Abstract
Background and Purpose: To evaluate the predictive power of the China-PAR model for cardiovascular disease (CVD).
Methods: Dominate databases, including PubMed, Web of Science, CNKI, Wanfang Data Knowledge Service Platform, Chinese Biomedical Literature Service System, and VIP self-built database, were searched from January 1, 2016 to February 22, 2022. The primary outcome included observed events and predicted events by China-PAR. Meta-analysis was performed using RevMan 5.3 software. Stroke, arteriosclerotic cardiovascular disease (ASCVD), male, and female were divided into subgroup analyses. Funnel plots were used to assess publication bias.
Results: A total of nine studies, which included 221,918 participants, were analyzed. Meta-analysis showed the combined observed incidence of CVD was 3.97%, and the combined predicted incidence was 9.59% by China-PAR. There was no significant difference between the observed and the predicted events. Subgroup analysis showed there was no statistical significance between the observed and the predicted events for stroke or for ASCVD. The difference between the observed and the predicted events by China-PAR was not statistically significant in either males or females.
Conclusions: China-PAR model has important public health significance to further improve the primary prevention strategy of CVD.
Cardiovascular disease (CVD) predominantly presents as angina pectoris, coronary heart disease, myocardial infarction, arrhythmia, heart failure, etc.1 The increasing burden of CVD has become a major public health problem. A report from the American Heart Association pointed out that in 2017 nearly 18 million deaths were attributable to CVD globally, amounting to an increase of over 21% compared with the previous decade.2 With the development of the social economy, the aging population, and the acceleration of urbanization in China, consequently the prevalence of CVD is increasing sharply. According to statistics, CVD is the leading cause of death in both urban and rural areas, accounting for 44.6% in rural areas and 42.1% in urban areas.3
Fortunately, CVD is a preventable and controllable disease, and early intervention can effectively control its progression. Definitely, it is important to improve primary prevention and health management of CVD and carry out risk assessment and stratification.4 Currently, CVD assessment tools mainly include the Framingham universal CVD equation5 in the United States, the Systematic Coronary Artery Risk Assessment (SCORE) model in Europe,6 and the QRISK in the United Kingdom,7 as well as the latest atherosclerotic CVD (ASCVD) Pooled Cohort Equation (PCE) reported in the American College of Cardiology (ACC) and American Heart Association (AHA) guidelines.8
The Chinese Atherosclerotic Cardiovascular Disease Risk Prediction (China-PAR) equation was derived and validated by the China-PAR Project, a large population-based prospective cohort study in China.9 In 2016, Xing et al.10 used a large sample of cohort data in prediction for ASCVD risk in China to develop the China-PAR model applicable to risk assessment of CVD and proposed a risk stratification standard suitable for Chinese. The equation has excellent performance in CVD risk prediction, and its accuracy has been confirmed by other Chinese cohorts,11,12 and the ACC/AHA Guidelines and the Chinese Guidelines have adopted it as a practical tool to promote CVD risk assessment and risk factor management.13,14 In the Chinese guidelines, the China-PAR model has been recommended for the general population and for patients with type 2 diabetes.14
Numerous researchers used China-PAR to conduct risk assessment on the study population. However, the predictive power of the model for CVD risk was inconsistent among different studies. Here, we pooled previous studies to perform this systemic review and to assess its predictability of China-PAR against CVD.
Methods
Search Strategy
One reviewer (CQF) searched PubMed, Web of Science, CNKI, Wanfang Data Knowledge Service Platform, Chinese Biomedical Literature Service System, and VIP self-built database from January 1, 2016 to February 22, 2022 independently. Search terms were listed as follows: #1 TS= (“Cardiovascular Disease” OR “CVD” “Atherosclerotic Cardiovascular Disease” OR “ASCVD” OR “stroke”); #2 TS= (China-PAR); #3 DOP= (2016-01-01/2022-2-22); #4 #1 AND #2 AND #3AND.
Inclusion and Exclusion Criteria
Inclusion criteria were: (1) cardiovascular and cerebrovascular diseases; (2) included China-PAR; and (3) cohort study. Exclusion criteria were: (1) no described outcomes; (2) no control groups; (3) literature with incomplete or no experimental data, duplicate published literature, reviews, and abstracts; (4) impossible to find original paper; and (5) the predicted and observed events cannot be extracted.
Data Extraction
For full text screening, the papers were divided in three different subsets for independent screening by one of the three researchers (CQF, LH, or PXY). They independently extracted the demographic data and outcome information. Baseline information extracted from nine studies contained the first author’s name, year of publication, design type, study subjects (sample size, age, male/female), scale used to assess dysfunction, and the average follow-up time. Besides, the primary outcomes, also included were observed events and predicted events by China-PAR.
Quality Assessment
The quality of studies was assessed using the Newcastle Ottawa scale (NOS), which generated a maximum of nine stars for each study, such as four stars for the selection of participants, two stars for the comparability of participants, and three stars for the assessment of outcomes. Quality was assigned according to the final scores, with 0–3 stars for low quality, 4–6 stars for middle quality, and 7–9 stars indicating high quality.15
Statistical Analysis
The meta-analysis was performed with the statistical software review manager (version 5.4, UK). The OR was defined with 95 % confidence interval (CI) as the effect size. Heterogeneity was assessed by Cochrane’s Q statistics (chi-square) or inverse variance (I2). If I2 < 50 %, and the P > 0.1, these studies could be considered homogeneous as assessed by a fixed-effects model; or else I2 ≥ 50 %, P < 0.10, the random effect model was used for meta-analysis. A P-value < 0.05 was considered statistically significant.
Results
Flowchart and Study Quality
A total of 351 papers with the China-PAR model (including documents, reviews, case reports, and repeated studies) were retrieved from each database. After 137 duplicate records were removed, the full text of each of the remaining 214 studies was read. Among those studies, two were excluded because the articles were reviews. There were 200 studies that did not have related titles and abstracts. The full text of the remaining 12 studies was read, and 3 were removed due to having incomplete data. According to the data extraction requirements, the remaining nine papers10,16-23 were used to extract the corresponding data. The literature screening process is shown in Figure 1. The basic characteristics of each included study is shown in Table 1.
PRISMA reference screen process
Basic characteristics of enrolled studies
Characteristic Data
Data related to the China-PAR model were extracted from nine studies, including sex, city, smoking status, anti-hypertensive treatment status, diabetes status, and family history of CVD (Table 2). The quantitative data of age, waist circumference, systolic blood pressure, diastolic blood pressure, total cholesterol, and high-density lipoprotein status are shown in Table 3. The predicted events were selected as the evaluation index, which was calculated based on the China-PAR model, and the actual occurrence events were extracted from each paper, as shown in Table 4.
Classification data extracted from nine studies
Quantitative data extracted from nine studies
The amount predicted and observed extracted from 9 studies
Quality of All Studies
For prospective studies, the NOS scores varied from 6 to 9 stars, which indicated that the articles included were of medium or high quality (Table 5).
Methodological quality assessments of included cohort studies by Newcastle Ottawa scale (NOS)
The Risk Prediction Ability of China-PAR for CVD
A total of nine studies reported the prediction of the number of CVD based on the China-PAR model. The combined observed incidence of CVD was 3.97% (8806/221918), and the combined predicted incidence was 9.59% (21287/221918) by China-PAR. There was no significant difference between the observed events and the predicted events by China-PAR [RR = 0.81, 95% CI (0.44, 1.48), P=0.49). As there was no statistical heterogeneity among the studies (I2 = 99%, P < 0.01), the random-effect model was used (Figure 2).
Forest plot showing the difference between observed events and predicted events by China PAR about cardiovascular disease
The Risk Prediction Ability of China-PAR for Stroke or Arteriosclerotic CVD
There was statistical heterogeneity among studies16,17,10 that for stroke (I2 = 93%, P < 0.01) and studies20,21,22 with ASCVD (I2 = 100%, P < 0.01), so the random-effect model was used. The observed incidence of stroke was 3.77% (1094/29006), and the predicted incidence was 3.67% (1062/29006) by China-PAR, which showed there was no statistical significance between the observed events and the predicted events by China-PAR [RR = 1.05, 95% CI (0.75, 1.47), P = 0.79]. The observed incidence of ASCVD was 5.41% (4688/86645), and the predicted incidence was 17.4% (15099/86645) by China-PAR. There was no statistical significance between the observed events and the predicted events by China-PAR [RR = 0.65, 95% CI (0.14, 3.06), P = 0.59] (Figure 3).
Forest plot showing the difference between observed events and predicted events by China PAR about stroke or arteriosclerotic cardiovascular disease
The Risk Prediction Ability of China-PAR for Gender
As not all included studies were able to extract the respective actual incidence of CVD for male and female subjects, meta-analysis was performed on eight studies according to sex. Among them, the incidence of CVD showed statistical heterogeneity among the studies (I2=98%, P < 0.01). Therefore, the random effect model was used, and the results showed the observed incidences of CVD were 4.48% (4801/107130) for male and 3.43% (3899/113585) for female. At the same time, the predicted incidence by China-PAR was 15.2% (16287/107130) and 4.58% (4906/107130), respectively. There was no statistical significance between the observed events and the predicted events by China-PAR both in male [RR=0.80, 95% CI (0.37, 1.76), P=0.59] and female (RR=0.91, 95%CI (0.68, 1.22), P=0.52), as shown in Figure 4.
Forest plot showing the difference between observed events and predicted events by China PAR about cardiovascular disease in male and female
Publication Bias
Funnel plot analysis showed the distribution of funnel plot in studies predicting the incidence of CVD based on the China-PAR model was asymmetric, as shown in Figure 5.
Funnel plot of cardiovascular disease incidence literature
Discussion
CVD is the leading cause of death worldwide and a major public health concern. Fortunately, CVD is a preventable and controllable disease, and early intervention can effectively control its progression. The use of cardiovascular risk assessment tools to comprehensively assess an individual’s risk of future development of CVD is an important part of cardiovascular primary prevention.24 Around the world, currently there were a number of CVD prediction models, such as the Framingham risk model,25 the QRISK cardiovascular risk assessment model, a cardiovascular disease risk score for the United Kingdom,26 and the PCE model recommended by ACC/AHA.27 For many years, there were no assessment tools specific to the Chinese population, so most of the large-scale CVD risk surveys in Chinese populations were conducted using foreign tools, which may overestimate or underestimate the overall CVD risk.28
China-PAR is the first risk prediction model based on ASCVD in China, which provides an important assessment tool for CVD risk prediction and cardiovascular primary prevention in Chinese population. It is based on previous models; the interaction between age and risk factors was analyzed according to the actual situation and epidemiological characteristics, and risk factors such as waist circumference, region, urbanization, and family history were added to improve the prediction efficiency of the model.9 In the 6 years since the model was built, more and more researchers have applied China-PAR to conducted risk assessment on CVD, but the predictive power of the model for CVD risk was inconsistent among different studies.
Compared with other CVD risk prediction assessment tools, FAST (Face, Arm, Speech Test) and BEFAST (Balance, Eyes, Face, Arm, Speech Test) were only suitable for predicting stroke. The diagnostic value of BEFAST in acute ischemic stroke was higher than in FAST.29 The Framingham Wilson, ATP (Adult Treatment Panel) III and PCE all have an overestimate of the risk of CVD,30 but the Framingham Wilson scale could be considered the gold standard when used together with QRISK Score.31
To the best of our knowledge, this is the first systematic review and meta-analysis to compare and evaluate the predictive power of the China-PAR model. In this study, a total of nine studies, including 221,918 participants were analyzed. There was no statistical difference between the combined observed incidence and the combined predicted incidence by China-PAR of CVD. In addition, the subgroup analysis by disease showed there was no statistical difference between the predicted incidence by China-PAR and actual incidence both in stroke and ASCVD. According to the assessment of predictability based on sex, it was found that the difference between observed incidence and the predicted incidence by China-PAR was not statistically significant, suggesting that China-PAR had good predictive ability.
The results of our study showed there was no statistical difference between the China-PAR prediction group and the actual observation group, and the subgroup analysis showed that China-PAR had a good predictive power regardless of disease or gender. Hence, the China-PAR model was useful in the prediction of CVD and had good predictive ability. It is of great importance in public health significance to further improve the primary prevention strategy of CVD.
Study Limitation
The overall number of studies included in this review was small, and there was asymmetry in the funnel diagram. China-PAR is a CVD risk assessment model tailored for Chinese, and it has only been established for 6 years, so the amount of literature is limited. In addition, the follow-up time of the included literature varied greatly. Significant heterogeneity existing among the included studies could reduce the statistical efficiency. All these shortcomings should be further investigated and addressed in future studies.
Acknowledgments
We thank postgraduate candidates Wang Meng and Zhang Yan, who performed numerous clinical examinations on the prediction of stroke in community hospitals and provided valuable suggestions for this study.
Footnotes
Disclosures: This study was supported by the Chengdu Science and Technology Bureau focus on research and development support plan (2019-YF09-00097-SN). The data presented in this study are available on request from the corresponding author. The authors declare they have no conflicts of interest.
Author Contributions: QC, HL, XP contributed equally to this work. YW, PZ, LH, JW, MH, FX QC designed and wrote the manuscript for FX. QC, HL, XP screened the references and extracted the data from the literature. JW and LH double checked the raw data. YW and PZ accessed the basic information from the study. QC and FX analyzed and interpreted the data. QC evaluated the quality of studies. FX and MH proofread the manuscript.
- Received April 24, 2023.
- Revision received January 18, 2024.
- Accepted January 19, 2024.
References
- 1.↵Yang L, Wu H, Jin X, Study of cardiovascular disease prediction model based on random forest in eastern China. Sci Rep. 2020;10(1):5245. doi:10.1038/s41598-020-62133-5.
- 2.↵Virani SS, Alonso A, Benjamin EJ, ; American Heart Association Council on Epidemiology and Prevention Statistics Committee and Stroke Statistics Subcommittee. Heart Disease and Stroke Statistics—2020 Update: A Report From the American Heart Association. Circulation. 2020;141(9):e139-e596. doi:10.1161/CIR.0000000000000757.
- 3.↵National Center for Cardiovascular Quality Improvement. 2021 Medical Quality Report of Cardiovascular Diseases in China: an Executive Summary. Chinese Circulation Journal. 2021;36(11):1041-1064.
- 4.↵Joint Task Force for Guideline on the Assessment and Management of Cardiovascular Risk in China. Zhonghua Yu Fang Yi Xue Za Zhi. 2019;53(1):13-35. doi:10.3760/cma.j.issn.0253-9624.2019.01.004
- 5.↵Kannel WB, McGee D, Gordon T. A general cardiovascular risk profile: The Framingham study. Am J Cardiol. 1976;38(1):46-51. doi:10.1016/0002-9149(76)90061-8.
- 6.↵D’Agostino RB Sr, Vasan RS, Pencina MJ, 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
- 7.↵Hippisley-Cox J, Coupland C, Vinogradova Y, Robson J, May M, Brindle P. Derivation and validation of QRISK, a new cardiovascular disease risk score for the United Kingdom: prospective open cohort study. BMJ. 2007;335(7611):136. doi:10.1136/bmj.39261.471806.55.
- 8.↵Goff DC Jr, Lloyd-Jones DM, Bennett G, 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines [published correction appears in Circulation. 2014 Jun 24;129(25 Suppl 2):S74-5]. Circulation. 2014;129(25 Suppl 2):S49-S73. doi:10.1161/01.cir.0000437741.48606.98
- 9.↵Yang X, Li J, Hu D, Predicting the 10-Year Risks of Atherosclerotic Cardiovascular Disease in Chinese Population: The China-PAR Project (Prediction for ASCVD Risk in China). Circulation. 2016;134(19):1430-1440. doi:10.1161/CIRCULATIONAHA.116.022367
- 10.↵Xing X, Yang X, Liu F, Predicting 10-Year and Lifetime Stroke Risk in Chinese Population. Stroke. 2019;50(9):2371-2378. doi:10.1161/STROKEAHA.119.025553.
- 11.↵Tang X, Zhang D, He L, Performance of atherosclerotic cardiovascular risk prediction models in a rural Northern Chinese population: Results from the Fangshan Cohort Study. Am Heart J. 2019;211:34-44. doi:10.1016/j.ahj.2019.01.009.
- 12.↵Zeng NM, Zheng XW, Peng H, Validation of the China-PAR Equations for Cardio-cerebrovascular Risk Prediction in the Inner Mongolian Population. Biomed Environ Sci. 2018;31(6):463-466. doi:10.3967/bes2018.061.
- 13.↵Arnett DK, Blumenthal RS, Albert MA, 2019 ACC/AHA Guideline on the Primary Prevention of Cardiovascular Disease: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2019;140(11):e596-e646. doi:10.1161/CIR.0000000000000678.
- 14.↵Joint Task Force for Guideline on the Assessment and Management of Cardiovascular Risk in China. Zhonghua Yu Fang Yi Xue Za Zhi. 2019;53(1):13-35. doi:10.3760/cma.j.issn.0253-9624.2019.01.004
- 15.↵Nunes KP, Labazi H, Webb RC. New insights into hypertension-associated erectile dysfunction. Curr Opin Nephrol Hypertens. 2012;21(2):163-170. doi:10.1097/MNH.0b013e32835021bd.
- 16.↵Tang X, Zhang DD, Liu XF, Beijing Da Xue Xue Bao Yi Xue Ban. 2020;52(3):444-450. doi:10.19723/j.issn.1671-167X.2020.03.008.
- 17.↵Zhang Y, Fang X, Guan S, Validation of 10-Year Stroke Prediction Scores in a Community-Based Cohort of Chinese Older Adults. Front Neurol. 2020;11:986. doi:10.3389/fneur.2020.00986
- 18.↵Qing-yun Tu. Comparison and application of CVD risk prediction models AMONG patients with hypertension. Master’s Thesis: Southeast University; 2020.
- 19.Ni-Mei Z. Validation of the China-PAR in predicting CVD among Mongolian and Association of Galectin-3 and FPG/HDL-C with the prognosis of ischemic stroke. Master’s Thesis: Soochow University; 2019.
- 20.↵Li HH, Huang S, Liu XZ, Zou DJ. Applying the China-PAR Risk Algorithm to Assess 10-year Atherosclerotic Cardiovascular Disease Risk in Populations Receiving Routine Physical Examinations in Eastern China. Biomed Environ Sci. 2019;32(2):87-95. doi:10.3967/bes2019.014.
- 21.↵Jiang Y, Ma R, Guo H, External validation of three atherosclerotic cardiovascular disease risk equations in rural areas of Xinjiang, China. BMC Public Health. 2020;20(1):1471. doi:10.1186/s12889-020-09579-4.
- 22.↵Tang X, Zhang D, He L, Performance of atherosclerotic cardiovascular risk prediction models in a rural Northern Chinese population: Results from the Fangshan Cohort Study. Am Heart J. 2019;211:34-44. doi:10.1016/j.ahj.2019.01.009.
- 23.↵Li J, Liu F, Yang X, Validating World Health Organization cardiovascular disease risk charts and optimizing risk assessment in China. Lancet Reg Health West Pac. 2021;8:100096. doi:10.1016/j.lanwpc.2021.100096
- 24.↵Lloyd-Jones DM, Braun LT, Ndumele CE, Use of Risk Assessment Tools to Guide Decision-Making in the Primary Prevention of Atherosclerotic Cardiovascular Disease: A Special Report From the American Heart Association and American College of Cardiology [published correction appears in J Am Coll Cardiol. 2019 Jun 25;73(24):3234]. J Am Coll Cardiol. 2019;73(24):3153-3167. doi:10.1016/j.jacc.2018.11.005
- 25.↵Cybulska B, Kłosiewicz-Latoszek L. Landmark studies in coronary heart disease epidemiology. The Framingham Heart Study after 70 years and the Seven Countries Study after 60 years. Kardiol Pol. 2019;77(2):173-180. doi:10.5603/KP.a2019.0017.
- 26.↵Hippisley-Cox J, Coupland C, Vinogradova Y, Robson J, May M, Brindle P. Derivation and validation of QRISK, a new cardiovascular disease risk score for the United Kingdom: prospective open cohort study. BMJ. 2007;335(7611):136. doi:10.1136/bmj.39261.471806.55.
- 27.↵Goff DC Jr, Lloyd-Jones DM, Bennett G, 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines [published correction appears in J Am Coll Cardiol. 2014 Jul 1;63(25 Pt B):3026]. J Am Coll Cardiol. 2014;63(25 Pt B):2935-2959. doi:10.1016/j.jacc.2013.11.005
- 28.↵Zhang M, Jiang Y, Wang LM, Prediction of 10-year Atherosclerotic Cardiovascular Disease Risk among Adults Aged 40-79 Years in China: a Nationally Representative Survey. Biomed Environ Sci. 2017;30(4):244-254. doi:10.3967/bes2017.034.
- 29.Chen X, Zhao X, Xu F, A Systematic Review and Meta-Analysis Comparing FAST and BEFAST in Acute Stroke Patients. Front Neurol. 2022;12:765069. doi:10.3389/fneur.2021.765069
- 30.↵Damen JA, Pajouheshnia R, Heus P, Performance of the Framingham risk models and pooled cohort equations for predicting 10-year risk of cardiovascular disease: a systematic review and meta-analysis. BMC Med. 2019;17(1):109. doi:10.1186/s12916-019-1340-7
- 31.↵Lucaroni F, Cicciarella Modica D, Macino M, Can risk be predicted? An umbrella systematic review of current risk prediction models for cardiovascular diseases, diabetes and hypertension. BMJ Open. 2019;9(12):e030234. doi:10.1136/bmjopen-2019-030234
