Age-stratified association between isolated diastolic hypertension and carotid intima-media thickness: results from the Cardiovascular and Metabolic Diseases Etiology Research Center (CMERC) Study

Ethics statement

The study’s protocol was approved by the institutional review boards of Severance Hospital, Yonsei University College of Medicine (No. 4-2013-0661) and Ajou University Hospital, Ajou University School of Medicine (No. AJIRB-BMR-SUR-13-272). Written informed consent was obtained from all participants. This study was conducted in compliance with the principles of the Declaration of Helsinki.

Study population

This study used baseline data from the Cardiovascular and Metabolic Diseases Etiology Research Center (CMERC) cohort, which recruited community-dwelling, middle-aged adults aged 30–64 years between 2013 and 2018. The cohort was designed to identify novel risk factors and develop prevention strategies for cardiovascular and metabolic diseases. Data were collected at two study sites in Korea (Yonsei University College of Medicine, Seoul; Ajou University College of Medicine, Suwon).

Participants were eligible if they had lived at their current residence for at least 8 months, had no plans to relocate within the following 2 years, and could provide informed consent. Exclusion criteria were a history of major CVD (myocardial infarction, heart failure, or stroke), a cancer diagnosis or treatment within the preceding two years, participation in clinical trials, and pregnancy at enrollment. Additional details of the CMERC study design and methodology have been described previously [14,15].

Of the 8,097 initial participants (2,808 men and 5,289 women), we excluded 18 participants with missing BP or carotid IMT data, one participant with missing blood chemistry information, and 1,320 participants who were taking antihypertensive medications. These exclusions ensured accurate BP classification and minimized confounding due to treatment effects. The final study population included 6,759 participants aged 30–64 years (Fig. 1).

Fig. 1.

Flowchart of the study population. CMERC, Cardiovascular and Metabolic Diseases Etiology Research Center; IMT, intima-media thickness.

BP measurements and classification of hypertension subtypes

BP was measured at both study sites following a standardized protocol using an automated oscillometric device (HEM-7080, Omron Health). Participants were asked to rest in a seated position for at least 5 minutes before measurement and to remain relaxed throughout the procedure. Trained research staff obtained three BP readings at 2-minute intervals, and the mean of the second and third measurements was used for analysis. Hypertension subtypes were defined according to the 2017 ACC/AHA guidelines: IDH as SBP <130 mmHg with DBP ≥80 mmHg; ISH as SBP ≥130 mmHg with DBP <80 mmHg; and SDH as SBP ≥130 mmHg with DBP ≥80 mmHg [2]. Based on these classifications, participants were assigned to one of four mutually exclusive categories: no hypertension (including both normal and elevated BP), IDH, ISH, or SDH.

Assessment of carotid IMT

Carotid ultrasonography was performed at Yonsei University College of Medicine using Accuvix XG (Samsung Medison) and at Ajou University College of Medicine using Logiq S8 (GE Healthcare). Trained sonographers conducted all examinations using a standardized protocol [15]. Quality control procedures were applied across sites to ensure consistency in participant positioning and measurement locations. Carotid IMT was assessed using high-resolution B-mode ultrasonography. Interreader and intrareader reliability analyses demonstrated acceptable reproducibility and good agreement among sonographers [14,16].

Carotid IMT was measured in a 1-cm segment of the distal common carotid artery proximal to the carotid bulb, with measurements taken at the R-wave on the electrocardiogram. Both mean and maximum carotid IMT values were obtained using dedicated image analysis software; however, the maximum IMT values from both the left and right carotid arteries were used for the primary analysis. IMT thickening was defined as a maximum carotid IMT greater than or equal to the 75th percentile of the study population, consistent with the American Society of Echocardiography criteria, which consider values above this threshold indicative of increased CVD risk. In this study, the 75th percentile cutoff for maximum IMT was 0.857 mm [17]. Age-adjusted carotid IMT values were further estimated using regression models with age as a covariate.

Covariates

At baseline, trained interviewers collected data on demographic characteristics, socioeconomic status, medical and medication history, and lifestyle factors using a structured questionnaire based on a standardized protocol. Educational attainment was classified as low for participants who had completed high school or less, and as high for those who had attended college or achieved higher education. Physical activity was assessed using the Korean short-form version of the International Physical Activity Questionnaire, which evaluated activity levels over the past 7 days and categorized participants into low, moderate, or high activity groups [18]. Smoking status was classified as never, past, or current. Body weight was measured to the nearest 0.1 kg using a digital scale (DB-150, CAS), and standing height was measured to the nearest 0.1 cm using a stadiometer (DS-102, Jenix). Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared (kg/m²). Fasting blood samples were obtained after at least eight hours of overnight fasting. Serum lipid profiles, including total cholesterol (TC) and high-density lipoprotein cholesterol (HDL-C), were measured using enzymatic methods. High-sensitivity C-reactive protein levels were assessed using a turbidimetric method on the ADVIA 1800 AutoAnalyzer (Siemens Medical Solutions). Diabetes and the use of lipid-lowering medications at baseline were self-reported and recorded as binary variables (yes or no).

Statistical analysis

Participant characteristics were summarized as mean and standard deviation for continuous variables and as frequency and percentage for categorical variables. For skewed continuous variables, median and interquartile range were reported. Baseline characteristics and carotid IMT values were compared across hypertension subtypes to evaluate group differences descriptively without formal hypothesis testing for statistical significance (Table 1).

Table 1.

Characteristics of participants according to hypertension subtypes (n=6,759)

Associations between hypertension subtypes and carotid IMT thickening were estimated using multivariable logistic regression models, with results expressed as odds ratios (ORs) and 95% confidence intervals (CIs). Models were adjusted for age, sex, educational attainment, physical activity, smoking, BMI, TC, HDL-C, high-sensitivity C-reactive protein, diabetes, lipid-lowering drug use, and study site. Age was categorized as <50 years and ≥50 years and further stratified in 10-year intervals, consistent with prior studies on age-related BP changes and the distribution of our study population (median age, 50 years) [7,19]. Age-stratified analyses were then performed to examine potential age-related associations between hypertension subtypes and carotid IMT thickening. Effect modification by age was tested using an interaction term in the regression models.

We performed additional sensitivity analyses to test the robustness of the results. First, participants taking antihypertensive medications (n=1,324) were included, and results were compared with those obtained after excluding these individuals. Second, associations between hypertension subtypes and carotid IMT were evaluated using mean IMT values, calculated as the average of left and right carotid artery measurements. Third, the main analyses were repeated using alternative IMT cutoffs, including the 80th and 90th percentiles, as well as age- and sex-specific thresholds [20,21].

All analyses were conducted using SAS ver. 9.4 (SAS Institute Inc) or R ver. 4.0.3 (R Foundation for Statistical Computing). A P-value of <0.05 was considered statistically significant.

Participants’ characteristics

The final analytical sample included 6,759 participants (mean age, 50.5±8.9 years; 66.9% women) who were not receiving antihypertensive treatment. Of these, 961 (14.2%), 192 (2.8%), and 1,151 participants (17.0%) were classified as having IDH, ISH, and SDH, respectively. Participants with IDH were younger (49.9±8.5 years) than those with ISH (55.7±8.4 years) or SDH (51.9±8.3 years), and they had lower fasting glucose, TC, and BMI levels. HDL-C levels, however, were comparable across groups. Carotid IMT was lower in the IDH group (0.783±0.136 mm) compared with the ISH group (0.882±0.187 mm) and SDH group (0.839±0.155 mm). Similarly, the prevalence of increased IMT (≥0.857 mm) was lower in participants with IDH (24.6%) than in those with ISH (46.9%) and SDH (39.4%) (Table 1).

A total of 4,093 participants (60.6%) were aged ≥50 years. Compared with those <50 years, older participants had higher SBP and DBP, while BMI and HDL-C levels were similar between groups. Participants <50 years had significantly lower IMT (0.694±0.120 mm) compared with those ≥50 years (0.802±0.192 mm), and the prevalence of increased IMT was also lower in the younger group (10.0% vs. 34.8%) (Table S1).

The prevalence of IDH exceeded that of SDH in participants aged 30–39 years (13.6% vs. 12.3%) and 40–49 years (15.3% vs. 13.8%). In contrast, SDH was more common than IDH in participants aged 50–59 years (15.0% vs. 11.3%) and 60–64 years (14.2% vs. 8.4%). A similar distribution was observed among participants with untreated hypertension. Notably, although IDH was predominantly observed in younger adults, many transitioned to SDH or ISH phenotypes over time as SBP increased with aging (Fig. 2) [7].

Fig. 2.

Age-specific proportions of hypertension subtypes by age stratum. (A) Overall participants. (B) Untreated hypertension. Data are presented as the percentage of participants exposed to each hypertension subtype within age groups (30–39, 40–49, 50–59, and 60–64 years). Hypertension subtypes were defined as follows: no hypertension, systolic blood pressure (SBP) <130 mmHg and diastolic blood pressure (DBP) <80 mmHg without antihypertensive treatment; treated hypertension, use of antihypertensive medication regardless of blood pressure level; isolated diastolic hypertension (IDH), SBP <130 mmHg and DBP ≥80 mmHg; isolated systolic hypertension (ISH), SBP ≥130 mmHg and DBP <80 mmHg; and systolic diastolic hypertension (SDH), SBP ≥130 mmHg and DBP ≥80 mmHg.

Association between BP subtypes and carotid IMT

The association between hypertension subtypes and increased carotid IMT varied significantly by age. In the multivariable-adjusted model, IDH was significantly associated with increased carotid IMT among participants <50 years (OR, 1.57; 95% CI, 1.10–2.26), whereas no significant association was observed among those ≥50 years (OR, 0.89; 95% CI, 0.72–1.09). There was a heterogeneity between age <50 versus ≥50 years in the association of IDH with increased carotid IMT (P for interaction=0.003) (Table 2).

Table 2.

Age-stratified analyses of the association between hypertension subtype and carotid IMT (≥0.857 mm vs. <0.857 mm)

Further analyses stratified by 10-year age intervals revealed consistent age-dependent patterns. In the multivariable-adjusted model, IDH was associated with higher carotid IMT in both the 30–39 and 40–49 age groups; however, the association reached statistical significance only in the 40–49 age group (OR, 1.52; 95% CI, 1.00–2.29). In contrast, IDH was inversely associated with carotid IMT in the 50–59 and 60–64 age groups, although these associations were not statistically significant. ISH demonstrated the strongest association with increased carotid IMT across most age groups, except for the 40–49 and 60–64 groups (Table S2).

Sensitivity analyses

Multiple sensitivity analyses were performed to confirm the robustness of the findings. First, when participants taking antihypertensive medications (n=1,324) were included, the positive association between IDH and increased carotid IMT among those <50 years remained statistically significant (Table S3). Second, when mean IMT values (calculated as the average of left and right carotid measurements) were used instead of maximum values, a similar association was observed in younger adults, although statistical significance was lost after multivariable adjustment (Table S4). Third, analyses using alternative IMT cutoffs showed that the 80th percentile (0.882 mm) produced results consistent with the main findings, whereas the 90th percentile (0.972 mm) attenuated the association in participants <50 years (Table S5). Finally, sex- and age-specific analyses demonstrated similar trends, with IDH associated with higher carotid IMT in both men and women <50 years, although these associations were not statistically significant. In contrast, among participants ≥50 years, the direction of association varied by sex (Table S6).

Strengths and limitations

This study has several strengths. The inclusion of a healthy population without clinical CVD allowed for accurate evaluation of IDH, which has often been regarded as conferring modest cardiovascular risk. Standardized protocols and rigorous quality control in the CMERC cohort minimized measurement bias and ensured reliable carotid IMT assessment across sites. Additionally, extensive adjustment for multiple confounders enhanced the robustness of our findings. As relatively few studies have investigated age-specific associations between IDH and carotid IMT in Asian populations, this study provides new insights into an important but understudied group. Nonetheless, several limitations should be noted. First, BP was measured during a single office visit rather than by 24-hour ambulatory BP monitoring, which may have led to misclassification of hypertension subtypes, particularly among younger adults more prone to white-coat or masked hypertension [4]. Second, despite comprehensive covariate adjustment, residual confounding cannot be excluded. Certain medications, such as hormone therapy and corticosteroids among older women, were not fully accounted for [37]. Other potentially important factors, including inflammatory biomarkers, dietary sodium intake, and psychosocial stress, were not incorporated into this analysis [38,39]. Finally, the cross-sectional design precludes causal inference regarding the relationship between IDH and carotid IMT.

Conclusions

These findings underscore the clinical importance of IDH in younger adults and support a proactive approach to its identification and management. Early detection and intervention may help prevent long-term vascular damage and reduce future CVD risk. Further research is warranted to improve age-specific risk stratification and to develop preventive strategies for younger adults with IDH.