2026 ESC Guidelines on Cardiac Rehabilitation

Authors, Journal, Affiliations, Type, DOI

• Authors: Bäck M, Wilhelm M (co-chairs, equal contribution); Marcin T, Machado FVC, Sandberg A (co-ordinators); and the ESC Guidelines Advisory Group
• Journal: European Heart Journal 2026;47:5631–5729
• Affiliations: ESC Task Force on cardiac rehabilitation, with ACNAP, ACVC, EACVI, EAPC, EAPCI, EHRA and the HFA; endorsed by the European Society of Physical and Rehabilitation Medicine (ESPRM) and the World Organization of Family Doctors (WONCA)
• Type: Clinical practice guideline — the first ESC guideline devoted solely to cardiac rehabilitation
• DOI: 10.1093/eurheartj/ehag099
• Grading: 30 recommendation tables; uses the new ESC 2026 levels-of-evidence framework for the first time; two ESC Patient Forum representatives sat on the Task Force with voting rights

Overview

The first dedicated ESC cardiac rehabilitation guideline redefines cardiac rehabilitation as an unqualified multimodal, multidisciplinary, patient-centred intervention delivered across Phases 0–III, and states that the term should no longer carry the qualifiers "exercise-based" or "comprehensive". It issues 30 recommendation tables across some 15 cardiac populations — ACS, CCS, HFrEF and HFpEF, LVAD and transplant, valve disease and TAVI, adult congenital heart disease, AF, CIEDs, cardiomyopathies, cancer, frailty and comorbidity — plus quality standards, twelve core components, telehealth, delivery models and participation. Four positions distinguish it: dose–response with no ceiling; a minimum standard of 24 sessions over 12 weeks; ventilatory-threshold-based exercise intensity in place of %HRpeak, %HRR or %VO2peak; and explicit endorsement of cardiac telerehabilitation and hybrid models as alternatives to centre-based rehabilitation. It is also unusually candid about the weakness of the mortality evidence and about the European capacity gap of one place per seven patients in need.

Keywords

Cardiac rehabilitation; prevention; patient-centred care; shared decision-making; risk factor management; exercise training; mental functioning; physical functioning; quality of life; healthy ageing; telemedicine

Key Takeaways

Definition, Phases and Scope

• Definition: cardiac rehabilitation is a multimodal, multidisciplinary, patient-centred intervention — not "exercise-based" or "comprehensive" rehabilitation. The Task Force's position is to address the misconceptions rather than rename the field.
• Four phases: prehabilitation (Phase 0), acute in-hospital (Phase I), centre- or home-based post-acute (Phase II), and long-term maintenance (Phase III). Supervised programmes are typically Phase II.
• European provision is fragmented — ~40% of programmes lack social-security funding, so patients pay partly or fully, or cannot attend.
• Out of scope: stroke, aortic aneurysm and peripheral artery disease, deferred to the 2024 ESC PAD/aortic guideline and the 2024 ESVS asymptomatic lower-limb PAD guideline.

Effectiveness, Cost and Safety

• Functioning and health-related quality of life are elevated to independent primary outcomes, alongside mortality and morbidity. The healthspan–lifespan gap widened from 8.5 to 9.6 years across 183 WHO Member States, with non-communicable disease the principal driver.
• The ICF framework may be considered (IIa C) to structure goal setting and outcome assessment.
• Adding cardiac rehabilitation to usual care is consistently cost-effective and often cost-saving; the economic signal is weaker for individual non-exercise components.
• Safety: exercise-related serious adverse events occur at ~1 per 53,770 patient-hours in remotely supervised sessions, ~1 per 11,333 patient-hours in centre-based HIIT, and have not been reported for centre-based moderate-intensity continuous training.
• Contraindications (Table 4): ACS within 2 days, untreated life-threatening arrhythmia, decompensated HF, uncontrolled hypertension, advanced AV block, acute myocarditis or pericarditis, symptomatic aortic stenosis, HCM with severe obstruction, acute systemic illness, intracardiac thrombus. An "increased risk" tier adds >1.8 kg weight gain over 1–3 days, NYHA IV, supine HR >100 bpm, complex ventricular arrhythmia and critical coronary stenosis.

Timing of Initiation

• Early initiation: I C; in patients hospitalised for decompensated HF, IIa B2; telemetry during exercise in selected high-risk patients, IIb C.
• After ACS begin ideally within 14 days and no later than 30 days. After CABG within 28 days and never beyond 42 days. After decompensated HF, once decongested, haemodynamically stable, discharge feasible and GDMT started.
• Non-exercise components — nutrition, psychological support, smoking cessation, pharmacotherapy optimisation — can begin immediately even when exercise is deferred.

Indications by Population

• ACS and CCS: a 2023 meta-analysis (23,172 patients) found reductions in cardiovascular mortality, MI and all-cause hospitalisation and improvement in the mental component of HRQoL, but no reduction in revascularisation or all-cause mortality, interpreted as reflecting modern background therapy. Benefit also extends to MINOCA.
• Heart failure: trials mostly enrolled LVEF <40%, while the 2026 ESC HF guideline defines HFrEF as <50%. Meta-analyses show no all-cause mortality reduction, though large registries suggest survival benefit. Early rehabilitation after decompensation is linked to lower in-hospital mortality and readmission, and a REHAB-HF secondary analysis showed prevention of sarcopenia and frailty.
• Valve disease: a 2021 meta-analysis of 10 small RCTs (580 patients) found the impact on mortality, hospitalisation and HRQoL uncertain; the Class I in aortic stenosis after SAVR/TAVI rests on functional endpoints.
• Cardiomyopathy: supervised moderate and high-intensity programmes in HCM are safe with improved fitness and no increase in arrhythmia, but benefits are lost several months after stopping. Moderate to high intensity is discouraged in arrhythmogenic cardiomyopathy. There are no data in laminopathies, metabolic or restrictive cardiomyopathy, or amyloidosis.
• Cancer: the CORE model — structured exercise, nutritional counselling, psychological support and cardiovascular risk assessment — with oncologists added to the team. Benefits persist months after the intervention and extend to disease-free and overall survival; cardioprotection against anthracycline or anti-HER2 dysfunction is not established.
• Frailty: cardiac rehabilitation must not be withheld because of frailty or comorbidity — an explicit Class III "do not do" recommendation.

Quality Standards

• Dose–response: higher session dose associates with lower mortality and morbidity with no clear plateau, so programme completion and session adherence are the key quality indicators.
• Phase II programmes should be managed according to EAPC structure- and process-based metrics, quality indicators and outcome assessment (I C).
• Team leadership may fall to any qualified member, but overall medical responsibility is ideally a board-certified cardiologist with cardiac rehabilitation experience, or a physical and rehabilitation medicine physician with formal cardiology linkage. Annual evaluation of delivery and outcomes is required.

Core Components

Assessment. Patient-led baseline assessment covering history, examination, diagnostic tests and patient-reported outcome measures, capturing culture, education, ethnicity and social determinants, female-specific risk factors (gestational diabetes, hypertensive disorders of pregnancy, premature menopause, PCOS, endometriosis) and sexual dysfunction. Tests include resting 12-lead ECG, imaging, lipids including Lp(a) if not previously measured, fasting glucose, HbA1c, uric acid, renal function and urinalysis. Cardiopulmonary exercise testing is the gold standard for risk stratification and prescription but is underused; 6MWT is the alternative, 1RM testing guides resistance prescription, and an arm ergometer serves lower-limb disability. Frailty tools: Edmonton Frail Scale, 7-point Clinical Frailty Scale, SPPB.

Self-management and education. Multi-behaviour interventions outperform single-behaviour ones. Effective components from an umbrella review of 865,000 people are a credible source, graded tasks, human coaching and personalisation, plus goals and planning, feedback and monitoring, social support and digital prompts. In-hospital education improved knowledge, QoL, readmissions and modestly LDL-C, blood pressure, attendance and adherence; standalone education does not reduce mortality. Programmes of ≥3 months work better. Education reduced clinical depression by 35% and anxiety by 60%. 40%–80% of health information is forgotten immediately, which is the case for teach-back. Recommendations: self-management with behaviour-change techniques I B1; goal-specific education I B1; health-literacy interventions I B2; psycho-education in ICD IIb B2.

Pharmacotherapy optimisation. EUROASPIRE V showed most post-ACS patients miss blood pressure, LDL-C and glucose targets, with only 31% of high or very-high-risk patients reaching 2019 ESC/EAS LDL-C goals. Rehabilitation is the framework for up-titration, down-titration, stopping, new prescription, reconciliation and deprescribing (I C), with STRONG-HF supporting rapid up-titration after decompensated HF. A sex disparity is documented across 175,187 STEMI admissions, and female sex independently predicts failure to reach LDL-C <1.8 mmol/L.

Physical activity and sedentary behaviour. Sedentary behaviour is waking behaviour at ≤1.5 METs while sitting, reclining or lying. WHO 2020 targets are 150–300 min/week of moderate or 75–150 min/week of vigorous aerobic activity, muscle strengthening on ≥2 days/week and multicomponent balance or strength work on ≥3 days/week. An extra 500–1000 steps/day associates with lower all-cause mortality and cardiovascular events, short-term goals of +5–10 min per session are practical, and sedentary time is an independent risk even in physically active people. Recommendations: activity trackers I B1; active lifestyle in CAD I B2; counselling aligned with WHO I C; reducing sedentary time IIa B2.

Exercise training. Session composition is two-thirds endurance, one-third resistance, with a median programme duration of 3–4 months.

• Intensity anchors: moderate intensity corresponds to 45%–85% VO2peak, 50%–70% Wpeak, 55%–90% HRpeak, 40%–85% HRR, or Borg 12–17/20. Low-capacity or recently decompensated patients start at 40% VO2peak and progress to VT1 (50%–60%) then VT2 (65%–90%).
• Ventilatory thresholds are preferred over %HRpeak, %HRR or %VO2peak, and MET-based intensity classification is discouraged because MET 4–6 may be supramaximal for many cardiac patients. Ventilatory thresholds can be estimated from resting and peak heart rate when gas exchange is unavailable.
• Resistance: %1RM is encouraged but 10RM may be safer in cardiac patients to avoid Valsalva; OMNI-RES and the talk test give subjective calibration; reassess regularly, especially with chronotropic incompetence.
• Risk-factor-specific dosing: obesity 250–420 min/week to exceed 2000 kcal/week; type 2 diabetes ≥5 days/week and ≥21 resistance sets/week; hypertension >30 min and ≥8 sets per session, aerobic first-line with or without isometric work; strength at 70%–79% 1RM (6–8 reps) gives the largest effect; in type 1 diabetes the HbA1c minimal clinically important difference is reached at 160 MET-min/week with a plateau near 1500 MET-min/week.
• HIIT is superior or more time-efficient for VO2peak in CCS, chronic HF and post-transplant, but protocol uncertainty and absent unsupervised safety data confine it to supervised, medically cleared patients (IIb B1). High-intensity resistance training shows no excess adverse events versus lower intensity, with caution in uncontrolled or very high blood pressure.
• Recommendations: tailored prescription for all patients IIa B2; exercise training in ACS/CCS I A for CV mortality, MI and hospitalisation and I B1 for mental functioning, HRQoL, VO2peak and lower costs; resistance training added I B1; early mobilisation after cardiothoracic surgery, possibly within 4 weeks, I B1; early exercise after MI for LV remodelling IIa B2; HIIT IIb B1; stable chronic HF, transplant and VAD I B1 each.

Nutrition and body composition. Mediterranean and low-fat diets reduce cardiovascular events and all-cause mortality (I B1); comprehensive weight-loss intervention I B1. Alcohol is capped at 100 g pure alcohol/week for both sexes, against an observational J-shaped curve with a nadir at 10–40 g/day, a wine-specific mortality signal, a linear rise in blood pressure with no safe threshold, and IARC Group 1 carcinogen status. Malnutrition affects ~60%–69% of chronic HF patients, and 47% of Phase II patients are malnourished or at risk. BMI is specific but insensitive, missing excess adiposity in ~half of affected people. More than 80% of CAD patients are overweight or obese, yet weight loss during rehabilitation is typically only 2%–3%. GLP-1 receptor agonists produce 15%–24% weight loss with MACE −13% and all-cause mortality −12%, but 50% discontinue within a year and muscle accounts for 25%–39% of weight lost, so protein-rich diet and resistance training are required — the drug is IIa A while muscle preservation is only I C.

Smoking cessation. Passive smoking confers a 1.3-fold CAD risk; smoking confers 25% higher CAD risk in women than men, and anti-smoking legislation cut acute MI by 20% within 12 months. Former heavy smokers who quit within 5 years reach HR 0.61, and quitting reduces the relative risk of death in CAD by 36%. Behavioural support should run at least one month and be combined with pharmacotherapy. Recommendations: cessation as an integral component I A; pharmacotherapy (NRT, varenicline, bupropion, cytisine) I A; structured behavioural counselling I B1; e-cigarettes as a quit aid IIb C.

Psychosocial management. Depression affects 31.3%, anxiety 32.9% and stress 57.7% of cardiac populations; heart disease confers a 1.79-fold depression risk and depression a 1.63-fold heart disease risk, and distressed patients are less adherent. A large trial found CBT reduced recurrent CVD by 41% and recurrent MI by 45%. Recommendations: psychological interventions in CAD or HF I B1; CBT in CAD, HF or ICD I B1; psychological screening I C; mindfulness IIa B2; SSRI for moderate-to-severe depression in CHD IIa B2.

Sexual counselling. Sexual disorders affect 62.6% of patients with CVD; erectile dysfunction predicts MI (1.55-fold), CVD (1.45) and all-cause mortality (1.25), and women with CVD have 1.51-fold sexual dysfunction rising to 2.07-fold after MI. Physical activity for erectile dysfunction is 160 min/week for 6 months, or 40 min of moderate-to-vigorous aerobic exercise four times weekly. Screening sexual function in all men and women with CVD is I B2; sexual rehabilitation IIa B2.

Frailty, comorbidity and environment. Harmonise rehabilitation with disease-specific programmes for COPD, CKD, rheumatoid arthritis, stroke, PAD and cancer, with more frequent monitoring in COPD, CKD, musculoskeletal disease and cancer (I C). Environmental risk factors are a new component — air pollution, noise, artificial light at night, extreme heat, desert dust and wildfires — with counselling on mitigation IIa C.

Return to work. Comprehensive rehabilitation is associated with return to work in 66% (60%–71%) versus 58% (47%–68%) of non-participants, with combined counselling and exercise training IIa B2. Return rates are lower in physically demanding jobs, and return to work is usually a secondary outcome with poor reporting.

Telehealth and Digital Delivery

• Cardiac telerehabilitation programmes used a mean of 2.6 delivery modes. A meta-analysis of 14 RCTs in 2,869 CHD patients found 6MWT +25.58 m, daily steps +1.05 k, exercise habits OR 2.28, depression SMD −0.16 and SF-12 mental and physical improvements. An HF telemedicine meta-analysis (34 studies, 13,269 patients) found CV mortality RR 0.83 and CV hospitalisation RR 0.71, mostly over short follow-up.
• Barriers include unvalidated devices, digital literacy and disparities, privacy, absent infrastructure, training gaps, no reimbursement tariff, medico-legal uncertainty and national regulatory variation. Cuffless blood pressure monitors lack accepted validation protocols, and the EU AI Act (Regulation 2024/1689) governs transparency, oversight and risk assessment.
• Recommendations: telehealth consultation plus telemonitoring in HF I B1; telehealth consultation to increase enrolment, adherence or completion IIa B2; self-operated wearables and self-monitoring apps IIa B1; AI for event detection or clinical decision-making IIb C; implantable pulmonary artery pressure monitoring IIb C.

Delivery Models

• Centre-based, home-based with or without digital support, cardiac telerehabilitation as full replacement (assessment and pre-exercise screening remain in person), and hybrid models.
• Minimum standard: 24 sessions over 12 weeks, two per week, regardless of mode.
• The evidence conflicts on mode. A Cochrane meta-regression found consistent benefit irrespective of delivery mode, and a full telerehabilitation versus hybrid review (28 RCTs, 2,662 patients) found full telerehabilitation slightly superior for exercise capacity with adherence 87% and severe adverse events at 1 per 23,823 patient-hours. But a network meta-analysis (36 RCTs, 3,534 CAD patients) found centre-based rehabilitation improved HRQoL and reduced depression and anxiety with exercise effectiveness considerably reduced when home-delivered, and an HF network meta-analysis (139 RCTs, 18,670 patients) found that only centre-based rehabilitation significantly reduced HF hospitalisation and mortality, by ~60%.
• Recommendations: centre-based I B1; inpatient rehabilitation in high-risk or unstable patients I C; telerehabilitation or hybrid as an alternative in ACS/CCS after PCI or CABG IIa B1, in HF IIa B1, in ACHD with impaired fitness IIa B2, in AF IIa B2 and in cancer IIa B2; home-based without digital support in non-high-risk patients IIb B1.

Participation and Equity

• Participation is 9%–50% of eligible patients internationally; referral and attendance range 4.5%–71.1%, and adherence 4%–62%.
• Capacity: rehabilitation is available in 91% of European countries but at one place per seven patients in need, requiring ~3 million additional places per year; globally only 1 in 18 patients has access. UK modelling suggests timely access would raise enrolment by 14.3% and completion by 1.9%.
• Automatic referral plus clinician liaison is the single most effective enrolment intervention — in 2,635 CAD inpatients it achieved 85.8% referral and 73.5% enrolment versus 32.2% and 29.0% with usual referral (OR 8.41). Automatic referral alone gave OR 3.27 and liaison alone OR 3.35. Invitation to an initial assessment gave OR 4.22 in men and 4.03 in women.
• Adherence: remote-delivered interventions are most effective, app users are 1.4× more adherent, and financial incentives nearly doubled completion in low-SES patients.
• Under-represented groups are women, ethnic minorities, older adults, low-SES patients and those with multimorbidity. Women are referred less often — especially after Takotsubo — are older, more often widowed and more functionally limited, and face caregiving, transport and financial barriers. In the USA Black patients participate less because of lower referral, persisting after socioeconomic adjustment. Older adults are under-represented despite higher disease prevalence, while younger patients perceive rehabilitation as "for older people".
• Recommendations: physician or health-professional endorsement I B2; automatic referral plus individual discussion I B2; invitation to initial assessment I B2; increasing programme availability I C; tailoring to women IIa B1; telerehabilitation or hybrid to improve enrolment and adherence IIa B1; person-centred theory-based behavioural interventions IIb B2.

Limitations of the document

• Scope excludes stroke and PAD/aortic disease, so rehabilitation for those populations is not covered despite shared infrastructure.
• LOE C dominates several recommendation tables, and many recommendations are IIb C or "may be considered".
• Blinding is impossible and trials are underpowered for mortality; the Task Force concedes this and downgrades the level of evidence accordingly, while still assigning Class I A to some mortality and MI endpoints.
• Evidence is predominantly European and North American, and trial populations under-represent every group the equity sections discuss — women, the very old, ethnic minorities, low-SES and multimorbid patients, and rare conditions such as SCAD and INOCA.
• Cardiomyopathy evidence is scarce or absent for restrictive and amyloid disease, laminopathy and metabolic cardiomyopathy; the pulmonary hypertension evidence is Group 1 only.
• Nearly all exercise dosing derives from meta-analyses in ACS/CCS and HF, and the dosing table is explicitly a reference frame rather than indication-specific prescription.
• Quality-standard detail, evidence tables and many definitions sit in supplementary data, not the print text.
• Conflict-of-interest declarations are in a separate document, and two patient representatives are a small qualitative input whose sections are narrative rather than evidence-graded.

Key Concepts Mentioned

• Comprehensive versus exercise-based cardiac rehabilitation; Phases 0–III
• Dose–response without a ceiling; session dose, completion and adherence as quality indicators
• The 24-session / 12-week minimum standard
• FITT-VP exercise prescription; ventilatory thresholds; 1RM and 10RM; HIIT and high-intensity resistance training; inspiratory muscle training
• Cardiopulmonary exercise testing, 6MWT, SPPB and the ICF framework
• COM-B and the Behaviour Change Wheel; motivational interviewing; OARS; teach-back; health literacy
• Cardiac telerehabilitation, hybrid models, mHealth, wearables and clinical decision support
• Malnutrition, sarcopenia, sarcopenic obesity and GLP-1 receptor agonist-associated muscle loss
• Psychosocial distress, CBT, mindfulness and SSRIs
• Frailty, multimorbidity, prehabilitation and return to work
• Equity by sex, ethnicity, age, socioeconomic status and digital access
• The exposome and environmental cardiovascular risk

Key Entities Mentioned

• entities/Heart-Failure · entities/HFrEF · entities/HFpEF
• entities/Acute-Coronary-Syndrome · entities/Chronic-Coronary-Disease
• entities/Atrial-Fibrillation · entities/ICD · entities/CRT
• concepts/Aortic-Stenosis · concepts/Valvular-Heart-Disease
• concepts/Pulmonary-Hypertension · concepts/Cardiac-Rehabilitation
• concepts/GLP-1-Receptor-Agonists · concepts/Cardio-Oncology
• entities/Frailty-in-Cardiovascular-Disease
• Trials cited: REHAB-HF, STRONG-HF, SELECT, REBOOT, MONITOR-HF, CopenHeartRFA, EUROASPIRE V
• Devices: CardioMEMS, ICD, CRT-D/P, ICM, LVAD

Wiki Pages Updated

• wiki/sources/CRP-ESC-2026.md — created
• wiki/concepts/Cardiac-Rehabilitation.md — rewritten around the ESC 2026 definition, phases, indications table, 24-session standard, ventilatory-threshold prescription, delivery-mode evidence, European capacity figures and new contradictions
• wiki/sourceindex.md — new row
• wiki/wikiindex.md — new row and rewritten Cardiac-Rehabilitation row

Still to propagate: the ACS/CCS (I A), HF (I B1), AF (IIa B1), post-TAVI aortic stenosis (I B1), PAH (IIa B1) and post-ICD/CRT (I B1) recommendations belong on entities/Acute-Coronary-Syndrome, entities/Chronic-Coronary-Disease, entities/Heart-Failure, entities/Atrial-Fibrillation, concepts/Aortic-Stenosis, concepts/Pulmonary-Hypertension, entities/ICD and entities/CRT. They are recorded here and on the rehabilitation page but have not yet been written into those pages.