Highlights
-
•
Heart failure patients with Class 3 obesity experience higher economic burden.
-
•
Patients with Class 3 obesity had more acute care visits than those with overweight.
-
•
Patients with Class 3 obesity had increased 30-day readmissions rates.
-
•
Significant economic burden is posed by severe obesity in the HFpEF population.
Keywords: Heart failure with preserved ejection fraction, BMI, Health care costs
Abstract
Introduction
Obesity-related HFpEF is linked to worse quality of life and health outcomes; however, the association of obesity with real-world healthcare resource utilization (HCRU) and economic burden in HFpEF is unknown. This study assessed the association between BMI and HCRU- and cost-related outcomes among patients with HFpEF.
Methods
This claims-based study utilized the Optum Market Clarity database to identify adult commercial, Medicare, and Medicaid patients with HFpEF based on EF ≥ 50% and ≥ 1 diagnosis of HF from Oct 01, 2016 to Mar 31, 2023. Baseline clinical characteristics, post-index rehospitalization rates, and per-patient-per-year HCRU and cost data were described. Associations between BMI and HCRU and cost outcomes was examined using generalized linear models adjusted for baseline characteristics.
Results
A total of 95,070 patients were included; 14.4% were with normal weight, 26.1% with overweight, 23.8% Class 1 obesity, 15.9% Class 2 obesity, and 19.9% Class 3 obesity. Across cohorts, patients with Class 2 and Class 3 obesity had high rates of all-cause and HF-related rehospitalizations. Patients with Class 2 and Class 3 obesity had higher rates of HF-related hospitalization/ER visits (both p < 0.001), compared to the overweight cohort, when adjusted for baseline covariates. Across comparisons, patients with Class 3 obesity had numerically higher HF-related costs compared with the overweight cohort.
Conclusions
This study demonstrated that patients with HFpEF and Class 3 obesity incurred higher healthcare costs, more frequent all-cause and HF-related acute care encounters, in general, and increased readmissions rates, compared to patients in other BMI groups.
1. Introduction
Heart failure with preserved ejection fraction (HFpEF) is a complex clinical syndrome affecting more than 3 million people in the US and up to 32 million people worldwide.[1] It is mainly characterized by the presence of diastolic dysfunction (DD) and elevated left ventricular (LV) filling pressure in the setting of a LV ejection fraction (LVEF) ≥ 50%, [2] and is generally associated with elevated natriuretic peptides, including B‑type natriuretic peptide (BNP) and N‑terminal proBNP (NT‑proBNP).[3] People with HFpEF are hospitalized approximately 1.4 times per year and have a mortality rate of approximately 15% per year.[1] The majority of patients with HFpEF also have obesity, and it has become clear that visceral adiposity contributes to the evolution and progression of HFpEF.[4], [5].
Among patients with HFpEF, the prevalence of overweight and obesity is particularly high. Data from the I-PRESERVE clinical trial, which evaluated the efficacy of irbesartan in older patients with HFpEF, revealed that more than 83% of the participants had overweight or obesity.[6] Obesity-related HFpEF is increasingly recognized as a distinct clinical phenotype, characterized by excess visceral and ectopic adiposity that drives inflammation, hypertension, insulin resistance, and dyslipidemia, while also impairing diastolic, systolic, arterial, skeletal muscle, and physical function.[7], [8], [9] The risk of HFpEF increases as body-mass index (BMI) increases.[4], [10] Weight loss interventions, like treatment with GLP-1 receptor agonist & bariatric surgery, have shown to reduce the risk of incident HFpEF and to alleviate symptoms in people with established HFpEF.[11], [12], [13].
Despite epidemiological evidence of a steadily increasing prevalence, both in absolute terms and relative to the entire HF population, a prompt diagnosis of HFpEF is still challenging. Consequently, HFpEF is associated with a substantial economic burden. In the United States, HF results in annual direct health care costs of approximately $39.2 billion to $60 billion.[14] In a 2020 systematic review, the median annual total health care costs associated with HF were estimated to be $24,383 per patient, with HF-specific hospitalizations accounting for a median of $15,879 per patient.[15].
Given the high and rising prevalence of obesity among patients with HFpEF and its potential adverse impact on health outcomes related to HFpEF, it is essential to understand the association of BMI with real-world healthcare utilization and economic burden associated with HFpEF in adult patients. The objectives of this study were 1) assess acute care events among patients with HFpEF by BMI level and stratified by insurance type, 2) assess rehospitalizations among patients with HFpEF by BMI level, overall and stratified by insurance type, and 3) determine the association between BMI level and health care resource utilization (HCRU)-based and economic outcomes, adjusting for clinical and demographic characteristics.
2. Methods
2.1. Study overview
This was a retrospective cohort study utilizing the Optum’s de-identified Market Clarity Data (Optum® Market Clarity) between October 1, 2015 and March 31, 2023. Optum® Market Clarity is an integrated, multi-source medical claims, pharmacy claims, and electronic health records (EHR) data set, which links EHR data − including lab results, vital signs and measurements, diagnoses, procedures and information derived from unstructured clinical notes using natural language processing − with linked historical administrative claims data − including pharmacy claims, physician claims, clinical information facility claims and medications prescribed and administered. Optum® Market Clarity is statistically de-identified under the HIPAA Privacy Rule’s Expert Determination method and managed according to Optum® customer data use agreements. Institutional review board approval or waiver of approval was not required for this study because the study data were secondary and de-identified in accordance with the United States Department of Health and Human Services Privacy Rule’s requirements for de-identification codified at 45C.F.R. § 164.514(b). Throughout the process, patient privacy was preserved, and researchers complied strictly with all applicable Health Insurance Portability and Accountability Act data management rules and the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
2.2. Study inclusion and exclusion criteria
The study included adult patients with commercial, Medicare, or Medicaid coverage who had ≥ 1 medical claim with HF diagnosis between October 1, 2016 and March 31, 2023 (identification period). To assess healthcare utilization and cost outcomes in patients with prevalent HF, the index date was defined as the date a patient accrued ≥ 12 months of continuous enrollment (baseline period) following their first observed HF diagnosis. Follow-up began on the index date and continued until the earliest of death, disenrollment, five years post-index, or the end of the study period. Patients were included if they had evidence of HFpEF, defined as LVEF ≥ 50%, based on the closest measurement within ± 12 months of the index date.[8], [14] Exclusion criteria included ≥ 1 medical claim for cardiomyopathy during baseline, unknown or underweight BMI (BMI < 20 kg/m2), and missing demographic data.
2.3. Study variables
2.3.1. Baseline variables
Patient demographics were assessed as of the index date. Clinical characteristics, including BMI, Charlson Comorbidity Index (CCI) score, select comorbidities, and New York Heart Association (NYHA) functional class (severity measure of HFpEF), were evaluated over the 12-month baseline period. For patients with multiple BMI values during this window (including the index date), the measurement closest to the index date was captured. NYHA class, available in EHR data for a subset of patients, was reported as a marker of HFpEF severity. Patients were stratified into five BMI-based cohorts: 1) Class 3 obesity (BMI ≥ 40 kg/m2), 2) Class 2 obesity (35 kg/m2 to < 40 kg/m2), 3) Class 1 obesity (30 kg/m2 to < 35 kg/m2), 4) Overweight (25 kg/m2 to < 30 kg/m2), and 5) Normal (20 kg/m2 to < 25 kg/m2).
2.3.2. Outcomes
All-cause and HF-related 30-day rehospitalization rates were reported among the subset of patients who experienced an initial hospitalization after the index date and had sufficient continuous enrollment through 30 days post-discharge.
All-cause and HF-related total HCRU and costs were assessed over the variable follow-up period among the subset of patients with at least 12 months of follow-up. Total costs included pharmacy and medical costs, the latter including ambulatory costs (physician office and hospital outpatient), emergency room (ER) costs, inpatient costs, and other medical costs. Hospitalization or ER visits were categorized as acute care events. HCRU and costs were considered HF-related if the claim had a diagnosis of HF in the primary position or if there was a fill or administration of an HFpEF-related medication. To account for variable follow-up durations, unadjusted acute care events, total all-cause healthcare costs, and HF-related healthcare costs were calculated as per-patient per-year (PPPY). PPPY was defined as the total number of events or costs divided by the total number of patient-years of follow-up. Costs were adjusted using the 2023 Consumer Price Index (CPI) for medical care.
2.3.3. Analyses
Baseline characteristics were summarized descriptively across BMI cohorts. Categorical variables were reported as frequencies and percentages, while continuous variables were described using means, medians, standard deviations, and interquartile ranges.
The association between BMI class and HF-related acute care events was assessed in the overall HFpEF population among those with 12 months of follow-up, using generalized linear regression model (GLM) with a negative binomial distribution, while HF-related total costs were assessed using GLM with a gamma distribution, stratified by insurance type (commercial, Medicare, and Medicaid). All models were adjusted for NYHA class, age, sex, race, region, index year, baseline CCI score, comorbid conditions, HFpEF-related medication use, and baseline costs. Given the distinct clinical profiles of patients with obesity-related HFpEF versus non-obesity related HFpEF, the cohort with overweight was selected as the reference cohort for comparisons across BMI categories. Where applicable, results were stratified by insurance type.
3. Results
A total of 95,070 patients with evidence of HFpEF (LVEF ≥ 50%), were included in the final sample (Supplemental Fig. 1). Of these, 23,825 (25.1%) had commercial insurance, 58,471 (61.5%) were covered by Medicare, and 12,774 (13.4%) were covered by Medicaid. Patients were categorized by BMI as follows: 1) Normal weight, 14.4%, 2) overweight, 26.1%, 3) Class 1 obesity, 23.8%, 4) Class 2 obesity, 15.9%, and 5) Class 3 obesity, 19.9%. Among patients with commercial insurance, the mean (SD) follow-up duration was 820.8 days (622.3). For those with Medicare and Medicaid, the mean (SD) follow-up durations were 850.4 (623.3) and 775.9 (606.0) days, respectively. Among those with commercial insurance, the mean (SD) age was 57.7 (9.2) years and 41.8% were female. For patients with Medicare and Medicaid insurance, the mean (SD) age in years were 72.9 (8.3) and 53.5 (10.3) respectively and 53.5% and 54.0% were female, respectively (Table 1). The mean (SD) CCI score was 3.5 (1.8) in the overall commercial population, 4.4 (2.0) in the Medicare population, and 4.2 (2.0) in the Medicaid population. While most comorbidities were more commonly found among patients with obesity, acute kidney failure, myocardial infarction, stroke, chronic obstructive pulmonary disease, and respiratory failure, were generally higher in the Normal cohort compared to the obesity cohorts pointing to the distinct nature of non-obesity related HFpEF with obesity-related HFpEF. Among patients with HFpEF severity information (NYHA class) available, Class II severity was the most common across all insurance groups − 45.5% in commercial, 44.4% in Medicare, and 42.0% in Medicaid populations. Commercial patients were more frequently classified into NYHA Class I (28.5%), while approximately 31% of Medicare and Medicaid patients were classified as NYHA Class III (full demographics in Supplemental Tables 1-3).
Table 1.
Patient demographic and clinical characteristics.
| Normal | Overweight | Class 1 obesity | Class 2 obesity | Class 3 obesity | |
|---|---|---|---|---|---|
| Commercial, n (%) | 2,739 (11.5) | 5,799 (24.3) | 5,789 (24.3) | 3,961 (16.6) | 5,537 (23.2) |
| Age, mean (SD) | 57.8 (11.3) | 58.7 (9.5) | 58.2 (8.6) | 57.8 (8.4) | 56.0 (8.6) |
| Female sex, n (%) | 1,290 (47.1) | 2,039 (35.2) | 2,060 (35.6) | 1,650 (41.7) | 2,927 (52.9) |
| CCI, mean (SD) | 3.7 (2.1) | 3.4 (1.9) | 3.4 (1.8) | 3.5 (1.7) | 3.6 (1.7) |
| NYHA functional class | |||||
| Class I | 102/266 (38.4) | 183/568 (32.2) | 171/535 (32.0) | 75/318 (23.6) | 79/452 (17.5) |
| Class II | 97/266 (36.5) | 262/568 (46.1) | 231/535 (43.2) | 163/318 (51.3) | 220/452 (48.7) |
| Class III | 53/266 (19.9) | 104/568 (18.3) | 120/535 (22.4) | 70/318 (22.0) | 136/452 (30.1) |
| Class IV | 14/266 (5.3) | 19/568 (3.4) | 13/535 (2.4) | 10/318 (3.1) | 17/452 (3.8) |
| BMI, kg/m2 | |||||
| Valid n | 2,684 | 5,598 | 5,600 | 3,821 | 5,091 |
| Mean (SD) | 23.0 (1.4) | 27.6 (1.4) | 32.4 (1.4) | 37.3 (1.4) | 47.6 (7.4) |
| Medicare, n (%) | 9,104 (15.6) | 16,199 (27.7) | 14,338 (24.5) | 9,159 (15.7) | 9,671 (16.5) |
| Age, mean (SD) | 74.7 (8.1) | 74.4 (7.7) | 73.3 (7.6) | 71.9 (8.0) | 68.9 (9.0) |
| Female sex, n (%) | 4,944 (54.3) | 7,624 (47.1) | 7,108 (49.6) | 5,162 (56.4) | 6,425 (66.4) |
| CCI, mean (SD) | 4.5 (2.2) | 4.3 (2.1) | 4.3 (2.0) | 4.4 (2.0) | 4.4 (1.9) |
| NYHA functional class | |||||
| Class I | 194/807 (24.0) | 342/1,427 (24.0) | 265/1,218 (21.8) | 132/807 (16.4) | 103/755 (13.6) |
| Class II | 345/807 (42.8) | 631/1,427 (44.2) | 555/1,218 (45.6) | 373/807 (46.2) | 322/755 (42.7) |
| Class III | 230/807 (28.5) | 405/1,427 (28.4) | 357/1,218 (29.3) | 264/807 (32.7) | 296/755 (39.2) |
| Class IV | 38/807 (4.7) | 49/1,427 (3.4) | 41/1,218 (3.4) | 38/807 (4.7) | 34/755 (4.5) |
| BMI, kg/m2 | |||||
| Valid n | 8,872 | 15,637 | 13,798 | 8,837 | 8,659 |
| Mean (SD) | 22.9 (1.4) | 27.5 (1.4) | 32.3 (1.4) | 37.2 (1.4) | 46.5 (6.5) |
| Medicaid, n (%) | 1,824 (14.3) | 2,774 (21.7) | 2,529 (19.8) | 1,964 (15.4) | 3,683 (28.8) |
| Age, mean (SD) | 53.9 (11.9) | 54.7 (10.7) | 54.4 (9.8) | 53.6 (9.7) | 51.7 (9.6) |
| Female sex, n (%) | 869 (47.6) | 1,289 (46.5) | 1,272 (50.3) | 1,116 (56.8) | 2,354 (63.9) |
| CCI, mean (SD) | 4.4 (2.2) | 4.2 (2.1) | 4.2 (2.0) | 4.1 (1.8) | 4.1 (1.8) |
| NYHA functional class | |||||
| Class I | 39/147 (26.5) | 66/251 (26.3) | 42/208 (20.2) | 42/172 (24.4) | 42/314 (13.4) |
| Class II | 55/147 (37.4) | 112/251 (44.6) | 89/208 (42.8) | 73/172 (42.4) | 130/314 (41.4) |
| Class III | 42/147 (28.6) | 58/251 (23.1) | 67/208 (32.2) | 46/172 (26.7) | 126/314 (40.1) |
| Class IV | 11/147 (7.5) | 15/251 (6.0) | 10/208 (4.8) | 11/172 (6.4) | 16/314 (5.1) |
| BMI, kg/m2 | |||||
| Valid n | 1,764 | 2,646 | 2,433 | 1,895 | 3,319 |
| Mean (SD) | 22.8 (1.4) | 27.5 (1.4) | 32.3 (1.5) | 37.3 (1.5) | 49.5 (9.0) |
SD = standard deviation, CCI = Charlson comorbidity index, NYHA = New York heart association
3.1. All-cause and HF-related 30-day rehospitalizations
Table 2 describes all-cause and HF-related 30-day rehospitalizations among all patients. Among commercially insured patients who experienced at least one all-cause hospitalization and had at least 30 days of follow-up time post discharge, the rate of all-cause 30-day rehospitalization ranged from 18.3% for Class 2 obesity cohort (n = 1,304) to 27.7% for Normal cohort (n = 838) (Table 2). The rate of HF-related rehospitalization ranged from 16.8% for the Class 2 obesity cohort to 24.5% for the Normal cohort (Table 2). In the Medicare population, the rate of all-cause 30-day rehospitalization ranged from 32.2% for the Overweight cohort to 37.5% for Class 3 obesity while the rate of HF-related rehospitalization ranged from 28.9% for the Overweight cohort to 33.9% for the Class 3 Obesity cohort (Table 2). In the Medicaid population, the rate of all-cause 30-day rehospitalization ranged from 35.2% for the Class 3 obesity cohort to 42.8% for the normal cohort while the rate of HF-related rehospitalization ranged from 30.5% for the Class 2 cohort to 35.2% for the Overweight cohort (Table 2).
Table 2.
Follow-up rehospitalizations by BMI levels among those with HFpEF1.
| Rehospitalization measure | Normal | Overweight | Class 1 obesity | Class 2 obesity | Class 3 obesity | |
|---|---|---|---|---|---|---|
| Commercial | ||||||
| All-cause re- hospitalization within 30 days of discharge |
valid n | 838 | 1,675 | 1,748 | 1,304 | 2,176 |
| n | 232 | 402 | 382 | 238 | 477 | |
| % | 27.7 | 24.0 | 21.9 | 18.3 | 21.9 | |
| HF-related re-hospitalization within 30 days of discharge |
valid n | 376 | 812 | 862 | 746 | 1,385 |
| n | 92 | 178 | 177 | 125 | 293 | |
| % | 24.5 | 21.9 | 20.5 | 16.8 | 21.2 | |
| Medicare | ||||||
| All-cause re- hospitalization within 30 days of discharge |
valid n | 4,615 | 7,953 | 7,281 | 4,772 | 5,600 |
| n | 1,558 | 2,562 | 2,433 | 1,605 | 2,101 | |
| % | 33.8 | 32.2 | 33.4 | 33.6 | 37.5 | |
| HF-related re- hospitalization within 30 days of discharge |
valid n | 2,864 | 5,038 | 4,940 | 3,431 | 4,381 |
| n | 832 | 1,455 | 1,471 | 1,052 | 1,483 | |
| % | 29.1 | 28.9 | 29.8 | 30.7 | 33.9 | |
| Medicaid | ||||||
| All-cause re- hospitalization within 30 days of discharge |
valid n | 1,016 | 1,373 | 1,237 | 972 | 1,999 |
| n | 435 | 555 | 441 | 355 | 704 | |
| % | 42.8 | 40.4 | 35.7 | 36.5 | 35.2 | |
| HF-related re- hospitalization within 30 days of discharge |
valid n | 531 | 730 | 730 | 614 | 1,418 |
| n | 171 | 257 | 235 | 187 | 458 | |
| % | 32.2 | 35.2 | 32.2 | 30.5 | 32.3 |
1Rehosiptalization data includes patients with hospitalization and sufficient follow-up to observe rehospitalization
HF = heart failure.
3.2. All-cause and HF-related acute care events-unadjusted
Supplemental Fig. 2 describes the unadjusted acute care events by BMI classes for each payer type. For the commercial population, mean (SD) HF-related follow-up all-cause acute care events ranged from 0.1 (0.4) for the overweight cohort to 0.2 (0.6) for the Class 3 obesity cohort (Supplemental Fig. 2). For the Medicare population, the mean (SD) HF-related acute care events ranged from 0.2 (0.6) for the overweight cohort to 0.4 (1.0) for the Class 3 obesity cohort (Supplemental Fig. 2). For the Medicaid population, the mean (SD) HF-related acute care events ranged from 0.3 (0.9) for the overweight cohort to 0.5 (1.4) for the Class 3 obesity cohort (Supplemental Fig. 2).
3.3. HF-related acute care events-adjusted
Fig. 1 describes acute care events by BMI classes for each payer type. In the overall population (Commercial, Medicare and Medicaid combined), Class 2 obesity patients had 11.4% higher and Class 3 obesity patients had 54.2% higher rate of HF-related acute care events (both p < 0.001), compared to the overweight cohort, when adjusted for other baseline covariates (Fig. 1). Patients with NYHA Class III/IV had 30.4% higher HF-related acute care events, compared with NYHA Class I patients (p = 0.004) (Fig. 1). Compared to commercially insured patients, the rate of acute care events was 89.6% higher among Medicare enrollees and twice as high among those with Medicaid coverage (both p < 0.001) (Fig. 1) (full model Supplemental Table 4).
Fig. 1.
HF-related PPPY acute care visits − Generalized Linear Model − Negative Binomial Distribution − Adjusted – all patients1,2. 1Model was adjusted for NYHA functional class, age, sex, insurance type, race, region, index year, baseline CCI score, baseline comorbid conditions, baseline HFpEF medication use and baseline costs. 2The overweight category was the reference group. HF = heart failure, PPPY = per-patient per-year, BMI = body mass index, CI = confidence interval, NYHA = New York Heart Association, HFpEF = heart failure with preserved injection fraction.
3.4. All-cause and HF-related total healthcare costs-unadjusted and adjusted
Supplemental Fig. 3 displays the unadjusted all-cause and HF related healthcare costs. Unadjusted costs generally increased with obesity severity. In commercial population, total PPPY costs rose from $45,559 (overweight) to $56,220 (Class 3), with HF-related costs highest in Class 3 ($8,669). Medicare showed a similar pattern, with HF‑related costs reaching $18,559 in Class 3. In Medicaid, HF‑related costs were also highest in Class 3 ($20,562), despite total costs peaking in normal weight. Fig. 2 describes the predicted HF-related total PPPY costs by insurance type (Full model in Supplemental Table 5-7). After adjustment, Class 3 obesity had the highest predicted HF‑related costs in commercial ($7,985) and Medicare ($16,722) populations, with Medicare showing a 52.9% increase vs. overweight. In Medicaid, the normal‑weight cohort had the highest adjusted HF‑related costs ($23,701).
Fig. 2.
HF-related PPPY total healthcare costs − Generalized Linear Model − Gamma Distribution – Adjusted1,2. 1Model was adjusted for NYHA, age, sex, race, region, index year, baseline CCI score, baseline comorbid conditions, baseline HFpEF medication use and baseline costs. 2The overweight category was the reference group. HF = heart failure, PPPY = per-patient per-year CI = confidence interval, CCI = Charlson comorbidity index, NYHA = New York Heart Association, HFpEF = heart failure with preserved injection fraction.
4. Discussion
The aim of this study was to evaluate the association of BMI and real-world hospitalization and ER events and economic burden among patients with HFpEF. Results indicated that patients with Class 3 obesity incurred higher healthcare expenditures and exhibited greater rates of all-cause acute care utilization and hospital readmissions, highlighting the significant burden associated with obesity in this population. After adjusting for baseline characteristics, Class 3 obesity was associated with numerically higher expected costs compared to the overweight group in the commercially insured population, and statistically significantly higher costs in the Medicare population, while in the Medicaid population, patients with Normal weight incurred higher adjusted all-cause costs than those classified as overweight. In terms of acute care events, patients with Class 2 and Class 3 obesity had significantly higher rates of HF-related acute care events in the overall population (Commercial, Medicare and Medicaid combined), compared to the Overweight cohort, after adjusting for baseline covariates (both p < 0.001). Regarding 30-day rehospitalizations, Medicare patients with Class 3 obesity had a higher rate of all-cause rehospitalizations and Medicaid patients with Class 3 obesity had a higher rate of HF-related rehospitalizations. Additionally, patients in the Normal cohort were often observed to have a higher rate of rehospitalizations compared with other cohorts, likely reflecting their distinct clinical profile at baseline. Evidence suggests that patients with a Normal BMI have a distinct clinical profile than other patients, potentially indicating a greater burden of comorbidities at baseline, which may explain their higher resource utilization and warrants further investigation.[16].
This study aligns with prior research documenting the substantial economic burden associated with HFpEF.[14], [15] Kilgore et al have reported significant costs and high readmission rates among Medicare patients with HF.[17] Reinhardt et al also demonstrated high costs associated with readmissions among patients with HF.[18] The current study builds on this body of evidence by examining more recent data and documenting high costs and readmission rates among patients with HFpEF across multiple payor types. Furthermore, this study also aligns with existing literature documenting the risks for patients with HFpEF with obesity.[19] To the best of our knowledge, this is the first study to stratify and document healthcare costs, acute care utilization, and readmissions by obesity class, building on the literature regarding HFpEF and obesity. This is relevant as it offers a more nuanced understanding of the clinical and economic burden associated with varying degrees of BMI classes, thereby informing interventions and resource planning to improve patient outcomes.
The results of this study suggest a higher risk of adverse outcomes in the Normal cohort and Class 3 obesity cohort compared with the other cohorts. The I-PRESERVE clinical trial examined the relationship between HFpEF and cardiovascular outcomes such as death and hospitalization for CVD[6], and identified a U-shaped relationship in which both the lowest and highest BMI categories were associated with increased rates of adverse outcomes such as mortality and hospitalization.[6] The results of the present study are consistent with this finding. Furthermore, NYHA functional class was comparable between patients with normal BMI and those with Class 3 obesity. These findings further support the concept of the “obesity paradox” in HF, particularly in HFpEF, where individuals with obesity, especially those with Class 2 obesity, often exhibit better short- and intermediate-term survival compared to those with normal BMI.[20] The “obesity paradox” has also been documented in a meta-analysis by Pawal et al.[21] There are several potential explanations for the obesity paradox including the fact that BMI is not a precise measure of obesity as well as the possibility of residual confounding.[22] This paradox has been partly attributed to the fact that individuals with more severe or chronic illness may experience unintentional weight loss. However, among those with more severe degrees of obesity, the risk of adverse events remains elevated.[23].
Contemporary cardiometabolic therapies with demonstrated benefits in HFpEF and obesity, including sodium glucose cotransporter2 inhibitors (SGLT2i) and GLP-1 RA, were underutilized in clinical practice during the study period (2015–2023). Real world claims data indicate that fewer than approximately 10–13% of patients with HFpEF were treated with SGLT2 inhibitors in the early years following pivotal trial publications and regulatory approvals.[24] Similarly, GLP1 receptor agonist use among patients with obesity and cardiometabolic disease remained low through much of the study window, with marked acceleration only after 2021 following expanded regulatory approvals and growing recognition of their weight loss and cardiovascular benefits. According to a cross-sectional study by Li et al, prescription rates increased markedly from 2.4% to 34.3% between 2010 and 2024 among populations with diabetes and obesity.[25] As utilization continues to increase beyond the end of the current study period, future real-world studies will be important to determine whether broader adoption of these therapies attenuates the excess burden observed in patients with HFpEF and obesity.
5. Study strengths and limitations
A key strength of this study is its inclusion of a broad population of insured US adults with heart failure with preserved ejection fraction, enhancing the generalizability of the findings to the broader insured HFpEF population in the US. However, there are several limitations to consider when evaluating the results of this study. First, the study findings may not be generalizable to HFpEF population outside of US. Identification of patients with HFpEF was based on HF claims and LVEF testing among patients with both claims and EHR data available in Optum’s Market Clarity database. The presence of a diagnosis code does not necessarily indicate presence of disease; to limit misidentification of HF patients and to classify HFpEF, LVEF results were required. Thus, the analysis relied on the availability of valid LVEF values in the EHR data and may have excluded true patients with HFpEF without available LVEF results. In addition, key objective diagnostic measures, including brain natriuretic peptide (BNP), N-terminal proBNP (NT-proBNP), and detailed echocardiographic indices, are unavailable in claims and sparsely captured in EHR data. As a result, our ability to confirm whether the study population fully met contemporary diagnostic criteria for HFpEF was limited. This study relied on BMI which is an imperfect proxy for adiposity and did not take into account weight changes over time. All HF patients included in this study were enrolled in a health plan in the United States (ie, commercial, Medicare, or Medicaid) during the study period; thus, findings from this study may not be generalizable to all patients with HF. The study period includes the COVID-19 pandemic period, during which changes in HCRU and medication filling patterns may have influenced the results. There is limited clinical granularity in claims data which are collected for the purpose of payment and there are inherent limitations to the use of claims databases for research. Coding errors may result in inaccurate or incomplete data, leading to potential misclassification of variables of interest and bias in research findings. A claim for a filled prescription is not an indication the medication was consumed or taken as prescribed. Also, physician-provided samples, samples taken as part of a clinical trial or over-the-counter medications were not observed in claims data. In evaluating rehospitalization outcomes, the study did not assess the time interval between the initial hospitalization and subsequent rehospitalizations, nor were the specific causes of these events examined. Lastly, unmeasured or residual confounding may exist in the relationship between BMI and study endpoints. Although the analysis attempted to control for this by adjusting for NYHA classification and key baseline characteristics, residual confounding cannot be entirely ruled out.
6. Conclusions
In this study, patients with Class 3 obesity and HFpEF experienced higher healthcare costs, more frequent all-cause acute care encounters, and increased readmissions rates, compared to the overweight cohort. These findings underscore the significant economic burden posed by severe obesity in the HFpEF population and highlight need for focused HFpEF management in this population. Future studies should examine the association between weight change, HF outcomes, and cost of care among patients with HFpEF.
Data sharing statement:
The data contained in the database used for the study contain proprietary elements owned by Optum and, therefore, cannot be broadly disclosed or made publicly available at this time. The disclosure of these data to third parties assumes certain data security and privacy protocols are in place and that the third party has executed a standard license agreement which includes restrictive covenants governing the use of the data.
Funding
This study was sponsored by Eli Lilly and Company.
CRediT authorship contribution statement
Navneet Upadhyay: Writing – review & editing, Methodology, Funding acquisition, Formal analysis, Conceptualization. Kirti Batra: Writing – review & editing, Methodology, Formal analysis, Conceptualization. Lisa Le: Writing – review & editing, Methodology, Formal analysis, Data curation. Andrea Steffens: Writing – review & editing, Supervision, Investigation. Maureen Carlyle: Writing – review & editing, Project administration. Rui Song: Writing – review & editing, Methodology, Investigation, Formal analysis. Thomas Horstman: Writing – review & editing, Methodology, Investigation, Formal analysis. Wenyu Ye: Writing – review & editing, Methodology, Investigation, Formal analysis. Arian Plat: Writing – review & editing, Supervision, Conceptualization.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgement
The authors would like to thank Gretchen Hultman of Optum for medical writing support.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.ijcha.2026.101939.
Appendix A. Supplementary data
The following are the Supplementary data to this article:
References
- 1.Redfield M.M., Borlaug B.A. Heart failure with Preserved Ejection Fraction: a Review. JAMA. 2023;329:827–838. doi: 10.1001/jama.2023.2020. [DOI] [PubMed] [Google Scholar]
- 2.Kittleson M.M., Panjrath G.S., Amancherla K., Davis L.L., Deswal A., Dixon D.L., et al. 2023 ACC Expert Consensus Decision Pathway on Management of Heart Failure with Preserved Ejection Fraction. JACC. 2023;81:1835–1878. doi: 10.1016/j.jacc.2023.03.393. [DOI] [PubMed] [Google Scholar]
- 3.Meijers W.C., van der Velde A.R., de Boer R.A. Biomarkers in heart failure with preserved ejection fraction. Neth Heart J. 2016;24:252–258. doi: 10.1007/s12471-016-0817-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Borlaug B.A., Jensen M.D., Kitzman D.W., Lam C.S.P., Obokata M., Rider O.J. Obesity and heart failure with preserved ejection fraction: new insights and pathophysiological targets. Cardiovasc. Res. 2023;118:3434–3450. doi: 10.1093/cvr/cvac120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Oguntade AS, Taylor H, Lacey B, Lewington S. Adiposity, fat-free mass and incident heart failure in 500 000 individuals. Open Heart. 2024;11. [DOI] [PMC free article] [PubMed]
- 6.Haass M., Kitzman D.W., Anand I.S., Miller A., Zile M.R., Massie B.M., et al. Body mass index and adverse cardiovascular outcomes in heart failure patients with preserved ejection fraction: results from the Irbesartan in Heart failure with Preserved Ejection Fraction (I-PRESERVE) trial. Circ. Heart Fail. 2011;4:324–331. doi: 10.1161/CIRCHEARTFAILURE.110.959890. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kitzman D.W., Shah S.J. The HFpEF Obesity Phenotype: the Elephant in the Room. J. Am. Coll. Cardiol. 2016;68:200–203. doi: 10.1016/j.jacc.2016.05.019. [DOI] [PubMed] [Google Scholar]
- 8.Miller W.L., Borlaug B.A. Impact of Obesity on volume Status in patients with Ambulatory Chronic Heart failure. J. Card. Fail. 2020;26:112–117. doi: 10.1016/j.cardfail.2019.09.010. [DOI] [PubMed] [Google Scholar]
- 9.Packer M., Kitzman D.W. Obesity-Related Heart failure with a Preserved Ejection Fraction: the Mechanistic Rationale for Combining Inhibitors of Aldosterone, Neprilysin, and Sodium-Glucose Cotransporter-2. JACC Heart Fail. 2018;6:633–639. doi: 10.1016/j.jchf.2018.01.009. [DOI] [PubMed] [Google Scholar]
- 10.Savji N., Meijers W.C., Bartz T.M., Bhambhani V., Cushman M., Nayor M., et al. The Association of Obesity and Cardiometabolic Traits with Incident HFpEF and HFrEF. JACC. Heart Failure. 2018;6:701–709. doi: 10.1016/j.jchf.2018.05.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Aminian A., Zajichek A., Arterburn D.E., Wolski K.E., Brethauer S.A., Schauer P.R., et al. Association of Metabolic Surgery with Major Adverse Cardiovascular Outcomes in patients with Type 2 Diabetes and Obesity. JAMA. 2019;322:1271–1282. doi: 10.1001/jama.2019.14231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kosiborod M.N., Abildstrøm S.Z., Borlaug B.A., Butler J., Rasmussen S., Davies M., et al. Semaglutide in patients with Heart failure with Preserved Ejection Fraction and Obesity. N. Engl. J. Med. 2023;389:1069–1084. doi: 10.1056/NEJMoa2306963. [DOI] [PubMed] [Google Scholar]
- 13.Packer M., Zile M.R., Kramer C.M., Baum S.J., Litwin S.E., Menon V., et al. Tirzepatide for Heart failure with Preserved Ejection Fraction and Obesity. N. Engl. J. Med. 2025;392:427–437. doi: 10.1056/NEJMoa2410027. [DOI] [PubMed] [Google Scholar]
- 14.Heidenreich P.A., Bozkurt B., Aguilar D., Allen L.A., Byun J.J., Colvin M.M., et al. 2022 AHA/ACC/HFSA Guideline for the Management of Heart Failure: Executive Summary: a Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice guidelines. Circulation. 2022;145:e876–e894. doi: 10.1161/CIR.0000000000001062. [DOI] [PubMed] [Google Scholar]
- 15.Urbich M., Globe G., Pantiri K., Heisen M., Bennison C., Wirtz H.S., et al. A Systematic Review of Medical costs Associated with Heart failure in the USA (2014-2020) Pharmacoeconomics. 2020;38:1219–1236. doi: 10.1007/s40273-020-00952-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Obokata M., Reddy Y.N.V., Pislaru S.V., Melenovsky V., Borlaug B.A. Evidence supporting the Existence of a Distinct Obese Phenotype of Heart failure with Preserved Ejection Fraction. Circulation. 2017;136:6–19. doi: 10.1161/CIRCULATIONAHA.116.026807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Kilgore M., Patel H.K., Kielhorn A., Maya J.F., Sharma P. Economic burden of hospitalizations of Medicare beneficiaries with heart failure. Risk Manag Healthc Policy. 2017;10:63–70. doi: 10.2147/RMHP.S130341. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Reinhardt S.W., Clark K.A.A., Xin X., Parzynski C.S., Riello Iii R.J., Sarocco P., et al. Thirty-Day and 90-Day Episode of Care spending following Heart failure Hospitalization among Medicare Beneficiaries. Circ. Cardiovasc. Qual. Outcomes. 2022;15 doi: 10.1161/CIRCOUTCOMES.121.008069. [DOI] [PubMed] [Google Scholar]
- 19.Ndumele C.E., Matsushita K., Lazo M., Bello N., Blumenthal R.S., Gerstenblith G., et al. Obesity and Subtypes of Incident Cardiovascular Disease. J. Am. Heart Assoc. 2016;5 doi: 10.1161/JAHA.116.003921. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Alebna P.L., Mehta A., Yehya A., daSilva-deAbreu A., Lavie C.J., Carbone S. Update on obesity, the obesity paradox, and obesity management in heart failure. Prog. Cardiovasc. Dis. 2024;82:34–42. doi: 10.1016/j.pcad.2024.01.003. [DOI] [PubMed] [Google Scholar]
- 21.Padwal R., McAlister F.A., McMurray J.J., Cowie M.R., Rich M., Pocock S., et al. The obesity paradox in heart failure patients with preserved versus reduced ejection fraction: a meta-analysis of individual patient data. Int. J. Obes. (Lond) 2014;38:1110–1114. doi: 10.1038/ijo.2013.203. [DOI] [PubMed] [Google Scholar]
- 22.Deswal A. THE OBESITY PARADOX IN HEART FAILURE: TIME TO MOVE FORWARD. Trans. Am. Clin. Climatol. Assoc. 2025;135:43–51. [PMC free article] [PubMed] [Google Scholar]
- 23.Kastorini C.-M., Panagiotakos D.B. The obesity paradox: Methodological considerations based on epidemiological and clinical evidence—New insights. Maturitas. 2012;72:220–224. doi: 10.1016/j.maturitas.2012.04.012. [DOI] [PubMed] [Google Scholar]
- 24.Gonzalez J., Dave C.V. Prescribing trends of SGLT2 inhibitors among HFrEF and HFpEF patients with and without T2DM, 2013-2021. BMC Cardiovasc. Disord. 2024;24:285. doi: 10.1186/s12872-024-03961-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Li P., Varghese J.S., Shah M.K., Galindo R.J., Pasquel F.J., Ali M.K., et al. Prescribing Trends of Glucagon-like Peptide 1 Receptor Agonists for Type 2 Diabetes or Obesity. JAMA Netw. Open. 2025;8 doi: 10.1001/jamanetworkopen.2025.40890. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.