Article Text

Download PDFPDF

High-sensitivity C reactive protein and risk of cardiovascular disease in China-CVD study
  1. Ying Dong,
  2. Xin Wang,
  3. Linfeng Zhang,
  4. Zuo Chen,
  5. Congyi Zheng,
  6. Jiali Wang,
  7. Yuting Kang,
  8. Lan Shao,
  9. Ye Tian,
  10. Zengwu Wang
  11. for the China CVD study investigators
    1. Division of Prevention and Community Health, National Center for Cardiovascular Disease, Fuwai Hospital, Pecking Union Medical College & Chinese Academy of Medical Sciences, Beijing, China
    1. Correspondence to Dr. Zengwu Wang, Division of Prevention and Community Health, National Center for Cardiovascular Disease, Fuwai Hospital, Pecking Union Medical College & Chinese Academy of Medical Sciences, Beijing 102308, China; wangzengwu{at}foxmail.com

    Abstract

    Background This study aimed to assess the association of high sensitivity C-reactive protein (hs-CRP) with cardiovascular disease (CVD) in middle-aged Chinese population.

    Methods The baseline was collected 2009–2010, and follow-up was conducted in 2016–2017. Data of hs-CRP were from baseline examination and re-examination in 2016–2017 using transmission turbidimetry with a measurement range of 0–42 000. The primary outcome was CVD including coronary heart disease events and stroke events.

    Results Among 8688 participants free from CVD (at baseline, mean age, 50.1 years, 3897 were males), there were 189 CVD events, occurred during a median follow-up of 6.34 years (54 685 person-years at risk). From the Kaplan-Meier curve, we found that there was a progressive increase in CVD event rates by hs-CRP tertiles (log-rank test, p<0.001). Baseline hs-CRP was linearly associated with CVD (p for trend=0.015) even after adjusting for known CVD risk factors. Furthermore, the net reclassification improvement when hs-CRP was added to a model based on traditional factors was 7.85% for CVD (p=0.003). In addition, the correlation between change of hs-CRP and CVD was conducted in a subgroup (n=4778). However, we did not find the correlation between hs-CRP change and CVD (correlation coefficient: −0.003, p=0.846).

    Conclusions In the middle-aged Chinese population, hs-CRP was associated with increased risk of developing CVD. Although there was no correlation between hs-CRP change and CVD, the level of hs-CRP was higher at follow-up than baseline even among those with CVD. More attention should be given to those with higher level of hs-CRP for CVD prevention.

    • cohort studies
    • cardiovascular disease
    • epidemiology
    View Full Text

    Statistics from Altmetric.com

    Introduction

    Cardiovascular disease (CVD) is a leading cause of mortality worldwide, especially in Asian countries.1 Statistics from WHO have indicated that CVD is the leading cause of death in China,2 3 accounting for nearly 45% of all deaths in 2014.4 As known, CVD has a significant inflammatory component.5 6 Although several inflammatory biomarkers exist, high-sensitivity C-reactive protein (hs-CRP) is considered optimal for assessment of cardiovascular risk according to the recommendation of the American Heart Association (AHA).7 In several recent clinical studies, the hs-CRP has been reported as a predictor of CVD or all-cause mortality.8–13 Although the hs-CRP has been found to predict the risk of CVD, the majority of studies are done in white populations.10 11 To the best of our knowledge, only few studies are done among Chinese population.14 Moreover, the concentrations of hs-CRP have been shown to vary by ethnic status.15 16 The Chinese adults seem to have relatively low hs-CRP levels.14 17 The Study of Women’s Health Across the Nation found that the median of hs-CRP was 0.7 mg/L in Chinese women compared with 1.5 mg/L in white women.15 Therefore, the association between hs-CRP and CVD, and the potential benefit of adding information on hs-CRP need to be fully studied in Chinese population. In addition, the relationship between changes of hs-CRP and CVD in general population is unclear.

    Therefore, the aim of the present study was to assess the relationship between baseline hs-CRP and follow-up on CVD. Moreover, we also had hs-CRP measurements during follow-up, providing the opportunity to assess the correlation between the hs-CRP change and CVD.

    Methods

    Subjects

    The study population consisted of participants who participated in the CVD risk factors study in China (China-CVD study) from 2009 to 2010 (n=11 623). Detailed methods used in the survey have been described previously.18 Briefly, for the present study, 9 out of 12 study fields (n=8965) completed the follow-up measurements until 2016–2017.

    For this study, individuals were excluded if they did not have measurement of hs-CRP at baseline (n=93). Subjects were also excluded for pre-existing CVD or self-reported active inflammatory or infectious conditions19 (n=172), no data on demographic characteristics or laboratory measurements (n=12). This resulted in a final analytical sample with 8688 subjects for analysis of the association between baseline hs-CRP and CVD; and 4778 participants with complete data from baseline examination in 2009–2010 and re-examination in 2016–2017 were eligible for the analysis of the correlation between the change of hs-CRP and CVD.

    Measurements

    Medical history, demographic data, anthropometric data and blood pressure (BP) measurements were collected by trained staff using a standardised protocol at baseline. BP was measured using a standard mercury sphygmomanometer three times on the right arm at the heart level after the participant had been seated for at least 5 min, with 30 s between each measurement. The average of three readings was used for final analysis.

    Blood samples were collected after 10 hours of fasting and analysed in the same core clinical laboratory (Beijing CIC Clinical Laboratory, Beijing, China). Fasting blood glucose (FBG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C) were measured using enzymatic techniques on a HITACHI 7080 autoanalyser (Hitachi, Tokyo, Japan). The data of hs-CRP were from baseline examination and re-examination using transmission turbidimetry (Advia 2400 autoanalyser, Sciences, Munich, Germany) with a measurement range of 0–42 000 mg/L, with an interassay variation coefficient of 2.51%.

    CVD assessment

    Follow-up was completed 2016–2017 by trained research personnel via clinical visit, telephone or delivery of medical records. In the present study, primary outcome was CVD. The CVD included coronary heart disease events (myocardial infarction (I21, I22)), and stroke events (subarachnoid haemorrhage (I60), intracerebral haemorrhage (I61), other non-traumatic intracranial haemorrhage (I62), cerebral infarction (I63) and stroke, not specified as haemorrhage or infarction (I64)). All outcomes were adjudicated via independent committee which was blinded to patients, using standardised definitions. Only the first event was used for final analysis.

    Risk factor definition

    Current smokers were defined as participants who have smoked at least 20 packages of cigarettes or 0.5 kg leaf tobacco in their lifetime and currently smoke cigarette. Alcohol consumption was defined as drinking once per week in the latest month. Diabetes mellitus was defined as FBG ≥7.0 mmol/L or self-reported current treated with antidiabetic medications (insulin or oral hypoglycaemic agents). Statin therapy was defined as participants who self-reported use of a prescription statin medication within 2 weeks at the time of the interview at the baseline.

    Statistical analysis

    Quantitative variables were presented as means and SDs, and qualitative variables were shown as numbers and percentages. One-way analysis of variance and χ2 tests were used to compare the characteristics of study participants at baseline by tertiles of hs-CRP.

    Kaplan-Meier curves were used to display event rates. Moreover, Cox proportional hazards models were used to estimate adjusted HRs (95% CI) for CVD. Commonly, hs-CRP used tertile categories. The referent group was the lowest category. Potential covariates, such as age, gender, smoking status, alcohol consumption, TC, HDL-C, systolic BP, BP treatment, statin therapy and diabetes mellitus, were included in the multivariable models. For testing linear trend, the tertile rank was used as a continuous variable in the regression model. The comparison of change in hs-CRP between participants with or without CVD was evaluated in the subgroup (n=4778, 55.0%) with repeated value of hs-CRP by Student’s t-test, and the correlation of hs-CRP with CVD was assessed by partial correlation analysis adjusting for traditional CVD risk factors. The change of hs-CRP was determined for follow-up hs-CRP minus baseline hs-CRP.

    Finally, the CVD risk was estimated using Cox proportional hazard models which included the traditional risk factors (age, gender, smoking status, alcohol consumption, TC, HDL-C, systolic BP, BP treatment, statin therapy and diabetes mellitus). The discriminatory ability and model fit were compared between the base prediction model (according to traditional risk factors) and the new prediction model (traditional risk factor+hs-CRP). Model discriminatory was assessed using the area under the operating characteristics curve (AUC), and model fit was evaluated using the Akaike Information Criterion (AIC). Furthermore, net reclassification improvement (NRI) was also calculated to test whether addition of hs-CRP concentration can improve risk classification. The NRI was analysed using the Z test.20 21 All analyses were performed using SAS V.9.4 (SAS Institute).

    Results

    Table 1 shows baseline clinical and laboratory characteristics by hs-CRP tertiles (n=8688, mean age: 50.1 years, males: 44.9% and females: 55.1%). Age, BP and TC were increased with increasing hs-CRP level (all p for trend <0.001). Also, participants with elevated hs-CRP levels had a greater proportion of males, current smokers, BP therapy, statin therapy and diabetes mellitus.

    Table 1

    Baseline characteristics for participants (n=8688)

    Median length of follow-up for population was 6.34 years (54 685 person-years at risk). During this time, 189 CVD occurred. Figure 1 illustrates the Kaplan-Meier curve for CVD by hs-CRP tertiles. There was a progressive increase in CVD event rates by hs-CRP tertiles (log-rank test, p<0.001). As seen in table 2, there was a significant association between hs-CRP and CVD (p for trend=0.015), even after adjustment for traditional CVD risk factors.

    Figure 1

    Kaplan-Meier curve for cardiovascular disease. T1–3: tertile 1, hs-CRP ≤0.5 mg/L; tertile 2, hs-CRP ≤1.3 mg/L; tertile 3, hs-CRP >1.3 mg/L. hs-CRP: high-sensitivity C-reactive protein.

    Table 2

    hs-CRP level HRs for CVD (n=8688)

    More than half of participants (n=4778, 55.0%) completed the repeated hs-CRP measurements during the follow-up. As online supplementary table 1 shows, the mean (±SD) of hs-CRP has been significantly higher at follow-up than at baseline, irrespective of participants with or without CVD. Comparing with non-CVD participants, we observed a greater increment of hs-CRP change among participants with CVD, although there were no significant differences. Furthermore, as seen in table 3, there was no significant correlation between hs-CRP change and CVD (p=0.846) after adjustment for traditional CVD risk factors.

    Supplementary file 1

    Table 3

    The correlation coefficients between hs-CRP change and CVD (n=4778)

    Online supplementary table 2 shows the AUC and AIC for CVD outcomes according to different prediction models. The model discrimination due to the addition of hs-CRP were improved by very small increments, with AUC from 0.784 to 0.786 for CVD (p=0.665). Likewise, very small AIC differences were observed in the models with and without hs-CRP (3232.7 vs 3235.0 for CVD). Table 4 displays NRI by adding hs-CRP to the risk factors independently associated with CVD. A total of 24 (9+0+15) participants who developed CVD were reclassified upward, and 6 (4+0+2) participants who developed an event were reclassified downward. The net estimate for the percentage classified upward was the difference between these two estimates divided by total number of events [(24-6)/189=9.52%]. Similarly, the net estimate for those not developing a CVD event was [((111+0+34)-(198+0+89))/8499=−1.67%]. The NRI was estimated by taking the sum of the net estimates for those who developed an event and those who did not develop an event which was 7.85% (9.52%–1.67%; p=0.003).

    Table 4

    NRI by adding hs-CRP to the risk factors independently associated with CVD

    Discussion

    This study demonstrates higher hs-CRP significantly increased the risk of developing CVD in a large middle-aged Chinese population. Moreover, the addition of hs-CRP to the traditional risk factors of CVD may improve the reclassification of risk. However, we did not find that there was a significant correlation between hs-CRP change and CVD.

    Comparing with participants with lower level of hs-CRP, those with higher hs-CRP level had 62% higher risk of CVD after adjusting age, gender, smoking status, alcohol consumption, systolic BP, BP therapy, statin therapy, TC, HDL-C and diabetes mellitus. Our result was similar to results of other previous studies in Preventive Services Task Force (from USA),10 Japan Collaborative Cohort study (from Japan),22 Korean Industrial Safety and Health Law cohort (from Korea)9 or Shanghai Diabetes Studies (from China).14 For the mechanism, hs-CRP was a non-specific biomarker of inflammation predominantly in hepatocytes. The role of hs-CRP in plaque deposition is highly complex; on the one hand, hs-CRP may facilitate monocyte adhesion and transmigration into the vessel wall, and on the other hand, M1 macrophage polarisation, catalysed by hs-CRP, leads to macrophage infiltration of both adipose tissue and atherosclerotic lesions.23

    A few studies assessed the association between hs-CRP change and CVD, such as Korean Industrial Safety and Health Law cohort9 (in Korea), Atherosclerosis Risk in Community (ARIC) Study24 (in USA), Anglo-Scandinavian Cardiac Outcomes Trial (in UK and Ireland)25 and the Heart protection study (in the UK).26 It was reported that comparing with only a single measurement, long-term measurement could correct the ‘regression dilution’ which resulted in stronger associations of hs-CRP with clinical outcomes.27 Thus, two measurements of hs-CRP were recommended by the Centers for Disease Control and Prevention and the AHA, optimally 2 weeks apart to increase stability of measurement values.6 We measured the hs-CRP again during the follow-up, and there was no significant correlation between hs-CRP change and CVD. However, comparing with the hs-CRP level at baseline, the hs-CRP level was significantly higher at follow-up. Moreover, there was a greater increment of hs-CRP among participants with CVD than those without CVD, although there were no significant differences. This indicated that the risk of hs-CRP on CVD has not attracted enough attention by the public; even among the participants with CVD, hs-CRP was not well controlled. Thus, more attention should be paid for CVD prevention in the population level.

    Moreover, in the current study, hs-CRP did not improve discriminatory ability to identify new cases of CVD during follow-up, since the change of AUC was minimal with the addition of hs-CRP.13 However, regarding to reclassification, a proportion of participants at intermediate risk of CVD upward or downward (NRI for CVD 7.85%). This finding is consistent with the European Society of Cardiology guideline which recommends that participants at moderate, but not low, risk of CVD should measure hs-CRP.28

    There are also some limitations of this study. First, because of the use of random cluster sampling might occur bias, but the economic and social development levels were considered in the sampling process. Second, as we only examined the participants aged 35–64 years with a similar ethnic background, the generalisability to other age and ethnic groups is unknown. Third, there were only 55% participants with hs-CRP data from baseline and re-examination, and there was significant different between participants who completed two measurements and those who did not complete two measurements in terms of gender and lifestyle habits (online supplementary table 3) which might have resulted in selection bias. Lastly, because of the relative small number of cardiovascular events, we had limited power to examination the correlation between the hs-CRP change and CVD. However, the China-CVD study is an ongoing longitudinal study, the results can be validated in further follow-up. Despite these limitations, the data from the present study clearly demonstrate that hs-CRP has adverse effect on CVD in middle-aged Chinese population.

    In conclusion, our results demonstrated that hs-CRP was associated with increased risk of developing CVD independent of the established risk factors. Although there was no correlation between hs-CRP change and CVD, the level of hs-CRP was higher at follow-up than baseline even among those with CVD. Thus, more attention should be given to those with higher level of hs-CRP for CVD prevention.

    What is already known on this subject

    • High-sensitivity C-reactive protein (hs-CRP), an inflammatory biomarker, is considered optimal for the assessment of cardiovascular risk according to the recommendation of the American Heart Association.

    • People with sustained elevation in hs-CRP were at highest risk of cardiovascular disease (CVD). However, the majority of the studies were done in white population.

    • The association of hs-CRP and its change with CVD in general Chinese population is unclear.

    What this study adds

    • We found that in the middle-aged Chinese population, hs-CRP is associated with increased risk of developing CVD, independent of the established risk factors.

    • hs-CRP offers moderate improvement in reclassification of risk.

    • Although there was no correlation between hs-CRP change and CVD, the level of hs-CRP was higher at follow-up than at baseline, even among those with CVD.

    Acknowledgments

    The investigators are grateful to all the colleagues involving in the China-CVD study, and all research staff, including Congyi Zheng, Jiali Wang, Yuting Kang, Zhouhai Yao for help in maintaining the data.

    References

    View Abstract

    Footnotes

    • Patient consent for publication Not required.

    • Contributors YD researched data, conducted data analyses, interpreted results and prepared the manuscript. XW, LZ and ZC conceptualised the study, aided in data analysis and critically revised the manuscript. CZ, JW, YK, LS and YT aided in result interpretation and critically revised the manuscript. ZW conceptualised the study, researched and critically revised the manuscript. All authors have read and approved the final version of the manuscript.

    • Funding This study was supported by the National Natural Science Foundation of China (81373070) and Chinese National Specific Fund for Health-scientific Research in Public Interest (200902001) and CAMS Innovation Fund for Medical Sciences (2017-I2M-1-004).

    • Competing interests None declared.

    • Ethics approval Ethics Review Board of Fuwai hospital approved this study.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Collaborators Investigators of China-CVD study: For a partial listing of colleagues see the follows (provinces sorted as alphabetical order): Huiqing Cao, Xiaoxia Wang and Tian Fang, Institute of Molecular Medicine, Pecking University, Beijing, China; Xiaoyan Han and Zhe Li, Chaoyang District Center for Disease Control and Prevention, Beijing, China. Heilongjiang: Ye Tian, Lihang Dong, Fengyu Sun and Fucai Yuan, First Affiliated Hospital of Harbin Medical University, Heilongjiang, China. Jiangsu: Xin Zhou, Yunyang Zhu, Yi He and Qingping Xi, Jintan Institute of Hygiene, Jiangsu, China. Shanxi: Ruihai Yang, Jun Yang, Yong Ren, Maiqi Dan, Yiyue Wang, Daming Yu and Ru Ju, Hanzhong Hospital, Shanxi, China. Shanxi: Dongshuang Guo, Yuxian Hospital, Shanxi, China. Sichuan: Dahua Tan, Zhiguo Zheng, Jingjing Zheng and Yang Xu, Deyang Institute of Hygiene, Sichuan, China. Xinjiang: Dongsheng Wang and Tao Chen, Autonomous Region Yining Center for Disease Control and Prevention, Xinjiang Uygur Autonomous Region, China. Yunnan: Meihui Su and Yongde Zhang, Yunnan Center for Disease Prevention and Control, Yunnan, China. Zhejiang: Zhanhang Sun and Chen Dai, Zhoushan Cardiovascular Institute, Zhejiang, China.

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.