Sex differences in the burden of multiple long-term conditions among Japanese individuals with heart failure Authors: Shuhei Tara, Yayoi Tetsuou Tsukada, Takuya Nishino, Katsuhito Kato, Takeshi Yamamoto, Mitsunori Maruyama, Eitaro Kodani, Nobuaki Kobayashi, Akihiro Shirakabe, Kuniya Asai Published: 2025 ABSTRACT AbstractBackgroundMultiple long-term conditions (MLTCs) are common among individuals with heart failure (HF); however, the influence of sex on clinical characteristics and prognosis remains unclear. This study investigated sex-related differences in clinical characteristics and prognosis in individuals with HF and a high MLTC burden.MethodsWe conducted a multicentre retrospective study of 3004 hospitalised patients with HF who survived to discharge. Comorbid chronic conditions were defined using the Charlson Comorbidity Index (CCI), and MLTC burden was quantified using the age-adjusted CCI. Patients were stratified into high (CCI≥6; n=1514) and low (CCI<6; n=1490) MLTC burden groups based on the median age-adjusted CCI, and comparisons were made by sex. The primary outcome was a composite of all-cause death or HF readmission within 1 year.ResultsThe prevalence of ≥2 comorbid conditions increased with age, peaking before age 85 and declining slightly thereafter. This trend differed by sex, with a steeper age-related increase observed in men. In the high MLTC burden group, females were older, had higher left ventricular ejection fraction and were prescribed fewer medications than men. Dementia and rheumatologic disease were more common in women. Although no significant sex differences in event-free survival were observed overall in either MLTC group, among individuals aged≥75 years with high MLTC burden, men had a significantly higher risk of adverse events than women (HR: 1.334; 95% CI 1.031 to 1.727). A spline-based three-way interaction analysis (age×sex×MLTC burden) demonstrated that male risk exceeded female risk after age 70, only in the high MLTC burden group.ConclusionIn individuals with HF and a high MLTC burden, age-dependent sex differences in prognosis were evident. These findings highlight the importance of incorporating age-specific and sex-specific approaches into personalised HF care.Trial registration numberUMIN000054854.Keywords: Heart Failure, Epidemiology, Women WHAT IS ALREADY KNOWN ON THIS TOPIC WHAT THIS STUDY ADDS This multicentre study found that among individuals with a high MLTC burden, older men had a significantly higher risk of adverse outcomes than women. This multicentre study found that among individuals with a high MLTC burden, older men had a significantly higher risk of adverse outcomes than women. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY These findings emphasise the need to incorporate age-specific and sex-specific considerations into HF risk stratification and personalised care strategies, particularly for older patients with multimorbidity. These findings emphasise the need to incorporate age-specific and sex-specific considerations into HF risk stratification and personalised care strategies, particularly for older patients with multimorbidity. Introduction In ageing societies, one of the major challenges is providing adequate healthcare for individuals with multiple long-term conditions (MLTCs), defined as the presence of two or more chronic conditions in an individual.1 2 Over 70% of adults develop cardiovascular disease (CVD) by the age of 70 years, and more than two-thirds of these individuals also experience non-CVD chronic conditions, such as diabetes, chronic kidney disease or chronic obstructive pulmonary disease.3 4 Consequently, MLTCs are particularly prevalent among older adults with CVD and are associated with a higher risk of hospital readmission, reduced quality of life and worse survival outcomes.5 Heart failure (HF) represents the terminal stage of many CVDs and is characterised by complex clinical manifestations. As a result, individuals with HF commonly exhibit MLTCs, further worsening their prognosis.6 In Japan, which has become a super-ageing society, the increase in individuals with HF, referred to as a ‘heart failure pandemic’, has placed a significant strain on the healthcare delivery system and led to a rise in medical expenses.7 8 Treatment strategies increasingly emphasise the need to consider diverse individual characteristics. However, the understanding and application of patient-specific factors among individuals with HF remain limited. For example, older women with HF are more likely to present with preserved ejection fraction and comorbid conditions, such as anaemia or depression, whereas men often exhibit reduced ejection fraction and higher rates of ischaemic heart disease.9 These differences are often underappreciated in clinical practice and may contribute to disparities in treatment outcomes. Given these considerations, we focused on the potential sex differences in MLTC burden among individuals with HF, as such differences may influence clinical outcomes. Therefore, this study aimed to investigate sex differences in the clinical characteristics and prognostic impact of MLTC burden among individuals with HF using the age-adjusted Charlson Comorbidity Index (CCI), a scoring system that incorporates specific chronic conditions to estimate prognosis.10 Methods Study design and data collection This multicentre retrospective cohort study was conducted using an inpatient database and follow-up data from four affiliated hospitals of Nippon Medical School: Nippon Medical School Hospital, Musashi Kosugi Hospital, Tama Nagayama Hospital and Chiba Hokusoh Hospital. The database was constructed using a combination of Diagnosis Procedure Combination (DPC) data and medical record data. This study is registered in the UMIN Clinical Trials Registry (URL: https://www.umin.ac.jp/ctr; unique identifier: UMIN000054854). The study population consisted of individuals aged 18 years or older with acute HF who were urgently admitted to the department of cardiovascular medicine or the cardiovascular intensive care unit of the four affiliated hospitals of Nippon Medical School between April 2018 and September 2023 (online supplemental figure 1). Individuals with acute HF were identified as those with an ICD-10 code (I50X) for HF, recorded as the condition requiring the most medical resources, the primary diagnosis at the time of hospitalisation or the condition leading to hospitalisation. After excluding individuals who died during hospitalisation, 3004 individuals were included in the analysis. The primary endpoint of this study was the composite outcome of all-cause death and readmission due to HF within 1 year of hospital discharge. Definition and classification of MLTC burden in individuals with HF Chronic conditions were defined using the CCI10 and identified using ICD-10 codes.11,13 All comorbid conditions included in the CCI were defined strictly according to the original CCI framework and its validated ICD-10 coding algorithms. Because all individuals in this cohort had HF as the index condition, HF itself was excluded from the CCI calculation to avoid double-counting HF-related pathologies. The CCI weights for all comorbidities, including AIDS/HIV, were applied exactly as defined in the original CCI and its validated ICD-10 coding algorithms, without modification. Therefore, in this study, 18 chronic conditions (ie, all CCI conditions other than HF), with weights assigned to each, were used to evaluate comorbid chronic conditions and to define MLTC burden in individuals with HF (online supplemental table 1). For epidemiological evaluation, individuals were categorised based on the number of comorbid chronic conditions (0, 1 or ≥2). MLTC burden was quantified using the age-adjusted CCI. The CCI was calculated by summing the weights assigned to each chronic condition, and additional points were added based on age (eg, 1 point for 50–59 years, 2 points for 60–69 years and 3 points for 70–79 years10). The cohort was stratified into two groups based on the median age-adjusted CCI: a low MLTC burden group (CCI<6, n=1490) and a high MLTC burden group (CCI≥6, n=1514) (online supplemental figure 1). Sex differences were evaluated within each group. Variables The DPC captured data on all hospitalised individuals, including demographics (age, sex), physical metrics (height, weight, body mass index (BMI) (calculated by dividing body weight in kilograms by the square of height in metres)) and clinical details (comorbidities, procedures during hospitalisation and medications).14 Ultrasound left ventricular ejection fraction (LVEF) and blood test data at discharge, defined as the most recent sampling within 14 days before discharge, were obtained from medical records. In accordance with major international HF guidelines, we classified LVEF using the standard threshold of <40% to define HF with reduced ejection fraction (HFrEF) for subgroup analysis. Although physiological sex differences in LVEF have been reported, sex-specific EF cut-offs have not been established or validated for clinical classification or prognostic stratification. Therefore, to maintain methodological consistency and comparability with existing HF literature, the same guideline-endorsed 40% threshold was applied to both sexes in this study. Statistical analysis Categorical variables are presented as numbers and percentages and were compared using the χ2 test. Continuous variables are presented as median values with IQRs and were analysed using the Mann-Whitney U test. For survival analysis, the Kaplan-Meier method was used to estimate event-free survival. Subgroup analyses were performed to evaluate sex differences in the primary endpoints across subgroups. Variables included in the subgroup analysis were age (<75 years vs ≥75 years), LVEF (<40% vs ≥40%), number of medications (<10 vs ≥10) and BMI (<18.5, 18.5–<25.0 and ≥25.0). HRs and 95% CIs for men versus women were calculated within each subgroup using Cox proportional hazards modelling. Interaction p values were calculated to assess the significance of sex-related interactions across subcategories. These analyses were conducted separately for the low and high MLTC burden groups. To evaluate the modifying effect of MLTC burden on the relationship between age and sex in predicting outcomes, a multivariate Cox proportional hazards model was constructed, incorporating a three-way interaction term (age×sex×MLTC burden group (low vs high)). Age was modelled as a continuous variable using restricted cubic splines (RCS) with four knots (5th, 35th, 65th and 95th percentiles) to account for potential nonlinear associations. Sex-specific predicted risk scores were calculated across age and visualised as trajectories with 95% CIs, stratified by MLTC burden group. Statistical significance was defined as a two-sided p value of <0.05. All analyses were performed using R software V.4.2.2 Patched (R Foundation for Statistical Computing, Vienna, Austria). Results Epidemiological assessment of comorbid chronic conditions The prevalence in each age group for the three categories classified according to the number of comorbid chronic conditions among individuals with HF (ie, 0, 1 and ≥2) is shown in figure 1A. The proportion of individuals with two or more chronic conditions steadily increased with age, peaking in the 71–75 year age group. However, after 85 years of age, this proportion declined slightly. Epidemiological assessments of comorbid chronic conditions in individuals with heart failure. Individuals were categorised into three groups based on the number of chronic conditions (0, 1 or ≥2). (A) Age-specific prevalence of the three groups across different age categories. (B) Sex-specific distribution of individuals within each age group, separately analysed for men and women. When the prevalence of the three groups, classified by the number of chronic conditions (ie, 0, 1 and ≥2), was evaluated by age group and sex, sex differences were observed (figure 1B). Among men, the number of individuals with two or more chronic conditions consistently exceeded those without chronic conditions across all age groups aged 51 years or older. This gap gradually widened with advancing age, peaking at approximately 81–85 years of age. In contrast, among women, the number of individuals with two or more chronic conditions was only slightly higher than those without chronic conditions. However, this trend reversed in individuals aged 81 years or older, with the number without chronic conditions surpassing those with two or more. Sex differences in clinical characteristics among individuals with a high MLTC burden Sex differences in clinical characteristics were evaluated among individuals with a high MLTC burden (CCI≥6). Compared with men, women were older (86 (81–90) vs 82 (75–86) years; p<0.001), had a higher prevalence of HF with preserved ejection fraction (LVEF 50% (36–64) vs 41% (30–57); p<0.001), a lower rate of haemodialysis (4.0% vs 8.3%; p=0.001) and were prescribed fewer medications at discharge (median 10 (8–12) vs 11 (9–13); p=0.001). Specifically, the prescription rates of β-blockers (64.8% vs 71.1%; p=0.010), sodium–glucose cotransporter 2 inhibitors (10.7% vs 21.7%; p<0.001), aspirin (20.6% vs 33.0%; p<0.001) and P2Y12 inhibitors (21.6% vs 30.7%; p<0.001) were significantly lower in women than in men (online supplemental table 2). In the ischaemic subgroup, defined according to CCI comorbidities of myocardial infarction, peripheral vascular disease or cerebrovascular disease, women had lower antiplatelet use than men (55.4% vs 67.9%; p<0.001). Among individuals with a high MLTC burden (CCI≥6), sex differences were also evident in the prevalence of specific chronic conditions. Compared with men, women had a higher prevalence of dementia (18.0% vs 5.9%; p<0.001) and rheumatologic disease (3.4% vs 1.3%; p=0.011), but a lower prevalence of chronic pulmonary disease (13.4% vs 19.0%; p=0.005), diabetes with chronic complications (11.7% vs 19.7%; p<0.001), myocardial infarction (20.6% vs 29.1%; p<0.001) and renal disease (23.3% vs 35.0%; p<0.001) (table 1). Table 1Sex difference in the prevalence of chronic conditions in high MLTC burdenComorbidity conditionsMale (n=858)Female (n=656)P valueAIDS/HIV1 (0.1)1 (0.2)1.000Cerebrovascular disease88 (10.3)83 (12.7)0.168Chronic pulmonary disease163 (19.0)88 (13.4)0.005Dementia51 (5.9)118 (18.0)<0.001Diabetes with chronic complications169 (19.7)77 (11.7)<0.001Diabetes without chronic complications299 (34.8)200 (30.5)0.083Hemiplegia or paraplegia5 (0.6)3 (0.5)1.000Metastatic solid tumour10 (1.2)4 (0.6)0.396Mild liver disease40 (4.7)32 (4.9)0.941Moderate or severe liver disease1 (0.1)0 (0.0)1.000Myocardial infarction250 (29.1)135 (20.6)<0.001Peptic ulcer disease234 (27.3)145 (22.1)0.025Peripheral vascular disease84 (9.8)51 (7.8)0.203Renal disease300 (35.0)153 (23.3)<0.001Rheumatology disease11 (1.3)22 (3.4)0.011Leukaemia9 (1.0)2 (0.3)0.166Lymphoma5 (0.6)6 (0.9)0.654Malignant tumour71 (8.3)39 (5.9)0.103Variables are presented as n (%).MLTC, multiple long-term condition. Sex difference in the prevalence of chronic conditions in high MLTC burden Variables are presented as n (%). MLTC, multiple long-term condition. Survival analysis During the 1-year follow-up period, 466 events occurred in the overall population, comprising 126 all-cause deaths and 340 HF readmissions. No significant sex differences in the 1-year event-free survival rate were observed in either the low or high MLTC burden groups (low MLTC burden: men, 83.1% vs women, 82.1%; p=0.638; high MLTC burden: men, 69.8% vs women, 72.7%; p=0.311) (online supplemental figure 2A,B). Subgroup analysis demonstrated that, among individuals with a low MLTC burden (CCI<6), no statistically significant association between sex and the primary endpoint was observed in any subgroup. However, within the age subgroup, a similar trend to that seen in the high MLTC burden group was noted: men aged≥75 years tended to have a higher risk of the primary endpoint compared with women (HR: 1.439, 95% CI 0.996 to 2.080), and no difference was observed among those aged<75 years (HR: 0.893, 95% CI 0.521 to 1.531). The interaction p value for age was 0.152, indicating no statistically significant interaction (figure 2A). In the high MLTC burden group (CCI≥6), men aged≥75 years showed a significantly higher risk of the primary endpoint compared with women (HR: 1.334, 95% CI 1.031 to 1.727), while no significant difference was observed in those aged<75 years (HR: 0.679, 95% CI 0.346 to 1.331). The interaction p value for age was 0.067, suggesting a potential modifying effect of sex, although statistical significance was not reached (figure 2B). In other subgroups, including those based on LVEF, number of medications and BMI, no significant associations between sex and the primary endpoints were observed in any category, and the interaction p values indicated no significant interactions between sex and these factors. Subgroup analysis assessing sex differences in the primary endpoints. HRs for men versus women were calculated within each subgroup. Interaction p values were also calculated to assess the significance of sex-related interactions across the categories. Analyses were performed for (A) the low MLTC burden group and (B) the high MLTC burden group. BMI, body mass index (kg/m2); LVEF, left ventricular ejection fraction; MLTC, multiple long-term conditions. To assess how age modifies the associations among sex, MLTC burden and clinical outcomes, we performed a three-way interaction analysis (age×sex×MLTC burden) using RCS for age. Chronological age was not included as a separate covariate because the CCI is an age-adjusted score, and including age independently would introduce multicollinearity within the analytic framework. In the three-way interaction model incorporating RCS for age, predicted risk scores showed a nonlinear association with age, with distinct trajectories stratified by MLTC burden and sex (figure 3). In the low MLTC burden group, the predicted risk increased sharply with advancing age, particularly among men, whereas women showed a more gradual increase in risk. In the high MLTC burden group, the predicted risk curves for men and women intersected around the age of 70 years, indicating a possible reversal of sex-related prognostic differences at older ages. Solid lines represent the estimated risk scores, and shaded areas represent 95% CIs for each sex-specific curve. Sex-specific predicted risk scores by age and MLTC burden, derived from a Cox proportional hazards model with a three-way interaction term (age×sex×MLTC burden). The model incorporated restricted cubic splines for age to allow for nonlinear effects. Predicted risk scores are shown across age for men (blue) and women (red), stratified by MLTC burden: low (left panel) and high (right panel). Solid lines represent the estimated risk scores, while the shaded areas indicate 95% CIs. The analysis illustrates how the prognostic impact of sex varies with age and MLTC burden. MLTC, multiple long-term conditions. Discussion Among individuals with HF, this study demonstrated that sex differences in clinical characteristics, including patterns of comorbid chronic conditions, were evident in those with a high MLTC burden. Among individuals with a high MLTC burden, no significant sex-specific differences were observed in the incidence of outcomes, including all-cause death and HF readmission. However, older men within this group had a higher risk of adverse outcomes compared with women. This finding may reflect age-dependent modification of sex effects on prognosis, as indicated by the spline-based three-way interaction analysis. Moreover, epidemiological assessments revealed that the age-related increase in the number of individuals with two or more comorbid chronic conditions was more pronounced in men than in women. Taken together, these findings suggest that interventions targeting MLTCs may be particularly beneficial in older men, who not only have a higher comorbid burden but also demonstrate a higher risk of poor outcomes. In contrast, older women tend to have fewer comorbid chronic conditions, and conditions such as dementia may reflect their older age distribution. As a result, the impact of MLTC-focused interventions may be relatively limited in this group. Therefore, incorporating age- and sex-specific strategies into risk stratification and the management of HF—especially focusing on MLTCs in older men—may help optimise clinical outcomes. In the high-MLTC burden group, women had a smaller proportion of HFrEF compared with men. Since guideline-directed medical therapy (GDMT) is more strongly supported in HFrEF, this difference in EF phenotype may partly explain the lower use of these medications in women. In addition, women with a high MLTC burden were also older than their male counterparts, which may have influenced prescribing decisions due to concerns regarding tolerability, renal function, hypotension or potential drug–drug interactions. Because clinical characteristics differed markedly between sexes, and the reasons for not prescribing these medications were not available, we were unable to isolate sex differences in GDMT use specifically within the HFrEF subgroup. Given these considerations, we did not restrict analyses to a single EF phenotype. EF phenotype is closely linked to age, sex and comorbidity structure; therefore, restricting analyses to a single EF category would have markedly reduced sample size and analytic stability within sex- and MLTC-stratified subgroups. Japan is one of the world’s most rapidly ageing societies. Individuals with HF in Japan are ageing, with the proportion of women also increasing.7 Comparative studies have shown that Japanese individuals with HF tend to be older and are more frequently women than individuals in other Asian or European populations.7 15 16 Although this study was conducted solely in Japan, its findings may offer valuable insights for other countries facing similar demographic transitions. However, differences in healthcare systems, socioeconomic contexts and disease management practices may limit the generalisability of the results. Therefore, future international comparative studies are warranted to determine whether the sex-specific patterns of MLTCs and HF outcomes identified in this study are also present in other healthcare settings. In addition to evaluating generalisability, our findings highlight the importance of understanding the mechanisms underlying sex differences in MLTCs among individuals with HF. A comprehensive management approach that incorporates MLTCs into HF care is essential, particularly given the diverse impact of comorbid chronic conditions. MLTCs in HF can be broadly categorised as concordant conditions (eg, HF and hypertension), which share pathophysiological pathways and treatment strategies, and discordant conditions (eg, HF and osteoarthritis), which have unrelated mechanisms but still influence overall care.5 17 Despite the need to address both types, current disease-specific guidelines often lack recommendations for managing discordant MLTCs in individuals with HF. This gap may be attributed to the complexity of individualised treatment planning and the limited evidence supporting integrated care models.5 Furthermore, the sex differences observed in individuals with a high MLTC burden emphasise the importance of incorporating sex-specific considerations into clinical guidelines and care strategies.18 However, this study was not designed to disentangle the specific biological, clinical, behavioural or social factors contributing to these sex-based differences.19 Future research should elucidate these mechanisms to support the development of more personalised and equitable therapeutic approaches. To quantify MLTC burden, we used the age-adjusted CCI.6 20 A previous study using Japanese nationwide claims data demonstrated a strong association between higher CCI and adverse outcomes.11 Furthermore, non-cardiac chronic conditions are known to exacerbate HF progression, with an increasing number of comorbid chronic conditions associated with higher hospitalisation rates.21 However, the prognostic impact of MLTCs varies depending on the type and severity of comorbid diseases, as well as patient age. Therefore, quantifying MLTC burden according to both comorbid chronic conditions and age is essential for accurate risk stratification, especially among elderly populations. In our cohort, individuals with an age-adjusted CCI of ≥6 were classified as having a high MLTC burden. This threshold is comparable to previous findings, suggesting that a higher age-adjusted CCI predicts adverse outcomes in HF.6 22 Minor discrepancies may be attributed to variations in population demographics, regional healthcare practices or operational definitions of chronic conditions. This study had some limitations. First, because this was a retrospective analysis, the findings are subject to potential selection bias. Second, reliance on ICD-10 coding may have led to underreporting or misclassification of chronic conditions. Third, the lack of detailed data on treatment regimens and their interaction with chronic conditions limits the interpretability of the prognostic outcomes. Future studies should use prospective designs to confirm causality and investigate the effect of integrated MLTC-focused interventions, ideally stratified by sex, on clinical outcomes in HF. Fourth, although sex-related physiological differences in LVEF have been described, sex-specific EF thresholds are not currently incorporated into international HF guidelines or validated for prognostic research. Therefore, we applied the widely accepted cut-off of 40% for both sexes. Fifth, because the study cohort included only individuals who survived to discharge, comorbidity information required to calculate MLTC burden was not available for those who died during the index hospitalisation. Therefore, the MLTC burden at admission among in-hospital deaths could not be assessed, and the influence of early mortality on the overall MLTC distribution remains unknown. Sixth, our dataset did not include detailed information on postdischarge medical expenditures, long-term care insurance claims or community-based social care services. As a result, we were unable to assess the downstream economic burden associated with MLTCs in individuals with HF. Seventh, although DPC data generally contain in-hospital resource utilisation information, our analysis dataset lacked cost components or resource use data for patients who died during the index admission. Consequently, healthcare resource utilisation among non-survivors could not be evaluated. Future studies linking DPC claims, long-term care insurance databases, or social care registries will be essential for a more comprehensive assessment of the economic impact of MLTCs in individuals with HF. In conclusion, sex differences in the clinical characteristics were evident among those with a high MLTC burden in individuals with HF. Among older individuals with a high MLTC burden, men were at a higher risk of adverse outcomes compared with women. Recognising and addressing age-dependent sex differences is essential for improving management strategies and outcomes in individuals with HF and MLTCs. Supplementary material Acknowledgements The authors extend their sincere thanks to all individuals involved in patient care, including emergency staff, technicians, medical engineers, nurses, pharmacists, physicians and surgeons at Nippon Medical School Hospital, Musashi-Kosugi Hospital, Tama Nagayama Hospital, Chiba Hokusoh Hospital and Nippon Medical School. Footnotes Funding: This study was supported by the Nippon Medical School Grant-in-Aid for Medical Research (grant number: 23K00002). Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Ethics approval: This study involves human participants and was approved by the central ethics review committee of Nippon Medical School (approval number: M-2024-178) and was conducted in accordance with the Declaration of Helsinki. Participants gave informed consent to participate in the study before taking part. Data availability statement No data are available. References Associated Data Supplementary Materials Data Availability Statement No data are available.