal 2-chamber strain (%) | −20.44 ± 3.83 |
| Circumferential strain (%) | −22.15 ± 4.40 |
| Radial strain (%) | 34.62 ± 21.49 |
| RV Function | |
| TAPSE (cm) | 2.25 ± 0.34 |
| Diastolic Function (mitral valve) | |
| Peak E with lateral Em (cm/s) | 88.62 ± 17.61 |
| Lateral Em (cm/s) | 14.31 ± 3.03 |
| Lateral E/Em (cm/s) | 6.48 ± 1.94 |
| Peak E with septal Em (cm/s) | 88.08 ± 17.55 |
| Septal Em (cm/s) | 10.75 ± 2.04 |
| Septal E/Em (cm/s) | 8.44 ± 2.06 |
| Mean septal-lateral E/Em (cm/s) | 7.48 ± 1.86 |
Composite arterial stiffness z-score represents an overall measure of arterial stiffness and is constructed by calculating a separate z-score ((x-μ)/σ) for the 3 PWV assessments and AIx and BrachD using the mean and standard deviation from the TODAY cohort. Z-scores were then summed and divided by the total number of measures assessed. The BrachD z-score value was reversed prior to summing since lower BrachD values are indicative of worsening.
In unadjusted models, higher composite arterial stiffness was associated with higher LV mass index (p<0.0001, Table 3). Higher composite arterial stiffness was also positively associated with relative wall thickness and LA diameter, but not left atrial area (all p<0.01; Supplemental Table 1). In a final model that included adjustment for demographic variables and time weighted mean BMI, mean arterial BP, and dysglycemia, composite arterial stiffness remained a significant independent predictor of LV mass index (p=0.005; Table 3) but not relative wall thickness or LA diameter (Supplemental Table 1).
Table 3.
Linear regression analysis of the effect of composite arterial stiffness z-score on echocardiographic measures assessed in unadjusted and adjusted models*
| Models | LV mass/height2.7 (g/m2.7) | Septal E/Em (cm/s) | Lateral E/Em (cm/s) | Mean septal-lateral E/Em (cm/s) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| β ± SE | P Value | R2 | β ± SE | P Value | R2 | β ± SE | P Value | R2 | β ± SE | P Value | R2 | |
| Unadjusted Model | ||||||||||||
| Composite arterial stiffness | 4.50 ± 0.89 | <.0001 | 7.3% | 0.41 ± 0.19 | 0.04 | 1.3% | 0.79 ± 0.18 | <.0001 | 5.5% | 0.60 ± 0.17 | 0.0006 | 3.6% |
| Adjusted Model | ||||||||||||
| Composite arterial stiffness | 2.41 ± 0.85 | 0.005 | 1.7% | 0.06 ± 0.21 | 0.76 | <0.1% | 0.45 ± 0.19 | 0.01 | 1.5% | 0.26 ± 0.18 | 0.15 | 0.6% |
| Age (year) | −0.34 ± 0.19 | 0.07 | 0.7% | 0.03 ± 0.05 | 0.52 | <0.1% | 0.01 ± 0.04 | 0.98 | <0.1% | 0.02 ± 0.04 | 0.71 | <0.1% |
| Female | −3.32 ± 1.05 | 0.002 | 2.2% | 0.47 ± 0.25 | 0.06 | 1.0% | 0.43 ± 0.22 | 0.06 | 1.0% | 0.46 ± 0.22 | 0.03 | 1.2% |
| Race-ethnicity | ||||||||||||
| Non-Hispanic Black | 3.93 ± 1.32 | 0.003 | 2.3% | 0.10 ± 0.31 | 0.74 | <0.1% | 0.24 ± 0.28 | 0.39 | 1.3% | 0.20 ± 0.27 | 0.46 | 0.5% |
| Hispanic | 2.26 ± 1.30 | 0.08 | 0.04 ± 0.31 | 0.91 | 0.04 ± 0.27 | 0.87 | 0.06 ± 0.27 | 0.81 | ||||
| Cigarette smoking | 1.98 ± 1.09 | 0.07 | 0.7% | −0.22 ± 0.26 | 0.39 | 0.2% | 0.16 ± 0.23 | 0.48 | 0.1% | −0.02 ± 0.23 | 0.96 | <0.1% |
| Mean HbA1c (%)† | −0.26 ± 0.26 | 0.31 | 0.2% | 0.10 ± 0.06 | 0.10 | 0.8% | 0.04 ± 0.05 | 0.43 | 0.2% | 0.07 ± 0.05 | 0.17 | 0.5% |
| Mean arterial BP (mm Hg)† | 0.05 ± 0.08 | 0.49 | 0.1% | 0.02 ± 0.02 | 0.22 | 0.4% | 0.03 ± 0.02 | 0.06 | 0.9% | 0.02 ± 0.02 | 0.13 | 0.6% |
| Mean BMI (kg/m2)† | 0.53 ± 0.07 | <.0001 | 14.0% | 0.06 ± 0.02 | <.0001 | 4.6% | 0.06 ± 0.01 | <.0001 | 5.2% | 0.06 ± 0.01 | <.0001 | 6.0% |
Composite arterial stiffness z-score from the arterial stiffness assessment (~study year 7) is related to echocardiography measures evaluated about 2 years later (~study year 9), before and after adjustment for potential confounders, in linear regression models. Both unadjusted and adjusted models are minimally adjusted for heart rate (beats per minute). Model coefficient estimates (β) ± standard error (SE), P-values, and partial R-square (R2) are presented for each characteristic. The partial R2 corresponds to the proportion of explained variance by the characteristic in the model. Cigarette smoking (defined as smoking in the past month), HbA1c, mean arterial BP, and BMI are characteristics evaluated at the time of the echocardiogram assessment. Reference groups are male for sex and non-Hispanic Whites for race-ethnicity.
Time-weighted mean. P-values<0.05 are bolded.
Composite arterial stiffness was not associated with LV ejection fraction, LV peak longitudinal 2 or 4 chamber strain, circumferential or radial strain in adjusted models (Supplemental Table 1). Similarly, composite arterial stiffness was not related to TAPSE. Collectively, after accounting for potential confounders, arterial stiffness was not related to measures of LV systolic or RV function two years later.
As shown in Table 3, composite arterial stiffness was positively associated with the three measures of LV diastolic function, as assessed by septal E/Em, lateral E/Em, and their average, mean septal-lateral E/Em in unadjusted models (all p<0.05). After adjustment for risk factors, the relationships were attenuated by cumulative (time weighted mean) BMI and composite arterial stiffness and remained significant with lateral E/Em only (p=0.01), driven by the association between Em with arterial stiffness.
Significant interactions were seen between composite arterial stiffness and mean arterial BP overtime in relation to lateral E/Em (p=0.0003) and mean septal-lateral E/Em (p=0.0005). To illustrate the interactions, a median split was applied to mean arterial BP with the relationships between composite arterial stiffness with lateral E/Em and mean septal-lateral E/Em plotted for two subgroups of equal size (mean arterial BP ≥ 90 mm vs < 90 mm Hg; Supplemental Figure 1). The interaction shows a greater adverse effect of composite arterial stiffness on diastolic outcomes (i.e., more positive association) among participants with higher levels of mean arterial BP overtime. No interactions were seen between composite arterial stiffness and time-weighted mean HbA1c or BMI.
Analyses were repeated using a composite arterial stiffness measure derived from published lean controls (12). This resulted in a composite z score that was shifted from a mean of 0.003 to 0.817 demonstrating increased stiffness in the TODAY cohort vs lean controls, but the relationships were essentially unchanged. Composite arterial stiffness was associated with slightly lower lateral E/Em in adjusted models (beta coefficient ± SE in adjusted model 0.27 ± 0.14, p=0.06).
Given carotid femoral PWV is a widely used arterial stiffness measure, we also examined the relationship between carotid femoral PWV (as opposed to composite arterial stiffness) and LV structure, systolic and diastolic function as above (Supplemental Tables 1 and 2). Supplemental Table 2 shows carotid femoral PWV associated with lateral E/Em and mean septal-lateral E/Em in unadjusted models, but after adjustment remained positively associated with lateral E/Em only (p=0.02). Similar to the composite arterial stiffness results, BMI attenuated the relationship for mean septal-lateral E/Em. A significant interaction was also observed between carotid femoral PWV and time weighted mean arterial BP in relation to lateral E/Em (p=0.01), suggestive of stronger association between arterial stiffness and diastolic function at higher BP levels.
Finally, we examined the clinical characteristics (evaluated at the time of the follow-up echocardiogram assessment) of participants with vascular stiffness, LV mass index or diastolic function in the upper quintile of the TODAY distribution in separate unadjusted models (Table 4). Participants with the highest stiffness, highest LV mass and worse diastolic function were more likely to be non-Hispanic black, have a higher time weighted mean BMI, systolic and diastolic BP, and more likely to be on anti-hypertensive medications. Cumulative exposure to HbA1c (time weighted mean from randomization to the echocardiogram assessment) was not different between the groups.
Table 4.
Characteristics at the time of the follow-up echocardiogram by groups of adverse clinical risk phenotypes*
| Characteristics at time of follow-up echocardiogram (n=350) | Echo and Vascular Risk (both in upper quintile) N=32, 9.1% | Vascular Risk (only vascular in upper quintile) N=91, 26.0% | Echo Risk (only echo in upper quintile) N=41, 11.7% | No Echo or Vascular Risk (none in upper quintile) N=186, 53.1% | P value |
|---|---|---|---|---|---|
| Duration in study (years) | 8.9 ± 1.4 | 9.2 ± 1.1 | 9.0 ± 1.4 | 9.0 ± 1.4 | 0.66 |
| Type 2 diabetes duration (years) | 9.5 ± 1.5 | 9.9 ± 1.3 | 9.7 ± 1.6 | 9.6 ± 1.5 | 0.31 |
| Age (years) | 22.5 ± 2.3 | 23.2 ± 2.2 | 23.3 ± 2.5 | 22.5 ± 2.6 | 0.06 |
| Female (%) | 71.9% | 60.4% | 70.7% | 66.7% | 0.53 |
| Non-Hispanic White | 9.7% | 16.7% | 27.5% | 24.1% | |
| Smoking in the past month (%) | 25.0% | 28.6% | 19.5% | 21.5% | 0.55 |
| Mean BMI (kg/m2)† | 42.9 ± 7.9 | 39.8 ± 7.9 | 36.6 ± 8.9 | 33.9 ± 5.8 | <.0001 a,b,c |
| Mean systolic BP (mm Hg)† | 123.4 ± 9.2 | 118.8 ± 7.6 | 120.2 ± 9.4 | 115.0 ± 8.8 | <.0001 a,c,d |
| Mean diastolic BP (mm Hg)† | 74.8 ± 7.5 | 71.4 ± 6.0 | 74.3 ± 7.9 | 69.9 ± 6.8 | <.0001 a,d |
| Mean arterial BP (mm Hg)† | 91.0 ± 7.7 | 87.2 ± 6.1 | 89.6 ± 8.0 | 84.9 ± 7.1 | <.0001 a,d |
| Heart rate (beats per minute) | 81.3 ± 8.8 | 76.2 ± 13.0 | 83.9 ± 12.3 | 79.2 ± 12.0 | 0.009 e |
| On anti-hypertensive medications (%) | 56.3% | 42.9% | 34.2% | 25.8% | 0.001 a,c |
| Mean total cholesterol (mg/dL)† | 163.4 ± 30.8 | 159.2 ± 27.8 | 160.2 ± 33.1 | 162.9 ± 32.3 | 0.94 |
| Mean LDL cholesterol (mg/dL)† | 90.5 ± 24.9 | 93.3 ± 23.5 | 91.7 ± 28.9 | 93.1 ± 25.7 | 0.63 |
| Mean HDL cholesterol (mg/dL)† | 41.9 ± 9.2 | 40.2 ± 8.6 | 41.9 ± 8.4 | 43.4 ± 9.3 | 0.06 |
| Mean triglycerides (mg/dL)† | 105 [78–192] | 100 [75–166] | 112 [80–157] | 113 [72–161] | 0.92 |
| Mean triglycerides to HDL-cholesterol ratio† | 5.8 ± 9.7 | 3.9 ± 3.7 | 4.7 ± 9.6 | 3.6 ± 3.5 | 0.16 |
| Mean HbA1c (%)† | 8.8 ± 2.1 | 7.9 ± 1.9 | 7.9 ± 1.8 | 8.0 ± 2.1 | 0.17 |
Mean ± SD, median [25th – 75th percentile] or percent. Based on data for n=350 participants; 48 participants could not be classified into one of the four adverse risk groups due to missing data for at least one of the outcomes used in the categorization. Categorization into the four groups was based on participants having at least one, both or neither vascular (composite score) and cardiac (LV mass adjusted for height or mean septal-lateral E/Em) values in the upper quintile (≥80% percentile within our population). Overall tests of difference comparing measurements at the time of the echocardiogram between the four groups of adverse risk phenotypes were carried out using analysis of variance (continuous factors), and logistic (binary factors; e.g., sex) or multinomial regression (categorical factors; e.g., race-ethnicity).
Time-weighted mean. Significant pairwise comparisons (P<0.05) are denoted by superscripts if overall difference across the four risk groups is significant:
risk of both vs. no risk;
risk of both vs. echo risk only;
vascular risk only vs. no risk;
echo risk only vs. no risk; and
vascular risk only vs. echo risk only.
Discussion
The main impact of higher arterial stiffness in youth-onset T2D on echocardiographic outcomes assessed 2 years later is an adverse change in LV diastolic function, as reflected by tissue Doppler assessment. This is important as it suggests that the vascular injury associated with T2D has a downstream effect on myocardial function, particularly diastolic function. Diastolic dysfunction is an important late consequence of long standing T2D, resulting in heart failure with preserved ejection fraction (14). This adverse effect was more pronounced among participants with higher mean arterial BP. LV systolic function and RV function were not impacted over the two-year follow-up interval. This study is novel in three aspects: the cohort is young, it is a high-risk population all with youth-onset T2D, and assess the relationship between arterial stiffness and follow-up cardiac structure and function building on prior data that is only cross-sectional. These results suggest that the path to future heart failure may begin early in life in this setting of youth-onset type 2 diabetes.
Prior studies in adolescents relating arterial stiffness with cardiac structure and function have been cross-sectional (12). In a similar age cohort, where data were stratified by high and normal arterial stiffness, those in the high PWV group were more likely to have higher LV mass and LV relative wall thickness as well as higher BP, BMI, and HbA1c (5). Cross-sectional work in adults also shows higher arterial stiffness associated with adverse changes in diastolic function; however, only stiffness in the brachial artery was assessed which may not represent stiffness in other arterial beds (15). In older adults, on average about 30 years older than the cohort described in our study, arterial stiffness was also associated with cardiac deformation as assessed by speckle tracking echocardiography (16). Collectively, these and other data suggest the major impact of higher arterial stiffness is on late systole and early diastole with slowing of late contraction and early filling (17), findings associated with future heart failure (15). However, here we show these relationships exist within 10 years of developing youth-onset T2D and arterial stiffness adversely impacts LV structure and function 2 years later.
Adiposity has been independently associated with higher stiffness in youth with T2D (12). Furthermore, we and others have previously reported progression of LV mass and diastolic dysfunction in the TODAY cohort associated with obesity (6). Given that arterial stiffness is a very early manifestation of vascular injury in diabetes, progression of LV mass and deterioration in LV diastolic function in youth with obesity and T2D is likely exacerbated by the adverse effects of diabetes on the vasculature. The TODAY study has reported a high rate of incident cardiovascular events in this cohort, including heart failure and atrial dysrhythmias, which are likely consequences of associations reported herein (18). The TODAY clinical trial was a treatment intervention trial. While no effect of treatment arm was seen in these analyses, the observed change in BMI in the lifestyle intervention arm (pre and post TODAY clinical trial BMI was 34 kg/m2) was minimal. Based on the data presented here (specifically in the models and interactions) and those previously described, clinical trials to prevent subclinical cardiovascular injury focused on weight management and BP should still be considered.
Afterload on the heart is impacted by reflected waves from the peripheral arterial tree with each heartbeat. The major impact is on late systolic workload. In older adults, maladaptive changes in the LV include LV hypertrophy and fibrosis (16, 19). The finding that the Em component of the tissue Doppler ratio drove the association is consistent with early stiffening of the left ventricle in these patients. In animal models, LV dysfunction has been demonstrated with higher arterial stiffness (20, 21). In this study of younger individuals, LV mass index and LV diastolic function were primarily impacted. These primary findings may be a result of the study design, with echocardiograms performed two years after arterial stiffness assessment; diastolic function and bigger hearts are likely downstream consequences of the increased afterload on the heart.
Strengths of this study include the well-phenotyped T2D cohort allowing the impact of chronic risk exposure over several years to be incorporated into statistical models. Echocardiography and arterial stiffness measures were collected using standardized protocols with data analysis at experienced central reading centers with strong quality control procedures. Finally, in addition to measuring carotid femoral PWV, we also obtained additional measures of stiffness reflecting a composite measure of the arterial tree.
Weaknesses of the study include a relatively narrow window of time between arterial stiffness and echocardiography assessment. The severely obese nature of this cohort limited image quality, which impacted data collection, particularly for speckle tracking and area measurements. Additionally, diastolic function assessment in this study was limited to the measures reported as current guidelines for diastolic function assessment were published after development of the echocardiography protocol for this study. The arterial stiffness and echocardiography measurements were taken post TODAY trial intervention limiting the ability to determine the effect of treatment in the trial on these outcomes. The standardization of values allowed the derivation of an unweighted arterial stiffness composite score but did not provide a summary measure maximizing the variance of the data. Furthermore, as these analyses are post-hoc, some of the statistical associations reported may occur by chance.
Conclusion
This study documents the impact of arterial stiffness as well as elevated BP and obesity on subsequent adverse changes in LV diastolic function in a cohort with T2D diagnosed in adolescence. The association of higher arterial stiffness with future left ventricular diastolic dysfunction suggests the path to future heart failure may begin early in life in this setting of youth-onset type 2 diabetes. Clinical trials focused on correction of obesity and hypertension are needed to test whether the course to cardiovascular complications can be altered in patients with T2D.
Supplementary Material
Highlights.
Youth-onset type 2 diabetes increases the risk for cardiovascular disease (CVD).
We found artery stiffening is associated with worse diastolic function 2 years later.
This relationship was exacerbated by higher blood pressure.
In this young cohort, CVD is observed with effects on the heart and vessels.
Acknowledgements
A complete list of individuals in the TODAY Study Group is presented in the Appendix of the Supplemental Material.
The TODAY Study Group thanks the following companies for donations in support of the study’s efforts: Becton, Dickinson and Company; Bristol-Myers Squibb; Eli Lilly and Company; GlaxoSmithKline; LifeScan, Inc.; Pfizer; Sanofi Aventis. We also gratefully acknowledge the participation and guidance of the American Indian partners associated with the clinical center located at the University of Oklahoma Health Sciences Center, including members of the Absentee Shawnee Tribe, Cherokee Nation, Chickasaw Nation, Choctaw Nation of Oklahoma, and Oklahoma City Area Indian Health Service; the opinions expressed in this paper are those of the authors and do not necessarily reflect the views of the respective Tribes and the Indian Health Service.
Funding
This work was completed with funding from NIDDK and the NIH Office of the Director through grants U01-DK61212, U01-DK61230, U01-DK61239, U01-DK61242, and U01-DK61254. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The NIDDK project office was involved in all aspects of the study, including: design and conduct; collection, management, analysis, and interpretation of the data; review and approval of the manuscript; and decision to submit the manuscript for publication.
List of abbreviations
- AIx
augmentation index
- BP
blood pressure
- BMI
body mass index
- BrachD
brachial distensibility
- CBL
Central Biochemistry Laboratory
- HbA1c
hemoglobin A1c
- HDL
high-density lipoprotein
- LA
left atrial
- LDL
low-density lipoprotein
- LV
left ventricular
- PWV
pulse wave velocity
- RV
right ventricular
- TAPSE
tricuspid annular plane systolic excursion
- T2D
type 2 diabetes
- TODAY
Treatment Options for type 2 Diabetes in Adolescents and Youth study
- TODAY2
Treatment Options for type 2 Diabetes in Adolescents and Youth observational follow-up study
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Ethics approval and consent to participate
The study was approved by institutional review boards at all 15 centers and all participants and guardians provided written informed assent and/or consent as appropriate for age and local guidelines.
Consent for publication
The authors consent to the publication of this article.
Competing interests
The authors declare that they have no competing interests to report.
Availability of data and materials
All data and materials are available upon request.
References
- 1.Levitt Katz L, Gidding SS, Bacha F, Hirst K, McKay S, Pyle L, et al. Alterations in left ventricular, left atrial, and right ventricular structure and function to cardiovascular risk factors in adolescents with type 2 diabetes participating in the TODAY clinical trial. Pediatr Diabetes. 2015:16:39–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Shah AS, El Ghormli L, Gidding SS, Bacha F, Nadeau KJ, Levitt Katz LE, et al. Prevalence of arterial stiffness in adolescents with type 2 diabetes in the TODAY cohort: Relationships to glycemic control and other risk factors. J Diabetes Complications. 2018:32:740–745. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Urbina EM, Gidding SS, Bao W, Pickoff AS, Berdusis K, Berenson GS. Effect of body size, ponderosity, and blood pressure on left ventricular growth in children and young adults in the Bogalusa Heart Study. Circulation. 1995:91:2400–6. [DOI] [PubMed] [Google Scholar]
- 4.Ikonomidis I, Aboyans V, Blacher J, Brodmann M, Brutsaert DL, Chirinos JA, et al. The role of ventricular-arterial coupling in cardiac disease and heart failure: assessment, clinical implications and therapeutic interventions. A consensus document of the European Society of Cardiology Working Group on Aorta & Peripheral Vascular Diseases, European Association of Cardiovascular Imaging, and Heart Failure Association. Eur J Heart Fail. 2019:21:402–424. [DOI] [PubMed] [Google Scholar]
- 5.Urbina EM, Dolan LM, McCoy CE, Khoury PR, Daniels SR, Kimball TR. Relationship between elevated arterial stiffness and increased left ventricular mass in adolescents and young adults. J Pediatr. 2011:158:715–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.TODAY Study Group. Longitudinal Changes in Cardiac Structure and Function From Adolescence to Young Adulthood in Participants With Type 2 Diabetes Mellitus: The TODAY Follow-Up Study. Circ Heart Fail. 2020:13:e006685. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ryder JR, Northrop E, Rudser KD, Kelly AS, Gao Z, Khoury PR, et al. Accelerated Early Vascular Aging Among Adolescents With Obesity and/or Type 2 Diabetes Mellitus. J Am Heart Assoc. 2020:9:e014891. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Zeitler P, Hirst K, Pyle L, Linder B, Copeland K, Arslanian S, et al. A clinical trial to maintain glycemic control in youth with type 2 diabetes. N Engl J Med. 2012:366:2247–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.TODAY Study Group. Rapid rise in hypertension and nephropathy in youth with type 2 diabetes: the TODAY clinical trial. Diabetes Care. 2013:36:1735–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Zeitler P, Epstein L, Grey M, Hirst K, Kaufman F, Tamborlane W, et al. Treatment options for type 2 diabetes in adolescents and youth: a study of the comparative efficacy of metformin alone or in combination with rosiglitazone or lifestyle intervention in adolescents with type 2 diabetes. Pediatric diabetes. 2007:8:74–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Shah AS, El Ghormli L, Vajravelu ME, Bacha F, Farrell RM, Gidding SS, et al. Heart Rate Variability and Cardiac Autonomic Dysfunction: Prevalence, Risk Factors, and Relationship to Arterial Stiffness in the Treatment Options for Type 2 Diabetes in Adolescents and Youth (TODAY) Study. Diabetes Care. 2019:42:2143–2150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Urbina EM, Kimball TR, Khoury PR, Daniels SR, Dolan LM. Increased arterial stiffness is found in adolescents with obesity or obesity-related type 2 diabetes mellitus. J Hypertens. 2010:28:1692–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr. 2015:28:1–39.e14. [DOI] [PubMed] [Google Scholar]
- 14.Jia G, DeMarco VG, Sowers JR. Insulin resistance and hyperinsulinaemia in diabetic cardiomyopathy. Nature reviews Endocrinology. 2016:12:144–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Chirinos JA, Segers P, Hughes T, Townsend R. Large-Artery Stiffness in Health and Disease: JACC State-of-the-Art Review. Journal of the American College of Cardiology. 2019:74:1237–1263. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Chirinos JA, Segers P, Rietzschel ER, De Buyzere ML, Raja MW, Claessens T, et al. Early and late systolic wall stress differentially relate to myocardial contraction and relaxation in middle-aged adults: the Asklepios study. Hypertension (Dallas, Tex : 1979). 2013:61:296–303. [DOI] [PubMed] [Google Scholar]
- 17.Vinereanu D, Nicolaides E, Boden L, Payne N, Jones CJ, Fraser AG. Conduit arterial stiffness is associated with impaired left ventricular subendocardial function. Heart. 2003:89:449–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Bjornstad P, Drews KL, Caprio S, Gubitosi-Klug R, Nathan DM, Tesfaldet B, et al. Long-Term Complications in Youth-Onset Type 2 Diabetes. N Engl J Med. 2021:385:416–426. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zamani P, Bluemke DA, Jacobs DR Jr., Duprez DA, Kronmal R, Lilly SM, et al. Resistive and pulsatile arterial load as predictors of left ventricular mass and geometry: the multi-ethnic study of atherosclerosis. Hypertension. 2015:65:85–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kobayashi S, Yano M, Kohno M, Obayashi M, Hisamatsu Y, Ryoke T, et al. Influence of aortic impedance on the development of pressure-overload left ventricular hypertrophy in rats. Circulation. 1996:94:3362–8. [DOI] [PubMed] [Google Scholar]
- 21.Gillebert TC, Lew WY. Influence of systolic pressure profile on rate of left ventricular pressure fall. Am J Physiol. 1991:261:H805–13. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All data and materials are available upon request.