Abstract
Depression and cancer are major public health concerns worldwide, particularly among aging populations. This study aims to investigate the association between depressive symptoms and incident cancer among middle-aged and older Chinese adults. We analyzed data from the 2015 to 2020 China Health and Retirement Longitudinal Study, a nationally representative cohort of Chinese adults aged 45 and older. Depressive symptoms were assessed using the 10-item Center for Epidemiologic Studies Depression Scale at baseline. Incident cancer cases were identified through self-reported physician diagnoses during follow-up waves. Cox proportional hazard models were used to estimate the association between baseline depressive symptoms and subsequent cancer incidence, adjusting for sociodemographic factors, lifestyles, and health status. Among 13,132 participants (median age = 60 years, 46.9% male) without a history of cancer at baseline, 250 new cancer cases were identified over a median follow-up of 5.0 years. After adjusting for potential confounders, individuals with clinically significant depressive symptoms (10-item Center for Epidemiologic Studies Depression Scale score ≥10) had a 19.3% higher risk of developing cancer compared to those without (hazard ratio = 1.193, 95% confidence interval = 0.905–1.572, P = .210). This association persisted when categorized by the severity of depressive symptoms. Our findings suggest that depressive symptoms showed a nonsignificant trend toward increased risk of incident cancer in middle-aged and older Chinese adults. This underscores the importance of mental health screening and intervention in this population, potentially as part of a comprehensive strategy for cancer prevention. Further research is needed to elucidate the underlying mechanisms and to develop targeted interventions.
Keywords: cancer, CHARLS, Chinese, depressive symptom
1. Introduction
Cancer is one of the leading causes of death worldwide, and its incidence and mortality rates have continued to rise over the past decades.[1] In China, cancer has become the second leading cause of death after cardiovascular diseases.[2] Previous studies have shown that lifestyle factors such as smoking, alcohol consumption, and physical inactivity, as well as chronic diseases such as type 2 diabetes mellitus, hypertension, and renal disease, may increase the risk of cancer in individuals.[3–5] However, increasing evidence suggests that psychological factors may also play an important role in cancer development and prognosis.[6,7]
Depression is a common psychiatric disorder, the main features of which include persistent low mood, reduced interest, and decreased vigor.[8] Unlike transient mood swings, depression may have widespread and long-lasting negative effects on an individual’s physiological functioning.[9] A growing number of studies have found an association between depressive symptoms and the development of malignant tumors.[10–12] A meta-analysis of 27 prospective studies showed that patients with depression had a 15% higher risk of developing cancer than those without depression.[12] However, the available evidence is mainly from developed countries in Europe and the United States, and this association has not been fully verified in Asian populations, especially in China.
Therefore, based on national cohort data, the present study was designed to assess the association between depressive symptoms and new-onset cancer and to provide a new basis for cancer prevention in the Chinese middle-aged and elderly population.
2. Methods
2.1. Study population
The China Health and Retirement Longitudinal Study (CHARLS) is a nationally representative, ongoing prospective cohort study that covers urban and rural areas across 28 provinces in China.[13] Established in 2011, CHARLS utilizes a multistage stratified probability sampling method to recruit participants aged 45 years and older. The survey collects comprehensive data on demographics, health status, socio-economic conditions, and lifestyle factors, which allows for robust analyses of aging-related health outcomes and ensures the study cohort’s representativeness of the middle-aged and older Chinese population.
The study complied with the Declaration of Helsinki and was approved by the Biomedical Ethics Committee of Peking University (IRB00001052-11015). This study followed Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines.[14]
Participants at the time of the 2015 survey were included as baseline and follow-up to 2020. The median follow-up period was 5.0 years, which is sufficient to capture the majority of incident cancer cases based on cancer latency and prior cohort studies. We excluded participants younger than 45 years of age; participants with a diagnosis of any type of cancer at baseline; participants with missing data on the assessment of depressive symptoms at baseline; and participants who were lost to follow-up or died during the follow-up period. A total of 13,132 participants were ultimately included in the analysis. Figure 1 illustrates the participant screening flowchart.
Figure 1.
Participants screen flowchart. CESD-10 = 10-item Center for Epidemiologic Studies Depression Scale, CHARLS = China Health and Retirement Longitudinal Study.
2.2. Ascertainment of depressive symptoms
Depressive symptoms were assessed at baseline using the 10-item Center for Epidemiologic Studies Depression Scale (CESD-10), a validated screening tool for depressive symptoms in older adults.[15] The CESD-10 asks participants to rate the frequency of depressive symptoms experienced in the past week on a scale from 0 (rarely or none of the time) to 3 (most or all of the time). Total scores range from 0 to 30, with higher scores indicating more severe depressive symptoms. Following established guidelines, we categorized participants with scores ≥10 as having clinically significant depressive symptoms.[16]
2.3. Ascertainment of incident cancer events
Incident cancer cases were identified through self-reported physician diagnoses at each follow-up wave. Participants were asked, “Have you been diagnosed with cancer or a malignant tumor (excluding minor skin cancers) by a doctor?” For participants who answered “yes,” the researchers then asked further questions about the type of cancer. All diagnosed new malignant tumor events were defined as study endpoints in this study.
2.4. Covariates
Based on previous studies, we selected the following potential confounders: demographic characteristics: gender, age (<60/≥60 years), marital status (married/other), education level (elementary school and below/junior high school/high school and above), and residence (urban/rural); lifestyle factors: smoking status (never/past/present), drinking status (never/past/present), sleep duration (<7/≥7 hours). Physical activity is excluded due to a disproportionately high percentage of missing; physical health status: diabetes mellitus (yes/no), hypertension (yes/no), renal disease (yes/no), body mass index (≤18.4/18.5–23.9/≥24.0 kg/m2), activities of daily living (impaired/normal). We imputed missing values in categorical covariates using single imputation with the mode (most frequent category). The proportion of missing data for each covariate was low (all <5%; Table S1), and no imputation was performed for the exposure (CESD-10 score) or outcome (incident cancer).
2.5. Statistical analysis
We first conducted descriptive analyses to characterize the study population, comparing baseline characteristics between those with and without clinically significant depressive symptoms using chi-square tests for categorical variables and the Wilcoxon rank sum test for continuous variables.
To examine the association between baseline depressive symptoms and incident cancer, we used Cox proportional hazards models. Person time was calculated from the date of the baseline interview to the date of cancer diagnosis or the end of the study period, whichever came first. We then examined depressive symptoms as a continuous variable, estimating the hazard ratio (HR) for each 1-unit or standard deviation increase in the CESD-10 score. We constructed 4 models: Model 1 was a crude model; Model 2 adjusted for age, gender, marital status, education level, and residence; Model 3 additionally adjusted for lifestyle factors, including smoking status, drinking status, and sleep duration; and Model 4 further adjusted for health status, including diabetes, hypertension, renal disease, body mass index, and activities of daily living.
To further investigate the relationship between the severity of depressive symptoms and cancer, we categorized scores of depressive symptoms into quintiles, which were then included in a Cox proportional hazards model with quintile 1 as the reference group. In addition, we explored potential nonlinear associations using 3-node (10th, 50th, and 90th) restricted cubic spline regression. We also conducted subgroup analyses to examine whether the potential association between depressive symptoms and cancer was influenced by the following demographic and clinical characteristics: gender, age, marital status, education level, and residence; lifestyle factors: smoking status, drinking status, and sleep duration. Physical activity is excluded due to a disproportionately high percentage of missing; physical health status: diabetes mellitus, hypertension, renal disease, body mass index, and activities of daily living. P values for interactions were assessed by interaction terms and likelihood ratio tests. In addition, we performed a post hoc power analysis to estimate the statistical power of our study to detect the observed association between depressive symptoms and incident cancer risk. The post hoc power was calculated using the observed hazard ratio, the sample size, the number of incident cancer cases, and a significance level of 0.05, with the powerSurvEpi package.
All statistical analyses were performed using Stata version 26.0 (Stata Corp) and R software (version 4.2.0). Two-sided P values <.05 were considered statistically significant.
3. Results
3.1. Baseline characteristics
Of the 13,132 participants finally included in the analysis, the median age was 60 years, and 46.9% were male. The median CESD-10 score was 6. Table 1 presents key demographic and health-related factors among the study participants. The analysis revealed significant differences in age distribution, with individuals in the depressive symptom group having a slightly higher median age of 61 compared to the normal group (P < .001). Gender distribution varied significantly, with a higher proportion of females (64.6%) in the depressive symptom group compared to males (35.4%), whereas the normal group had a more balanced gender distribution. Marital status differed significantly between the groups, with a lower proportion of married individuals in the depressive symptom group (83.5%) compared to the normal group (89.9%). Rural residence was more common in the depressive symptom group (72.1%) than in the normal group (60.2%). Other factors, such as drinking, smoking, hypertension, diabetes, and kidney issues, demonstrated significant differences between the groups.
Table 1.
Participant’s demographics and baseline characteristics.
| Characteristic | Depressive status | P value | SMD | 95% CI | ||
|---|---|---|---|---|---|---|
| Overall (N = 13,132) | Normal (N = 8750) | Depressive symptoms (N = 4382) | ||||
| CESD-10 | 6 (3–12) | 4 (2–6) | 14 (12–18) | <.001* | −2.9 | −2.9 to −2.8 |
| Age (continuous), yr | 60 (53–67) | 60 (53–66) | 61 (53–67) | <.001* | −0.1 | −0.1 to 0.0 |
| Age (category), yr | <.001† | 0.1 | 0.0 to 0.1 | |||
| <60 | 6186 (47.1%) | 4245 (48.5%) | 1941 (44.3%) | |||
| ≥60 | 6946 (52.9%) | 4505 (51.5%) | 2441 (55.7%) | |||
| Gender | <.001† | 0.4 | 0.3 to 0.4 | |||
| Female | 6974 (53.1%) | 4144 (47.4%) | 2830 (64.6%) | |||
| Male | 6158 (46.9%) | 4606 (52.6%) | 1552 (35.4%) | |||
| Marital status | <.001† | 0.2 | 0.2 to 0.2 | |||
| Other | 1607 (12.2%) | 883 (10.1%) | 724 (16.5%) | |||
| Married | 11,525 (87.8%) | 7867 (89.9%) | 3658 (83.5%) | |||
| Residence | <.001† | 0.3 | 0.2 to 0.3 | |||
| Rural | 8425 (64.2%) | 5264 (60.2%) | 3161 (72.1%) | |||
| Urban | 4707 (35.8%) | 3486 (39.8%) | 1221 (27.9%) | |||
| Education | <.001† | 0.2 | 0.1 to 0.2 | |||
| Nonformal education | 2930 (22.3%) | 1983 (22.7%) | 947 (21.6%) | |||
| Primary school and below | 5745 (43.7%) | 3340 (38.2%) | 2405 (54.9%) | |||
| Secondary school | 2890 (22.0%) | 2135 (24.4%) | 755 (17.2%) | |||
| High school and above | 1567 (11.9%) | 1292 (14.8%) | 275 (6.3%) | |||
| Former drinking | <.001† | 0.2 | 0.1 to 0.2 | |||
| No | 7091 (54.0%) | 4504 (51.5%) | 2587 (59.0%) | |||
| Yes | 6041 (46.0%) | 4246 (48.5%) | 1795 (41.0%) | |||
| Current drinking | <.001† | 0.2 | 0.2 to 0.2 | |||
| No | 8493 (64.7%) | 5378 (61.5%) | 3115 (71.1%) | |||
| Yes | 4639 (35.3%) | 3372 (38.5%) | 1267 (28.9%) | |||
| Former smoking | <.001† | 0.2 | 0.2 to 0.2 | |||
| No | 7463 (56.8%) | 4675 (53.4%) | 2788 (63.6%) | |||
| Yes | 5669 (43.2%) | 4075 (46.6%) | 1594 (36.4%) | |||
| Current smoking | <.001† | 0.1 | 0.1 to 0.2 | |||
| No | 9488 (72.3%) | 6148 (70.3%) | 3340 (76.2%) | |||
| Yes | 3644 (27.7%) | 2602 (29.7%) | 1042 (23.8%) | |||
| Sleep duration | <.001† | 0.4 | 0.3 to 0.4 | |||
| <7 | 6948 (52.9%) | 4104 (46.9%) | 2844 (64.9%) | |||
| ≥7 | 6184 (47.1%) | 4646 (53.1%) | 1538 (35.1%) | |||
| Body mass index, kg/m2 (continuous) | 24.0 (21.8–25.7) | 24.0 (21.9–25.8) | 24.0 (21.5–25.4) | <.001* | 0.1 | 0.0 to 0.1 |
| Body mass index (category), kg/m2 | <.001† | 0.1 | 0.1 to 0.2 | |||
| ≤18.4 | 540 (4.1%) | 302 (3.5%) | 238 (5.4%) | |||
| 18.4–23.9 | 5222 (39.8%) | 3396 (38.8%) | 1826 (41.7%) | |||
| >23.9 | 7370 (56.1%) | 5052 (57.7%) | 2318 (52.9%) | |||
| Kidney disease | <.001† | 0.2 | 0.2 to 0.2 | |||
| No | 11,941 (90.9%) | 8123 (92.8%) | 3818 (87.1%) | |||
| Yes | 1191 (9.1%) | 627 (7.2%) | 564 (12.9%) | |||
| Activities of daily living | <.001† | 0.6 | 0.5 to 0.6 | |||
| None | 10,683 (81.4%) | 7777 (88.9%) | 2906 (66.3%) | |||
| Impaired | 2449 (18.6%) | 973 (11.1%) | 1476 (33.7%) | |||
| Hypertension | .033† | 0.0 | 0.0 to 0.1 | |||
| No | 6972 (53.1%) | 4703 (53.7%) | 2269 (51.8%) | |||
| Yes | 6160 (46.9%) | 4047 (46.3%) | 2113 (48.2%) | |||
| Diabetes | <.001† | 0.1 | 0.0 to 0.1 | |||
| No | 11,120 (84.7%) | 7478 (85.5%) | 3642 (83.1%) | |||
| Yes | 2012 (15.3%) | 1272 (14.5%) | 740 (16.9%) | |||
Data were presented as median (interquartile range) or n (%).
CESD-10 = 10-item Center for Epidemiologic Studies Depression Scale, CI = confidence interval, SMD = standardized mean difference.
Wilcoxon rank sum test.
Pearson’s chi-squared test.
3.2. Association between depressive symptoms and incident cancer
A total of 250 (1.9%) participants were diagnosed with new-onset cancer during the 5-year follow-up period. The results of Cox proportional risk modeling analyses showed (Table 2) that the risk of new cancer was significantly higher in persons with positive depressive symptoms compared to normal persons (HR = 1.253, 95% CI = 0.972–1.615, P = .082). This result was maintained after adjustment for demographic factors (Model 1), lifestyle factors (Model 2), and physical health status (Model 3), although not statistically significant. Similar results were found when modeling the total CESD-10 scores as quintiles (Table 2). After adjusting for confounders, by comparing quintile 5 (15–30 scores) with quintile 1 (0–3 scores), the adjusted HRs were 1.337 (95% CI = 0.890–2.010, P = .162) for incident cancer.
Table 2.
Association of depressive symptom and cancer risk.
| Variables | Event/total | Model 1 | Model 2 | Model 3 | Model 4 |
|---|---|---|---|---|---|
| Per 1 unit | 250/13,132 | 1.016 (95% CI = 0.997–1.035), 0.100 | 1.018 (95% CI = 0.998–1.037), 0.08 | 1.016 (95% CI = 0.996–1.036), 0.122 | 1.011 (95% CI = 0.990–1.033), 0.294 |
| Per 1 SD | 250/13,132 | 1.104 (95% CI = 0.981–1.242), 0.100 | 1.117 (95% CI = 0.987–1.263), 0.08 | 1.105 (95% CI = 0.974–1.253), 0.122 | 1.074 (95% CI = 0.940–1.226), 0.294 |
| CESD-10 | |||||
| Normal | 152/8750 | Reference | Reference | Reference | Reference |
| Depressive symptoms | 98/4382 | 1.253 (95% CI = 0.972–1.615), 0.082 | 1.278 (95% CI = 0.983–1.661), 0.067 | 1.252 (95% CI = 0.960–1.634), 0.098 | 1.193 (95% CI = 0.905–1.572), 0.210 |
| CESD-10 | |||||
| 0–3 | 62/3806 | Reference | Reference | Reference | Reference |
| 4–6 | 59/2881 | 1.243 (95% CI = 0.870–1.775), 0.232 | 1.249 (95% CI = 0.873–1.785), 0.223 | 1.238 (95% CI = 0.866–1.772), 0.242 | 1.231 (95% CI = 0.860–1.762), 0.256 |
| 7–9 | 31/2063 | 0.907 (95% CI = 0.590–1.396), 0.659 | 0.926 (95% CI = 0.600–1.430), 0.729 | 0.909 (95% CI = 0.588–1.405), 0.668 | 0.88 (95% CI = 0.568–1.363), 0.567 |
| 10–14 | 46/2269 | 1.212 (95% CI = 0.828–1.775), 0.323 | 1.247 (95% CI = 0.846–1.837), 0.264 | 1.218 (95% CI = 0.823–1.801), 0.324 | 1.17 (95% CI = 0.786–1.740), 0.439 |
| 15–30 | 52/2135 | 1.446 (95% CI = 1.000–2.090), 0.05 | 1.49 (95% CI = 1.015–2.187), 0.042 | 1.444 (95% CI = 0.978–2.133), 0.065 | 1.337 (95% CI = 0.890–2.010), 0.162 |
Data were presented as hazard ratio (95% confidence interval), and P value. Model 1: unadjusted. Model 2: adjusted for age, gender, marital status, education level, and residence. Model 3: additionally adjusted for smoking status, drinking status, and sleep duration. Model 4: further adjusted for diabetes, hypertension, renal disease, body mass index, and activities of daily living.
CESD-10 = 10-item Center for Epidemiologic Studies Depression Scale, SD = standard deviation.
When the CESD-10 was included as a continuous variable in the multivariate adjustment model, the risk of cancer was increased by 1.1% for every 1-unit increase in CESD-10 score (HR = 1.011, 95% CI = 0.990–1.033, P = .294), whereas the risk of cancer incidence was increased by 7.4% for every 1-standard deviation increase in CESD-10 (HR = 1.074, 95% CI = 0.940–1.226, P = .294; Table 2). Restricted cubic spline regression revealed an approximately linear association between depressive symptom scores and incident cancer risk across the entire range of scores (Pfor nonlinearity = .850; Fig. 2). There was no visible threshold or nonlinear relationship, suggesting that higher depressive symptoms were consistently associated with increased cancer risk without a specific inflection point.
Figure 2.
Adjusted HRs of cancer events risk, according to depressive symptoms scores. The HR from the multivariate Cox proportional risk models were adjusted for the variables of Model 4 in Table 2. CESD-10 = 10-item Center for Epidemiologic Studies Depression Scale, CI = confidence interval, HR = hazard ratio.
3.3. Subgroup analysis
We further evaluated the association between depressive symptoms and cancer risk in different population subgroups. The results showed that the association was held in both male (HR = 1.060, 95% CI = 0.680–1.650, P = .796) and female (HR = 1.260, 95% CI = 0.880–1.800, P = .884) subgroups. Stratified by age, depressive symptoms were significantly associated with cancer risk in those <60 years of age (HR = 1.560, 95% CI = 1.000–2.410, P = .048), whereas this association was weaker in the older subgroup (HR = 1.010, 95%CI = 0.710–1.450, P = .939). In addition, the results were consistent in the subgroups of BMI (18.4–23.9 kg/m2; HR = 1.030, 95% CI = 0.660–1.620, P = .884), >23.9 kg/m2 (HR = 1.280, 95% CI = 0.890–1.820, P = .184). The association between depressive symptoms and the risk of cancer development was not affected by the presence or absence of comorbid hypertension (Pfor interaction = .490) or kidney disease (Pfor interaction = .277; Fig. 3).
Figure 3.
Subgroup analysis forest plot. ADL = activities of daily living, BMI = body mass index, CI = confidence interval, HR = hazard ratio.
4. Discussion
4.1. Principal findings
Our study, utilizing data from the CHARLS, provides evidence of a positive association between depressive symptoms and incident cancer risk among middle-aged and older Chinese adults. After adjusting for a comprehensive set of potential confounders, individuals with clinically depressive symptoms at baseline had a 19.3% higher risk of developing cancer during the follow-up period compared to those without such symptoms. A post hoc power analysis indicated that, with our sample size and the observed HR, the study power was modest and did not reach the generally recommended level of 80%. This may partly explain the lack of statistical significance.
4.2. Comparison with previous studies
Our findings are consistent with several large-scale prospective studies conducted in Western populations. For instance, a meta-analysis by Jia et al found that depression was associated with a 15% increased risk of overall cancer incidence. Similarly, Gross et al[12] reported a 29% higher risk of cancer among individuals with depression in a 24-year follow-up study. Interestingly, in the French GAZEL cohort,[11] no significant association was also found between depressive symptoms and the risk of cancer development.
The discrepancies in findings across studies may be attributed to differences in study populations, measures of depression, follow-up duration, and cancer ascertainment methods. Notably, our study aims to examine this association in a large, representative sample of Chinese adults, providing valuable insights into the depression-cancer relationship in a non-Western context. It should be noted that we observed a nonsignificant association between depressive symptoms and incident cancer; several factors may have contributed to this result. These include the limited number of incident cancer cases, resulting in insufficient statistical power to detect a modest effect, as well as potential residual confounding by unmeasured or imprecisely measured variables. In addition, the reliance on self-reported depression and cancer diagnoses may have introduced outcome misclassification, further affecting the statistical significance of our findings.
4.3. Potential mechanism
Several biological and behavioral mechanisms may explain the observed association between depressive symptoms and increased cancer risk. Depression has been linked to dysregulation of the hypothalamic-pituitary-adrenal axis and the sympathetic nervous system, leading to altered cortisol patterns and increased inflammation.[17] Chronic inflammation and oxidative stress associated with depression may contribute to DNA damage and impaired DNA repair mechanisms, potentially promoting carcinogenesis.[18] In addition, Depressive symptoms are associated with elevated levels of oxidative stress, and excessive free radical production can lead to genetic mutations, which in turn promote tumorigenesis.[19]
Moreover, depression is often associated with unhealthy behaviors such as smoking, excessive alcohol consumption, and physical inactivity, which are established risk factors for various cancers.[20] While we adjusted for these factors in our analyses, residual confounding may still exist. In addition, individuals with depression may be less likely to engage in preventive health behaviors or adhere to cancer screening recommendations, potentially leading to delayed cancer detection and a more advanced stage at diagnosis.[21]
4.4. Implications for clinical practice and future research
Our findings have important implications for clinical practice and public health strategies in China. The observed association between depressive symptoms and increased cancer risk underscores the importance of mental health screening and intervention in middle-aged and older adults. Integration of mental health services into primary care settings could facilitate early detection and treatment of depression, potentially contributing to cancer prevention efforts.
Furthermore, our results highlight the need for a holistic approach to cancer prevention that considers both physical and mental health. Healthcare providers should be aware of the potential link between depression and cancer risk and consider this when assessing patients’ overall health and cancer risk profiles.
Future research should focus on elucidating the specific biological mechanisms underlying the depression-cancer association, particularly in the Chinese population. Longitudinal studies with repeated measures of depressive symptoms and more detailed cancer outcome data would be valuable. In addition, investigating the potential impact of depression treatment on subsequent cancer risk could provide insights into whether managing depression might serve as a strategy for cancer prevention.
4.5. Strengths and limitations
Our study has several strengths, including its large, nationally representative sample, prospective design, and use of a validated measure of depressive symptoms. The availability of detailed information on potential confounders allowed for comprehensive adjustments in our analyses.
However, some limitations should be considered when interpreting our results. First, cancer diagnoses were based on self-reports, which may be subject to recall bias or misclassification. While self-reported cancer diagnoses have shown good agreement with medical records in previous studies,[22] future research using validated cancer registries would be valuable. Second, we assessed depressive symptoms only at baseline, and changes in depressive status during the follow-up period were not accounted for. Third, baseline differences were observed between participants with and without depressive symptoms, such as age, smoking status, and comorbidities, which may have introduced residual confounding. Although we adjusted for these variables in our models, the possibility of unmeasured or inadequately measured confounders cannot be completely excluded. In addition, we recognize that performing multiple subgroup analyses may inflate the risk of false-positive findings; thus, interpretation of these results should be cautious. Although we planned to use propensity score matching to further control for confounding, the relatively low event rate and baseline imbalance limited its utility in our dataset. Furthermore, because detailed information on the exact timing of cancer diagnoses was limited, we were unable to perform sensitivity analyses excluding cases occurring in the first 1 to 2 years of follow-up. As a result, we cannot fully exclude the possibility of reverse causality, whereby preclinical cancer may have contributed to depressive symptoms at baseline. Another limitation is that we used single imputation with the mode for missing covariate data, which does not account for imputation uncertainty and may underestimate variability compared with multiple imputation approaches. Although the proportion of missing data was relatively low, this method may still introduce some bias. Finally, the present study is based on data from a Chinese population, and the generalizability of the results to populations of other races or regions needs to be further verified. Last but not least, this study focused only on new cancer events and failed to assess the impact of depressive symptoms on tumor prognosis. Follow-up studies should further explore the potential impact of depressive symptoms on cancer survival.
5. Conclusion
In conclusion, our study provides evidence of an association between depressive symptoms and increased cancer risk in middle-aged and older Chinese adults. These findings underscore the importance of addressing mental health as part of comprehensive cancer prevention strategies in China. Future studies with larger samples and objective cancer diagnoses are needed to confirm our findings and clarify the underlying mechanisms. In particular, linkage with population-based cancer registries would allow more accurate and complete ascertainment of cancer outcomes, and repeated CESD-10 assessments over time would better capture changes in depressive symptoms and their temporal relationship with cancer development. Such enhancements, together with more detailed information on cancer subtypes and stages, could help elucidate causal pathways and identify high-risk subgroups who might benefit most from integrated mental and physical health interventions.
Acknowledgments
We thanked the China Center for Economic Research, and the National School of Development of Peking University for providing the data.
Author contributions
Conceptualization: Yuanjing Zhao, Lei Zhou, Lexia Lei, Ye Zhang, Huachun Zhang.
Data curation: Yuanjing Zhao, Lei Zhou, Qiuzi Sun, Lexia Lei, Ye Zhang.
Writing – original draft: Yuanjing Zhao, Lei Zhou, Lexia Lei.
Methodology: Qiuzi Sun, Cuiya Tang.
Project administration: Cuiya Tang, Maoying Guan.
Supervision: Huachun Zhang.
Writing – review & editing: Huachun Zhang, Maoying Guan.
Funding acquisition: Maoying Guan, Lei Zhou.
Abbreviations:
- CESD-10
- 10-item Center for Epidemiologic Studies Depression Scale
- CHARLS
- China Health and Retirement Longitudinal Study
- CI
- confidence interval
- HR
- hazard ratio
This study was supported by Shanghai Clinical Medical Research Center of Chinese Medicine and Tumor (21MC1930500), Shanghai Municipal Health Commission Scientific Research Project (20244Y0111), National Medical Center Clinical Research Project, Longhua Hospital of Shanghai University of Traditional Chinese Medicine (GFB2401-1-1), Shanghai New Three-Year Action Plan Research Project (2025–2027; GFB2502), National Medical Center Clinical Research Project, National Administration of Traditional Chinese Medicine (ZYYZDXK-2023063), National Program for the Construction of Advantageous Traditional Chinese Medicine Specialty (XH40202-20250904), and Health Commission of Pudong New Area, Shanghai Municipality, Joint Research Project (PW2024D-13).
The authors have no conflicts of interest to disclose.
The datasets generated during and/or analyzed during the current study are publicly available.
Supplemental Digital Content is available in the online version of this article (http://dx.doi.org/10.1097/MD.0000000000048791).
How to cite this article: Zhao Y, Zhou L, Sun Q, Lei L, Zhang Y, Tang C, Zhang H, Guan M. Depressive symptoms and incident cancer risk in middle-aged and older Chinese adults: Findings from CHARLS. Medicine 2026;105:21(e48791).
Contributor Information
Yuanjing Zhao, Email: mmjing08@163.com.
Lei Zhou, Email: lucyzhou100@163.com.
Qiuzi Sun, Email: ratatt@163.com.
Lexia Lei, Email: 1843299776@qq.com.
Ye Zhang, Email: lhhlky@163.com.
Cuiya Tang, Email: 1320694129@qq.com.
Huachun Zhang, Email: lhhlky@163.com.
References
- [1].Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68:394–424. [DOI] [PubMed] [Google Scholar]
- [2].Chen W, Zheng R, Baade PD, et al. Cancer statistics in China, 2015. CA Cancer J Clin. 2016;66:115–32. [DOI] [PubMed] [Google Scholar]
- [3].Islami F, Goding Sauer A, Miller KD, et al. Proportion and number of cancer cases and deaths attributable to potentially modifiable risk factors in the United States. CA Cancer J Clin. 2018;68:31–54. [DOI] [PubMed] [Google Scholar]
- [4].Spei ME, Samoli E, Bravi F, La Vecchia C, Bamia C, Benetou V. Physical activity in breast cancer survivors: a systematic review and meta-analysis on overall and breast cancer survival. Breast. 2019;44:144–52. [DOI] [PubMed] [Google Scholar]
- [5].Stein KB, Snyder CF, Barone BB, et al. Colorectal cancer outcomes, recurrence, and complications in persons with and without diabetes mellitus: a systematic review and meta-analysis. Dig Dis Sci. 2010;55:1839–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Chida Y, Hamer M, Wardle J, Steptoe A. Do stress-related psychosocial factors contribute to cancer incidence and survival? Nat Clin Pract Oncol. 2008;5:466–75. [DOI] [PubMed] [Google Scholar]
- [7].Satin JR, Linden W, Phillips MJ. Depression as a predictor of disease progression and mortality in cancer patients: a meta-analysis. Cancer. 2009;115:5349–61. [DOI] [PubMed] [Google Scholar]
- [8].Malhi GS, Mann JJ. Depression. Lancet. 2018;392:2299–312. [DOI] [PubMed] [Google Scholar]
- [9].Penninx BW, Milaneschi Y, Lamers F, Vogelzangs N. Understanding the somatic consequences of depression: biological mechanisms and the role of depression symptom profile. BMC Med. 2013;11:129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Cuijpers P, Vogelzangs N, Twisk J, Kleiboer A, Li J, Penninx BW. Comprehensive meta-analysis of excess mortality in depression in the general community versus patients with specific illnesses. Am J Psychiatry. 2014;171:453–62. [DOI] [PubMed] [Google Scholar]
- [11].Gross AL, Gallo JJ, Eaton WW. Depression and cancer risk: 24 years of follow-up of the Baltimore epidemiologic catchment area sample. Cancer Causes Control. 2010;21:191–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Jia Y, Li F, Liu YF, Zhao JP, Leng MM, Chen L. Depression and cancer risk: a systematic review and meta-analysis. Public Health. 2017;149:138–48. [DOI] [PubMed] [Google Scholar]
- [13].Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China Health and Retirement Longitudinal Study (CHARLS). Int J Epidemiol. 2014;43:61–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Vandenbroucke JP, von Elm E, Altman DG, et al. Strengthening the reporting of observational studies in epidemiology (STROBE): explanation and elaboration. Epidemiology. 2007;18:805–35. [DOI] [PubMed] [Google Scholar]
- [15].Andresen EM, Malmgren JA, Carter WB, Patrick DL. Screening for depression in well older adults: evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale). Am J Prev Med. 1994;10:77–84. [PubMed] [Google Scholar]
- [16].Song Q, Wang H, Burr JA. Depressive symptoms among rural “left-behind” older adults: a test of the high outmigration penalty hypothesis. J Gerontol B Psychol Sci Soc Sci. 2022;77:592–603. [DOI] [PubMed] [Google Scholar]
- [17].Sephton SE, Sapolsky RM, Kraemer HC, Spiegel D. Diurnal cortisol rhythm as a predictor of breast cancer survival. J Natl Cancer Inst. 2000;92:994–1000. [DOI] [PubMed] [Google Scholar]
- [18].Currier MB, Nemeroff CB. Depression as a risk factor for cancer: from pathophysiological advances to treatment implications. Annu Rev Med. 2014;65:203–21. [DOI] [PubMed] [Google Scholar]
- [19].Reuter S, Gupta SC, Chaturvedi MM, Aggarwal BB. Oxidative stress, inflammation, and cancer: how are they linked? Free Radic Biol Med. 2010;49:1603–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Russ TC, Stamatakis E, Hamer M, Starr JM, Kivimäki M, Batty GD. Association between psychological distress and mortality: individual participant pooled analysis of 10 prospective cohort studies. BMJ. 2012;345:e4933. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Carney CP, Jones LE. The influence of type and severity of mental illness on receipt of screening mammography. J Gen Intern Med. 2006;21:1097–104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Bergmann MM, Calle EE, Mervis CA, Miracle-McMahill HL, Thun MJ, Heath CW. Validity of self-reported cancers in a prospective cohort study in comparison with data from state cancer registries. Am J Epidemiol. 1998;147:556–62. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.