Abstract
Aims
Obesity is common and associated with unique phenotypic features in heart failure with preserved ejection fraction (HFpEF). Therefore, understanding the efficacy and safety of new therapies in HFpEF patients with obesity is important. The effects of dapagliflozin were examined according to body mass index (BMI) among patients in the Dapagliflozin Evaluation to Improve the LIVEs of Patients With PReserved Ejection Fraction Heart Failure trial.
Methods and results
Body mass index was analysed by World Health Organization (WHO) categories and as a continuous variable using restricted cubic splines. Body mass index ranged from 15.2 to 50 kg/m2 with a mean value of 29.8 (standard deviation ± 6.1) kg/m2. The proportions, by WHO category, were: normal weight 1343 (21.5%); overweight 2073 (33.1%); Class I obesity 1574 (25.2%); Class II obesity 798 (12.8%); and Class III obesity 415 (6.6%). Compared with placebo, dapagliflozin reduced the risk of the primary outcome to a similar extent across these categories: hazard ratio (95% confidence interval): 0.89 (0.69–1.15), 0.87 (0.70–1.08), 0.74 (0.58–0.93), 0.78 (0.57–1.08), and 0.72 (0.47–1.08), respectively (P-interaction = 0.82). The placebo-corrected change in Kansas City Cardiomyopathy Questionnaire total symptom score with dapagliflozin at 8 months was: 0.9 (−1.1, 2.8), 2.5 (0.8, 4.1), 1.9 (−0.1, 3.8), 2.7 (−0.5, 5.8), and 8.6 (4.0, 13.2) points, respectively (P-interaction = 0.03). The placebo-corrected change in weight at 12 months was: –0.88 (−1.28, –0.47), –0.65 (−1.04, –0.26), –1.42 (−1.89, –0.94), –1.17 (−1.94, –0.40), and –2.50 (−4.4, –0.64) kg (P-interaction = 0.002).
Conclusions
Obesity is common in patients with HFpEF and is associated with higher rates of heart failure hospitalization and worse health status. Treatment with dapagliflozin improves cardiovascular outcomes across the spectrum of BMI, leads to greater symptom improvement in patients with obesity, compared with those without, and has the additional benefit of causing modest weight loss.
Keywords: Heart failure, Obesity, Body mass index, SGLT2 inhibitor
Structured Graphical Abstract
Structured graphical abstract.
Summary of the key background and findings of this study. BMI, body mass index; CV, cardiovascular; DELIVER, Dapagliflozin Evaluation to Improve the LIVEs of Patients With PReserved Ejection Fraction Heart Failure; HFmrEF, heart failure with mildly reduced ejection fraction; HFpEF, heart failure with preserved ejection fraction; KCCQ-TSS, Kansas City Cardiomyopathy Questionnaire Total Symptom Score.
See the editorial comment for this article ‘‘Normal weight' is really not normal in chronic disease’, by Stefan D. Anker et al., https://doi.org/10.1093/eurheartj/ehac486.
Introduction
Obesity is a risk factor for the development of heart failure (HF), especially HF with preserved ejection fraction (HFpEF), more so in women than men.1–5 Consequently, many patients with HF are obese and obesity is more often a concomitant problem in patients with HFpEF. Indeed, in the USA, >80% of patients with HFpEF are overweight or obese and in contemporary clinical trials, the proportion of HF patients with severe obesity [World Health Organization (WHO) Class III] is almost 20%.6,7 Obese patients generally have worse symptoms and greater functional limitations than their non-obese counterparts.8 Recent studies have highlighted several potential explanations for this, including a variety of systemic effects of inflammatory cytokines and other mediators secreted by adipose tissue, direct cardiac consequences such as pericardial restraint due to local deposition of adipose tissue, greater concentric left-ventricular remodelling and right-ventricular dilatation and dysfunction, and even renal dysfunction due to excess adipose tissue in and around the kidney.4,8–13 Consequently, finding treatments that are efficacious in HFpEF patients with concomitant obesity are important. Sodium-glucose cotransporter 2 (SGLT2) inhibitors may be a particularly attractive treatment in these patients, especially as obesity is part of a common triad which includes dysglycaemia and hypertension and SGLT2 inhibitors lead to modest reductions in weight, glycated haemoglobin, and blood pressure, in addition to their benefits on HF symptoms and outcomes.14–20
Therefore, the effect of SGLT2 inhibition was examined according to body mass index (BMI) in the Dapagliflozin Evaluation to Improve the LIVEs of Patients With PReserved Ejection Fraction Heart Failure (DELIVER) trial.19,21,22 DELIVER included 6263 patients with HF and mildly reduced and preserved ejection fraction and showed that dapagliflozin, compared with placebo, reduced the risk of worsening HF events or cardiovascular death, and improved symptoms.19,21
Methods
The DELIVER was a double-blind, placebo-controlled trial which examined the efficacy and safety of dapagliflozin 10 mg once daily compared with matched placebo in patients with HF and mildly reduced and preserved ejection fraction. Randomization was stratified by the presence or absence of Type 2 diabetes. The design, baseline characteristics, and primary results are published.19,21,22 The protocol was approved by an Ethics Committee at each participating centre and all patients provided written informed consent.
Study patients
Key inclusion criteria included age ≥40 years, HF diagnosis ≥6 weeks, and a requirement for treatment with at least intermittent diuretic, New York Heart Association (NYHA) functional Classes II–IV, left-ventricular ejection fraction (LVEF) >40%, evidence of structural heart disease (either left-atrial enlargement or left-ventricular hypertrophy), and an N-terminal pro-B-type natriuretic peptide (NT-proBNP) concentration ≥300 pg/mL (≥600 pg/mL if atrial fibrillation/flutter on the electrocardiogram at enrolment). Patients could be enrolled in both the outpatient and inpatient setting. Patients were excluded if BMI was >50 kg/m2. Other key exclusion criteria included Type 1 diabetes; estimated glomerular filtration rate (eGFR) <25 mL/min/1.73 m2; and systolic blood pressure <95 mmHg. A complete list of inclusion and exclusion criteria is provided in the design paper.21
Body mass index
Body mass index was calculated as weight in kilograms divided by height in meters squared, using measurements made at the trial enrolment visit. Patients categorized using WHO definitions, i.e. underweight (<18.5 kg/m2); normal weight (18.5–24.9 kg/m2); overweight (25.0–29.9 kg/m2); obesity Class I (30.0–34.9 kg/m2); obesity Class II (35.0–39.9 kg/m2), and obesity Class III (≥40 kg/m2).23
Outcomes
The primary outcome in DELIVER was the composite of time to the first occurrence of a worsening HF event (HF hospitalization or urgent outpatient HF visit requiring intravenous diuretic therapy) or cardiovascular death. The primary outcome was assessed in the full population and patients with an ejection fraction of <60% in a dual primary analysis. Secondary outcomes included in this analysis were the total number of worsening HF events (including first and recurrent events) and cardiovascular deaths; change in self-reported severity of HF symptoms at 8 months based on the Kansas City Cardiomyopathy Questionnaire total symptom score (KCCQ-TSS); worsening HF events, cardiovascular death, and all-cause death.
Prespecified safety analyses included serious adverse events, adverse events leading to discontinuation of trial treatment, and selected adverse events, including volume depletion, renal adverse events, amputation, major hypoglycaemia, and diabetic ketoacidosis for consistency across reporting in trials. These analyses included only patients who had taken at least one dose of study medication.
Statistical analyses
Baseline characteristics are reported for each BMI category as means ± standard deviation (SD), median with interquartile range (IQR; Q1–Q3) and proportions, as appropriate. The Cochran–Armitage test was used to test-for-trend across groups for binary variables and the Jonckheere–Terpstra test for continuous variables.
The association between BMI category and each outcome, adjusted for randomized treatment, was compared between BMI groups in a Cox regression model with overweight as the reference. This was repeated with additional adjustments for age, sex, race, region, heart rate, systolic blood pressure, glycated haemoglobin (HbA1c), creatinine, history of HF hospitalization, NYHA class, LVEF, atrial fibrillation/flutter, hypertension, myocardial infarction, coronary artery bypass graft, stroke, and (log-transformed) NT-proBNP. The associations between BMI, as a continuous variable, and outcomes were modelled using restricted cubic splines with adjustment for randomized treatment with median population BMI as reference. The 5 knots were placed at default positions according to percentiles of BMI (5, 27.5, 50, 72.5, and 95 centile). This was repeated with the additional adjustments listed above, in both the whole population and in males and females separately.
The efficacy of dapagliflozin, compared with placebo, in each BMI category, was examined using Kaplan–Meier estimates and Cox regression models. Event rates per 100 person-years and hazard ratios (HRs) are reported for each BMI category (these and all other models were stratified by diabetes status). The presence of an interaction between BMI category and treatment on the occurrence of each outcome was examined using a likelihood ratio test. The effect of randomized treatment across baseline BMI as a continuous variable was modelled flexibly using restricted cubic splines with 3 knots (at 10th, 50th, and 90th percentile of BMI), in the whole population and subgroups of interest (males and females; LVEF <60 and ≥60; presence or absence of diabetes; and patients of White and Asian race). Analysis of recurrent events was by the Lin, Wei, Ying, and Yang method, a semi-parametric proportional-rates model.24 Mean change in KCCQ-TSS at 8 months was calculated using a mixed model for repeated measurements including all time points within each BMI category. A three-way interaction between treatment, time, and BMI was assessed in the mixed model at 8-month follow up.
A mixed model for repeated measurement was used to examine change in weight over time according to baseline BMI (adjusted for baseline values, randomized treatment, and interaction of treatment and visit, with a random intercept and slope per patient). A three-way interaction between treatment, time, and BMI was assessed at 12-month follow up.
The interaction between BMI category and randomized treatment on the occurrence of the safety outcomes was tested in a logistic regression model.
All analyses were conducted using STATA version 17.0 (College Station, TX, USA).
Results
Of the 6263 patients randomized in DELIVER, 6 had missing data for BMI (Structured Graphical Abstract). Body mass index ranged from 15.2 to 50 kg/m2 with a mean value of 29.8 (SD ± 6.1) kg/m2 and a median value of 29.1 (Q1–Q3 25.4–33.4) kg/m2 and an approximately normal distribution (see Supplementary material online, Figure S1). In total, 54 (0.9%) patients were classified as underweight, 1343 (21.5%) as normal weight, 2073 (33.1%) as overweight, 1574 (25.2%) as Class I obesity, 798 (12.8%) as Class II obesity, and 415 (6.6%) as Class III obesity. Due to the small number of patients in the underweight category, these participants were not included in the main results but are presented in the Supplementary material online, Tables S1–S4.
Patient characteristics
Baseline characteristics according to BMI class are presented in Table 1. Compared with patients with normal BMI, those with obesity were younger, more often female and more likely to be White (and less likely to be of Asian race). They had higher systolic and diastolic blood pressure, a higher HbA1c and were more likely to have a diagnosis of Type 2 diabetes than patients with a normal weight. There was an inverse association between BMI and NT-proBNP with the lowest level of NT-proBNP in Class III obesity—862 (IQR 539–1311) ng/L compared with 1155 (IQR 667–2055) ng/L in patients in the normal weight category. A higher proportion of patients with obesity were in NYHA Class III or IV and patients with obesity had lower (worse) KCCQ-TSS. Patients with higher BMI had a higher mean LVEF compared with those with normal weight. Patients with greater obesity had more hypertension but less coronary heart disease. Patients with obesity were more often treated with a beta-blocker, renin–angiotensin system blocker, calcium channel blocker, and a loop diuretic, but were less often treated with a mineralocorticoid receptor antagonist.
Table 1.
Baseline characteristics according to body mass index category
| Normal weight | Overweight | Obesity Class I | Obesity Class II | Obesity Class III | P-value | |
|---|---|---|---|---|---|---|
| n = 1343 | n = 2073 | n = 1574 | n = 798 | n = 415 | ||
| Age (years) | 73.3 ± 10.0 | 72.4 ± 9.5 | 71.4 ± 9.0 | 69.3 ± 9.1 | 67.8 ± 9.0 | <0.001 |
| Female sex | 578 (43.0) | 828 (39.9) | 676 (42.9) | 380 (47.6) | 255 (61.4) | <0.001 |
| Race | <0.001 | |||||
| White | 587 (43.7) | 1427 (68.8) | 1333 (84.7) | 711 (89.1) | 362 (87.2) | |
| Asian | 633 (47.1) | 453 (21.9) | 119 (7.6) | 24 (3.0) | 8 (1.9) | |
| Black or African American | 24 (1.8) | 47 (2.3) | 32 (2.0) | 30 (3.8) | 25 (6.0) | |
| Other | 99 (7.4) | 146 (7.0) | 90 (5.7) | 33 (4.1) | 20 (4.8) | |
| Smoking status | 0.58 | |||||
| Current | 115 (8.6) | 158 (7.6) | 116 (7.4) | 65 (8.1) | 21 (5.1) | |
| Former | 475 (35.4) | 750 (36.2) | 579 (36.8) | 281 (35.2) | 158 (38.1) | |
| Never | 753 (56.1) | 1165 (56.2) | 879 (55.8) | 452 (56.6) | 236 (56.9) | |
| Vital signs | ||||||
| Pulse (beats/min) | 71.2 ± 12.1 | 70.9 ± 11.8 | 71.7 ± 11.2 | 72.4 ± 11.7 | 72.8 ± 12.0 | <0.001 |
| SBP(mmHg) | 124.9 ± 15.6 | 127.8 ± 14.7 | 130.1 ± 14.7 | 130.6 ± 15.9 | 130.6 ± 16.1 | <0.001 |
| DBP (mmHg) | 72.1 ± 10.4 | 73.5 ± 9.8 | 75.1 ± 10.1 | 75.6 ± 10.3 | 75.4 ± 11.9 | <0.001 |
| Laboratory values | ||||||
| HbA1c (%) | 6.0 (5.6–6.5) | 6.1 (5.7–6.8) | 6.2 (5.8–7.1) | 6.4 (5.9–7.6) | 6.5 (5.9–7.7) | <0.001 |
| Creatinine (μmol/L) | 100.0 ± 30.4 | 102.4 ± 30.5 | 103.8 ± 31.4 | 105.1 ± 32.7 | 100.3 ± 30.8 | 0.01 |
| eGFR (mL/min/1.73 m2) | 62.1 ± 19.4 | 61.2 ± 19.0 | 60.2 ± 18.5 | 60.0 ± 19.4 | 62.0 ± 20.6 | 0.03 |
| eGFR < 60 mL/min/1.73 m2 (%) | 643 (47.9) | 1004 (48.4) | 786 (49.9) | 405 (50.8) | 201 (48.4) | 0.28 |
| NT-proBNP (ng/L) | 1155.0 (667.0–2055.0) | 1014.0 (623.0–1764.0) | 981.0 (616.0–1690.0) | 934.5 (601.0–1524.0) | 862.0 (539.0–1311.0) | <0.001 |
| NT-proBNP (ng/L) if baseline ECG in AF/flutter | 1606.0 (1100.0–2573.0) | 1447.0 (996.0–2311.0) | 1404.5 (956.0–2155.0) | 1242.0 (865.0–1974.0) | 1108.0 (865.0–1616.0) | <0.001 |
| NT-proBNP (ng/L) if baseline ECG not in AF/flutter | 810.0 (497.0–1580.0) | 757.0 (492.0–1308.0) | 663.0 (447.0–1182.0) | 627.0 (433.5–1154.0) | 612.0 (426.0–936.0) | <0.001 |
| Heart failure characteristics | ||||||
| Prior HF hospitalization | 586 (43.6) | 817 (39.4) | 625 (39.7) | 309 (38.7) | 173 (41.7) | 0.10 |
| NYHA class | <0.001 | |||||
| II | 1064 (79.2) | 1633 (78.8) | 1160 (73.7) | 560 (70.2) | 245 (59.0) | |
| III/IV | 279 (20.8) | 439 (21.2) | 414 (26.3) | 238 (29.8) | 170 (41.0) | |
| KCCQ-TSS | 80.2 (62.5–93.8) | 75.0 (59.4–89.6) | 69.8 (54.2–84.4) | 66.7 (50.0–83.3) | 58.3 (38.5–72.9) | <0.001 |
| Baseline LVEF (%) | 53.0 (46.0–60.0) | 53.0 (46.0–60.0) | 54.0 (47.0–60.0) | 55.0 (47.0–60.0) | 55.0 (49.0–60.0) | 0.02 |
| Clinical history | ||||||
| T2DM | 421 (31.3) | 885 (42.7) | 782 (49.7) | 463 (58.0) | 240 (57.8) | <0.001 |
| Atrial fibrillation | 720 (53.6) | 1060 (51.1) | 929 (59.0) | 486 (60.9) | 235 (56.6) | <0.001 |
| Hypertension | 1071 (79.7) | 1812 (87.4) | 1474 (93.6) | 750 (94.0) | 398 (95.9) | <0.001 |
| Myocardial infarction | 344 (25.6) | 635 (30.6) | 398 (25.3) | 180 (22.6) | 71 (17.1) | <0.001 |
| CABG | 126 (9.4) | 283 (13.7) | 207 (13.2) | 89 (11.2) | 37 (8.9) | 0.97 |
| Stroke | 147 (10.9) | 220 (10.6) | 140 (8.9) | 66 (8.3) | 22 (5.3) | <0.001 |
| Medical therapy | ||||||
| Beta-blocker | 1033 (77.0) | 1694 (81.7) | 1353 (86.0) | 699 (87.6) | 352 (84.8) | <0.001 |
| Calcium channel blockers | 353 (26.3) | 594 (28.7) | 491 (31.2) | 313 (39.2) | 150 (36.1) | <0.001 |
| ACEi, ARB, or ARNI | 941 (70.2) | 1594 (76.9) | 1289 (81.9) | 644 (80.7) | 329 (79.3) | <0.001 |
| Mineralocorticoid receptor antagonist | 618 (46.1) | 869 (41.9) | 679 (43.1) | 308 (38.6) | 167 (40.2) | 0.004 |
| Loop diuretics | 980 (73.1) | 1513 (73.0) | 1227 (78.0) | 674 (84.5) | 373 (89.9) | <0.001 |
| Other (non-loop) diuretics | 243 (18.1) | 448 (21.6) | 356 (22.6) | 189 (23.7) | 95 (22.9) | 0.003 |
Data are presented as mean ± SD or median (IQR) for continuous measures, and n (%) for categorical measures.
ACEi, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; ARNI, angiotensin receptor neprilysin inhibitor CABG, coronary artery bypass graft; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; HbA1c, glycated haemoglobin; HF, heart failure; KCCQ-TSS, Kansas City Cardiomyopathy Questionnaire—Total Symptom Score; LVEF, left-ventricular ejection fraction; NT-proBNP, N-terminal pro-B-type natriuretic peptide; NYHA, New York Heart Association, SBP, systolic blood pressure; T2DM, Type 2 diabetes mellitus.
Outcomes according to body mass index
There was a J-shaped relationship between BMI categories and the crude (unadjusted) incidence of the primary outcome and worsening HF events. The crude incidence of both outcomes was lowest in the overweight category and increased with increasing BMI category above this, driven by an increasing incidence of worsening HF events (Table 2, Figures 1 and 2). The relationship between BMI category and cardiovascular and all-cause death was different, with the lowest crude rates among patients with Class III obesity (Table 2, Figures 1 and 2). Examination of BMI as a continuous variable confirmed these patterns.
Table 2.
Outcomes according to body mass index category
| Normal weight (n = 1343) | Overweight (n = 2073) | Obesity Class I (n = 1574) | Obesity Class II (n = 798) | Obesity Class III (n = 415) | |
|---|---|---|---|---|---|
| Primary outcome | |||||
| N (%) | 233 (17.4) | 331 (16.0) | 298 (18.9) | 154 (19.3) | 92 (22.2) |
| Rate (95% CI) | 8.7 (7.6–9.9) | 7.7 (6.9–8.6) | 9.1 (8.1–10.2) | 9.2 (7.9–10.8) | 11.0 (9.0–13.6) |
| Unadjusted (95% CI)a | 1.17 (0.99–1.38) | REF | 1.16 (0.99–1.36) | 1.15 (0.95–1.39) | 1.37 (1.09–1.73) |
| Additional adjustmentb (95% CI) | 1.12 (0.94–1.34) | REF | 1.21 (1.03–1.43) | 1.26 (1.03–1.54) | 1.68 (1.32–2.15) |
| Worsening heart failure event | |||||
| N (%) | 163 (12.1) | 227 (11.0) | 234 (14.9) | 122 (15.3) | 72 (17.4) |
| Rate (95% CI) | 6.1 (5.2–7.1) | 5.3 (4.6–6.0) | 7.1 (6.3–8.1) | 7.3 (6.1–8.7) | 8.6 (6.9–10.9) |
| Unadjusted (95% CI)a | 1.19 (0.97–1.45) | REF | 1.33 (1.11–1.60) | 1.34 (1.07–1.67) | 1.57 (1.20–2.04) |
| Additional adjustmentb (95% CI) | 1.07 (0.86–1.32) | REF | 1.42 (1.17–1.71) | 1.48 (1.18–1.87) | 1.94 (1.47–2.57) |
| Cardiovascular death | |||||
| N (%) | 125 (9.3) | 149 (7.2) | 110 (7.0) | 67 (8.4) | 29 (7.0) |
| Rate (95% CI) | 4.4 (3.7–5.2) | 3.3 (2.8–3.8) | 3.1 (2.6–3.7) | 3.7 (2.9–4.7) | 3.1 (2.2–4.5) |
| Unadjusted (95% CI)a | 1.39 (1.09–1.76) | REF | 0.93 (0.72–1.19) | 1.09 (0.81–1.45) | 0.92 (0.62–1.38) |
| Additional adjustmentb (95% CI) | 1.48 (1.15–1.90) | REF | 0.98 (0.76–1.26) | 1.27 (0.94–1.71) | 1.25 (0.82–1.89) |
| All-cause death | |||||
| N (%) | 239 (17.8) | 325 (15.7) | 255 (16.2) | 125 (15.7) | 63 (15.2) |
| Rate (95% CI) | 8.4 (7.4–9.5) | 7.1 (6.4–7.9) | 7.2 (6.3–8.1) | 6.8 (5.7–8.2) | 6.8 (5.3–8.7) |
| Unadjusted (95% CI)a | 1.29 (1.02–1.43) | REF | 0.99 (0.84–1.16) | 0.93 (0.75–1.14) | 0.92 (0.70–1.20) |
| Additional adjustmentb (95% CI) | 1.29 (1.08–1.54) | REF | 1.03 (0.87–1.21) | 1.05 (0.85–1.30) | 1.21 (0.92–1.61) |
| Total heart failure hospitalizations and cardiovascular deaths | |||||
| No | 385 | 511 | 514 | 283 | 155 |
| Rate (95% CI) | 13.6 (11.7–15.8) | 11.2 (9.9–12.8) | 14.5 (12.6–16.8) | 15.6 (12.9–18.9) | 16.8 (13.4–21.3) |
| Unadjusted (95% CI)a | 1.25 (1.02–1.53) | REF | 1.26 (1.04–1.53) | 1.33 (1.05–1.67) | 1.42 (1.09–1.85) |
| Additional adjustmentb (95% CI) | 1.14 (0.94–1.39) | REF | 1.35 (1.11–1.64) | 1.52 (1.21–1.91) | 1.83 (1.38–2.41) |
Rates are given per 100 patient-years.
Baseline model adjusted for randomized treatment and stratified by diabetes status.
Age, sex, race, region, heart rate, systolic blood pressure, HbA1c, creatinine, history of heart failure hospitalization, NYHA class, LV ejection fraction, atrial fibrillation, hypertension, MI, CABG, stroke, NT-proBNP (log-transformed).
Figure 1.
Cumulative incidence of key outcomes according to body mass index. Risk of each outcome in patients grouped by baseline body mass index.
Figure 2.
Effect of body mass index as continuous variable on the hazard of each outcome. Risk of each outcome according to body mass index with reference to the population median (29 kg/m2). The baseline (blue) line is adjusted for randomized treatment and stratified for diabetes status (blue-shaded area represents the 95% confidence interval). The red line includes additional adjustment for age, sex, race, heart rate, pulse, systolic blood pressure, glycated haemoglobin, creatinine, history of heart failure hospitalization, New York Heart Association class, left-ventricular ejection fraction, atrial fibrillation, hypertension, myocardial infarction, coronary artery bypass graft, stroke, and (log-transformed) N-terminal pro-B-type natriuretic peptide.
Adjustment for prognostically important variables, including NT-proBNP, did not fundamentally alter the patterns observed in the unadjusted categorical or continuous analyses described above (Table 2 and Figure 2).
Repeating this analysis in males and females separately showed the nadir in risk for the primary endpoint and HF hospitalization occurred at a slightly lower BMI in men compared with women (see Supplementary material online, Figure S2). There was no difference between sexes for mortality outcomes.
Effects of dapagliflozin on clinical outcomes according to body mass index
The HR for the primary outcome was 0.82 [95% confidence interval (CI) 0.73–0.92] in the full population and 0.83 (95% CI 0.73–0.95) in the <60% ejection fraction subgroup. Dapagliflozin reduced the risk of the primary outcome to a similar extent across BMI categories: HR (95% CI) 0.89 (0.69–1.15) for normal weight, 0.87 (0.70–1.08) for overweight; 0.74 (0.58–0.93) for obesity Class I, 0.78 (0.57–1.08) for obesity Class II, and 0.72 (0.47–1.08) for obesity Class III (P-interaction = 0.82; Table 3). Examined as a continuous variable, there was no significant interaction between BMI and randomized treatment on the primary outcome (P-interaction = 0.68; Figure 3). Results in the subgroup of patients with ejection fraction <60% were consistent with that of the full population (see Supplementary material online, Table S5 and Figure S3).
Table 3.
Effect of randomized treatment on outcomes according to body mass index category
| Overall (n = 6263)a | Normal weight (n = 1343) | Overweight (n = 2073) | Obesity Class I (n = 1574) | Obesity Class II (n = 798) | Obesity Class III (n = 415) | P-int | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Placebo | Dapagliflozin | Placebo | Dapagliflozin | Placebo | Dapagliflozin | Placebo | Dapagliflozin | Placebo | Dapagliflozin | Placebo | Dapagliflozin | ||
| n | 3132 | 3131 | 670 | 673 | 1048 | 1025 | 802 | 772 | 386 | 412 | 204 | 211 | |
| Primary endpoint | |||||||||||||
| N (%) | 610 (19.5) | 512 (16.4) | 122 (18.2) | 111 (16.5) | 176 (16.8) | 155 (15.1) | 172 (21.5) | 126 (16.3) | 84 (21.8) | 70 (17.0) | 52 (25.5) | 40 (19.0) | |
| Rate (95% CI) | 9.6 (8.9–10.4) | 7.8 (7.2–8.5) | 9.2 (7.7–11.0) | 8.2 (6.8–9.9) | 8.2 (7.1–9.5) | 7.2 (6.1–8.4) | 10.5 (9.0–12.2) | 7.7 (6.5–9.2) | 10.3 (8.3–12.8) | 8.2 (6.5–10.3) | 13.2 (10.0–17.3) | 9.1 (6.7–12.5) | |
| HR | 0.82 (0.73–0.92) | 0.89 (0.69–1.15) | 0.87 (0.70–1.08) | 0.74 (0.58–0.93) | 0.78 (0.57–1.08) | 0.72 (0.47–1.08) | 0.82 | ||||||
| Worsening heart failure event | |||||||||||||
| N (%) | 455 (14.5) | 368 (11.8) | 83 (12.4) | 80 (11.9) | 126 (12.0) | 101 (9.9) | 136 (17.0) | 98 (12.7) | 66 (17.1) | 56 (13.6) | 43 (21.1) | 29 (13.7) | |
| Rate (95% CI) | 7.2 (6.5–7.8) | 5.6 (5.1–6.2) | 6.3 (5.0–7.8) | 5.9 (4.7–7.3) | 5.9 (5.0–7.0) | 4.7 (3.8–5.7) | 8.3 (7.0–9.8) | 6.0 (4.9–7.3) | 8.1 (6.4–10.3) | 6.5 (5.0–8.5) | 10.9 (8.1–14.7) | 6.6 (4.6–9.5) | |
| HR | 0.79 (0.69–0.91) | 0.94 (0.69–1.28) | 0.79 (0.61–1.03) | 0.72 (0.56–0.94) | 0.80 (0.56–1.14) | 0.63 (0.39–1.00) | 0.66 | ||||||
| Cardiovascular death | |||||||||||||
| N (%) | 261 (8.3) | 231 (7.4) | 66 (9.9) | 59 (8.8) | 81 (7.7) | 68 (6.6) | 64 (8.0) | 46 (6.0) | 33 (8.6) | 34 (8.3) | 13 (6.4) | 16 (7.6) | |
| Rate (95% CI) | 3.8 (3.3–4.3) | 3.3 (2.9–3.8) | 4.7 (3.7–5.9) | 4.1 (3.2–5.3) | 3.5 (2.8–4.4) | 3.0 (2.4–3.8) | 3.5 (2.8–4.5) | 2.6 (2.0–3.5) | 3.7 (2.6–5.2) | 3.7 (2.6–5.1) | 2.9 (1.7–5.0) | 3.3 (2.0–5.5) | |
| HR | 0.88 (0.74–1.05) | 0.88 (0.62–1.25) | 0.84 (0.61–1.16) | 0.75 (0.52–1.10) | 1.00 (0.62–1.61) | 1.17 (0.56–2.44) | 0.89 | ||||||
| All-cause death | |||||||||||||
| N (%) | 526 (16.8) | 497 (15.9) | 126 (18.8) | 113 (16.8) | 171 (16.3) | 154 (15.0) | 134 (16.7) | 121 (15.7) | 58 (15.0) | 67 (16.3) | 31 (15.2) | 32 (15.2) | |
| Rate (95% CI) | 7.6 (9.7–8.3) | 7.2 (6.6–7.8) | 8.9 (7.5–10.6) | 7.9 (6.5–9.5) | 7.4 (6.4–8.6) | 6.8 (5.8–8.0) | 7.4 (6.2–8.7) | 6.9 (5.8–8.3) |
6.4 (5.0–8.3) | 7.2 (5.7–9.2) | 6.8 (4.8–9.7) |
6.7 (4.7–9.5) |
|
| HR | 0.94 (0.83–1.07) | 0.88 (0.69–1.14) | 0.91 (0.73–1.13) | 0.94 (0.74–1.20) | 1.13 (0.79–1.60) | 0.97 (0.59–1.60) | 0.82 | ||||||
| Total heart failure hospitalizations and cardiovascular deaths | |||||||||||||
| N (%) | 1057 | 815 | 198 | 187 | 288 | 223 | 323 | 191 | 154 | 129 | 89 | 66 | |
| Rate (95% CI) | 15.3 (13.9–16.9) | 11.8 (10.7–13.1) | 14.0 (11.4–17.5) | 13.1 (10.6–16.3) | 12.6 (10.6–15.1) | 9.9 (8.2–12.0) | 17.8 (14.6–21.9) | 11.0 (9.1–13.4) | 17.2 (13.5–22.2) | 14.0 (10.5–19.1) | 19.8 (14.8–27.1) |
13.9 (9.8–20.4) |
|
| Rate ratio | 0.77 (0.67–0.89) | 0.93 (0.69–1.25) | 0.78 (0.60–1.01) | 0.62 (0.47–0.82) | 0.80 (0.55–1.18) | 0.71 (0.45–1.12) | 0.44 | ||||||
| KCCQ-TSS | |||||||||||||
| Mean change in KCCQ at 8 m | 5.6 (4.9–6.3) |
7.9 (7.2–8.6) | 3.9 (2.6–5.3) |
4.8 (3.5–6.2) | 5.6 (4.4–6.7) | 8.0 (6.9–9.2) | 6.1 (4.7–7.5) | 8.0 (6.6–9.4) | 7.2 (5.0–9.5) | 9.9 (7.7–12.1) | 5.9 (2.6–9.1) | 14.5 (11.3–17.7) | |
| Placebo-corrected change at 8 mb | 2.4 (1.4, 3.4) | 0.9 (−1.1, 2.8) | 2.5 (0.8, 4.1) | 1.9 (−0.1, 3.8) | 2.7 (−0.5, 5.8) | 8.6 (4.0, 13.2) | 0.03 | ||||||
Rates are given per 100 patient-years.
Full trial population, including the six patients with missing BMI at baseline.
Mixed-effect models for repeated measurements adjusted for baseline value, visit (Months 1, 4, and 8), randomized treatment, and interaction of treatment and visit.
Figure 3.
Treatment effect of dapagliflozin on the main study outcomes according to baseline body mass index. Baseline body mass index (5–95 centile) is shown on the x-axis and the hazard ratio for the effect of dapagliflozin compared with placebo is shown on the y-axis. The horizontal black line shows a hazard ratio of 1 (unity). The blue line represents a continuous hazard ratio and the blue-shaded areas the 95% confidence interval. A hazard ratio <1 indicates a benefit of dapagliflozin over placebo.
The effects of dapagliflozin on the other outcomes (cardiovascular death, a worsening HF event, all-cause mortality, total HF events, and cardiovascular death) were also consistent across BMI categories (P for interaction for all outcomes ≥0.4; Table 3). The results were also consistent when BMI was modelled as a continuous variable (P for interaction all ≥0.2; Figure 3). There was no treatment-by-sex-by-BMI interaction (P = 0.76 for the primary endpoint). The effect of treatment in men and women separately is shown in Supplementary material online, Figure S4 and of other subgroups of interest in Supplementary material online, Figure S5, with no significant variation of treatment effect by BMI in these subgroups.
Change in KCCQ-TSS according to baseline body mass index
Overall, 5792 patients (93%) had KCCQ-TSS recorded at baseline, and 4485 (71.7%) had a measurement at 8 months (219 missing due to death). The improvement in KCCQ-TSS at 8 months with dapagliflozin, compared with placebo, was greater in patients with higher BMI: placebo-corrected change 0.9 (−1.1, 2.8), 2.5 (0.8, 4.1), 1.9 (−0.1, 3.8), 2.7 (−0.5, 5.8), and 8.6 (4.0, 13.2) points, in normal weight, overweight, Class I obesity, Class II obesity, and Class III obesity, respectively (P-interaction = 0.03; Table 3 and Structured Graphical Abstract).
Change in weight according to baseline body mass index
Patients with a higher baseline BMI lost a greater amount of weight (at 12 months) with dapagliflozin. The placebo-corrected weight loss with dapagliflozin was: normal weight –0.88 (−1.28, –0.47) kg; overweight –0.65 (−1.04, –0.26) kg; Class I obesity –1.42 (−1.89, –0.94) kg; Class II obesity –1.17 (−1.94, –0.40) kg; and Class III obesity –2.5 (−4.4, –0.64) kg (P for interaction = 0.002).
Safety analyses
There was no significant interaction between BMI categories and the occurrence of adverse events according to randomized treatment (Table 4).
Table 4.
Adverse events
| Normal weight | Overweight | Obesity Class I | Obesity Class II | Obesity Class III | P-interaction | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Placebo | Dapagliflozin | Placebo | Dapagliflozin | Placebo | Dapagliflozin | Placebo | Dapagliflozin | Placebo | Dapagliflozin | ||
| N | 669 | 673 | 1046 | 1023 | 801 | 770 | 385 | 411 | 204 | 211 | |
| AE leading to discontinuation of randomized treatment | 42 (6.3) | 41 (6.1) | 61 (5.8) | 54 (5.3) | 45 (5.6) | 53 (6.9) | 21 (5.5) | 19 (4.6) | 12 (5.9) | 12 (5.7) | 0.79 |
| Amputation | 2 (0.3) | 2 (0.3) | 9 (0.9) | 9 (0.9) | 8 (1.0) | 5 (0.6) | 4 (1.0) | 2 (0.5) | 3 (1.5) | 1 (0.5) | 0.84 |
| Definite or probable DKAa | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (0.1) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (0.5) | n/a |
| Major hypoglycaemic event | 0 (0.0) | 1 (0.1) | 4 (0.4) | 3 (0.3) | 0 (0.0) | 2 (0.3) | 2 (0.5) | 1 (0.2) | 1 (0.5) | 1 (0.5) | 0.92 |
| Volume depletion SAE/DAE | 9 (1.3) | 8 (1.2) | 14 (1.3) | 18 (1.8) | 9 (1.1) | 14 (1.8) | 3 (0.8) | 2 (0.5) | 2 (1.0) | 6 (2.8) | 0.64 |
| Renal SAE/DAE | 16 (2.4) | 16 (2.4) | 27 (2.6) | 23 (2.2) | 23 (2.9) | 28 (3.6) | 18 (4.7) | 10 (2.4) | 6 (2.9) | 6 (2.8) | 0.58 |
AE, adverse event; SAE, serious adverse event; DAE, adverse event leading to discontinuation of randomized treatment; DKA, diabetic ketoacidosis.
Confirmed by independent adjudication committee.
Discussion
The key finding of this study was that dapagliflozin was equally efficacious in reducing the primary composite outcome of worsening HF or cardiovascular death across the spectrum of BMI in DELIVER, including among participants who were obese. Treatment with dapagliflozin also led to an improvement in symptoms measured with the KCCQ-TSS and which was greater in patients with a higher BMI. In addition, dapagliflozin treatment led to a modest but significantly larger reduction in weight in more obese patients than others.
In keeping with epidemiological observations, 45% of participants in this study were found to be obese and 78% were obese or overweight, despite the exclusion of patients with a BMI >50 kg/m2. As expected, the prevalence of diabetes and history of hypertension were higher (as was blood pressure) in patients with obesity. Participants with obesity had less evidence of coronary disease, which might also appear paradoxical but has been described previously and may reflect their younger age and the higher proportion of women. The association between higher BMI and lower NT-proBNP was confirmed and LVEF tended to be higher in participants with a higher BMI. Notably, patients with obesity had worse NYHA class, with almost twice as many in functional Class III or IV compared with participants with a normal weight. This was mirrored in self-reported symptoms and health-related quality of life, with a striking difference of over 20 points between patients with Class III obesity, compared with a normal weight, at baseline.
Regarding clinical outcomes during follow up, a ‘U-’ or ‘J-shaped’ relationship was found between BMI and mortality in patients with HFpEF (as in HF with reduced ejection fraction) with the nadir in crude risk among patients with Class I obesity (BMI range 30.0–34.9 kg/m2) and the highest risk of death in patients who were normal or underweight (although there were very few patients in the latter category).25–27 Interestingly, the pattern was different for worsening HF, where risk increased with increasing obesity, and was highest in patients with Class II/III obesity. Why the relationships between BMI and fatal, compared with non-fatal, outcomes diverged in this way is uncertain. The well-known, but poorly understood association between higher BMI and lower natriuretic peptides is a confounding factor relevant to these findings and in a fully adjusted analysis, including NT-proBNP, the association of obesity with lower mortality was eliminated. By contrast, the association with higher rates of worsening HF were not, possibly because the relationship between natriuretic peptides and plasma volume in patients with obesity may be different than in non-obese patients (obesity, unusually, is associated with low natriuretic peptide levels despite a greater expansion of plasma volume and higher filling pressures, especially during exercise).28
Dapagliflozin reduced the risk of the primary composite endpoint (worsening HF or cardiovascular death), with no interaction between the effect of treatment and BMI, examined as either a categorical or continuous variable, for this or any other outcome. Conservatively, applying the HR for the trial overall to each BMI category gave an NNT for the primary outcome in patients with Class III obesity of only 22 (over the median DELIVER follow up of 2.3 years), compared with 31 in patients with a normal weight, indicating a greater absolute benefit in obese individuals because of their higher absolute risk of this outcome.
As described above, patients with greater obesity had markedly worse health status at baseline and, importantly, this was improved following randomization to dapagliflozin. Indeed, the improvement in KCCQ-TSS was greatest in patients with the highest BMI. This finding is consistent with the hypothesis that the results of the PRESERVED-HF were more positive than EMPERIAL-Preserved because of the much higher prevalence of obesity in the former trial.29,30
While the main reason to use SGLT2 inhibitors in HFpEF is to improve symptoms and reduce worsening HF, their modest weight-reducing effect may be a useful additional attribute in patients with obesity. By increasing urinary excretion of glucose and calories, these agents have been shown to reduce weight in a range of patient populations, including patients with HF.14,15,27 The overall weight reduction at 12 months in DELIVER was just over 1 kg, slightly larger than in dapagliflozin and prevention of adverse outcomes in heart failure (DAPA-HF).27 The reduction in DELIVER was greater in patients with a higher BMI; the most obese participants with nearly three times the loss of those who had a normal weight (2.5 vs. 0.88 kg). Although modest, this additional benefit of SGLT2 inhibition may augment other strategies to reduce weight in HFpEF patients with obesity. One example is caloric restriction and aerobic exercise training which was shown to lead to a larger decrease in weight, and improvement in peak oxygen consumption but no consistent improvement in health-related quality of life.31,32 Glucagon-like peptide-1 receptor agonists and related treatments are currently under investigation as weight-loss treatments in HFpEF patients with obesity (www.ClinicalTrials.gov, NCT04788511, NCT04847557, and NCT04916470).
As with similar reports, our study had some limitations. Body mass index does not take into account the location of body fat or its amount, relative to muscle, or the weight of the skeleton, which may often differ according to sex, age, and race.33,34 Therefore, the conventional definition of obesity based on this metric may not account for these differences across populations, although our models were adjusted for race and region. Patients with extreme obesity (a BMI >50 kg/m2) were excluded and only 54 patients who were ‘underweight’ were enrolled, precluding any meaningful analysis, although this is a rare group of patients, at least in most countries.
Conclusion
In conclusion, the benefit of dapagliflozin on clinical outcomes was consistent across the spectrum of BMI in DELIVER, without any safety concerns. Treatment with dapagliflozin led to an improvement in symptoms and a modest reduction in weight, both of which were amplified in patients with a higher BMI.
Supplementary material
Supplementary material is available at European Heart Journal online.
Supplementary Material
Contributor Information
Carly Adamson, British Heart Foundation Cardiovascular Research Centre, University of Glasgow, 126 University Place, Glasgow G12 8TA, UK.
Toru Kondo, British Heart Foundation Cardiovascular Research Centre, University of Glasgow, 126 University Place, Glasgow G12 8TA, UK; Department of Cardiology, Nagoya University Graduate School of Medicine, Nagoya, Japan.
Pardeep S Jhund, British Heart Foundation Cardiovascular Research Centre, University of Glasgow, 126 University Place, Glasgow G12 8TA, UK.
Rudolf A de Boer, Department of Cardiology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands.
Jose Walter Cabrera Honorio, Clínica Vesalio, San Borja, Peru.
Brian Claggett, Cardiovascular Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA.
Akshay S Desai, Cardiovascular Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA.
Marco Antonio Alcocer Gamba, Centro de Estudios Clínicos de Querétaro (CECLIQ), Querétaro, México.
Waleed Al Habeeb, Cardiac Sciences Department, King Saud University, Riyadh, Saudi Arabia.
Adrian F Hernandez, Duke Clinical Research Institute, Duke University School of Medicine, Durham, NC, USA; Division of Cardiology, Department of Medicine, Duke University School of Medicine, Durham, NC, USA.
Silvio E Inzucchi, Section of Endocrinology, Yale University School of Medicine, New Haven, CT, USA.
Mikhail N Kosiborod, Saint Luke’s Mid America Heart Institute, University of Missouri-Kansas City, Kansas City, MO, USA.
Carolyn S P Lam, Department of Cardiology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands; National Heart Centre Singapore & Duke-National University of Singapore, Singapore, Singapore.
Anna Maria Langkilde, Late-Stage Development, Cardiovascular, Renal, and Metabolism, BioPharmaceuticals R&D, AstraZeneca, Gothenburg, Sweden.
Daniel Lindholm, Late-Stage Development, Cardiovascular, Renal, and Metabolism, BioPharmaceuticals R&D, AstraZeneca, Gothenburg, Sweden.
Erasmus Bachus, Late-Stage Development, Cardiovascular, Renal, and Metabolism, BioPharmaceuticals R&D, AstraZeneca, Gothenburg, Sweden.
Sheldon E Litwin, Division of Cardiology Medical University of South Carolina and Ralph H. Johnson, Veterans Affairs Medical Center, Charleston, SC, USA.
Felipe Martinez, Cordoba National University, Cordoba, Argentina.
Magnus Petersson, Late-Stage Development, Cardiovascular, Renal, and Metabolism, BioPharmaceuticals R&D, AstraZeneca, Gothenburg, Sweden.
Sanjiv J Shah, Division of Cardiology, Department of Medicine, and Bluhm Cardiovascular Institute, Northwestern University Feinberg School of Medicine, Chicago, IL, USA.
Muthiah Vaduganathan, Cardiovascular Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA.
Pham Nguyen Vinh, Cardiovascular Center, Tam Anh hospital, Tan Tao University, Ho Chi Minh City, Vietnam.
Ulrica Wilderäng, Late-Stage Development, Cardiovascular, Renal, and Metabolism, BioPharmaceuticals R&D, AstraZeneca, Gothenburg, Sweden.
Scott D Solomon, Cardiovascular Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA.
John J V McMurray, British Heart Foundation Cardiovascular Research Centre, University of Glasgow, 126 University Place, Glasgow G12 8TA, UK.
Funding
The DELIVER trial was funded by AstraZeneca. C.A. and J.J.V.M. are supported by a British Heart Foundation Centre of Research Excellence Grant RE/18/6/34217.
Data availability
The data sharing policy of the DELIVER trial sponsor, AstraZeneca, is described at https://astrazenecagrouptrials.pharmacm.com/ST/Submission/Disclosure.
References
- 1. Kenchaiah S, Evans JC, Levy D, Wilson PWF, Benjamin EJ, Larson MG, et al. Obesity and the risk of heart failure. N Engl J Med 2002;347:305–313. [DOI] [PubMed] [Google Scholar]
- 2. Aune D, Sen A, Norat T, Janszky I, Romundstad P, Tonstad S, et al. Body mass index, abdominal fatness, and heart failure incidence and mortality. Circulation 2016;133:639–649. [DOI] [PubMed] [Google Scholar]
- 3. Ho JE, Enserro D, Brouwers FP, Kizer JR, Shah SJ, Psaty BM, et al. Predicting heart failure with preserved and reduced ejection fraction: the International Collaboration on Heart Failure Subtypes. Circ Heart Fail 2016;9:e003116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Pandey A, Patel KV, Vaduganathan M, Sarma S, Haykowsky MJ, Berry JD, et al. Physical activity, fitness, and obesity in heart failure with preserved ejection fraction. JACC Heart Fail 2018;6:975–982. [DOI] [PubMed] [Google Scholar]
- 5. Savji N, Meijers WC, Bartz TM, Bhambhani V, Cushman M, Nayor M, et al. The association of obesity and cardiometabolic traits with incident HFpEF and HFrEF. JACC Heart Fail 2018;6:701–709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Kitzman DW, Lam CSP. Obese heart failure with preserved ejection fraction phenotype. Circulation 2017;136:20–23. [DOI] [PubMed] [Google Scholar]
- 7. Joyce E, Lala A, Stevens SR, Cooper LB, AbouEzzeddine OF, Groarke JD, et al. Prevalence, profile and prognosis of severe obesity in contemporary hospitalized heart failure trial populations. JACC Heart Fail 2016;4:923–931. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Reddy YNV, Rikhi A, Obokata M, Shah SJ, Lewis GD, AbouEzzedine OF, et al. Quality of life in heart failure with preserved ejection fraction: importance of obesity, functional capacity, and physical inactivity. Eur J Heart Fail 2020;22:1009–1018. [DOI] [PubMed] [Google Scholar]
- 9. Packer M. Derangements in adrenergic–adipokine signalling establish a neurohormonal basis for obesity-related heart failure with a preserved ejection fraction. Eur J Heart Fail 2018;20:873–878. [DOI] [PubMed] [Google Scholar]
- 10. Packer M, Kitzman DW. Obesity-Related heart failure with a preserved ejection fraction. JACC Heart Fail 2018;6:633–639. [DOI] [PubMed] [Google Scholar]
- 11. Neeland IJ, Gupta S, Ayers CR, Turer AT, Rane JE, Das SR, et al. Relation of regional fat distribution to left ventricular structure and function. Circ Cardiovasc Imaging 2013;6:800–807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Patel VB, Shah S, Verma S, Oudit GY. Epicardial adipose tissue as a metabolic transducer: role in heart failure and coronary artery disease. Heart Fail Rev 2017;22:889–902. [DOI] [PubMed] [Google Scholar]
- 13. Rao VN, Fudim M, Mentz RJ, Michos ED, Felker GM. Regional adiposity and heart failure with preserved ejection fraction. Eur J Heart Fail 2020;22:1540–1550. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Cai X, Ji L, Chen Y, Yang W, Zhou L, Han X, et al. Comparisons of weight changes between sodium-glucose cotransporter 2 inhibitors treatment and glucagon-like peptide-1 analogs treatment in type 2 diabetes patients: a meta-analysis. J Diabetes Investig 2017;8:510–517. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Lee PC, Ganguly S, Goh SY. Weight loss associated with sodium-glucose cotransporter-2 inhibition: a review of evidence and underlying mechanisms. Obes Rev 2018;19:1630–1641. [DOI] [PubMed] [Google Scholar]
- 16. Zaccardi F, Webb DR, Htike ZZ, Youssef D, Khunti K, Davies MJ. Efficacy and safety of sodium-glucose co-transporter-2 inhibitors in type 2 diabetes mellitus: systematic review and network meta-analysis. Diabetes Obes Metab 2016;18:783–794. [DOI] [PubMed] [Google Scholar]
- 17. Baker WL, Smyth LR, Riche DM, Bourret EM, Chamberlin KW, White WB. Effects of sodium-glucose co-transporter 2 inhibitors on blood pressure: a systematic review and meta-analysis. J Am Soc Hypertens 2014;8:262–275. [DOI] [PubMed] [Google Scholar]
- 18. Anker SD, Butler J, Filippatos G, Ferreira JP, Bocchi E, Böhm M, et al. Empagliflozin in heart failure with a preserved ejection fraction. N Engl J Med 2021;385:1451–1461. [DOI] [PubMed] [Google Scholar]
- 19. Solomon SD, McMurray JJV, Claggett BL, de Boer RA, DeMets D, Hernandez AF, et al. Dapagliflozin in heart failure with a mildly reduced or preserved ejection fraction. N Engl J Med 2022. doi: 10.1056/NEJMoa2206286. [DOI] [PubMed] [Google Scholar]
- 20. Butler J, Anker SD, Filippatos G, Khan MS, Ferreira JP, Pocock SJ, et al. EMPEROR-reduced trial committees and investigators. Empagliflozin and health-related quality of life outcomes in patients with heart failure with reduced ejection fraction: the EMPEROR-reduced trial. Eur Heart J 2021;42:1203–1212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Solomon SD, de Boer RA, DeMets D, Hernandez AF, Inzucchi SE, Kosiborod MN, et al. Dapagliflozin in heart failure with preserved and mildly reduced ejection fraction: rationale and design of the DELIVER trial. Eur J Heart Fail 2021;23:1217–1225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Solomon SD, Vaduganathan M, Claggett BL, de Boer RA, DeMets D, Hernandez AF, et al. Baseline characteristics of patients with HF with mildly reduced and preserved ejection fraction. JACC Heart Fail 2022;10:184–197. [DOI] [PubMed] [Google Scholar]
- 23. World Health Organisation . Obesity: Preventing and managing the global epidemic. Report of a WHO Consultation (WHO Technical Report Series 894). 2000.
- 24. Lin DY, Wei LJ, Yang I, Ying Z. Semiparametric regression for the mean and rate functions of recurrent events. J R Stat Soc Series B Stat Methodol 2000;62:711–730. [Google Scholar]
- 25. Horwich TB, Fonarow GC, Clark AL. Obesity and the obesity paradox in heart failure. Prog Cardiovasc Dis 2018;61:151–156. [DOI] [PubMed] [Google Scholar]
- 26. Lavie CJ, Sharma A, Alpert MA, De Schutter A, Lopez-Jimenez F, Milani RV, et al. Update on obesity and obesity paradox in heart failure. Prog Cardiovasc Dis 2016;58:393–400. [DOI] [PubMed] [Google Scholar]
- 27. Adamson C, Jhund PS, Docherty KF, Bělohlávek J, Chiang CE, Diez M, et al. Efficacy of dapagliflozin in heart failure with reduced ejection fraction according to body mass index. Eur J Heart Fail 2021;23:1662–1672. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Miller WL, Borlaug BA. Impact of obesity on volume Status in patients with ambulatory chronic heart failure. J Card Fail 2020;26:112–117. [DOI] [PubMed] [Google Scholar]
- 29. Abraham WT, Lindenfeld J, Ponikowski P, Agostoni P, Butler J, Desai AS, et al. Effect of empagliflozin on exercise ability and symptoms in heart failure patients with reduced and preserved ejection fraction, with and without type 2 diabetes. Eur Heart J 2021;42:700–710. [DOI] [PubMed] [Google Scholar]
- 30. Nassif ME, Windsor SL, Borlaug BA, Kitzman DW, Shah SJ, Tang F, et al. The SGLT2 inhibitor dapagliflozin in heart failure with preserved ejection fraction: a multicenter randomized trial. Nat Med 2021;27:1954–1960. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. El Hajj EC, El Hajj MC, Sykes B, Lamicq M, Zile MR, Malcolm R, et al. Pragmatic weight management program for patients with obesity and heart failure with preserved ejection fraction. J Am Heart Assoc 2021;10:e022930. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Kitzman DW, Brubaker P, Morgan T, Haykowsky M, Hundley G, Kraus WE, et al. Effect of caloric restriction or aerobic exercise training on peak oxygen consumption and quality of life in obese older patients with heart failure with preserved ejection fraction: a randomized clinical trial. JAMA 2016;315:36–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Heymsfield SB, Peterson CM, Thomas DM, Heo M, Schuna JM Jr. Why are there race/ethnic differences in adult body mass index–adiposity relationships? A quantitative critical review. Obes Rev 2016;17:262–275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Li C, Engström G, Hedblad B, Calling S, Berglund G, Janzon L. Sex differences in the relationships between BMI, WHR and incidence of cardiovascular disease: a population-based cohort study. Int J Obes 2006;30:1775–1781. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data sharing policy of the DELIVER trial sponsor, AstraZeneca, is described at https://astrazenecagrouptrials.pharmacm.com/ST/Submission/Disclosure.