OBJECTIVE

Insulin resistance (IR) and dysglycemia can induce cardiac remodeling in adulthood, but little evidence exists with respect to cardiac remodeling in youth with and without evidence of new-onset glucose metabolic alterations. This study investigated whether changes in metabolic status from adolescence to young adulthood are associated with the risk of progressive cardiac remodeling and examined potential mechanistic pathways.

RESEARCH DESIGN AND METHODS

From the Avon Longitudinal Study of Parents and Children (ALSPAC), U.K. cohort, 1,595 adolescents, mean (SD) age 17.7 (0.4) years, who had data on fasting plasma glucose and insulin levels, and echocardiography left ventricular (LV) mass indexed for height raised to the power of 2.7 (LVMI2.7) and in whom these factors repeatedly were measured at a clinic visit when they were aged 24 years were included. HOMA-IR was computed, hyperglycemia was defined as glucose concentration of ≥5.6 mmol/L and ≥6.1 mmol/L, and LV hypertrophy was defined as LVMI2.7 ≥51g/m2.7.

RESULTS

The prevalence of LV hypertrophy increased from 2.4% at baseline to 7.1% at follow-up. Each unit increase of glucose (β = 0.37 g/m2.7 [95% CI 0.23–0.52]; P < 0.001) and HOMA-IR (1.10 g/m2.7 [0.63–1.57]; P < 0.001) was independently associated with increased LVMI2.7 over 7 years. Persistent hyperglycemia of 5.6 mmol/L and 6.1 mmol/L was associated with higher odds (odds ratio [OR] 1.46 [95% CI 1.35–1.47], P < 0.001; and 3.10 [95% CI 1.19–8.08], P = 0.021, respectively) of worsening LV hypertrophy over 7 years. Increased fat mass (62% mediation) significantly mediated the association of increased HOMA-IR with increased LVMI2.7.

CONCLUSIONS

Persistent adolescent hyperglycemia and worsening IR were associated with the risk of worsening structural and functional cardiac damage, and these were largely explained by increased fat mass.

Insulin resistance (IR) and dysglycemia are precursors of young-onset type 2 diabetes and have been associated with cardiac and vascular alterations in the presence of young-onset type 2 diabetes (1–5). We recently observed that fat mass in mid-adolescence may be unidirectionally associated with later IR whereas in late adolescence and young adulthood, fat mass and IR are likely bidirectionally reinforcing, with three-fourths of the vicious pathologic cycle driven by fat mass and one-fourth by IR (6). Despite the increasing incidence of young-onset type 2 diabetes globally (3,4), the U.S. Preventive Services Task Force deemed there was insufficient evidence to assess the benefits of screening for type 2 diabetes in children and adolescents and requested further evidence on cardiovascular outcomes in the young population (7).

A key evidence point is whether IR- and dysglycemia-induced cardiac remodeling exist in apparently healthy adolescents. This point remains unexamined due to the scarcity of large-scale and long-term studies with repeated echocardiography assessments (5,8). Other competing risk factors for premature cardiac remodeling in apparently healthy youth are elevated lipid levels, blood pressure, obesity, tobacco smoking, and sedentary behavior, which could co-occur with glycemic traits (9–12). Therefore, longitudinal data to examine discordant changes in glycemic versus coexisting factors on changes in subclinical cardiac structure are critically necessary to determine the precursors of atherosclerotic cardiovascular events (13,14).

Using data from the Avon Longitudinal Study of Parents and Children (ALSPAC) birth cohort (England, U.K.), this study examined 1) the longitudinal associations of persistent hyperglycemia and increased IR with the risk of changes in cardiac structure and function and premature progressive cardiac damage from ages 17–24 years, and 2) the mediating roles of lipids, high-sensitivity C-reactive protein (hsCRP), systolic blood pressure, total body fat mass, and lean mass in the associations of dysglycemia and IR with cardiac changes.

Details of the ALSPAC birth cohort have been published (15–17). Data from the ALSPAC birth cohort are used to investigate factors that influence childhood development and growth. Altogether, 14,541 pregnancies of women residing in Avon, southwestern England, U.K., who had a total of 14,676 fetuses were enrolled between 1 April 1991 and 31 December 1992. Of these, 13,988 children were alive at 1 year of age. From age 7 years, an attempt was made to bolster the initial sample with eligible cases who did not join the study originally. The phases of enrolment are described in more detail in the cohort profile report and its update (15,16). The total sample size from the age of 7 years, therefore, is 15,447 pregnancies, resulting in 15,658 fetuses, of which 14,901 children were alive at 1 year of age. Annual face-to-face clinics commenced from the age of 7 years and are ongoing. Altogether, 2,079 participants had cardiac structural measures at the age 17-year clinic visit and 1,957 participants had cardiac measures at 24 years (Supplementary Fig. 1).

For the present analysis, 1,595 participants who had complete cardiac and blood sample measures at age 24 years were included, approximating 82% of participants with cardiac measures at the age 24-year clinic visit. The characteristics of the included 1,595 participants were similar those of the excluded 363 participants (19%) (Supplementary Table 1).

Ethical approval for the study was obtained from the ALSPAC Ethics and Law Committee and the Local Research Ethics Committees. Informed consent for the use of data collected via questionnaires and clinics was obtained from participants, following the recommendations of the ALSPAC Ethics and Law Committee at the time. Study data at 24 years were collected and managed using REDCap electronic data capture tools (18).

Anthropometric, Cardiometabolic, and Lifestyle Factors Measures

Anthropometry (height and weight) of participants at ages 17 and 24 years was assessed by observing standard protocols, and BMI was computed as weight in kilograms per height in meters squared. Overweight/obesity was categorized as BMI >24.99 kg/m2. Heart rate and systolic and diastolic blood pressures were measured at ages 17 and 24 years using an Omron 705-IT monitor, as previously detailed (17,19). Fasting blood samples at ages 17 and 24 years were collected, spun, and frozen at –80°C, and later assayed for glucose and insulin, as detailed previously (17,19–21). Fasting insulin was measured using an ultrasensitive, automated microparticle enzyme immunoassay (Mercodia), which does not cross-react with proinsulin; the sensitivity of the immunoassay was 0.07 mU/L (8,19). Participants with fasting glucose levels >5.6 mmol/L and fasting insulin levels >11.78 mU/L (∼85th percentile) were categorized as at risk for hyperglycemia and hyperinsulinemia (8,19,22). We also applied a glucose cut point of 6.1 mmol/L to reduce misclassifications, because >70% of adolescents in our cohort who were classified has having prediabetes at age 15 years when the population glucose concentration was high returned to normoglycemia by age 17 years (8). Nearly 98% of our adolescents at age 15 years had attained puberty, and all adolescents attained puberty by age 17 years (8). The HOMA-IR was calculated as (fasting insulin × fasting glucose/22.5) (23). The >75th percentile of HOMA-IR was categorized as high or elevated IR. Persistent hyperglycemia, hyperinsulinemia, and high IR were described as having values above the cut points at both time periods (adolescence and young adulthood).

Fasting plasma triglyceride, HDL cholesterol (HDL-c), LDL cholesterol (LDL-c), and hsCRP levels were assessed at both the 17- and 24-year clinic visits, in line with standard protocols (17,19,20). Total fat mass and lean mass were assessed using a DEXA scanner (GE Medical Systems, Madison, WI) at 17-year and 24-year clinic visits. Questionnaires to assess smoking behavior were administered at the 17-year and 24-year clinic visits. A specific question regarding whether participants smoked in the past 30 days was used as an indicator of current smoking status (11). At the 17-year clinic visit, participants were briefly asked about their personal and family (mother, father, and siblings) medical history, such as a history of hypertension, diabetes, high cholesterol, and vascular disease. The participant’s mother’s socioeconomic status was grouped according to the 1991 British Office of Population and Census Statistics classification (19,24). Sedentary time, light physical activity, and moderate to vigorous physical activity were assessed with waist-worn ActiGraph accelerometer devices for seven consecutive days at the 15-year clinic visits and four consecutive days at the age 24 years clinic visits, with valid wear days including at least two weekdays and one weekend day (8,19,24–26).

Cardiac Structure and Function Measures

At mean (SD) age 17.7 (0.4) years clinic visit, echocardiography was performed according to American Society of Echocardiography guidelines (13) by one of two experienced echocardiographers using an HDI 5000 ultrasound machine (Phillips Healthcare, Amsterdam, the Netherlands) equipped with a P4‐2 Phased Array ultrasound transducer. At mean (SD) age 24.4 (0.6) years clinic visit, echocardiography was performed by two experienced echocardiographers using a Philips EPIQ 7G Ultrasound System equipped with a X5-1 transducer. The Philips Q-station was used for the M-mode, two-dimensional, and Doppler echo analyses, and TomTec software was used for the three-dimensional echo analyses. Measures of cardiac structure were left ventricular mass index for height raised to the power of 2.7 (LVMI2.7) and relative wall thickness (RWT) computed from septal wall thickness, posterior wall thickness, and LV diastolic diameter (LVDD). Measures of cardiac function were LVD function (LVDF) early (E) to atrial (A) wave ratio and LV filling pressure (LVFP) E/é wave ratio (the ratio of the velocities of early diastolic mitral inflow [the E-wave] to early diastolic mitral annual tissue [e′-wave]) (14,27). Pulsed Doppler examination of transmitral flow was recorded from the apical four-chamber view. For LV measurements, the sample volume was positioned between the mitral annulus and the tips of the mitral leaflets, with the position adjusted to maintain the sample volume at an angle as near parallel to transmitral flow as possible with the participant in passive end-expiration (13). The peak flow velocity of the E and A waves were measured from the three consecutive cardiac cycles displaying the highest measurable velocity profiles (13). A similar measurement (e′) was conducted at the tricuspid valve. Tissue Doppler echocardiography was performed in the four-chamber view on the lateral, inferior, and septal LV walls to obtain myocardial wall velocities. Data were acquired with the beam parallel to the wall of interest and with optimal settings to ensure no over-gain of the low-velocity signals. A 5-mm sample volume was placed at the level of the mitral valve annulus and a loop of 8–10 cardiac cycles was recorded. The reproducibility of echocardiographic examinations was assessed by recalling 30 participants and repeating their measurements. The intraclass correlation of repeated measurements ranged from 0.75 to 0.93 (intraobserver) and 0.78 to 0.93 (interobserver) (27). LV hypertrophy was defined as LVMI2.7 ≥51 g/m2.7 because that predicts hard cardiovascular events, increased RWT was defined as ≥0.44, LVDD as E/A <1.5, and increased LVFP (hiLVFP) as E/e′ ≥8 (14).

Statistical Analysis

The participants’ descriptive characteristics were summarized as means and SDs, medians and interquartile ranges, or frequencies and percentages. Change in variables over time was tested with a paired t test for normally distributed continuous variables, Wilcoxon signed-rank test for skewed continuous variables, and McNemar change tests for categorical variables.

The separate associations of the 7-year changes in glucose, insulin, and HOMA-IR values with each of the progressive changes in LVMI2.7, RWT, LVDF, and LVFP were examined using generalized linear mixed-effect models for repeated measures. Model 1 was unadjusted. Model 2 was adjusted for sex and other time-varying covariates measured at both baseline and follow-up such as age; hsCRP; heart rate; systolic blood pressure; fat mass; lean mass; smoking status; family history of hypertension, diabetes, high cholesterol, or vascular disease; socioeconomic status; sedentary time; light physical activity; moderate to vigorous physical activity; LDL-c, HDL-c, triglyceride, glucose, or insulin levels, depending on the predictor. When HOMA-IR was the predictor, there were no further adjustments for glucose and insulin. The persistence of altered metabolic status from adolescence through young adulthood predicting the odds of cardiac damage progression was examined with fully adjusted, generalized, linear mixed-effect models with logit links. Sex-based analyses are presented in Supplementary Table 2 and Supplementary Fig. 2.

Mediating path analyses using structural equation models separately examined the mediating role of each change in LDL-c, triglyceride, hsCRP, systolic blood pressure, total body fat mass, and lean mass on the longitudinal associations of each change in glucose and HOMA-IR with the progression in LVMI2.7. The mediation analysis was conducted in line with the Guideline for Reporting Mediation Analyses of Randomized Trials and Observational Studies (AGReMA) (28). The examined mediation mechanism between metabolic and cardiac indices is partly based on previous studies in which worsening cardiometabolic profiles were associated with cardiac diseases (10,14,29). Analyses were adjusted for sex, family history of hypertension, diabetes, high cholesterol, or vascular disease; socioeconomic status; and time-varying covariates measured at both baseline and follow-up, such as age; heart rate; smoking status; sedentary time; light physical activity; moderate to vigorous physical activity; HDL-c, LDL-c, and triglyceride levels; fat mass; lean mass; and hsCRP level; in addition to insulin, depending on the predictor or mediator.

The path models had three equations per regression analysis: the longitudinal associations of change in glucose or HOMA-IR with change in LDL-c, triglyceride, and hsCRP levels, systolic blood pressure, and total body fat mass or lean mass (Eq. 1); the longitudinal associations of change in LDL-c, triglyceride, and hsCRP levels, systolic blood pressure, and total body fat mass or lean mass with LVMI2.7 (Eq. 2); and the longitudinal associations of change in glucose or HOMA-IR with change in LVMI2.7 (Eq. 3; total effect). Equation 3′ (direct effect) accounted for the mediating role of change in LDL-c, triglyceride, and hsCRP levels; systolic blood pressure; total body fat mass; or lean mass on the longitudinal associations of change in glucose or HOMA-IR with changes in LVMI2.7. The proportion of mediating or suppressing roles was estimated as the ratio of the difference between Eq. 3 and Eq. 3′ or the multiplication of Eqs. 1 and 2 divided by Eq. 3 and expressed as a percentage.

A mediating or indirect role is confirmed when there are statistically significant associations between the predictor and mediator, the predictor and outcome, the mediator and outcome, and when the longitudinal association between the predictor and outcome variable was attenuated upon inclusion of the mediator (30). However, when the magnitude of the longitudinal association between the predictor and outcome is increased upon inclusion of a third variable, a suppression is confirmed (30). This means that suppression occurs when the mediational path has an opposite effect (i.e., instead of a decrease in the point estimate of the direct effect between an exposure and an outcome in relation to the total effect, there is an increase in the direct effect above the total effect’s point estimate) (30). We considered a statistically significant mediation or suppression of <1% as minimal, and of ≥1% as partial. Path analyses were conducted with 1,000 bootstrapped samples (31). We additionally conducted sex-based mediation analyses of the role of fat mass in the relationships between HOMA-IR and LVMI2.7.

Differences and associations with a two-sided P value <0.05 were considered statistically significant, with interpretations aided with effect estimates and their CIs or SEs. Analyses involving 20% of a sample of 10,000 children in the ALSPAC at 0.8 statistical power, 0.05 α, and two-sided P value would show a minimum detectable effect size of 0.062 SDs if they had relevant exposure for a normally distributed quantitative variable (32). Missing data were handled with 20 cycles of multiple imputations, because it has >98% efficiency in simulating real data in our cohort (17,19,24). Multiple comparisons were adjusted for using sequential Sidak correction. All statistical analyses were performed using SPSS statistics software, version 29.0 (IBM Corp, Armonk, NY) and mediation analyses was performed with IBM AMOS, version 29.

Data and Resource Availability

The informed consent obtained from ALSPAC participants does not allow the data to be made freely available through any third-party–maintained public repository. However, data used for this submission can be made available on request to the ALSPAC executive. The ALSPAC data management plan describes in detail the policy regarding data sharing, which is through a system of managed open access. Full instructions for applying for data access are given at https://www.bristol.ac.uk/alspac/researchers/access/. The ALSPAC study website contains details of all the data that are available (https://www.bristol.ac.uk/alspac/researchers/our-data/).

Cohort Study Characteristics

Among 1,595 adolescents (mean [SD] age, 17.74 [0.38] years; 955 [59.8%] female participants]) glucose, insulin, and HOMA-IR levels increased in the total cohort and in both male and female participants. The prevalence of hyperglycemia increased fivefold during growth from age 17 through 24 years irrespective of the glucose cut points (Table 1). The prevalence of LV hypertrophy increased threefold in the total cohort from 2.4 to 7.1%, and in both sexes (Table 1 and Supplementary Table 2). The prevalence of LVDD increased in the total cohort from 9.2 to 15.8%, and in both sexes over a period of 7 years (Table 1 and Supplementary Table 2). The mean (SD) BMI for the total cohort at baseline was 21.5 (3.82) kg/m2 and 23.1 (4.6) kg/m2 at follow-up. Approximately 96% of the cohort are White. Other characteristics are described in Table 1 and Supplementary Table 2.

Table 1

Descriptive characteristics of cohort participants (n = 1,595)

Age 17 years, mean (SD)Age 24 years, mean (SD)P value
Anthropometry/demography    
 Female sex, n (%) 955 (59.9) NA  
 Age at clinic visit (years) 17.74 (0.39) 24.38 (0.62) <0.001 
 Height (m) 1.71 (0.09) 1.72 (0.09) <0.001 
 Weight, median (IQR), kg 63.40 (14.55) 68.70 (18.6) <0.001 
White race, n (%) 1,415 (96.3) NA  
Body composition, median (IQR)    
 Total fat mass (kg) 15.59 (10.64) 19.20 (9.58) <0.001 
 Lean mass (kg) 41.98 (15.58) 44.76 (14.48) <0.001 
 BMI (kg/m221.50 (3.82) 23.11 (4.57) <0.001 
Overweight and obese ≥25 kg/m2, n (%) 214 (15.6) 490 (30.7) <0.001 
Fasting plasma metabolic indices    
 HDL, mmol/L 1.30 (0.30) 1.58 (0.42) <0.001 
 LDL, mmol/L 2.12 (0.63) 2.37 (0.73) <0.001 
 Triglyceride, median (IQR), mmol/L 0.73 (0.35) 0.82 (0.46) <0.001 
 Glucose (mmol/L) 5.01 (0.42) 5.35 (0.73) <0.001 
 Hyperglycemia >5.6 mmol/L, n (%) 64 (6.2) 429 (26.9) <0.001 
 Hyperglycemia >6.1 mmol/L, n (%) 11 (1.1) 90 (5.6) <0.001 
 Insulin, median (IQR), mU/L 6.57 (4.30) 7.39 (4.87) <0.001 
 Hyperinsulinemia >11.78 mU/L, n (%) 126 (12.2) 275 (17.2) 0.003 
 HOMA-IR, median (IQR) 1.46 (1.01) 1.74 (1.23) <0.001 
 hsCRP, median (IQR), mg/L 0.50 (0.91) 0.78 (1.64) <0.001 
Vascular measures    
 Heart rate (beat/min) 65 (10) 66 (10) <0.001 
 Systolic blood pressure (mmHg) 114 (10) 115 (11) <0.001 
 Diastolic blood pressure (mmHg) 64 (6) 66 (7) <0.001 
Cardiac measure    
 LV mass indexed for height (g/m2.735.36 (7.30) 38.14 (8.53) <0.001 
 LV hypertrophy ≥51 g/m2.7, n (%) 18 (2.4) 114 (7.1) <0.001 
 RWT (cm) 0.38 (0.06) 0.36 (0.06) <0.001 
 High RWT ≥0.44 cm, n (%) 78 (10.4) 121 (7.6) 0.010 
 LV diastolic function (E/A) 1.95 (0.38) 2.00 (0.57) <0.001 
 LV diastolic dysfunction E/A <1.5, n (%) 68 (9.2) 237 (15.8) <0.001 
 LVFP (E/e′) 4.86 (1.05) 5.03 (1.02) 0.003 
 High LVFP E/e′ ≥8, n (%) 11 (1.5) 14 (1.0) 0.774 
Lifestyle factor    
 Smoked cigarette in the past 30 days, n (%) 298 (24.7) 417 (26.4) 0.276 
 Family history of H-D-C-V, n (%) 408 (30.2) NA  
 Sedentary time (min/day) 478 (82) 528 (85) <0.001 
 Light physical activity (min/day) 277 (63) 147 (52) <0.001 
 Moderate to vigorous physical activity (min/day) 46 (27) 51 (31) 0.045 
Maternal social economic status, n (%)  NA  
 Professional occupation 53 (7.3)   
 Managerial and technical 305 (41.9)   
 Skilled nonmanual 249 (34.2)   
 Skilled manual 11 (1.5)   
 Partly skilled 95 (13.0)   
 Unskilled 15 (2.1)   
Age 17 years, mean (SD)Age 24 years, mean (SD)P value
Anthropometry/demography    
 Female sex, n (%) 955 (59.9) NA  
 Age at clinic visit (years) 17.74 (0.39) 24.38 (0.62) <0.001 
 Height (m) 1.71 (0.09) 1.72 (0.09) <0.001 
 Weight, median (IQR), kg 63.40 (14.55) 68.70 (18.6) <0.001 
White race, n (%) 1,415 (96.3) NA  
Body composition, median (IQR)    
 Total fat mass (kg) 15.59 (10.64) 19.20 (9.58) <0.001 
 Lean mass (kg) 41.98 (15.58) 44.76 (14.48) <0.001 
 BMI (kg/m221.50 (3.82) 23.11 (4.57) <0.001 
Overweight and obese ≥25 kg/m2, n (%) 214 (15.6) 490 (30.7) <0.001 
Fasting plasma metabolic indices    
 HDL, mmol/L 1.30 (0.30) 1.58 (0.42) <0.001 
 LDL, mmol/L 2.12 (0.63) 2.37 (0.73) <0.001 
 Triglyceride, median (IQR), mmol/L 0.73 (0.35) 0.82 (0.46) <0.001 
 Glucose (mmol/L) 5.01 (0.42) 5.35 (0.73) <0.001 
 Hyperglycemia >5.6 mmol/L, n (%) 64 (6.2) 429 (26.9) <0.001 
 Hyperglycemia >6.1 mmol/L, n (%) 11 (1.1) 90 (5.6) <0.001 
 Insulin, median (IQR), mU/L 6.57 (4.30) 7.39 (4.87) <0.001 
 Hyperinsulinemia >11.78 mU/L, n (%) 126 (12.2) 275 (17.2) 0.003 
 HOMA-IR, median (IQR) 1.46 (1.01) 1.74 (1.23) <0.001 
 hsCRP, median (IQR), mg/L 0.50 (0.91) 0.78 (1.64) <0.001 
Vascular measures    
 Heart rate (beat/min) 65 (10) 66 (10) <0.001 
 Systolic blood pressure (mmHg) 114 (10) 115 (11) <0.001 
 Diastolic blood pressure (mmHg) 64 (6) 66 (7) <0.001 
Cardiac measure    
 LV mass indexed for height (g/m2.735.36 (7.30) 38.14 (8.53) <0.001 
 LV hypertrophy ≥51 g/m2.7, n (%) 18 (2.4) 114 (7.1) <0.001 
 RWT (cm) 0.38 (0.06) 0.36 (0.06) <0.001 
 High RWT ≥0.44 cm, n (%) 78 (10.4) 121 (7.6) 0.010 
 LV diastolic function (E/A) 1.95 (0.38) 2.00 (0.57) <0.001 
 LV diastolic dysfunction E/A <1.5, n (%) 68 (9.2) 237 (15.8) <0.001 
 LVFP (E/e′) 4.86 (1.05) 5.03 (1.02) 0.003 
 High LVFP E/e′ ≥8, n (%) 11 (1.5) 14 (1.0) 0.774 
Lifestyle factor    
 Smoked cigarette in the past 30 days, n (%) 298 (24.7) 417 (26.4) 0.276 
 Family history of H-D-C-V, n (%) 408 (30.2) NA  
 Sedentary time (min/day) 478 (82) 528 (85) <0.001 
 Light physical activity (min/day) 277 (63) 147 (52) <0.001 
 Moderate to vigorous physical activity (min/day) 46 (27) 51 (31) 0.045 
Maternal social economic status, n (%)  NA  
 Professional occupation 53 (7.3)   
 Managerial and technical 305 (41.9)   
 Skilled nonmanual 249 (34.2)   
 Skilled manual 11 (1.5)   
 Partly skilled 95 (13.0)   
 Unskilled 15 (2.1)   

The values are means (SD) unless otherwise indicated. Change in variables over time was tested with a paired sample t test for normally distributed continuous variables, Wilcoxon signed-rank test for skewed continuous variables, and McNemar change test for categorical variables. A two-sided P < 0.05 was considered statistically significant. H-D-C-V, hypertension/diabetes/high cholesterol/vascular disease; NA, not available or applicable.

Longitudinal Changes in Glucose, Insulin, and HOMA-IR With Changes in Cardiac Structure and Function

In fully adjusted whole-cohort (Table 2) and sex-specific analyses (Supplementary Table 3) from ages 17 to 24 years, increasing glucose concentration was significantly associated with increasing LVMI2.7 and RWT. Although increased glucose concentration was associated with decreased LVFP in the whole cohort and male participants, it was associated with decreased LVDF in female participants but increased LVDF in male participants (Table 2 and Supplementary Table 3). Increased insulin concentration in the whole cohort was significantly associated with increased LVFP (Table 2). Increased HOMA-IR was associated with increased LVMI2.7 and RWT (Table 2).

Table 2

Longitudinal associations of changes in glucose, insulin, and IR with changes in cardiac structure and function from ages 17 through 24 years (n = 1,595)

LVMI2.7 (g/m2.7)RWT (cm)LVDF (E/A)LVFP (E/e′)
β (95% CI)P valueβ (95% CI)P valueβ (95% CI)P valueβ (95% CI)P value
Glucose (mmol/L)         
 Model 1 0.627 (0.457–0.797) <0.001 −0.004 (−0.006 to −0.003) <0.001 0.022 (0.011–0.034) <0.001 −0.026 (−0.0483 to −0.003) 0.027 
 Model 2 0.374 (0.227–0.521) <0.001 0.002 (0.001–0.004) <0.001 −0.007 (−0.018–0.005) 0.263 −0.027 (−0.0513 to −0.004) 0.021 
Insulin (mU/L)         
 Model 1 1.717 (1.230–2.203) <0.001 0.002 (−0.003 to 0.007) 0.405 0.004 (−0.031–0.038) 0.839 0.058 (−0.0213 to 0.136) 0.149 
 Model 2 −0.001 (−0.045–0.043) 0.961 0.000 (0.000–0.000) 0.599 0.002 (−0.001–0.005) 0.294 0.011 (0.004–0.018) 0.002 
IR         
 Model 1 1.988 (1.502–2.473) <0.001 −0.002 (−0.006 to 0.003) 0.445 0.021 (−0.014–0.055) 0.234 0.035 (−0.0433 to 0.114) 0.378 
 Model 2 1.101 (0.632–1.571) <0.001 0.006 (0.002–0.010) 0.005 0.005 (−0.027–0.038) 0.745 0.066 (−0.0133 to 0.145) 0.103 
  LV hypertrophy  High RWT  LVDF  High LVFP  
  OR (95% CI)  OR (95% CI)  OR (95% CI)  OR (95% CI)  
Persistent hyperglycemia (>5.6 mmol/L) 1.46 (1.35–1.57) <0.001 1.30 (1.21–1.40) <0.001 1.12 (1.04–1.20) 0.002 1.33 (1.24–1.43) <0.001 
Persistent hyperglycemia (>6.1 mmol/L) 3.10 (1.19–8.08) 0.021 1.98 (1.55–2.45) <0.001 1.31 (1.08–1.60) 0.006 0.95 (0.74–1.22) 0.667 
Persistent hyperinsulinemia 1.34 (0.40–4.54) 0.635 0.95 (0.43–2.12) 0.905 1.61 (0.87–2.98) 0.132 1.22 (1.23–1.32) <0.001 
Persistent high insulin resistance 1.10 (1.02–1.18) 0.011 1.11 (1.04–1.20) 0.004 1.28 (1.20–1.37) <0.001 1.07 (0.99–1.15) 0.060 
LVMI2.7 (g/m2.7)RWT (cm)LVDF (E/A)LVFP (E/e′)
β (95% CI)P valueβ (95% CI)P valueβ (95% CI)P valueβ (95% CI)P value
Glucose (mmol/L)         
 Model 1 0.627 (0.457–0.797) <0.001 −0.004 (−0.006 to −0.003) <0.001 0.022 (0.011–0.034) <0.001 −0.026 (−0.0483 to −0.003) 0.027 
 Model 2 0.374 (0.227–0.521) <0.001 0.002 (0.001–0.004) <0.001 −0.007 (−0.018–0.005) 0.263 −0.027 (−0.0513 to −0.004) 0.021 
Insulin (mU/L)         
 Model 1 1.717 (1.230–2.203) <0.001 0.002 (−0.003 to 0.007) 0.405 0.004 (−0.031–0.038) 0.839 0.058 (−0.0213 to 0.136) 0.149 
 Model 2 −0.001 (−0.045–0.043) 0.961 0.000 (0.000–0.000) 0.599 0.002 (−0.001–0.005) 0.294 0.011 (0.004–0.018) 0.002 
IR         
 Model 1 1.988 (1.502–2.473) <0.001 −0.002 (−0.006 to 0.003) 0.445 0.021 (−0.014–0.055) 0.234 0.035 (−0.0433 to 0.114) 0.378 
 Model 2 1.101 (0.632–1.571) <0.001 0.006 (0.002–0.010) 0.005 0.005 (−0.027–0.038) 0.745 0.066 (−0.0133 to 0.145) 0.103 
  LV hypertrophy  High RWT  LVDF  High LVFP  
  OR (95% CI)  OR (95% CI)  OR (95% CI)  OR (95% CI)  
Persistent hyperglycemia (>5.6 mmol/L) 1.46 (1.35–1.57) <0.001 1.30 (1.21–1.40) <0.001 1.12 (1.04–1.20) 0.002 1.33 (1.24–1.43) <0.001 
Persistent hyperglycemia (>6.1 mmol/L) 3.10 (1.19–8.08) 0.021 1.98 (1.55–2.45) <0.001 1.31 (1.08–1.60) 0.006 0.95 (0.74–1.22) 0.667 
Persistent hyperinsulinemia 1.34 (0.40–4.54) 0.635 0.95 (0.43–2.12) 0.905 1.61 (0.87–2.98) 0.132 1.22 (1.23–1.32) <0.001 
Persistent high insulin resistance 1.10 (1.02–1.18) 0.011 1.11 (1.04–1.20) 0.004 1.28 (1.20–1.37) <0.001 1.07 (0.99–1.15) 0.060 

Model 1 was unadjusted. Model 2 was adjusted for sex and other time-varying covariates measured at both baseline and follow-up, such as age, hsCRP, heart rate, systolic blood pressure, fat mass, lean mass, smoking status, family history of hypertension, diabetes, high cholesterol, vascular disease, socioeconomic status, sedentary time, light physical activity, moderate to vigorous physical activity, LDL-c, HDL-c, and triglyceride and glucose or insulin levels, depending on the predictor. When IR was included as a predictor, there were no further adjustments for insulin and glucose. Regression coefficients (β) were computed from a generalized linear mixed-effect model for repeated measures. ORs were computed from generalized logit mixed-effect model for repeated measures. Participants with LVMI2.7 ≥51 g/m2.7, RWT ≥0.44, LVD function <1.5, and LVFP ≥8 were categorized as having LV hypertrophy, high RWT, LVDF, and high LVFP, respectively. Glucose >5.6 mmol/L and >6.1 mmol/L, insulin >11.78 mU/L, and 75th percentile IR were categorized as hyperglycemia, hyperinsulinemia, and high IR, respectively. HOMA-IR was calculated from (fasting insulin × fasting glucose/22.5). A two-sided P < 0.05 was considered statistically significant. Multiple testing was corrected with Sidak correction. Multiple imputations (n = 20 cycles) were used to account for missing variables. A 1-unit change in exposure is associated with the point estimate a 1-unit increase in the continuous outcome.

Persistent Hyperglycemia and Hyperinsulinemia With the Risk of Progressive Cardiac Damage

In a fully adjusted model, persistent hyperglycemia of >5.6 mmol/L (odds ratio [OR] 1.46 [95% CI 1.35–1.57]; P < 0.001) and >6.1 mmol/L (OR, 3.10; 95% CI, 1.19–8.08; P = 0.021), were separately associated with higher odds of LV hypertrophy progression over 7 years (Table 2). Persistent hyperglycemia of >5.6 mmol/L was associated with worsening high RWT, LVDD, and high LVFP over 7 years (Table 2). There were no statistically significant relationships of persistent insulinemia with LV hypertrophy, high RWT, and LVDD, except for high LVFP (Table 2). Persistently high IR was significantly associated with worsening LV hypertrophy, high RWT, and LVDD.

Mediation Effect of Body Composition, Lipids, Blood Pressure, and Inflammation on the Associations of Glucose and HOMA-IR With Cardiac Structure and Function

Increased fat mass significantly mediated the associations of increased HOMA-IR with increased LVMI2.7 (62.3% mediation effect) (Table 3 and Fig. 1). Increased LDL-c partly mediated the associations of the increased glucose with increased LVMI2.7 (6.3% mediation effect), but increased hsCRP suppressed the association by the same amount (Table 3). Increased triglyceride levels and systolic blood pressure partly mediated the associations of increased HOMA-IR with increased LVMI2.7 (11.8–18.6% mediation effect) (Table 3).

Table 3

Mediating or suppressing role of fat mass, lean mass, blood pressure, and fasting lipids and inflammation on the longitudinal associations of change in glucose and IR with LV mass progression from ages 17 through 24 years

MediatorLV mass from ages 17–24 years
Total effect β (95% CI)P valueDirect effectβ (95% CI)P valueIndirect effect β (95% CI)P valueMediation or suppression (%)
Glucose        
 LDL-c 0.144 (0.101–0.199) 0.003 0.135 (0.088–0.191) 0.004 0.009 (0.004–0.015) 0.001 6.3 mediation 
 Triglyceride 0.134 (0.092–0.183) 0.004 0.129 (0.086–0.184) 0.003 0.005 (−0.002–0.011) 0.144 3.7 
 Systolic blood pressure 0.099 (0.060–0.147) 0.003 0.101 (0.058–0.152) 0.004 −0.002 (−0.012–0.009) 0.814 2.0 
 hsCRP 0.142 (0.097–0.196) 0.004 0.151 (0.109–0.206) 0.003 −0.009 (−0.014–−0.005) 0.002 6.3 suppression 
 Fat mass 0.132 (0.095–0.180) 0.003 0.133 (0.096–0.183) 0.003 −0.001 (−0.010–0.008) 0.834 0.8 
 Lean mass 0.077 (0.036–0.126) 0.004 0.074 (0.039–0.119) 0.003 0.004 (−0.002–0.011) 0.211 5.2 
IR        
 LDL-c 0.130 (0.096–0.165) 0.002 0.134 (0.101–0.169) 0.002 −0.004 (−0.011–0.000) 0.060 3.1 
 Triglyceride 0.140 (0.106–0.173) 0.002 0.113 (0.079–0.152) 0.002 0.026 (0.017–0.036) 0.003 18.6 mediation 
 Systolic blood pressure 0.127 (0.094–0.160) 0.002 0.112 (0.078–0.144) 0.002 0.015 (0.006–0.026) 0.003 11.8 mediation 
 hsCRP 0.133 (0.099–0.169) 0.002 0.137 (0.101–0.172) 0.002 −0.003 (−0.009–0.001) 0.098 2.3 
 Fat mass 0.114 (0.078–0.147) 0.002 0.043 (0.002–0.079) 0.041 0.071 (0.059–0.084) 0.002 62.3 mediation 
 Lean mass 0.147 (0.114–0.178) 0.002 0.152 (0.121–0.185) 0.002 −0.005 (−0.012–0.001) 0.123 3.4 
MediatorLV mass from ages 17–24 years
Total effect β (95% CI)P valueDirect effectβ (95% CI)P valueIndirect effect β (95% CI)P valueMediation or suppression (%)
Glucose        
 LDL-c 0.144 (0.101–0.199) 0.003 0.135 (0.088–0.191) 0.004 0.009 (0.004–0.015) 0.001 6.3 mediation 
 Triglyceride 0.134 (0.092–0.183) 0.004 0.129 (0.086–0.184) 0.003 0.005 (−0.002–0.011) 0.144 3.7 
 Systolic blood pressure 0.099 (0.060–0.147) 0.003 0.101 (0.058–0.152) 0.004 −0.002 (−0.012–0.009) 0.814 2.0 
 hsCRP 0.142 (0.097–0.196) 0.004 0.151 (0.109–0.206) 0.003 −0.009 (−0.014–−0.005) 0.002 6.3 suppression 
 Fat mass 0.132 (0.095–0.180) 0.003 0.133 (0.096–0.183) 0.003 −0.001 (−0.010–0.008) 0.834 0.8 
 Lean mass 0.077 (0.036–0.126) 0.004 0.074 (0.039–0.119) 0.003 0.004 (−0.002–0.011) 0.211 5.2 
IR        
 LDL-c 0.130 (0.096–0.165) 0.002 0.134 (0.101–0.169) 0.002 −0.004 (−0.011–0.000) 0.060 3.1 
 Triglyceride 0.140 (0.106–0.173) 0.002 0.113 (0.079–0.152) 0.002 0.026 (0.017–0.036) 0.003 18.6 mediation 
 Systolic blood pressure 0.127 (0.094–0.160) 0.002 0.112 (0.078–0.144) 0.002 0.015 (0.006–0.026) 0.003 11.8 mediation 
 hsCRP 0.133 (0.099–0.169) 0.002 0.137 (0.101–0.172) 0.002 −0.003 (−0.009–0.001) 0.098 2.3 
 Fat mass 0.114 (0.078–0.147) 0.002 0.043 (0.002–0.079) 0.041 0.071 (0.059–0.084) 0.002 62.3 mediation 
 Lean mass 0.147 (0.114–0.178) 0.002 0.152 (0.121–0.185) 0.002 −0.005 (−0.012–0.001) 0.123 3.4 

Mediation structural equation model was adjusted for sex, family history of hypertension, diabetes, high cholesterol, vascular disease, socioeconomic status, and time-varying covariates measured at both baseline and follow-up, such as age, heart rate, smoking status, sedentary time, light physical activity, moderate to vigorous physical activity; fat mass, lean mass, insulin, hsCRP, HDL, and triglyceride levels, depending on the predictor or mediator. When IR was the predictor, glucose and insulin were not included as covariates. The standardized regression coefficient is β. P < 0.05 was considered statistically significant. When the magnitude of the longitudinal association between the predictor and outcome is increased upon inclusion of a third variable, a suppression is confirmed, but when decreased, it is mediation. HOMA-IR was calculated from (fasting insulin × fasting glucose/22.5). LV mass was indexed for height2.7.

Figure 1

Mediating role of increased fat mass in the longitudinal relationships of insulin resistance with left ventricular mass. Mediation structural equation model estimating natural direct and indirect effects was adjusted for sex, family history of hypertension, diabetes, high cholesterol, vascular disease, and socioeconomic status, in addition to time-varying covariates such as age, hsCRP, heart rate, systolic blood pressure, smoking status, sedentary time, light physical activity, moderate to vigorous physical activity, HDL cholesterol, LDL cholesterol, triglyceride level, and lean mass. The standardized regression coefficient is indicated by β. Two-sided P < 0.05 was considered statistically significant. HOMA-IR was calculated as (fasting insulin × fasting glucose/22.5). When the magnitude of the longitudinal association between the predictor and outcome is decreased upon inclusion of a third variable, a mediation is confirmed. LVMI2.7, left ventricular mass indexed for height raised to the 2.7 power.

Figure 1

Mediating role of increased fat mass in the longitudinal relationships of insulin resistance with left ventricular mass. Mediation structural equation model estimating natural direct and indirect effects was adjusted for sex, family history of hypertension, diabetes, high cholesterol, vascular disease, and socioeconomic status, in addition to time-varying covariates such as age, hsCRP, heart rate, systolic blood pressure, smoking status, sedentary time, light physical activity, moderate to vigorous physical activity, HDL cholesterol, LDL cholesterol, triglyceride level, and lean mass. The standardized regression coefficient is indicated by β. Two-sided P < 0.05 was considered statistically significant. HOMA-IR was calculated as (fasting insulin × fasting glucose/22.5). When the magnitude of the longitudinal association between the predictor and outcome is decreased upon inclusion of a third variable, a mediation is confirmed. LVMI2.7, left ventricular mass indexed for height raised to the 2.7 power.

Close modal

Among male participants, increased total fat mass significantly suppressed the associations of increased HOMA-IR with increased LVMI2.7 (178% suppression effect) (Supplementary Fig. 2). The total effect standardized β of 0.069 (95% CI 0.016–0.120; P = 0.008) and the indirect effect of β = 0.123 (95% CI 0.098–0.150; P = 0.002) were statistically significant (Supplementary Fig. 2). Among female participants, increased total fat mass significantly mediated the associations of increased HOMA-IR with increased LVMI2.7 (94.5% mediation effect) (Supplementary Fig. 2). The total effect standardized β of 0.153 (95% CI 0.107–0.199; P = 0.003) and the indirect effect β of 0.145 (95% CI 0.114–0.175; P = 0.003) were statistically significant (Supplementary Fig. 2).

In a large adolescent population, persistent hyperglycemia was associated with threefold higher odds of progressive cardiac structural damage, whereas increased glucose levels and IR were associated with worsening cardiac remodeling and diastolic dysfunction indices. The relationship between increased glucose and worse cardiac structural change was consistent across sexes. Increased triglyceride levels, systolic blood pressure, and fat mass partly mediated the relationship of IR with increased cardiac mass. Given previous literature on pediatric obesity, elevated blood pressure, and dyslipidemia as predictors of LV remodeling, these findings add worsening hyperglycemia and IR as predictors of structural and functional cardiac alterations necessitating primordial and primary preventive approaches to mitigating these changes in the young population (1–5).

Effect of Increased Hyperglycemia and IR on Changes in Cardiac Structure

In the Treatment Options for Type 2 Diabetes in Adolescents and Youth (TODAY) study, which included 411 young adults with type 2 diabetes, mean age of 23 years, and who were followed up for 5 years, the prevalence of LV hypertrophy was 15.8%, which was twice as high as in the present study follow-up age of 24 years (2). Unfortunately, treatment failed to reverse the cardiac organ injury observed in the TODAY study, which worsened (2). After 10 years follow-up of the TODAY study, 17 major cardiovascular events occurred and one death due to myocardial infarction, consistent with previous literature indicating LV hypertrophy predicts future atherosclerotic cardiovascular diseases (33). In our cohort, increased glucose level and IR were both associated with increased LVM and RWT, in addition to hyperglycemia-induced increased odds for LV hypertrophy suggesting the likelihood of concentric hypertrophy (34). These findings in an apparently healthy cohort extend the evidence already documented in adolescents and young adults diagnosed with young-onset type 2 diabetes (1–5). In adult studies such as the Coronary Artery Risk Development in Young Adults (CARDIA) and Framingham Heart studies, long-standing adult-onset type 2 diabetes has been longitudinally associated with adverse cardiac remodeling (35,36). The consistent effect of elevated glucose levels on cardiac structure and functional indices in youth extends previously reported evidence among 95,783 adults in whom progressively increased glucose levels were associated with cardiovascular risk, including fatal or nonfatal cardiovascular outcomes (e.g., sudden death, stroke, or acute myocardial infarction) during a 12.4-year follow-up period (37). In the adult study, fasting and 2-h glucose levels of 6.1 mmol/L and 7.8 mmol/L were associated with cardiovascular event risk of 1.33 (95% CI 1.06–1.67) and 1.58 (95% CI 1.19–2.10), respectively (37). In the present study, the odds of persistent hyperglycemia-associated worsening LV hypertrophy was twofold higher when the ≥6.1 mmol/L hyperglycemia cut point (OR = 3.01) was the predictor, compared with the ≥5.6 mmol/L cut point (OR = 1.45). These findings may indicate one of the earliest and covert, yet deleterious, consequences of uncontrolled high glucose concentration on the cardiovascular system of young people.

Youth-onset type 2 diabetes has a more extreme metabolic phenotype than does adult-onset type 2 diabetes, with greater IR and deterioration of β-cell function (5). Dietary intervention and light physical activity from childhood may be able to reverse metabolic alterations, thereby primordially preventing premature cardiac structural damage (5,21,38). Although hyperglycemia seems to worsen cardiac structure in both sexes, it appears that cardiac function response is sex differential. Each unit increase in glucose was associated with a fivefold larger effect estimate of increased LVM in female participants compared with male participants. Evidence in adults suggests that women with type 2 diabetes have a greater relative risk of cardiovascular disease than men (39). Worsening glycemia was associated with improved LVFP in men but worse LVDF in women. The alterations of these cardiac functional parameters are precursors of heart failure with preserved ejection fraction in adults (14). Moreover, an aggressive hyperglycemia-induced cardiac structural and functional damage in women relative to men might have begun several decades before a clinical diagnosis of type 2 diabetes, likely in the mid-40s age range (39). In addition, higher blood glucose levels could also be a marker for IR or metabolic syndrome (23).

The long-term persistence of hyperglycemic stress on cardiomyocytes even after blood glucose normalization triggers mitochondrial dysfunction and endoplasmic reticulum stress–induced inflammation (29). This promotes reactive oxygen species and increases the risk of cellular damage (29). Persistent hyperglycemia results in advanced glycation end-product generation, which accumulates in the vessel wall, compromising the structural integrity of the extracellular matrix, inducing endothelium damage, and nitric oxide reduction (29). Hyperglycemia promotes increased inflammation, leading to increased IR and β-cell dysfunction, perpetuating a vicious cycle of hyperglycemia and subsequent IR (29). In the presence of hyperglycemia and IR, cardiac metabolic flexibility is lost, resulting in reduced use of free fatty acids, and in intramyocardial lipid accumulation. This may contribute to cardiomyocyte enlargement with consequent hypertrophy, along with metabolic alterations through impaired mitochondrial function that lead to contractile dysfunction (29). In the present study, we observed that the pathway through which IR exerts its influence on cardiac structure is mainly through increased fat mass, but we also observed important sex differences. Among female participants, 95% of the longitudinal association of IR with cardiac mass was explained by increased fat mass. On the contrary, fat mass suppressed the association of IR with increased cardiac mass in male participants, which also is referred to as inconsistent mediation. In the cohort, female participants have more fat mass relative to male participants, but the effect of fat mass on cardiac indices seems significant despite having an average BMI in both male and female participants at baseline and follow-up within the normal healthy BMI range (6,24,40). Obesity in youth is known to be a hypervolemic and hyperdynamic state, which could induce LV remodeling and induce dysglycemia (39). The vicious cycle of increased fat mass and IR appears to begin in adolescence, and the findings of the present study indicate metabolic-induced cardiac alterations might have been concurrently initiated (6). Triglycerides also had an approximately 20% mediating effect on the relationship of IR with increased cardiac mass, which may indicate the role of diet; however, triglyceride levels also may be a risk factor for increased IR (38). This evidence collectively supports the need for primordial and primary prevention of hyperglycemia, IR, and excess fat mass in the young population to decrease the risk of young-onset type 2 diabetes and cardiovascular disease sequelae (5,21,24,26).

Strengths and Limitations

In an extensively phenotyped large birth cohort (ALSPAC) with repeated measures of variables during late adolescence and early adulthood, we examined the associations of metabolic alterations on cardiac structure and function in an asymptomatic young population. Several objectively and repeatedly measured confounders were controlled for, such as fat mass, lean mass, physical activity, sedentary time, and several lifestyle factors, like smoking, family history, and socioeconomic status, which increased the internal validity of the study. Nonetheless, we may not exclude the existence of residual biases from unmeasured confounders, such as the unavailability of dietary records. The measure of insulin sensitivity and secretion using gold standard methods such as the clamp test was unavailable, given that the feasibility of these measures in large epidemiologic studies is limited. The participants were mostly White; thus, findings may not be generalizable to other ethnic groups. Last, establishing causation with observational studies is challenging and warrants further experimental studies.

Conclusion

Persistent hyperglycemia and increased IR were independently associated with the risk of worsening cardiac structural and functional damage in youth, with excess fat mass explaining >60% of the relationship. Hyperglycemia-induced cardiac alterations may affect female youth worse than male youth. Efforts targeted at preventing and treating adolescent obesity may help disrupt the pathogenesis of IR, young-onset type 2 diabetes, and cardiovascular alterations in later life.

This article contains supplementary material online at https://doi.org/10.2337/figshare.28635524.

Acknowledgments. The authors are extremely grateful to all the families who took part in this study, the midwives for their help in recruiting them, and the whole ALSPAC team, which includes interviewers, computer and laboratory technicians, clerical workers, research scientists, volunteers, managers, receptionists, and nurses. They appreciate Jennifer Lyn Baker for her helpful suggestions during manuscript preparation.

Funding. The UK Medical Research Council and Wellcome (grant 217065/Z/19/Z) and the University of Bristol provide core support for ALSPAC. The British Heart Foundation (grant CS/15/6/31468) funded blood pressure and Actigraph activity monitoring device measurement at 24 years. The Medical Research Council (grant MR/M006727/1) supported smoking data collection. A comprehensive list of grant funding is available on the ALSPAC website (https://www.bristol.ac.uk/alspac/external/documents/grant-acknowledgements.pdf); A.O.A.’s research group (UndeRstanding FITness and Cardiometabolic Health In Little Darlings: urFIT-child) has been funded by the Jenny and Antti Wihuri Foundation (grant 00180006); the North Savo regional and central Finnish Cultural Foundation (grants 65191835, 00200150, 00230190, and 00250189); Orion Research Foundation sr; Aarne Koskelo Foundation; Antti and Tyyne Soininen Foundation; Kuopio University Foundation; Paulo Foundation; Paavo Nurmi Foundation; Yrjö Jahnsson Foundation (grant 20217390); Ida Montin Foundation (grant 20230289); Fund of Eino Räsänen and Fund of Matti and Vappu Maukonen via the Faculty of Health Sciences University of Eastern Finland; Pediatric Research Foundation (grant 240417); the Finnish Foundation for Cardiovascular Research (grants 220021, 230012, and 240003); the Alfred Kordelin Foundation (grant 230082); and the 2024 European Association for the Study of Obesity-Novo Nordisk Foundation New Investigator Award for Childhood Obesity (grant NNF24SA0090437). The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. A.O.A. had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. A.O.A. contributed to the study concept and design, data acquisition, analysis, and interpretation, statistical analysis, funding, and drafting the manuscript. A.O.A., J.P.Z., A.R.B., C.A.W., D.V., C.S., and T.-P.T. critically reviewed and revised the manuscript for important intellectual content.

Prior Presentation. An abstract from this study was presented at the European Society of Cardiology Congress, London, U.K., 1 September 2024, and published in the European Heart Journal (https://doi.org/10.1093/eurheartj/ehae6 66.2898).

Handling Editors. The journal editors responsible for overseeing the review of the manuscript were Steven E. Kahn and Kristen J. Nadeau.

1.
Shah
AS
,
El ghormli
L
,
Gidding
SS
, et al.;
TODAY Study Group
.
Longitudinal changes in vascular stiffness and heart rate variability among young adults with youth-onset type 2 diabetes: results from the follow-up observational treatment options for type 2 diabetes in adolescents and youth (TODAY) study
.
Acta Diabetol
2022
;
59
:
197
205
2.
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
3.
Maahs
DM
,
Daniels
SR
,
de Ferranti
SD
, et al.;
American Heart Association Atherosclerosis, Hypertension and Obesity in Youth Committee of the Council on Cardiovascular Disease in the Young, Council on Clinical Cardiology, Council on Cardiovascular and Stroke Nursing, Council for High Blood Pressure Research, and Council on Lifestyle and Cardiometabolic Health
.
Cardiovascular disease risk factors in youth with diabetes mellitus: a scientific statement from the American Heart Association
.
Circulation
2014
;
130
:
1532
1558
4.
Misra
S
,
Ke
C
,
Srinivasan
S
, et al
.
Current insights and emerging trends in early-onset type 2 diabetes
.
Lancet Diabetes Endocrinol
2023
;
11
:
768
782
5.
Bjornstad
P
,
Chao
LC
,
Cree-Green
M
, et al
.
Youth-onset type 2 diabetes mellitus: an urgent challenge
.
Nat Rev Nephrol
2023
;
19
:
168
184
6.
Agbaje
AO
,
Saner
C
,
Zhang
J
,
Henderson
M
,
Tuomainen
T-P.
DXA-based fat mass with risk of worsening insulin resistance in adolescents: a 9-year temporal and mediation study
.
J Clin Endocrinol Metab
2024
;
109
:
e1708
e1719
7.
Mangione
CM
,
Barry
MJ
,
Nicholson
WK
, et al.;
US Preventive Services Task Force
.
Screening for prediabetes and type 2 diabetes in children and adolescents: US Preventive Services Task Force Recommendation Statement
.
JAMA
2022
;
328
:
963
967
8.
Agbaje
AO
,
Zachariah
JP
,
Bamsa
O
,
Odili
AN
,
Tuomainen
T-P.
Cumulative insulin resistance and hyperglycemia with arterial stiffness and carotid IMT progression in 1,779 adolescents: a 9-yr longitudinal cohort study
.
Am J Physiol Endocrinol Metab
2023
;
324
:
E268
E278
9.
Agbaje
AO.
Elevated blood pressure and worsening cardiac damage during adolescence
.
J Pediatr
2023
;
257
:
113374
10.
Agbaje
AO.
Increasing lipids with risk of worsening cardiac damage in 1595 adolescents: a 7-year longitudinal and mediation study
.
Atherosclerosis
2024
;
389
:
117440
11.
Agbaje
AO.
Incidental and progressive tobacco smoking in childhood and subsequent risk of premature cardiac damage
.
J Am Coll Cardiol
2025
;
85
:
546
549
12.
Agbaje
AO.
Accelerometer-based Sedentary time and physical activity from childhood through young adulthood with progressive cardiac changes: a 13-year longitudinal study
.
Eur J Prev Cardiol
2024
;
31
:
1480
1492
13.
Nagueh
SF
,
Smiseth
OA
,
Appleton
CP
, et al
.
Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging
.
J Am Soc Echocardiogr
2016
;
29
:
277
314
14.
Mitter
SS
,
Shah
SJ
,
Thomas
JD.
A test in context: E/A and E/e' to assess diastolic dysfunction and LV filling pressure
.
J Am Coll Cardiol
2017
;
69
:
1451
1464
15.
Boyd
A
,
Golding
J
,
Macleod
J
, et al
.
Cohort profile: the 'children of the 90s'-the index offspring of the Avon Longitudinal Study of Parents and Children
.
Int J Epidemiol
2013
;
42
:
111
127
16.
Northstone
K
,
Lewcock
M
,
Groom
A
, et al
.
The Avon Longitudinal Study of Parents and Children (ALSPAC): an update on the enrolled sample of index children in 2019
.
Wellcome Open Res
2019
;
4
:
51
17.
Agbaje
AO
,
Barker
AR
,
Tuomainen
T-P.
Effects of arterial stiffness and carotid intima-media thickness progression on the risk of overweight/obesity and elevated blood pressure/hypertension: a cross-lagged cohort study
.
Hypertension
2022
;
79
:
159
169
18.
Harris
PA
,
Taylor
R
,
Thielke
R
,
Payne
J
,
Gonzalez
N
,
Conde
JG.
Research electronic data capture (REDCap)-a metadata-driven methodology and workflow process for providing translational research informatics support
.
J Biomed Inform
2009
;
42
:
377
381
19.
Agbaje
AO
,
Barker
AR
,
Mitchell
GF
,
Tuomainen
T-P.
Effect of Arterial stiffness and carotid intima-media thickness progression on the risk of dysglycemia, insulin resistance, and dyslipidemia: a temporal causal longitudinal study
.
Hypertension
2022
;
79
:
667
678
20.
Agbaje
AO
,
Lloyd-Jones
DM
,
Magnussen
CG
,
Tuomainen
T-P.
Cumulative dyslipidemia with arterial stiffness and carotid IMT progression in asymptomatic adolescents: a simulated intervention longitudinal study using temporal inverse allocation model
.
Atherosclerosis
2023
;
364
:
39
48
21.
Agbaje
AO.
The interactive effects of sedentary time, physical activity, and fat mass on insulin resistance in the young population
.
J Clin Endocrinol Metab
2024
;
110
:
e117
e131
22.
American Diabetes Association
.
2. Classification and diagnosis of diabetes: Standards of Medical Care in Diabetes-2020
.
Diabetes Care
2020
;
43
(
Suppl. 1
):
S14
S31
23.
Wallace
TM
,
Levy
JC
,
Matthews
DR.
Use and abuse of HOMA modeling
.
Diabetes Care
2004
;
27
:
1487
1495
24.
Agbaje
AO
,
Perng
W
,
Tuomainen
T-P.
Effects of accelerometer-based sedentary time and physical activity on DEXA-measured fat mass in 6059 children
.
Nat Commun
2023
;
14
:
8232
25.
Agbaje
AO.
Longitudinal mediating effect of fat mass and lipids on sedentary time, light PA, and MVPA with inflammation in youth
.
J Clin Endocrinol Metab
2023
;
108
:
3250
3259
26.
Agbaje
AO.
Associations of sedentary time and physical activity from childhood with lipids: a 13-year mediation and temporal study
.
J Clin Endocrinol Metab
2024
;
109
:
e1494
e1505
27.
Agbaje
AO.
Longitudinal left ventricular mass indexing for DEXA-measured lean mass and fat mass: novel normative reference centiles in postpubertal adolescents and young adults
.
Am J Physiol Heart Circ Physiol
2023
;
324
:
H571
H577
28.
Lee
H
,
Cashin
AG
,
Lamb
SE
, et al.;
AGReMA group
.
A guideline for reporting mediation analyses of randomized trials and observational studies: the AGReMA statement
.
JAMA
2021
;
326
:
1045
1056
29.
Ormazabal
V
,
Nair
S
,
Elfeky
O
,
Aguayo
C
,
Salomon
C
,
Zuñiga
FA.
Association between insulin resistance and the development of cardiovascular disease
.
Cardiovasc Diabetol
2018
;
17
:
122
30.
MacKinnon
DP
,
Krull
JL
,
Lockwood
CM.
Equivalence of the mediation, confounding and suppression effect
.
Prev Sci
2000
;
1
:
173
181
31.
Preacher
KJ
,
Hayes
AF.
Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models
.
Behav Res Methods
2008
;
40
:
879
891
32.
Golding
J
,
Pembrey
M
,
Jones
R
,
ALSPAC Study Team
.
ALSPAC-the Avon Longitudinal Study of Parents and Children. I. Study methodology
.
Paediatr Perinat Epidemiol
2001
;
15
:
74
87
33.
Bjornstad
P
,
Drews
KL
,
Caprio
S
, et al.;
TODAY Study Group
.
Long-Term complications in youth-onset type 2 diabetes
.
N Engl J Med
2021
;
385
:
416
426
34.
Gaasch
WH
,
Zile
MR.
Left ventricular structural remodeling in health and disease: with special emphasis on volume, mass, and geometry
.
J Am Coll Cardiol
2011
;
58
:
1733
1740
35.
Gidding
SS
,
Liu
K
,
Colangelo
LA
, et al
.
Longitudinal determinants of left ventricular mass and geometry: the Coronary Artery Risk Development in Young Adults (CARDIA) Study
.
Circ Cardiovasc Imaging
2013
;
6
:
769
775
36.
Cheng
S
,
Xanthakis
V
,
Sullivan
LM
, et al
.
Correlates of echocardiographic indices of cardiac remodeling over the adult life course: longitudinal observations from the Framingham Heart Study
.
Circulation
2010
;
122
:
570
578
37.
Coutinho
M
,
Gerstein
HC
,
Wang
Y
,
Yusuf
S.
The relationship between glucose and incident cardiovascular events. A metaregression analysis of published data from 20 studies of 95,783 individuals followed for 12.4 years
.
Diabetes Care
1999
;
22
:
233
240
38.
Pahkala
K
,
Laitinen
TT
,
Niinikoski
H
, et al
.
Effects of 20-year infancy-onset dietary counselling on cardiometabolic risk factors in the Special Turku Coronary Risk Factor Intervention Project (STRIP): 6-year post-intervention follow-up
.
Lancet Child Adolesc Health
2020
;
4
:
359
369
39.
Kautzky-Willer
A
,
Leutner
M
,
Harreiter
J.
Sex differences in type 2 diabetes
.
Diabetologia
2023
;
66
:
986
1002
40.
Agbaje
AO.
Waist-circumference-to-height-ratio had better longitudinal agreement with DEXA-measured fat mass than BMI in 7237 children
.
Pediatr Res
2024
;
96
:
1369
1380
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/journals/pages/license.