Fontan haemodynamics in adults with obesity compared with overweight and normal body mass index: a retrospective invasive exercise study

Abstract

Aims

The effects of obesity on Fontan haemodynamics are poorly understood. Accordingly, we assessed its impact on exercise invasive haemodynamics and exercise capacity.

Methods and results

Seventy-seven adults post-Fontan undergoing exercise cardiac catheterization (supine cycle protocol) were retrospectively identified using an institutional database and categorized according to the presence of obesity [body mass index (BMI) > 30 kg/m²] and overweight/normal BMI (BMI ≤ 30 kg/m²). There were 18 individuals with obesity (BMI 36.4 ± 3 kg/m²) and 59 with overweight/normal BMI (BMI 24.1 ± 3.6 kg/m²). Peak oxygen consumption (VO₂) on non-invasive cardiopulmonary exercise testing was lower in patients with obesity (15.6 ± 3.5 vs. 19.6 ± 5.8 mL/kg/min, P = 0.04). At rest, systemic flow (Q_s) [7.0 (4.8; 8.3) vs. 4.8 (3.9; 5.8) L/min, P = 0.001], pulmonary artery (PA) pressure (16.3 ± 3.5 vs. 13.1 ± 3.5 mmHg, P = 0.002), and PA wedge pressure (PAWP) (11.7 ± 4.4 vs. 8.9 ± 3.1 mmHg, P = 0.01) were higher, while arterial O₂ saturation was lower [89.5% (86.5; 92.3) vs. 93% (90; 95)] in obesity compared with overweight/normal BMI. Similarly, patients with obesity had higher exercise PA pressure (29.7 ± 6.5 vs. 24.7 ± 6.8 mmHg, P = 0.01) and PAWP (23.0 ± 6.5 vs. 19.8 ± 7.3 mmHg, P = 0.047), but lower arterial O₂ saturation [82.4 ± 7.0% vs. 89% (85; 92), P = 0.003].

Conclusion

Adults post-Fontan with obesity have worse aerobic capacity, increased Q_s, higher filling pressures, and decreased arterial O₂ saturation compared with those with overweight/normal BMI, both at rest and during exercise, mirroring the findings observed in the obesity phenotype of heart failure with preserved ejection fraction. Whether treating obesity and its cardiometabolic sequelae in Fontan patients will improve haemodynamics and outcomes requires further study.

Graphical Abstract

Graphical Abstract.

See the editorial comment for this article ‘The obesity epidemic meets complex congenital heart disease: challenges for prevention’, by W.A. Helbing, https://doi.org/10.1093/eurjpc/zwae322.

Introduction

The progressive increase in the prevalence of obesity is a worldwide phenomenon. As patients with repaired congenital heart disease age, they are invariably at risk of acquired disorders, including obesity and its related comorbidities. In agreement with prior reports,¹ obesity was present in almost 30% of adults seen in our adult congenital heart disease practice (mean age <40 years), with hypertension and/or hyperlipidaemia present in more than in one-third of them and diabetes seen in 12–21% of individuals according to the degree of obesity.² Alarmingly, weight gain was accompanied by an increased risk of cardiovascular events during follow-up. Achieving weight loss in this population is challenging due to the frequent coexistence of exercise intolerance and deconditioning. Albeit preliminary, the use of glucagon-like peptide-1 receptor (GLP-1) agonists appears to be safe in adults with congenital heart disease and can be effective in weight loss.³

Although patients post-Fontan are often underweight in childhood, it has been reported that ∼40% of adult Fontan patients in the USA are overweight or obese,⁴ with rates similar to the general population. Currently, data on the clinical and prognostic implications of obesity in adults post-Fontan are limited, and higher rates of overweight and obesity have been reported among those with lower performance on cardiopulmonary exercise testing (CPET) compared with those with more preserved exercise capacity.⁵

Obesity-related changes to cardiac physiology in patients with biventricular circulation have been well described, including increased cardiac output and ventricular size/mass, elevation in ventricular filling pressures, and reduced functional capacity.⁶ Fontan physiology is universally associated with increased central venous pressure and low-normal cardiac index. In addition to the inherent limitations due to the passive nature of venous return, other non-cardiac factors, such as abnormal respiratory mechanics and myopenia, further contribute to exercise intolerance documented in post-Fontan patients.⁷ Thus, it would be intuitive to expect the co-occurrence of obesity in adults post-Fontan to impact the haemodynamic milieu and functional capacity further. However, the implications of obesity to Fontan physiology and, particularly, exercise haemodynamics are scarce.

To address this gap in knowledge, the goals of the present study were to assess the impact of obesity among adults post-Fontan palliation on: (i) resting and exercise invasive haemodynamics and (ii) exercise capacity, reflected by peak oxygen consumption (VO₂) at the time of non-invasive and invasive CPET.

Methods

In this retrospective cross-sectional study, adults (age ≥18 years) undergoing exercise cardiac catheterization at Mayo Clinic, MN, USA, between November 2018 and June 2023 were retrospectively identified using an electronic institutional database. The Institutional Review Board approved the study, and given the retrospective design, the need for individual written consent was waived. Seventy-eight individuals were initially identified; one patient was subsequently excluded due to a lack of prior research authorization for the use of the medical chart for research purposes. Accordingly, the final retrospective cohort included 77 patients. Patients then were categorized according to body mass index (BMI) at the time of catheterization as having obesity (BMI >30 kg/m²) and overweight/normal BMI (BMI ≤30 kg/m²) according to the following formula: BMI = weight (kg)/[height (m)]². Due to the small number of patients with underweight (BMI <18.5 kg/m²; n=5), these patients were analyzed and referred to as having normal BMI.

Exercise cardiac catheterization was done as clinically indicated at the discretion of the referring provider. The procedure was undertaken in a fasting state under mild sedation without discontinuation of regular medications. Venous catheterization was performed via jugular approach using a 7 Fr balloon-tipped catheter inserted through an 8 Fr sheath to allow for simultaneous superior vena cava pressure (SVC) measurement. Following standard resting cardiac catheterization and before administration of iodine contrast, exercise was performed using a supine cycle protocol, as previously reported.^8,9 Firstly, the patient’s feet were allowed to rest comfortably on the bicycle pedals (feet-up phase). Workload was then increased by 20 W intervals every 2 min until maximal perceived exhaustion (Borg exertion scale ≥17 and/or respiratory exchange ratio >1.0). Pressure measurements were obtained at rest, during feet-up, and during every stage of exercise, while oximetry data were obtained at rest and during peak exercise. Resting and exercise systemic (Q_s) and pulmonary (Q_p) flows were calculated by the direct Fick principle with VO₂ measured throughout the study. Arterial O₂ saturation was assessed by direct arterial sampling or pulse oximetry if radial access was not obtained. Pressure recordings correspond to a computer-generated average of ≥5 consecutive beats, measured throughout the entire respiratory cycle during spontaneous breathing. Reflecting the standards of our laboratory, pressure records were individually reviewed by two technicians and by the responsible invasive cardiologist to ensure the accuracy of the measurements. Elevated pulmonary artery (PA) pressure was defined as ≥14 mmHg¹⁰ and elevated PA wedge pressure (PAWP) as ≥12 mmHg.¹¹

Angiography data obtained at the time of cardiac catheterization were individually reviewed offline by one of the co-authors (W.R.M.), who was blinded to the haemodynamic data. Veno-venous collaterals correspond to the connections between the supra- or infra-diaphragmatic systemic veins and the pulmonary veins/functional left atrium. Fontan conduit/pathway obstruction was defined as an inferior vena cava to PA pressure gradient ≥3 mmHg at rest and ≥5 mmHg during exercise.¹²

Clinical data were abstracted from the electronic records and represent the most recent values prior to cardiac catheterization. Non-invasive CPET was performed using a treadmill or semi-recumbent bicycle protocol (53 and 6 patients, respectively).¹³ Ideal body weight was calculated by the following formula: a + b × (height [cm] − 150), where a = 48 in men and 45 in women, while b = 1.1 in men and 0.9 in women.¹⁴ Estimated plasma volume was calculated as: (1−hematocrit [%]/100) × (a + [b × weight (kg)]), where a = 1530 for men and 864 for women, while b = 41 for men and 47.9 for women.¹⁵ Serologic liver scores were calculated as previously described.¹⁶

Statistical analyses

Continuous data were presented as mean ± standard deviation or median (25^th and 75^th percentile). Nominal data were presented as counts (%). Between-group comparisons were made using Fisher’s exact test or the Wilcoxon rank test. N-terminal pro-brain natriuretic peptide (NT-pro-BNP) levels were analysed as log-transformed (ln). Multivariable logistic regression analyses were performed to assess the association between obesity and haemodynamics while adjusting for potential confounders. For these analyses, only haemodynamic variables with a P < 0.1 on between-group comparisons were included; co-variables (age, presence of pacemaker, ventricular morphology, ventricular ejection fraction, and atrioventricular valve regurgitation) were chosen a priori based on clinical importance and potential impact on Fontan haemodynamics as well as relationship with clinical outcomes on previous studies.^17–20 JMP software (SAS Institute Inc., Cary, NC, USA) was used for statistical analyses, and a P-value <0.05 was considered statistically significant.

Results

Clinical characteristics

There were 18 (23.7%) patients with obesity and 59 (76.3%) with overweight/normal BMI. Clinical data for the entire cohort are presented in Table 1. By design, BMI was higher in the group with obesity compared with the rest of the cohort (36.4 ± 3.3 vs. 24.1 ± 3.6 kg/m²), while ideal body weight did not differ [71.8 (55.4; 78.0) vs. 67.8 ± 12.7 kg, P = 0.92]. Estimated plasma volume was significantly lower in patients without obesity (3178 ± 399 vs. 2309 ± 363 mL, P < 0.001). There was no difference in age at the time of catheterization [32.7 (29.5; 37.1) vs. 31.5 (23.7; 37.2) years, P = 0.36] or at the time of Fontan palliation [3 (2; 8) vs. 4 (2.8; 5) years, P = 0.81] between groups. Prior pacemaker implantation was less prevalent among individuals with obesity compared with those without obesity (5.6 vs. 50.8%, P < 0.001), but there were no other differences in clinical characteristics, ventricular systolic function, or degree of atrioventricular valve regurgitation between groups.

Table 1.

Clinical characteristics

	Obesity (n = 18)	Overweight or normal BMI (n = 59)	P-value
Age, years	32.7 (29.5; 37.1)	31.5 (23.7; 37.2)	0.36
Body mass index, kg/m2	36.4 ± 3.3	24.1 ± 3.6	—
Ideal weight, kg	71.8 (55.4; 78.0)	67.8 ± 12.7	0.92
Estimated plasma volume, mL	3178 ± 399	2309 ± 363	<0.001
Female sex	8 (44.4%)	23 (39.0%)	0.79
Patent fenestration	4 (22.2%)	8 (13.6%)	0.46
NYHA classes III or IV	8 (44.4%)	20 (33.9%)	0.42
Hypertension	1 (5.6%)	6 (10.2%)	0.99
Diabetes	2 (11.1%)	2 (3.4%)	0.23
Pacemaker	1 (5.6%)	30 (50.8%)	<0.001
History of atrial arrhythmias	10 (55.6%)	28 (47.5%)	0.60
Creatinine clearance <60 mL/min	0	5 (8.5%)	0.59
Symptoms
Dyspnoea	12 (66.7%)	37 (62.7%)	0.99
Fatigue	12 (66.7%)	24 (40.7%)	0.06
Oedema	8 (44.4%)	22 (37.3%)	0.59
Medications
Diuretics	9 (50%)	30 (50.8%)	0.99
Beta-blocker	8 (44.4%)	19 (32.2%)	0.40
ACEi/ARB	11 (61.1%)	32 (54.2%)	0.79
Aldosterone antagonist	5 (27.8%)	19 (32.2%)	0.78
Digitalis	0	9 (15.3%)	0.11
Antiarrhythmic agent	7 (38.9%)	15 (25.4%)	0.37
Phosphodiesterase type 5 inhibitor	7 (38.9%)	13 (22.0%)	0.22
Laboratory/imaging
NT-pro-BNP, pg/mL	183 (125.5; 573.5)	322 (148.5; 633.5)	0.78
MELD-XI	9.4 (9.4; 12.7)	10.6 (9.4; 12.4)	0.53
APRI	0.4 (0.3; 0.6)	0.4 (0.3; 0.6)	0.49
Spleen size, mm	14.0 ± 2.0	13.3 ± 2.6	0.15
Echocardiography
Ventricular ejection fraction, %	53.8 (50; 57.5)	52.5 (45; 57.5%)	0.63
≥moderate AV regurgitation	1 (5.6%)	13 (22.0%)	0.17
Non-invasive cardiopulmonary exercise test
Respiratory exchange ratio	1.08 ± 0.1	1.15 ± 0.1	0.07
Peak VO2, mL/min	1713 ± 539	1391 (1171; 1718)	0.15
Minimal O2 saturation by pulse oximetry, %	90 (83; 92)	91 (86; 94)	0.72
Peak VO2, mL/kg/min	15.5 ± 3.5	19.9 ± 5.8	0.03
Peak VO2, % of predicted	41.6 ± 10.9	49.6 ± 13.2	0.07
VE/VCO2 nadir	33.7 ± 6.0	33.7 ± 5.8	0.99
Peak ventilation, % predicted	50.2 ± 11.5	46.4 ± 16.7	0.33

ACEi/ARB, angiotensin-converting enzyme inhibitor/angiotensin receptor blocker; APRI, aspartate aminotransferase to platelet ratio index; AV, atrioventricular valve; MELD-XI, model for end-stage liver disease excluding international normalized ratio; NT-pro-BNP, N-terminal pro-brain natriuretic peptide; NYHA, New York Heart Association; VE, minute ventilation; VO₂, oxygen consumption; VCO₂, carbon dioxide output.

Values in bold indicate statistical significance (defined as P-value <0.05).

Underlying congenital lesions in the group with obesity were as follows: tricuspid atresia and double-inlet left ventricle in 5 (27.8%) each, hypoplastic left heart syndrome in 3 (16.7%), pulmonary atresia/intact ventricular septum and double-outlet right ventricle in one (6%) each, and other in 3 (16.5%). Among patients with overweight/normal BMI, congenital defects were as follows: double-inlet left ventricle in 15 (25.4%), double-outlet right ventricle in 12 (20.3%), pulmonary atresia/intact ventricular septum, hypoplastic left heart syndrome, and tricuspid atresia in 10 (16.5% each), and other in 2 (3.4%). There was no difference in predominant right ventricular morphology between groups (obesity 27.8% vs. overweight/normal BMI 40.7%, P = 0.41).

Fontan connections among patients with obesity were as follows: lateral tunnel in 7 (38.9%), extracardiac conduit in 6 (33.3%), and atriopulmonary in 5 (27.8%). Among patients without obesity, Fontan types were as follows: extracardiac conduit in 25 (42.3%), lateral tunnel in 19 (32.2%), atriopulmonary in 7 (11.2%), intra-atrial in 5 (8%), and other in 3 (5.2%). Fontan conduit obstruction was present in 12.5% of individuals with obesity and 14.5% of patients without obesity (P = 0.99), while a patent fenestration was seen in 22.2 and 13.6% (P = 0.46), respectively.

There was no difference in the New York Heart Association functional Classes III or IV between groups (obesity 44.4% vs. overweight/normal BMI 33.9%; P = 0.42), but fatigue tended to be more common in individuals with obesity (66.7 vs. 40.7%, P = 0.06). On non-invasive CPET, peak VO₂ indexed to body mass was lower in patients with obesity (15.5 ± 3.5 vs. 19.9 ± 5.6 mL/kg/min, P = 0.03), while no differences in minimal O₂ saturation were seen [90% (83; 92) vs. 91% (86; 94); P = 0.72]. There were no statistically significant differences in NT-pro-BNP levels [obesity 183 (125.5; 573.5) vs. overweight/normal BMI 322 (148.5; 633.5) pg/mL, P = 0.78].

Resting haemodynamics

Table 2 and Figure 1 present resting haemodynamic data. Superior vena cava pressure (17.0 ± 3.1 vs. 13.8 ± 3.5 mmHg, P = 0.002), PA pressure (16.3 ± 3.5 vs. 13.1 ± 3.5 mmHg, P = 0.002), and PAWP (11.7 ± 4.4 vs. 8.9 ± 3.1 mmHg, P = 0.01) were higher in patients with obesity compared with the rest of the cohort. Elevated PAWP (55.6 vs. 24.6%, P = 0.02) and PA (83.3 vs. 41.4%, P = 0.003) pressures were more prevalent in the group obesity. Arterial O₂ saturation was lower in patients with obesity [89.5% (86.5; 92.3) vs. 93% (90; 95), P = 0.006], and veno-venous collaterals tended to be more prevalent in this group than in individuals without obesity (66.7 vs. 37.3%, P = 0.07).

Table 2.

Resting cardiac catheterization

	Obesity (n = 18)	Overweight or normal BMI (n = 59)	P-value
Resting haemodynamics
Haemoglobin, g/dL	14.2 ± 1.7	14.1 ± 2.2	0.97
Arterial O2 saturation, %	89.5 (86.5; 92.3)	93 (90; 95)	0.006
SVC, mmHg	17.0 ± 3.1	13.8 ± 3.5	0.002
PA, mmHg	16.3 ± 3.5	13.1 ± 3.5	0.002
PAWP, mmHg	11.7 ± 4.4	8.9 ± 3.1	0.01
Arterial systolic pressure, mmHg	120.7 ± 17.7	121.9 ± 20.3	0.80
Arterial diastolic pressure, mmHg	61.5 ± 15.8	65.2 ± 12.0	0.41
Arterial mean pressure, mmHg	78.1 ± 16.2	83.7 ± 13.9	0.16
Qp, L/min	5.1 (3.7; 6.1)	4.3 ± 1.3	0.08
Qs, L/min	7.0 (4.8; 8.3)	4.8 (3.9; 5.8)	0.001
Cardiac index, L/min/m2	3.3 (2.3; 3.8)	2.7 (2.1; 3.1)	0.06
Stroke volume, mL	102.5 ± 48.2	68.3 (52.0; 83.7)	0.008
Stroke volume index, mL/m2	44.8 ± 17.8	39.1 ± 11.0	0.27
Heart rate, b.p.m.	72.7 ± 17.8	71.5 ± 10.4	0.97
PVRi, U.m2	2.0 ± 0.9	1.8 (1.2; 2.3)	0.60
PVR, WU	0.9 (0.6; 1.2)	0.9 (0.7; 1.3)	0.56
SVR, dynes/s/cm−5	817 ± 494	1201 ± 415	0.007
SVRi, dynes/s/cm−5/m2	385 ± 238	620 (483; 821)	0.001
VO2, mL/min/kg	269.6 ± 52.6	207.9 ± 46.9	<0.001
VO2, mL/min	2.6 ± 0.4	3.0 ± 0.6	0.003
Angiographya
Veno-venous collateral	10 (66.7%)	19 (37.3%)	0.07

PA, pulmonary artery; PAWP, pulmonary artery wedge pressure; PVR, pulmonary vascular resistance; PVRi, pulmonary vascular resistance index; Q_p, pulmonary flow; Q_s, systemic flow; SVC, superior vena cava; SVR, systemic vascular resistance; SVRi, systemic vascular resistance index; VO₂, oxygen consumption.

Values in bold indicate statistical significance (defined as P-value <0.05).

^aAngiography available/performed in 66 individuals.

Figure 1.

Resting haemodynamics in obese and non-obese adults post-Fontan. PA, pulmonary artery; PAWP, PA wedge pressure; Q_s, systemic flow; SVi, stroke volume index; SVR, systemic vascular resistance.

Q _s [7.0 (4.8; 8.3) vs. 4.8 (3.9; 5.8) L/min, P = 0.001] and stroke volume [102.5 ± 48.2 vs. 68.3 (52.0; 83.7) mL, P= 0.008] were higher in individuals with obesity. While there was a trend towards higher cardiac index in the group with obesity [3.3 (2.3; 3.8) vs. 2.7 (2.1; 3.1) L/min/m², P = 0.06], there was no difference in stroke volume index (44.8 ± 17.8 vs. 39.1 ± 11.0 mL/m², P = 0.27). Systemic vascular resistance (SVR) was lower among individuals with obesity (817 ± 494 vs. 1201 ± 415 dynes/s/cm⁻⁵; P = 0.007).

The results of logistic regression analyses are presented in Table 4. After adjusting for confounders, obesity was independently associated with SVC pressures, PA pressures, PAWP, arterial O₂ saturation, Q_s, stroke volume, indexed and absolute VO₂, and indexed and non-indexed SVR. Similarly, obesity was independently associated with the presence of veno-venous collaterals [odds ratio 6.5 (95% confidence interval 1.4; 30.1); P = 0.02].

Table 4.

Multivariable logistic regression

	β Coefficient ± SEa	P-value
Resting haemodynamics
Arterial O2 saturation, %	−1.9 ± 0.8	0.02
SVC, mmHg	1.7 ± 0.5	0.003
PA, mmHg	1.6 ± 0.6	0.005
PAWP, mmHg	1.5 ± 0.6	0.01
Qp, L/min	0.4 ± 0.2	0.05
Qs, L/min	0.9 ± 0.3	0.01
Cardiac index, L/min/m2	0.2 ± 0.1	0.13
Stroke volume, mL	15.3 ± 4.8	0.002
SVR, dynes/s/cm−5	−203 ± 83	0.02
SVRi, dynes/s/cm−5/m2	−141 ± 52	0.01
VO2, mL/min/kg	−0.2 ± 0.1	0.01
VO2, mL/min	27.8 ± 6.8	<0.001
Exercise haemodynamics
Arterial O2 saturation, %	−2.7 ± 1.2	0.03
SVC, mmHg	4.0 ± 1.0	<0.001
PA, mmHg	3.0 ± 1.0	0.004
PAWP, mmHg	3.1 ± 1.1	0.006
Qs, L/min	0.5 ± 0.4	0.24
Stroke volume, mL	15.9 ± 4.2	<0.001
SVR, dynes/s/cm−5	−94 ± 47	0.05
SVRi, dynes/s/cm−5/m2	−60 ± 32	0.07
VO2, mL/min/kg	−1.9 ± 0.5	<0.001
Angiography
	Odds ratio (95% CI) a	P-value
Veno-venous collateral	6.5 (1.4; 30.1)	0.02

CI, confidence interval; PA, pulmonary artery; PAWP, pulmonary artery wedge pressure; Q_p, pulmonary flow; Q_s, systemic flow; SVC, superior vena cava; SE, standard error; SVR, systemic vascular resistance; SVRi, systemic vascular resistance index; VO₂, oxygen consumption.

Values in bold indicate statistical significance (defined as P-value <0.05).

^aβ Coefficient and odds ratio represent the association between obesity (vs. non-obesity) as the predictor variable and individual haemodynamic measurements on the outcome variable, while adjusting for age, pacemaker, ventricular ejection, ventricular morphology, and atrioventricular valve regurgitation.

Exercise haemodynamics

The results of exercise haemodynamics are presented in Table 3 and Figure 2. There was no difference in load [80 (60; 100) vs. 60 (60; 80) W, P = 0.56] or respiratory exchange ratio (1.02 ± 0.1 vs. 1.01 ± 0.1, P = 0.47) between patients with or without obesity at peak exercise, but, similar to the non-invasive CPET, patients with obesity had lower peak VO₂ indexed to body mass during cardiac catheterization (10.4 ± 2.9 vs. 13.5 ± 3.9 mL/min/kg, P = 0.002). Patients with obesity had higher SVC pressure (30.6 ± 5.0 vs. 24.4 ± 7.3 mmHg, P = 0.001), PA pressure (29.7 ± 6.5 vs. 24.7 ± 6.8 mmHg, P = 0.01), and PAWP (23.0 ± 6.5 vs. 19.8 ± 7.3 mmHg, P = 0.047). There was no difference in ΔPAWP/ΔQ_s [3.2 (1.8; 9.6) vs. 3.9 (1.1; 7.2) mmHg/L/min, P = 0.45] or ΔPA/ΔQ_p [4.2 (2.7; 9.5) vs. 4.2 (1.9; 7.4) mmHg/L/min, P = 0.35] between patients with or without obesity.

Table 3.

Haemodynamics during feet-up and peak exercise

	Obesity (n = 18)	Overweight or normal BMI (n = 59)	P-value
Feet-up
SVC, mmHg	20.3 ± 5.8	16.2 ± 4.3	0.008
PAWP, mmHg	14.5 (10.8; 17.8)	12.1 ± 3.6	0.05
Exercise
Arterial O2 saturation, %	82.4 ± 7.0	89 (85; 92)	0.003
SVC, mmHg	30.6 ± 5.0	24.4 ± 7.3	0.001
PA, mmHg	29.7 ± 6.5	24.7 ± 6.8	0.01
PAWP, mmHg	23.0 ± 6.5	19.8 ± 7.3	0.047
Arterial systolic pressure, mmHg	154.3 ± 33.7	156.3 ± 35.1	0.91
Arterial diastolic pressure, mmHg	68.8 ± 11.3	76.1 ± 13.4	0.10
Arterial mean pressure, mmHg	97.5 ± 15.7	103.0 ± 18.2	0.41
Qp, L/min	8.3 ± 2.6	7.6 ± 2.5	0.34
Qs, L/min	10.3 ± 2.8	8.5 ± 2.5	0.04
Stroke volume, mL	111.3 ± 36.9	80.2 ± 21.9	0.001
Stroke volume index, mL/m2	50.3 ± 14.7	44.5 ± 11.1	0.24
Heart rate, b.p.m.	100.5 (78; 120.3)	109.3 ± 27.1	0.11
Qs response, % predicted	65.8 (40.2; 93.1)	82.3 (54.6; 104.0)	0.17
Qp response, % predicted	58.1 ± 24.9	66.3 (51.6; 90.1)	0.20
PVRi, U.m2	1.6 ± 0.8	1.3 ± 0.8	0.15
SVR, dynes/s/cm−5	529 ± 155	785 ± 238	0.002
SVRi, dynes/s/cm−5/m2	245 ± 84	415 (328; 560)	<0.001
ΔPAWP/ΔQs, mmHg/L/min	3.2 (1.8; 9.6)	3.9 (1.1; 7.2)	0.45
ΔPA/ΔQp, mmHg/L/min	4.2 (2.7; 9.5)	4.2 (1.9; 7.4)	0.35
VO2, mL/min/kg	10.4 ± 2.9	13.5 ± 3.9	0.002
VO2, mL/min	1098 ± 357	948 ± 329	0.12
Load, W	80 (60; 100)	60 (60; 80)	0.56
Respiratory exchange ratio	1.02 ± 0.1	1.01 ± 0.1	0.47

PA, pulmonary artery; PAWP, pulmonary artery wedge pressure; PVR, pulmonary vascular resistance; PVRi, pulmonary vascular resistance index; Q_p, pulmonary flow; Q_s, systemic flow; SVC, superior vena cava; SVR, systemic vascular resistance; SVRi, systemic vascular resistance index; VO₂, oxygen consumption.

Values in bold indicate statistical significance (defined as P-value <0.05).

Figure 2.

Exercise haemodynamics in obese and non-obese adults post-Fontan. PA, pulmonary artery; PAWP, PA wedge pressure; Q_s, systemic flow; SVi, stroke volume index; SVR, systemic vascular resistance.

Exercise Q_s (10.3 ± 2.8 vs. 8.5 ± 2.5 L/min, P = 0.04) and stroke volume (111.3 ± 36.9 vs. 80.2 ± 21.9 mL, P = 0.001) were higher in patients with obesity, while no differences were observed for stroke volume index (50.3 ± 14.7 vs. 44.5 ± 11.1 mL/m², P = 0.24). Arterial O₂ saturation [82.4 ± 7.0% vs. 89% (85; 92), P = 0.003] and SVR (529 ± 155 vs. 785 ± 238 dynes/s/cm⁻⁵, P = 0.002) were lower in individuals with obesity.

The results of regression analyses including exercise haemodynamics are presented in Table 4. Obesity was independently associated with SVC pressures, PA pressures, PAWP, arterial O₂ saturation, stroke volume, and indexed VO₂ after adjusting for confounders.

Discussion

To the best of our knowledge, this is the first attempt to delineate the impact of obesity on exercise haemodynamics in adults post-Fontan. Compared with individuals with overweight or normal BMI, adults with obesity post-Fontan demonstrated (Graphical Abstract): (i) lower aerobic capacity on non-invasive and invasive CPET; (ii) increased resting Q_s as well resting and exercise stroke volume; (iii) higher resting and exercise systemic venous and ventricular filling pressures; (iv) lower resting SVR; and (v) more prominent degrees of arterial desaturation at catheterization and increased prevalence of veno-venous collaterals.

Pathologic cardiac remodelling associated with obesity was recognized almost a century ago,²¹ with subsequent autopsy studies confirming the secondary increases in cardiac mass and ventricular enlargement in those individuals.²² The interest in the interplay between obesity and cardiac physiology has recently grown, given the overlap between the former and heart failure with preserved ejection fraction (HFpEF).⁶ A distinct phenotype of patients with HFpEF and obesity has been described,²³ which includes lower levels of serum natriuretic peptides, higher plasma volume, increased cardiac output and ventricular size, and higher resting and exercise filling pressures compared with subjects without obesity. Importantly, causality between obesity and HFpEF has now been evidenced by randomized trials where weight loss has demonstrated meaningful improvements in functional status and quality of life, wherein the magnitude of weight loss was associated with the degree of clinical improvement.^24,25 We have demonstrated that exercise haemodynamics in adults post-Fontan share several similarities with HFpEF.⁹ It would then be intuitive to expect obesity to significantly impair the delicate balance of a Fontan circuit, but data on the haemodynamic impact of weight gain in this population are very limited. This represents an important knowledge gap given the prevalence of overweight and obesity among adults post-Fontan now being similar to the general population⁴ and the fact that long-term complications of the Fontan are intimately related to underlying haemodynamics.

Yogeswaran et al.²⁶ reported higher resting Fontan and ventricular filling pressures in Fontan patients with overweight or obesity compared with those with normal BMI. Our current findings agree with their observations. Alarmingly, a PA pressure ≥14 mmHg, a degree of elevation in Fontan pressures that has been associated with increased mortality in adults,¹⁰ was present in >80% of our subgroup with obesity. Increased plasma volume and metabolic demands secondary to obesity can contribute to chronically elevated cardiac output in patients with obesity and biventricular circulation. Given the inherent relationships between flow, chamber volume, and operative compliance, this increase in cardiac output predisposes patients with obesity and even normal heart function to elevated filling pressures and high-output heart failure.²⁷ With underlying myocardial disease and abnormal diastolic function, obesity results in an additional haemodynamic insult. Our data suggest a similar phenomenon in patients with Fontan palliation and obesity, where excess body mass appears to contribute to a rise in plasma volume and increased stroke volume/cardiac output with higher filling pressures during rest and exercise compared with individuals without obesity.

Classically, Q_s in Fontan physiology was felt to be well below the normal range for biventricular circulation. However, this premise is based on early observations derived from patients following atriopulmonary Fontan procedure. Contemporary data, including subsequent modifications of the procedure, suggest the expected resting cardiac index post-Fontan to be ∼2.5 L/min/m²²⁸ (low-normal for a biventricular circulation). In contrast, the median cardiac index among individuals with obesity in our cohort was 3.3 L/min/m². Ohuchi et al.²⁹ reported a high cardiac index (defined in their study as >3 L/min/m²) to be associated with increased mortality post-Fontan. Those with high cardiac index and elevated central venous pressure had the highest 5-year mortality of all groups analysed. We have reported similar findings, albeit using different cut-offs.³⁰ Various mechanisms have been implicated in the higher-than-expected Q_s seen in some Fontan patients and the associated adverse outcomes, including worsening right-to-left shunt and underlying cirrhosis. Our results suggest that obesity should also be incorporated into the potential mechanisms of these haemodynamic profiles.

We have reported similar SVR values between adults post-Fontan with normal haemodynamics and controls.³¹ In contrast, Fontan patients with unfavourable haemodynamic profiles and systemic venous hypertension have been found to have low resting SVR.^30,32 Fontan-associated liver disease and the frequent co-existent worsening cyanosis have been implicated in reductions in SVR manifested by these individuals. Obesity has also been associated with reduced vascular tone, possibly related to the effects of adipokines.³³ Accordingly, patients with obesity and HFpEF have been found to have lower SVR compared with those with lower BMI values.²³ The current findings align with those observations, as Fontan patients with obesity demonstrated lower resting SVR than the group with overweight or normal BMI despite having similar arterial pressure. Noteworthy, there was no difference in serologic liver scores and spleen size between groups with and without obesity, refuting more advanced liver disease in the former. Collectively, these observations suggest that obesity is another contributor to the low SVR values observed in some adults post-Fontan.

Peak VO₂ indexed to body mass was lower in the group with obesity, indicating worse functional capacity compared with those with lower BMI values. This finding agrees with data from patients with biventricular circulation.^23,34 Importantly, reduced peak VO₂ on CPET has been associated with an increased risk of cardiovascular outcomes in patients with congenital heart disease³⁵ and those post-Fontan palliation.³⁶ In a prior publication, we reported the association between abnormal Fontan and/or ventricular filling pressures during exercise catheterization and reduced % predicted VO₂ on CPET.¹³ Data from the Pediatric Heart Network Fontan 3 cohort reported that the highest rates of overweight and obesity were seen among patients (mean age 23 years) with the lowest peak VO₂ tertile.⁵ Although the pathophysiology of exercise limitation post-Fontan is complex and invariably multifactorial, our observations provide a pathophysiologic underpinning for their group’s observations.

The lower resting and exercise arterial O₂ saturations at cardiac catheterization among adults post-Fontan with obesity also deserve highlighting. Mild systemic arterial desaturation is expected in most Fontan patients, given coronary sinus drainage into the functional left atrium. Several mechanisms have been implicated in worsening arterial desaturation post-palliation, including abnormal pulmonary mechanics, presence of Fontan fenestration, pulmonary arteriovenous malformation/fistulae, and veno-venous collaterals. In our cohort, the prevalence of patent fenestration did not differ between groups. Although data on the occurrence of pulmonary arteriovenous malformation were unavailable, the presence of veno-venous collaterals was independently associated with obesity, potentially explaining the lower O₂ saturation in these individuals. Lastly, the ventilatory efficiency and peak ventilation data on non-invasive CPET would argue against significant differences in underlying pulmonary function between groups. However, the lack of between-group differences in pulse oximetry values on upright exercise suggests that worsening ventilation–perfusion mismatch while supine (due to decreased ventilation/atelectasis) and/or increased right-to-left shunt might also contribute to worsening arterial desaturation among individuals with obesity.

Future directions

Weight loss has been shown to decrease cardiac output and filling pressures and promote favourable cardiac remodelling in patients with obesity and biventricular circulation^37,38 with haemodynamic benefits of weight loss during rest and exercise seen even among individuals with obesity and no heart failure. However, the benefits of weight loss in patients post-Fontan are unclear and deserve further investigation. Similarly, the current findings support the need for studies regarding the use of sodium-glucose cotransporter-2 (SGLT2) inhibitors and GLP-1 receptor agonists in patients post-Fontan with obesity. Indeed, SGLT2 inhibitors have been shown to lower cardiac filling pressures in patients with obesity-related HFpEF, and the magnitude of improvement was found to be related to the degree of weight loss.³⁹ Lastly, our observations underscore the need for recognizing and managing obesity in childhood,⁴⁰ as obesity has been shown to impact functional capacity among those with congenital heart disease even prior to adulthood.⁴¹

N-terminal pro-brain natriuretic peptide levels <300 pg/dL have been associated with less adverse haemodynamic profiles in adults post-Fontan.¹³ Like patients with HFpEF²³ and obesity, despite the higher Fontan and ventricular filling pressures, patients with obesity in our cohort had numerically lower NT-pro-BNP levels compared with the rest of the cohort, though this was not statistically significant. Accordingly, providers should not be falsely reassured by the absence of elevated natriuretic peptide levels in symptomatic Fontan patients with obesity. Given the wide clinical use, a better understanding of the interpretation of natriuretic peptide levels in those post-Fontan is critically needed.

Limitations

We acknowledge the study’s limitations, including its retrospective design and sample size, which might have resulted in type II error. The pitfalls in using BMI as a marker of obesity compared with direct measures of body composition (particularly in patients post-Fontan⁴²) deserve highlighting. Given the anatomical complexity, ventricular volume and mass by echocardiography between groups could not be compared. Due to the low yield in our practice, agitated saline injection at the time of catheterization is no longer routinely performed. Thus, the prevalence of pulmonary arteriovenous malformation could not be ascertained. The assessment of veno-venous collaterals was performed at the discretion of the operator. Co-existent ventilation–perfusion mismatch might have led to an underestimation of Q_p in some individuals. Finally, the reason for the higher prevalence of pacemakers among patients without obesity is unclear, but we do not expect this to have significantly affected our results, given the lack of differences in chronotropic response between groups.

Conclusions

Adults post-Fontan with obesity have increased cardiac output, higher filling pressures, decreased SVR, and arterial O₂ saturation compared with individuals with overweight or normal BMI, both at rest and exercise. Moreover, obesity was associated with a decrease in peak VO₂ in both the upright and supine positions compared with Fontan patients without obesity. These findings share several similarities with the obesity phenotype of HFpEF. Further studies are warranted to determine the role and impact of weight loss on haemodynamics and clinical outcomes among obese adults post-Fontan.

Author contribution

D.N.O. and W.R.M. drafted the manuscript. C.C.J., A.C.E., B.A.B., Y.V.R., H.M.C., K.M.L.-B., and R.C. critically revised the manuscript. W.R.M. was also responsible for data acquisition, analysis, and interpretation. All gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy.

Funding

A.C.E. was supported by the National Heart, Lung, and Blood Institute (NHLBI) grants K23 HL141448, R01 HL158517, and R01 160761. B.A.B. was supported by the NHLBI grants R01 HL128526, R01 HL162828, and U01 HL160226 and the US Department of Defense grant W81XWH2210245. Y.V.R. was supported by the National Institute of Health grant K23HL164901.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.