Development and Validation of a Coronary Heart Disease Risk Prediction Model in Snorers with Hypertension: A Retrospective Observed Study

Introduction

Snoring is a common disorder with an overall prevalence of 20–40%1–3 and may be a predictor of obstructive sleep apnea (OSA).4,5 Several studies have reported the correlation between OSA signs and coronary heart disease (CHD).6 Current research shows that snoring can increase the intima–media thickness of the carotid arteries,7 and the risk of carotid stenosis,8 cardiovascular disease (CVD)9 and CHD.10 A large Chinese population-based cohort study estimated an 18% increased risk of ischemic heart disease among habitual snorers aged younger than 50 years.11 A meta-analysis of prospective studies found that snoring was associated with CVD (hazard ratio [HR] 1.26; 95% confidence interval [CI] 0.98–1.62) and CHD (HR 1.15; 95% CI 1.05–1.27).12 The mechanism involved might include vibratory stimuli, an inflammatory cascade, OSA and atherosclerosis.

A positive relationship between snoring and hypertension has been shown in several studies,13,14 especially in men,15,16 which may elevate the risk of CVD, with CHD as one of its manifestations. The association may result from intermittent hypoxia-induced sympathetic activation, oxidative stress and endothelial dysfunction.17 Hypertension is a recognized risk factor for CHD and CVD,18,19 and is a serious clinical and public health problem. The most important cornerstone of CHD prevention is to identify high-risk individuals and intervene as early as possible.

Various countries have population-based screening models for CHD or CVD.20–24 Framingham is a classic model for predicting the incidence of CHD; it can effectively predict individuals at high risk and has been widely used and validated in clinical practice.25 However, pre-existing models developed for the general population might not be suitable for snorers with hypertension, since they do not account for specific characteristics, such as suffocation, sleep apnea, hypopnea.

Accordingly, we aimed to develop and internally validate a new risk prediction model for CHD in snorers with hypertension, which included clinical data and polysomnography results. We also aimed to create a nomogram score to identify snorers with hypertension at high risk of CHD.

Materials and Methods Study Design and Participants

We conducted a retrospective cohort study from January 2011 to December 2013 using a database of patients with suspected OSA and hypertension. Inclusion criteria for participants were as follows: no CHD at baseline; self-reported snoring or family members’ complaints of snoring; and availability of specific polysomnography monitoring data (including apnea-hypopnea index [AHI] and minimum oxygen saturation [minimum SaO2]). The follow-up time was from enrollment (January 2011) to the date of the following events: the diagnosis of the endpoint event (CHD) confirmed by medical documentation or the end date of the study (December 2020). The health status or clinical outcomes of registered patients were confirmed by outpatient visits, inpatient medical records, or telephone calls with patients.

Ethical Approval

The research was authorized by the Medical Ethics Committee of the People’s Hospital of Xinjiang Uygur Autonomous Region (No. 2019030662) and was conducted in strict compliance with the ethical standards set forth in the Declaration of Helsinki and its subsequent amendments. Written informed consent was submitted by all patients or their legal relatives participating in this study.26

Outcomes

The outcome was CHD that occurred during follow-up, including hospitalized angina, myocardial infarction, coronary revascularization and coronary death.27

Data Collection

We screened data for the following conventional risk factors for CVD: sex, age, smoking status, diabetes, body mass index (BMI), systolic blood pressure(SBP), diastolic blood pressure(DBP), serum creatinine (Scr), total cholesterol (TC), triglycerides, high-density lipoprotein (HDL), low-density lipoprotein (LDL), high-sensitivity C-reactive protein (hsCRP), and fasting plasma glucose (FPG). We also incorporated other widely available variables, such as neck circumference (NC), waist circumference (WC), AHI, minimum SaO2 and kinds of antihypertensive drugs at discharge.

Smoking status was divided into non-smokers (including individuals who had not smoked for at least 1 year) and smokers. BMI was calculated by dividing body weight (kg) by the square of height (m). Glomerular filtration rate (GFR; mL/min/1.73 m2) was estimated by the MDRD study equation, . AHI indicates the number of apneas plus the number of hypopneas per hour.

Development and Validation Study

The training and validation sets were created by randomly splitting the total data (70:30). Model development was accomplished in the training set and model validation was realized in the validation set.

Statistical Analyses

Descriptive statistics were performed on all variables, with continuous variables expressed as median (interquartile range), and categorical variables expressed as relative frequencies and percentages. Non-parametric tests were computed for continuous variables and chi-square tests for categorical variables. Univariate and multivariate Cox regression analysis was used to identify the HRs of the candidate factors associated with the incidence of CHD.

We performed multiple imputation to replace missing values using information from the other candidate factors and created five datasets. We converted negative values to the smallest extreme value.

LASSO (least absolute shrinkage and selector operator) regression was used to extract for potential predictors of CHD in the training set. LASSO regression involves variable selection and regularization to improve the predictive accuracy and interpretability of the statistical model it generates. It can also reduce model complexity and avoid overfitting of the prediction model.28 In the training set, we used multivariate Cox regression to develop a new prognostic model of CHD for snorers with hypertension, displayed using a nomogram score.

The discrimination of the model was evaluated by the area under the receiver operating curve (AUC) with 95% CI at 5, 7 and 9 years in the training and validation sets. A value of AUC less than 0.5 indicates no prediction capability, 0.6–0.75 suggests that the prediction model can help distinguish CHD from non-CHD and an AUC ≥0.75 suggests a good degree of discrimination.29 The nomogram score was evaluated by resampling it 1000 times using the regression coefficients of the selected independent variables. The calibration of the model was assessed by the consistency at 5, 7 and 9 years between the observed CHD-free and nomogram score predicted CHD-free probabilities. A well-calibrated model shows predictions lying on or around the 45-degree line of the calibration plot.

We used decision curve analysis to compare the CHD predictive value of the nomogram score, China-PAR and Framingham models in snorers with hypertension at median follow-up time. We also calculated the AUC, sensitivity, specificity, integrated discrimination improvement index (IDI) and net reclassification improvement index (NRI). IDI and NRI are measures of the predictive capability of the new model versus the old model.30

Risk Groups

The nomogram score was used to obtain individual total risk scores. An optimal cutoff point for the risk score in the training set, which was derived using the Youden index, was computed to stratify patients into low-risk or high-risk groups. Tests were two-sided and p<0.05 was considered statistically significant. Statistical analyses were performed using IBM SPSS version 22.0 (IBM Corp., Armonk, NY, USA) or R version 4.0.5 (The R Project for Statistical Computing, Vienna, Austria). This article was prepared in accordance with the TRIPOD reporting checklist and the methods were assessed with PROBAST.31,32

Results Baseline Characteristics of Participants

A total of 2810 snorers with hypertension aged ≥18 years without CHD were finally included in our study. We randomly assigned 1967 individuals (70%) to the training set and the remaining 843 (30%) to the validation set. Information was missing in the research cohort about NC (8.3%), WC (0.2%), Scr (1.1%), TC (2.3%), triglycerides (2.3%), HDL (2.3%), LDL (2.3%), hsCRP (3.2%) and FPG (3.4%). We used multiple imputation to replace missing values. There was no significant difference in baseline characteristics between the training and validation sets (Table 1). The median follow-up time of the two groups was 7 years (6.82 ± 0.04 years). During the follow-up period, 227 patients had CHD, including 152 in the training set and 75 in the validation set. Table 1 shows the baseline characteristics between the subgroups with and without CHD in the training and validation sets.

Table 1 Characteristics of Snorers with Hypertension in the Training and Validation Set

Cox Regression and LASSO Regression to Extract Predictors

Twenty variables were filtered by the Cox regression and LASSO regression method (Table 2). In univariate analysis, sex, age, duration of hypertension, smoking, BMI, NC, WC, SBP, diabetes, TC, HDL, LDL, FPG, AHI, minimum SaO2 and kinds of antihypertensive drugs at discharge were associated with the occurrence of CHD. However, after multivariate analysis, only age, smoking, WC, diabetes, LDL and minimum SaO2 remained independently associated with the occurrence of CHD. LASSO regression reduced 20 variables to five potential predictors. Next, a coefficient distribution diagram was generated, as shown in Figure 1A. The cross-validation error diagram is shown in Figure 1B, which indicates the path of the coefficients included in the model with varying logarithmic transformation λ values. The model included five independent predictors: age, smoking, WC, diabetes and minimum SaO2 (Supplementary Table 1). Among them, age was the strongest predictor of CHD (p<0.0001), followed by smoking, WC, and diabetes, while minimum SaO2 was the least predictive.

Table 2 Cox Regression and Lasso Regression of Candidate Factors of CHD in Patients in the Training Set (n = 1967)

Figure 1 Lasso regression was used to extract predictors. (A) Lasso coefficients of 20 factors. (B) Optimal penalty coefficient in lasso model(λ) is achieved by 10 times cross validation and minimum criterion. The vertical line on the left represents the minimum error, and the vertical line on the right represents the cross validation error within 1 standard error of the minimum error.

Development of a CHD Nomogram Score

A nomogram score was generated by assigning a weighted score to each of the final five independent predictors of CHD (age, smoking, WC, diabetes and minimum SaO2). This presentation format is a graphical presentation of the original mathematical regression formula. For the variable selected in the nomogram score, the value of the variable corresponds to the fraction on the integral line at the top of the nomogram score through the projection of the vertical line. The total points can be obtained by adding the corresponding scores of each variable. From the total point at the bottom of the nomogram score, the projection of the vertical line corresponds to the probability of an individual being CHD-free at 5, 7 and 9 years. The distribution of risk scores is shown in Figure 2.

Figure 2 Nomogram score. For the variable selected in the nomogram score, the value of the variable corresponds to the fraction on the integral line at the top of the nomogram score through the projection of the vertical line. The total points can be obtained by adding the corresponding scores of each variable. From the total point at the bottom of the nomogram score, the projection of the vertical line corresponds to the probability of an individual being CHD-free at 5, 7 and 9 years.

Performance of the CHD Nomogram Score

The AUCs of the nomogram score to predict CHD at 5, 7 and 9 years were 0.706 (95% CI 0.649–0.763), 0.703 (95% CI 0.655–0.751) and 0.669 (95% CI 0.593–0.744), respectively, in the training set (Figure 3A). These values represent good performance and were higher than those in the internal validation set at 5, 7 and 9 years, which were 0.682 (95% CI 0.607–0.758), 0.689 (95% CI 0.618–0.760) and 0.664 (95% CI 0.539–0.789), respectively (Figure 3B). With the passage of time, the diagnostic efficiency of the nomogram score decreased gradually. Furthermore, the nomogram score was well calibrated for predicted and actual values related to participants being CHD-free at 5, 7, and 9 years in the training set; however, deviation was observed in the validation set (Figure 3C and D).

Figure 3 Nomogram score accuracy evaluation. AUC in the training set (A) and validation set (B). Calibration in the training set (C) and validation set (D). The CHD-free risk predicted by the nomogram score is plotted on the x-axis and the observed CHD-free risk is plotted on the y-axis.

Comparison of the Nomogram Score, China-PAR and Framingham Models

In decision curve analysis with a 7-year survival estimation, the nomogram score had a higher overall net benefit compared with the China-PAR and Framingham models (Figure 4). The discriminatory capability of the nomogram score to predict CHD risk (0.699, 95% CI 0.658–0.739) was higher than that of the China-PAR (0.669, 95% CI 0.624–0.713) and Framingham (0.646, 95% CI 0.603–0.689) models. The sensitivity and specificity of the nomogram score to predict CHD were 70% and 55%, respectively. We compared the diagnostic accuracy and differential capability of the nomogram score, China-PAR and Framingham models to predict the risk of CHD by calculating the IDI and NRI. The findings indicated that the nomogram had better predictive and discriminatory capability than the China-PAR (1.9%) or Framingham models (2.0%). The results are shown in Supplementary Table 2.

Figure 4 AUC (A) and Decision curve analysis (B) of Nomogram, China-PAR and Framingham models based on 7-year survival estimation.

Application of the Nomogram Score

Patients were stratified into two risk groups (high risk vs low risk) using the calculated optimal cutoff for the risk scores in the training set (≤121 points for being CHD-free in the nomogram). For the application of the nomogram scores, input the corresponding values of the five predictor variables first, and then the points of every predictors can be obtained in the “update the domain“. The total points in the table and CHD probability at 5-, 7-and 9-year “update the domain” in the same way to obtain the total points for risk stratification and the probability of an individual’s risk of CHD at 5, 7 and 9 years. As an example, a man who snores and has hypertension who is 30 years of age (20 points), has diabetes (24 points), smokes (27 points), has a WC of 120 cm (67 points) and a minimum SaO2 of 86 (7 points) has a high risk of CHD (total score 145; more than the ≤121 cutoff). This individual’s risk probability of CHD at 5, 7 and 9 years is 9%, 13% and 16%, respectively (Table 3).

Table 3 Nomogram Score for Predicting CHD in Snorers with Hypertension

Discussion

CHD is the most common type of CVD and is a major cause of death worldwide.33,34 In this study, 227 cases of CHD were detected among 2810 snorers with hypertension (8.1% prevalence), which is higher than the prevalence of CHD in the general population in China (6.3%).35 We constructed a new nomogram score to evaluate the risk probability of CHD for snorers with hypertension at 5, 7 and 9 years.

Our analysis used regression models with LASSO screened predictors. Unlike other statistical modelling methods, LASSO uses shrinkage property, which results in more stable variable selection.36,37 The new nomogram score, which includes age, smoking, WC, diabetes and minimum SaO2 as predictive factors, helps to determine an individual’s risk of CHD at 5, 7 and 9 years. This score could help clinicians to identify high-risk patients early enough to intervene and prevent CHD.

Our nomogram score contains classic CVD risk factors such as age, smoking and diabetes. However, Framingham, QRISK and Reynold risk scores do not include WC, whereas our risk score does. Increasing evidence indicates that WC is a CVD risk marker that is independent of BMI.38,39 In fact, WC is a highly significant predictor in our nomogram score. Our previous studies have demonstrated that WC is an important risk factor for CHD in snorers with uncontrolled hypertension.27 We think that this is because WC is a better indicator of abdominal obesity in snorers than BMI. Epidemiological data show that snorers have increased abdominal obesity1 and WC is positively associated with CHD.40 WC is a simple and effective anthropometric index, which has low acquisition cost and is easy to implement in clinical practice, and this makes the nomogram score easy to obtain.

New markers for predicting the risk of CHD are emerging. We included polysomnography results, including AHI and minimum SaO2, as candidate factors in our risk prediction model. Minimum SaO2 was one of five predictors in the nomogram score. A decrease in minimum SaO2 increases the risk of CHD. Repeated cycles of hypoxemia and reoxygenation elicited by OSA result in oxidative stress and systemic inflammation, which contribute to coronary atherosclerosis and acute myocardial infarction events.41 Although minimum SaO2 is weak in predicting CHD in the nomogram, our findings suggest that common but non-classical CVD risk factors should be evaluated in future studies.

To enable the nomogram score to be used in clinical settings, we integrated our findings into a risk score that can automatically predict patient outcomes. Patients were assigned scores based on their age, smoking, WC, diabetes and minimum SaO2, and a total score was obtained. The probability of CHD in different years was calculated according to the underlying mathematical formula of the nomogram score. Our nomogram score can also be stratified using high-risk and low-risk groupings by total score. The patient described in a clinical example of how to calculate the nomogram total score (Table 3) developed CHD during the third year of follow-up. The China-PAR model was developed to predict the 10-year risk of atherosclerotic CVD in the Chinese population.42 However, patients assessed as at intermediate risk of atherosclerotic CVD with the China-PAR model are assessed as low risk with the Framingham model. This inconsistency suggests that the based population model may not be applicable to specific patient groups. Alternatively, it may be because age <40 years is a protective factor in Framingham. The performance of our nomogram score was slightly better than that of the China-PAR and Framingham models in snorers with hypertension, and could help in the early identification and management of these patients. In particular, hypoxia burden should be identified in snorers, as it seems to be clinically important.

Our study has several important limitations. First, this was a retrospective study completed in snorers with hypertension, so the findings may not be applicable for widespread use. Second, we included minimum SaO2 in our nomogram model. However, sleep monitoring and data analysis might be different in other medical institutions. Lastly, the new nomogram score requires external or prospective validation. In the future, we need to further verify the performance of the nomogram model.

Conclusion

In conclusion, our nomogram score is simple to use and can predict the risk probability of CHD in snorers with hypertension. The nomogram score can be used to identify high-risk patients, who may benefit from early intervention.

Data Sharing Statement

The datasets used and/or analyzed in the current study are available from the corresponding author upon reasonable request.

Ethics Approval and Consent to Participate

This study was conducted according to the Declaration of Helsinki guidelines and was approved by the Ethics Committee of the People’s Hospital of Xinjiang Uygur Autonomous Region. All participants agreed to participate in this study and provided informed written consent.

Acknowledgments

The authors are very grateful to Department of Hypertensive sleep monitoring room, Xinjiang Uygur Autonomous Region People’s Hospital for their contribution to this study.

Funding

This study is supported by a grant from Special scientific research project of Health young medical science and technology talents of Autonomous Region (WJWY-202201).

Disclosure

The authors declare no conflicts of interest in this work.

References

1. Wei YX, Shen ZW, Yu CQ, et al. [Epidemiological characteristics and correlated factors of habitual snoring among Chinese aged 30 to 79 year-old]. Zhonghua Liu Xing Bing Xue Za Zhi. 2019;40(8):917–923. Chinese. doi:10.3760/cma.j.issn.0254-6450.2019.08.009

2. Torzsa P, Keszei A, Kalabay L, et al. Socio-demographic characteristics, health behaviour, co-morbidity and accidents in snorers: a population survey. Sleep Breath. 2010;15(4):809–818. doi:10.1007/s11325-010-0442-4

3. Lechner M, Breeze CE, Ohayon MM, et al. Snoring and breathing pauses during sleep: interview survey of a United Kingdom population sample reveals a significant increase in the rates of sleep apnoea and obesity over the last 20 years - data from the UK sleep survey. Sleep Med. 2018;54:250–256. doi:10.1016/j.sleep.2018.08.029

4. Sowho M, Sgambati F, Guzman M, et al. Snoring: a source of noise pollution and sleep apnea predictor. Sleep. 2020;43(6). doi:10.1093/sleep/zsz305

5. Baguet JP, Courand PY, Lequeux B, et al. Snoring but not sleepiness is associated with increased aortic root diameter in hypertensive patients. The SLEEPART study. Int J Cardiol. 2015;202:131–132. doi:10.1016/j.ijcard.2015.03.319

6. Javadi HR, Jalilolghadr S, Yazdi Z, Rezaie Majd Z. Correlation between obstructive sleep apnea syndrome and cardiac disease severity. Cardiovasc Psychiatry Neurol. 2014;2014:631380. doi:10.1155/2014/631380

7. Deeb R, Judge P, Peterson E, et al. Snoring and carotid artery intima-media thickness. Laryngoscope. 2014;124(6):1486–1491. doi:10.1002/lary.24527

8. Deeb R, Smeds MR, Bath J, et al. Snoring and carotid artery disease: a new risk factor emerges. Laryngoscope. 2018;129(1):265–268. doi:10.1002/lary.27314

9. Sands M, Loucks EB, Lu B, et al. Self-reported snoring and risk of cardiovascular disease among postmenopausal women (from the women’s health initiative). Am J Cardiol. 2013;111(4):540–546. doi:10.1016/j.amjcard.2012.10.039

10. Yunus FM, Khan S, Mitra DK, et al. Relationship of sleep pattern and snoring with chronic disease: findings from a nationwide population-based survey. Sleep Health. 2018;4(1):40–48. doi:10.1016/j.sleh.2017.10.003

11. Wei Y, Lv J, Guo Y, et al. Age-specific associations between habitual snoring and cardiovascular diseases in China: a 10-year Cohort Study. Chest. 2021;160(3):1053–1063. doi:10.1016/j.chest.2021.04.070

12. Li D, Liu D, Wang X, et al. Self-reported habitual snoring and risk of cardiovascular disease and all-cause mortality. Atherosclerosis. 2014;235(1):189–195. doi:10.1016/j.atherosclerosis.2014.04.031

13. Kim J, Yi H, Shin KR, et al. Snoring as an independent risk factor for hypertension in the nonobese population: the Korean Health and Genome Study. Am J Hypertens. 2007;20(8):819–824. doi:10.1016/j.amjhyper.2007.03.007

14. Khazaie H, Negahban S, Ghadami MR, et al. Among middle-aged adults, snoring predicted hypertension independently of sleep apnoea. J Int Med Res. 2018;46(3):1187–1196. doi:10.1177/0300060517738426

15. Sekizuka H, Miyake H. Relationship between snoring and lifestyle-related diseases among a Japanese occupational population. Intern Med. 2020;59(18):2221–2228. doi:10.2169/internalmedicine.4723-20

16. Lee SK, Choi K, Chang YH, et al. Increased risk for new-onset hypertension in midlife male snorers: the 14-year follow-up study. J Sleep Res. 2018;28(5):e12757. doi:10.1111/jsr.12757

17. Nagayoshi M, Tanigawa T, Yamagishi K, et al. Self-reported snoring frequency and incidence of cardiovascular disease: the Circulatory Risk in Communities Study (CIRCS). J Epidemiol. 2012;22(4):295–301. doi:10.2188/jea.je20110109

18. Bundy JD, Mills KT, He J. Comparison of the 2017 ACC/AHA hypertension guideline with earlier guidelines on estimated reductions in cardiovascular disease. Curr Hypertens Rep. 2019;21(10):76. doi:10.1007/s11906-019-0980-5

19. Han M, Chen Q, Liu L, et al. Stage 1 hypertension by the 2017 American College of Cardiology/American Heart Association hypertension guidelines and risk of cardiovascular disease events: systematic review, meta-analysis, and estimation of population etiologic fraction of prospective cohort studies. J Hypertens. 2020;38(4):573–578. doi:10.1097/HJH.0000000000002321

20. D’Agostino RB, Vasan RS, Pencina MJ, 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

21. Conroy RM, Pyörälä K, Fitzgerald AP, et al. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J. 2003;24(11):987–1003. doi:10.1016/s0195-668x(03)00114-3

22. Hippisley-Cox J, Coupland C, Vinogradova Y, et al. 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

23. Ridker PM, Buring JE, Rifai N, et al. Development and validation of improved algorithms for the assessment of global cardiovascular risk in women: the Reynolds risk score. JAMA. 2007;297(6):611–619. doi:10.1001/jama.297.6.611

24. Woodward M, Brindle P, Tunstall-Pedoe H. Adding social deprivation and family history to cardiovascular risk assessment: the ASSIGN score from the Scottish Heart Health Extended Cohort (SHHEC). Heart. 2006;93(2):172–176. doi:10.1136/hrt.2006.108167

25. Liu J, Hong Y, D’Agostino RB, et al. Predictive value for the Chinese population of the Framingham CHD risk assessment tool compared with the Chinese Multi-Provincial Cohort Study. JAMA. 2004;291(21):2591–2599. doi:10.1001/jama.291.21.2591

26. Cai X, Li N, Hu J, et al. Nonlinear relationship between Chinese visceral adiposity index and new-onset myocardial infarction in patients with hypertension and obstructive sleep apnoea: insights from a Cohort Study. J Inflamm Res. 2022;15:687–700. doi:10.2147/JIR.S351238

27. Wang M, Heizhati M, Li N, et al. Development and validation of a prognostic model to predict high-risk patients for coronary heart disease in snorers with uncontrolled hypertension. Front Cardiovasc Med. 2022;9. doi:10.3389/fcvm.2022.777946

28. Xiang ZJ, Wang Y, Ramadge PJ. Screening tests for lasso problems. IEEE Trans Pattern Anal Mach Intell. 2016;39(5):1008–1027. doi:10.1109/TPAMI.2016.2568185

29. Alba AC, Agoritsas T, Walsh M, et al. Discrimination and calibration of clinical prediction models: users’ guides to the medical literature. JAMA. 2017;318(14):1377–1384. doi:10.1001/jama.2017.12126

30. Pepe MS, Feng Z, Gu JW. Comments on ‘evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond’ by M. J. Pencina et al, statistics in medicine. Stat Med. 2008;27(2):173–181. doi:10.1002/sim.2991

31. Collins GS, Reitsma JB, Altman DG, et al. Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD): the TRIPOD statement. Eur Urol. 2015;67(6):1142–1151. doi:10.1016/j.eururo.2014.11.025

32. Wolff RF, Moons KGM, Riley RD, et al. PROBAST: a tool to assess the risk of bias and applicability of prediction model studies. Ann Intern Med. 2019;170(1):51–58. doi:10.7326/M18-1376

33. Flora GD, Nayak MK. A brief review of cardiovascular diseases, associated risk factors and current treatment regimes. Curr Pharm Des. 2019;25(38):4063–4084. doi:10.2174/1381612825666190925163827

34. Dalen JE, Alpert JS, Goldberg RJ, et al. The epidemic of the 20(th) century: coronary heart disease. Am J Med. 2014;127(9):807–812. doi:10.1016/j.amjmed.2014.04.015

35. Zhu KF, Wang YM, Zhu JZ, et al. National prevalence of coronary heart disease and its relationship with human development index: a systematic review. Eur J Prev Cardiol. 2015;23(5):530–543. doi:10.1177/2047487315587

36. Mullah MAS, Hanley JA, Benedetti A. LASSO type penalized spline regression for binary data. BMC Med Res Methodol. 2021;21(1):83. doi:10.1186/s12874-021-01234-9

37. Zhao X, Liu C, Zhou P, et al. Estimation of major adverse cardiovascular events in patients with myocardial infarction undergoing primary percutaneous coronary intervention: a risk prediction score model from a derivation and validation study. Front Cardiovasc Med. 2020;7:603621. doi:10.3389/fcvm.2020.603621

38. Powell-Wiley TM, Poirier P, Burke LE, et al. Obesity and cardiovascular disease: a scientific statement from the American Heart Association. Circulation. 2021;143(21):e984–e1010. doi:10.1161/CIR.0000000000000973

39. Chang KT, Chen CH, Chuang HH, et al. Which obesity index is the best predictor for high cardiovascular disease risk in middle-aged and elderly population? Arch Gerontol Geriatr. 2018;78:165–170. doi:10.1016/j.archger.2018.06.002

40. Chen Q, Li L, Yi J, et al. Waist circumference increases risk of coronary heart disease: evidence from a Mendelian randomization study. Mol Genet Genomic Med. 2020;8(4):e1186. doi:10.1002/mgg3.1186

41. Yeghiazarians Y, Jneid H, Tietjens JR, et al. Obstructive sleep apnea and cardiovascular disease: a scientific statement from the American Heart Association. Circulation. 2021;144(3):e56–e67. doi:10.1161/CIR.0000000000000988

42. Yang X, Li J, Hu D, et al. 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

留言 (0)

沒有登入
gif