Abstract
Background
This large-scale study aims to comprehensively examine associations between cardiovascular health (CVH), arterial injury and clinical cardiovascular disease (CVD) events in primary prevention populations, and quantify the potential mediating role of carotid intima-media thickness (cIMT) in Asia population.
Methods
A total of 358,650 adults aged 35-75 years without baseline CVD were included from a national cohort in China. CVH factors were assessed via Life’s Simple 7 (LS7) metrics. Multilevel regression models were employed to identify cIMT risk factors. Time-to-event analyses for new onset CVD outcomes were conducted using frailty models. Mediation analysis was performed to quantify the proportion of the association between CVH and CVD events mediated through cIMT.
Results
Participants in the highest LS7 quartile had a 0.014 mm lower mean-max cIMT and a 12 % lower risk (OR: 0.88, 95 % CI: 0.86-0.90) of elevated cIMT compared to those in the lowest quartile. Each 0.1 mm higher in mean-max cIMT was associated with a 6 % increased risk of CVD (HR: 1.06, 95 % CI: 1.06-1.07). Mediation analysis demonstrated that cIMT mediated a significant but modest portion of the overall CVH-CVD relationship, accounting for 3.64 % (95 % CI: 2.76 %-4.80 %) of the effect. This mediation effect showed notable heterogeneity, being substantially stronger for factors directly affecting arterial structure and in male participants.
Conclusions
cIMT serves as a partial mediator in the relationship between CVH and incident CVD. Sex-specific and risk-stratified findings highlight cIMT’s potential for personalized CVD prevention.
Keywords: Carotid intima-media thickness, Cardiovascular health, CVD, Mediation
Graphical abstract
Key message.
What is already known on this topic
Despite the recognized significance of cIMT as a cardiovascular risk marker, there is a notable lack of data from specific regions, such as Asia. Inconsistent results are reported in prior studies examining the associations between CVH factors and cIMT, particularly lifestyle behaviours. More evidence is needed for the potential mediating role of cIMT in the relationship between CVH factors and CVD outcomes involving primary prevention populations in Asia
What this study adds
In this observational study based on a large national cohort from China, a more favourable overall CVH status was significantly correlated with lower cIMT levels and a diminished risk of cIMT thickening, although this correlation varied among specific CVH factors. The cIMT mediated partial of cardiovascular health’s effects on CVD events, with stronger effects in males and high-risk groups.
How this study might affect research, practice or policy
The cIMT accounts for only a small portion of the pathway through which CVH factors influence CVD in a primary prevention population., but mediation effects vary depending on specific risk factors, population subgroups, and clinical endpoints. A potential efficient public health strategy is to use a targeted approach that combines CVH promotion for the whole population with selective cIMT assessment in specific population, to help decide on personalized therapy.
1. Introduction
Carotid intima-media thickness (cIMT) can be evaluated via noninvasive carotid ultrasound and is a vital marker of arterial injury. The thicken of cIMT has been proposed as signify unique phenotypes of vascular remodeling besides plaque[1]. Guidelines recommend using cIMT in risk stratification in individuals classified as having intermediate risk[2,3].
Despite the recognized significance of cIMT as a cardiovascular risk marker, there is a notable lack of data from specific regions, such as Asia, where cIMT measurements are generally lower compared to European and American populations[[4], [5], [6]]. The significant associations among cardiovascular health (CVH) factors, cIMT, and cardiovascular disease (CVD) outcomes have been widely investigated individually. However, studies examining the associations between CVH factors and cIMT[[7], [8], [9]], particularly lifestyle behaviours such as diet and physical activity, have reported inconsistent results[10,11], which may be attributed to the challenges in obtaining high-quality data from standard health assessments or routine clinical sources. Additionally, prior research has suggested a potential mediating role of cIMT in the relationship between CVH factors and CVD outcomes based on multi-ethnic cohort[12]; nevertheless, this finding requires validation through larger-scale prospective cohorts, especially those involving primary prevention populations in Asia. It is crucial to emphasize that studying primary prevention populations is important in regions with limited medical resources, where low-risk individuals rarely undergo carotid artery ultrasound examinations during routine care. To the best of our knowledge, previous studies addressing this issue within the Chinese population have predominantly used single-center data[13,14], consequently constraining the variety of risk factors analyzed16[15].
This study aims to address the existing knowledge gap by using data from a cohort encompassing all provinces of China’s mainland. The study aims to investigate the comprehensive associations between cIMT and various CVH factors, as well as related CVD outcomes among primary prevention populations for CVD.
2. Methods
2.1. Study design and participants
The ChinaHEART project is a government-funded public health program designed to screen and intervene in CVD risk among community-based populations throughout China. The project has been described previously[16]. A total of 555,627 participants who underwent carotid ultrasonography at baseline were selected from the entire cohort. After excluding 12,156 (2.19 %) participants who did not complete their questionnaires, 87,760 (15.79 %) with a reported CVD history, and 97,061 (17.47 %) classified as low risk for CVD in 10 years[17], a total of 358,650 participants were included. Missing data were identified in 22,273 (6.21 %) participants, primarily related to physical activity and blood glucose. We used a widely accepted random forest method to impute both categorical and continuous variables (supplementary methods).
This study was performed in line with the principles of the Declaration of Helsinki. The central ethics committee at the China National Center for Cardiovascular Diseases approved this study (approval NO. 2014-574). Informed consent was obtained from all individual participants included in the study.
2.2. Assessment of cIMT
Participants underwent carotid ultrasonography performed by a certified physician at local hospitals following standardized protocols (supplementary methods). Measurement images were taken from the posterior wall of the carotid artery at its minimum inner diameter during the diastolic phase of the cardiac cycle. Measurement sites included both the left and right common carotid arteries (CCA), with three segments assessed on each side: the proximal segment, the middle segment, and the distal segment. Each segment was measured once, resulting in a total of six measurement points for IMT. The commonly used mean-maximum cIMT (mean-max cIMT) was primarily utilized in this study[18]. Specifically, the largest cIMT values on each side were first calculated separately, and the mean-max cIMT represents the average of these two values. Elevated cIMT (thickening) was defined as a value greater than 1.0 mm according to the Chinese expert consensus[19].
2.3. Measures and definitions of CVH factors
For each participant, a standardized interview was conducted in person by trained personnel to collect information on socioeconomic status (SES), lifestyle factors and medical history at baseline. Additionally, blood pressure (BP), fasting blood glucose (FBG), total cholesterol (TC), and body mass index (BMI) were measured at baseline using standardized protocols (supplementary method). The aforementioned CVH factors were categorized and scored according to the Life's Simple 7 (LS7) grading definitions introduced by the American Heart Association (AHA). Each participant's LS7 score was calculated by assigning 2 points for ideal status, 1 point for intermediate status, and 0 points for poor status for each factor, including current smoking, BMI, PA, healthy diet score, TC, BP, and FBG[20]. This total score (ranging from 0 to 14) was then divided into four quartiles, with Q1 (the poorest status) serving as the control group for analysis.
2.4. Outcome assessment
CVD events were collected through a combination of passive and active follow-up methods. Passive follow-up was used to obtain mortality information by linking individual data with the National Mortality Surveillance System and Vital Registration of China Center for Disease Control and Prevention (CDC). Active follow-up involved collecting patients' self-reported hospitalizations for cardiovascular diseases via questionnaires administered during follow-up visits. The primary outcome of this study was the incidence of new CVD events. Secondary outcomes included ischemic heart disease (IHD), ischemic stroke, hemorrhagic stroke and all-cause mortality. A new event was defined as a hospitalization reported at the follow-up visit with the specified cause or a primary cause of death recorded in the CDC surveillance data. All-cause mortality was derived from the surveillance system (supplementary methods).
2.5. Statistical analysis
The standardized mean difference (SMD) was used to compare participant characteristics between sexes. The associations between various CVH factors and cIMT levels or elevated cIMT were analyzed using multilevel regression models, with cIMT levels or the presence of thickening as the dependent variables and province as a random effect. Distal factors (e.g., age, sex, SES, geographic factors such as rural or urban areas, geographical regions, GDP per capita, and average temperature) were adjusted as covariates[8]. The associations between CVH factors and CVD outcomes, as well as between cIMT and CVD outcomes, were analyzed using frailty models, with province as the clustering variable. Mediation analysis, based on the SAS ‘ %mediate’ macro, was used to explore the mediating effect of cIMT on the relationship between CVH and CVD outcomes. Mediation effects were reported, if present, as the proportion of the relationship mediated through cIMT. Details are provided in the supplementary methods.
Subgroup analyses were conducted across sexes, age groups, and CVD risk stratifications. Besides, in order to adjust the baseline health and medication status, stratified analyses were also conducted based on whether the subjects had hypertension, diabetes, dyslipidemia, and taking medications, as well as whether they had carotid artery plaques. P value for interaction among subgroups were also calculated. Statistical analyses were performed using SAS 9.4 (SAS Institute Inc., Cary, North Carolina) and R (version 4.4.2). Statistical significance was defined as a two-sided p-value < 0.05.
3. Results
3.1. Study participant characteristics
The study enrolled a total of 358,650 participants, with a mean age of 61.5 ± 8.17 years, of whom 49.21 % were male. Differences (absolute SMD > 0.2) were observed between sexes, including age, education level, marital status, history of hypertension and dyslipidemia, as well as blood pressure and lipid levels. Females had a superior CVH status, as indicated by a LS7 score of 5.84 ± 1.70 compared to 5.27 ± 2.01 for males. The average mean-max cIMT level in the overall population was 0.9 ± 0.2 mm, with 25.56 % of participants exhibiting levels exceeding 1 mm (Table 1). Differences were observed based on sex and age. Specifically, as age increased, the average mean-max cIMT rose from 0.76 ± 0.17 mm in the 35 to 39 age group to 0.96 ± 0.2 mm in the 70 to 75 age group. Generally, mean-max cIMT values were higher in males than in females, and the average cIMT level and prevalence of elevated cIMT in men were comparable to those of women approximately 5 to 10 years older (Fig. 1).
Table 1.
Characteristics of the study participants according to sex.
| Characteristics | Aggregate | Male | Female | SMD |
|---|---|---|---|---|
| N=358650 | N=176481 | N=182169 | ||
| Age, year, mean [SD] | 61.5±8.17 | 60.31±8.91 | 62.64±7.18 | 0.288* |
| 35-44 | 9214(2.57 %) | 7956(4.51 %) | 1258(0.69 %) | -0.242* |
| 45-54 | 63837(17.80 %) | 39336(22.29 %) | 24501(13.45 %) | -0.232* |
| 55-64 | 133821(37.31 %) | 60887(34.50 %) | 72934(40.04 %) | 0.115 |
| 65-75 | 151778(45.32 %) | 68302(38.70 %) | 83476(45.82 %) | 0.145 |
| Area | ||||
| Urban | 129845 (36.20 %) | 65685 (37.22 %) | 64160 (35.22 %) | -0.042 |
| Rural | 228805 (63.80 %) | 110796 (62.78 %) | 118009 (64.78 %) | 0.042 |
| Region | ||||
| East | 137210 (38.26 %) | 65668 (37.21 %) | 71542 (39.27 %) | 0.042 |
| Central | 107545 (29.99 %) | 52294 (29.63 %) | 55251 (30.33 %) | 0.015 |
| West | 113895 (31.76 %) | 58519 (33.16 %) | 55376 (30.40 %) | -0.059 |
| High school or above | 61271 (17.08 %) | 41050 (23.26 %) | 20221 (11.10 %) | -0.327* |
| Married | 326497 (91.03 %) | 166528 (94.36 %) | 159969 (87.81 %) | -0.231* |
| Household income>50k RMB yuan per year | 64146 (17.89 %) | 36935 (20.93 %) | 27211 (14.94 %) | -0.157 |
| Having medical insurance | 351531 (98.02 %) | 172943 (98.00 %) | 178588 (98.03 %) | 0.003 |
| Occupation: farmer | 194681 (54.28 %) | 91634 (51.92 %) | 103047 (56.57 %) | 0.093 |
| Disease history | ||||
| Hypertension | 328980 (91.73 %) | 155625 (88.18 %) | 173355 (95.16 %) | 0.255* |
| Type 2 diabetes | 128746 (35.90 %) | 57506 (32.58 %) | 71240 (39.11 %) | 0.136 |
| Dyslipidemia | 92300 (25.74 %) | 33283 (18.86 %) | 59017 (32.40 %) | 0.314* |
| Physical examination or lab test | ||||
| SBP, mmHg, mean [SD] | 162.81±18.97 | 159.24±19.91 | 166.28±17.31 | 0.377* |
| DBP, mmHg, mean [SD] | 91.1±12.3 | 92.73±12.62 | 89.51±11.77 | -0.264* |
| TC, mmol/L, mean [SD] | 5.05±1.38 | 4.68±1.32 | 5.4±1.34 | 0.546* |
| LDL-c, mmol/L, mean [SD] | 2.83±1.14 | 2.6±1.09 | 3.06±1.14 | 0.414* |
| HDL-c, mmol/L, mean [SD] | 1.36±0.45 | 1.27±0.45 | 1.45±0.42 | 0.405* |
| FBG, mmol/L, mean [SD] | 6.77±2.1 | 6.67±2.03 | 6.87±2.15 | 0.096 |
| BMI, kg/m2, mean [SD] | 25.77±3.5 | 25.68±3.39 | 25.86±3.6 | 0.054 |
| CVH factors | ||||
| Current smoking: Ideal | 254800 (71.04 %) | 77933 (44.16 %) | 176867 (97.09 %) | 1.428* |
| Intermediate | 2869 (0.80 %) | 2756 (1.56 %) | 113 (0.06 %) | -0.168 |
| Poor | 100981 (28.16 %) | 95792 (54.28 %) | 5189 (2.85 %) | -1.385* |
| BMI: Ideal | 150964 (42.09 %) | 74600 (42.27 %) | 76364 (41.92 %) | -0.007 |
| Intermediate | 162872 (45.41 %) | 81882 (46.40 %) | 80990 (44.46 %) | -0.039 |
| Poor | 44814 (12.50 %) | 19999 (11.33 %) | 24815 (13.62 %) | 0.069 |
| Physical activity: Ideal | 124109 (34.60 %) | 59966 (33.98 %) | 64143 (35.21 %) | 0.026 |
| Intermediate | 6005 (1.67 %) | 2955 (1.67 %) | 3050 (1.67 %) | -0.000 |
| Poor | 228536 (63.72 %) | 113560 (64.35 %) | 114976 (63.12 %) | -0.026 |
| Healthy diet: Ideal | 31878 (8.89 %) | 14659 (8.31 %) | 17219 (9.45 %) | 0.040 |
| Intermediate | 189955 (52.96 %) | 92664 (52.51 %) | 97291 (53.41 %) | 0.018 |
| Poor | 136817 (38.15 %) | 69158 (39.19 %) | 67659 (37.14 %) | -0.042 |
| TC: Ideal | 42759 (11.92 %) | 32001 (18.13 %) | 10758 (5.91 %) | -0.383* |
| Intermediate | 59705 (16.65 %) | 37822 (21.43 %) | 21883 (12.01 %) | -0.254* |
| Poor | 256186 (71.43 %) | 106658 (60.44 %) | 149528 (82.08 %) | 0.493* |
| Blood pressure: Ideal | 5541 (1.54 %) | 4813 (2.73 %) | 728 (0.40 %) | -0.188 |
| Intermediate | 15824 (4.41 %) | 10873 (6.16 %) | 4951 (2.72 %) | -0.168 |
| Poor | 337285 (94.04 %) | 160795 (91.11 %) | 176490 (96.88 %) | 0.245* |
| FBG: Ideal | 89391 (24.92 %) | 46919 (26.59 %) | 42472 (23.31 %) | -0.076 |
| Intermediate | 158798 (44.28 %) | 79729 (45.18 %) | 79069 (43.40 %) | -0.036 |
| Poor | 110461 (30.80 %) | 49833 (28.24 %) | 60628 (33.28 %) | 0.109 |
| Life's Simple 7 score, mean [SD] | 5.56±1.88 | 5.27±2.01 | 5.84±1.7 | 0.306* |
| Q1(0-4) | 105090 (29.30 %) | 64542 (36.57 %) | 40548 (22.26 %) | -0.318* |
| Q2(5-6) | 143415 (39.99 %) | 64325 (36.45 %) | 79090 (43.42 %) | 0.143 |
| Q3(7) | 54853 (15.29 %) | 23221 (13.16 %) | 31632 (17.36 %) | 0.117 |
| Q4(8-14) | 55292 (15.42 %) | 24393 (13.82 %) | 30899 (16.96 %) | 0.087 |
| FRS risk classification | ||||
| Middle | 131702 (36.72 %) | 37785 (21.41 %) | 93917 (51.55 %) | 0.659* |
| High | 226948 (63.28 %) | 138696 (78.59 %) | 88252 (48.45 %) | -0.659* |
| Carotid ultrasound | ||||
| Mean-max cIMT, mm, mean [SD] | 0.9±0.2 | 0.92±0.21 | 0.89±0.19 | -0.127 |
| Elevated cIMT: >1mm | 91671 (25.56 %) | 49885 (28.27 %) | 41786 (22.94 %) | -0.122 |
| Carotid plaque | 186343 (51.96 %) | 95951 (54.37 %) | 90392 (49.62 %) | -0.095 |
a. Abbreviations: BMI, body mass index; cIMT, carotid intima-media thickness; SBP, systolic blood pressure; DBP, diastolic blood pressure; FBG, fasting blood glucose; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; CVH, cardiovascular health; FRS, Framingham Risk Score; SMD, standardized mean difference.
Data are presented as mean±SD or number (percentage) of participants.
b. * absolute SMD>0.2
Fig. 1.
Distribution of mean-max cIMT by age and sex.
Scatters and lines showed the mean level of cIMT (mm) overall and by sex, with error bars showing the SD. Vertical bars showed the proportion of cIMT>1mm overall and by sex, with error bars showing the 95 % CI. The average cIMT level and prevalence of elevated cIMT in males were comparable to those of females who were about 5 to 10 years older.
3.2. Individual CVH factors associated with mean-max cIMT
Individuals in the highest quartile of the Life's Simple 7 score had a mean cIMT approximately 0.014 mm lower than those in the lowest quartile. This association remained significant among males (-0.022 mm), females (-0.007 mm), and high-risk populations (-0.012 mm) when analyzed separately but was no longer significant in the middle-risk subgroups (Table 2 and Table S1). Specifically, an ideal smoking status was associated with a 0.010 mm reduction in mean-max cIMT. Conversely, both an ideal level of physical activity and a healthy diet were linked to a 0.004 mm higher in cIMT. An intermediate blood glucose level was associated with a lower cIMT (-0.009 mm) in the overall population, whereas an ideal blood glucose status was associated with higher cIMT in females only (0.005 mm; Table S1).
Table 2.
Association between CVH factors and mean-max cIMT.
| Risk factors | Estimate (SE) | p value | OR (95 % CI) | p value |
|---|---|---|---|---|
| Current smoking | ||||
| Poor | Ref. | Ref. | ||
| Intermediate | 0.001(0.004) | 0.874 | 1.01(0.93,1.1) | 0.819 |
| Ideal | -0.010(0.001) | <0.001a | 0.91(0.89,0.93) | <0.001* |
| BMI | ||||
| Poor | Ref. | Ref. | ||
| Intermediate | -0.008(0.001) | <0.001a | 0.94(0.92,0.97) | <0.001* |
| Ideal | -0.021(0.001) | <0.001a | 0.85(0.82,0.87) | <0.001* |
| Physical activity | ||||
| Poor | Ref. | Ref. | ||
| Intermediate | -0.004(0.003) | 0.077 | 1(0.93,1.06) | 0.917 |
| Ideal | 0.004(0.001) | <0.001a | 1.06(1.04,1.08) | <0.001* |
| Healthy diet | ||||
| Poor | Ref. | Ref. | ||
| Intermediate | -0.002(0.001) | 0.023a | 0.97(0.95,0.99) | <0.001* |
| Ideal | 0.004(0.001) | 0.003a | 1.03(1,1.06) | 0.089 |
| TC | ||||
| Poor | Ref. | Ref. | ||
| Intermediate | -0.007(0.001) | <0.001a | 0.92(0.9,0.94) | <0.001* |
| Ideal | -0.011(0.001) | <0.001a | 0.9(0.88,0.93) | <0.001* |
| Blood pressure | ||||
| Poor | ||||
| Intermediate | -0.023(0.002) | <0.001a | 0.82(0.78,0.85) | <0.001* |
| Ideal | -0.034(0.003) | <0.001a | 0.74(0.69,0.8) | <0.001* |
| FBG | ||||
| Poor | Ref. | Ref. | ||
| Intermediate | -0.009(0.001) | <0.001a | 0.89(0.88,0.91) | <0.001* |
| Ideal | 0.001(0.001) | 0.096 | 0.99(0.97,1.01) | 0.468 |
| Life's Simple 7 score | ||||
| Q1 | Ref. | Ref. | ||
| Q2 | -0.009(0.001) | <0.001a | 0.92(0.9,0.94) | <0.001* |
| Q3 | -0.011(0.001) | <0.001a | 0.91(0.89,0.94) | <0.001* |
| Q4 | -0.014(0.001) | <0.001a | 0.88(0.86,0.9) | <0.001* |
a. BMI, body mass index; FBG, fasting blood glucose; TC, total cholesterol; SE, standard error; OR, odds ratio.
b. Obtained by using multivariable mixed-effect logistic regression analysis with province as random effect.
c. Adjusted for age, sex, rural or urban area, geographic region, gross domestic product per capita, average temperature, and social-economic status.
d. * p value < 0.05
The association between CVH factors and elevated cIMT was consistent with cIMT levels (Table 2). For example, individuals in the highest quartile of the Life's Simple 7 score had approximately a 12 % lower risk of elevated cIMT compared to those with the poorest status (OR: 0.88, 95 % CI: 0.86-0.9), with a stronger effect observed in males (OR: 0.83, 95 % CI: 0.8-0.87) than in females (OR: 0.93, 95 % CI: 0.9-0.97; Table S1).
3.3. Association between mean-max cIMT and events
Among 333,223 participants who had at least one follow-up visit, there were 25,577 CVD events during a median follow-up of approximately 49 months. A 0.1 mm higher in mean-max cIMT was associated with a 6 % higher risk of CVD (HR: 1.06, 95 % CI: 1.06-1.07), and elevated cIMT was associated with a 22 % higher risk of CVD (HR: 1.22, 95 % CI: 1.19-1.26). Hazard ratios were consistent across sexes and FRS risk classifications (p for interaction>0.05), with a decreasing trend observed with advancing age (Fig. 2). Similar associations were found for IHD (HR: 1.09, 95 % CI: 1.08- 1.1 for a 0.1 mm higher in cIMT; HR: 1.33, 95 % CI: 1.27-1.39 for elevated cIMT), ischemic stroke (HR: 1.04, 95 % CI: 1.03-1.05 for a 0.1 mm higher in cIMT; HR: 1.13, 95 % CI: 1.09-1.18 for elevated cIMT), hemorrhagic stroke (HR: 1.07, 95 % CI: 1.05-1.09 for a 0.1 mm higher in cIMT; HR: 1.26, 95 % CI: 1.17-1.36 for elevated cIMT), and all-cause mortality (HR: 1.06, 95 % CI: 1.05-1.06 for a 0.1 mm higher in cIMT; HR: 1.22, 95 % CI: 1.18-1.26 for elevated cIMT), as shown in Fig S1.
Fig. 2.
Association between mean-max cIMT and CVD occurrence.
The HRs and CIs of cIMT level and elevated cIMT on CVD among all including participants and subgroups were showed in this forest plot. HRs and CIs were calculated based on frailty model, with fully adjusted for age, sex, regional factors, CVH factors and disease history. An 0.1 mm higher in mean-max cIMT was linked to 6 % rise in CVD, and elevated cIMT was associated with a 22 % increase in CVD. Hazard ratios were similar across sexes, age groups and FRS risk classifications.
3.4. Mediation effect of cIMT on the relationship between CVH factors and events
Based on the associations described above and the links between CVH factors and outcomes (Table S2), the mediation effect of cIMT on the relationship between CVH factors and CVD events is presented in Table 3. In the overall population, the proportion of the effect mediated by cIMT on the CVH score for each event was 3.64 % (95 % CI: 2.76 %- 4.8 %) for CVD, 4.51 % (95 % CI: 3.36 %-6.03 %) for IHD, 3.12 % (95 % CI: 2.14 %-4.53 %) for ischemic stroke, 4.15 % (95 % CI: 2.12 %-7.96 %) for hemorrhagic stroke, and 2.13 % (95 % CI: 1.42 %-3.18 %) for all-cause mortality. The mediation effect was observed for smoking, diet, BMI, blood pressure, TC, and FBG, with the largest mediation effect seen for TC, followed by BMI and blood pressure. No mediation effect of cIMT was found for physical activity in relation to these events. Similar findings were noted in subgroup analyses (Table S3 and S4). Overall, males exhibited higher mediation proportions than females (e.g., 6.54 % vs. 2.44 % for the CVH score’s effect on IHD, p for interaction=0.03), and younger participants exhibited higher mediation proportions than elders (e.g., 12.72 % for participants aged 35-44 vs. 2.82 % for participants aged 65-75, p for interaction<0.01). In the plaque-free subgroup, the mediation effects were weaker but still significant.
Table 3.
Mediation effect of cIMT on the relationship between CVH factors and CVD events.
| CVD | IHD | Ischemic stroke | Hemorrhagic stroke | All-cause mortality | |
|---|---|---|---|---|---|
| CVH score (Q1) | 3.64(2.76,4.80) * | 4.51(3.36,6.03) * | 3.12(2.14,4.53) * | 4.15(2.12,7.96) * | 2.13(1.42,3.18)* |
| CVH factors | |||||
| Current smoking | 3.95(2.73,5.68) * | 4.00(2.80,5.69) * | 3.59(2.19,5.83) * | -48.64(NA,NA) | 2.48(1.62,3.77)* |
| Healthy diet | 3.22(1.66,6.13) * | 5.28(1.83,14.30) * | 1.12(0.55,2.29) * | 25.03(0.01,99.87) * | -10.86(NA,NA) |
| BMI | 6.75(4.92,9.20) * | 7.75(5.38,11.03) * | 6.28(3.70,10.45) * | 5.13(2.74,9.39) * | 3.82(2.49,5.80)* |
| Physical activity | -9.76(NA,NA) | 87.65(0.00,100.00) | -8.96(NA,NA) | -2.48(NA,NA) | -3.13(NA,NA) |
| Blood pressure | 4.72(3.45,6.42) * | 23.05(6.27,57.32) * | 2.62(1.81,3.78) * | 1.79(1.17,2.73) * | 12.93(4.56,31.55)* |
| TC | 30.98(7.65,70.86) * | 10.12(6.52,15.38) * | 35.66(0.79,97.46) * | -4.90(NA,NA) | -13.47(NA,NA) |
| FBG | 1.84(1.23,2.73) * | 2.28(1.57,3.29) * | 1.53(0.95,2.43) * | 3.00(1.23,7.15) * | 0.85(0.48,1.50)* |
a. BMI, body mass index; FBG, fasting blood glucose; TC, total cholesterol; CVH, cardiovascular health; CVD, cardiovascular disease; IHD, ischemic heart disease.
b. Mediators’ effects are reported as the proportion (95 % CI) of the relationship between each CVH factor (poorest status) and CVD outcomes mediated through cIMT(mm).
c. * p value<0.05
4. Discussion
This study comprehensively examined the relationships among cardiovascular health factors, the arterial injury marker cIMT, and incident CVD events within a large, nationwide primary prevention cohort. There are several key findings in our study: First, a more favourable overall CVH status was significantly correlated with lower cIMT levels and a diminished risk of cIMT thickening, although this correlation varied among specific CVH factors. Second, cIMT levels exhibited a positive and significant correlation with future CVD risk, a relationship that remained consistent across various CVD event types, genders, and risk strata. Third, mediation analysis indicated that cIMT served as a modest mediator in the relationship between CVH factors and CVD events, with the mediation proportion varying.
Our findings indicate that higher Life's Simple 7 scores, reflecting more ideal cardiovascular health, are associated with lower cIMT levels[7,8,21,22]. The association persisted in the high-risk FRS group but was not significant in the middle-risk group, indicating that in this population, cIMT may be influenced by unmeasured factors beyond traditional LS7 components, such as inflammatory processes[23], or that the middle-risk group is not unhealthy enough to exhibit vascular changes. Analysis of individual CVH components revealed a more nuanced pattern. The robust protective effect of optimal smoking status is in close alignment with current evidence[24], whereas intermediate smoking status exhibited no protective effect, highlighting tobacco's vascular toxicity[25]. Regarding the unexpected finding that the best physical activity and healthy diet statuses were associated with marginal elevations in cIMT levels, these findings may reflect residual confounding, measurement bias or reverse causality. Specifically, given that cIMT and CVH factors were cross sectional measured in this study, and considering that self-reported physical activity and food frequency questionnaires are subject to greater measurement bias, these methodological limitations could partly explain the observed associations. Physiological adaptation mechanisms may also contribute[26,27], however, additional mechanistic research is needed to clarify the underlying pathways. several explanations beyond residual confounding are plausible. It is crucial to emphasize that these findings do not against the well-established cardiovascular benefits of physical activity and healthy diets, as these benefit could be acquired through multiple pathways beyond structural arterial injury, such as inflammatory, hemostatic factors and blood pressure[28]. The inconsistent relationship between optimal blood glucose levels in males and females necessitates further examination, possibly linked to hormonal fluctuations and distinct glucose metabolism in postmenopausal women[29].
The study quantified the mediating effect of cIMT within the CVH to CVD pathway through mediation analysis. The observed differences in mediation proportions between our study and the prior Multi-Ethnic Study of Atherosclerosis(MESA)[12] (3.64 % vs 7.4 %) likely reflect important methodological and population differences. The studies used different CVH assessments, that we employed the LS7 score whereas MESA used the Framingham Risk Score, which incorporates additional factors such as age and sex but lack of BMI, physical activities and blood glucose. Besides, the different follow-up duration differed (7.5 years in MESA vs. 4 years in our study) may affect the observed mediation effects. While both MESA and our study recruited participants from community settings, differences in sampling strategies and risk profiles may also contribute to the observed variations in mediation proportions. MESA enrolled a multi-ethnic, population-based sample of adults free of clinical CVD, whereas our study focused on individuals identified through a public health screening program. These methodological variations highlight the importance of conducting studies within specific designs and population characteristics.
Further analysis revealed substantial heterogeneity, providing insights into the mechanisms of different risk factors. TC, BMI, and blood pressure showed the highest mediation proportions, consistent with their direct roles in promoting atherosclerotic lesions with prior study[12]. However, the divergence in mediation proportions for specific factors (e.g., 6.75 % in our study vs. 17.7 % in the cited work for BMI) additionally highlights the specific risk factor distributions and pathophysiological pathways in the primary prevention cohort from China compared to other populations. The mediation effect was stronger in men (5.1 %) than in women (2.15 %), implicating a greater role of the arterial injury pathway in the male CVH to CVD association, which might be related to estrogen's protective effect[30].
Our findings carry particular significance for developing primary prevention strategies in Asian populations, especially in resource-limited settings. The observed HRs for cIMT predicting future events are modest from a clinical perspective in this study. However, the significant heterogeneity results in this study provides supports for a targeted approach rather than a cost-effective universal screening for general population. Specifically, the mediation effect of cIMT was strongest for factors such as total cholesterol, BMI, and blood pressure. As a visual aid in clinical practice, cIMT may help patients understand and stick to changes in their lifestyles and medications, especially for those already identified with dyslipidemia, hypertension, or obesity. A potential efficient public health strategy is to use a targeted approach that combines CVH promotion for the whole population with selective cIMT assessment in specific population, to help decide on personalized therapy.
The primary strength of this study lies in constructing a comprehensive pathway model linking behavioural and health factors to a subclinical marker (cIMT) and subsequently to clinical endpoints within a large primary prevention Chinese population. Additionally, the study quantified cIMT’s mediating role in this pathway, providing critical evidence regarding the causal chain of cardiovascular disease in an Asian population. The study’s objective was not to provide nationally representative estimates of CVH status or cIMT levels in the general Chinese population, but rather to investigate the relationships among CVH, cIMT, and CVD events within a sufficiently powered large cohort. Several limitations should be considered. First, the observational design precludes definitive causal inference. Second, inter-operator variability in carotid ultrasound is a recognized challenge. The lack of ECG gating, potential variability in diastolic phase identification (particularly in arrhythmia), and dependence on visual angle adjustment may have introduced non-differential measurement error. Although systematic internal reliability assessments were not conducted across the 385 sites, all procedures followed unified standardized protocols and were performed by trained sonographers from local secondary or higher-level hospitals, maximizing data reliability and comparability. Third, despite adjustment for numerous covariates, residual confounding cannot be entirely excluded, particularly for complex behavioural factors such as diet and physical activity, as well as the effects of various drug use and their potential changes over time. Forth, self-report non-fatal events without universal adjudication by medical records is a limitation and may introduce some degree of misclassification. While such misclassification would bias the observed associations toward the null, making our risk estimates more conservative. Fifth, due to data availability, we used the Framingham Risk Score for risk stratification. Future studies with more comprehensive data may evaluate other newer estimators such as PREVENT in Chinese populations. Finally, while the cohort recruited participants from diverse communities across China, it was not a random probability sample. Therefore, caution is warranted when generalizing these findings to the general population, particularly for inferences about population-wide screening strategies.
In summary, this study delineates the pathway linking CVH, arterial injury, and clinical CVD events in a primary prevention population. CVH status, especially smoking, weight management, blood pressure, and lipids, is strongly associated with cIMT. Moreover, cIMT thickening significantly increases the risk of future CVD. However, cIMT accounts for only a small portion of the pathway through which CVH factors influence CVD. The mediation effects vary depending on specific risk factors, population subgroups, and clinical endpoints.
Data sharing
The data are not publicly available. The China Health Evaluation And risk Reduction through nationwide Teamwork (ChinaHEART) project only provides conditional data access for qualified researchers with legitimate requests; a formal application and research proposal is required. Please contact cvd-project@nccd.org.cn to seek approval for data access.
Authors agreement statements
Sources of funding
This work was supported by National Key R&D Program of China [2023YFC2509400].
None of the funders contributed to this manuscript.
Authorship requirements
All authors fulfill the role as an author.
CRediT authorship contribution statement
Chaoqun Wu: Writing – review & editing, Writing – original draft, Formal analysis. Xiaoyan Zhang: Writing – review & editing, Formal analysis. Xueke Bai: Writing – review & editing, Visualization, Validation. Jiapeng Lu: Writing – review & editing, Project administration, Methodology. Yang Yang: Writing – review & editing, Project administration, Investigation. Jianlan Cui: Writing – review & editing, Project administration, Investigation. Lijuan Song: Writing – review & editing, Project administration, Investigation. Wei Xu: Writing – review & editing, Project administration, Investigation. Yan Zhang: Writing – review & editing, Project administration, Investigation. Wenyan He: Writing – review & editing, Project administration, Investigation. Hao Yang: Writing – review & editing, Project administration, Investigation. Dianjianyi Sun: Writing – review & editing, Funding acquisition. Xi Li: Writing – review & editing, Resources, Methodology, Conceptualization. Jun Lv: Writing – review & editing, Methodology, Conceptualization.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Footnotes
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ajpc.2026.101531.
Contributor Information
Xi Li, Email: xi.li@nccd.org.cn.
Jun Lv, Email: epi.lvjun@vip.163.com.
Appendix. Supplementary materials
References
- 1.Tschiderer L., Klingenschmid G., Seekircher L., Willeit P. Carotid intima-media thickness predicts carotid plaque development: meta-analysis of seven studies involving 9341 participants. Eur J Clin Investig. 2020;50(4) doi: 10.1111/eci.13217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Touboul P.J., Hennerici M.G., Meairs S., et al. Mannheim carotid intima-media thickness and plaque consensus (2004-2006-2011). An update on behalf of the advisory board of the 3rd, 4th and 5th watching the risk symposia, at the 13th, 15th and 20th European Stroke Conferences, Mannheim, Germany, 2004, Brussels, Belgium, 2006, and Hamburg, Germany, 2011. Cerebrovasc Dis. 2012;34(4):290–296. doi: 10.1159/000343145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Stein J.H., Korcarz C.E., Hurst R.T., et al. Use of carotid ultrasound to identify subclinical vascular disease and evaluate cardiovascular disease risk: a consensus statement from the American Society of echocardiography carotid intima-media thickness task force. Endorsed by the society for vascular medicine. J Am Soc Echocardiogr: off Publ Am Soc Echocardiogr. 2008;21(2):93–111. doi: 10.1016/j.echo.2007.11.011. quiz 189-90. [DOI] [PubMed] [Google Scholar]
- 4.Liao X., Norata G.D., Polak J.F., et al. Normative values for carotid intima media thickness and its progression: are they transferrable outside of their cohort of origin? Eur J Prev Cardiol. 2016;23(11):1165–1173. doi: 10.1177/2047487315625543. [DOI] [PubMed] [Google Scholar]
- 5.Clarke R., Du H., Kurmi O., et al. Burden of carotid artery atherosclerosis in Chinese adults: implications for future risk of cardiovascular diseases. Eur J Prev Cardiol. 2017;24(6):647–656. doi: 10.1177/2047487317689973. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Fu J., Deng Y., Ma Y., et al. National and provincial-level prevalence and risk factors of carotid atherosclerosis in Chinese adults. JAMA Netw Open. 2024;7(1) doi: 10.1001/jamanetworkopen.2023.51225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Santos I.S., Goulart A.C., Pereira A.C., Lotufo P.A., Benseñor IM. Association between cardiovascular health score and carotid intima-Media thickness: cross-sectional analysis of the Brazilian Longitudinal study of adult health (ELSA-Brasil) baseline assessment. J Am Soc Echocardiogr: Off Publ Am Soc Echocardiogr. 2016;29(12):1207–1216.e4. doi: 10.1016/j.echo.2016.09.001. [DOI] [PubMed] [Google Scholar]
- 8.Nonterah E.A., Crowther N.J., Oduro A., et al. Poor cardiovascular health is associated with subclinical atherosclerosis in apparently healthy sub-Saharan African populations: an H3Africa AWI-Gen study. BMC Med. 2021;19(1):30. doi: 10.1186/s12916-021-01909-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lorenz M.W., Polak J.F., Kavousi M., et al. Carotid intima-media thickness progression to predict cardiovascular events in the general population (the PROG-IMT collaborative project): a meta-analysis of individual participant data. Lancet. 2012;379(9831):2053–2062. doi: 10.1016/s0140-6736(12)60441-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wang Y., Wu H., Sun J., et al. Effect of exercise on carotid artery intima-Media thickness in adults: A systematic review and meta-analysis. J Phys Act Health. 2022;19(12):855–867. doi: 10.1123/jpah.2022-0372. [DOI] [PubMed] [Google Scholar]
- 11.Bhat S., Mocciaro G., Ray S. The association of dietary patterns and carotid intima-media thickness: A synthesis of current evidence. Nutr Metab Cardiovasc Dis: NMCD. 2019;29(12):1273–1287. doi: 10.1016/j.numecd.2019.08.014. [DOI] [PubMed] [Google Scholar]
- 12.Yeboah J., Delaney J.A., Nance R., et al. Mediation of cardiovascular risk factor effects through subclinical vascular disease: the Multi-Ethnic Study of Atherosclerosis. Arterioscler Thromb Vasc Biol. 2014;34(8):1778–1783. doi: 10.1161/atvbaha.114.303753. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Wang L., Niu J.Y., Zhao Z.Y., et al. Ideal cardiovascular health is inversely associated with subclinical atherosclerosis: A prospective analysis. Biomed Environ Sci: BES. 2019;32(4):260–271. doi: 10.3967/bes2019.036. [DOI] [PubMed] [Google Scholar]
- 14.Liu Q., Cui H., Chen S., Zhang D., Huang W., Wu S. Association of baseline life's Essential 8 score and trajectories with carotid intima-media thickness. Front Endocrinol. 2023;14 doi: 10.3389/fendo.2023.1186880. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Li X., Wu C., Lu J., et al. Cardiovascular risk factors in China: a nationwide population-based cohort study. Lancet Public Health. 2020;5(12):e672–e681. doi: 10.1016/s2468-2667(20)30191-2. [DOI] [PubMed] [Google Scholar]
- 16.Wang R., Yang Y., Lu J., et al. Cohort profile: ChinaHEART (Health Evaluation And risk Reduction through nationwide Teamwork) Cohort. Int J Epidemiol. 2023;52(5):e273–e282. doi: 10.1093/ije/dyad074. [DOI] [PubMed] [Google Scholar]
- 17.D'Agostino R.B., Sr., Vasan R.S., Pencina M.J., et al. General cardiovascular risk profile for use in primary care: the Framingham Heart Study. Circulation. 2008;117(6):743–753. doi: 10.1161/circulationaha.107.699579. [DOI] [PubMed] [Google Scholar]
- 18.Ling Y., Wan Y., Barinas-Mitchell E., et al. Varying definitions of carotid intima-Media thickness and future cardiovascular disease: A systematic review and meta-analysis. J Am Heart Assoc. 2023;12(23) doi: 10.1161/jaha.123.031217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Prevention TPCoVUoS, Expert T, Committee of the National Health Commission tPCoSO, Engineering PVUotCMU, Craniocerebral tPCo, Engineering CVUotCMU Expert consensus on some problems of cerebral and carotid vascular ultrasonography(part of carotid) Chin J Cerebrovasc Dis. 2020;17(06):346–353. [Google Scholar]
- 20.Lloyd-Jones D.M., Hong Y., Labarthe D., et al. Defining and setting national goals for cardiovascular health promotion and disease reduction: the American Heart Association's strategic impact goal through 2020 and beyond. Circulation. 2010;121(4):586–613. doi: 10.1161/circulationaha.109.192703. [DOI] [PubMed] [Google Scholar]
- 21.Talegawkar S.A., Jin Y., Kandula N.R., Kanaya A.M. Cardiovascular health metrics among South Asian adults in the United States: prevalence and associations with subclinical atherosclerosis. Prev Med. 2017;96:79–84. doi: 10.1016/j.ypmed.2016.12.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Nève G., Wagner J., Knaier R., et al. Ideal Life's simple 7 score relates to macrovascular structure and function in the healthy population. Nutrients. 2022;14(17) doi: 10.3390/nu14173616. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Engström G., Melander O., Hedblad B. Carotid intima-media thickness, systemic inflammation, and incidence of heart failure hospitalizations. Arterioscler Thromb Vasc Biol. 2009;29(10):1691–1695. doi: 10.1161/atvbaha.109.193490. [DOI] [PubMed] [Google Scholar]
- 24.Mitra S., Biswas R.K., Hooijenga P., et al. Carotid intima-media thickness, cardiovascular disease, and risk factors in 29,000 UK Biobank adults. Am J Prev Cardiol. 2025;22 doi: 10.1016/j.ajpc.2025.101011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Münzel T., Hahad O., Kuntic M., Keaney J.F., Deanfield J.E., Daiber A. Effects of tobacco cigarettes, e-cigarettes, and waterpipe smoking on endothelial function and clinical outcomes. Eur Heart J. 2020;41(41):4057–4070. doi: 10.1093/eurheartj/ehaa460. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Staudt A., Stock K., Gande N., et al. Impact of lifestyle and cardiovascular risk factors on early atherosclerosis in a large cohort of healthy adolescents: the Early Vascular Ageing (EVA)-Tyrol Study. Atherosclerosis. 2020;305:26–33. doi: 10.1016/j.atherosclerosis.2020.05.011. [DOI] [PubMed] [Google Scholar]
- 27.Ried-Larsen M., Grøntved A., Kristensen P.L., Froberg K., Andersen LB. Moderate-and-vigorous physical activity from adolescence to adulthood and subclinical atherosclerosis in adulthood: prospective observations from the European Youth Heart Study. Br J Sports Med. 2015;49(2):107–112. doi: 10.1136/bjsports-2013-092409. [DOI] [PubMed] [Google Scholar]
- 28.Mora S., Cook N., Buring J.E., Ridker P.M., Lee IM. Physical activity and reduced risk of cardiovascular events: potential mediating mechanisms. Circulation. 2007;116(19):2110–2118. doi: 10.1161/circulationaha.107.729939. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Wei M., Gibbons L.W., Mitchell T.L., Kampert J.B., Stern M.P., Blair SN. Low fasting plasma glucose level as a predictor of cardiovascular disease and all-cause mortality. Circulation. 2000;101(17):2047–2052. doi: 10.1161/01.cir.101.17.2047. [DOI] [PubMed] [Google Scholar]
- 30.Holder S.M., Brislane Á, Dawson E.A., et al. Relationship between endothelial function and the eliciting shear stress stimulus in women: changes across the lifespan differ to men. J Am Heart Assoc. 2019;8(4) doi: 10.1161/jaha.118.010994. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.