Depression and long-term mortality among 5-year breast cancer survivors in Korea: a retrospective population-based cohort study

Abstract

To examine the association of depression after breast cancer diagnosis with long-term mortality risk in a Korean cohort. We conducted a retrospective population-based cohort study using data from the National Health Insurance Service (NHIS) cancer patient’s cohort of South Korea. We included women aged 40 years or older who were diagnosed with breast cancer between 2007 and 2013, survived at least 5 years, and had no history of depression prior to the breast cancer diagnosis. Depression was defined as hospitalization for more than 2 days with a primary diagnosis of depression (ICD-10 codes F32–F33). We evaluated all-cause, cancer-specific, and non-cancer-specific mortality using the Cox proportional hazards model while adjusting for covariates. Among 30,873 eligible women (mean age, 56.5 years), 502 were diagnosed with new-onset depression during the 5-year survival period, whereas 30,371 did not develop depression. During follow-up after the 5-year survivor period, a total of 1,904 deaths occurred. New-onset depression was associated with higher risks of all-cause mortality (adjusted hazard ratio [aHR], 1.38; 95% CI, 1.03–1.86; p = 0.033) and non-cancer mortality (aHR, 1.81; 95% CI, 1.14–2.86; p = 0.011), while no significant association was observed for cancer-specific mortality. The association was particularly pronounced among patients aged 65 years or older (aHR, 1.96; 95% CI, 1.27–3.03; p = 0.002). Depression was linked to increased mortality among 5-year breast cancer survivors, especially for non-cancer causes. This study implies the need for depression screening and treatment among breast cancer patients, especially in non-Western settings where depression may be underdiagnosed and undertreated.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-36919-y.

Introduction

Women with breast cancer have to cope with many difficulties that affect their quality of life in both psychological and physical aspects, as they face various sources of stress, fear, and emotional distress related to their condition, treatment options, and outcomes^1,2. Depression is one of the most common mental health problems among breast cancer patients, affecting up to 40% of them at some point during or after their treatment³. Depression can have negative impacts on the quality of life, adherence to treatment, and survival outcomes of breast cancer patients^4,5. The mechanisms by which depression influences breast cancer survival are complex and multifactorial, involving both behavioral and biological pathways^6,7. Depression has been associated with impaired immune function, increased inflammation, changes in hormonal levels, and the potential to impact cytokine homeostasis, consequently enhancing angiogenesis and metastasis⁸. Moreover, depression may reduce the likelihood of receiving guideline-recommended treatment, which can affect survival⁷.

Substantial evidence gathered from observational and clinical studies strongly indicates a link between the simultaneous occurrence of depression and poorer survival outcomes among women diagnosed with breast cancer, as noted across short- and long-term follow-ups^4,9,10. However, previous research has presented conflicting outcomes regarding the connection between depression and survival in breast cancer¹¹. Some observational studies have even suggested the absence of a significant relationship between depression and mortality^11–13. Moreover, most of these studies have primarily centered on Western, high-income populations, where patterns of breast cancer care, accessibility of mental health services, and cultural attitudes toward depression may differ substantially from those in non-Western or low- and middle-income settings. Meta-analytic data indicate that the prevalence of depression in breast cancer patients is substantially higher in middle-income than in developed countries, with the highest rates reported in the Eastern Mediterranean region⁵, and international surveys show that structured psycho-oncology services are far less available in low- and middle-income settings^14,15. Furthermore, studies from East Asian and African contexts highlight how cancer- and mental-illness–related stigma and fatalistic beliefs can exacerbate depressive symptoms and deter help-seeking among survivors^16,17. Consequently, uncertainty remains regarding any definitive association between depression and the progression or mortality rates related to breast cancer.

To address this evidence gap, we investigated the association between depression occurrences following breast cancer diagnosis and death from all causes, non-cancer causes, and cancer-related deaths using a population-based study involving 5-year breast cancer survivors in Korea.

Methods

Data source

The study, conducted as a retrospective cohort study, utilized the National Health Insurance Service (NHIS-2023-1-289) cancer patients cohort database in South Korea, spanning from January 1st, 2006, to December 31st, 2022¹⁸. South Korea’s national healthcare system allows for the utilization of healthcare data from a majority of the population for research purposes. The NHIS has systematically collected health screening information for Koreans, encompassing age, gender, insurance grade (as a substitute for income), residential region details, health examination dates, medical practice history, prescription data, dental treatment status, oriental medicine treatment, and disease history diagnosed with ICD-10 codes¹⁹, which represents the 10th revision of the International Statistical Classification of Diseases. Health checkups are regularly conducted by the NHIS for adults over 40 years old every two years. Utilizing this data allows for determining various factors, including patients’ smoking status, alcohol consumption, weekly physical activity frequency, blood pressure, total cholesterol, body mass index, and more¹⁹.

Study population and design

The NHIS cancer patient’s cohort database covers the years 2006 to 2022. However, we excluded a group of patients who underwent breast cancer treatment between January 1st, 2006, and December 31st, 2006, as our focus was on individuals newly diagnosed with breast cancer. In Fig. 1, we initially identified 63,014 women with a new diagnosis of breast cancer during the period from January 1, 2007, to December 31, 2013, who survived for at least five years. We excluded individuals with missing health examination data (N = 27,471) and those with missing demographic variables (N = 4,256). To establish a depression-free cohort at baseline, we further excluded women with a recorded history of depression prior to their breast cancer diagnosis (N = 253). Male patients (N = 161) were also excluded. The final analytic cohort consisted of 30,873 women who survived at least five years after their diagnosis and had no history of depression before the breast cancer onset. Women were eligible if they were aged ≥ 40 years, newly diagnosed with breast cancer between 2007 and 2013, survived at least five years after diagnosis, and had no history of depression prior to the date of newly diagnosed with breast cancer. In summary, eligible participants were women aged ≥ 40 years who were newly diagnosed with breast cancer between 2007 and 2013, survived at least five years after diagnosis, and had no recorded history of depression prior to the breast cancer diagnosis.

Fig. 1.

Flow diagram of the study subjects.

Key variables

We identified breast cancer using the ICD-10 code C50 and depression using ICD-10 codes F32–F33²⁰. Variables of interest included body mass index, blood pressure (sistolic and diastolic), fasting blood glucose, and total cholesterol. Covariates comprised sex, age, income level, drinking frequency or quantity, physical activity frequency, smoking status, and the Charlson Comorbidity Index (CCI). The frequency of moderate-intensity physical activity was calculated as the sum of weekly activities lasting ≥ 20 min and ≥ 30 min.

In South Korea, the majority of patients with depression are treated in outpatient settings, whereas inpatient admission is generally reserved for cases requiring closer clinical monitoring, such as those with marked functional impairment, suicidality, or other clinically significant presentations. In claims-based research, outpatient diagnostic codes may include provisional or rule-out diagnoses, which can increase the risk of outcome misclassification. Therefore, consistent with methodological recommendations for studies using Korean administrative health data, we restricted the operational definition of depression to inpatient admissions to improve diagnostic specificity and capture clinically meaningful depressive episodes rather than subthreshold conditions or provisional diagnoses²¹.

To further reduce potential misclassification related to very brief or administrative admissions, we required a minimum inpatient stay of ≥ 2 days. This criterion was applied as a pragmatic operational definition to exclude short observational stays that are unlikely to reflect true psychiatric hospitalization, rather than as a direct measure of depression severity. Such approaches, using hospitalization and additional utilization criteria to enhance the validity of case identification, have been commonly adopted in claims-based epidemiologic studies²².

Because individual antidepressant prescription data could not be fully ascertained for all patients, hospitalization for depression was used as a proxy for clinically significant depressive episodes. We did not assume that all hospitalized patients uniformly received antidepressant therapy; rather, this definition reflects an operational choice based on the structure and limitations of the claims database, aiming to enhance diagnostic validity while minimizing misclassification^21,22.

Accordingly, depression was defined as hospitalization for ≥ 2 days with a primary diagnosis of depression (ICD-10 codes F32–F33) occurring during the 5-year survivor period following the initial breast cancer diagnosis. Follow-up for mortality outcomes began after completion of this 5-year survivor period. All-cause death, cancer-cause death, and non-cancer-cause death were defined as deaths occurring from the end of the survivor period through December 31, 2021.

Primary outcomes

This study evaluated three primary outcomes: all-cause mortality, cancer-specific mortality, and non-cancer mortality. Mortality outcomes were ascertained using linked national death registry data. Follow-up for mortality outcomes began at the end of the 5-year survivor period after the initial breast cancer diagnosis and continued until death or December 31, 2021, whichever occurred first.

Statistical analysis

Following the diagnosis of breast cancer, characteristics of patients who survived for at least 5 years were categorized based on their history of depression during the survivor period. Continuous variables exhibiting a normal distribution were presented using standard deviation and average, while dichotomous and categorical data were summarized using frequency and ratio. Comparative analyses of clinical characteristics between breast cancer patients in the non-depression and depression groups involved chi-square tests for categorical and dichotomous variables, and independent sample t-tests for normally distributed continuous variables.

Cox proportional hazards regression models were utilized in the comprehensive analysis, accounting for possible covariates²³. The primary outcomes were all-cause mortality, cancer-specific mortality, and non-cancer mortality. The final results were analyzed after accounting for patients’ personal information (such as sex, age, and income) and health examination data (including smoking habits and frequency, blood pressure, Charlson Comorbidity Index, alcohol frequency and quantity, body mass index [BMI], total cholesterol, fasting serum glucose). The study assessed all-cause mortality, cancer-related mortality, and non-cancer-related mortality, analyzing the incidence of death in the two groups of breast cancer patients over person-years of follow-up. Hazard ratios (HR) with 95% confidence intervals (CI) were calculated to interpret the results²⁴, Hazard ratios (HRs) with 95% confidence intervals (CIs) were reported, and exact p-values are presented without using a dichotomous significance threshold. Subjects with missing values were excluded before the analysis, and SAS Enterprise Guide version 7.1 was utilized.

Results

Baseline characteristics

Table 1 presents the baseline characteristics of 30,371 patients without depression and 502 patients with new-onset depression during the 5-year survival period following a breast cancer diagnosis. All participants were women, with a mean age of 56.5 years in both groups. Based on NHIS premium data, the highest income quartile comprised 11,296 women (37.2%) in the non-depressed group and 169 women (33.7%) in the depressed group. Baseline characteristics were generally similar between groups, with small differences in blood pressure and fasting glucose levels. Regarding lifestyle factors within the two-year period after the 5-year survivorship mark, more than 94% of patients in both groups were non-smokers, and over 88% reported abstaining from weekly alcohol consumption. More than half of patients reported no weekly physical activity, whereas approximately 10–12% in each group engaged in physical activity one to five times per week. The distribution of Charlson Comorbidity Index (CCI) scores indicated that most individuals in both groups had a score of 1 or higher.

Table 1.

Descriptive statistics of the participants in the National health insurance Service.

	Patients who have not been diagnosed with depression since breast cancer diagnosis	New depression diagnostic patients after breast cancer diagnosis	p value
	(n = 30,371)	(n = 502)
Age, years, mean (SD)	56.5 (9.7)	56.5 (9.0)	0.018
Household incomea, n (%)			0.267
First (highest)	11,296 (37.2)	169 (33.7)
Second	7,180 (23.6)	135 (26.9)
Third	5,646 (18.6)	92 (18.3)
Fourth (lowest)	6,249 (20.6)	106 (21.1)
Body mass index, kg/m2	23.5 (3.3)	23.5 (3.1)	0.095
Systolic blood pressure, mmHg	121.1 (15.1)	118.7 (15.2)	0.839
Diastolic blood pressure, mmHg	74.4 (9.8)	74.3 (9.9)	0.590
Total cholesterol, mg/dL	192.5 (36.7)	194.1 (35.6)	0.348
Fasting serum glucose, mg/dL	97.7 (20.2)	97.4 (18.7)	0.020
Cigarette smoking, n (%)			0.004
Never smoker	29,372 (96.7)	473 (94.2)
Past smoker	636 (2.1)	21 (4.2)
Current smoker	363 (1.2)	8 (1.6)
Alcohol consumption, n (%)			0.042
None	27,038 (89.0)	444 (88.5)
1–2 times/week	2,947 (9.7)	51 (10.2)
3–4 times/week	287 (0.9)	7 (1.4)
≥5 times/week	99 (0.3)	0 (0.0)
Physical activity, n (%)			0.844
None	17,295 (57.0)	291 (58.1)
1–2 times/week	4,799 (15.8)	82 (16.4)
3–4 times/week	3,934 (13.0)	63 (12.6)
≥5 times/week	4,326 (14.3)	65 (13.0)
Charlson comorbidity index, n (%)			< 0.0001
0	971 (3.2)	3 (0.6)
1	14,792 (48.7)	164 (32.7)
≥2	14,608 (48.1)	335 (66.7)

Data are presented as median (interquartile range) unless otherwise specified.

^aProxy for socioeconomic status based on the insurance premium of the National Health Insurance Service.

Depression in survivor period and mortality outcomes among 5-year survivors breast cancer patients

During follow-up after the 5-year survivor period, a total of 1,904 deaths occurred. Follow-up began at the end of the 5-year survivor period and continued until death or December 31, 2021. Table 2 presents the primary outcomes of this study, examining the association between new-onset depression during the 5-year survivor period following a breast cancer diagnosis and subsequent mortality outcomes. Multivariable Cox proportional hazards models were applied with sequential adjustments. Model 1 adjusted for age only, and Model 2 additionally included household income. Model 3 included age, income, smoking status (past, current and never), alcohol intake (days per week), and moderate-intensity physical activity (days per week). Model 4 incorporated all patient information, including age, household income, alcohol intake (days per week), smoking status (past, current and never), moderate-intensity physical activity, body mass index, systolic blood pressure, fasting blood glucose, total cholesterol, and Charlson Comorbidity Index (CCI).

Table 2.

HRs of all-cause mortality, cancer-cause mortality and no cancer-cause mortality by diagnosed with depression after newly diagnosed with breast cancer.

			HR (95% CI)
	Event, n	Per 1,000 person-y	Model 1	Model 2	Model 3	Model 4
All-cause mortality
Patients who have not been diagnosed with depression since breast cancer diagnosis (n = 30,371)	1,859	183.4	1.00 (Ref.)	1.00 (Ref.)	1.00 (Ref.)	1.00 (Ref.)
New depression diagnostic patients after breast cancer diagnosis (n = 502)	45	2.8	1.61 (1.20–2.17)	1.61 (1.19–2.16)	1.57 (1.17–2.12)	1.38 (1.03–1.86)
p-value			0.002	0.002	0.003	0.033
Cancer-cause mortality
Patients who have not been diagnosed with depression since breast cancer diagnosis (n = 30,371)	1,146	183.4	1.00 (Ref.)	1.00 (Ref.)	1.00 (Ref.)	1.00 (Ref.)
New depression diagnostic patients after breast cancer diagnosis (n = 502)	26	2.8	1.48 (1.00-2.18)	1.47 (1.00-2.17)	1.45 (0.98–2.14)	1.20 (0.81–1.77)
p-value			0.050	0.051	0.067	0.360
No cancer-cause mortality
Patients who have not been diagnosed with depression since breast cancer diagnosis (n = 30,371)	713	183.4	1.00 (Ref.)	1.00 (Ref.)	1.00 (Ref.)	1.00 (Ref.)
New depression diagnostic patients after breast cancer diagnosis (n = 502)	19	2.8	1.81 (1.15–2.86)	1.80 (1.14–2.84)	1.76 (1.12–2.78)	1.81 (1.14–2.86)
p-value			0.011	0.012	0.015	0.011

CI = confidence interval, HR = hazard ratio, Crude HR, relative hazard ratio.

*Incidence rate, per 1000 person-year.

Model 1: adjusted for age.

Model 2: adjusted for age and household income.

Model 3: adjusted for age, household income, smoking status, alcohol consumption and physical activity.

Model 4: adjusted for age, household income, smoking status, alcohol consumption, physical activity, body mass index, systolic blood pressure, fasting serum glucose, total cholesterol, and Charlson comorbidity index.

In Model 4, new-onset depression during the survivor period was associated with a 38% higher risk of all-cause mortality (aHR 1.38, 95% CI 1.03–1.86, p = 0.033). Although cancer-specific mortality showed a similar trend, the association did not reach statistical significance (aHR 1.20, 95% CI 0.81–1.77, p = 0.360). In contrast, non-cancer mortality demonstrated a strong and statistically significant association, with depressed patients experiencing an 81% higher risk compared with those without depression (aHR 1.81, 95% CI 1.14–2.86, p = 0.011).

Subgroup analysis

For the three subgroup analyses involving mortality (all-cause, cancer-cause, and non-cancer-cause), the study divided participants into groups based on the CCI (< 2, ≥ 2), age (< 65, ≥ 65), income (< 50%, ≥ 50%), physical activity (exercise or no exercise), drink habits (non-drinker or drinker), and smoking status (non-smoker or smoker). The same adjustment variables as those in Model 4 were applied.

In summary, the Subgroup analysis in Table 3 showed that, for all-cause mortality, depression during the survivor period was associated with significantly elevated risk among patients aged ≥ 65 years (aHR 1.96, 95% CI 1.27–3.03, p = 0.002), those with lower income (aHR 1.57, 95% CI 1.02–2.43, p = 0.041), and those who did not engage in physical activity (aHR 1.59, 95% CI 1.12–2.25, p = 0.010).

Table 3.

Subgroup analysis of the association between diagnosed with depression during the survivor period among breast cancer and risk of all-cause mortality according to age, income, physical activity, drink habits, and smoking status.

	Patients who have not been diagnosed with depression since breast cancer diagnosis	New depression diagnostic patients after breast cancer diagnosis		Patients who have not been diagnosed with depression since breast cancer diagnosis	New depression diagnostic patients after breast cancer diagnosis
Age < 65	24,079	421	Age ≥ 65	6,292	81
Events, n	1,053	24	Events, n	806	21
Per 1,000 person-y	146.8	2.4	Per 1,000 person-y	36.6	0.4
aHR (95% CI)	1.00 (Ref.)	1.13 (0.75–1.69)	aHR (95% CI)	1.00 (Ref.)	1.96 (1.27–3.03)
p-value		0.556			0.002
Income higher		18,476	304	Income lower	11,895	198
Events, n		1,169	24	Events, n	690	21
Per 1,000 person-y		112.2	1.7	Per 1,000 person-y	71.2	1.2
aHR (95% CI)		1.00 (Ref.)	1.23 (0.82–1.85)	aHR (95% CI)	1.00 (Ref.)	1.57 (1.02–2.43)
p-value			0.315			0.041
Physical activity higher		13,059	210	No physical activity	17,295	291
Events, n		683	12	Events, n	1,176	33
Per 1,000 person-y		78.4	1.2	Per 1,000 person-y	104.9	1.7
aHR (95% CI)		1.00 (Ref.)	1.01 (0.57–1.78)	aHR (95% CI)	1.00 (Ref.)	1.59 (1.12–2.25)
p-value			0.972			0.010
No drink		27,038	444	Drink higher	3,333	584
Events, n		1,749	41	Events, n	110	41
Per 1,000 person-y		163.3	2.5	Per 1,000 person-y	20.1	0.3
aHR (95% CI)		1.00 (Ref.)	1.36 (0.99–1.85)	aHR (95% CI)	1.00 (Ref.)	1.87 (0.68–5.13)
p-value			0.057			0.225
No smoking		29,372	473	Smoker	999	29
Events, n		1,790	41	Events, n	69	4
Per 1,000 person-y		177.6	2.7	Per 1,000 person-y	2.9	0.2
aHR (95% CI)		1.00 (Ref.)	1.35 (0.99–1.83)	aHR (95% CI)	1.00 (Ref.)	2.28 (0.82–6.37)
p-value			0.058			0.113
Charlson comorbidity index < 2		15,763	167	Charlson comorbidity index ≥ 2	14,608	335
Events, n		475	8	Events, n	1,384	37
Per 1,000 person-y		94.9	0.9	Per 1,000 person-y	88.5	1.9
aHR (95% CI)		1.00 (Ref.)	1.83 (0.91–3.69)	aHR (95% CI)	1.00 (Ref.)	1.33 (0.96–1.84)
p-value			0.089			0.087

CI = confidence interval, HR = hazard ratio, Crude HR, relative hazard ratio, aHR = adjusted hazard ratio.

Charlson comorbidity index information is the set at the time of new breast cancer diagnosis.

Hazard ratios calculated by Cox proportional hazards regression analysis: adjusted for adjusted for age, household income, smoking status, alcohol consumption, physical activity, body mass index, systolic blood pressure, fasting serum glucose, total cholesterol, and Charlson comorbidity index.

For cancer-cause mortality, a significant association with depression was observed only among patients aged ≥ 65 years (aHR 1.48, 95% CI 1.00–2.18, p ≈ 0.050), whereas no other subgroup demonstrated statistically significant associations.

For non-cancer-cause mortality, depression was associated with increased risk across several subgroups, including patients with CCI < 2 (aHR 1.83, 95% CI 0.91–3.69, p = 0.089, borderline), those aged ≥ 65 years (aHR 1.96, 95% CI 1.27–3.03, p = 0.002), those with lower income (aHR 1.57, 95% CI 1.02–2.43, p = 0.041), and those without physical activity (aHR 1.59, 95% CI 1.12–2.25, p = 0.010). Although higher alcohol consumption also showed an increased point estimate (aHR 1.87, 95% CI 0.68–5.13), the association was not statistically significant (p = 0.225).

Subgroup analysis examining the association between depression during the survivor period and the risk of cancer and non-cancer mortality by age, income, physical activity, drinking, and smoking status are provided in Tables S1 and S2.

Discussion

This longitudinal study conducted on a population basis demonstrated a notable association between post-breast cancer diagnosis depression and extended-term mortality among 5-year cancer survivors (CS). Notably, this research is the initial one to illustrate that depression after a breast cancer diagnosis is correlated with a heightened probability of mortality, especially non-cancer-related mortality, among 5-year CS in Korea.

Our finding of a significant association between post-breast cancer depression and increased mortality among 5-year cancer survivors is in line with previous evidence on the adverse impact of depression on cancer prognosis. In a meta-analysis of 31 prospective cohorts, mortality was 25% higher among cancer patients experiencing depressive symptoms and a 39% increased mortality rate in individuals diagnosed with major depression, even after accounting for prognostic factors²⁵. However, these meta-analyses faced limitations due to significant heterogeneity stemming from different cancer types²⁶. Additionally, considering breast cancer’s hormone-dependent nature, its reaction to mental health conditions might differ from that of other cancers⁴. A systematic analysis of 17 studies involving 282,203 breast cancer patients revealed that depression was linked to cancer recurrence(pooled risk ratio [RR] 1.24, 95% CI 1.07–1.43), all-cause mortality( RR 1.30, 95% CI 1.23–1.36), and cancer-specific mortality(RR 1.29, 95% CI 1.11–1.49)⁴. However, most of these studies were conducted in Western countries, and few studies have examined the effect of depression on non-cancer-related mortality in breast cancer survivors. Our study is the first to demonstrate that depression after a breast cancer diagnosis is associated with a higher risk of non-cancer-related mortality among 5-year cancer survivors in Korea.

Currently, evidence on the association between depression and breast cancer mortality is limited to observational and genetic studies. One study in the UK Biobank cohort found that post-diagnostic use of selective serotonin reuptake inhibitors (SSRIs), a common antidepressant medication, was linked to a 27% higher breast cancer mortality risk, although the estimate was markedly reduced in analyses limited to patients with prior depression or in comparisons with other antidepressant classes²⁷. This study suggested a possible causal relationship between depression and breast cancer mortality, but also highlighted the potential confounding by indication. A randomized trial involving 125 women with metastatic breast cancer (MBC) revealed cortisol dysregulation in MBC patients compared to controls, predicting shorter survival⁷. This study suggests that irregular cortisol patterns in both MBC and depression may indicate an inadequate response to inflammation related to cancer. By disrupting cortisol regulation, tumors using inflammatory mediators may cause resistance to glucocorticoids, but addressing depression could potentially break this cycle, relieving related symptoms²⁸. While these studies have demonstrated a biological link between depression and breast cancer mortality, the evidence base is still dominated by cohorts from high-income North American and European populations^4,29. In these settings, routine screening and referral pathways for depression and anxiety are increasingly incorporated into oncology care, whereas psycho-oncology infrastructure and specialist mental health services remain limited or fragmented in many non-Western and low- and middle-income contexts^14,30. In East Asian settings, including Korea and China, strong cancer- and mental-health-related stigma and family-centred concerns, such as fears of burdening or bringing shame to one’s family, have been shown to increase depressive symptoms and deter help-seeking^31,32. Previous work using Korean National Health Insurance Service data has already shown that comorbid depressive and anxiety disorders are linked to higher all-cause mortality among patients with breast cancer, and that antidepressant treatment may attenuate this excess risk³³. Building on this evidence, our study specifically focused on women who had survived at least 5 years after their breast cancer diagnosis and examined how incident depression during survivorship was differentially associated across all-cause, cancer-specific, and non-cancer mortality outcomes.

Depression may increase non-cancer-related deaths after breast cancer diagnosis through both biological and behavioral pathways, including metabolic and immune-inflammatory alterations, autonomic imbalance, and dysregulation of the hypothalamic–pituitary–adrenal (HPA) axis dysregulation, as well as poorer adherence to treatment and delayed recognition of physical symptoms that complicate the management of comorbid conditions^34–36. In Korea, where cancer is the leading cause of death, depression and anxiety are common among breast cancer patients and survivors but often underdiagnosed and undertreated³⁷. Although the Korean health system offers universal coverage for cancer treatment, the accessibility and quality of mental health care and support systems for health and social welfare among cancer survivors remain inadequate³⁸. Our study of the Korean cohort further extends evidence on the long-term mortality impact of post-breast cancer diagnosis depression among 5-year cancer survivors, highlighting the need for adequate recognition and treatment of clinically significant depression among breast cancer survivors, especially in Asian populations where depression may be overlooked and undertreated³⁹. The strengths of our study stand out, which encompassed analyses of a large and representative study population linked to the Korean NHIS and death registry data to capture not only the diagnosis of depression and health service utilization but also the cause and date of death. Moreover, we addressed a diverse set of confounders for sociodemographic factors, health status, health behavior, comorbidities, and clinical characteristics for adjustment and Subgroup analysis to minimize potential bias.

Limitations

Our study has several limitations. First, we did not have information on the type and duration of antidepressant medication used by breast cancer survivors with depression. Instead, we used hospital-diagnosed depression as a proxy for clinically significant depressive episodes that typically involve pharmacologic treatment, although individual antidepressant prescriptions could not be fully ascertained from the database. We did not assume that all hospitalized patients uniformly received antidepressants; rather, hospitalization for depression was used as a pragmatic indicator of clinically significant depressive episodes in the context of the available claims data. Nonetheless, previous studies have shown that antidepressant medication can reduce the risk of mortality in breast cancer patients with depression³³. Second, as this was a retrospective cohort study based on administrative claims and health screening data, residual confounding, misclassification of depression and causes of death, and selection bias cannot be fully excluded; therefore, our findings should be interpreted with caution, and no causal inferences can be drawn. Third, we did not account for the severity of the disease, duration since diagnosis, and the specific treatment undergone by the women with breast cancer. These variables also affect the depression levels of these patients⁴⁰. Thus, the validity of our findings is limited by these confounding factors. However, we tried to reduce the effect of deaths caused by this factor by selecting only 5-year survivors for our study. A meta-analysis of 76 prospective studies with 105 samples showed that depression’s impact on mortality is independent of the disease stage in cancer patients⁴¹. Fourth, it should be noted that our study measured depression only once at the baseline. Therefore, whether the breast cancer survivors in our cohort had persistent or recurrent depression during the follow-up period remains unknown. Nonetheless, our results showed a significant association between depression and mortality, especially non-cancer-related mortality, in breast cancer survivors. This finding is consistent with previous studies that reported a higher risk of death from cardiovascular diseases, respiratory diseases, and suicide in breast cancer patients with depression.

Conclusions

In conclusion, in this large nationwide Korean cohort of 5-year breast cancer survivors, incident hospital-diagnosed depression after breast cancer diagnosis was associated with increased long-term risks of all-cause and non-cancer mortality, highlighting the importance of appropriate recognition and treatment of clinically significant depression as part of survivorship care, particularly in Asian settings where depression may be overlooked and undertreated. As this was a retrospective observational cohort study based on claims and screening data, our findings should not be interpreted as causal, and further research in diverse populations and across the spectrum of depressive symptom severity is warranted.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

S.K.L. and S.P. wrote the main manuscript text. S.K.L. and S.P. conducted the statistical analysis. S.K.L. and S.P. obtained funding and provided administrative, technical, or material support. S.K.L., S.P., and S.M.P. were responsible for the acquisition, analysis, or interpretation of data. S.K.L. and S.P. critically reviewed the manuscript for important intellectual content. S.M.P. supervised the study and had full access to all of the study’s data, ensuring its integrity and the accuracy of the data analysis. All authors reviewed the manuscript. S.K.L. and S.P. contributed equally.

Funding

Sangwoo Park received a BK21 FOUR education program scholarship. It was offered by the National Research Foundation of Korea(2021R1F1A1063346). The Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (RS-2023-00244084).

Data availability

Data from the Korea NHIS support the study findings, but access to them is restricted. They are used under license for this current study and are thus not publicly accessible. However, the authors can obtain the data upon reasonable request and with approval from the National Health Insurance System (https://nhiss.nhis.or.kr/bd/ab/bdaba000eng.do).

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

This study was reviewed by the Institutional Review Board of the Seoul National University Hospital Biomedical Research Institute and was granted an exemption from ethical approval (IRB No.: E-2204-038-1314). The requirement for informed consent was waived by the Institutional Review Board because the study used de-identified, routinely collected administrative data. All methods were performed in accordance with the relevant guidelines and regulations, including the principles of the Declaration of Helsinki.