Study population

The study population included 56 adult patients supervised by the Diabetes Outpatient Clinic of the M. Konopnicka University Pediatrics Centre in Łódź who were between 18 and 28 years old, had T1DM diagnosed in childhood or adolescence, and signed an informed consent to participate in the study and diagnostics.

Based on the duration of T1DM, the patients were divided into group 1 (n = 25; T1DM duration ≤10 years) and group 2 (n = 31; T1DM duration >10 years). All patients had their glycated hemoglobin (HbA_1c) level measured (between 3 weeks before and 3 weeks after a noninvasive testing). Based on the HbA_1c levels, the patients were divided into 2 groups with HbA_1c concentration equal to or below 7.5% and above 7.5%. Additional post hoc 2 × 2 factorial analysis was performed combining a circadian blood pressure (BP) profile (dipper and nondipper) as well as the duration of T1DM dichotomized for the longer (>10 years) and shorter period (≤10 years).

Definitions of medical conditions

For the purpose of this study, overweight was defined as body mass index (BMI) between 25 and 29.9 kg/m². Obesity was defined as a BMI greater than or equal to 30 kg/m². Nonproliferative diabetic retinopathy was defined as microaneurysm, hard or soft exudate, blot hemorrhages, and microinfarcts. Proliferative retinopathy referred to neovascularization, which appeared as a tortuous collection of blood vessels.

Noninvasive diagnostics

All tests assessing endothelial function (PWV, FMD, ankle‑brachial index [ABI]), as well as echocardiography and ultrasound of the carotid arteries, were performed at the Department of Cardiology of the Medical University of Lodz. All patients underwent noninvasive diagnostics including: 1) 24‑hour ambulatory blood pressure monitoring (ABPM), 2) 2‑dimensional transthoracic echocardiography (TTE) including myocardial speckle tracking, 3) aortic PWV measurement, 4) ultrasound of the carotid arteries, 5) ABI measurement, and 6) assessment of brachial artery FMD.

Twenty‑four‑hour ambulatory blood pressure measurement

Out‑of‑hospital 24‑hour ABPM was performed using the 90227 OnTrak recorder (Spacelabs Healthcare, Snoqualmie, Washington, United States). Based on the measurements, the presence of hypertension and the dipper or nondipper circadian BP profile were assessed. Hypertension was diagnosed when mean arterial pressure was above 135/85 mm Hg during periods of activity, above 120/70 mm Hg at rest, and above 130/80 mm Hg during the day. The dipper profile was diagnosed when the decrease in mean BP during bedrest was 10%–20%, as compared with mean for active / day hours. The nondipper profile was reported when the decrease in mean values during night rest was below 10%.

Echocardiography

Two‑dimensional TTE was performed using the Vivid 9 system (GE Healthcare, Horten, Norway). During examination, a minimum of 3 heart cycles were recorded for the apical 2-, 3-, and 4‑chamber views, maintaining the frame rate in the range of 60–80 per second and adjusting the sector width to the size of the myocardium. The recorded image loops were then analyzed offline on an Echo‑PAC v203 workstation (GE Healthcare, United States). Left ventricular global longitudinal strain (GLS) was calculated using automated functional imaging package, and the result was presented as a bull’s‑eye polar map in a 17‑segment model.

Aortic pulse wave velocity

PWV was measured after a 10‑minute rest in the supine position. PWV was determined using a SphygmoCor device (AtCor Medical, Sydney, Australia), based on the distance from the femoral artery to the place where the pulse wave was measured on the common carotid artery.

Ankle‑brachial index

The test was performed by using a specialized ABPI MD device (MESI, Ljubljana, Slovenia) allowing for simultaneous measurements of blood pressure at the arm and around both ankles. The ABI index was calculated automatically by dividing the systolic BP (SBP) at the ankle by the SBP at the arm.

Flow‑dependent vasodilation of the brachial artery

During examination, a sphygmomanometer cuff was placed on the upper arm, proximal to the examination site. A linear Doppler transducer (Vivid, GE Healthcare) was used to visualize the brachial artery by measuring the artery diameter and blood flow velocity using pulsed Doppler. Then, the cuff of the blood pressure monitor was inflated and maintained for 5 minutes at 50 mm Hg above the SBP value. After that time, the cuff was deflated, releasing the pressure. Approximately 60 seconds after releasing the pressure, the maximum diameter of the vessel was measured. FMD expresses the percentage change in the diameter of the brachial artery after ischemia in relation to the initial diameter of the vessel.

Statistical analysis

Statistical analysis was performed using MedCalc software version 12.1.4 (MedCalc, Ostend, Belgium). Normality of quantitative variable distribution was tested using the Shapiro–Wilk test. For comparison of 2 quantitative variables, the t test was employed for those with a normal distribution, and the Mann–Whitney test for those with a non‑normal distribution. Qualitative data expressed as a percentage were examined as a comparison of proportions using the χ² test. The results were considered significant if P value was below 0.05. Correlations were assessed with the Pearson linear correlation test and expressed using the correlation coefficient r. One‑way analysis of variance was used to evaluate differences in the examined parameters between subgroups depending on the daily BP profile and duration of T1DM. Due to the homogeneity of variance in the subgroups, the Scheffe test was used for the post hoc tests. The size of the group was estimated on the basis of available literature.^8-10 Published studies were sourced for data on a difference in the mean value and SD for PWV and ABI in T1DM patients. Calculations using MedCalc software, with α equal to 0.05 as the level of significance were made. They showed that a minimum group size of 56 patients is enough for demonstrating differences in GLS, IMT, and PWV, assuming a 1:1 ratio of the compared groups.

Study group characteristics

A total of 56 patients (48% men) with T1DM were included, at a median (interquartile range [IQR]) age of 22 (18–28) years. The study group characteristics are presented in Supplementary material, Table S1. Median (IQR) duration of T1DM was 11.3 (6–22) years. No patient in the study group had a history of myocardial infarction, coronary angioplasty, or stroke. Hypertension was present in 9 (16%) and thyroid disease in 10 (18%) of the study participants. In the total group of 56 patients, 10 (18%) were overweight (BMI, 25–29.9 kg/m²) and 2 (4%) were obese (BMI, 30–34.9 kg/m²). Routine ophthalmologic examination detected abnormal fundus in 4 (7%), nonproliferative retinopathy in 3 (5%), and preproliferative retinopathy in 1 (2%) patient.

Comparison of noninvasive test results depending on diabetes duration

PWV values were higher in the group 1 (T1DM duration ≤10 years) than in the group 2 (T1DM duration >10 years) (5.3 m/s vs 5.6 m/s; P = 0.01). In contrast, the absolute GLS value was lower in the group 2 than in the group 1 (19.55% vs 20.48%; P = 0.04). In addition, a linear relationship of T1DM duration with GLS and PWV was examined using the Spearman correlation analysis. The results showed a weak but significant negative correlation between T1DM duration and GLS (P = 0.04; r = –0.28), while no correlation was found between T1DM duration and PWV (P = 0.03; r = 0.03).

Comparison of noninvasive test results depending on glycated hemoglobin levels

Between 3 weeks before and 3 weeks after the noninvasive testing, all patients had their HbA_1c level measured. No considerable differences in absolute values were detected between patients with HbA_1c level equal to or below 7.5% (n = 27) and those with HbA_1c level above 7.5% (n = 29) for GLS (20% vs 19.3%; P = 0.16), PWV (5.7 vs 5.7 m/s; P = 0.9), ABI (1.03 vs 1.03; P = 0.82), IMT (0.65 vs 0.66 mm; P = 0.91), and FMD (8.1% vs 7.9%; P = 0.72).

Additional analyses

The number of patients in individual subgroups differed, as there were 17 patients with diabetes duration above 10 years and dipper profile, 14 patients with diabetes duration above 10 years and nondipper profile, 12 patients with diabetes duration up to or equal to 10 years and dipper profile, and 13 patients with diabetes duration up to or equal to 10 years and nondipper profile. The one‑way analysis of variance was performed for GLS, IMT, ABI, and FMD, as normality of their distribution was established. Differences between the subgroups were confirmed for GLS (P = 0.04) and IMT (P = 0.01) (Supplementary material, Table S2). To assess which of the compared subgroups differed from each other, a post hoc analysis with the Scheffe test was performed, which identified differences between: 1) the patients with diabetes duration up to or equal to 10 years and dipper or nondipper profile for GLS (P = 0.04; Figure 1A); 2) the patients with diabetes duration above 10 years and nondipper profile vs those with diabetes duration up to or equal to 10 years and nondipper profile for IMT score (P = 0.01; Figure 1B).

Study limitations

The group we studied came from a single center. This fact ensures a fairly homogeneous therapeutic approach; however, it may differ in other regions of Poland, Europe, and on the global scale. In addition, the study group consisted of relatively young people with a low proportion of complications. Supplementing the observations with cardiac magnetic resonance imaging would add valuable information on the health status of the patients.

Conclusions

Patients with T1DM duration of more than 10 years had significantly lower absolute GLS and significantly higher PWV. In addition, GLS showed a significant linear correlation with T1DM duration. A post hoc analysis showed that the patients with the nondipper daily BP profile and T1DM duration of more than 10 years had a higher mean IMT than those with the nondipper daily BP profile and T1DM duration of up to or equal to 10 years. In contrast, in the group with a shorter duration of T1DM, the nondippers had significantly lower mean absolute GLS values than the dippers. As subclinical target organ damage can be detected even in young adults with T1DM, in the group of patients suffering from T1DM for more than 10 years, echocardiography and PWV evaluation are recommended.