Abstract
Purpose
This study investigated dietary patterns among populations at high risk for cardiovascular disease (CVD) in Hunan Province and examined their associations with mortality.
Methods
Data were obtained from the China Health Evaluation and Risk Reduction through Nationwide Teamwork (ChinaHEART) project conducted between 2014 and 2023. A total of 46,239 individuals aged 35–75 years at high risk for CVD were included in the analysis. Major dietary patterns were identified using exploratory factor analysis with principal components analysis and Varimax rotation. Factor scores were calculated using the regression method. Each participant was then assigned to the dietary pattern with the highest score, and adherence to each pattern was further evaluated by quartiles of factor scores (Q1–Q4). Cox proportional hazards regression models were fitted to estimate hazard ratios (HRs) and 95% confidence intervals (CIs), with sequential adjustment for demographic characteristics, lifestyle factors, and cardiometabolic conditions.
Results
During a median follow-up of 45 months (range: 3–81 months), 792 all-cause deaths were recorded among 46,239 participants. Compared with the “Rice Pattern”, greater adherence to the “Meat, Eggs, and Fresh Vegetables Pattern” was associated with lower risks of all-cause mortality (HR = 0.64, 95% CI: 0.50–0.82; P < 0.001) and chronic cardiovascular disease mortality (HR = 0.44, 95% CI: 0.24–0.77; P < 0.01). Similar patterns of association were observed in subgroup analyses and were generally consistent in sensitivity analyses.
Conclusion
These findings provide evidence-based support for dietary guidance in high-risk populations and have important implications for the prevention and management of cardiovascular disease.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12872-026-05989-1.
Keywords: Cardiovascular diseases, High-risk population, Dietary patterns, All-cause mortality, Specific cause mortality
Introduction
Cardiovascular disease (CVD) remains one of the most important causes of mortality and disease burden worldwide [1]. By 2023, an estimated 18 million deaths annually were attributable to CVD, accounting for approximately 32% of all global deaths [2]. The China Cardiovascular Health and Disease Report 2024 indicates that the number of individuals living with CVD in China has exceeded 350 million [3]. Driven by population aging and the increasing prevalence of unhealthy lifestyles, both the prevalence of cardiovascular disease and the related mortality risk in China remain high [4]. Moreover, the burden of cardiovascular disease in central and western regions has been reported to exceed that in eastern coastal areas [5]. Epidemiological evidence further indicates that the burden of noncommunicable diseases, including cardiovascular disease and diabetes, attributable to elevated body mass index (BMI ≥ 25.0 kg/m²) has steadily increased in Hunan Province over the past three decades [6]. Therefore, strengthening early prevention strategies and health management for populations at high risk of cardiovascular disease is crucial for the effective prevention and control of chronic cardiovascular diseases in China [7].
High-risk cardiovascular individuals are defined as those with a substantially elevated risk of future major adverse cardiovascular events (MACE) or cardiovascular mortality [8]. This population is typically characterized by heightened chronic inflammation, metabolic dysfunction, and vascular impairment and constitutes a major source of cardiovascular mortality and recurrent events [9]. Updated U.S. dietary guidelines emphasize diet as a key strategy for reducing the burden of chronic diseases [10]. As a safe, cost-effective, and sustainable strategy, adherence to healthy dietary patterns has been shown to improve blood pressure and lipid profiles [11], thereby reducing the risk of cardiovascular and cerebrovascular events [12] and alleviating overall disease burden [13]. Moreover, dietary pattern–based interventions have also been recognized as having important implications for the primary prevention of cardiovascular disease among populations without established high cardiovascular risk [14].
Hunan Province, located in south-central China, has a hot, humid climate. The local diet mainly revolves around rice, with cooking methods that use a lot of oil and salt, often including pickled and high-fat foods. Previous studies have indicated that unhealthy dietary patterns-marked by excessive intake of oil, salt, fat, and sugar, alongside low consumption of vegetables and fruits-not only contribute to overweight and obesity [15], but also promote chronic inflammation, impair vascular endothelial function, and increase the risk of cardiovascular mortality [16]. Dietary patterns such as the Mediterranean diet [17], the Dietary Approaches to Stop Hypertension (DASH) diet [18], and plant-based diets [19] have been shown to reduce the incidence and mortality of cardiovascular disease by improving cardiometabolic indicators and exerting anti-inflammatory and antioxidant effects. However, most research on these dietary patterns has been done in Western populations. The cardiovascular effects of the Hunan dietary pattern, characterized by rice as the staple food and high consumption of oil, salt, and spices, remain incompletely elucidated. Furthermore, existing studies have mainly examined the general population, with limited evidence on dietary patterns in high-risk cardiovascular groups.
In this context, a large-scale population cohort was utilized to characterize dietary patterns among high-risk cardiovascular populations in Hunan Province, China. Associations between different dietary patterns and risks of all-cause and cause-specific mortality were examined, providing evidence to inform personalized dietary intervention strategies for high-risk cardiovascular populations.
Materials and methods
Study population
This study was conducted based on the China Health Evaluation And risk Reduction through nationwide Teamwork (ChinaHEART) project, a nationwide initiative aimed at early screening and comprehensive intervention among populations at high risk of cardiovascular disease in China [20]. The project was implemented in 14 prefecture-level cities and autonomous prefectures across Hunan Province between 2014 and 2023. This study population consisted of permanent residents aged 35–75 years who had lived in their communities for at least six months during the 12 months preceding enrollment.
Baseline data, including questionnaires, physical examinations, and laboratory tests, were collected from all enrolled participants. Cardiovascular high-risk individuals were identified based on predefined criteria. After excluding those with missing dietary records (n = 354) or missing survival records (n = 328) within the high-risk group, a total of 46,239 eligible participants were included in the final analysis (Fig. 1).
Fig. 1.
Flow chart of selection of eligible participants
Definition of high-risk cardiovascular individuals
According to the technical specifications of the national project, individuals meeting any of the following criteria were classified as being at high risk for cardiovascular disease: (1) History of major cardiovascular events; (2) Abnormal blood pressure or lipid levels; (3) A 10-year predicted risk of cardiovascular disease ≥ 20% as estimated using the World Health Organization (WHO) 2008 cardiovascular disease risk prediction chart; or (4) Family history of cardiovascular disease or hereditary cardiovascular conditions. Detailed criteria for the identification of high-risk cardiovascular individuals are provided in Table S1.
Dietary assessment
Dietary intake was assessed using a semi-quantitative food frequency questionnaire (FFQ) that recorded the frequency of consumption of major food groups over the past 12 months [21]. The FFQ was developed based on commonly consumed foods in Chinese populations and has been widely used in large-scale epidemiological surveys in China [20, 22]. Common foods were categorized into 12 major groups: rice, wheat, grain, meat, poultry, seafood, eggs, fresh vegetables, pickle, fresh fruits, bean, and dairy. Consumption frequency for each food group was recorded on a five-level scale: “daily = 1”, “4–6 days per week = 2”, “1–3 days per week = 3”, “1–3 days per month = 4”, and “never or rarely consumed = 5”. These scores were subsequently used for dietary pattern analysis.
Specifically, rice included not only cooked rice but also rice-based foods such as porridge and rice noodles, whereas wheat comprised wheat-based products including noodles, steamed buns, and pancakes. Grain represented coarse grains other than rice and wheat (e.g., millet, corn, sorghum, and sweet potatoes). Meat referred to livestock meat and related products, such as fresh or processed pork, beef, mutton, and organ meats, while poultry included chicken, duck, and goose. Seafood encompassed freshwater and marine fish, shellfish, and other aquatic products. Fresh vegetables included a wide variety of unprocessed vegetables, while pickled vegetables referred to those preserved by salting, fermenting, or drying. Bean products included soybean-based foods such as tofu and soy milk, and dairy products included milk, yogurt, cheese, and milk powder.
Mortality outcomes
Follow-up time was calculated from the date of enrollment to the date of death or the end of follow-up (December 31, 2023), whichever occurred first. Participants who were still alive at the end of follow-up were censored.
Causes of death were classified according to the International Classification of Diseases, 10th Revision (ICD-10) codes recorded in the death registry [23]. Based on clinical manifestations, pathophysiological characteristics, and disease progression patterns reported in previous epidemiological studies, cause-specific mortality was categorized into three groups [24]: acute cardiovascular mortality (ICD-10 codes I21–I22, I46, I49.0, and I60–I63), chronic cardiovascular mortality (ICD-10 codes I20, I30–I52, and I70–I79), and non-cardiovascular mortality (all causes excluding ICD-10 codes I00–I99). Detailed ICD-10 classifications are presented in Table S2.
Statistical analysis
To account for differences in measurement scales, dietary scores were standardized before analysis. Exploratory factor analysis was performed using principal components analysis (PCA) in SPSS 25.0, with Varimax rotation (Kaiser-normalized) to improve factor interpretability and independence. The number of retained factors (dietary patterns) was determined based on eigenvalues greater than 1, examination of scree plots, cumulative variance explained exceeding 50%, and interpretability of the patterns. Following previous studies, food groups with absolute factor loadings ≥ 0.5 were used to label each dietary pattern [25]. The robustness of the identified patterns was assessed through sensitivity analysis using Promax rotation. Factor structures and major food group loadings were compared with those obtained from Varimax rotation, and correlations among Promax-derived factors were examined to assess factor independence. For each participant, factor scores for the identified dietary patterns were calculated using the regression method based on the rotated factor loading matrix. Participants were then classified according to the dietary pattern with the highest factor score (dominant-pattern classification), representing the pattern that most strongly characterized their overall dietary intake. To assess the degree of adherence to each dietary pattern, scores were divided into quartiles, with Q1 representing the lowest level of adherence and Q4 representing the highest.
Baseline characteristics of participants across the four dietary patterns were analyzed using RStudio 4.4.3. Categorical variables were summarized as frequencies and percentages, whereas continuous variables were expressed as means ± standard deviations or medians with interquartile ranges, as appropriate. Kaplan-Meier survival curves were constructed to compare survival probabilities among high-risk individuals across dietary patterns, with the “Rice Pattern” selected as the reference due to the predominance of rice-based dietary habits in Hunan Province. Associations between dietary patterns and all-cause or cause-specific mortality were evaluated using Cox proportional hazards regression models. An unadjusted model and three sequentially adjusted models were fitted, with adjustments for demographic characteristics, lifestyle factors, and physiological indicators, respectively, to account for potential confounding. Missing covariates were handled using a complete-case approach, whereby participants with missing values for covariates in each model were excluded from the corresponding analysis. Most covariates had low missing rates (< 5%), and a detailed summary of missing data is provided in Table S3. Results are reported as hazard ratios (HRs) with corresponding 95% confidence intervals (CIs). Further stratified analyses were performed according to age (> 60 years vs. ≤60 years), sex (male vs. female), smoking status (yes vs. no), drinking status (yes vs. no), BMI (≥ 28 vs. <28), and physical activity (≥ 127 vs. <127 MET·h/week), and interaction terms were tested to evaluate the consistency of associations across subgroups.
Results
Types and characteristics of dietary patterns
The results of the exploratory factor analysis indicated that the Kaiser-Meyer-Olkin (KMO) value was 0.809 (> 0.6), and Bartlett’s sphericity test was significant (X² = 106,204.556, P < 0.001), demonstrating that the data were suitable for factor analysis. Four dietary patterns were extracted using PCA with Varimax rotation, with eigenvalues of 3.393, 1.482, 1.046, and 1.038, cumulatively explaining 57.99% of the total variance (Table S4). The scree plot supported retention of four factors (Fig. S1).
Based on factor loadings derived from the 12 food groups, four distinct dietary patterns were identified (Table 1). The “Balanced Pattern” was characterized by higher intakes of wheat, grain, poultry, fresh fruits, bean, and dairy, reflecting a relatively balanced nutritional profile. The “Meat, Eggs, and Fresh Vegetables Pattern” was characterized by higher consumption of meat, seafood, eggs, and fresh vegetables, corresponding to increased protein and dietary fiber intake. The “Pickle Pattern” was predominantly defined by pickled food consumption, whereas the “Rice Pattern” was primarily characterized by high rice intake.
Table 1.
Factor-loading matrix for the major dietary patterns identified by factor analysis
| Balanced Pattern | Meat, Eggs, and Fresh Vegetables Pattern | Pickle Pattern | Rice Pattern | |
|---|---|---|---|---|
| rice | 0.089 | 0.039 | -0.012 | 0.931 |
| wheat | 0.567 | 0.210 | -0.267 | -0.092 |
| grain | 0.713 | -0.193 | 0.279 | 0.061 |
| meat | 0.063 | 0.681 | 0.261 | 0.274 |
| poultry | 0.569 | 0.297 | 0.367 | 0.054 |
| seafood | 0.241 | 0.531 | 0.316 | -0.279 |
| eggs | 0.465 | 0.570 | -0.171 | -0.066 |
| fresh vegetables | -0.190 | 0.628 | -0.093 | -0.006 |
| pickle | 0.123 | 0.053 | 0.827 | -0.020 |
| fresh fruits | 0.671 | 0.114 | -0.028 | 0.098 |
| bean | 0.729 | 0.028 | 0.229 | -0.006 |
| dairy | 0.737 | -0.072 | 0.100 | 0.012 |
Extraction Method: Principal Component Analysis
Rotation Method: Kaiser-normalized Varimax rotation
The component matrix displays factor loadings after convergence through seven iterations of rotation. Bold values indicate food groups with absolute factor loadings >0.5, which were used to characterize and label each dietary pattern
Sensitivity analysis using Promax rotation yielded factor structures largely consistent with those derived from Varimax rotation (Table S5). Major food group loadings on each dietary pattern remained similar, and factor correlations were low to moderate (Table S6), indicating that the patterns were largely independent and robust. These results confirm the stability and reproducibility of the identified dietary patterns.
Baseline characteristics
A total of 46,239 participants were included in the analysis, of whom 41% were male, and the mean age of the cohort was 61 years. Baseline characteristics of the participants are summarized in Table 2. Stratified by dietary pattern, the “Balanced Pattern” accounted for the largest proportion of participants (approximately 30%) and was characterized by a higher representation of rural residents, individuals with lower educational attainment and income levels, and those who reported smoking or alcohol consumption. Participants in the “Meat, Eggs, and Fresh Vegetables Pattern” were predominantly urban residents, had higher educational attainment, and were more likely to be non-smokers and non-drinkers. The “Pickle Pattern” was primarily composed of Han Chinese participants and exhibited a relatively higher proportion of individuals with high income compared with other groups. The “Rice Pattern” accounted for the smallest proportion of participants (18%) and was characterized by relatively higher educational levels and a greater prevalence of habitual salt addition during meals.
Table 2.
Basic characteristics of the study population for the four dietary patterns
| Overall (n = 46,239) |
Balanced Pattern (n = 13,796) | Meat, Eggs, and Fresh Vegetables Pattern (n = 10,900) |
Pickle Pattern (n = 13,217) | Rice Pattern (n = 8,326) |
|
|---|---|---|---|---|---|
| Age, median (IQR), year | |||||
| 61.00(54.00–68.00) | 62.00 (54.00–67.00) | 62.00(54.00–68.00) | 61.00(54.00–67.00) | 60.00(53.00–67.00) | |
| Sex, n (%) | |||||
| Male | 18,815(41%) | 5,613 (41%) | 4,312 (40%) | 5,440 (41%) | 3,450 (41%) |
| Female | 27,424(59%) | 8,183 (59%) | 6,588 (60%) | 7,777 (59%) | 4,876 (59%) |
| Urbanity, n (%) | |||||
| Rural | 34,517(75%) | 12,298 (89%) | 6,747 (62%) | 9,517 (72%) | 5,955 (72%) |
| Urban | 11,722(25%) | 1,498 (11%) | 4,153 (38%) | 3,700 (28%) | 2,371 (28%) |
| Ethnicity, n (%) | |||||
| Han | 42,952(93%) | 12,745 (93%) | 9,334 (86%) | 12,848 (98%) | 8,025 (97%) |
| Ethnic minority | 3,112(6.8%) | 998 (7.3%) | 1,540 (14%) | 314 (2.4%) | 260 (3.1%) |
| Annual income (CNY), n (%) | |||||
| >50,000 | 10,280(22%) | 2,154 (16%) | 2,685 (25%) | 3,662 (28%) | 1,779 (21%) |
| ≦50,000 | 35,959(78%) | 11,642 (84%) | 8,215 (75%) | 9,555 (72%) | 6,547 (79%) |
| Education level, n (%) | |||||
| Below high school | 39,468(85%) | 12,564 (91%) | 9,143 (84%) | 10,982 (83%) | 6,779 (81%) |
| High school or above | 6,771(15%) | 1,232 (8.9%) | 1,757 (16%) | 2,235 (17%) | 1,547 (19%) |
| Smoking status, n (%) | |||||
| No | 35,530(77%) | 10,381 (75%) | 8,510 (78%) | 10,319 (78%) | 6,320 (76%) |
| Yes | 10,709(23%) | 3,415 (25%) | 2,390 (22%) | 2,898 (22%) | 2,006 (24%) |
| Alcohol consumption, n (%) | |||||
| No | 41,592(90%) | 12,195 (88%) | 10,007 (92%) | 11,912 (90%) | 7,478 (90%) |
| Yes | 4,647(10%) | 1,601 (12%) | 893 (8.2%) | 1,305 (9.9%) | 848 (10%) |
| Additional salt added, n (%) | |||||
| No | 38,691(84%) | 12,135 (88%) | 9,002 (83%) | 11,077 (84%) | 6,477 (78%) |
| Yes | 7,434(16%) | 1,651 (12%) | 1,876 (17%) | 2,116 (16%) | 1,791 (22%) |
| BMI, n (%) | |||||
| <28 | 38,630(84%) | 11,331 (82%) | 9,198 (84%) | 11,121 (84%) | 6,980 (84%) |
| ≥28 | 7,609(16%) | 2,465 (18%) | 1,702 (16%) | 2,096 (16%) | 1,346 (16%) |
| BP, median (IQR), mmHg | |||||
| SBP | 159.00(136.00-169.00) | 161.00 (139.00-171.00) | 158.50 (135.00-169.00) | 159.00 (135.00-169.00) | 154.00 (134.00-167.50) |
| DBP | 86.50(77.50–96.00) | 87.00 (78.50–96.50) | 86.00 (77.00-95.50) | 86.50 (77.00–96.00) | 86.50 (78.00-96.50) |
| TC, median (IQR), mmol/L | |||||
| 5.12(4.28–6.19) | 5.10 (4.26–6.11) | 5.09 (4.22–6.20) | 5.08 (4.27–6.14) | 5.26 (4.38–6.38) | |
| Glu, median (IQR), mmol/L | |||||
| 5.80(5.30–6.70) | 5.80 (5.30–6.70) | 5.90 (5.30–6.70) | 5.80 (5.30–6.60) | 5.80 (5.30–6.70) | |
| WC, median (IQR), cm | |||||
| 84.00(78.00–90.00) | 84.00 (78.00–90.00) | 84.00 (79.00–90.00) | 84.00 (78.00–90.00) | 84.00 (78.00–90.00) | |
| Height, median (IQR), cm | |||||
| 156.50(151.50-162.50) | 155.50 (155.00-161.50) | 157.00(151.00-162.80) | 157.00(152.00-162.50) | 158.00(152.50-163.10) | |
| Weight, median (IQR), kg | |||||
| 60.20(54.00–68.00) | 60.00 (53.40, 67.40) | 60.00 (53.90, 67.80) | 60.20 (54.00, 68.00) | 61.10 (54.50, 68.90) | |
Participants were classified into mutually exclusive dietary pattern groups according to the pattern for which they had the highest factor score (dominant-pattern classification)
Survival analysis of all-cause mortality
During follow-up, 792 all-cause deaths were observed among 46,239 participants over a median duration of 45 months (range: 3–81 months). Kaplan-Meier survival analysis demonstrated statistically significant differences in survival curves among the four dietary patterns (log-rank test, P < 0.0001) (Fig. 2). Restricted mean survival time (RMST) analysis indicated that, compared with the “Rice Pattern”, RMST was prolonged by 0.05 months for the “Balanced Pattern”, 0.03 months for the “Meat, Eggs, and Fresh Vegetables Pattern”, and 0.10 months for the “Pickle Pattern” (Table S7). The “Meat, Eggs, and Fresh Vegetables Pattern” exhibited the highest cumulative survival rates among the four dietary patterns at both the 3-year and 5-year follow-up time points, reaching 99.25% (95% CI: 99.07%–99.53%) and 98.17% (95% CI: 97.48%–98.97%), respectively. The survival rate for the “Pickle Pattern” was higher than that of the reference group (“Rice Pattern”) during years 1–4, but lower during years 5–6 (Table S8).
Fig. 2.
Kaplan–Meier curves for all-cause mortality across dietary patterns in high-risk cardiovascular populations. Reference group: “Rice Pattern”
Association between dietary patterns and mortality
Multivariate Cox proportional hazards models were applied to estimate hazard ratios and corresponding 95% confidence intervals for the associations between the four dietary patterns and cause-specific mortality. The “Meat, Eggs, and Fresh Vegetables Pattern” was associated with lower risks of all-cause mortality (HR = 0.64, 95%CI: 0.50–0.82, P < 0.001), Chronic CVD mortality (HR = 0.44, 95%CI: 0.24–0.77, P < 0.01), and non-CVD mortality (HR = 0.71, 95% CI: 0.51–0.98, P < 0.05) (Table 3). After full adjustment, a low risk of non-CVD mortality was observed for the “Balanced Pattern” (HR = 0.70, 95%CI: 0.53–0.93, P < 0.05). Compared with the lowest quartile (Q1), individuals in the highest quartile (Q4) of adherence to the “Meat, Eggs, and Fresh Vegetables” pattern showed a lower risk of all-cause mortality (HR = 0.785, 95%CI: 0.64–0.97, P = 0.023). A similar trend between higher adherence and lower mortality risk was suggested for other causes of death; however, no statistically significant differences were found (Table S9–S12).
Table 3.
Person-years, mortality rates (per 1000 PY), and hazard ratios (95% CI) for all-cause and cause-specific mortality across dietary patterns
| Balanced Pattern | Meat, Eggs, and Fresh Vegetables Pattern | Pickle Pattern | Rice Pattern | |
|---|---|---|---|---|
| Person-years (PY) | 54,323.8 | 36,111.9 | 46,377.7 | 26,876.8 |
| All-cause mortality | ||||
| No. of deaths | 304 | 124 | 229 | 135 |
| Mortality rate (/1000PY) | 5.60 | 3.43 | 4.94 | 5.02 |
| Model 1a | 1.08(0.88–1.33) | 0.66(0.51–0.84) *** | 0.99(0.80–1.22) | 1(Ref.) |
| Model 2 b | 0.82(0.67–1.01) | 0.64(0.50–0.83) *** | 0.90(0.73–1.12) | 1(Ref.) |
| Model 3 c | 0.82(0.67–1.01) | 0.64(0.50–0.83) *** | 0.90(0.73–1.12) | 1(Ref.) |
| Model 4 d | 0.83(0.68–1.02) | 0.64(0.50–0.82) *** | 0.91(0.74–1.13) | 1(Ref.) |
| Acute CVD mortality | ||||
| No. of deaths | 89 | 27 | 59 | 28 |
| Mortality rate(/1000PY) | 1.64 | 0.75 | 1.27 | 1.04 |
| Model 1a | 1.54(1.01–2.35) * | 0.68(0.40–1.15) | 1.22 (0.78–1.92) | 1(Ref.) |
| Model 2 b | 1.19(0.77–1.84) | 0.67(0.39–1.15) | 1.14(0.72–1.78) | 1(Ref.) |
| Model 3 c | 1.20(0.78–1.84) | 0.67(0.39–1.15) | 1.13(0.72–1.78) | 1(Ref.) |
| Model 4 d | 1.20(0.78–1.85) | 0.69(0.40–1.17) | 1.14(0.73–1.80) | 1(Ref.) |
| Chronic CVD mortality | ||||
| No. of deaths | 71 | 23 | 45 | 28 |
| Mortality rate(/1000PY) | 1.31 | 0.64 | 0.97 | 1.04 |
| Model 1a | 1.24(0.80–1.92) | 0.56(0.32–0.98) * | 0.93(0.58–1.49) | 1(Ref.) |
| Model 2 b | 0.80(0.51–1.26) | 0.45(0.26–0.80) ** | 0.80(0.50–1.29) | 1(Ref.) |
| Model 3 c | 0.82(0.52–1.28) | 0.50(0.25–0.79) ** | 0.80(0.50–1.28) | 1(Ref.) |
| Model 4 d | 0.81(0.52–1.28) | 0.44(0.24–0.77) ** | 0.81(0.50–1.30) | 1(Ref.) |
| Non-CVD mortality | ||||
| No. of deaths | 144 | 74 | 125 | 79 |
| Mortality rate(/1000PY) | 2.65 | 2.05 | 2.70 | 2.94 |
| Model 1a | 0.87(0.66–1.14) | 0.68(0.50–0.94) * | 0.92(0.70–1.22) | 1(Ref.) |
| Model 2 b | 0.69(0.52–0.91) ** | 0.72(0.52–0.99) * | 0.86(0.65–1.14) | 1(Ref.) |
| Model 3 c | 0.69(0.52–0.91) ** | 0.72(0.52–0.99) * | 0.86 (0.65–1.14) | 1(Ref.) |
| Model 4 d | 0.70(0.53–0.93) * | 0.71(0.51–0.98) * | 0.86(0.65–1.15) | 1(Ref.) |
PY = person-years of follow-up
Mortality rate = deaths per 1000 PY
*: p < 0.05, **: p < 0.01, ***p < 0.001. Statistically significant results are shown in bold
a Model 1: No adjustments made
b Model 2: Adjusted for age, sex, urbanity, ethnicity, annual household income, and education level
c Model 3: Model 2 further adjusted for smoking status and alcohol consumption
d Model 4: Model 3 further adjusted for BMI, hypertension, dyslipidemia, and diabetes
Subgroup and interaction analyses
Subgroup analyses were performed based on the fully adjusted Model 4, with the corresponding stratification variable excluded from each model. No evidence of heterogeneity was observed across subgroups in the overall association between dietary patterns and mortality outcomes (Fig. 3). The “Meat, Eggs, and Fresh Vegetables” pattern was associated with lower risks of all-cause and chronic cardiovascular mortality in several subgroups.
Fig. 3.
Forest plots of hazard ratios (HRs) and 95% confidence intervals (CIs) for the associations between dietary patterns and different causes of death. Hazard ratios were estimated using Model 4, which included all covariates except the stratification variable corresponding to each subgroup (age, sex, urbanity, ethnicity, annual household income, education level, smoking status, alcohol consumption, BMI, hypertension, dyslipidemia, and diabetes). Reference group: “Rice Pattern”
Across a range of demographic, socioeconomic, and clinical characteristics (including age, sex, urbanity, income, education, smoking, alcohol consumption, hypertension, dyslipidemia, diabetes, physical activity, waist circumference, and BMI), no statistically significant interactions with dietary patterns were detected (P interaction >0.05) (Table S13–S16).
Sensitivity analyses
After excluding participants with follow-up durations of less than one year, no substantial changes were observed in the estimated risk of all-cause mortality, acute cardiovascular mortality, chronic cardiovascular mortality, or non-cardiovascular mortality across the four dietary patterns (Table S17–S20).
Discussion
Compared with the reference group (“Rice Pattern”), greater adherence to the “Meat, Eggs, and Fresh Vegetables Pattern” exhibited longer median survival times and higher 3-year and 5-year survival rates. In multivariable Cox models adjusting for sociodemographic characteristics, lifestyle factors, and cardiometabolic conditions (Models 1–4), this pattern was associated with lower risks of all-cause mortality, chronic CVD mortality, and non-CVD mortality. Similar associations were observed across subgroups and were generally consistent after excluding participants with follow-up shorter than 1 year, in line with previous studies [26, 27].
From a broader perspective, the dietary pattern identified in this study shows similarities to several dietary models that have been linked to favorable cardiovascular outcomes, including the Mediterranean diet [28], the DASH diet [29], and the Oriental Healthy Dietary Pattern (OHDP) [30]. Higher intake of vegetables and balanced protein sources is commonly emphasized in these dietary models. Previous epidemiological studies have reported that dietary patterns rich in vegetables and moderate protein sources are associated with lower risks of cardiovascular events and mortality [31].
Several biological mechanisms have been proposed that may help explain these observed associations. Diets with higher-quality protein and lower intake of refined carbohydrates have been linked to improved glycemic control, more favorable lipid profiles [32], and maintenance of skeletal muscle mass [33]. These changes potentially reduce metabolic burden. In addition, fresh vegetables provide polyphenols and antioxidants that may help reduce systemic inflammation and oxidative stress, both of which are implicated in the progression of cardiovascular disease [34]. In contrast, pickled vegetables, a traditional dietary component in Hunan Province, were reported to be associated with elevated blood pressure, possibly due to increased sodium intake, plasma volume expansion, and activation of the renin–angiotensin system [35].
Overall, greater adherence to the “Meat, Eggs, and Fresh Vegetables Pattern” appeared to be associated with more favorable survival outcomes. It should be noted that this pattern combines heterogeneous food items, including different types of meat and eggs. Therefore, it may reflect broader dietary diversity rather than the effect of any single food item [36]. Moreover, the dietary patterns derived in this study primarily indicate the regularity of food consumption rather than absolute intake levels, as dietary assessment was based on a semi-quantitative food frequency questionnaire. Therefore, they may not fully capture actual consumption quantities, which should be considered when interpreting the observed associations. Currently, methods for assessing dietary intake are increasingly utilizing artificial intelligence technologies such as ChatGPT [37] and convolutional neural network [38] models to identify intake more accurately.
Participants classified in the “Meat, Eggs, and Fresh Vegetables” pattern exhibited more advantageous baseline profiles, characterized by a higher proportion of urban residents, greater educational attainment, and lower prevalence of smoking and alcohol consumption. Despite adjustment for a range of socioeconomic and behavioral covariates, residual confounding cannot be ruled out. Unmeasured factors—such as access to healthcare, health literacy, medication adherence, and other health-related behaviors—are likely to differ systematically across dietary pattern groups and may contribute to the observed associations [39]. Besides, comprehensive lifestyle management—including tobacco control, moderate alcohol intake, and regular physical activity—remains essential for reducing cardiovascular risk [40]. Accordingly, the reduced mortality risk observed in this group may, at least in part, reflect an overall healthier socioeconomic and behavioral profile rather than an independent effect of the dietary pattern itself [41].
Several limitations of this study should be acknowledged. First, the observational design and reliance on baseline dietary data may not adequately capture concurrent adherence to multiple dietary patterns or changes in dietary behavior over time. In populations at high cardiovascular risk, dietary habits are likely to evolve in response to clinical advice, disease progression, or lifestyle modifications; therefore, baseline patterns may not fully represent long-term exposure, potentially leading to non-differential misclassification and attenuation of the observed associations. Furthermore, the study population consisted of individuals at elevated cardiovascular risk, many of whom likely met multiple risk criteria simultaneously. The lack of stratified analyses across specific high-risk subgroups may therefore limit the interpretability of the findings. Finally, although the overall sample size was relatively large, the number of deaths, particularly for cause-specific outcomes such as acute and chronic cardiovascular mortality, was limited. This may have affected the precision and stability of hazard ratio estimates, especially in fully adjusted models, and may have constrained the robustness of subgroup and interaction analyses. Further research is warranted to provide additional evidence on the relationship between dietary patterns and mortality risk, with an emphasis on repeated dietary assessments, stratification of high-risk populations, objective biomarkers, and precision nutrition approaches [42].
Conclusion
In summary, four dietary patterns were identified among individuals at high cardiovascular risk in Hunan Province. The main features of these patterns were described in this study. Greater adherence to the “Meat, Eggs, and Fresh Vegetables Pattern” was associated with more favorable survival outcomes. This pattern also showed similarities to the principles of the Oriental Healthy Dietary Pattern. These findings underscore the potential importance of overall dietary patterns in cardiovascular prognosis and provide evidence to inform dietary recommendations for populations at high cardiovascular risk.
Supplementary Information
Acknowledgements
The authors thank all participants and staff involved in the China Health Evaluation And risk Reduction through nationwide Teamwork (ChinaHEART) project for their contributions to this study.
Authors’ contributions
**Yilin Lv: ** Writing – original draft, Visualization, Formal analysis, Data curation. **Xian Xie: ** Validation, Data curation, Visualization. **Yuan Liu: ** Project administration, Conceptualization, Software. **Lei Yin: ** Resources, Supervision. **Li Yin: ** Writing - review and editing, Data curation. **Xingli Li** : Conceptualization, Funding acquisition, Supervision, Writing – review and editing.
Funding
The study is funded by Noncommunicable Chronic Disease-National Science and Technology Major Project (2026ZD0558000, 2026ZD0558002).
Data availability
The datasets analyzed during the current study are available from the corresponding author upon reasonable request (Prof. Xingli Li; lixingli@csu.edu.cn).
Declarations
Ethics approval and consent to participate
The ChinaHEART project has been approved by the Central Ethics Committee of Fuwai Hospital (2014 − 574) and registered on ClinicalTrials.gov (NCT02536456). All participants provided written informed consent at enrollment. All procedures involving human participants were performed in accordance with the ethical standards of the institutional research committee and with the Declaration of Helsinki.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Li Yin, Email: 8273620@qq.com.
Xingli Li, Email: lixingli@csu.edu.cn.
References
- 1.Angela P, Danilo NG. The burden of hypercholesterolemia and ischemic heart disease in an ageing world. Pharmacol Res. 2023;193:106814–106814. [DOI] [PubMed] [Google Scholar]
- 2.I. TCWAAWAZ, Pankaj ACAMA, L. AC, Z. BSCMBA, B.A. K., B.A. E. CM, Yvonne EMSV, Chete EKREN, Setri F, Giuliano G, Swapnil HDGH, Rizwan HJEK, Darae KDSK, Junxiu LDAL, Jun M, D. MJWME, M.M., Remy P, Mary RH, Pierre RGASNSSOM, T.E., Yuh VSSVJHWN, Kristine Y. M.S. S., Heart Disease and Stroke Statistics—2023 Update: A Report From the American Heart Association, CirculationE., N.S. D., P.N. I. 2023.
- 3.Liu MB, He XY, Yang XH, Wang ZW. Interpretation of Annual Report on Cardiovascular Health and Diseases in China 2024. Biomed Environ Sci: BES. 2025;38:893–917. 10.3967/bes2025.099. [DOI] [PubMed]
- 4.H. Center For Cardiovascular Diseases The Writing Committee Of The Report On Cardiovascular, N. Diseases In China, Report on Cardiovascular Health and Diseases in China 2023: An Updated Summary. Biomed Environ Sci. 2024;37:949–92. 10.3967/bes2024.162. [DOI] [PubMed] [Google Scholar]
- 5.Zheng C, Wang X, Gu R, Wang W, Tian Y, Pei X, Jia Q, Song Q, Cao X, Tian Y, Tuerdi N, Sun M, Liu S, Zhang Y, Wang Z. The Burden of Cardiovascular Disease in China: Incidence Estimates From National Surveillance (2023). J Am Coll Cardiol. 2025. 10.1016/j.jacc.2025.08.060. [DOI] [PubMed] [Google Scholar]
- 6.He X, Zhang X, Mao F, Yao Y, Zhang R, Li J. Disease burden of non-communicable diseases attributable to high body mass index in China and its provinces, 1990–2021: A systematic analysis for the Global Burden of Disease Study 2021. Chin Med J (Engl). 2025. 10.1097/cm9.0000000000003900. [DOI] [PubMed] [Google Scholar]
- 7.Li Y, Zhang C, Chen F, Zhang J. Prevalence and influencing factors of subjects at high risk for cardiovascular disease in different regions of Gansu province, China: a cross-sectional study of 100,725 residents from 2017 to 2022. Front Cardiovasc Med. 2024;11:1373123–1373123. 10.3389/fcvm.2024.1373123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Mach F, Baigent C, Catapano AL, Koskinas KC, Casula M, Badimon L, Chapman MJ, De Backer GG, Delgado V, Ference BA, Graham IM, Halliday A, Landmesser U, Mihaylova B, Pedersen TR, Riccardi G, Richter DJ, Sabatine MS, Taskinen MR, Tokgozoglu L, Wiklund O. 2019 ESC/EAS Guidelines for the management of dyslipidaemias: lipid modification to reduce cardiovascular risk. Eur Heart J. 2020;41:111–88. 10.1093/eurheartj/ehz455. [DOI] [PubMed] [Google Scholar]
- 9.Nurul Izzati A, Fatin Syazwani AM, Kahar F, Amrina MA, Izzudin MP, Sazlina SG, Sabariah MN. Integrating CCL2 and TNF-α into the Framingham Risk Score for cardiovascular risk prediction: a cross-sectional study in a Malaysian cohort. Med J Malaysia. 2025;80:133–40. [PubMed] [Google Scholar]
- 10.U.S.D. o.A.a.U.S.D.o.H.a.H. Services, Dietary Guidelines for Americans, 2025–2030, Washington, DC, 2026.
- 11.Tian W, M KC, Sophie C, Sayan M, V RR, Shane J, Andrius M, M SA, Luigi F. Vegetarian Dietary Patterns and Cardiometabolic Risk in People With or at High Risk of Cardiovascular Disease: A Systematic Review and Meta-analysis. JAMA Netw open. 2023;6:e2325658–2325658. 10.1001/jamanetworkopen.2023.25658. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cherian L, Agarwal P, Agrawal S, James BD, Yang D, Wagner M, Leurgans SE, Bennett DA, Aggarwal NT, Schneider JA. Dietary Patterns Associated With Risk of Intracranial Atherosclerosis in Older Adults With Hypertension or Myocardial Infarction. Neurology. 2025;105:e214147. 10.1212/wnl.0000000000214147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Shreyaa R, Twisha P, sai k, Vaidheesh V, AHMED SS, Ridham P, Gunjan K, Anusha P. Dineshbhai, s. deepika, Abstract 4371831: Global Burden and Trends in Cardiovascular Disease Attributable to Dietary Risk Factors Among Women from 1990–2021: A Global Benchmarking Systematic Analysis. Circulation. 2025;152:A4371831–4371831. 10.1161/circ.152.Suppl_3.4371831. [Google Scholar]
- 14.Daneshzad E, Goodarzi S, Ghoreshi R, Shahrestanaki E, Hatami SE, Assar O, Esmaeili Abdar Z, Sadeghi T, Zarif Yeganeh M, Mahmoodi Z, Qorbani M. The effect of a lifestyle intervention package on cardiovascular risk factors in the first grade relatives of hospitalized patients with cardiovascular diseases: a protocol of parallel-randomized trial. J Diabetes Metab Disord. 2025;24:188. 10.1007/s40200-025-01697-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Hu Y, Zhang Y, Zhong J, Wang Y, Zhou E, Hong F. Association between obesity phenotypes and dietary patterns: A two-step cluster analysis based on the China multi-ethnic cohort study. Prev Med. 2024;187:108100. 10.1016/j.ypmed.2024.108100. [DOI] [PubMed] [Google Scholar]
- 16.Wilcox NS, Amit U, Reibel JB, Berlin E, Howell K, Ky B. Cardiovascular disease and cancer: shared risk factors and mechanisms, Nature reviews. Cardiology. 2024;21:617–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.R.I.-H.S.C.i.S.P. Cardiovascular Program-ICCC, Barcelona, Spain., M. CIBERCV Instituto de Salud Carlos III, Spain., R.I.-H.S.C.i.S.P. Cardiovascular Program-ICCC, Barcelona, Spain., D.o. Food, U.F.d.S.M. Nutrition, Bolsista do CNPq, Brazil., R.I.-H.S.C.i.S.P. Cardiovascular Program-ICCC, Barcelona, Spain., Diet and Cardiovascular Disease: Effects of Foods and Nutrients in Classical and Emerging Cardiovascular Risk Factors, Current medicinal chemistry. 2019;26:3639-3651. [DOI] [PubMed]
- 18.Maneeth M, Madiha K, Lakshmi MK, Prachi D, Teja RD, Namratha E, DEEPALI S, Ravi M. Abstract TAC134: Impact of Mediterranean Diet on Hypertension in Patients with High Cardiovascular disease Risk: A Systematic Review&Meta-Analysis of Randomized Controlled Trials. Hypertension. 2025;82. 10.1161/hyp.82.Suppl_1.Tac134.
- 19.Correia PE, Bisi L, Zhang M, Sun Y, Martins BB, Porepp OSC, Colpani V, Kunzler LB, Teixeira PP, Ferrari GT, Zajdenverg L, Brietzke E, Socal MP, Gerchman F. Plant and Animal-Based Dietary Patterns and Cardiometabolic Diseases in the Brazilian Population: Cross-Sectional Analysis of the Brazilian National Health Survey. Nutrients. 2025;17:3448–3448. 10.3390/nu17213448. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wang R, Yang Y, Lu J, Cui J, Xu W, Song L, Wu C, Zhang X, Dai H, Zhong H, Jin B, He W, Zhang Y, Yang H, Wang Y, Zhang X, Li X, Hu S. Cohort Profile: ChinaHEART (Health Evaluation And risk Reduction through nationwide Teamwork) Cohort. Int J Epidemiol. 2023;52:e273–82. 10.1093/ije/dyad074. [DOI] [PubMed] [Google Scholar]
- 21.Lu J, Xuan S, Downing NS, Wu C, Li L, Krumholz HM, Jiang L. Protocol for the China PEACE (Patient-centered Evaluative Assessment of Cardiac Events) Million Persons Project pilot. BMJ Open. 2016;6:e010200. 10.1136/bmjopen-2015-010200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Chen S, Hu S, Shen S, Zhang J, Xu X, Yu M, Xia Y, Cai Q, Yu W, Lu A, Li ZM, Gunarathne R, Lu J. Dietary patterns and cardiovascular diseases among Chinese high-risk population aged 35 years and older: A 6-year cohort study. Curr Res Food Sci. 2025;10:100960. 10.1016/j.crfs.2024.100960. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Organization WH. International statistical classification of diseases and related health problems: alphabetical index: world health organization. 2004.
- 24.Boeddinghaus J, Lopez-Ayala P, Thurston AJF, Lee KK, Tew YY, Doudesis D, Kadesjö E, Roos A, Raphael CE, Sandoval Y, Gulati R, Taggart C, Wereski R, Koechlin L, Anand A, Mahfoud F, Neumann JT, Strauß B, Sörensen NA, Smilowitz NR, Newby DE, Mueller C, Jaffe A, Mills NL, Vallejos CA, Chapman AR. Cardiovascular Risk Across Myocardial Injury and Infarction Categories Using the Universal Definition: An Individual Patient-Level Data Meta-Analysis. J Am Coll Cardiol. 2025. 10.1016/j.jacc.2025.11.043. [DOI] [PubMed] [Google Scholar]
- 25.Winter JCFd, Dodou D. Common Factor Analysis versus Principal Component Analysis: A Comparison of Loadings by Means of Simulations, Communications in Statistics -. Simul Comput. 2016;45:299–321. 10.1080/03610918.2013.862274. [Google Scholar]
- 26.Hu F, Qin W, Xu L. Association between Dietary Patterns and Cardiometabolic Multimorbidity among Chinese Rural Older Adults. Nutrients. 2024;16. 10.3390/nu16172830. [DOI] [PMC free article] [PubMed]
- 27.Vidal PM, Tsampasian V, Cassidy A, Zoccai GB, Chrysohoou C, Koskinas K, Verschuren WMM, Czapla M, Kavousi M, Kouvari M, Vassiliou VS, Panagiotakos D. Diet and nutrition in cardiovascular disease prevention: a scientific statement of the European Association of Preventive Cardiology and the Association of Cardiovascular Nursing & Allied Professions of the European Society of Cardiology. Eur J Prev Cardiol. 2025;32:1540–52. 10.1093/eurjpc/zwaf310. [DOI] [PubMed] [Google Scholar]
- 28.Belokon CO, Palacios SG, Gabucio LMC, Cánovas AO, Hera MGdl, Vioque J, Collado LT. Adherence to Three Mediterranean Dietary Indexes and All-Cause, Cardiovascular, and Cancer Mortality in an Older Mediterranean Population. Nutrients. 2025;17:2956–2956. 10.3390/nu17182956. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Sun J, Jiang X, Li Z, Shen Y. Impact of dietary patterns on the survival outcomes of patients with cardiovascular disease. Front Nutr. 2025;12:1535174–1535174. 10.3389/fnut.2025.1535174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Wang X, Xuan J, Chen H, Lin B, Yuan C. Eastern healthy dietary patterns in relation to all-cause mortality. Eur J Nutr. 2025;64:286. 10.1007/s00394-025-03807-6. [DOI] [PubMed] [Google Scholar]
- 31.Conceição ARd, Juvanhol LL, Marcadenti A, Ferreira ÂCB, Weber B, Bressan J. Diet quality scores and incidence of cardiovascular events: A 4-year prospective study of patients in cardiology secondary care (BALANCE Program Trial). Br J Nutr. 2025;1–28. 10.1017/s000711452510559x. [DOI] [PubMed]
- 32.Zhang W, Zou L, Chen J. The impact of physical activity and dietary habits on glycolipemic metabolism and inflammatory markers in the elderly: a cross-sectional study. Front Nutr. 2025;12:1600038–1600038. 10.3389/fnut.2025.1600038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Arslan S, Aydın A. The weight-loss paradox in older adults: balancing fat loss with muscle preservation. Bull Natl Res Centre. 2025;49:79–79. 10.1186/s42269-025-01374-8. [Google Scholar]
- 34.Fioroni N, Rivier CM, Meudec E, Cheynier V, Boudard F, Hemery Y, Babot CL. Antioxidant Capacity of Polar and Non-Polar Extracts of Four African Green Leafy Vegetables and Correlation with Polyphenol and Carotenoid Contents. Antioxidants. 2023;12:1726. 10.3390/antiox12091726. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Nishimoto M, Griffin K, Wynne BM, Fujita T. Salt-Sensitive Hypertension and the Kidney. Hypertension (Dallas, Tex.: 1979). 2024. 10.1161/hypertensionaha.123.21369. [DOI] [PubMed]
- 36.Holscher HD. Diving into dietary pattern and dietary diversity analyses. Am J Clin Nutr. 2024;119:1095–6. 10.1016/j.ajcnut.2024.03.014. [DOI] [PubMed] [Google Scholar]
- 37.Lo FP, Qiu J, Wang Z, Chen J, Xiao B, Yuan W, Giannarou S, Frost G, Lo B. Dietary Assessment With Multimodal ChatGPT: A Systematic Analysis. IEEE J Biomed Health Inf. 2024;28:7577–87. 10.1109/jbhi.2024.3417280. [DOI] [PubMed] [Google Scholar]
- 38.Nogay HS, Nogay NH, Adeli H. Image-based food groups and portion prediction by using deep learning. J Food Sci. 2025;90:e70116. 10.1111/1750-3841.70116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Pietiläinen O, Vähäsarja L, Etholén A, Teppo E, Boch J, Speyer P, Jousilahti P, Harkko J, Lallukka T. Contribution of health behaviours to cardiovascular disease by socioeconomic position in Finland. Eur J Pub Health. 2025;35. 10.1093/eurpub/ckaf161.1134.
- 40.Xia Y, Wu X, Zhang C, Hong X. Distribution and influencing factors of cardiovascular health among community inhabitants: Based on the scale of life’s Essential 8. Chin Gen Pract J. 2025;2:100071–100071. 10.1016/j.Cgpj.2025.100071. [Google Scholar]
- 41.Laddu D, Arena R, Pronk NP, Woodard C. Unhealthy dietary patterns in the American Nations: A crisis with cultural distinctions. Curr Probl Cardiol. 2024;49:102558. 10.1016/j.cpcardiol.2024.102558. [DOI] [PubMed] [Google Scholar]
- 42.Park S, Kang S. Interaction of Genetics and Dietary Patterns Scored by the High Healthy Eating Index in Hyperhomocysteinaemia Influencing Cardiovascular Disease Risk. Nutr Bull. 2025;50:311–25. 10.1111/nbu.70007. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets analyzed during the current study are available from the corresponding author upon reasonable request (Prof. Xingli Li; lixingli@csu.edu.cn).