You are here

  • Home
  • Archive
  • Volume 13, Issue 7
  • Bidirectional association between depressive symptoms and cardiovascular disease in the middle-aged and elderly Chinese: a 5-year longitudinal study

Article Text

Article menu
Mental health
Original research
Bidirectional association between depressive symptoms and cardiovascular disease in the middle-aged and elderly Chinese: a 5-year longitudinal study
  1. Zhang Ziwei,
  2. Yumeng Hua,
  3. Aiping Liu
  1. Department of Social Medicine and Health Education, Peking University School of Public Health, Beijing, China
  1. Correspondence to Professor Aiping Liu; apingliu{at}163.com

Abstract

Objective Depressive symptoms and cardiovascular diseases (CVDs) are important issues affecting the health of the middle-aged and elderly population in China. This study aimed to investigate the bidirectional association between depressive symptoms and CVD in middle-aged and elderly people in China.

Design A 5-year longitudinal study.

Setting and participants We included 6702 middle-aged and elderly participants from China Health and Retirement Longitudinal Study (CHARLS), which is a nationwide longitudinal household survey that started in 2011 (T1) and followed up every 2 years in 2013 (T2) and 2015 (T3).

Outcome measures Depressive symptoms were measured by the Center for Epidemiological Studies Depression Scale. Binary logistic regression was used to identify the influencing factors of depressive symptoms and CVD at T1. The cross-lagged panel model was used to analyse the association between depressive symptoms and CVD at T1, T2 and T3.

Results The CHARLS is a representative longitudinal survey of people aged ≥45 years. Using data extracted from the CHARLS, overall, at T1, 2621 (39.10%) participants had depressive symptoms and 432 (6.4%) had CVD, and at T3, 2423 (36.2%) had depressive symptoms and 760 (11.3%) had CVD, respectively. Depressive symptoms at T1 had a effect on CVD at T2 (β=0.015, p=0.009), and depressive symptoms at T2 had an effect on CVD at T3 (β=0.015, p=0.034). CVD at T1 predicted depressive symptoms at T2 (β=0.036, p=0.002).

Conclusions There is a bidirectional predictive effect between depressive symptoms and CVD. The effect of depressive symptoms on CVD is stable, and CVD has an effect on depressive symptoms in a short period of time.

  • depression & mood disorders
  • cardiology
  • mental health

Data availability statement

Data are available in a public, open access repository. CHARLS data of the study will be available to investigators at the CHARLS website (http://charls.pku.edu.cn/en).

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/bmjopen-2022-071175

Statistics from Altmetric.com

    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.

    STRENGTHS AND LIMITATIONS OF THIS STUDY

    • This study is based on China Health and Retirement Longitudinal Study, a large sample study, which aims to explore the impact of time through multiple time points.

    • A sensitivity analysis was conducted to ensure the stability of the results.

    • The psychology method cross-lagged panel model was used to assess the interaction effect.

    • We only examined every 2-year relationship as the time interval, more accurate time effects require more time point follow-up.

    Introduction

    With rapid economic growth and lifestyle changes, China has completed an epidemiological paradigm shift from infectious to chronic diseases.1 Of the total expected personal health expenditure, 7.8% was used for depression, which has become a major factor in the decline in quality of life and is also associated with premature death.1 2

    Cardiovascular diseases (CVDs) include coronary heart disease (CHD) and stroke.3 CVD is a public health issue and a major cause of death and disability worldwide. The prevalence rate of CVD in China has been increasing since 2006. In 2018, there were about 290 million patients with CVD in China, including 13 million patients who had a stroke and 11 million patients with CHD. In terms of mortality, two-fifths of deaths in China are attributed to CVD, which is higher than cancer or other diseases.4 5 It is estimated that stroke causes 20% of the total deaths in China, which is much higher than 11% globally, and 6% in the USA.5 Although the prevention, diagnosis and early intervention of CVD have been gradually improving, more effective health management strategies are urgently needed.6

    The mechanism underlying the association between depressive symptoms and increased risk of CVD is complex and is influenced by multiple factors, and is currently not fully understood.7 The exploration of the association between depressive symptoms and CVD has always been a causal problem of ‘chicken or egg first’. On the one hand, persons with depressive symptoms are more likely to eventually develop CVD. Previous studies have reported that depressive symptoms are an independent risk factor of CVD.8 9 Patients with depressive symptoms had an increased risk of CHD (HR: 1.28; 95% CI: 1.26 to 1.30).9 High Center for Epidemiological Studies Depression Scale (CES-D) scores can predict a higher incidence rate of stroke (p<0.05).10 On the other hand, patients with CVD have a high incidence of depressive symptoms.11–13 Epidemiological studies show that CVD is a risk factor for depressive symptoms, and the prevalence of depressive symptoms is higher in patients with CVD.14 Early longitudinal studies found that the occurrence of CVD led to the destruction of the subfrontal cortex marginal network involved in emotion regulation, which in turn led to the occurrence of depressive symptoms.15 In addition, some evidence suggests gender differences in psychological and physical responses. A study combining three large cross-sectional datasets has demonstrated the association between CVD and the severity of depressive symptoms, with a higher prevalence of depressive symptoms among female patients with CVD.11

    Currently, existing epidemiological studies on the relationship between depressive symptoms and CVD have concluded that there is a one-way association between the two. However, no longitudinal study has been used to evaluate the bidirectional association between depressive symptoms and CVD. Compared with other previous studies, expanding the meaning of the two-way relationship and using longitudinal analysis to test the potential causal relationship are undoubtedly valuable. The cross-lagged panel model (CLPM) is a statistical analysis method based on longitudinal tracking data, combining cross-sectional design with time-delay models. The basic principle is that causal effects have a time lag, and the causal variable is influenced by time to infer more detailed and closer to the true causal relationships.16 Currently, CLPM research has mostly been used to explore the relationship between factors measured by scales, such as psychological research related to depressive symptoms,17–19 and also used in research on the relationship between depressive symptoms and diseases.20 This study used national survey data from the China Health and Retirement Longitudinal Study (CHARLS) at three time points in 2011, 2013 and 2015 to explore the bidirectional association between depressive symptoms and CVD at different time points using a cross-lagged model.

    Methods

    Study design

    The study is a longitudinal study. The data come from the national survey data of CHARLS in 2011, 2013 and 2015. The project was carried out and implemented by the National Development Research Institute of Peking University to conduct household surveys on middle-aged and elderly people (45+) in China. The survey started in 2011 (T1) and followed up every 2 years in 2013 (T2) and 2015 (T3).

    Population

    The baseline survey of the CHARLS involved 17 692 people who participated in T1. Exclusion criteria for subjects were as follows: (1) people with mental retardation or malignant tumour at T1 (N=673); (2) missing covariates information (age, physical disability, education level, smoking history, hypertension, diabetes, dyslipidaemia) at T1 (N=1132); (3) lack of information on depressive symptoms or CVD at T1 (N=1456); (4) the diagnosis time of people with CVD is more than 2 years from T1 (N=1114); (5) lack of information on depressive symptoms or CVD in T2 (N=5074); (6) lack of information on depressive symptoms or CVD in T3 (N=1541). Due to many missing values of body mass index (BMI) in CHARLS and because they are only used as covariates, in order to preserve the sample size and make the sample population more representative of the core variables depressive symptoms and CVD, we first replace BMI outliers found due to errors in manually recording weight and height (BMI<Q1−1.5×IQR=13.53, or BMI>Q3+1.5×IQR=37.35) with the blank BMI. Then, the mean value was used to fill in the blank BMI of different genders (mean of BMIfemale=27.80, mean of BMImale=22.75); 826 cases (Nfemale=398, Nmale=428) whose baseline BMI was blank were interpolated to form the study population (N=6702). At the same time, in the absence of interpolation, the deletion of the blank BMI population constituted a sensitivity analysis of the sensitivity study population (N=5876). Comparing the distribution of depressive symptoms and CVD at T1, T2 and T3 between the BMI-not filled population (N=5876) and BMI-filled population (N=6702) showed that there was no difference in the distribution (p>0.05) (see online supplemental table 1). Therefore, the data after the interpolation of covariate BMI were used. A total of 6702 subjects were included (see figure 1).

    Supplemental material

    Figure 1
    Figure 1

    Flow chart of participant inclusion and exclusion. CHARLS, China Health and Retirement Longitudinal Study; CVD, cardiovascular disease.

    Measures

    Measured at T1:

    1. Sociodemographic information, which includes gender, age, urban and rural areas, marital status, education level, physical disability, smoking and BMI. Marital status is divided into married, unmarried, divorced or widowed. Education level is divided into primary school and below, junior high school, senior high school, junior college and above. Smoking is divided into never smoking, quit smoking and smoking.

    2. Cardiometabolic disease, which includes hypertension, diabetes and dyslipidaemia. Hypertension is diagnosed by hypertension history in the CHARLS questionnaire or blood test data with systolic/diastolic blood pressure ≥140/90 mm Hg. Diabetes is diagnosed by diabetes history in the CHARLS questionnaire or fasting blood glucose >7.0 mmol/L in the blood test data. Dyslipidaemia is diagnosed by answering having dyslipidaemia in the CHARLS questionnaire or meeting the requirements of total cholesterol ≥6.2 mmol/L and (or) triglycerides (TG) ≥2.3 mmol/L and (or) high-density lipoprotein cholesterol <1.0 mmol/L and (or) low-density lipoprotein cholesterol ≥4.1 mmol/L in the blood test data.

    Measured at T1, T2 and T3:

    1. Depressive symptoms. The depressive symptom measurements at three time points used the CES-D-10 in the CHARLS survey, which was developed by the Epidemiology Research Center and consists of 10 items. The total score of the scale is 30 points. The critical value is 10. A CES-D-10 score of 10 or greater is considered a depressive symptom. CES-D-10 is a reliable and effective self-evaluation measurement tool for depressive symptoms.21

    2. CVD, which includes heart disease or stroke. It is determined by answering ‘yes’ to the statements ‘whether any doctor has ever told you of heart disease’ and ‘whether any doctor has ever told you of stroke’ in the 2011 CHARLS questionnaire. The response of ‘When was the first time you were diagnosed with CVD?’ was ‘in 2010’ or ‘in 2011’ in 2011 CHARLS questionnaire means that the patient has suffered from CVD at T1 (to explore the time effect, we only include patients diagnosed with CVD 2 years from the baseline). Any response of ‘yes’ in the statement ‘Since the last visit (the past 2 years), has any doctor ever told you that you have heart disease or stroke’ in the 2013 (T2) and 2015 (T3) CHARLS questionnaire means that the patient has suffered from CVD at that time point.

    Statistical analysis

    R V.4.1.0 was used for data sorting, while SPSS V.25.0 and Mplus V.8.3 were used for data description and data analysis. Continuous variables were expressed as means and SD, and t-test was used for comparison between groups. Categorical variables were expressed by N (%), and then the statistical differences by groups were examined with the χ2 tests. We used binary logistic regression to study covariates that affect CVD and depressive symptoms at T1. Pearson correlation analysis was used to explore the correlation between depressive symptoms and CVD at three time points (T1, T2, T3). Mplus V.8.3 was used to fit a CLPM to analyse the relationship between depressive symptoms and CVD; covariates affecting CVD and depressive symptoms were controlled for in the model. Stratified by gender, a CLPM was used to analyse the relationship between depressive symptoms and CVD at T1 and T2. P values less than 0.05 were considered statistically significant.

    Patient and public involvement

    Patients and the public were not involved in the research.

    Results

    Characteristics of the participants

    In 2011 (T1), among 6702 middle-aged and elderly people, 2621 (39.10%) suffered from depressive symptoms and 432 (6.4%) suffered from CVD. The prevalence of depressive symptoms at baseline (T1) was lower in males (32.8%) than females (45.2%), urban residents (31.8%) than rural residents (43.2%), those with high education level than those with low education level, married residents (37.5%) than divorced ones (53.5%), and those who were healthy (38.5%) than those with physical disability (56.8%). The age and BMI of people with depressive symptoms were higher (p<0.05). The prevalence of depressive symptoms was different among people with smoking conditions (p<0.05).

    The prevalence of CVD was lower in in males (5.2%) than females (7.6%), rural residents (5.6%) than urban residents (8.0%), married people (6.2%) than divorced people (8.8%) (p<0.05), and healthy people (6.3%) than those with physical disability (9.9%). The prevalence of CVD was higher in people with hypertension, diabetes and dyslipidaemia (p<0.05). The age of people with CVD were higher, and the prevalence of CVD was different among people with different education levels and smoking conditions (p<0.05) (table 1).

    View this table:
    Table 1

    Baseline characteristics of different depressive symptoms and CVD in 2011

    Depressive symptoms and CVD at T1

    The binary logistic regression analysis showed that females, rural residents, people with low education level, divorced people, people who were smoking, people with physical disabilities and people with high BMI had higher prevalence risk of depressive symptoms (p<0.05). Gender, age, education, smoking, BMI, hypertension, diabetes and dyslipidaemia influenced the prevalence risk of CVD (p<0.05) (table 2).

    View this table:
    Table 2

    Binary logistic regression analysis of depressive symptoms and CVD in 2011

    Prevalence of depressive symptoms and CVD at T1, T2 and T3

    There are 2621 (39.1%) respondents with depressive symptoms, including 237 respondents with CVD at T1. At T2, 2226 (33.2%) had depressive symptoms, and among them, 244 had CVD. There are 2423 (36.2%) respondents who had depressive symptoms, of which, 320 are patients with CVD at T3 (table 3).

    View this table:
    Table 3

    Prevalence of depressive symptoms and CVD in middle-aged and elderly people in China in 2011 (T1), 2013 (T2) and 2015 (T3)

    Correlation analysis between depressive symptoms and CVD at T1, T2 and T3

    Correlation analysis showed that depressive symptoms at T1, T2 and T3 were significantly positively correlated with CVD. Simultaneity and time continuity are significantly correlated, showing a certain degree of stability at three time points (p<0.05), and there is a multidirectional causal relationship, which initially meets the conditions for building CLPM (see online supplemental table 2).

    The CLPM analysis between depressive symptoms and CVD

    The CLPM takes gender, region, education level, marital status, smoking, physical disabilities and BMI as depressive symptoms control variables, and takes gender, age, education level, smoking, BMI, hypertension, diabetes and dyslipidaemia as CVD control variables. CLPM was used to predict the relationship between depressive symptoms and CVD at three time points. The model fit for the CLPM was acceptable: χ2/ df=475.870, CFI=0.985 (CFI: acceptable >0.90, good >0.95), TLI=0.910, RMSEA=0.055 (90% CI: 0.051, 0.060) (RMSEA: acceptable <0.08, good <0.05), SRMR=0.020 (SRMR: acceptable <0.08, good <0.05) (CFI/TLI/RMSEA/SRMR is the fitting index for the CLPM).22 In order to simplify the CLPM graph, the effect of control variables on depressive symptoms and CVD was not shown (see figure 2).

    Figure 2
    Figure 2

    The cross-lagged panel model between depressive symptoms and CVD. Non-significant paths are denoted via dotted lines and significant paths via solid lines. CVD, cardiovascular disease; T1, 2011; T2, 2013; T3, 2015.

    The results showed that depressive symptoms at T1 had a positive effect on CVD at T2 (β=0.015, p=0.009), and CVD at T1 had a positive effect on depressive symptoms at T2 (β=0.036, p=0.002). Depressive symptoms at T2 had a stable effect on CVD at T3 (β=0.015, p=0.034) (see figure 2).

    Stratified analysis

    After stratification by gender, CLPM analysis was performed for the bidirectional association between depressive symptoms and CVD at T1 and T2. The results showed that there was no correlation between depressive symptoms and CVD in the male population (see online supplemental figure 1), while there was a bidirectional relationship between depressive symptoms and CVD in the female population (see online supplemental figure 2).

    Sensitivity analysis

    A CLPM analysis based on the data before BMI interpolation (N=5876) showed that the relationship between depressive symptoms and CVD is consistent with the data after BMI interpolation (see online supplemental figure 3). Depressive symptoms have an effect on CVD 2 years later (βT1–T2=0.009, p=0.007; βT2–T3=0.010, p=0.047); CVD at T1 has an effect on depressive symptoms at T2 (βT1–T2=0.061, p=0.012).

    At the same time, for the study population with complete information in 2011 (T1) and 2013 (T2) (N=8243), the CLPM analysis of depressive symptoms and CVD was constructed. The results show that depressive symptoms can predict overall prevalence of CVD in a short time (β=0.013, p<0.001). CVD also has an effect on depressive symptoms in a short time (β=0.039, p<0.001), which is consistent with the main result conclusion (see online supplemental figure 4).

    Discussion

    In the present study, the prevalence of depressive symptoms in 2011, 2013 and 2015 was 39.1%, 33.2%, 36.2%, respectively. On the other hand, the prevalence of CVD among middle-aged and elderly people in China was 6.4% in 2011, 8.0% in 2013 and 11.3% in 2015. China’s Cardiovascular Health and Disease Report 2020 points out that the prevalence rate of CVD in China is on the rise,23 which is consistent with the results of this study.

    The research results show that there is a bidirectional correlation between CVD and depressive symptoms, and both of them can positively predict the occurrence of another disease at the next time point in a short period of time (p<0.05). The positive predictive effect of depressive symptoms on CVD is relatively stable. However, the prediction of CVD on depressive symptoms only has an impact in a short time.

    Psychosocial factors are considered as risk factors for CVD. In a large prospective Eastern European health-related study (HAPIEE), depression is a powerful predictor of central vascular disease and all-cause mortality in the population.24 This study measured depressive symptoms and CVD every 2 years, and found that depressive symptoms can positively predict overall prevalence of CVD (βT1–T2=0.015, p=0.009; βT2–T3=0.015, p=0.034), with two predictive coefficients of 0.015, indicating that the effect of depressive symptoms on CVD has a relatively stable mechanism. Furthermore, multiple similar studies have found similar results.25 26 There are several plausible mechanisms underlying this association. The inflammation caused by depressive symptoms may lead to CVD through a separate physiological pathway, such as increased interleukin-6 upstream of C reactive protein, elevated TG, stress-induced hyperactivity of the hypothalamus–pituitary–adrenal axis, endothelial dysfunction and platelet activation, and excessive activation of the sympathetic nervous system.27–29 At the same time, other behavioural factors such as smoking, lack of exercise, obesity, insufficient drug compliance and unhealthy dietary patterns may also mediate the relationship between depressive symptoms and CVD.30 31 Our present study found that CVD has a positive predictive effect on depressive symptoms in a short period of time. Since CVD is an irreversible chronic disease, after diagnosis, in addition to the physical health damage caused by the disease itself, the cost of drugs for the long-term treatment will also bring an economic burden to the patient. At the same time, the suddenness and severity of the disease will also cause psychological burden to patients, which may also lead to depressive symptoms. At the T1 time point, patients with CVD may cause depressive symptoms at the T2 time point due to disease burden. However, some patients with CVD at T2 time point no longer suffer from depressive symptoms because they received health guidance and strengthened their adaptability to the disease, then CVD at T2 did not affect depressive symptoms at T3. Therefore, timely health management and psychological counselling for patients with CVD can eliminate fear and play an important role in the prognosis of CVD.

    In this study, we found gender differences in the association between depression and CVD. The risk of prevalence of depressive symptoms in women is 1.594 times higher than that in men (95% CI: 1.333, 1.906, p<0.001), and the risk of CVD in women is higher than that in men (OR=2.477, 95% CI: 1.719, 3.568, p<0.001). Previous research evidence also suggests that there is a gender difference in the relationship between depressive symptoms and CVD. Depression is more closely associated with CHD in women than in men.7 ,24 The risk of stroke in male patients with mild depression was 2.77 (95% CI: 2.27 to 3.39, p<0.001), the risk of CHD in patients with moderate and major depressive disorder was 2.21 (95% CI: 2.02 to 2.42, p<0.001) and the risk of stroke in female patients with mild depression was 2.78 (95% CI: 1.52 to 5.08, p<0.001). The risk of CHD in women with moderate and major depressive disorder was 2.28 (95% CI: 1.99 to 2.61, p<0.001).11 This cross-lagged analysis stratified by gender showed that depressive symptoms at T1 in women can positively predict CVD after 2 years in a short period of time (βfemale T1–T2=0.013, p=0.004), while CVD can positively predict depressive symptoms after 2 years (βfemale T1–T2=0.085, p=0.004). However, no interaction between the two was found in men, which is consistent with our findings on gender differences between depression and CVD.

    This study has many advantages. First, we used cross-lagged analysis, a widely used method in the field of psychology, to model the two categorical variables of depressive symptoms and CVD, which is the innovation of the article. The cross-lagged analysis results further explored the interaction between depressive symptoms and CVD. Second, we prospectively analysed the bidirectional association between depression and CVD, fully estimating the association at different time points, which has certain guiding significance for the management of CVD and psychological health maintenance of middle-aged and elderly people in China. Third, from the perspective of the research population, CHARLS has a nationally representative and universally applicable large sample population, which can reflect the health status of middle-aged and elderly people in China.

    However, there are still some limitations to the research. First, the measurement of depressive symptoms in this study used a self-reported scale. Although CES-D-10 is a validated tool for screening depressive symptoms, further diagnosis is still needed in combination with clinical practice. Furthermore, the CHARLS database does not collect information on the duration of depressive symptoms, and further experiments are needed to confirm this association. Second, the diagnosis time of CVD was self-reported, which may lead to recall bias. Finally, the follow-up of depressive symptoms and CVD in this study is based on a 2-year interval. Future research needs to increase the tracking of time nodes, change the time interval and explore the optimal time lag effect nodes.

    Conclusion

    In conclusion, depressive symptoms and CVDs are important factors affecting the health of middle-aged and elderly people in China. The study conducted a cross-lagged analysis through longitudinal panel data at three time points, and the results show that depressive symptoms and CVDs have a two-way predictive effect. Meanwhile, the positive predictive effect of depressive symptoms on CVD is relatively stable, while CVD has a positive predictive effect on depressive symptoms in a short period of time. Timely health guidance can alleviate depressive symptoms in patients with CVD at the early stage of the disease burden, which plays an important role in the prognosis of CVD.

    Data availability statement

    Data are available in a public, open access repository. CHARLS data of the study will be available to investigators at the CHARLS website (http://charls.pku.edu.cn/en).

    Ethics statements

    Patient consent for publication

    Ethics approval

    The Medical Ethics Review Committee of Peking University approved this study, and all participants provided written informed consent before participating. This study is a secondary analysis of a public dataset and does not require ethics approval again.

    Acknowledgments

    We are grateful to the CHARLS team for making the data publicly available.

    References

    1. ↵
      2. Hsieh C-R ,
      3. Qin X
      . Depression hurts, depression costs: the medical spending attributable to depression and depressive symptoms in China. HEALTH ECON 2018;27:525–44. doi:10.1002/hec.3604
      OpenUrlPubMed
    2. ↵
      2. Cuijpers P ,
      3. Smit F
      . Excess mortality in depression: a meta-analysis of community studies. J Affect Disord 2002;72:227–36. doi:10.1016/s0165-0327(01)00413-x
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      2. Xia S ,
      3. Du X ,
      4. Guo L , et al
      . Sex differences in primary and secondary prevention of cardiovascular disease in China. Circulation 2020;141:530–9. doi:10.1161/CIRCULATIONAHA.119.043731
      OpenUrlCrossRefPubMed
    4. ↵
      2. Ma L-Y ,
      3. Chen W-W ,
      4. Gao R-L , et al
      . China cardiovascular diseases report 2018: an updated summary. J Geriatr Cardiol 2020;17:1–8. doi:10.11909/j.issn.1671-5411.2020.01.001
      OpenUrlPubMed
    5. ↵
      2. Roth GA ,
      3. Mensah GA ,
      4. Johnson CO , et al
      . Global burden of cardiovascular diseases and risk factors, 1990-2019: update from the GBD 2019 study. J Am Coll Cardiol 2020;76:2982–3021. doi:10.1016/j.jacc.2020.11.010
      OpenUrlCrossRefPubMed
    6. ↵
      2. Wu S ,
      3. Zhang S ,
      4. Wu X , et al
      . M6A RNA methylation in cardiovascular diseases. Molecular Therapy 2020;28:2111–9. doi:10.1016/j.ymthe.2020.08.010
      OpenUrl
    7. ↵
      2. Vaccarino V ,
      3. Badimon L ,
      4. Bremner JD , et al
      . Depression and coronary heart disease: 2018 position paper of the ESC working group on coronary pathophysiology and Microcirculation. Eur Heart J 2020;41:1687–96. doi:10.1093/eurheartj/ehy913
      OpenUrl
    8. ↵
      2. Case SM ,
      3. Sawhney M ,
      4. Stewart JC
      . Atypical depression and double depression predict new-onset cardiovascular disease in U.S. adults. Depress Anxiety 2018;35:10–7. doi:10.1002/da.22666
      OpenUrl
    9. ↵
      2. Daskalopoulou M ,
      3. George J ,
      4. Walters K , et al
      . Depression as a risk factor for the initial presentation of twelve cardiac, cerebrovascular, and peripheral arterial diseases: data linkage study of 1.9 million women and men. PLoS One 2016;11:e0153838. doi:10.1371/journal.pone.0153838
    10. ↵
      2. Colantonio A ,
      3. Kasi SV ,
      4. Ostfeld AM
      . Depressive symptoms and other Psychosocial factors as predictors of stroke in the elderly. Am J Epidemiol 1992;136:884–94. doi:10.1093/aje/136.7.884
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      2. Piwoński J ,
      3. Piwońska A ,
      4. Zdrojewski T , et al
      . Association between cardiovascular diseases and depressive symptoms in adults. A pooled analysis of Population‐Based surveys: WOBASZ, NATPOL 2011, and WOBASZ II. Pol Arch Intern Med 2021;131:503–11. doi:10.20452/pamw.15945
      OpenUrl
    12. ↵
      2. Bianco F ,
      3. Bucciarelli V ,
      4. Gallina S
      . The association between cardiovascular diseases and depressive symptoms in adults: the classic "chicken or egg" causality dilemma. Pol Arch Intern Med 2021;131:497–8. doi:10.20452/pamw.16038
      OpenUrl
    13. ↵
      2. Bucciarelli V ,
      3. Caterino AL ,
      4. Bianco F , et al
      . Depression and cardiovascular disease: the deep blue sea of women's heart. Trends Cardiovasc Med 2020;30:170–6. doi:10.1016/j.tcm.2019.05.001
      OpenUrl
    14. ↵
      2. Clarke DM ,
      3. Currie KC
      . Depression, anxiety and their relationship with chronic diseases: a review of the epidemiology, risk and treatment evidence. Med J Aust 2009;190:S54–60. doi:10.5694/j.1326-5377.2009.tb02471.x
      OpenUrlPubMedWeb of Science
    15. ↵
      2. Aizenstein HJ ,
      3. Baskys A ,
      4. Boldrini M , et al
      . Vascular depression consensus report - a critical update. BMC Med 2016;14:161. doi:10.1186/s12916-016-0720-5
    16. ↵
      2. Yuan S ,
      3. Cao WR ,
      4. Zhang MY , et al
      . Towards more accurate causal analysis: new advances in cross lag models. Human Resources Development of China 2021;38:23–41.
      OpenUrl
    17. ↵
      2. Yang X ,
      3. Guo W-J ,
      4. Tao Y-J , et al
      . A Bidirectional association between Internet addiction and depression: A large-sample longitudinal study among Chinese university students. J Affect Disord 2022;299:416–24. doi:10.1016/j.jad.2021.12.013
      OpenUrl
    18. ↵
      2. Zhou H ,
      3. Dang L ,
      4. Lam LW , et al
      . A cross-lagged panel model for testing the Bidirectional relationship between depression and Smartphone addiction and the influences of maladaptive Metacognition on them in Chinese adolescents. Addict Behav 2021;120:S0306-4603(21)00163-5. doi:10.1016/j.addbeh.2021.106978
    19. ↵
      2. Zhang K ,
      3. Guo H ,
      4. Wang T , et al
      . A Bidirectional association between Smartphone addiction and depression among college students: A cross-lagged panel model. Front Public Health 2023;11:1083856. doi:10.3389/fpubh.2023.1083856
    20. ↵
      2. Birdi G ,
      3. Larkin M ,
      4. Knibb RC
      . Prospective analysis of the temporal relationship between psychological distress and Atopic dermatitis in female adults: a preliminary study. Healthcare (Basel) 2022;10:1913. doi:10.3390/healthcare10101913
    21. ↵
      2. Boey KW
      . Cross-validation of a short form of the CES-D in Chinese elderly. Int J Geriatr Psychiatry 1999;14:608–17. doi:10.1002/(sici)1099-1166(199908)14:8<608::aid-gps991>3.0.co;2-z
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      2. Grimm KJ ,
      3. Helm J ,
      4. Rodgers D , et al
      . Analyzing cross-lag effects: A comparison of different cross-lag modeling approaches. New Dir Child Adolesc Dev 2021;2021:11–33. doi:10.1002/cad.20401
      OpenUrl
    23. ↵
      1. The Writing Committee of the Report on Cardiovascular Health and Diseases in China 2021
      . Key points of report on cardiovascular health and diseases in China 2021. Chinese Journal of Cardiovascular Research 2022;7:577–96.
      OpenUrl
    24. ↵
      2. Kozela M ,
      3. Bobak M ,
      4. Besala A , et al
      . The Association of depressive symptoms with cardiovascular and all-cause mortality in central and Eastern Europe: prospective results of the HAPIEE study. Eur J Prev Cardiol 2016;23:1839–47. doi:10.1177/2047487316649493
      OpenUrlCrossRefPubMed
    25. ↵
      2. Pan A ,
      3. Sun Q ,
      4. Okereke OI , et al
      . Depression and risk of stroke morbidity and mortality: a meta-analysis and systematic review. JAMA 2011;306:1241–9. doi:10.1001/jama.2011.1282
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      2. Thomson W
      . Rate of stroke death after depression: a 40-year longitudinal study extension of Chichester/Salisbury catchment area study. J Stroke Cerebrovasc Dis 2014;23:1837–42. doi:10.1016/j.jstrokecerebrovasdis.2014.03.013
      OpenUrlCrossRefPubMed
    27. ↵
      2. Ridker PM
      . From C-reactive protein to Interleukin-6 to Interleukin-1: moving upstream to identify novel targets for Atheroprotection. Circ Res 2016;118:145–56. doi:10.1161/CIRCRESAHA.115.306656
      OpenUrlAbstract/FREE Full Text
    28. ↵
      2. Chen H ,
      3. Yiu K-H ,
      4. Tse H-F
      . Relationships between vascular dysfunction, circulating endothelial progenitor cells, and psychological status in healthy subjects. Depress Anxiety 2011;28:719–27. doi:10.1002/da.20839
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      2. Fenger-Grøn M ,
      3. Vestergaard M ,
      4. Pedersen HS , et al
      . Depression, antidepressants, and the risk of non-valvular atrial fibrillation: A nationwide Danish matched cohort study. Eur J Prev Cardiol 2019;26:187–95. doi:10.1177/2047487318811184
      OpenUrlPubMed
    30. ↵
      2. Carod-Artal FJ
      . Are mood disorders a stroke risk factor Stroke 2007;38:1–3. doi:10.1161/01.STR.0000251674.14708.9b
      OpenUrlFREE Full Text
    31. ↵
      2. Musselman DL ,
      3. Evans DL ,
      4. Nemeroff CB
      . The relationship of depression to cardiovascular disease: epidemiology, biology, and treatment. Arch Gen Psychiatry 1998;55:580–92. doi:10.1001/archpsyc.55.7.580
      OpenUrlCrossRefPubMedWeb of Science

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Contributors ZZ and AL designed the study. ZZ conducted the analysis and wrote the draft. YH managed the literature searches. AL revised the manuscript and is the guarantor. All authors contributed to and have approved the final manuscript.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests None declared.

    • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

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

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.