Prognostic Impact of Cardiovascular‐Kidney‐Metabolic Overlap Among Patients With Acute Heart Failure: Insights From a Real‐World Multinational Cohort

Abstract

Background

The prognostic significance of cardiovascular‐kidney‐metabolic (CKM) multimorbidity in acute heart failure (AHF) remains unclear. We evaluated its association with clinical outcomes in a real‐world, multinational cohort of patients with AHF.

Methods

We included patients with AHF hospitalized from 2005 to 2020 at 2 tertiary care centers in Switzerland and Kyrgyzstan. Patients were categorized based on the number of CKM conditions in addition to HF (0–3). CKM conditions included prior myocardial infarction, impaired kidney function (estimated glomerular filtration rate <60 mL/min/m²), and diabetes. The association between the number of CKM conditions and the primary outcome of HF hospitalization or all‐cause death was assessed with Cox models.

Results

Among 1745 patients (69% Swiss, 44% female, age 74±13 years, left ventricular ejection fraction 43±16%), at baseline, 18.9% had HF alone, 39.0% had 1 CKM condition, 30.7% had 2, 11.4% had 3. Impaired kidney function was the most prevalent CKM condition (n=1011), followed by prior myocardial infarction (n=742) and diabetes (n=595). At 1 year, 731 patients (41.9%) experienced the primary outcome. Compared with HF alone, an increasing number of CKM conditions was associated with stepwise increases in the risk of the primary outcome (adjusted hazard ratios [HRs], 1.27 [95% CI, 1.01–1.59], 1.50 [95% CI, 1.19–1.89], and 2.10 [95% CI, 1.59–2.77] for 1, 2, and 3 CKM conditions). These associations remained consistent regardless of sex, age, and left ventricular ejection fraction (P _interaction>0.05 for all).

Conclusions

In this real‐world cohort of hospitalized patients with AHF, CKM multimorbidity was common and associated with a stepwise increase in the risk of adverse outcomes. These data underscore the importance of comprehensive CKM assessment in an AHF setting.

Nonstandard Abbreviations and Acronyms

AHF

acute heart failure

CKM

cardiovascular‐kidney‐metabolic

SGLT2i

sodium‐glucose cotransporter‐2 inhibitor

Clinical Perspective.

What Is New?

• Cardiovascular‐kidney‐metabolic multimorbidity, long recognized as prognostically important in chronic heart failure (HF), is also highly prevalent and confers a stepwise increase in risk of death or HF hospitalization among patients admitted with acute HF.

What Are the Clinical Implications?

• Cardiovascular‐kidney‐metabolic burden provides a simple and powerful framework for risk stratification during acute HF hospitalizations, independent of age, sex, and left ventricular ejection fraction.

• Hospitalization for acute HF represents a critical window to identify high‐risk patients with cardiovascular‐kidney‐metabolic multimorbidity and to implement integrated, multidisciplinary strategies and guideline‐directed therapies targeting cardiovascular, renal, and metabolic dysfunction.

Cardiovascular‐kidney‐metabolic (CKM) syndrome is a recently recognized multidimensional construct that captures the intertwined burden of cardiovascular disease, chronic kidney disease (CKD), and metabolic disorders such as type 2 diabetes. ¹ This syndrome reflects not only the coexistence of risk factors but also their synergistic contribution to end‐organ dysfunction and adverse clinical outcomes. In patients with chronic heart failure (HF), particularly those with mildly reduced or preserved ejection fraction (EF), CKM multimorbidity is highly prevalent and associated with stepwise increases in the risk of hospitalizations and mortality. ² , ³ , ⁴

Despite increasing recognition of CKM multimorbidity and recent efforts to define its stages and guide integrated care strategies, ¹ most evidence to date has been derived from patients in an ambulatory or chronic setting. In contrast, its relevance in the acute setting remains largely unexplored. Acute HF (AHF) is characterized by rapid decompensation necessitating hospitalization and stabilization, whereas chronic HF reflects a progressive syndrome managed longitudinally. ⁵ Given these divergent trajectories, the prognostic implications of CKM multimorbidity in AHF warrant dedicated investigation, as hospitalizations for AHF represent periods of heightened clinical vulnerability where the cumulative burden of CKM conditions may further exacerbate the risk of adverse outcomes. Whether the stepwise risk associated with CKM multimorbidity observed in chronic ambulatory HF populations similarly applies to the AHF setting is unknown and warrants dedicated investigation.

We aimed to evaluate the prevalence and prognostic impact of CKM multimorbidity in a large, real‐world cohort of patients hospitalized for AHF across 2 tertiary care centers in Switzerland and Kyrgyzstan.

METHODS

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Study Design and Population

This was a prospective cohort study including consecutive patients hospitalized for AHF between 2005 and 2020 at 2 tertiary care centers with comprehensive HF programs: Lausanne University Hospital (Switzerland) and the National Center of Cardiology and Internal Medicine (Kyrgyzstan). These centers provide full‐spectrum HF care, including ambulatory, ward‐based, intensive, and interventional services. AHF was diagnosed according to European HF guidelines ⁶ applicable at the time of the study. The original study aimed to analyze 1‐year outcomes in patients hospitalized with AHF, comparing cohorts from Kyrgyzstan and Switzerland under European Society of Cardiology guideline‐based management. ⁷

Patients were included in the cohort if they had a confirmed hospitalization for AHF and available data on coronary artery disease, renal function at admission, and diabetes status or antidiabetic medication use. Patients were excluded if they were admitted for (1) acute pulmonary embolism, (2) severe valvular disease requiring percutaneous or surgical intervention, (3) chronic obstructive pulmonary disease exacerbation, (4) complex congenital heart disease, or (5) cardiac surgery within 3 months. Additional exclusions applied to patients with acute infectious, toxic, or metabolic conditions, pregnant individuals, or those with advanced malignancy and life expectancy <1 year.

Patient data, including laboratory values and comorbidities, were extracted from electronic medical records at both centers. Whenever feasible, variables were mapped to a common set of definitions (eg, AHF diagnosis according to European Society of Cardiology guidelines criteria ⁶ ) to enable pooled analyses. Echocardiography was performed or reviewed by board‐certified cardiologists. Data accuracy was verified in 20% of the study population, yielding a concordance rate of 99.7%. ⁷ Because SGLT2 (sodium‐glucose cotransporter‐2) inhibitors (SGLT2is) were not guideline recommended for HF during the study period, their use was considered as part of oral antidiabetic therapy.

Ethical Approval

The study protocol was approved by the respective institutional review boards at the 2 centers (National Center of Cardiology and Internal Medicine: Local Ethics Committee 2019‐7; Lausanne University Hospital: CER‐VD 2019‐1158) and conducted in accordance with the Declaration of Helsinki. All participants provided written informed consent.

Cardiovascular‐Kidney‐Metabolic Syndrome Definition

CKM multimorbidity at the time of the index AHF hospitalization was defined by the presence of one or more of the following:

• prior myocardial infarction (MI), defined as a documented diagnosis in the medical record based on history, discharge summaries, or cardiology reports, and included electrocardiogram, biomarker, or imaging evidence consistent with previous MI,

• impaired kidney function at admission (estimated glomerular filtration rate: <60 mL/min/1.73 m²),

• or documented type 2 diabetes (including use of insulin or oral antidiabetic medication).

Patients with a primary diagnosis of acute MI were excluded unless the hospitalization was coded as AHF; those with prior MI presenting with acute decompensated HF were included. Patients were categorized based on the number of CKM conditions at baseline in addition to AHF (0, 1, 2, or 3 conditions), using an approach consistent with earlier work. ² , ³ , ⁴

Study Outcomes

The primary outcome was a composite of all‐cause death or hospitalization for HF. Secondary outcomes were the individual components of the primary outcome—namely, all‐cause death alone, and HF hospitalization alone. All outcomes were analyzed using a time‐to‐first event approach. Data on subsequent HF admissions were obtained from each hospital’s electronic records, and 1‐year all‐cause mortality data were retrieved from national death registries in Kyrgyzstan and Switzerland (last accessed in December 2021).

We also assessed the use of prognostic HF therapies — including angiotensin‐converting enzyme inhibitors (ACEis), angiotensin receptor blockers, angiotensin receptors/neprilysin inhibitors, mineralocorticoid receptor antagonists (MRAs), and beta blockers — as well as the use of loop diuretics at hospital admission and discharge, and the changes between these time points. This analysis was performed among patients who were alive at both admission and discharge, in the overall cohort and according to left ventricular EF (LVEF, <40% versus ≥40%).

Statistical Analysis

Continuous variables are expressed as mean±SD or as median with interquartile range, as appropriate. Group comparisons for continuous variables were performed using the Kruskal–Wallis test for nonnormally distributed data and 1‐way ANOVA for normally distributed data. Categorical variables are presented as counts with percentages and compared using Fisher exact test. P values for trend were reported for comparisons across CKM categories. The association between the number of CKM conditions and outcomes of interest was assessed using Cox proportional hazards models, with the group without CKM comorbidities serving as the reference category. Kaplan–Meier curves were generated to illustrate survival across CKM subgroups. Analyses were stratified by country (Kyrgyzstan or Switzerland), and were both unadjusted and adjusted for age, sex, body mass index, atrial fibrillation or flutter, and hypertension. The selection of these covariates was guided by baseline differences between categories and clinical judgment. In sensitivity analyses, we defined impaired kidney function as an estimated glomerular filtration rate <60 mL/min/1.73 m² both at admission and at discharge. We then repeated the analyses assessing the association between the number of CKM conditions and the specified outcomes, restricting the population to patients who were alive at discharge. All statistical tests were 2 sided, with a significance threshold of P<0.05, and were performed using STATA, version 18 (StataCorp, TX, USA).

RESULTS

Patient Baseline Characteristics

A total of 1745 patients (age 74±13 years, 44% female, 69% from Switzerland, mean LVEF 43±16%) were included.

With respect to CKM distribution, 18.9% (n=330) had AHF without additional CKM conditions, 39.0% (n=681) had 1 CKM comorbidity, 30.7% (n=535) had 2, and 11.4% (n=199) had all 3. Impaired kidney function was the most prevalent CKM component (57.9%, n=1011), followed by prior MI (42.5%, n=742) and type 2 diabetes (34.1%, n=595). The most frequent combination of CKM conditions was impaired kidney function and prior MI (24.9%, n=434) (Figure 1).

Figure 1. Distribution of patients according to cardiovascular‐kidney‐metabolic conditions.

CKM indicates cardiovascular‐kidney‐metabolic; and MI, myocardial infarction.

Across increasing CKM multimorbidity groups (from 0 to 3 conditions), age and body mass index both increased with CKM burden (P for trend <0.001 for both, Table 1). Patients with more CKM conditions had lower heart rates at admission (P for trend <0.001) and were more frequently enrolled in Switzerland (P for trend=0.003). Key comorbidities—such as atrial fibrillation or flutter, prior HF hospitalization, and hypertension—were significantly more prevalent with increasing CKM burden (all P for trend <0.001). Smoking and dyslipidemia were also more common among those with more CKM comorbidity (P for trend=0.018 and <0.001, respectively).

Table 1.

Patient Baseline Characteristics by Number of Cardiovascular‐Kidney‐Metabolic Conditions

	0 (n=330)	1 (n=681)	2 (n=535)	3 (n=199)	P for trend
Age, y	68.8±15.7	75.7±13.4	75.6±11.9	73.9±9.8	<0.001
Swiss cohort	202 (61.2)	478 (70.2)	378 (70.7)	150 (75.4)	0.003
Female sex	127 (38.5)	349 (51.2)	210 (39.3)	74 (37.2)	0.09
Kyrgyz cohort	38 (29.7)	82 (40.4)	60 (38.2)	28 (57.1)
Swiss cohort	89 (44.1)	267 (55.9)	150 (39.7)	46 (30.7)
Body mass index, kg/m2	26.8±7.5	27.2±6.0	28.2±6.2	29.8±5.9	<0.001
Systolic blood pressure, mm Hg	134.1±28.6	135.0±27.2	135.5±27.7	141.4±27.3	0.011
Diastolic blood pressure, mm Hg	80.3±16.7	80.1±17.6	77.8±17.0	79.4±18.2	0.09
Heart rate, beats/min	101.4±30.0	93.8±26.8	89.9±25.3	86.2±20.6	<0.001
Comorbidities
Previous myocardial infarction	0 (0.0)	194 (28.5)	349 (65.2)	199 (100.0)	By design
Atrial fibrillation/flutter	193 (58.5)	399 (58.6)	288 (53.8)	80 (40.2)	<0.001
Prior heart failure hospitalization	221 (67.0)	480 (70.5)	436 (81.5)	178 (89.4)	<0.001
Hypertension	213 (64.5)	537 (78.9)	455 (85.0)	186 (93.5)	<0.001
Diabetes	0 (0.0)	94 (13.8)	297 (55.5)	199 (100.0)	By design
Chronic obstructive pulmonary disease	76 (23.0)	155 (22.8)	126 (23.6)	53 (26.6)	0.38
Smoking	138 (41.8)	240 (35.3)	215 (40.2)	104 (52.3)	0.018
Dyslipidemia	112 (33.9)	304 (44.7)	315 (58.9)	155 (77.9)	<0.001
Hemoglobin, g/L	134.0±21.3	128.6±22.7	125.6±23.7	119.6±22.8	<0.001
Hemoglobin A1C, %	5.7±0.4	7.0±1.4	7.1±1.7	6.9±0.8	0.11
N‐terminal pro‐B‐type natriuretic peptide, pg/mL*	2933 [1524, 5582]	4659 [2319, 9628]	6113 [2941, 13 212]	5285 [2462, 10 896]	<0.001
Creatinine, umol/L	84.7±17.6	119.0±62.4	153.2±104.5	172.6±88.4	By design
Estimated glomerular filtration rate, mL/min/1.73 m2	77.9±13.2	56.3±21.0	47.1±21.0	37.9±13.1	By design
Echo parameters
LVEF, %	44.0±16.6	44.8±16.2	42.1±15.6	40.8±13.8	0.002
LVEF <40%	129 (39.1)	269 (39.5)	244 (45.6)	88 (44.2)	0.039
LV end‐diastolic diameter, mm	56.6±11.4	53.7±11.8	55.2±9.4	56.7±8.3	0.79
LV mass index, g/m2	124.0±47.6	119.2±43.6	120.3±36.7	125.9±33.4	0.81
Treatment
Statin	54 (16.4)	180 (26.4)	224 (41.9)	115 (57.8)	<0.001
Beta blocker	118 (35.8)	300 (44.1)	286 (53.5)	122 (61.3)	<0.001
Angiotensin‐converting enzyme inhibitor	74 (22.4)	186 (27.3)	178 (33.3)	76 (38.2)	<0.001
Angiotensin receptor blocker	59 (17.9)	135 (19.8)	122 (22.8)	55 (27.6)	0.004
Angiotensin receptor/neprilysin inhibitor	0 (0.0)	4 (0.6)	1 (0.2)	1 (0.5)	0.68
Mineralocorticoid receptor antagonist	52 (15.8)	121 (17.8)	108 (20.2)	35 (17.6)	0.25
Loop diuretics	116 (35.2)	325 (47.7)	320 (59.8)	129 (64.8)	<0.001
Oral antidiabetics	0 (0.0)	52 (7.6)	141 (26.4)	84 (42.2)	By design
Insulin therapy	0 (0.0)	27 (4.0)	114 (21.3)	101 (50.8)	By design
Implantable cardioverter‐defibrillator	9 (2.7)	12 (1.8)	20 (3.7)	14 (7.0)	0.003
Pacemaker	20 (6.1)	59 (8.7)	55 (10.3)	23 (11.6)	0.014

Results are expressed as n (%), mean±SD, or median [25th percentile, 75th percentile].

^*Missing in 901 (51.6%) patients.

LV indicates left ventricular; and LVEF, left ventricular ejection fraction.

From a biological standpoint, NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide) significantly increased, and hemoglobin levels decreased with higher CKM burden. Patients with greater CKM burden had significantly lower LVEF (P for trend=0.002); no significant differences were observed for LV end‐diastolic diameter and LV mass index. Baseline patient characteristics by country are provided in Tables S1 and S2.

Outcomes

The primary composite outcome occurred in 731 patients (41.9%) at 1 year. Compared with patients with AHF alone, the risk of the primary outcome increased stepwise with a greater number of overlapping CKM conditions, both in unadjusted and adjusted analyses, with adjusted hazard ratios (HRs) of 1.27 (95% CI, 1.01–1.59) for 1 condition, 1.50 (95% CI, 1.19–1.89) for 2, and 2.10 (95% CI, 1.59–2.77) for 3 (Table 2, Figures 2 and 3). These associations remained consistent regardless of sex (P _interaction=0.57), age (categorized at median: <77 and ≥77 years, P _interaction=0.43), and LVEF (<40% and ≥40%, P _interaction=0.42). They were also observed in both the Swiss and Kyrgyz cohorts, with generally stepwise increases in risk across CKM categories (Table S3). Moreover, these findings were generally consistent when first HF hospitalization and all‐cause death were analyzed separately, showing a stepwise increase in risk with a higher number of CKM conditions (Table 2, Figure 3, Figures S1 and S2). However, the presence of a single CKM condition was not significantly associated with either outcome in adjusted models. These results remained generally consistent when defining impaired kidney function as low estimated glomerular filtration rate at both admission and discharge (n=1654, Table S4).

Table 2.

Event Numbers, Event Rates, and Hazard Ratios Associated With Number of Cardiovascular‐Kidney‐Metabolic Conditions

	Events, n (%)	Event rate (95% CI), per 100 py	Unadjusted HR (95% CI)	Adjusted HR (95% CI), Model*
First HFH or all‐cause death
0 (n=330)	110 (33.3)	42.0 (34.9–50.7)	1 (reference)	1 (reference)
1 (n=681)	275 (30.4)	57.7 (51.3–64.9)	1.33 (1.07–1.66)	1.27 (1.01–1.59)
2 (n=535)	240 (44.9)	68.5 (60.3–77.7)	1.56 (1.24–1.95)	1.50 (1.19–1.89)
3 (n=199)	106 (53.3)	87.8 (72.6–106.2)	1.96 (1.50–2.56)	2.10 (1.59–2.77)
First HFH
0 (n=330)	77 (23.3)	27.0 (21.6–33.7)	1 (reference)	1 (reference)
1 (n=681)	203 (29.8)	36.5 (31.8–41.8)	1.31 (1.00–1.70)	1.25 (0.96–1.64)
2 (n=535)	163 (30.5)	38.3 (32.9–44.7)	1.37 (1.04–1.79)	1.32 (1.00–1.74)
3 (n=199)	75 (37.7)	51.3 (40.9–64.4)	1.80 (1.31–2.47)	1.80 (1.30–2.51)
All‐cause death
0 (n=330)	56 (17.0)	19.0 (14.7–24.7)	1 (reference)	1 (reference)
1 (n=681)	144 (21.1)	25.7 (21.8–30.2)	1.31 (0.96–1.79)	1.23 (0.90–1.69)
2 (n=535)	136 (25.4)	31.8 (26.9–37.6)	1.61 (1.18–2.20)	1.55 (1.13–2.14)
3 (n=199)	52 (26.1)	31.9 (24.3–41.9)	1.59 (1.09–2.32)	1.86 (1.25–2.76)

P _interaction between sex and categories of number of cardiovascular‐kidney‐metabolic conditions regarding first HFH or all‐cause death, 0.57. P _interaction between age categories (<77 years, ≥77 years) and number of cardiovascular‐kidney‐metabolic conditions regarding first HFH or all‐cause death, 0.43. P _interaction between left ventricular ejection fraction (<40%, ≥40%) and number of cardiovascular‐kidney‐metabolic conditions regarding first HFH or all‐cause death, 0.42. HFH indicates heart failure hospitalization; HR, hazard ratio; and py, patient‐years.

^*Adjusted for age, sex, body mass index, atrial fibrillation or flutter, hypertension. All analyses are stratified by country (Switzerland or Kyrgyzstan).

Figure 2. Cumulative incidence of the composite outcome of first heart failure hospitalization or all‐cause death according to the number of cardiovascular‐kidney‐metabolic conditions.

CKM indicates cardiovascular‐kidney‐metabolic; and HFH, heart failure hospitalization.

Figure 3. Hazard ratios for outcomes stratified by number of cardiovascular‐kidney‐metabolic conditions. Same adjustment model as in Table 2.

CKM indicates cardiovascular‐kidney‐metabolic; and HFH, heart failure hospitalization.

Impaired kidney function—whether present alone or in combination with diabetes or prior MI—was consistently associated with the highest risk across all outcomes, including the primary composite, HF hospitalization, and all‐cause death. Adjusted HRs for the primary outcome were 1.35 (95% CI, 1.05–1.74) for impaired kidney function alone, 1.59 (95% CI, 1.19–2.14) when combined with diabetes, and 1.53 (95% CI, 1.17–2.00) when combined with prior MI. In contrast, diabetes or prior MI alone were not significantly associated with excess risk. The combination of all 3 CKM conditions was associated with the greatest risk across all end points, with adjusted HRs of 2.09 (95% CI, 1.58–2.76) for the primary outcome, 1.81 (95% CI, 1.30–2.51) for HF hospitalization, and 1.84 (95% CI, 1.24–2.74) for all‐cause death. Notably, the combination of impaired kidney function and prior MI was also strongly associated with excess all‐cause mortality (HR, 1.77; 95% CI, 1.24–2.53) (Table 3).

Table 3.

Rate of Primary and Secondary End Points Stratified by Type of Cardiovascular‐Kidney‐Metabolic Overlap

Outcome	None (n=330)	Diabetes only (n=96)	Impaired kidney function only (n=391)	Prior MI only (n=194)	Diabetes and impaired kidney function (n=186)	Diabetes and prior MI (n=114)	Impaired kidney function and prior MI (n=235)	All 3 conditions (n=199)
Primary composite
Events	110	26	169	80	85	45	110	106
Event rates (95% CI)	42.0 (34.9–50.7)	33.7 (22.9–49.5)	65.7 (56.5–76.4)	56.2 (45.1–69.9)	70.9 (57.3–87.7)	52.1 (38.9–69.8)	76.2 (63.3–91.9)	87.8 (72.6–106.2)
Adjusted HR (95% CI)	1 (reference)	0.91 (0.59–1.41)	1.35 (1.05–1.74)	1.29 (0.97–1.73)	1.59 (1.19–2.14)	1.32 (0.93–1.88)	1.53 (1.17–2.00)	2.09 (1.58–2.76)
Heart failure hospitalization
Events	77	19	125	59	57	39	67	75
Event rates (95% CI)	27.0 (21.6–33.7)	22.5 (14.3–35.2)	40.5 (33.9–48.2)	36.2 (28.0–46.7)	39.1 (30.2–50.7)	42.8 (31.3–58.5)	35.5 (28.0–45.1)	51.3 (40.9–64.4)
Adjusted HR (95% CI)	1 (reference)	0.82 (0.49–1.38)	1.35 (1.00–1.81)	1.29 (0.92–1.82)	1.36 (0.96–1.94)	1.57 (1.06–2.33)	1.18 (0.84–1.65)	1.81 (1.30–2.51)
All‐cause death
Events	56	12	97	35	45	17	74	52
Event rates (95% CI)	19.0 (14.7–24.7)	14.0 (8.0–24.7)	31.4 (25.7–38.3)	21.0 (15.1–29.3)	29.6 (22.1–39.7)	16.5 (10.3–26.6)	42.7 (34.0–53.7)	31.9 (24.3–41.9)
Adjusted HR (95% CI)	1 (reference)	0.95 (0.50–1.78)	1.37 (0.97–1.95)	1.11 (0.72–1.69)	1.60 (1.06–2.40)	1.01 (0.58–1.75)	1.77 (1.24–2.53)	1.8 (1.24–2.74)

Same adjustment model as in Table 2. HR indicates hazard ratio; and MI, myocardial infarction.

Heart Failure Medication Use According to CKM Burden

A total of 1649 patients (94.5% of 1745) were alive at discharge and had available data on medication use at both admission and discharge.

At admission, 115 patients (35.7%) with no CKM condition were on beta blockers, compared with 287 (45.1%), 270 (54.0%), and 115 (60.5%) among those with 1, 2, or 3 CKM conditions, respectively (P value <0.001, Table 4). A similar stepwise increase was observed for loop diuretics, with use rising from 34.2% to 64.7% across the CKM spectrum (P value <0.001). ACEi use gradually increased from 22.7% in patients without CKM conditions to 38.9% in those with 3 (P for trend <0.001). Angiotensin receptor blocker, angiotensin receptor/neprilysin inhibitor and MRA use did not differ significantly across groups.

Table 4.

Number, Proportion, and Relative Changes in Heart Failure Prognostic Medications and Loop Diuretics by Number of Cardiovascular‐Kidney‐Metabolic Conditions, Patients Alive at Admission and Discharge

	Number of CKM conditions				P value
	0 (n=322)	1 (n=637)	2 (n=500)	3 (n=190)
Beta blocker
At admission	118 (35.7)	287 (45.1)	270 (54.0)	115 (60.5)	<0.001
At discharge	220 (68.3)	425 (65.7)	352 (70.4)	143 (75.3)	0.14
Relative change (in %)	+91.3	+48.1	+30.4	+24.3
Angiotensin‐converting enzyme inhibitor
At admission	73 (22.7)	172 (27.0)	167 (33.4)	74 (38.9)	<0.001
At discharge	124 (38.5)	254 (39.9)	239 (47.8)	84 (44.2)	0.020
Relative change (in %)	+69.9	+47.7	+43.1	+13.5
Angiotensin receptor blocker
At admission	59 (18.3)	131 (20.6)	108 (21.6)	52 (27.4)	0.11
At discharge	63 (19.6)	138 (21.7)	107 (21.4)	45 (23.7)	0.74
Relative change (in %)	+6.8	+5.3	−0.9	−13.5
Angiotensin receptor/neprilysin inhibitor
At admission	0 (0.0)	3 (0.5)	1 (0.2)	1 (0.5)	0.56
At discharge	1 (0.3)	3 (0.5)	2 (0.4)	1 (0.5)	0.98
Relative change (in %)	Not applicable	0.0	+100.0	0.0
Mineralocorticoid receptor antagonist
At admission	49 (15.2)	114 (17.9)	98 (19.6)	34 (17.9)	0.46
At discharge	163 (50.6)	253 (39.7)	155 (31.0)	47 (24.7)	<0.001
Relative change (in %)	+232.7	+121.9	+58.2	+38.2
Loop diuretics
At admission	110 (34.2)	296 (46.5)	294 (58.8)	123 (64.7)	<0.001
At discharge	277 (86.0)	550 (86.3)	438 (87.6)	168 (88.4)	0.80
Relative change (in %)	+151.8	+85.8	+49.0	+36.6

Results are expressed as n (%) unless specified otherwise. The relative change for each medication class was calculated as: (proportion at discharge ‐ proportion at admission)/proportion at admission. CKM indicates cardio‐kidney‐metabolic.

Between admission and discharge, beta‐blocker use increased from 35.7% to 68.3% (+91.3%) in patients who were CKM naïve, compared with a more modest rise in patients with 3 CKM conditions (60.5% to 75.3%; +24.3%) (Table 4). Similar trends were observed for ACEi and loop diuretics, where relative increases in in‐hospital implementation were most pronounced in those without CKM multimorbidity, with diminishing increments as the CKM burden increased.

Among patients with HF and an LVEF <40%, the use of guideline‐directed therapies (beta blockers, ACEis/ angiotensin receptor blockers, MRAs), and loop diuretics at admission, was significantly higher with increasing CKM burden (Table S5). Beta‐blocker use at admission increased from 32.6% in patients with no CKM condition to 54.2% in those with 3 CKM conditions (P value <0.001). Similar results were observed for loop diuretics and angiotensin receptor blockers, whereas the use of MRAs and ACEis did not differ significantly. Despite higher baseline use in patients with high CKM burden, the change in in‐hospital implementation between admission and discharge was most marked in individuals who were CKM naïve—for example, beta‐blocker use increased from 32.6% to 81.4% (+150%) in those without CKM compared with an increase from 54.2% to 71.1% (+31%) in those with all 3 CKM conditions. Similar patterns were found for the other medication classes.

In patients with HF with an LVEF ≥40%, the same stepwise increase in baseline medication use with CKM burden was observed—beta blockers ranged from 37.8% (0 CKM) to 65.4% (3 CKM conditions), and loop diuretics from 34.7% to 63.6% (both P values <0.001, Table S6). ACEi use also significantly increased with CKM burden. Again, the greatest increase in in‐hospital implementation was seen in patients without CKM multimorbidity for all medication classes.

DISCUSSION

In this multinational, real‐world cohort of patients hospitalized for AHF, a greater burden of CKM multimorbidity was associated with a clear stepwise increase in the risk of adverse clinical events. This gradient of risk was consistent across age groups, sex, and HF phenotypes, highlighting the broad prognostic relevance of CKM burden in the acute setting. Moreover, although patients with higher CKM burden were more likely to be on guideline‐recommended treatments at admission, they were significantly less likely to receive in‐hospital initiation of new therapies. Stepwise risk increments across the CKM spectrum were observed in both reduced and mildly/preserved LVEF.

Although the CKM construct has been increasingly recognized as a major determinant of prognosis in chronic HF, particularly in populations with HF with mildly reduced EF and HF with preserved EF, ² , ³ , ⁴ its relevance in AHF has remained unexplored. Our study addresses this gap by showing that CKM multimorbidity is not only prevalent but also strongly prognostic in the acute setting, with a clear stepwise increase in the risk of death or HF hospitalization as the number of overlapping CKM conditions increases. The magnitude of risk increase observed with each additional CKM condition was broadly in line with prior reports in chronic heart failure populations. ² , ³ , ⁴ These findings extend the relevance of the CKM framework into this period of heightened vulnerability and underscore the need for earlier detection and integrated management strategies across the spectrum of HF care. ¹

The acute setting may, however, differ significantly from the chronic context, particularly with regard to renal function. Indeed, worsening renal function in the context of AHF is not equivalent to CKD in a stable clinical setting. In our study, impaired kidney function may have reflected either preexisting CKD or worsening renal function. The prognostic significance of worsening renal function in AHF is complex and largely influenced by the patient’s volume status ⁸ , ⁹ —an aspect we were unable to assess. Still, baseline renal function was independently associated with worse outcomes in our study, even when considered in isolation. This likely reflects that patients presenting with impaired kidney function may have had underlying CKD or worsening renal function as part of a cardiorenal syndrome at admission—both of which are well‐established markers of poor prognosis. ⁸ Importantly, our findings remained consistent when using a more conservative definition of impaired kidney function, based on reduced estimated glomerular filtration rate at both admission and discharge.

Interestingly, patients with higher CKM multimorbidity were more likely to receive prognostic HF medications at baseline. However, despite this higher baseline treatment rate, these patients experienced significantly less in‐hospital implementation of key therapies (notably beta blockers, MRAs, and loop diuretics), compared with those with less CKM multimorbidity. This may be largely attributable to impaired kidney function, which remains a major limiting factor for initiating and titrating prognostic HF therapies. ¹⁰ , ¹¹ Other factors, such as frailty (which was not assessable in our study), may also hinder optimal implementation of therapy in this particular population.

Taken together, our findings have several implications. First, the CKM framework provides an important risk stratification tool not only in chronic HF but also in AHF, supporting the need for early identification and prevention strategies before decompensation occurs. Second, hospitalization for AHF may represent a critical window of opportunity to proactively address CKM multimorbidity—not only by initiating or optimizing guideline‐directed therapies, but also by implementing multidisciplinary strategies aimed at managing metabolic, renal, and cardiovascular comorbidities in a coordinated and integrated fashion. Several therapies that were not yet available at the time of our study have since demonstrated both efficacy and safety—not only in individual CKM conditions but also across the broader CKM spectrum, both in HF with reduced EF and preserved EF. For instance, SGLT2is have shown consistent cardiovascular benefits in patients with HF with or without diabetes, ¹² , ¹³ irrespective of baseline renal function, ¹⁴ , ¹⁵ , ¹⁶ and across the entire CKM continuum. ³ Similar findings have been observed with finerenone ¹⁷ , ¹⁸ , ¹⁹ and sacubitril/valsartan. ² , ²⁰ Moreover, these agents suggest potential kidney‐protective effects in the long run, ¹⁴ , ¹⁵ , ¹⁶ , ²¹ , ²² and the use of SGLT2is, sacubitril/valsartan, and finerenone has been shown to be associated with better glycemic control and a lower risk of incident diabetes among patients with HF. ²³ , ²⁴ , ²⁵ SGLT2is, sacubitril/valsartan, and MRAs are considered safe to initiate even during acute HF episodes, ²⁶ , ²⁷ , ²⁸ and finerenone is currently being tested among patients with AHF in the ongoing REDEFINE‐HF (A Study to Determine the Efficacy and Safety of Finerenone on Morbidity and Mortality Among Hospitalized Heart Failure Patients) trial (clinicaltrials.gov ID: NCT06008197). Initiating these treatments during the acute phase may therefore offer a dual opportunity: improving HF outcomes while also addressing the broader metabolic and renal dysfunction characteristic of CKM syndrome. Patients with high CKM burden may require alternative care models (eg, interdisciplinary or transitional care programs) to overcome barriers to therapeutic optimization during and after hospitalization. ¹ Finally, new CKM outcomes trials are underway, specifically in patients presenting with worsening HF such as the BalanceD‐HF (Study to Evaluate the Effect of Balcinrenone/Dapagliflozin in Patients With Heart Failure and Impaired Kidney Function) trial (ClinicalTrials.gov ID: NCT06307652), which is enrolling patients with both HF and impaired kidney function.

Our study has several limitations. First, data on specific cardiometabolic therapies—including SGLT2is and GLP‐1 (glucagon‐like peptide‐1) receptor agonists—were not systematically collected, and their use was likely low in this historical cohort. Similarly, angiotensin receptor/neprilysin inhibitor use was minimal. As a result, the prognostic implications for patients with similar clinical profiles but higher use of these contemporary therapies remain uncertain. Second, the long study period (2005–2020) encompassed evolving diagnostic criteria and treatment strategies for HF, which may have influenced outcomes despite stratification and adjustment for potential confounders. Third, due to data constraints, our definition of CKM conditions was deliberately stringent and did not include other relevant components such as peripheral artery disease, potentially underestimating the true prevalence of CKM multimorbidity. Nevertheless, approximately 81% of patients had at least 1 CKM comorbidity, a proportion that approaches contemporary estimates. ²⁹ Fourth, we could not ascertain whether all HF therapies at baseline were prescribed specifically for HF or for other concomitant indications (such as hypertension, or atrial fibrillation or flutter), which may limit the precision of interpreting treatment patterns. Finally, although we assessed the presence of prognostic therapies at admission and discharge, we lacked information on the doses used, limiting our ability to evaluate the quality and aggressiveness of therapeutic optimization.

CONCLUSIONS

CKM syndrome was common in this real‐world cohort of more than 1700 patients hospitalized for AHF and was associated with a stepwise increase in the risk of adverse cardiovascular events. These findings emphasize the prognostic relevance of CKM syndrome in the acute setting, extending its significance beyond chronic HF. Hospitalization for acute HF may represent a critical window of opportunity to proactively address CKM multimorbidity—not only by identifying patients at highest risk but also by implementing coordinated, multidisciplinary strategies that target the intertwined cardiovascular, renal, and metabolic components of this syndrome.

Sources of Funding

This study was supported by funds from the Centre hospitalier universitaire vaudois, enabling prospective data collection for the cohort 2005–2009. The data collection for the study cohort 2015–2018 was made possible by funds from the Schweizerische Herzstiftung attributed to Pierre Monney, MD, and Roger Hullin, MD.

Disclosures

Dr Lu has received research grant support or served as advisory boards for Bayer, AstraZeneca, Boehringer Ingelheim, Cytokinetics and Abbott. Dr Pitta Gros has received research grant support from Amgen, Novo Nordisk, and Abbott. Dr Vaduganathan has received research grant support, served on advisory boards, or had speaker engagements with American Regent, Amgen, AstraZeneca, Bayer AG, Baxter Healthcare, BMS, Boehringer Ingelheim, Chiesi, Cytokinetics, Fresenius Medical Care, Idorsia Pharmaceuticals, Lexicon Pharmaceuticals, Merck, Milestone Pharmaceuticals, Novartis, Novo Nordisk, Pharmacosmos, Relypsa, Roche Diagnostics, Sanofi, and Tricog Health, and participates on clinical trial committees for studies sponsored by Amgen, AstraZeneca, Galmed, Novartis, Bayer AG, Occlutech, and Impulse Dynamics. The other authors do not have any conflict of interest in relation with this article.

Supporting information

Data S1