Association of Depressive Symptoms With Incident Cardiovascular Diseases in Middle-Aged and Older Chinese Adults

1Depressive symptoms are common among middle-aged and older adults., China, like much of Asia, is experiencing an increase in older adults. The prevalence of depressive symptoms in older adults has become an increasingly important public health priority.,, A national survey in China showed that approximately 30% of men and 43% of women aged 45 years and older experienced depressive symptoms. Epidemiological studies,, have found that depressive symptoms are associated with a range of negative health outcomes, such as coronary heart diseases, stroke, and all-cause mortality.

2Some studies,, found that both the history and new onset of depressive symptoms were associated with a series of cardiovascular events, such as angina, myocardial infarction, stroke, heart failure, and coronary death. A meta-analysis of 24 prospective cohort studies found that depressive symptoms could be associated with a 30% excess risk for coronary heart disease. The association between depression and incident cardiovascular disease (CVD) may vary across different depression subtypes and may be bidirectional; for example, depressive symptoms are associated with an increased risk of CVD, whereas cardiovascular risk factors also are associated with depression or depressive symptoms., Therefore, to reduce the risk of CVD, it is important to understand its association with depressive symptoms.

3Depressive symptoms are usually assessed using validated rating scales with established cutoffs, such as the Center for Epidemiologic Studies Depression Scale (CES-D) and the Geriatric Depression Scale. Although these scales cannot be used to establish the diagnosis of major depression, they are widely used in both research and daily practice, and most confirmatory factor analysis studies support 2 clusters of symptoms: emotional or affective (eg, felt depressed, happy, or lonely) and somatic symptoms (eg, fatigue, appetite, or sleep). Most previous studies,,,, on the association between depressive symptoms and CVD risk only analyzed the presence of depressive symptoms as a binary variable (eg, depressed or not) or used its total score, with the assumption that all depressive symptoms are equally good severity indicators, despite the lack of evidence to support this.

4To date, the contribution of individual depressive symptoms to incident CVD is still unknown, which gives us the impetus to examine the association between specific depressive symptoms and incident CVD among middle-aged and older adults in China on the basis of the China Health and Retirement Longitudinal Study (CHARLS). We hypothesized that elevated depressive symptoms and certain specific depressive symptoms would be associated with increased risk of CVD.

5This cohort study was a secondary analysis of the data set of the CHARLS, which is an ongoing nationally representative cohort study. Details of the study design have been described elsewhere., In brief, a total of 17 708 participants in 10 257 households were recruited from 150 counties or districts and 450 villages within 28 provinces in China between June 2011 and March 2012, using the multistage stratified probability-proportional-to-size sampling technique. All participants underwent an assessment using a standardized questionnaire to collect data on sociodemographic and lifestyle factors and health-related information. The response rate of the baseline survey was 80.5%. All participants were followed up every 2 years after the baseline survey.

6The CHARLS study was approved by the institutional review board of Peking University. Written informed consent was obtained from all participants. This study was conducted following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

7In the baseline survey of the CHARLS, depressive symptoms were measured using the CES-D short form, which is a widely used self-report measure on depressive symptoms in population-based studies. The CES-D short form consists of 10 items: (1) bothered by little things, (2) had trouble concentrating, (3) felt depressed, (4) everything was an effort, (5) felt hopeful, (6) felt fearful, (7) sleep was restless, (8) felt happy, (9) felt lonely, and (10) could not get going. Depressive symptoms in the past week were measured from 0 (rarely or none of the time [<1 day]) to 3 (most or all of the time [5-7 days]). Before summing item scores, items 5 and 8 need to be reverse scored. The CES-D total score varies from 0 to 30, with a higher score indicating more depressive symptoms. The CES-D short form has excellent psychometric properties among Chinese older adults. A total score of 12 or higher was used as the cutoff for having elevated depressive symptoms.

8The study outcome was incident CVD events. In accordance with previous studies,, incident CVD events were assessed by the following standardized questions: “Have you been told by a doctor that you have been diagnosed with a heart attack, coronary heart disease, angina, congestive heart failure, or other heart problems?” or “Have you been told by a doctor that you have been diagnosed with a stroke?” Participants who reported heart disease or stroke during the follow-up period were defined as having incident CVD. The date of CVD diagnosis was recorded as being between the date of the last interview and that of the interview reporting an incident CVD.,

9At baseline, trained interviewers collected information on sociodemographic status and health-related factors using a structured questionnaire, including age, sex, living residence, marital status, and educational level. Educational level was classified as no formal education, primary school, middle or high school, and college or above. Health-related factors included self-reported smoking and drinking status (never, former, or current), self-reported physician-diagnosed medical conditions (diabetes, hypertension, dyslipidemia, and chronic kidney disease), and use of medications for diabetes, hypertension, and dyslipidemia. Marital status was classified into 2 groups: married and other marital status (never married, separated, divorced, and widowed). Height, weight, and blood pressure were measured by a trained nurse. Body mass index was calculated as weight in kilograms divided by height in meters squared.

10A subcohort of 8696 CHARLS participants underwent metabolic examinations, including fasting plasma glucose, total cholesterol, triglycerides, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, high-sensitivity C-reactive protein, and serum creatinine. The estimated glomerular filtration rate was calculated using the Chronic Kidney Disease Epidemiology Collaboration’s 2009 creatinine equation.

11Statistical analysis was conducted from April 25, 2018, to December 13, 2018. Data were described as means and SDs for normally distributed continuous variables, and as medians and interquartile ranges for nonnormally distributed continuous variables. Frequency with percentage was used to describe categorical variables. Baseline characteristics are summarized according to depressive symptoms and compared between participants with and without elevated depressive symptoms using the χ2 test, analysis of variance, or Mann-Whitney U test, as appropriate. Eighteen percent (2235 of 12 417) of total data items were missing, were assumed to be missing at random, and, thus, were imputed with the multiple imputation of chained equations method using the baseline characteristics. We created 10 imputed data sets and pooled the results using the Stata statistical software version 15.1 (StataCorp) command “mi estimate.”

12We computed the person-time of follow-up for each participant from the date of the 2011 to 2012 survey (baseline) to the dates of the CVD diagnosis, death, loss to follow-up (605 of 12 417 [4.9%]), or the end of follow-up (June 31, 2015), whichever came first. Incidence rates of CVD events per 1000 person-years were calculated by depressive symptoms. To examine the association between depressive symptoms and incident CVD events, Cox proportional hazards models were used to calculate hazard ratios (HRs) with 95% CIs. Three models were estimated: in model 1, age and sex were adjusted; in model 2, age, sex, residence, marital status, educational level, smoking status, and drinking status were adjusted; and in model 3, the variables in model 2 plus history of diabetes, hypertension, dyslipidemia, and chronic kidney disease; systolic blood pressure and body mass index; and use of hypertension medications, diabetes medications, and lipid-lowering therapy were adjusted. To examine the association between specific depressive symptoms (eg, CES-D individual items) and incident CVD events, using the method of Jokela et al, we coded the items as dichotomous variables by defining the responses as occasionally or a moderate amount of time (3-4 days) and all of the time (5-7 days) as having the specific symptoms. All 10 items were entered simultaneously in model 3.

13To further examine the association between the severity of depressive symptoms and incident CVD events, scores of depressive symptoms were split into quintiles and then were included in Cox proportional hazards models with quintile 1 as the reference group. In addition, we explored the potential nonlinear associations using 3-knotted restricted cubic spline regression. Subgroup analyses were conducted to examine whether the potential association between depressive symptoms and CVD events was moderated by the following demographic and clinical characteristics: age, sex, residence, marital status, educational level, smoking status, drinking status, diabetes (defined as fasting plasma glucose ≥126 mg/dL [to convert to millimoles per liter, multiply by 0.0555], current use of antidiabetic medication, or self-reported history of diabetes), hypertension (defined as systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg, current use of the antihypertensive medication, or self-reported history of hypertension), dyslipidemia (defined as total cholesterol ≥240 mg/dL [to convert to millimoles per liter, multiply by 0.0259], triglycerides ≥150 mg/dL, low-density lipoprotein cholesterol ≥160 mg/dL, high-density lipoprotein cholesterol <40 mg/dL, current use of lipid-lowering medication, or self-reported history of dyslipidemia), chronic kidney disease (defined as estimated glomerular filtration rate <60 mL/min/1.73 m2 or self-reported history of chronic kidney disease), and body mass index. P values for interaction were evaluated using interaction terms and likelihood ratio tests.

14Three sensitivity analyses were conducted as follows: (1) further adjusting for metabolic biomarkers in model 3 in the sample of 8696 participants who underwent metabolic examinations; (2) repeating all analyses using the complete data set (10 186 participants) without multiple imputations; and (3) using the Fine and Gray competing risk model to account for competing risks due to mortality. Two-sided P <.05 was considered as statistically significant. All analyses were performed using Stata statistical software version 15.1 (StataCorp) and R statistical software version 3.6.1 (R Foundation).

15Of the 17 708 CHARLS participants at study baseline, we excluded 1841 individuals younger than 45 years, 2789 with heart disease or stroke at baseline, 1878 with incomplete information on depressive symptoms, and 140 with no answers for the questions on the physician-diagnosis CVD during follow-up. Finally, 12 417 participants were included for analysis, and 8696 (70.0%) of them provided blood samples at baseline. A comparison of baseline characteristics between participants included and those who were not included in the analysis is shown in eTable 1 in the Supplement.

16A total of 12 417 adults were included in the analyses. The mean (SD) age at baseline was 58.40 (9.51) years; 6113 (49.2%) of the participants were men and 6304 (50.8%) were women. Table 1 shows the characteristics of the participants. At baseline, 3223 participants (26.0%) had elevated depressive symptoms (CES-D total score ≥12). Univariate analysis revealed that compared with those without depressive symptoms, participants with depressive symptoms were more likely to be older (mean [SD] age, 60.10 [9.85] vs 57.81 [9.32]; difference, 2.29 years; 95% CI, 19.91 to 2.67 years; P <.001), be female (60.9% vs 47.2%; difference, 13.7%; 95% CI, 11.7% to 15.6%; P <.001), live in a rural setting (rural residence, 71.9% vs 57.2%; difference, 14.7%; 95% CI, 12.8% to 16.5%; P <.001), be unmarried (23.8% vs 14.2%; difference 9.6%; 95% CI, 8.0% to 11.2%; P <.001), have no formal education (59.1% vs 39.2%; difference, 19.9%; 95% CI, 17.9% to 21.8%; P <.001), be never smokers (64.7% vs 58.0%; difference, 6.7%; 95% CI, 4.7% to 8.6%; P <.001), be never drinkers (63.1% vs 56.2%; difference, 6.9%; 95% CI, 5.0% to 8.9%; P <.001), have a history of hypertension (24.6% vs 21.6%; difference, 3.0%; 95% CI, 1.3% to 4.7%; P <.001) and chronic kidney disease (8.0% vs 4.1%; difference, 3.9%; 95% CI, 2.9% to 5.0%; P <.001), use medications for diabetes (3.8% vs 3.0%; difference, 0.8%; 95% CI, 0.1% to 1.6%; P =.02) and hypertension (17.3% vs 15.2%; difference, 2.1%; 95% CI, 0.6% to 3.6%; P =.01), and have lower diastolic blood pressure (mean [SD], 75.02 [12.27] mm Hg vs 76.20 [12.07] mm Hg; difference, −1.18 mm Hg; 95% CI, −1.70 to −0.66 mm Hg; P <.001), body mass index (mean [SD], 22.76 [3.78] vs 23.55 [3.86]; difference, −0.78; 95% CI, −0.95 to −0.62; P <.001), and estimated glomerular filtration rate (mean [SD], 92.25 [14.88] mL/min/1.73 m2 vs 93.11 [14.67] mL/min/1.73 m2; difference, −0.87 mL/min/1.73 m2; 95% CI, −1.57 to −0.16 mL/min/1.73 m2; P <.02) but higher high-density lipoprotein cholesterol (mean [SD], 52.53 [15.75] mg/dL vs 50.84 [15.20] mg/dL; difference, 1.68 mg/dL; 95% CI, 0.95 to 2.42 mg/dL; P <.001).

17During the follow-up period between 2011 and 2015, 1088 participants experienced incident CVD (heart disease, 929 cases; stroke, 190 cases). The incidence rate of CVD was 29.18 per 1000 person-years among participants with elevated depressive symptoms and 20.55 per 1000 person-years among participants without elevated depressive symptoms. Table 2 shows the associations between depressive symptoms and incident CVD events. After adjusting for potential confounders (in model 3), the presence of elevated baseline depressive symptoms was independently associated with a 39.0% increased risk of incident CVD (CVD: adjusted HR, 1.39; 95% CI, 1.22-1.58; heart disease: adjusted HR, 1.36; 95% CI, 1.18-1.57; stroke: adjusted HR, 1.45; 95% CI, 1.06-1.99). Similar results were found when modeling the total CES-D scores as quintiles (Table 2). After adjusting for confounders, by comparing quintile 5 with quintile 1, the adjusted HRs were 1.75 (95% CI, 1.45-2.11) for incident CVD, 1.73 (95% CI, 1.42-2.12) for heart disease, and 1.69 (1.09-2.64) for stroke. A linear and positive association between the CES-D total score and risk of incident CVD events using restricted cubic spline regression was also found (for nonlinearity, P =.30 for CVD, P =.40 for heart disease, and P =.82 for stroke) (Figure 1).

18Of the individual depressive symptoms, the most common ones included feeling hopeless (35.9%), restless sleep (32.7%), and being bothered by little things (31.7%) (Table 3). Table 3 shows the associations of individual depressive symptoms and incident CVD events. When entering all 10 individual depressive symptoms as measured by the CES-D items in model 3, only 2 symptoms (sleep was restless: adjusted HR, 1.21; 95% CI, 1.06-1.39; and felt lonely: adjusted HR, 1.21; 95% CI, 1.02-1.44) were significantly associated with incident CVD, after adjusting for potential confounders. Feeling lonely was also significantly associated with stroke (HR, 2.10; 95% CI, 1.43-3.10).

19Figure 2 shows the association between depressive symptoms and incident CVD events stratified by potential risk factors. The association between depressive symptoms and incident CVD was more pronounced among participants without hypertension (adjusted HR, 1.68; 95% CI, 1.36-2.07), compared with those with hypertension (adjusted HR, 1.27; 95% CI, 1.07-1.50) at baseline (P =.04 for interaction). The results did not significantly change after further adjusting for total cholesterol, triglycerides, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, fasting plasma glucose, estimated glomerular filtration rate, or high-sensitivity C-reactive protein levels (eTable 2 in the Supplement). Similar results were found when complete data analyses were conducted (eTable 3 in the Supplement). In addition, elevated depressive symptoms were significantly associated with 1.41-fold (95% CI, 1.23-1.62) incident CVD when using the Fine and Gray model with death as competing risk event (eTable 4 in the Supplement).

20This study examined the associations between depressive symptoms and incident CVD in a nationally representative cohort of 12 417 adults in China aged 45 years and older with over 4 years of follow-up. At baseline, 26.0% of the participants experienced depressive symptoms. Depressive symptoms were associated with a 39.0% increased risk of CVD. The presence of certain depressive symptoms (restless sleep and loneliness) were independently associated with incident CVD.

21Increasing evidence suggests that the presence of depressive symptoms is associated with an increased risk of CVD.,,,, The Jackson Heart Study found an almost 2-fold increase in the risk of coronary heart disease and stroke in patients with major depression. The Reasons for Geographical and Racial Differences in Stroke Study also found that severe depressive symptoms were associated with an increased risk of coronary heart disease, stroke, and mortality. A recent cohort study suggested that time-dependent depressive symptoms were associated with 1.4-fold risk of cardiovascular mortality. In addition, several meta-analyses and systematic reviews,, found a positive association between severe depressive symptoms and increased risk of CVD. As expected, the risk of CVD including heart disease and stroke was associated with depressive symptoms in this study.

22Although the association between depressive symptoms and incident CVD has been widely examined, the contribution of specific depressive symptoms to incident cardiovascular events is still not clear. This study found that restless sleep and loneliness were independently associated with incident CVD and stroke. Restless sleep or insomnia previously have been associated with CVD.,, In the Health and Retirement Study Cohort, 2 individual depressive symptoms (everything was an effort and restless sleep) were independently associated with an increased risk of stroke. A recent study also found that a combination of depressive symptoms and sleep problems was associated with an increase in the odds of coronary heart disease at 6-year follow-up. The mechanisms that underlie the association between sleep problems and increased risk of CVD have been widely examined. Some studies found that short sleep duration or nonrestorative sleep could lead to metabolic or endocrine changes through elevated levels of inflammatory cytokines and sympathetic activation,, poor sleep quality with low levels of slow-wave sleep could impair glucose tolerance and then increase the risk of type 2 diabetes, and sleep curtailment could increase cortisol secretion and alter circulating levels of growth hormone, leptin, and ghrelin, all of which are associated with an increased risk of CVD.

23In the cross-sectional Netherlands Study of Depressed Older Persons, Hegeman et al found an independent association between loneliness and CVD in women (odds ratio, 1.13; 95% CI, 1.06-1.21). On the basis of a secondary analysis of the English Longitudinal Study of Aging, Valtorta et al found that loneliness was associated with a 1.27-fold increased risk of CVD (HR, 1.27; 95% CI, 1.01-1.57). Loneliness could increase the activity of the hypothalamic-pituitary-adrenal axis, leading to higher levels of circulating cortisol.,, Increased cortisol levels further affect blood pressure and vascular endothelial function through the vascular nitric oxide system, which may account for the association between loneliness and CVD. However, 2 longitudinal studies,, of older adults in the United States and United Kingdom reported that loneliness did not have cardiometabolic effects. Apart from differences in methods, we assume that different sociocultural contexts across these countries may partly explain the different findings between studies,, although convincing findings are needed to support this notion. In China, loneliness among elders may take on a specific relevance. As China continues to develop, it is less common for young people to comport with traditional Confucian ideals of filial piety, and this, coupled with massive internal migration, may disruptive to care structures for older adults.

24The underlying mechanisms of the association between depressive symptoms and CVD are multifactorial, involving autonomic nerve dysfunction, inflammation, endothelial dysfunction, platelet activation and thrombosis, life behavior, and cardiac metabolic risk factors. For example, a meta-analysis found that depression may be associated with dietary patterns, which could change the gut microbiome and then increase CVD risk. Furthermore, the association between elevated levels of inflammatory markers, such as high-sensitivity C-reactive protein, and depression are well documented., In the current study, after adjusting for high-sensitivity C-reactive protein, the association between depressive symptoms and increased risk of CVD remained, which suggests these findings are robust. In this study, apart from restless sleep and loneliness, other individual depressive symptoms as measured by the CES-D were not significantly associated with incident CVD. We have no clear explanation about the associations between different individual depressive symptoms and incident CVD except for assuming that the effect size between both restless sleep and loneliness and CVD is, perhaps, greater than those for other individual depressive symptoms because of biological and environmental factors., In addition, the associations between other depressive symptoms and CVD may be undetected because of the short study period (4 years).

25The strengths of this study included the prospective design, the long follow-up period, and the inclusion of specific depressive symptoms. However, several limitations need to be addressed. First, some confounding factors of the association between depression and CVD, such as income, social support, isolation, and joblessness,, were not adjusted in this study. Second, similar to other studies,, for logistical reasons, the diagnosis of CVD was self-reported. Medical records were not available in the CHARLS; however, some other large-scale studies, such as the English Longitudinal Study of Aging, found that self-reported incident coronary heart disease had a good agreement with medical records (accuracy, 77.5%). Third, only participants from China were involved in this study; thus the findings may not fully generalize to other countries. In addition, time-varying exposures were not included in the present analysis, so residual confounding is a concern.

26This study found that the presence of certain depressive symptoms, such as restless sleep and loneliness, could be associated with an increased risk of CVD in middle-aged and older Chinese adults. To reduce the risk of CVD, effective treatment and psychosocial interventions should be delivered targeting these symptoms.