Interacting and joint effects of frailty and depressive symptoms in relation to risk of cardio-oncology comorbidity in older adults from three prospective cohorts Authors: Chenxi Li, Yan Zhang, Donghui Li, Jie Gao, Jianhua Zhu, Jiahong Liu Published: 2026 ABSTRACT BackgroundWhile cardiovascular disease and cancer share common risk factors and exhibit complex bidirectional interplay, evidence regarding the interacting and joint effects of frailty and depressive symptoms on the risk of cardio-oncology comorbidity in older adults remains scarce.ObjectivesThis study aimed to examine the association between frailty and depressive symptoms with the risk of cardio-oncology comorbidity, and to explore their interacting and joint effects.MethodsThis study harmonized data from three prospective, nationally representative cohort studies: China Health and Retirement Longitudinal Study (CHARLS), the English Longitudinal Study of Ageing (ELSA), and the Health and Retirement Study (HRS). The final analytical sample included 10,937 participants aged ≥ 65 years without baseline cardio-oncology comorbidity. Frailty and depressive symptoms were assessed at baseline. Cardio-oncology comorbidity was ascertained based on self-reported physician diagnoses of cardiovascular disease and cancer. Cox proportional hazards models were utilized to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). Multiplicative and additive interaction analyses were conducted, followed by joint association evaluations.ResultsFrailty (CHARLS HR = 1.71; ELSA HR = 1.41; HRS HR = 1.38) and depressive symptoms (CHARLS HR = 1.27; ELSA HR = 1.20; HRS HR = 1.18) were independently associated with significantly increased cardio-oncology comorbidity risk. While interaction analyses demonstrated the absence of significant multiplicative or additive interactions, the joint effect analysis revealed that cardio-oncology comorbidity risk was elevated among individuals with both frailty and depressive symptoms by 80% in CHARLS, 57% in ELSA, and 47% in HRS compared to those with neither condition.ConclusionsConcurrent frailty and depressive symptoms constitute the highest-risk phenotype for cardio-oncology comorbidity in community-dwelling older adults. These findings suggest that combined assessment of frailty and depressive symptoms may improve baseline risk stratification of cardio-oncology comorbidity.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12877-026-07834-8. Introduction Cardiovascular disease (CVD) and cancer rank as the first and second leading causes of death worldwide, respectively [1]. Epidemiological evidence further indicates that cancer patients face a significantly elevated risk of major atherosclerotic events, such as myocardial infarction and stroke [2]. Moreover, the coexistence of cardiovascular or metabolic comorbidities substantially increases both all-cause and cancer-specific mortality in this population [3]. The pathophysiological mechanisms underlying CVD and cancer exhibit complex bidirectional interplay. On one hand, they share several common risk factors, including chronic inflammation, and metabolic dysregulation [4]. On the other hand, a profound “bidirectional pathological crosstalk” exists. For instance, cardiometabolic dysfunction may promote tumor progression through mechanisms like chronic inflammation and cellular senescence [5, 6]. Conversely, the tumor microenvironment can influence cardiac structural and functional remodeling via immune modulation [7]. In terms of clinical management, the first European Society of Cardiology guideline on cardio-oncology emphasizes the need for standardized definitions, risk stratification, and a patient-centered, multidisciplinary integrated care pathway [8]. It is noteworthy that traditional risk factors including age, smoking, and alcohol use do not fully account for the high risk of CVD-cancer comorbidity. Consequently, investigating risk mechanisms through the novel lenses of “physical frailty” and “psychosomatic integration” has attracted growing research interest. Longitudinal analyses from multinational cohorts including the China Health and Retirement Longitudinal Study (CHARLS), the English Longitudinal Study of Ageing (ELSA), and the Health and Retirement Study (HRS), indicate that frailty and its dynamic progression are independent predictors of CVD [9], with depressive symptoms playing a partial mediating role in this association [10]. Concurrently, frailty has been significantly linked to an elevated risk of cancer. For instance, the China Kadoorie Biobank study (N = 484,225) found that a higher frailty status was associated with an increased risk of overall cancer, and a healthy lifestyle substantially attenuated this risk [11]. Analyses from HRS cohort further revealed that even minimal frailty in long-term cancer survivors is associated with a significantly elevated risk of developing CVD and type 2 diabetes, suggesting that frailty may serve as a key pathophysiological node linking cardiac and oncological conditions [12]. Separately, depressive symptoms are an established cardiovascular risk factor, associated with increased risks of myocardial infarction, stroke, heart failure, and CVD mortality [13]. The prevalence of depression is also markedly higher among cancer patients than in the general population [14]. Clinically, frailty and depressive symptoms frequently co-occur in older adults with chronic conditions, exhibit a significant bidirectional association, and advanced age and depressive symptoms itself are independent risk factors for frailty progression [15]. Mechanistically, existing evidence suggests that frailty and depressive symptoms may jointly elevate the risk of cardio-oncology comorbidity through shared biological pathways, such as chronic inflammation [16, 17]. Nevertheless, despite these established links, evidence regarding the interacting and joint effects of frailty and depressive symptoms on the risk of cardio-oncology comorbidity in older adults remains scarce, and their explicit epidemiological patterns and underlying biological mechanisms await further elucidation. Therefore, using data from three prospective aging cohorts including CHARLS, ELSA, and HRS, this study aimed to examine the associations of frailty and depressive symptoms, assessed as distinct but related baseline risk markers, with the risk of cardio-oncology comorbidity. Additionally, we explored the interaction and joint effects of frailty and depressive symptoms in relation to cardio-oncology comorbidity risk. Findings from this study may help clarify whether the combined assessment of physical vulnerability and psychological vulnerability can improve baseline risk stratification for cardio-oncology comorbidity in older adults. Methods Study design and population This study harmonized data from three nationally representative aging cohorts: CHARLS from China, ELSA from England, and HRS from the United States. Baseline was defined as wave 1 (2011) for CHARLS, wave 2 (2004–2005) for ELSA, and wave 8 (2006) for HRS, with follow-up through wave 4 (2018), wave 9 (2018–2019), and wave 15 (2020), respectively. All studies received ethical approval from their respective institutional review boards (Peking University for CHARLS, London Multi-Centre Research Ethics Committee for ELSA, and University of Michigan for HRS), and all participants provided written informed consent. From the initial pooled sample of 20,655 adults aged ≥ 65 years, we excluded 3,581 participants with cardio-oncology comorbidity, 5,448 participants with missing baseline frailty or depressive symptoms measurements, and 689 participants without cardio-oncology comorbidity follow-up information. The final analytical sample included 10,937 participants (Figure S1). Measures Assessment of Frailty: Frailty was assessed using the 27-item Frailty Index (FI), which aggregates age-related health deficits including chronic diseases (excluding heart disease, stroke and cancer), self-reported health, functional limitations, and cognition. All items except cognition were dichotomized (1 = deficit, 0 = no deficit). Cognition was retained as one component because cognitive impairment is a commonly recognized age-related deficit and contributes to multidimensional vulnerability in older adults. Cognition was treated as a continuous variable (range 0–1, higher scores indicating poorer function) [10, 18]. The FI score was calculated as the sum of deficits divided by the total number of items (range 0-100), with higher values indicating greater frailty. Participants were categorized as non-frailty (FI < 25) or frailty (FI ≥ 25) (Table S8). Depressive symptoms were not included in the frailty index but were assessed separately to avoid conceptual overlap and to evaluate the independent and joint contributions of physical/functional vulnerability and psychological vulnerability to cardio-oncology comorbidity risk. Assessment of Depressive Symptoms: The Center for Epidemiologic Studies Depression Scale (CES-D) and its short forms have been widely used to assess depressive symptoms in population-based studies of older adults [19]. CHARLS used the CESD-10 (range 0–30), while HRS and ELSA used the CESD-8 (range 0–8) [19, 20]. Thresholds for depressive symptoms were ≥ 10 for CHARLS and ≥ 3 for HRS/ELSA (Table S9). Because instruments and cut-offs were not identical across cohorts, differences in measurement may partly explain variation in effect magnitude, although the direction of association was consistent. Assessment of cardio-oncology comorbidity: The heart disease was ascertained based on self-reported physician diagnosed angina, heart attack, congestive heart failure or other heart problems. The stroke was ascertained based on self-reported physician diagnosis. The participants were asked whether they had ever been told by a doctor that they had cancer or a malignant tumor (excluding non-melanoma skin cancer). Participants who answered ‘yes’ were considered cancer survivors. Incident cardio-oncology comorbidity was defined as the occurrence of both cancer and cardiovascular disease (heart disease or stroke) during follow-up. For participants who developed the two conditions at different times, the event date was defined as the date of the second diagnosis, when coexistence was first established. During sample selection, only participants without cardio-oncology comorbidity at baseline were retained. Survival time was calculated as the year of disease onset minus the baseline survey year. For participants without cardio-oncology comorbidity during the study period, survival time was defined as the last valid follow-up survey or study cutoff date minus the baseline year [21–23]. Assessment of covariates: Based on previous research, the covariates selected for this study included demographic factors (sex and age), socioeconomic factors (education), living conditions (marital status), behavioral habits (smoking, drinking). For consistency across studies, marital status was categorized as married/partnered or other. Education was classified as below high school, high school, or college/above. Smoking and drinking status were each dichotomized (never vs. ever). Physical activity level was categorized as vigorous, moderate, or light. Statistical analyses Continuous variables are presented as means (standard deviations) if normally distributed, or as medians (interquartile ranges) otherwise. Categorical variables are described as frequencies (percentages). Between-group comparisons at baseline were performed with chi-square tests, analysis of variance, or Kruskal-Wallis rank sum tests, as appropriate. Missing data, assumed to be missing at random, were imputed using multiple imputation by chained equations based on baseline characteristics. Cox proportional hazards models were used to estimate hazard ratios (HRs) with 95% confidence intervals (CIs) for outcomes associated with frailty and depressive symptoms. The proportional hazards assumption was evaluated separately for CHARLS, ELSA, and HRS using Schoenfeld residuals; no meaningful violation of the assumption was observed. The associations were also analyzed in the different age, gender and BMI subgroups. Interaction analyses included multiplicative interaction (assessed by including a product term of frailty and depressive symptoms; HR ≠ 1 indicating interaction) and additive interaction (quantified using the relative excess risk due to interaction [RERI], attributable proportion [AP], and synergy index [SI]). Values of RERI = 0, AP = 0, and SI = 1 indicate no interaction; RERI > 0, AP > 0, and SI > 1 suggest synergy; RERI < 0, AP < 0, and SI < 1 suggest antagonism [24, 25]. Joint associations were assessed by categorizing participants into four groups based on frailty-depressive symptoms combinations, with HRs referenced to the group with no frailty and no depressive symptoms [26]. All models were adjusted for sex, age, education, marital status, smoking, drinking, and BMI. All statistical analyses were performed using R software (version 4.1.2). A two-sided p value of 0.05 was used to define statistical significance. Results Characteristics of participants According to inclusion and exclusion criteria, a total of 2993 participants from CHARLS, 2627 from ELSA, and 5317 from HRS were included in this study. Baseline characteristics of these participants are presented in Table 1. In these three cohort studies, compared with participants without cardio-oncology comorbidity during follow-up, those with cardio-oncology comorbidity incidence exhibited elevated BMI levels (e.g., 28.3 vs. 27.6 kg/m2 in HRS), aligning with a higher prevalence of females (55.0% vs. 47.8% in CHARLS). Meanwhile, we observed that reduced physical activity (e.g., 19.7% vs. 26.0% vigorous physical activity in CHARLS) and higher prevalence of ever drinking (e.g., 51.4% vs. 46.3% in HRS) compared to those who without cardio-oncology comorbidity in the ELSA and HRS. Baseline characteristics of included and excluded participants were also described (Table S7). Table 1Baseline characteristics of participants by cardio-oncology comorbidity status during studyCHARLS (N = 2993)ELSA (2627)HRS (5317)VariablesNo COC incidenceCOC incidenceNo COC incidenceCOC incidenceNo COC incidenceCOC incidenceNumbers21378561594103328952422Age, mean (SD), years71.9 (5.7)71.2 (5.1)73.3 (6.4)73.0 (5.9)73.2 (7.0)72.9 (6.4)BMI, mean (SD), kg/m222.1 (4.0)23.0 (4.1)27.5 (4.5)27.8 (4.6)27.6 (5.3)28.3 (5.4)Female, n (%)1022 (47.8%)471 (55.0%)940 (59.0%)564 (54.6%)1937 (66.9%)1448 (59.8%)Ever smokers, n (%)909 (42.5%)339 (39.6%)192 (12.0%)106 (10.3%)298 (10.3%)250 (10.3%)Ever drinkers, n (%)845 (39.5%)330 (38.6%)1368 (85.8%)904 (87.5%)1339 (46.3%)1245 (51.4%)Marital status, n (%) Married or partnered1509 (70.6%)609 (71.1%)922 (57.8%)644 (62.3%)2797 (96.6%)2369 (97.8%) Other marital status628 (29.4%)247 (28.9%)672 (42.2%)389 (37.7%)98 (3.4%)53 (2.2%)Education, n (%) Below high school2014 (94.2%)776 (90.7%)929 (58.3%)557 (53.9%)895 (30.9%)622 (25.7%) High school86 (4.0%)57 (6.7%)250 (15.7%)177 (17.1%)940 (32.5%)853 (35.2%) College or above37 (1.7%)23 (2.7%)415 (26.0%)299 (28.9%)1060 (36.6%)947 (39.1%)Physical activity, n (%) Light997 (46.7%)390 (45.6%)278 (17.4%)173 (16.7%)551 (19.0%)407 (16.8%) Moderate584 (27.3%)297 (34.7%)781 (49.0%)467 (45.2%)1349 (46.6%)1073 (44.3%) Vigorous556 (26.0%)169 (19.7%)535 (33.6%)393 (38.0%)995 (34.4%)942 (38.9%)Frailty index, mean (SD)23.3 (15.0)27.4 (15.7)15.1 (12.5)16.2 (13.0)17.1 (12.4)17.8 (12.3)Depressive symptoms (CES-D), mean (SD)8.5 (6.2)9.5 (6.8)1.4 (1.8)1.5 (1.8)1.3 (1.8)1.3 (1.8)CHARLS China Health and Retirement Longitudinal Study, ELSA English Longitudinal Study of Ageing, HRS Health and Retirement Study, COC cardio-oncology comorbidity Table 1Baseline characteristics of participants by cardio-oncology comorbidity status during studyCHARLS (N = 2993)ELSA (2627)HRS (5317)VariablesNo COC incidenceCOC incidenceNo COC incidenceCOC incidenceNo COC incidenceCOC incidenceNumbers21378561594103328952422Age, mean (SD), years71.9 (5.7)71.2 (5.1)73.3 (6.4)73.0 (5.9)73.2 (7.0)72.9 (6.4)BMI, mean (SD), kg/m222.1 (4.0)23.0 (4.1)27.5 (4.5)27.8 (4.6)27.6 (5.3)28.3 (5.4)Female, n (%)1022 (47.8%)471 (55.0%)940 (59.0%)564 (54.6%)1937 (66.9%)1448 (59.8%)Ever smokers, n (%)909 (42.5%)339 (39.6%)192 (12.0%)106 (10.3%)298 (10.3%)250 (10.3%)Ever drinkers, n (%)845 (39.5%)330 (38.6%)1368 (85.8%)904 (87.5%)1339 (46.3%)1245 (51.4%)Marital status, n (%) Married or partnered1509 (70.6%)609 (71.1%)922 (57.8%)644 (62.3%)2797 (96.6%)2369 (97.8%) Other marital status628 (29.4%)247 (28.9%)672 (42.2%)389 (37.7%)98 (3.4%)53 (2.2%)Education, n (%) Below high school2014 (94.2%)776 (90.7%)929 (58.3%)557 (53.9%)895 (30.9%)622 (25.7%) High school86 (4.0%)57 (6.7%)250 (15.7%)177 (17.1%)940 (32.5%)853 (35.2%) College or above37 (1.7%)23 (2.7%)415 (26.0%)299 (28.9%)1060 (36.6%)947 (39.1%)Physical activity, n (%) Light997 (46.7%)390 (45.6%)278 (17.4%)173 (16.7%)551 (19.0%)407 (16.8%) Moderate584 (27.3%)297 (34.7%)781 (49.0%)467 (45.2%)1349 (46.6%)1073 (44.3%) Vigorous556 (26.0%)169 (19.7%)535 (33.6%)393 (38.0%)995 (34.4%)942 (38.9%)Frailty index, mean (SD)23.3 (15.0)27.4 (15.7)15.1 (12.5)16.2 (13.0)17.1 (12.4)17.8 (12.3)Depressive symptoms (CES-D), mean (SD)8.5 (6.2)9.5 (6.8)1.4 (1.8)1.5 (1.8)1.3 (1.8)1.3 (1.8)CHARLS China Health and Retirement Longitudinal Study, ELSA English Longitudinal Study of Ageing, HRS Health and Retirement Study, COC cardio-oncology comorbidity Baseline characteristics of participants by cardio-oncology comorbidity status during study CHARLS China Health and Retirement Longitudinal Study, ELSA English Longitudinal Study of Ageing, HRS Health and Retirement Study, COC cardio-oncology comorbidity Associations of frailty and depressive symptoms with incident cardio-oncology comorbidity Compared to participants without frailty, those with frailty showed significantly increased cardio-oncology comorbidity risk: CHARLS (HR = 1.71, 95% CI 1.48, 1.97), ELSA (HR = 1.41, 95% CI 1.19, 1.67), HRS (HR = 1.38, 95% CI 1.24, 1.53) after full covariate adjustment (Table 2). Similarly, participants with depressive symptoms had higher cardio-oncology comorbidity risk than participants without depressive symptoms (CHARLS: HR = 1.27, 95% CI 1.10, 1.46; ELSA: HR = 1.20, 95% CI 1.00, 1.44; HRS: HR = 1.18, 95% CI 1.04, 1.34) (Table 2). Table 2Associations of frailty and depressive symptoms with incident cardio-oncology comorbidityOutcomeCHARLSELSAHRSHR (95%CI)P-valueHR (95%CI)P-valueHR (95%CI)P-valueFrailty Model 11.66 (1.46, 1.90)< 0.0011.54 (1.32, 1.79)< 0.0011.40 (1.27, 1.54)< 0.001 Model 21.63 (1.42, 1.87)< 0.0011.50 (1.29, 1.75)< 0.0011.39 (1.26, 1.53)< 0.001 Model 31.72 (1.49, 1.97)< 0.0011.50 (1.28, 1.76)< 0.0011.43 (1.30, 1.58)< 0.001 Model 41.71 (1.48, 1.97)< 0.0011.41 (1.19, 1.67)< 0.0011.38 (1.24, 1.53)< 0.001Depressive symptoms Model 11.24 (1.08, 1.42)0.0021.24 (1.04, 1.48)0.0141.14 (1.01, 1.29)0.041 Model 21.20 (1.04, 1.38)0.0121.25 (1.05, 1.49)0.0141.18 (1.05, 1.34)0.008 Model 31.24 (1.07, 1.43)0.0031.26 (1.05, 1.50)0.0131.21 (1.07, 1.37)0.003 Model 41.27 (1.10, 1.46)0.0011.20 (1.00, 1.44)0.0561.18 (1.04, 1.34)0.011CHARLS China Health and Retirement Longitudinal Study, ELSA English Longitudinal Study of Ageing, HRS Health and Retirement Study, BMI body mass indexModel 1: UnadjustedModel 2: Adjusted for age and sexModel 3: Adjusted for model 2 plus education, marital status, smoking status, drinking statusModel 4: Adjusted for model 3 plus BMI Table 2Associations of frailty and depressive symptoms with incident cardio-oncology comorbidityOutcomeCHARLSELSAHRSHR (95%CI)P-valueHR (95%CI)P-valueHR (95%CI)P-valueFrailty Model 11.66 (1.46, 1.90)< 0.0011.54 (1.32, 1.79)< 0.0011.40 (1.27, 1.54)< 0.001 Model 21.63 (1.42, 1.87)< 0.0011.50 (1.29, 1.75)< 0.0011.39 (1.26, 1.53)< 0.001 Model 31.72 (1.49, 1.97)< 0.0011.50 (1.28, 1.76)< 0.0011.43 (1.30, 1.58)< 0.001 Model 41.71 (1.48, 1.97)< 0.0011.41 (1.19, 1.67)< 0.0011.38 (1.24, 1.53)< 0.001Depressive symptoms Model 11.24 (1.08, 1.42)0.0021.24 (1.04, 1.48)0.0141.14 (1.01, 1.29)0.041 Model 21.20 (1.04, 1.38)0.0121.25 (1.05, 1.49)0.0141.18 (1.05, 1.34)0.008 Model 31.24 (1.07, 1.43)0.0031.26 (1.05, 1.50)0.0131.21 (1.07, 1.37)0.003 Model 41.27 (1.10, 1.46)0.0011.20 (1.00, 1.44)0.0561.18 (1.04, 1.34)0.011CHARLS China Health and Retirement Longitudinal Study, ELSA English Longitudinal Study of Ageing, HRS Health and Retirement Study, BMI body mass indexModel 1: UnadjustedModel 2: Adjusted for age and sexModel 3: Adjusted for model 2 plus education, marital status, smoking status, drinking statusModel 4: Adjusted for model 3 plus BMI Associations of frailty and depressive symptoms with incident cardio-oncology comorbidity CHARLS China Health and Retirement Longitudinal Study, ELSA English Longitudinal Study of Ageing, HRS Health and Retirement Study, BMI body mass index Model 1: Unadjusted Model 2: Adjusted for age and sex Model 3: Adjusted for model 2 plus education, marital status, smoking status, drinking status Model 4: Adjusted for model 3 plus BMI Interaction between the frailty and depressive symptoms on cardio-oncology comorbidity Interaction analyses were conducted to evaluate potential additive and multiplicative effects between frailty and depressive symptoms on cardio-oncology comorbidity risk (Table 3). After full covariate adjustment, all four models consistently showed that the 95% CIs of RERI and AP spanned 0, whereas SI and the multiplicative effect encompassed 1, demonstrating absence of significant interactions on either scale between all frailty and depressive symptoms. Table 3Interaction between the frailty and depressive symptoms on cardio-oncology comorbidityDatasetsInteractive effects (95% CI)Multiplicative scaleRERIAPSICHARLS Model 11.19 (0.88, 1.61)0.25 (-0.16, 0.61)0.14 (-0.10, 0.33)1.52 (0.71, 3.23) Model 21.19 (0.88, 1.61)0.23 (-0.17, 0.59)0.14 (-0.11, 0.33)1.52 (0.68, 3.40) Model 31.12 (0.83, 1.52)0.18 (-0.25, 0.56)0.10 (-0.15, 0.30)1.3 (0.68, 2.48) Model 41.11 (0.82, 1.51)0.19 (-0.24, 0.57)0.11 (-0.14, 0.30)1.31 (0.7, 2.46)ELSA Model 11.02 (0.7, 1.48)0.08 (-0.45, 0.60)0.05 (-0.34, 0.29)1.13 (0.48, 2.66) Model 21.07 (0.74, 1.56)0.15 (-0.37, 0.68)0.09 (-0.28, 0.32)1.29 (0.53, 3.16) Model 31.06 (0.73, 1.54)0.12 (-0.39, 0.64)0.08 (-0.31, 0.32)1.26 (0.47, 3.35) Model 41.06 (0.73, 1.54)0.12 (-0.39, 0.63)0.08 (-0.31, 0.32)1.26 (0.46, 3.45)HRS Model 10.99 (0.76, 1.28)-0.01 (-0.33, 0.3)-0.01 (-0.27, 0.18)0.97 (0.45, 2.11) Model 20.99 (0.77, 1.28)0.01 (-0.32, 0.34)0.01 (-0.25, 0.20)1.03 (0.49, 2.15) Model 30.99 (0.77, 1.28)0.02 (-0.31, 0.36)0.02 (-0.24, 0.20)1.05 (0.53, 2.07) Model 41.00 (0.77, 1.29)0.03 (-0.3, 0.36)0.02 (-0.24, 0.21)1.07 (0.52, 2.21)RERI relative excess risk due to interaction, AP attributable proportion, SI synergy index, CHARLS China Health and Retirement Longitudinal Study, ELSA English Longitudinal Study of Ageing, HRS Health and Retirement Study, BMI body mass indexModel 1: UnadjustedModel 2: Adjusted for age and sexModel 3: Adjusted for model 2 plus education, marital status, smoking status, drinking statusModel 4: Adjusted for model 3 plus BMI; RERI=0, AP=0, and SI=1 indicate no additive interaction; RERI>0, AP>0, and SI>1 suggest positive additive interaction; RERI<0, AP<0, and SI<1 suggest negative additive interaction Table 3Interaction between the frailty and depressive symptoms on cardio-oncology comorbidityDatasetsInteractive effects (95% CI)Multiplicative scaleRERIAPSICHARLS Model 11.19 (0.88, 1.61)0.25 (-0.16, 0.61)0.14 (-0.10, 0.33)1.52 (0.71, 3.23) Model 21.19 (0.88, 1.61)0.23 (-0.17, 0.59)0.14 (-0.11, 0.33)1.52 (0.68, 3.40) Model 31.12 (0.83, 1.52)0.18 (-0.25, 0.56)0.10 (-0.15, 0.30)1.3 (0.68, 2.48) Model 41.11 (0.82, 1.51)0.19 (-0.24, 0.57)0.11 (-0.14, 0.30)1.31 (0.7, 2.46)ELSA Model 11.02 (0.7, 1.48)0.08 (-0.45, 0.60)0.05 (-0.34, 0.29)1.13 (0.48, 2.66) Model 21.07 (0.74, 1.56)0.15 (-0.37, 0.68)0.09 (-0.28, 0.32)1.29 (0.53, 3.16) Model 31.06 (0.73, 1.54)0.12 (-0.39, 0.64)0.08 (-0.31, 0.32)1.26 (0.47, 3.35) Model 41.06 (0.73, 1.54)0.12 (-0.39, 0.63)0.08 (-0.31, 0.32)1.26 (0.46, 3.45)HRS Model 10.99 (0.76, 1.28)-0.01 (-0.33, 0.3)-0.01 (-0.27, 0.18)0.97 (0.45, 2.11) Model 20.99 (0.77, 1.28)0.01 (-0.32, 0.34)0.01 (-0.25, 0.20)1.03 (0.49, 2.15) Model 30.99 (0.77, 1.28)0.02 (-0.31, 0.36)0.02 (-0.24, 0.20)1.05 (0.53, 2.07) Model 41.00 (0.77, 1.29)0.03 (-0.3, 0.36)0.02 (-0.24, 0.21)1.07 (0.52, 2.21)RERI relative excess risk due to interaction, AP attributable proportion, SI synergy index, CHARLS China Health and Retirement Longitudinal Study, ELSA English Longitudinal Study of Ageing, HRS Health and Retirement Study, BMI body mass indexModel 1: UnadjustedModel 2: Adjusted for age and sexModel 3: Adjusted for model 2 plus education, marital status, smoking status, drinking statusModel 4: Adjusted for model 3 plus BMI; RERI=0, AP=0, and SI=1 indicate no additive interaction; RERI>0, AP>0, and SI>1 suggest positive additive interaction; RERI<0, AP<0, and SI<1 suggest negative additive interaction Interaction between the frailty and depressive symptoms on cardio-oncology comorbidity RERI relative excess risk due to interaction, AP attributable proportion, SI synergy index, CHARLS China Health and Retirement Longitudinal Study, ELSA English Longitudinal Study of Ageing, HRS Health and Retirement Study, BMI body mass index Model 1: Unadjusted Model 2: Adjusted for age and sex Model 3: Adjusted for model 2 plus education, marital status, smoking status, drinking status Model 4: Adjusted for model 3 plus BMI; RERI=0, AP=0, and SI=1 indicate no additive interaction; RERI>0, AP>0, and SI>1 suggest positive additive interaction; RERI<0, AP<0, and SI<1 suggest negative additive interaction Joint effects of frailty and depressive symptoms with incident cardio-oncology comorbidity Cumulative hazard plots demonstrated the highest cumulative hazard in participants with frailty and depressive symptoms, while the lowest cumulative hazard was observed in those without frailty and without depressive symptoms (all log-rank P < 0.001) (Fig. 1). In detail, cardio-oncology comorbidity risk was elevated among individuals with both frailty and depressive symptoms by 80% in CHARLS (HR: 1.80, 95% CI: 1.51, 2.14), 57% in ELSA (HR: 1.57, 95% CI: 1.21, 2.02), and 47% in HRS (HR: 1.47, 95% CI: 1.24, 1.74), compared to the those without frailty and without depressive symptoms (Table 4). Fig. 1Cumulative hazard plots of cardio-oncology comorbidity by frailty and depressive symptoms. A: Cumulative hazard plots in CHARLS, B: Cumulative hazard plots in ELSA, C: Cumulative hazard plots in HRS. CHARLS: China Health and Retirement Longitudinal Study, ELSA: English Longitudinal Study of Ageing, HRS: Health and Retirement Study Cumulative hazard plots of cardio-oncology comorbidity by frailty and depressive symptoms. A: Cumulative hazard plots in CHARLS, B: Cumulative hazard plots in ELSA, C: Cumulative hazard plots in HRS. CHARLS: China Health and Retirement Longitudinal Study, ELSA: English Longitudinal Study of Ageing, HRS: Health and Retirement Study Table 4Joint effects of the frailty and depressive symptoms on cardio-oncology comorbidityDatasetsNo frailty andno depressive symptomsNo frailty anddepressive symptomsFrailty andno depressive symptomsFrailty anddepressive symptomsP for trendCHARLS Model 11 [Reference]0.94 (0.75, 1.18)1.54 (1.28, 1.85)1.73 (1.46, 2.03)< 0.001 Model 21 [Reference]0.92 (0.73, 1.16)1.52 (1.26, 1.83)1.67 (1.41, 1.98)< 0.001 Model 31 [Reference]0.97 (0.77, 1.23)1.62 (1.34, 1.95)1.78 (1.50, 2.12)< 0.001 Model 41 [Reference]1.01 (0.80, 1.27)1.60 (1.33, 1.93)1.80 (1.51, 2.14)< 0.001ELSA Model 11 [Reference]1.08 (0.85, 1.38)1.49 (1.24, 1.79)1.65 (1.30, 2.09)< 0.001 Model 21 [Reference]1.08 (0.85, 1.38)1.44 (1.19, 1.73)1.67 (1.31, 2.12)< 0.001 Model 31 [Reference]1.10 (0.86, 1.40)1.44 (1.19, 1.73)1.68 (1.32, 2.15)< 0.001 Model 41 [Reference]1.09 (0.85, 1.40)1.36 (1.11, 1.66)1.57 (1.21, 2.02)< 0.001HRS Model 11 [Reference]1.01 (0.85, 1.21)1.40 (1.25, 1.56)1.40 (1.19, 1.64)< 0.001 Model 21 [Reference]1.06 (0.89, 1.28)1.37 (1.23, 1.53)1.45 (1.23, 1.70)< 0.001 Model 31 [Reference]1.09 (0.91, 1.30)1.41 (1.26, 1.58)1.53 (1.29, 1.80)< 0.001 Model 41 [Reference]1.08 (0.90, 1.30)1.35 (1.20, 1.52)1.47 (1.24, 1.74)< 0.001CHARLS: China Health and Retirement Longitudinal Study, ELSA: English Longitudinal Study of Ageing, HRS: Health and Retirement StudyModel 1: Unadjusted;Model 2: Adjusted for age and sex;Model 3: Adjusted for model 2 plus education, marital status, smoking status, drinking status;Model 4: Adjusted for model 3 plus BMI Table 4Joint effects of the frailty and depressive symptoms on cardio-oncology comorbidityDatasetsNo frailty andno depressive symptomsNo frailty anddepressive symptomsFrailty andno depressive symptomsFrailty anddepressive symptomsP for trendCHARLS Model 11 [Reference]0.94 (0.75, 1.18)1.54 (1.28, 1.85)1.73 (1.46, 2.03)< 0.001 Model 21 [Reference]0.92 (0.73, 1.16)1.52 (1.26, 1.83)1.67 (1.41, 1.98)< 0.001 Model 31 [Reference]0.97 (0.77, 1.23)1.62 (1.34, 1.95)1.78 (1.50, 2.12)< 0.001 Model 41 [Reference]1.01 (0.80, 1.27)1.60 (1.33, 1.93)1.80 (1.51, 2.14)< 0.001ELSA Model 11 [Reference]1.08 (0.85, 1.38)1.49 (1.24, 1.79)1.65 (1.30, 2.09)< 0.001 Model 21 [Reference]1.08 (0.85, 1.38)1.44 (1.19, 1.73)1.67 (1.31, 2.12)< 0.001 Model 31 [Reference]1.10 (0.86, 1.40)1.44 (1.19, 1.73)1.68 (1.32, 2.15)< 0.001 Model 41 [Reference]1.09 (0.85, 1.40)1.36 (1.11, 1.66)1.57 (1.21, 2.02)< 0.001HRS Model 11 [Reference]1.01 (0.85, 1.21)1.40 (1.25, 1.56)1.40 (1.19, 1.64)< 0.001 Model 21 [Reference]1.06 (0.89, 1.28)1.37 (1.23, 1.53)1.45 (1.23, 1.70)< 0.001 Model 31 [Reference]1.09 (0.91, 1.30)1.41 (1.26, 1.58)1.53 (1.29, 1.80)< 0.001 Model 41 [Reference]1.08 (0.90, 1.30)1.35 (1.20, 1.52)1.47 (1.24, 1.74)< 0.001CHARLS: China Health and Retirement Longitudinal Study, ELSA: English Longitudinal Study of Ageing, HRS: Health and Retirement StudyModel 1: Unadjusted;Model 2: Adjusted for age and sex;Model 3: Adjusted for model 2 plus education, marital status, smoking status, drinking status;Model 4: Adjusted for model 3 plus BMI Joint effects of the frailty and depressive symptoms on cardio-oncology comorbidity CHARLS: China Health and Retirement Longitudinal Study, ELSA: English Longitudinal Study of Ageing, HRS: Health and Retirement Study Model 1: Unadjusted; Model 2: Adjusted for age and sex; Model 3: Adjusted for model 2 plus education, marital status, smoking status, drinking status; Model 4: Adjusted for model 3 plus BMI Subgroup and sensitivity analysis Across the three cohorts (CHARLS, ELSA, and HRS), we consistently found that the joint presence of frailty and depressive symptoms was associated with an increased risk of cardio-oncology comorbidity compared to those with neither condition. However, the magnitude of these associations varied slightly across demographic and clinical subgroups. Specifically, participants with both frailty and depressive symptoms exhibited the highest risk across nearly all strata. In the CHARLS cohort, this risk was particularly pronounced in males (HR = 2.16, 95% CI: 1.65, 2.82) and those aged < 70 years (HR = 1.88, 95% CI: 1.46, 2.44), although the p value for interaction for age, gender, and BMI remained non-significant (p value > 0.05). Regarding gender differences, a significant interaction was observed in the HRS cohort ( p value for interaction = 0.041), where the joint effect was more prominent in females (HR = 1.61, 95% CI: 1.33, 1.95) than in males (HR = 1.09, 95% CI: 0.74, 1.60). In contrast, no such gender-specific moderation was identified in the ELSA or CHARLS cohorts. For BMI subgroups, the elevated risk associated with the co-occurrence of frailty and depressive symptoms remained consistent across all cohorts, with no significant interaction detected (all p value for interaction > 0.05) (Table 5). Although the magnitude of the joint association varied across age, sex, and BMI strata, most tests for interaction were non-significant. Therefore, subgroup-specific estimates should be interpreted as exploratory and should not be taken as evidence of consistent effect modification. Across age, sex, and BMI subgroups, frailty and depressive symptoms were generally associated with an increased risk of cardio-oncology comorbidity, with largely non-significant interaction tests suggesting no consistent evidence of effect modification by these subgroup factors (Table S4- S6). We observed similar associations with the primary analyses after additionally adjusting medication information (Table S1). Additional exclusion of incident cardio-oncology comorbidity cases within the first follow-up did not alter the observed associations (Table S2). The findings were consistent when we restricted the analyses to participants without missing data on covariates (Table S3). Multiple sensitivity analyses demonstrated the robustness of our main findings. Table 5Subgroup analyses of the joint associations of frailty and depressive symptoms with cardio-oncology comorbidityDatasetsNo frailty andno depressive symptomsNo frailty anddepressive symptomsFrailty andno depressive symptomsFrailty anddepressive symptomsP for interactionCHARLS Age < 701 [Reference]1.12 (0.81, 1.54)1.69 (1.26, 2.25)1.88 (1.46, 2.44)0.643 Age ≥ 701 [Reference]0.88 (0.63, 1.23)1.51 (1.18, 1.93)1.71 (1.35, 2.16) Male1 [Reference]0.82 (0.57, 1.20)1.53 (1.16, 2.03)2.16 (1.65, 2.82)0.170 Female1 [Reference]1.13 (0.83, 1.53)1.64 (1.27, 2.11)1.63 (1.30, 2.05) BMI < 30 kg/m21 [Reference]0.98 (0.77, 1.24)1.61 (1.33, 1.96)1.81 (1.51, 2.16)0.779 BMI≥30 kg/m21 [Reference]1.92 (0.46, 8.11)1.54 (0.64, 3.71)1.24 (0.49, 3.17)ELSA Age < 701 [Reference]1.12 (0.71, 1.76)1.23 (0.83, 1.81)1.36 (0.83, 2.24)0.919 Age ≥ 701 [Reference]1.07 (0.79, 1.44)1.44 (1.14, 1.82)1.68 (1.24, 2.27) Male1 [Reference]1.17 (0.76, 1.81)1.41 (1.04, 1.92)1.88 (1.18, 3.00)0.870 Female1 [Reference]1.06 (0.78, 1.44)1.33 (1.03, 1.73)1.43 (1.05, 1.95) BMI < 30 kg/m21 [Reference]1.05 (0.79, 1.40)1.38 (1.07, 1.79)1.59 (1.15, 2.19)0.754 BMI≥30 kg/m21 [Reference]1.24 (0.76, 2.02)1.35 (0.99, 1.86)1.64 (1.08, 2.51)HRS Age < 701 [Reference]1.12 (0.85, 1.49)1.32 (1.08, 1.61)1.65 (1.24, 2.21)0.927 Age ≥ 701 [Reference]1.07 (0.84, 1.36)1.46 (1.26, 1.68)1.43 (1.16, 1.77) Male1 [Reference]1.15 (0.84, 1.57)1.21 (0.99, 1.48)1.09 (0.74, 1.60)0.041 Female1 [Reference]1.06 (0.85, 1.32)1.45 (1.26, 1.68)1.61 (1.33, 1.95) BMI < 30 kg/m21 [Reference]1.16 (0.94, 1.43)1.43 (1.23, 1.67)1.41 (1.11, 1.78)0.345 BMI≥30 kg/m21 [Reference]0.93 (0.65, 1.33)1.27 (1.06, 1.52)1.52 (1.19, 1.95) Table 5Subgroup analyses of the joint associations of frailty and depressive symptoms with cardio-oncology comorbidityDatasetsNo frailty andno depressive symptomsNo frailty anddepressive symptomsFrailty andno depressive symptomsFrailty anddepressive symptomsP for interactionCHARLS Age < 701 [Reference]1.12 (0.81, 1.54)1.69 (1.26, 2.25)1.88 (1.46, 2.44)0.643 Age ≥ 701 [Reference]0.88 (0.63, 1.23)1.51 (1.18, 1.93)1.71 (1.35, 2.16) Male1 [Reference]0.82 (0.57, 1.20)1.53 (1.16, 2.03)2.16 (1.65, 2.82)0.170 Female1 [Reference]1.13 (0.83, 1.53)1.64 (1.27, 2.11)1.63 (1.30, 2.05) BMI < 30 kg/m21 [Reference]0.98 (0.77, 1.24)1.61 (1.33, 1.96)1.81 (1.51, 2.16)0.779 BMI≥30 kg/m21 [Reference]1.92 (0.46, 8.11)1.54 (0.64, 3.71)1.24 (0.49, 3.17)ELSA Age < 701 [Reference]1.12 (0.71, 1.76)1.23 (0.83, 1.81)1.36 (0.83, 2.24)0.919 Age ≥ 701 [Reference]1.07 (0.79, 1.44)1.44 (1.14, 1.82)1.68 (1.24, 2.27) Male1 [Reference]1.17 (0.76, 1.81)1.41 (1.04, 1.92)1.88 (1.18, 3.00)0.870 Female1 [Reference]1.06 (0.78, 1.44)1.33 (1.03, 1.73)1.43 (1.05, 1.95) BMI < 30 kg/m21 [Reference]1.05 (0.79, 1.40)1.38 (1.07, 1.79)1.59 (1.15, 2.19)0.754 BMI≥30 kg/m21 [Reference]1.24 (0.76, 2.02)1.35 (0.99, 1.86)1.64 (1.08, 2.51)HRS Age < 701 [Reference]1.12 (0.85, 1.49)1.32 (1.08, 1.61)1.65 (1.24, 2.21)0.927 Age ≥ 701 [Reference]1.07 (0.84, 1.36)1.46 (1.26, 1.68)1.43 (1.16, 1.77) Male1 [Reference]1.15 (0.84, 1.57)1.21 (0.99, 1.48)1.09 (0.74, 1.60)0.041 Female1 [Reference]1.06 (0.85, 1.32)1.45 (1.26, 1.68)1.61 (1.33, 1.95) BMI < 30 kg/m21 [Reference]1.16 (0.94, 1.43)1.43 (1.23, 1.67)1.41 (1.11, 1.78)0.345 BMI≥30 kg/m21 [Reference]0.93 (0.65, 1.33)1.27 (1.06, 1.52)1.52 (1.19, 1.95) Subgroup analyses of the joint associations of frailty and depressive symptoms with cardio-oncology comorbidity Discussion In this study of three multinational cohorts, frailty was significantly associated with an elevated risk of cardio-oncology comorbidity. This association demonstrated consistency across different national and population cohorts. Concurrently, this study also identified an association between depressive status and an increased risk of cardio-oncology comorbidity. Although no statistically significant multiplicative interaction between frailty and depressive symptoms was detected, the joint effect analysis revealed that individuals concurrently experiencing both frailty and depressive symptoms constituted the subgroup at the highest risk of developing cardio-oncology comorbidity, compared to those with neither condition. After full adjustment for potential confounders, frailty was significantly associated with an elevated risk of cardio-oncology comorbidity in the present study. This association demonstrated consistency across different national and population cohorts. Our findings align with evidence from multiple large-scale cohort studies, which have established frailty as an independent and dynamically modifiable risk factor for incident CVD and cancer. For instance, analyses from three multinational prospective cohorts (n = 19,685) including CHARLS, ELSA, and HRS showed that progression in frailty status (from robust to pre-frail or frail) was significantly associated with an increased risk of incident CVD [9]. Data from the UK Biobank (n = 453,144) and Screening Across the Lifespan Twin study (n = 36,888) indicated that frailty was associated with a higher risk of overall cancer (UK Biobank: FI-HR = 1.22, 95% CI: 1.17, 1.28) [27]. The analysis of HRS further revealed that baseline frailty was associated with a higher cancer risk (HR = 1.61, 95% CI: 1.27, 2.02), that risk was further elevated with frailty progression [28]. Notably, healthy lifestyle significantly attenuated the adverse effect of frailty on cancer risk, with a protective effect observed particularly for lung, colorectal, and liver cancers (pooled HR = 0.74, 95% CI: 0.71, 0.78) [11]. Concurrently, this study also identified an association between depressive status and an increased risk of cardio-oncology comorbidity. The UK Biobank study showed that individuals with depressive symptoms alone had a significantly higher risk of incident CVD [29]. An meta-analysis including 26 studies of 1,957,621 individuals indicated that depressive symptoms is associated with an increased risk of incident stroke, heart disease, and CVD [30]. Results from the PSY-CA consortium including data from 18 cohorts (n = 319,613) reported that depressive symptoms and anxiety (symptoms and diagnoses) were associated with the incidence of lung cancer and smoking-related cancers, though no associations were found between depressive symptoms or anxiety and overall cancers [31]. The use of three multinational prospective cohorts strengthens the external validity of this study. However, CHARLS, ELSA, and HRS were conducted in different national, socioeconomic, and healthcare contexts, and participants differed in demographic composition, lifestyle characteristics, and baseline risk profiles. Therefore, our findings should not be interpreted as requiring strict equivalence in effect size across cohorts. Rather, the broadly consistent direction of associations across cohorts suggests that frailty and depressive symptoms may represent relevant risk markers for cardio-oncology comorbidity across diverse older populations. Although no statistically significant multiplicative interaction between frailty and depressive symptoms was detected, the joint effect analysis revealed that individuals concurrently experiencing both frailty and depressive symptoms constituted the subgroup at the highest risk of developing cardio-oncology comorbidity, compared to those with neither condition. Therefore, the absence of formal interaction should not be interpreted as evidence that combined assessment is uninformative. Rather, it indicates that the observed risk elevation was compatible with the expected combined contribution of two independent risk markers, while the absolute and relative risk remained greatest among participants with both conditions. Direct evidence evaluating the joint effects of frailty and depressive symptoms on the risk of incident CVD or cancer is currently lacking, a pooled analysis (n = 220,947) including UK Biobank, ELSA and HRS reported that participants with both physical frailty and depressive symptoms exhibited the highest dementia risk (pooled HR: 3.23) and 17.1% of dementia risk attributed to their interactive effects [32]. Mendelian randomization studies have confirmed a bidirectional causal relationship between frailty and depressive symptoms [33, 34]. An analysis of data from 352,277 UK Biobank participants further indicated that inflammatory markers (e.g., C-reactive protein, neutrophil count) and brain structural changes (e.g., reduced hippocampal volume) mediate this association between frailty and depressive symptoms [35], which also sheds light on the biological pathways through which they may jointly influence cardio-oncology comorbidity. Frailty and depressive symptoms share core pathophysiological pathways implicated in both cardiovascular disease and cancer, primarily involving chronic inflammation and metabolic oxidative stress pathways. A review supported that the link between depressive symptoms, frailty, and cardiovascular risk may involve a state of sustained overproduction of pro-inflammatory cytokines [36]. Evidence from animal models further elucidates shared mechanisms related to oxidative stress. Results from studies in mouse models indicate that a depressive state can alter gut microbiota, leading to aberrant tryptophan metabolism and reduced levels of the cardioprotective metabolite indole-3-propionic acid, which in turn exacerbates myocardial ferroptosis and cardiac dysfunction via the NRF2/System xc⁻/GPX4 pathway [37]. In older adults with cancer, systemic inflammation (e.g., neutrophil-to-lymphocyte ratio ≥ 3) is positively correlated with frailty status, and the co-existence of both is associated with a significantly increased risk of all-cause mortality [38, 39]. Analysis in breast cancer patients has found that depressive symptoms are closely associated with a panel of immune-metabolic biomarkers, a combination that can identify a depressive symptoms subtype driven by a dysregulated immune-metabolic milieu [40]. Cumulatively, this evidence suggests that frailty and depressive symptoms may share a critical immune-metabolic dysregulation pathway in cancer pathogenesis. Building on the finding that the co-occurrence of frailty and depressive symptoms identified the subgroup with the highest observed risk, our results may have implications for baseline risk stratification. Combined assessment of frailty and depressive symptoms may help identify older adults with coexisting physical and psychological vulnerability who warrant closer monitoring and more comprehensive evaluation, including frailty-oriented functional assessment, mental health assessment, and multidisciplinary cardio-oncology risk assessment. However, because this was an observational study, our findings should not be interpreted as direct evidence that combined screening or integrated interventions improve clinical outcomes. Future implementation studies and clinical trials are needed to determine whether combined screening followed by tailored physical, psychological, or multidisciplinary interventions can improve cardio-oncology outcomes. This study is the first to evaluate the joint effect of frailty and depressive symptoms on cardio-oncology comorbidity in a multinational older population, identifying concurrent “frailty and depressive symptoms” as the highest-risk phenotype for comorbidity, thereby providing key epidemiological rationale for integrated biopsychosocial interventions. Limitations of this study should be considered. Disease outcomes were based on self-reported physician diagnoses, which may introduce outcome misclassification. If such misclassification was nondifferential with respect to baseline frailty or depressive symptoms, the estimated hazard ratios would likely be biased toward the null. In addition, the available cohort data did not provide sufficient detail on cancer type, stage, treatment history, or specific cardiovascular subtypes, limiting our ability to evaluate heterogeneity across cardio-oncology disease profiles. Frailty and depressive symptoms were assessed only at baseline. Because both conditions may change during follow-up, unmeasured transitions in frailty or depressive symptom status may have introduced exposure misclassification and attenuated the observed associations. Future studies with repeated assessments are needed to evaluate how trajectories or changes in frailty and depressive symptoms relate to cardio-oncology comorbidity risk. Despite adjustment for numerous confounders, residual confounding (e.g., detailed genetic factors and dietary information) remains possible, and the observational design cannot establish causality. Our study did not directly evaluate whether separate identification of frailty and depressive symptoms leads to differential clinical management or improved patient outcomes. Future implementation studies and clinical trials are needed to determine whether combined screening followed by tailored physical, psychological, or multidisciplinary interventions can improve cardio-oncology outcomes. Residual measurement heterogeneity may remain despite harmonization, as the frailty index included cognition as continuous variable and depressive symptoms were assessed using different CES-D versions and cohort-specific cut-off values. Such differences may have contributed to between-cohort variation in prevalence and effect estimates, and thus cross-cohort comparisons should be interpreted cautiously. Several contextual differences across cohorts should be considered when interpreting the generalizability of our findings. Differences in healthcare systems, outcome ascertainment, population structure, socioeconomic context, and baseline risk factor distributions may have contributed to between-cohort heterogeneity. Therefore, the transportability and contextual applicability of these findings should be further evaluated in additional populations and healthcare settings. The absence of biomarker data precluded direct within-sample validation of hypothesized mechanisms such as inflammation or metabolic dysregulation. This study provides the first multinational evidence that the co-occurrence of frailty and depressive symptoms constitutes the highest-risk phenotype for incident cardio-oncology comorbidity in community-dwelling older adults. This finding offers a pivotal epidemiological rationale for establishing an integrated “physical-mental” risk assessment framework in geriatric care. Future studies are needed to determine whether such risk stratification can guide tailored monitoring or multidisciplinary interventions to reduce cardio-oncology comorbidity risk. Supplementary Information Supplementary Material 1. Supplementary Material 1.