각종 난치병 환자 '운동치료' 효과.. 란셋, jama, nejm 논문!!

[문형철]

노화(aging)는

분자·세포 수준의 손상이 쌓여 조직과 장기 기능이 저하되는

복잡한 생물학적 과정입니다.

노화는

인지 기능 저하를 동반하며,

알츠하이머병(가장 흔한 치매 형태)의 가장 큰 위험인자입니다.

주요 메시지:

• 앉아서 생활(sedentary lifestyle)과 불건강한 생활습관은 뇌 노화를 가속화합니다.
• 반대로 규칙적인 신체 활동(physical activity), 특히 높은 심폐지구력(Cardiorespiratory Fitness, CRF) 또는 이 둘의 조합은 인지 저하를 완화하고 치매 위험을 줄일 수 있습니다.

이 리뷰의 목적은:

1. 지구력 운동(endurance exercise)의 신경보호(neuroprotective) 기전을 탐구
2. CRF(심폐지구력)가 건강한 뇌 노화에 얼마나 중요한지 강조
3. 공중보건 차원에서 CRF를 높이는 지구력 운동을 치매 예방 전략으로 제안

주요 발견 (Key Findings)

리뷰에서 강조하는 운동의 신경보호 효과는 주로 CRF 향상을 통해 매개됩니다.

핵심 기전은 다음과 같습니다:

• 뇌 혈류 개선 (Improved cerebral blood flow): 운동으로 뇌로 가는 혈액 공급이 증가해 영양·산소 공급이 좋아짐
• 염증 감소 (Reduced inflammation): 전신 및 뇌 내 만성 염증이 줄어듦
• 신경가소성 향상 (Enhanced neuroplasticity): 새로운 신경 연결 형성, 시냅스 가소성 증가
• 기타: 산화 스트레스 감소, 신경영양인자(예: BDNF) 증가, 미토콘드리아 기능 개선, 성인 신경발생(neurogenesis) 촉진 등

이러한 기전들이 종합적으로 작용해

뇌 노화 속도를 늦추고,

인지 기능을 보호하며,

알츠하이머병·치매 위험을 낮춥니다.

일부 연구에서는

강도 높은 단기 활동으로

치매 위험이 최대 40%까지 감소할 수 있다는 결과도 언급됨

https://www.thelancet.com/journals/lanrhe/article/piis2665-9913(22)00156-4/fulltext

염증성 류마티스 질환
(류마티스 관절염, 강직성 척추염, 결합조직질환 등) 환자들의
가장 흔하고 힘든 증상 중 하나가 만성 피로(fatigue)입니다.

기존 치료(약물)로도 잘 조절되지 않고,
삶의 질을 크게 떨어뜨리며
일상·업무에 지장을 줍니다.

이 연구는
대면이 아닌 원격(전화)으로 제공하는 두 가지 중재
— 인지행동치료(Cognitive Behavioural Approach, CBA)와
개인화된 운동 프로그램(Personalised Exercise Programme, PEP) — 이
피로를 효과적으로 줄일 수 있는지를 확인했습니다.

특히
코로나19 팬데믹 기간에
대면 치료가 어려운 상황을 고려한 실용적 연구입니다.

연구 방법 (LIFT 시험)

• 연구 디자인: 다기관, 무작위배정, 대조군, 개방형(open-label), 평행군 연구
• 대상자: 367명 (평균 연령 57.5세, 여성 75%)
  • 안정된 염증성 류마티스 질환 환자 (류마티스 관절염 55%, 축성 척추관절염 20%, 결합조직질환 21% 등)
  • 지속적이고 임상적으로 의미 있는 피로 (0~10점 중 ≥6점, 3개월 이상)
• 무작위 배정 (1:1:1):
  1. 대조군 (Usual care): 피로 자가관리 교육 책자만 제공
  2. PEP 군 (Personalised Exercise Programme): 전화로 14주 동안 최대 7회 세션 + 22주 부스터. 환자 상태에 맞춘 개인화된 운동 (점진적 강도 증가, graded exposure)
  3. CBA 군 (Cognitive Behavioural Approach): 전화로 최대 7회 세션 + 부스터. 피로에 대한 비합리적 생각·행동 교정
• 중재 제공자: 류마티스 전문 보건의료인(간호사, 물리치료사 등)이 훈련받아 전화로 진행
• 주요 평가 시점: 56주 (약 1년 후) — 효과의 지속성을 확인

주요 결과 (56주 시점)

두 중재 모두 대조군보다 피로를 유의하게 개선했습니다.

• 피로 중증도 (Chalder Fatigue Scale, 0~33점):
  • PEP: –3.03점 (97.5% CI –5.05 ~ –1.02, p=0.0007)
  • CBA: –2.36점 (97.5% CI –4.28 ~ –0.44, p=0.0058)
• 피로가 삶에 미치는 영향 (Fatigue Severity Scale, 1~9점):
  • PEP: –0.64점 (95% CI –0.95 ~ –0.33, p<0.0001)
  • CBA: –0.58점 (95% CI –0.87 ~ –0.28, p<0.0001)

효과
크기(effect size)는 중간 정도(~0.5~0.6)로 임상적으로도 의미 있는 수준입니다.
효과는 56주까지 유지되었습니다.

기타 긍정 효과:

• 다차원 피로 개선
• 정신 건강 관련 삶의 질 향상
• 수면 장애 감소
• PEP 군에서는 우울 증상 감소, 업무 장애 감소, 가치 있는 일상 활동 개선까지 추가 효과

결론 

“전화로 제공된 인지행동치료(CBA)와 개인화된 운동 프로그램(PEP)은
다양한 염증성 류마티스 질환에서
피로의 중증도와 삶에 미치는 영향을 통계적·임상적으로 유의하게 줄였으며,
그 효과가 1년까지 유지되었다

https://www.nejm.org/doi/full/10.1056/NEJMoa2026141

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure | NEJM

연구 배경

급성 심부전(특히 decompensated heart failure)으로

입원하는 고령 환자들은

대부분 신체 기능이 매우 약해져 있습니다.

입원 기간 동안 침상 안정으로 인해

근력·균형·이동 능력이 더 떨어지고,

퇴원 후에도 회복이 느리며 재입원율이 높습니다.

기존 심부전 치료 지침은

약물·장치 중심이었고,

입원 직후 신체 재활에 대한 대규모 연구가 부족했습니다.

이 연구는

“고령 급성 심부전 환자에게 입원 중 또는 퇴원 직후부터 시작하는

맞춤형·점진적 신체 재활 프로그램이 신체 기능을 개선하고 재입원을 줄일 수 있는가?”를 확인하기 위해

진행되었습니다.

연구 방법 (REHAB-HF 시험)

• 연구 디자인: 다기관, 무작위배정, 단일맹검, 대조군 연구 (349명 참여)
• 대상: 60세 이상, 급성 심부전으로 입원한 환자 (좌심실 박출률(EF)에 관계없이 포함, 즉 HFrEF와 HFpEF 모두)
• 중재군 (175명):
  • 입원 중 또는 퇴원 직후부터 시작
  • 4가지 영역을 모두 다루는 맞춤형 재활: 근력(strength), 균형(balance), 이동성(mobility), 지구력(endurance)
  • 병원 → 외래(주 3회, 60분씩, 총 12주) → 가정 운동으로 전환
  • 환자 상태에 따라 4단계로 점진적으로 난이도 조절
• 대조군 (174명): 기존 표준 치료 (필요 시 일반 물리치료 가능, 하지만 연구 특화 재활은 받지 않음)

주요 결과

• 주요 평가변수 (3개월 후): Short Physical Performance Battery (SPPB) 점수
• → 중재군: 8.3점 vs 대조군: 6.9점 (차이 1.5점, 95% CI 0.9~2.0, P<0.001)
• → 균형, 보행 속도, 의자에서 일어나기 모두 유의하게 개선
• 기타 긍정적 효과:
  • 6분 보행 거리: +34m 개선
  • 보행 속도: +0.12 m/s 개선
  • 허약(frailty) 정도 감소
  • 삶의 질(KCCQ 점수): +7.1점 개선
  • 우울 증상 감소
• 재입원과 사망률 (6개월):
  • 모든 원인 재입원율: 중재군 1.18 vs 대조군 1.28 (비율비 0.93, 통계적으로 유의미한 차이 없음)
  • 사망률: 중재군에서 약간 높았으나 유의미한 차이 없음 (21명 vs 16명)
• 안전성: 중재군에서 낙상 등 심각한 부작용 증가 없이 안전하게 진행됨.

결론 (논문에서 직접 인용)

“다양한 고령 급성 심부전 환자 집단에서,

조기·전환적·맞춤형·점진적 재활 중재(여러 신체 기능 영역 포함)는

기존 치료보다 신체 기능을 더 크게 개선시켰다.”

재입원이나 사망률은 유의하게 줄이지 못했지만,

신체 기능·삶의 질·허약 개선 효과는 뚜렷했습니다.

이는 고령 심부전 환자 관리에서

“운동/재활”의 중요성을 보여주는 근거가 됩니다.

임상적 의미

• 고령 심부전 환자에게 입원 직후부터 적극적인 재활을 고려할 수 있는 근거가 마련됨.
• 특히 HFpEF(박출률 보존 심부전) 환자에서도 효과가 확인되어 의미가 큽니다.
• 이후 경제성 분석에서도 비용 대비 삶의 질 향상 효과가 괜찮은 것으로 평가됨.

이 연구는 심부전 치료 패러다임에서 “약물뿐 아니라 기능 회복”을 강조한 대표적인 논문 중 하나

심부전 재활(심장재활, Cardiac Rehabilitation)에 대한
주요 국제 및 국내 가이드라인을 최신 기준(2026년 4월 현재)으로 정리

1. 주요 국제 가이드라인 요약

• 2022 AHA/ACC/HFSA Guideline for the Management of Heart Failure (미국 심장학회/미국심장협회/심부전학회, 아직도 가장 대표적인 기준)
  • 운동 훈련(regular physical activity 또는 exercise training): Class I, Level A → 심부전 환자(참여 가능한 경우)에게 기능 상태, 운동 능력, 삶의 질(QOL) 개선을 위해 강력 추천.
  • 심장재활 프로그램(Cardiac Rehabilitation): Class 2a, Level B-NR → 기능 용량, 운동 내성, 건강 관련 삶의 질 개선에 유용함.
  • 재활 내용: 운동 + 교육 + 상담(자기관리, 식이, 약물 순응 등) 포함.
  • HFrEF뿐만 아니라 HFpEF, HFmrEF에도 적용 가능.
  • 입원 후 조기 재활(퇴원 직후 시작) 강조.
• ESC (유럽심장학회) Heart Failure Guidelines (2021년 이후 업데이트된 내용 반영, 2025년 별도 HF 전용 대규모 업데이트는 아직 없음)
  • 운동 기반 심장재활을 강력 권고 (Class I).
  • 특히 고령자, 입원 환자, 허약(frailty) 환자에서 기능 회복과 재입원 감소 효과 강조.
• JCS/JHFS 2025 Guideline on Diagnosis and Treatment of Heart Failure (일본순환기학회/일본심부전학회, 2025년 발표)
  • 입원 중 및 외래 심장재활 적극 권고.
  • 운동 훈련으로 peak VO₂(최대 산소 섭취량) 향상 → 장기 심장 사건 감소.
  • 하이브리드(원격+대면) 재활, 가정 운동 프로그램도 언급.
• 기타 (2025~2026)
  • HFAI (Heart Failure Association of India) 2025 Guidelines: 운동 훈련과 심장재활을 재활 섹션에서 명확히 추천.
  • 많은 국가에서 입원 후 심부전 환자에게 심장재활 referral을 성과 지표(quality measure)로 사용 중.

2. 한국 상황 (대한심장학회·대한심부전학회 기준) 한국심부전학회(KHFS)와 관련 지침·교육 자료에서는 다음과 같이 권고

• 심부전 환자에서 심장재활의 목표:
  1. 심폐지구력 향상 → 삶의 질 개선
  2. 심부전 악화로 인한 재입원 및 사망률 감소
• 운동 기반 심장재활을 강력 추천 (국제 가이드라인과 일치).
• 특히 REHAB-HF 연구(이전 설명드린 NEJM 논문)처럼 입원 중/퇴원 직후 시작하는 맞춤형·점진적 재활(근력 + 균형 + 이동성 + 지구력)을 권장.
• 한국에서는 건강보험 급여 적용(심장재활 프로그램)이 확대되어, 병원급 이상에서 외래 심장재활을 받을 수 있습니다.

3. 심부전 재활 프로그램의 일반적인 구성 (가이드라인 공통)

1. 평가 단계: 운동 부하 검사(CPET 또는 6분 보행 검사), 심기능 평가, 허약/영양/우울증 screening.
2. 운동 처방:
   • 유산소 운동: 주 3~5회, 중강도 (Borg scale 12~14 또는 최대심박수의 60~80%)
   • 저항(근력) 운동: 주 2~3회, 가벼운 덤벨/밴드
   • 균형·유연성 운동 (고령자 필수)
   • 점진적 증가 (입원 중 침상 운동 → 병동 보행 → 외래 supervised exercise → 가정 운동)
     • 지속이 중요: 12주 프로그램 끝난 후에도 주 150분 이상 중강도 운동 유지 권고.
     
     
     
3. 교육 및 상담:
   • 저염식, 체중/증상 일지, 약물 순응, 금연, 백신 접종
   • 성생활, 일상생활 복귀 조언
4. 기간: 보통 8~12주 (총 36회 정도), 이후 유지 운동( lifelong)

4. 효과 (근거 기반)

• 운동 능력(6MWT, peak VO₂) 개선
• 삶의 질(KCCQ 점수) 향상
• 재입원율 감소 (특히 운동 기반 재활)
• 사망률 감소 경향
• 안전성: 제대로 감독된 프로그램에서는 부작용(낙상, 부정맥 등) 매우 낮음

5. 실제 적용 팁 (환자·보호자 관점)

• 퇴원 전에 주치의에게 “심장재활 referral” 요청하세요.
• 고령·허약 환자도 참여 가능 (맞춤형으로 강도 조절).
• 코로나 이후 하이브리드/원격 재활(텔레재활)도 많이 도입됨.
• 지속이 중요: 12주 프로그램 끝난 후에도 주 150분 이상 중강도 운동 유지 권고.

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Authors: Dalane W. Kitzman, M.D., David J. Whellan, M.D., M.H.S., Pamela Duncan, P.T., Ph.D., Amy M. Pastva, P.T., Ph.D., Robert J. Mentz, M.D., Gordon R. Reeves, M.D., M.P.T., M. Benjamin Nelson, M.S. https://orcid.org/0000-0002-1701-6071, +5 , and Christopher M. O’Connor, M.D.Author Info & Affiliations

Published May 16, 2021

N Engl J Med 2021;385:203-216

DOI: 10.1056/NEJMoa2026141

VOL. 385 NO. 3

Copyright © 2021

• Abstract
• Methods
• Results
• Discussion
• Notes
• Supplementary Material
• References

• Information & Authors
• Metrics & Citations
• View Options
• References
• Media
• Tables
• Share

AbstractBackground

Older patients who are hospitalized for acute decompensated heart failure have high rates of physical frailty, poor quality of life, delayed recovery, and frequent rehospitalizations. Interventions to address physical frailty in this population are not well established.

Methods

We conducted a multicenter, randomized, controlled trial to eval‎uate a transitional, tailored, progressive rehabilitation intervention that included four physical-function domains (strength, balance, mobility, and endurance). The intervention was initiated during, or early after, hospitalization for heart failure and was continued after discharge for 36 outpatient sessions. The primary outcome was the score on the Short Physical Performance Battery (total scores range from 0 to 12, with lower scores indicating more severe physical dysfunction) at 3 months. The secondary outcome was the 6-month rate of rehospitalization for any cause.

Research Summary

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Results

A total of 349 patients underwent randomization; 175 were assigned to the rehabilitation intervention and 174 to usual care (control). At baseline, patients in each group had markedly impaired physical function, and 97% were frail or prefrail; the mean number of coexisting conditions was five in each group. Patient retention in the intervention group was 82%, and adherence to the intervention sessions was 67%. After adjustment for baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001). At 6 months, the rates of rehospitalization for any cause were 1.18 in the intervention group and 1.28 in the control group (rate ratio, 0.93; 95% CI, 0.66 to 1.19). There were 21 deaths (15 from cardiovascular causes) in the intervention group and 16 deaths (8 from cardiovascular causes) in the control group. The rates of death from any cause were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27).

Conclusions

In a diverse population of older patients who were hospitalized for acute decompensated heart failure, an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains resulted in greater improvement in physical function than usual care. (Funded by the National Institutes of Health and others; REHAB-HF ClinicalTrials.gov number, NCT02196038.)

Quick Take

Physical Rehabilitation for Heart Failure

1m 56s

Acute decompensated heart failure is the leading cause of hospitalization among older persons in the United States1 and is associated with poor health-related quality of life, frequent rehospitalizations, high mortality, and costs exceeding $39 billion per year.1,2 Most intervention trials in acute decompensated heart failure have had neutral results, which suggests that outcomes may be driven in part by mechanisms that have been overlooked.3-5

Among older patients with acute heart failure, physical function is markedly impaired, and frailty rates and the burden of coexisting conditions are high.5-8 Even among older patients with stable and well-compensated heart failure, severe impairments in physical function are often present owing to the combined effects of aging, cardiovascular dysfunction, and skeletal-muscle dysfunction.9,10 As patients with chronic heart failure transition to acute decompensated heart failure, physical function worsens further, and this decline is exacerbated by hospitalization and bed rest.8 These deficits often persist. Many patients never recover baseline function, lose independence, and have high risks of rehospitalization and death after discharge (sometimes referred to as “post-hospital syndrome”).4,5,11–14

However, management guidelines do not address physical dysfunction in patients hospitalized for heart failure,15 and previous exercise training trials excluded patients with heart failure who had recently been hospitalized.10,16 To address these issues, we conducted the Rehabilitation Therapy in Older Acute Heart Failure Patients (REHAB-HF) trial, a multicenter, randomized, single-blind, controlled trial of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains. We hypothesized that the intervention would improve physical function and reduce rates of rehospitalization for any cause at 6 months.

MethodsTrial Design and Oversight

Details of the trial design and intervention methods have been described previously.17,18 The organizational structure is shown in Figure S1 in the Supplementary Appendix, available with the full text of this article at NEJM.org. The steering committee designed the trial and oversaw operations. The protocol, which is available at NEJM.org, was approved by the institutional review board at each site. An independent data and safety monitoring committee eval‎uated patient safety. The authors vouch for the accuracy and completeness of the data and analyses and for the fidelity of the trial to the protocol.

Patients and Randomization

Patients were screened at the time of hospital admission and were enrolled before discharge. Patients were eligible for participation if they were 60 years of age or older, if they had been admitted for acute decompensated heart failure regardless of ejection fraction, if they could walk at least 4 m at enrollment (with or without the aid of an assistive device), if they were functionally independent before admission, and if they were expected to be discharged home. Full details of the inclusion and exclusion criteria are provided in the Supplementary Appendix.17 After eligible patients provided written informed consent and completed baseline testing, they were randomly assigned with equal probability to the rehabilitation intervention (intervention group) or to usual care (control group) by a centralized, Web-based system, with the use of block randomization. Randomization was stratified according to ejection fraction (<45% vs. ≥45%) and clinical site. The patients in both trial groups received usual care, as recommended by their medical providers, which could include inpatient or outpatient physical therapy and standard cardiac rehabilitation.

Trial Procedures

The trial intervention (for the intervention group) was an early, transitional, tailored, progressive physical rehabilitation program that had been developed for frail, older patients with acute decompensated heart failure.18,19 The intervention focused on four physical-function domains (strength, balance, mobility, and endurance) and progressed through four prespecified functional levels within each domain (Table S1). The progression of exercise intensity and the types of exercises at each session were individualized on the basis of the patient’s performance level within each domain.18 A key goal was to increase each patient’s endurance (duration of walking); doing this safely required first addressing deficits in balance, strength, and mobility.

The intervention was initiated in the hospital when feasible and was subsequently transitioned to an outpatient facility as soon as possible after discharge. If needed, home-based sessions were provided by interventionists until the patient was physically able to attend the facility-based outpatient sessions. Outpatient sessions were 60 minutes long, occurred 3 days per week for 12 weeks (or 36 sessions), and were conducted at a 1:1 interventionist–patient ratio. Outpatient sessions were complemented by home exercise (low-intensity walking, which was gradually increased to up to 30 minutes daily, and strengthening exercises) on nonprogram days. The home exercise component of the intervention was initiated only after a visit to the patient’s home by an interventionist to eval‎uate the home environment.18

A key goal of the intervention during the first 3 months (the outpatient phase) was to prepare the patient to transition to the independent maintenance phase (months 4 through 6). At the 3-month visit, patients were provided with individualized exercise prescriptions and were subsequently followed every 4 weeks by telephone contact. Patient retention in the intervention group and adherence to the intervention sessions were reviewed and discussed every 2 weeks by a dedicated committee in accordance with the recommendations of the National Institutes of Health Behavior Change Consortium Treatment Fidelity Workgroup.18 Additional details regarding the intervention are provided in the Supplementary Appendix.

Patients who had been randomly assigned to the control group received a telephone call every 2 weeks and had in-person clinic visits at 1 month and 3 months after discharge from the index hospitalization.17 Information regarding the occurrence of symptoms or clinical events and the receipt of rehabilitation therapy unrelated to the trial was collected. Patients received no specific recommendations with respect to exercise, but they were encouraged to adhere to prescribed usual-care therapy and follow-up appointments. Additional details regarding the control group are provided in the Supplementary Appendix.

Trial Outcomes

Outcome measures of physical and cognitive function were assessed by personnel who were unaware of the trial-group assignments. Physical function, quality of life, depression, and cognitive function were assessed at baseline in the hospital and at 3 months.17 Clinical events were ascertained throughout follow-up from monthly interviews and from review of medical records.

The primary outcome was the score on the Short Physical Performance Battery at 3 months. The Short Physical Performance Battery is a standardized, reproducible measure of global physical function that has been validated in frail, older persons and predicts a wide range of clinical outcomes.20-22 It has three components: a standing balance test, a gait-speed (4-m walk) test, and a strength test (as assessed by the time needed to rise from a chair five times). Each component is scored on a scale of 0 to 4; the sum of the scores ranges from 0 to 12, with lower scores indicating more severe physical dysfunction.

The secondary outcome was the rate of rehospitalization for any cause at 6 months, with rehospitalization defined as any hospital stay longer than 24 hours. The reasons for rehospitalization were categorized as noncardiovascular cause, heart failure, or another cardiovascular cause by an independent adjudicator who was unaware of the trial-group assignments.

Additional physical-function outcomes included 6-minute walk distance, frailty status (assessed according to modified Fried criteria6), hand-grip strength, and gait speed at 3 months. Quality of life was assessed at 3 months with the use of the Kansas City Cardiomyopathy Questionnaire (KCCQ) and the EQ-5D-5L (also known as the European Quality of Life 5-Dimension 5-Level questionnaire) visual-analogue scale. Other outcomes included the Geriatric Depression Scale–15 score and the Montreal Cognitive Assessment score.

Statistical Analysis

On the basis of the results from a pilot study,19 we estimated that 258 patients who could be eval‎uated for efficacy would provide the trial with 80% power to detect a 10% difference (equivalent to a difference of 0.6 points) between the intervention group and the control group in the score on the Short Physical Performance Battery at 3 months (the primary outcome); the enrollment of 334 patients who could be eval‎uated for efficacy would be needed to detect a 25% difference in the rate of rehospitalization at 6 months (the secondary outcome). We planned to enroll 360 patients in order to allow for approximately 7% of the patients to withdraw from the trial.

Baseline characteristics are presented as means and standard deviations for continuous variables and as counts and percentages for categorical variables. To account for deaths and loss to follow-up, joint models of continuous and survival outcomes were used to assess differences between the intervention group and the control group in the 3-month outcomes (including the primary outcome), with adjustment for baseline measures.23 Differences between the two groups in the rate of rehospitalization (for any cause and for heart failure) and in the number of days of rehospitalization were assessed with the use of joint models similar to those described above, with a Poisson distribution for clinical events based on counts and a negative binomial distribution for days of rehospitalization for any cause to account for overdispersion. Differences between the two groups in the rate of death and in the rate of combined rehospitalization for any cause and death were assessed with the use of generalized linear models, with a Poisson distribution. Differences in proportion-based (binary) clinical measures were analyzed with the use of logistic regression. All the models were adjusted for clinical site, ejection fraction category, age, and sex. The secondary outcome was adjusted for the baseline score on the Short Physical Performance Battery. The potential consistency of intervention effects among prespecified subgroups for the primary outcome was examined with the use of forest plots.

For the primary outcome, a two-tailed P value of less than 0.05 was considered to indicate statistical significance. For all other outcomes, effect-size estimates and 95% confidence intervals are reported without P values. The widths of the confidence intervals were not adjusted for multiple comparisons, so the intervals should not be used to infer definitive treatment effects for the secondary outcome or other outcomes. Analyses were performed with SAS Enterprise Guide, version 7.11, and SAS software, version 9.4.

ResultsTrial Population and Baseline Characteristics

The first patient was enrolled in September 2014, and the last patient was enrolled in September 2019. A total of 27,300 hospital admission records (which included multiple repeat admissions and thus did not represent unique patients) were electronically reviewed. Ultimately, 410 patients were screened for eligibility, and 349 were enrolled; 175 were randomly assigned to the intervention group and 174 to the control group. Data were available for analysis of the primary and secondary outcomes for 87% and 99% of the patients, respectively (Figure 1 and Fig. S2).

Figure 1

Screening and Randomization.

The demographic and clinical characteristics of the patients at baseline are shown in Table 1 and in Table S3. The mean (±SD) age was 72.7±8.1 years, 52% of the patients were women, and 49% were non-White (of whom 94% were Black). The cause of heart failure was ischemic heart disease in 35% of the patients, and 53% had preserved ejection fraction. Patients had high burdens of coexisting conditions, and the incidences of previous hospitalization for any cause and previous hospitalization for heart failure were high (approximately 45% and 25%, respectively). Most (97%) of the patients were assessed as frail or prefrail, according to the modified Fried criteria. Urinary incontinence, falls, and depression were common. The incidence of diabetes mellitus was higher in the intervention group than in the control group (58% vs. 47%). At baseline, the patients were assessed as having severely impaired physical function, poor quality of life, and at least mild cognitive dysfunction (Table 2).

Table 1

Demographic and Clinical Characteristics of the Patients at Baseline.

Table 2

Trial Outcomes.

Follow-up and Trial Outcomes

The last follow-up visit was in March 2020. Among the 175 patients who had been randomly assigned to the rehabilitation intervention, 12 died before completing the intervention, 14 were lost to follow-up for the analysis of the primary outcome, and 16 permanently discontinued the intervention but were included in the analyses of the primary and secondary outcomes. In an analysis that excluded patients who died, patient retention in the intervention group was 82%, and patients completed a mean (±SE) of 24.3±1.0 outpatient intervention sessions; adherence to the sessions was 67±3% (Fig. S3 and Table S4). After adjustment for sessions missed because of medical appointments and illness, adherence to the intervention was 78±3%. Patients generally progressed to higher functional levels in each domain during the course of the intervention (Figure 2). A key goal was to increase each patient’s exercise endurance (duration of walking); among patients who participated in the first and last sessions, the mean (±SD) endurance doubled from 10.7±5.9 minutes in the first session to 22.0±11.1 minutes in the last session. Additional data regarding exercise during the hospitalization, outpatient, and maintenance phases are provided in Tables S5 through S7; information regarding usual-care exercise therapy not associated with the trial is provided in Table S8.

Figure 2

Progression of Physical Function over Time in the Intervention Group, According to Exercise Domain.

After adjustment for the baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001) (Table 2 and Fig. S4). This effect appeared to be relatively uniform across a wide variety of prespecified subgroups (Figure 3). The results for each of the three components of the Short Physical Performance Battery are shown in Table 2. The benefit of the rehabilitation intervention persisted after post hoc adjustment for baseline imbalances in diagnoses of diabetes and peripheral vascular disease; the least-squares mean score was 8.3±0.2 in the intervention group and 6.8±0.2 in the control group (mean between-group difference, 1.5; 95% CI, 0.9 to 2.1).

Figure 3

Prespecified Subgroup Analysis of the Primary Outcome.

The secondary outcome, the rate of rehospitalization for any cause at 6 months, showed no appreciable difference between the intervention group and the control group, with rates of 1.18 and 1.28, respectively (rate ratio, 0.93; 95% CI, 0.66 to 1.19) (Table 2). The exploratory outcomes, including 6-minute walk distance, gait speed, hand-grip strength, frailty status, quality of life, cognition, depression, and clinical events including falls and rehospitalizations, are shown in Table 2 and Figure S5.

There were 21 deaths in the intervention group and 16 deaths in the control group; the rates of death were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27) (Table 2). Among these deaths, 15 in the intervention group and 8 in the control group were from cardiovascular causes (Table S9). Serious and nonserious adverse events are summarized in Tables S10 and S11. Chest pain, hypertension, dizziness, hyperglycemia, and hypoglycemia were more common in the intervention group than in the control group, and falls and heart failure were more common in the control group.

Discussion

The REHAB-HF trial examined the effects of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains in frail, older patients who were hospitalized for acute decompensated heart failure. The intervention group had significantly greater improvement in physical function, as assessed by the score on the Short Physical Performance Battery at 3 months, than the control group. The results of the analyses of 6-minute walk distance, frailty status, quality of life, and depression also suggested clinical benefits of the intervention. Over the course of 6 months, the incidence of rehospitalization for any cause, rehospitalization for heart failure, and death was high in both groups.

Our trial was designed to address several critical evidence gaps regarding physical rehabilitation in patients with heart failure. Most previous trials excluded patients who had been hospitalized within the previous 6 weeks — a period during which the severity of physical dysfunction and the risk of clinical events are highest; those trials also involved few older, frail patients with multiple coexisting conditions in whom different approaches may be appropriate.3,10 In previous early trials of rehabilitation after heart failure, enrollment of the patients and initiation of the intervention began, on average, 7 weeks after hospital discharge; traditional endurance exercise training was commonly used24,25; the enrolled patients were younger and much less frail and diverse than those in our trial25; there was no control group24,26; and the trials were unblinded25,27 and often small, single-center trials.27 One of the largest of such trials involving recently hospitalized patients, EJECTION-HF (Exercise Joins Education: Combined Therapy to Improve Outcomes in Newly-Discharged Heart Failure), showed no benefit of the intervention over usual care with respect to 6-minute walk distance, rehospitalization, and death, but adherence to the intervention was low (43%).25

Physical dysfunction, frailty, and depression are often unrecognized clinically in older patients hospitalized for heart failure,8,28 are generally not addressed in clinical care pathways,10,29 and probably contribute to delayed, incomplete recovery and high rates of rehospitalization, death, and long-term loss of independence after hospital discharge.2,11,12,14 Physical-function impairments in the patients in the REHAB-HF trial were broader and more severe than those observed in patients with chronic heart failure.28 For example, the mean baseline 6-minute walk distance in the REHAB-HF trial was half that observed in the HF-ACTION (Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training) trial,30 and severe leg weakness prevented nearly one third of the patients from standing even once from a seated position without the use of their arms. The patients in our trial also had severe deficits in balance and mobility, and a history of falls and other geriatric conditions was common — findings that are not typically seen in patients with chronic heart failure and are not addressed by conventional cardiac rehabilitation. The initiation of standard endurance exercise training in frail, older patients without first addressing deficits in balance and mobility can limit efficacy31 and increase the risk of injuries and falls.32,33

The intervention-related benefits seen in the REHAB-HF trial generally exceeded previously reported values for the minimal clinically important difference. The mean difference between the groups in the Short Physical Performance Battery score (1.5 points) was three times as large as the reported minimal clinically important difference (0.5 points).21,34 All three of the components of the Short Physical Performance Battery — corresponding to balance, strength, and mobility — showed greater improvement in the intervention group than in the control group. The apparent benefits in 6-minute walk distance (34 m) and KCCQ score (7.1 points) were also larger than the reported minimal clinically important differences (30 m and 5 points, respectively).35,36 The suggested benefit for depression is of interest, since depression is common among patients with heart failure and is associated with frequent rehospitalization,37 and trials targeting depression in patients with heart failure have had neutral results.

The greater improvements in physical function relative to the control group were seen despite the receipt of routine physical or occupational therapy or traditional cardiac or pulmonary rehabilitation as part of usual care by 43% of the patients in the control group. At 6 months, 83% of the patients in the intervention group who were alive and were being followed by telephone contact reported regular home exercise, which suggested that behavioral change — a requisite for long-term adherence — may have occurred.

The intervention-related benefits may be related to both the severity of the baseline deficits and the robust, broad systemic effects of physical exercise, which favorably alters energy metabolism, oxidative stress, inflammation, tissue repair, growth-factor response, and regulatory pathways.38 Older patients with heart failure can have severe skeletal-muscle myopathy that contributes to physical dysfunction and abates with exercise.9

The number of deaths, including deaths from cardiovascular causes, was higher in the intervention group than in the control group, although the numbers and differences were small and may have been due to chance. A meta-analysis of trials of exercise-based rehabilitation for heart failure showed no significant effect on death from any cause during follow-up for up to 12 months.39 In the EJECTION-HF trial, fewer deaths were reported in the intervention group than in the control group at 12 months.25 Given the wide confidence interval for the rate ratio for death from any cause, we cannot rule out the possibility of an increase (or decrease) in risk with an early exercise regimen among some patients.

Our trial has other important limitations. First, the results did not show a beneficial effect on clinical events. However, a study that examined patient preferences in patients with heart failure indicated that improving physical function and maintaining independence are highly valued, independent of clinical events.40 Second, although the staff members who assessed the primary outcome were unaware of the trial-group assignments, it was not possible for patients to be unaware of the group to which they had been randomly assigned. Third, the benefits of the intervention over usual care may have been moderated owing to the usual-care exercise therapy received by the control group. Fourth, differences between the groups in the amount of caregiver attention could have influenced outcomes. Fifth, the long-term durability of the benefit of the intervention is uncertain. Finally, many patients were ineligible or unable or unwilling to participate, and some discontinued the intervention.

Among patients who were hospitalized for acute decompensated heart failure, a transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains and that began during, or early after, hospitalization and continued for 12 weeks after hospital discharge resulted in significantly greater improvement in physical function than usual care.

Notes

This article was published on May 16, 2021, at NEJM.org.

A data sharing statement provided by the authors is available with the full text of this article at NEJM.org.

Supported by research grants from the National Institutes of Health (R01AG045551, R01AG18915, P30AG021332, P30AG028716, and U24AG059624), the Kermit Glenn Phillips II Chair in Cardiovascular Medicine, and the Oristano Family Fund at Wake Forest School of Medicine.

Dr. Kitzman reports receiving steering committee fees and data and safety monitoring board fees from AbbVie, grant support, paid to Wake Forest Baptist Health, and advisory board fees from AstraZeneca and Bayer, advisory board fees from Boehringer Ingelheim and Corvia Medical, consulting fees from Keyto and Merck, and grant support, paid to Wake Forest Baptist Health, and consulting fees from Novartis; Dr. Whellan, receiving grant support, paid to Thomas Jefferson University, from Amgen, end-point committee fees from CVRx and FibroGen, steering committee fees from Cytokinetics, and grant support, paid to Thomas Jefferson University, and advisory board fees from Novo Nordisk; Dr. Duncan, owning stock in Care Directions; and Dr. Mentz, receiving grant support from American Regent and Cytokinetics, consulting fees from Amgen, Boehringer Ingelheim, and Relypsa, research support and honoraria from Abbott, American Regent, Amgen, Boehringer Ingelheim, Eli Lilly, Boston Scientific, Cytokinetics, FAST BioMedical, Gilead Sciences, Innolife, Medtronic, Relypsa, Respicardia, Roche, Sanofi, Vifor Pharma, and Windtree Therapeutics, grant support, research support, and honoraria from AstraZeneca, and research support, consulting fees, and honoraria from Bayer, Merck, and Novartis. No other potential conflict of interest relevant to this article was reported.

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

We thank Alain Bertoni, M.D., Peter Brubaker, Ph.D., and Jack Rejeski, Ph.D. for their critical review and editing of an earlier version of the manuscript; and Timothy Morgan, Ph.D. (deceased), for his biostatistical expertise.

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Authors: Dalane W. Kitzman, M.D., David J. Whellan, M.D., M.H.S., Pamela Duncan, P.T., Ph.D., Amy M. Pastva, P.T., Ph.D., Robert J. Mentz, M.D., Gordon R. Reeves, M.D., M.P.T., M. Benjamin Nelson, M.S. https://orcid.org/0000-0002-1701-6071, +5 , and Christopher M. O’Connor, M.D.Author Info & Affiliations

Published May 16, 2021

N Engl J Med 2021;385:203-216

DOI: 10.1056/NEJMoa2026141

VOL. 385 NO. 3

Copyright © 2021

• Abstract
• Methods
• Results
• Discussion
• Notes
• Supplementary Material
• References

• Information & Authors
• Metrics & Citations
• View Options
• References
• Media
• Tables
• Share

AbstractBackground

Older patients who are hospitalized for acute decompensated heart failure have high rates of physical frailty, poor quality of life, delayed recovery, and frequent rehospitalizations. Interventions to address physical frailty in this population are not well established.

Methods

We conducted a multicenter, randomized, controlled trial to eval‎uate a transitional, tailored, progressive rehabilitation intervention that included four physical-function domains (strength, balance, mobility, and endurance). The intervention was initiated during, or early after, hospitalization for heart failure and was continued after discharge for 36 outpatient sessions. The primary outcome was the score on the Short Physical Performance Battery (total scores range from 0 to 12, with lower scores indicating more severe physical dysfunction) at 3 months. The secondary outcome was the 6-month rate of rehospitalization for any cause.

Research Summary

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Results

A total of 349 patients underwent randomization; 175 were assigned to the rehabilitation intervention and 174 to usual care (control). At baseline, patients in each group had markedly impaired physical function, and 97% were frail or prefrail; the mean number of coexisting conditions was five in each group. Patient retention in the intervention group was 82%, and adherence to the intervention sessions was 67%. After adjustment for baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001). At 6 months, the rates of rehospitalization for any cause were 1.18 in the intervention group and 1.28 in the control group (rate ratio, 0.93; 95% CI, 0.66 to 1.19). There were 21 deaths (15 from cardiovascular causes) in the intervention group and 16 deaths (8 from cardiovascular causes) in the control group. The rates of death from any cause were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27).

Conclusions

In a diverse population of older patients who were hospitalized for acute decompensated heart failure, an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains resulted in greater improvement in physical function than usual care. (Funded by the National Institutes of Health and others; REHAB-HF ClinicalTrials.gov number, NCT02196038.)

Quick Take

Physical Rehabilitation for Heart Failure

1m 56s

Acute decompensated heart failure is the leading cause of hospitalization among older persons in the United States1 and is associated with poor health-related quality of life, frequent rehospitalizations, high mortality, and costs exceeding $39 billion per year.1,2 Most intervention trials in acute decompensated heart failure have had neutral results, which suggests that outcomes may be driven in part by mechanisms that have been overlooked.3-5

Among older patients with acute heart failure, physical function is markedly impaired, and frailty rates and the burden of coexisting conditions are high.5-8 Even among older patients with stable and well-compensated heart failure, severe impairments in physical function are often present owing to the combined effects of aging, cardiovascular dysfunction, and skeletal-muscle dysfunction.9,10 As patients with chronic heart failure transition to acute decompensated heart failure, physical function worsens further, and this decline is exacerbated by hospitalization and bed rest.8 These deficits often persist. Many patients never recover baseline function, lose independence, and have high risks of rehospitalization and death after discharge (sometimes referred to as “post-hospital syndrome”).4,5,11–14

However, management guidelines do not address physical dysfunction in patients hospitalized for heart failure,15 and previous exercise training trials excluded patients with heart failure who had recently been hospitalized.10,16 To address these issues, we conducted the Rehabilitation Therapy in Older Acute Heart Failure Patients (REHAB-HF) trial, a multicenter, randomized, single-blind, controlled trial of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains. We hypothesized that the intervention would improve physical function and reduce rates of rehospitalization for any cause at 6 months.

MethodsTrial Design and Oversight

Details of the trial design and intervention methods have been described previously.17,18 The organizational structure is shown in Figure S1 in the Supplementary Appendix, available with the full text of this article at NEJM.org. The steering committee designed the trial and oversaw operations. The protocol, which is available at NEJM.org, was approved by the institutional review board at each site. An independent data and safety monitoring committee eval‎uated patient safety. The authors vouch for the accuracy and completeness of the data and analyses and for the fidelity of the trial to the protocol.

Patients and Randomization

Patients were screened at the time of hospital admission and were enrolled before discharge. Patients were eligible for participation if they were 60 years of age or older, if they had been admitted for acute decompensated heart failure regardless of ejection fraction, if they could walk at least 4 m at enrollment (with or without the aid of an assistive device), if they were functionally independent before admission, and if they were expected to be discharged home. Full details of the inclusion and exclusion criteria are provided in the Supplementary Appendix.17 After eligible patients provided written informed consent and completed baseline testing, they were randomly assigned with equal probability to the rehabilitation intervention (intervention group) or to usual care (control group) by a centralized, Web-based system, with the use of block randomization. Randomization was stratified according to ejection fraction (<45% vs. ≥45%) and clinical site. The patients in both trial groups received usual care, as recommended by their medical providers, which could include inpatient or outpatient physical therapy and standard cardiac rehabilitation.

Trial Procedures

The trial intervention (for the intervention group) was an early, transitional, tailored, progressive physical rehabilitation program that had been developed for frail, older patients with acute decompensated heart failure.18,19 The intervention focused on four physical-function domains (strength, balance, mobility, and endurance) and progressed through four prespecified functional levels within each domain (Table S1). The progression of exercise intensity and the types of exercises at each session were individualized on the basis of the patient’s performance level within each domain.18 A key goal was to increase each patient’s endurance (duration of walking); doing this safely required first addressing deficits in balance, strength, and mobility.

The intervention was initiated in the hospital when feasible and was subsequently transitioned to an outpatient facility as soon as possible after discharge. If needed, home-based sessions were provided by interventionists until the patient was physically able to attend the facility-based outpatient sessions. Outpatient sessions were 60 minutes long, occurred 3 days per week for 12 weeks (or 36 sessions), and were conducted at a 1:1 interventionist–patient ratio. Outpatient sessions were complemented by home exercise (low-intensity walking, which was gradually increased to up to 30 minutes daily, and strengthening exercises) on nonprogram days. The home exercise component of the intervention was initiated only after a visit to the patient’s home by an interventionist to eval‎uate the home environment.18

A key goal of the intervention during the first 3 months (the outpatient phase) was to prepare the patient to transition to the independent maintenance phase (months 4 through 6). At the 3-month visit, patients were provided with individualized exercise prescriptions and were subsequently followed every 4 weeks by telephone contact. Patient retention in the intervention group and adherence to the intervention sessions were reviewed and discussed every 2 weeks by a dedicated committee in accordance with the recommendations of the National Institutes of Health Behavior Change Consortium Treatment Fidelity Workgroup.18 Additional details regarding the intervention are provided in the Supplementary Appendix.

Patients who had been randomly assigned to the control group received a telephone call every 2 weeks and had in-person clinic visits at 1 month and 3 months after discharge from the index hospitalization.17 Information regarding the occurrence of symptoms or clinical events and the receipt of rehabilitation therapy unrelated to the trial was collected. Patients received no specific recommendations with respect to exercise, but they were encouraged to adhere to prescribed usual-care therapy and follow-up appointments. Additional details regarding the control group are provided in the Supplementary Appendix.

Trial Outcomes

Outcome measures of physical and cognitive function were assessed by personnel who were unaware of the trial-group assignments. Physical function, quality of life, depression, and cognitive function were assessed at baseline in the hospital and at 3 months.17 Clinical events were ascertained throughout follow-up from monthly interviews and from review of medical records.

The primary outcome was the score on the Short Physical Performance Battery at 3 months. The Short Physical Performance Battery is a standardized, reproducible measure of global physical function that has been validated in frail, older persons and predicts a wide range of clinical outcomes.20-22 It has three components: a standing balance test, a gait-speed (4-m walk) test, and a strength test (as assessed by the time needed to rise from a chair five times). Each component is scored on a scale of 0 to 4; the sum of the scores ranges from 0 to 12, with lower scores indicating more severe physical dysfunction.

The secondary outcome was the rate of rehospitalization for any cause at 6 months, with rehospitalization defined as any hospital stay longer than 24 hours. The reasons for rehospitalization were categorized as noncardiovascular cause, heart failure, or another cardiovascular cause by an independent adjudicator who was unaware of the trial-group assignments.

Additional physical-function outcomes included 6-minute walk distance, frailty status (assessed according to modified Fried criteria6), hand-grip strength, and gait speed at 3 months. Quality of life was assessed at 3 months with the use of the Kansas City Cardiomyopathy Questionnaire (KCCQ) and the EQ-5D-5L (also known as the European Quality of Life 5-Dimension 5-Level questionnaire) visual-analogue scale. Other outcomes included the Geriatric Depression Scale–15 score and the Montreal Cognitive Assessment score.

Statistical Analysis

On the basis of the results from a pilot study,19 we estimated that 258 patients who could be eval‎uated for efficacy would provide the trial with 80% power to detect a 10% difference (equivalent to a difference of 0.6 points) between the intervention group and the control group in the score on the Short Physical Performance Battery at 3 months (the primary outcome); the enrollment of 334 patients who could be eval‎uated for efficacy would be needed to detect a 25% difference in the rate of rehospitalization at 6 months (the secondary outcome). We planned to enroll 360 patients in order to allow for approximately 7% of the patients to withdraw from the trial.

Baseline characteristics are presented as means and standard deviations for continuous variables and as counts and percentages for categorical variables. To account for deaths and loss to follow-up, joint models of continuous and survival outcomes were used to assess differences between the intervention group and the control group in the 3-month outcomes (including the primary outcome), with adjustment for baseline measures.23 Differences between the two groups in the rate of rehospitalization (for any cause and for heart failure) and in the number of days of rehospitalization were assessed with the use of joint models similar to those described above, with a Poisson distribution for clinical events based on counts and a negative binomial distribution for days of rehospitalization for any cause to account for overdispersion. Differences between the two groups in the rate of death and in the rate of combined rehospitalization for any cause and death were assessed with the use of generalized linear models, with a Poisson distribution. Differences in proportion-based (binary) clinical measures were analyzed with the use of logistic regression. All the models were adjusted for clinical site, ejection fraction category, age, and sex. The secondary outcome was adjusted for the baseline score on the Short Physical Performance Battery. The potential consistency of intervention effects among prespecified subgroups for the primary outcome was examined with the use of forest plots.

For the primary outcome, a two-tailed P value of less than 0.05 was considered to indicate statistical significance. For all other outcomes, effect-size estimates and 95% confidence intervals are reported without P values. The widths of the confidence intervals were not adjusted for multiple comparisons, so the intervals should not be used to infer definitive treatment effects for the secondary outcome or other outcomes. Analyses were performed with SAS Enterprise Guide, version 7.11, and SAS software, version 9.4.

ResultsTrial Population and Baseline Characteristics

The first patient was enrolled in September 2014, and the last patient was enrolled in September 2019. A total of 27,300 hospital admission records (which included multiple repeat admissions and thus did not represent unique patients) were electronically reviewed. Ultimately, 410 patients were screened for eligibility, and 349 were enrolled; 175 were randomly assigned to the intervention group and 174 to the control group. Data were available for analysis of the primary and secondary outcomes for 87% and 99% of the patients, respectively (Figure 1 and Fig. S2).

Figure 1

Screening and Randomization.

The demographic and clinical characteristics of the patients at baseline are shown in Table 1 and in Table S3. The mean (±SD) age was 72.7±8.1 years, 52% of the patients were women, and 49% were non-White (of whom 94% were Black). The cause of heart failure was ischemic heart disease in 35% of the patients, and 53% had preserved ejection fraction. Patients had high burdens of coexisting conditions, and the incidences of previous hospitalization for any cause and previous hospitalization for heart failure were high (approximately 45% and 25%, respectively). Most (97%) of the patients were assessed as frail or prefrail, according to the modified Fried criteria. Urinary incontinence, falls, and depression were common. The incidence of diabetes mellitus was higher in the intervention group than in the control group (58% vs. 47%). At baseline, the patients were assessed as having severely impaired physical function, poor quality of life, and at least mild cognitive dysfunction (Table 2).

Table 1

Demographic and Clinical Characteristics of the Patients at Baseline.

Table 2

Trial Outcomes.

Follow-up and Trial Outcomes

The last follow-up visit was in March 2020. Among the 175 patients who had been randomly assigned to the rehabilitation intervention, 12 died before completing the intervention, 14 were lost to follow-up for the analysis of the primary outcome, and 16 permanently discontinued the intervention but were included in the analyses of the primary and secondary outcomes. In an analysis that excluded patients who died, patient retention in the intervention group was 82%, and patients completed a mean (±SE) of 24.3±1.0 outpatient intervention sessions; adherence to the sessions was 67±3% (Fig. S3 and Table S4). After adjustment for sessions missed because of medical appointments and illness, adherence to the intervention was 78±3%. Patients generally progressed to higher functional levels in each domain during the course of the intervention (Figure 2). A key goal was to increase each patient’s exercise endurance (duration of walking); among patients who participated in the first and last sessions, the mean (±SD) endurance doubled from 10.7±5.9 minutes in the first session to 22.0±11.1 minutes in the last session. Additional data regarding exercise during the hospitalization, outpatient, and maintenance phases are provided in Tables S5 through S7; information regarding usual-care exercise therapy not associated with the trial is provided in Table S8.

Figure 2

Progression of Physical Function over Time in the Intervention Group, According to Exercise Domain.

After adjustment for the baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001) (Table 2 and Fig. S4). This effect appeared to be relatively uniform across a wide variety of prespecified subgroups (Figure 3). The results for each of the three components of the Short Physical Performance Battery are shown in Table 2. The benefit of the rehabilitation intervention persisted after post hoc adjustment for baseline imbalances in diagnoses of diabetes and peripheral vascular disease; the least-squares mean score was 8.3±0.2 in the intervention group and 6.8±0.2 in the control group (mean between-group difference, 1.5; 95% CI, 0.9 to 2.1).

Figure 3

Prespecified Subgroup Analysis of the Primary Outcome.

The secondary outcome, the rate of rehospitalization for any cause at 6 months, showed no appreciable difference between the intervention group and the control group, with rates of 1.18 and 1.28, respectively (rate ratio, 0.93; 95% CI, 0.66 to 1.19) (Table 2). The exploratory outcomes, including 6-minute walk distance, gait speed, hand-grip strength, frailty status, quality of life, cognition, depression, and clinical events including falls and rehospitalizations, are shown in Table 2 and Figure S5.

There were 21 deaths in the intervention group and 16 deaths in the control group; the rates of death were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27) (Table 2). Among these deaths, 15 in the intervention group and 8 in the control group were from cardiovascular causes (Table S9). Serious and nonserious adverse events are summarized in Tables S10 and S11. Chest pain, hypertension, dizziness, hyperglycemia, and hypoglycemia were more common in the intervention group than in the control group, and falls and heart failure were more common in the control group.

Discussion

The REHAB-HF trial examined the effects of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains in frail, older patients who were hospitalized for acute decompensated heart failure. The intervention group had significantly greater improvement in physical function, as assessed by the score on the Short Physical Performance Battery at 3 months, than the control group. The results of the analyses of 6-minute walk distance, frailty status, quality of life, and depression also suggested clinical benefits of the intervention. Over the course of 6 months, the incidence of rehospitalization for any cause, rehospitalization for heart failure, and death was high in both groups.

Our trial was designed to address several critical evidence gaps regarding physical rehabilitation in patients with heart failure. Most previous trials excluded patients who had been hospitalized within the previous 6 weeks — a period during which the severity of physical dysfunction and the risk of clinical events are highest; those trials also involved few older, frail patients with multiple coexisting conditions in whom different approaches may be appropriate.3,10 In previous early trials of rehabilitation after heart failure, enrollment of the patients and initiation of the intervention began, on average, 7 weeks after hospital discharge; traditional endurance exercise training was commonly used24,25; the enrolled patients were younger and much less frail and diverse than those in our trial25; there was no control group24,26; and the trials were unblinded25,27 and often small, single-center trials.27 One of the largest of such trials involving recently hospitalized patients, EJECTION-HF (Exercise Joins Education: Combined Therapy to Improve Outcomes in Newly-Discharged Heart Failure), showed no benefit of the intervention over usual care with respect to 6-minute walk distance, rehospitalization, and death, but adherence to the intervention was low (43%).25

Physical dysfunction, frailty, and depression are often unrecognized clinically in older patients hospitalized for heart failure,8,28 are generally not addressed in clinical care pathways,10,29 and probably contribute to delayed, incomplete recovery and high rates of rehospitalization, death, and long-term loss of independence after hospital discharge.2,11,12,14 Physical-function impairments in the patients in the REHAB-HF trial were broader and more severe than those observed in patients with chronic heart failure.28 For example, the mean baseline 6-minute walk distance in the REHAB-HF trial was half that observed in the HF-ACTION (Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training) trial,30 and severe leg weakness prevented nearly one third of the patients from standing even once from a seated position without the use of their arms. The patients in our trial also had severe deficits in balance and mobility, and a history of falls and other geriatric conditions was common — findings that are not typically seen in patients with chronic heart failure and are not addressed by conventional cardiac rehabilitation. The initiation of standard endurance exercise training in frail, older patients without first addressing deficits in balance and mobility can limit efficacy31 and increase the risk of injuries and falls.32,33

The intervention-related benefits seen in the REHAB-HF trial generally exceeded previously reported values for the minimal clinically important difference. The mean difference between the groups in the Short Physical Performance Battery score (1.5 points) was three times as large as the reported minimal clinically important difference (0.5 points).21,34 All three of the components of the Short Physical Performance Battery — corresponding to balance, strength, and mobility — showed greater improvement in the intervention group than in the control group. The apparent benefits in 6-minute walk distance (34 m) and KCCQ score (7.1 points) were also larger than the reported minimal clinically important differences (30 m and 5 points, respectively).35,36 The suggested benefit for depression is of interest, since depression is common among patients with heart failure and is associated with frequent rehospitalization,37 and trials targeting depression in patients with heart failure have had neutral results.

The greater improvements in physical function relative to the control group were seen despite the receipt of routine physical or occupational therapy or traditional cardiac or pulmonary rehabilitation as part of usual care by 43% of the patients in the control group. At 6 months, 83% of the patients in the intervention group who were alive and were being followed by telephone contact reported regular home exercise, which suggested that behavioral change — a requisite for long-term adherence — may have occurred.

The intervention-related benefits may be related to both the severity of the baseline deficits and the robust, broad systemic effects of physical exercise, which favorably alters energy metabolism, oxidative stress, inflammation, tissue repair, growth-factor response, and regulatory pathways.38 Older patients with heart failure can have severe skeletal-muscle myopathy that contributes to physical dysfunction and abates with exercise.9

The number of deaths, including deaths from cardiovascular causes, was higher in the intervention group than in the control group, although the numbers and differences were small and may have been due to chance. A meta-analysis of trials of exercise-based rehabilitation for heart failure showed no significant effect on death from any cause during follow-up for up to 12 months.39 In the EJECTION-HF trial, fewer deaths were reported in the intervention group than in the control group at 12 months.25 Given the wide confidence interval for the rate ratio for death from any cause, we cannot rule out the possibility of an increase (or decrease) in risk with an early exercise regimen among some patients.

Our trial has other important limitations. First, the results did not show a beneficial effect on clinical events. However, a study that examined patient preferences in patients with heart failure indicated that improving physical function and maintaining independence are highly valued, independent of clinical events.40 Second, although the staff members who assessed the primary outcome were unaware of the trial-group assignments, it was not possible for patients to be unaware of the group to which they had been randomly assigned. Third, the benefits of the intervention over usual care may have been moderated owing to the usual-care exercise therapy received by the control group. Fourth, differences between the groups in the amount of caregiver attention could have influenced outcomes. Fifth, the long-term durability of the benefit of the intervention is uncertain. Finally, many patients were ineligible or unable or unwilling to participate, and some discontinued the intervention.

Among patients who were hospitalized for acute decompensated heart failure, a transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains and that began during, or early after, hospitalization and continued for 12 weeks after hospital discharge resulted in significantly greater improvement in physical function than usual care.

Notes

This article was published on May 16, 2021, at NEJM.org.

A data sharing statement provided by the authors is available with the full text of this article at NEJM.org.

Supported by research grants from the National Institutes of Health (R01AG045551, R01AG18915, P30AG021332, P30AG028716, and U24AG059624), the Kermit Glenn Phillips II Chair in Cardiovascular Medicine, and the Oristano Family Fund at Wake Forest School of Medicine.

Dr. Kitzman reports receiving steering committee fees and data and safety monitoring board fees from AbbVie, grant support, paid to Wake Forest Baptist Health, and advisory board fees from AstraZeneca and Bayer, advisory board fees from Boehringer Ingelheim and Corvia Medical, consulting fees from Keyto and Merck, and grant support, paid to Wake Forest Baptist Health, and consulting fees from Novartis; Dr. Whellan, receiving grant support, paid to Thomas Jefferson University, from Amgen, end-point committee fees from CVRx and FibroGen, steering committee fees from Cytokinetics, and grant support, paid to Thomas Jefferson University, and advisory board fees from Novo Nordisk; Dr. Duncan, owning stock in Care Directions; and Dr. Mentz, receiving grant support from American Regent and Cytokinetics, consulting fees from Amgen, Boehringer Ingelheim, and Relypsa, research support and honoraria from Abbott, American Regent, Amgen, Boehringer Ingelheim, Eli Lilly, Boston Scientific, Cytokinetics, FAST BioMedical, Gilead Sciences, Innolife, Medtronic, Relypsa, Respicardia, Roche, Sanofi, Vifor Pharma, and Windtree Therapeutics, grant support, research support, and honoraria from AstraZeneca, and research support, consulting fees, and honoraria from Bayer, Merck, and Novartis. No other potential conflict of interest relevant to this article was reported.

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

We thank Alain Bertoni, M.D., Peter Brubaker, Ph.D., and Jack Rejeski, Ph.D. for their critical review and editing of an earlier version of the manuscript; and Timothy Morgan, Ph.D. (deceased), for his biostatistical expertise.

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Authors: Dalane W. Kitzman, M.D., David J. Whellan, M.D., M.H.S., Pamela Duncan, P.T., Ph.D., Amy M. Pastva, P.T., Ph.D., Robert J. Mentz, M.D., Gordon R. Reeves, M.D., M.P.T., M. Benjamin Nelson, M.S. https://orcid.org/0000-0002-1701-6071, +5 , and Christopher M. O’Connor, M.D.Author Info & Affiliations

Published May 16, 2021

N Engl J Med 2021;385:203-216

DOI: 10.1056/NEJMoa2026141

VOL. 385 NO. 3

Copyright © 2021

• Abstract
• Methods
• Results
• Discussion
• Notes
• Supplementary Material
• References

• Information & Authors
• Metrics & Citations
• View Options
• References
• Media
• Tables
• Share

AbstractBackground

Older patients who are hospitalized for acute decompensated heart failure have high rates of physical frailty, poor quality of life, delayed recovery, and frequent rehospitalizations. Interventions to address physical frailty in this population are not well established.

Methods

We conducted a multicenter, randomized, controlled trial to eval‎uate a transitional, tailored, progressive rehabilitation intervention that included four physical-function domains (strength, balance, mobility, and endurance). The intervention was initiated during, or early after, hospitalization for heart failure and was continued after discharge for 36 outpatient sessions. The primary outcome was the score on the Short Physical Performance Battery (total scores range from 0 to 12, with lower scores indicating more severe physical dysfunction) at 3 months. The secondary outcome was the 6-month rate of rehospitalization for any cause.

Research Summary

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Results

A total of 349 patients underwent randomization; 175 were assigned to the rehabilitation intervention and 174 to usual care (control). At baseline, patients in each group had markedly impaired physical function, and 97% were frail or prefrail; the mean number of coexisting conditions was five in each group. Patient retention in the intervention group was 82%, and adherence to the intervention sessions was 67%. After adjustment for baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001). At 6 months, the rates of rehospitalization for any cause were 1.18 in the intervention group and 1.28 in the control group (rate ratio, 0.93; 95% CI, 0.66 to 1.19). There were 21 deaths (15 from cardiovascular causes) in the intervention group and 16 deaths (8 from cardiovascular causes) in the control group. The rates of death from any cause were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27).

Conclusions

In a diverse population of older patients who were hospitalized for acute decompensated heart failure, an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains resulted in greater improvement in physical function than usual care. (Funded by the National Institutes of Health and others; REHAB-HF ClinicalTrials.gov number, NCT02196038.)

Quick Take

Physical Rehabilitation for Heart Failure

1m 56s

Acute decompensated heart failure is the leading cause of hospitalization among older persons in the United States1 and is associated with poor health-related quality of life, frequent rehospitalizations, high mortality, and costs exceeding $39 billion per year.1,2 Most intervention trials in acute decompensated heart failure have had neutral results, which suggests that outcomes may be driven in part by mechanisms that have been overlooked.3-5

Among older patients with acute heart failure, physical function is markedly impaired, and frailty rates and the burden of coexisting conditions are high.5-8 Even among older patients with stable and well-compensated heart failure, severe impairments in physical function are often present owing to the combined effects of aging, cardiovascular dysfunction, and skeletal-muscle dysfunction.9,10 As patients with chronic heart failure transition to acute decompensated heart failure, physical function worsens further, and this decline is exacerbated by hospitalization and bed rest.8 These deficits often persist. Many patients never recover baseline function, lose independence, and have high risks of rehospitalization and death after discharge (sometimes referred to as “post-hospital syndrome”).4,5,11–14

However, management guidelines do not address physical dysfunction in patients hospitalized for heart failure,15 and previous exercise training trials excluded patients with heart failure who had recently been hospitalized.10,16 To address these issues, we conducted the Rehabilitation Therapy in Older Acute Heart Failure Patients (REHAB-HF) trial, a multicenter, randomized, single-blind, controlled trial of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains. We hypothesized that the intervention would improve physical function and reduce rates of rehospitalization for any cause at 6 months.

MethodsTrial Design and Oversight

Details of the trial design and intervention methods have been described previously.17,18 The organizational structure is shown in Figure S1 in the Supplementary Appendix, available with the full text of this article at NEJM.org. The steering committee designed the trial and oversaw operations. The protocol, which is available at NEJM.org, was approved by the institutional review board at each site. An independent data and safety monitoring committee eval‎uated patient safety. The authors vouch for the accuracy and completeness of the data and analyses and for the fidelity of the trial to the protocol.

Patients and Randomization

Patients were screened at the time of hospital admission and were enrolled before discharge. Patients were eligible for participation if they were 60 years of age or older, if they had been admitted for acute decompensated heart failure regardless of ejection fraction, if they could walk at least 4 m at enrollment (with or without the aid of an assistive device), if they were functionally independent before admission, and if they were expected to be discharged home. Full details of the inclusion and exclusion criteria are provided in the Supplementary Appendix.17 After eligible patients provided written informed consent and completed baseline testing, they were randomly assigned with equal probability to the rehabilitation intervention (intervention group) or to usual care (control group) by a centralized, Web-based system, with the use of block randomization. Randomization was stratified according to ejection fraction (<45% vs. ≥45%) and clinical site. The patients in both trial groups received usual care, as recommended by their medical providers, which could include inpatient or outpatient physical therapy and standard cardiac rehabilitation.

Trial Procedures

The trial intervention (for the intervention group) was an early, transitional, tailored, progressive physical rehabilitation program that had been developed for frail, older patients with acute decompensated heart failure.18,19 The intervention focused on four physical-function domains (strength, balance, mobility, and endurance) and progressed through four prespecified functional levels within each domain (Table S1). The progression of exercise intensity and the types of exercises at each session were individualized on the basis of the patient’s performance level within each domain.18 A key goal was to increase each patient’s endurance (duration of walking); doing this safely required first addressing deficits in balance, strength, and mobility.

The intervention was initiated in the hospital when feasible and was subsequently transitioned to an outpatient facility as soon as possible after discharge. If needed, home-based sessions were provided by interventionists until the patient was physically able to attend the facility-based outpatient sessions. Outpatient sessions were 60 minutes long, occurred 3 days per week for 12 weeks (or 36 sessions), and were conducted at a 1:1 interventionist–patient ratio. Outpatient sessions were complemented by home exercise (low-intensity walking, which was gradually increased to up to 30 minutes daily, and strengthening exercises) on nonprogram days. The home exercise component of the intervention was initiated only after a visit to the patient’s home by an interventionist to eval‎uate the home environment.18

A key goal of the intervention during the first 3 months (the outpatient phase) was to prepare the patient to transition to the independent maintenance phase (months 4 through 6). At the 3-month visit, patients were provided with individualized exercise prescriptions and were subsequently followed every 4 weeks by telephone contact. Patient retention in the intervention group and adherence to the intervention sessions were reviewed and discussed every 2 weeks by a dedicated committee in accordance with the recommendations of the National Institutes of Health Behavior Change Consortium Treatment Fidelity Workgroup.18 Additional details regarding the intervention are provided in the Supplementary Appendix.

Patients who had been randomly assigned to the control group received a telephone call every 2 weeks and had in-person clinic visits at 1 month and 3 months after discharge from the index hospitalization.17 Information regarding the occurrence of symptoms or clinical events and the receipt of rehabilitation therapy unrelated to the trial was collected. Patients received no specific recommendations with respect to exercise, but they were encouraged to adhere to prescribed usual-care therapy and follow-up appointments. Additional details regarding the control group are provided in the Supplementary Appendix.

Trial Outcomes

Outcome measures of physical and cognitive function were assessed by personnel who were unaware of the trial-group assignments. Physical function, quality of life, depression, and cognitive function were assessed at baseline in the hospital and at 3 months.17 Clinical events were ascertained throughout follow-up from monthly interviews and from review of medical records.

The primary outcome was the score on the Short Physical Performance Battery at 3 months. The Short Physical Performance Battery is a standardized, reproducible measure of global physical function that has been validated in frail, older persons and predicts a wide range of clinical outcomes.20-22 It has three components: a standing balance test, a gait-speed (4-m walk) test, and a strength test (as assessed by the time needed to rise from a chair five times). Each component is scored on a scale of 0 to 4; the sum of the scores ranges from 0 to 12, with lower scores indicating more severe physical dysfunction.

The secondary outcome was the rate of rehospitalization for any cause at 6 months, with rehospitalization defined as any hospital stay longer than 24 hours. The reasons for rehospitalization were categorized as noncardiovascular cause, heart failure, or another cardiovascular cause by an independent adjudicator who was unaware of the trial-group assignments.

Additional physical-function outcomes included 6-minute walk distance, frailty status (assessed according to modified Fried criteria6), hand-grip strength, and gait speed at 3 months. Quality of life was assessed at 3 months with the use of the Kansas City Cardiomyopathy Questionnaire (KCCQ) and the EQ-5D-5L (also known as the European Quality of Life 5-Dimension 5-Level questionnaire) visual-analogue scale. Other outcomes included the Geriatric Depression Scale–15 score and the Montreal Cognitive Assessment score.

Statistical Analysis

On the basis of the results from a pilot study,19 we estimated that 258 patients who could be eval‎uated for efficacy would provide the trial with 80% power to detect a 10% difference (equivalent to a difference of 0.6 points) between the intervention group and the control group in the score on the Short Physical Performance Battery at 3 months (the primary outcome); the enrollment of 334 patients who could be eval‎uated for efficacy would be needed to detect a 25% difference in the rate of rehospitalization at 6 months (the secondary outcome). We planned to enroll 360 patients in order to allow for approximately 7% of the patients to withdraw from the trial.

Baseline characteristics are presented as means and standard deviations for continuous variables and as counts and percentages for categorical variables. To account for deaths and loss to follow-up, joint models of continuous and survival outcomes were used to assess differences between the intervention group and the control group in the 3-month outcomes (including the primary outcome), with adjustment for baseline measures.23 Differences between the two groups in the rate of rehospitalization (for any cause and for heart failure) and in the number of days of rehospitalization were assessed with the use of joint models similar to those described above, with a Poisson distribution for clinical events based on counts and a negative binomial distribution for days of rehospitalization for any cause to account for overdispersion. Differences between the two groups in the rate of death and in the rate of combined rehospitalization for any cause and death were assessed with the use of generalized linear models, with a Poisson distribution. Differences in proportion-based (binary) clinical measures were analyzed with the use of logistic regression. All the models were adjusted for clinical site, ejection fraction category, age, and sex. The secondary outcome was adjusted for the baseline score on the Short Physical Performance Battery. The potential consistency of intervention effects among prespecified subgroups for the primary outcome was examined with the use of forest plots.

For the primary outcome, a two-tailed P value of less than 0.05 was considered to indicate statistical significance. For all other outcomes, effect-size estimates and 95% confidence intervals are reported without P values. The widths of the confidence intervals were not adjusted for multiple comparisons, so the intervals should not be used to infer definitive treatment effects for the secondary outcome or other outcomes. Analyses were performed with SAS Enterprise Guide, version 7.11, and SAS software, version 9.4.

ResultsTrial Population and Baseline Characteristics

The first patient was enrolled in September 2014, and the last patient was enrolled in September 2019. A total of 27,300 hospital admission records (which included multiple repeat admissions and thus did not represent unique patients) were electronically reviewed. Ultimately, 410 patients were screened for eligibility, and 349 were enrolled; 175 were randomly assigned to the intervention group and 174 to the control group. Data were available for analysis of the primary and secondary outcomes for 87% and 99% of the patients, respectively (Figure 1 and Fig. S2).

Figure 1

Screening and Randomization.

The demographic and clinical characteristics of the patients at baseline are shown in Table 1 and in Table S3. The mean (±SD) age was 72.7±8.1 years, 52% of the patients were women, and 49% were non-White (of whom 94% were Black). The cause of heart failure was ischemic heart disease in 35% of the patients, and 53% had preserved ejection fraction. Patients had high burdens of coexisting conditions, and the incidences of previous hospitalization for any cause and previous hospitalization for heart failure were high (approximately 45% and 25%, respectively). Most (97%) of the patients were assessed as frail or prefrail, according to the modified Fried criteria. Urinary incontinence, falls, and depression were common. The incidence of diabetes mellitus was higher in the intervention group than in the control group (58% vs. 47%). At baseline, the patients were assessed as having severely impaired physical function, poor quality of life, and at least mild cognitive dysfunction (Table 2).

Table 1

Demographic and Clinical Characteristics of the Patients at Baseline.

Table 2

Trial Outcomes.

Follow-up and Trial Outcomes

The last follow-up visit was in March 2020. Among the 175 patients who had been randomly assigned to the rehabilitation intervention, 12 died before completing the intervention, 14 were lost to follow-up for the analysis of the primary outcome, and 16 permanently discontinued the intervention but were included in the analyses of the primary and secondary outcomes. In an analysis that excluded patients who died, patient retention in the intervention group was 82%, and patients completed a mean (±SE) of 24.3±1.0 outpatient intervention sessions; adherence to the sessions was 67±3% (Fig. S3 and Table S4). After adjustment for sessions missed because of medical appointments and illness, adherence to the intervention was 78±3%. Patients generally progressed to higher functional levels in each domain during the course of the intervention (Figure 2). A key goal was to increase each patient’s exercise endurance (duration of walking); among patients who participated in the first and last sessions, the mean (±SD) endurance doubled from 10.7±5.9 minutes in the first session to 22.0±11.1 minutes in the last session. Additional data regarding exercise during the hospitalization, outpatient, and maintenance phases are provided in Tables S5 through S7; information regarding usual-care exercise therapy not associated with the trial is provided in Table S8.

Figure 2

Progression of Physical Function over Time in the Intervention Group, According to Exercise Domain.

After adjustment for the baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001) (Table 2 and Fig. S4). This effect appeared to be relatively uniform across a wide variety of prespecified subgroups (Figure 3). The results for each of the three components of the Short Physical Performance Battery are shown in Table 2. The benefit of the rehabilitation intervention persisted after post hoc adjustment for baseline imbalances in diagnoses of diabetes and peripheral vascular disease; the least-squares mean score was 8.3±0.2 in the intervention group and 6.8±0.2 in the control group (mean between-group difference, 1.5; 95% CI, 0.9 to 2.1).

Figure 3

Prespecified Subgroup Analysis of the Primary Outcome.

The secondary outcome, the rate of rehospitalization for any cause at 6 months, showed no appreciable difference between the intervention group and the control group, with rates of 1.18 and 1.28, respectively (rate ratio, 0.93; 95% CI, 0.66 to 1.19) (Table 2). The exploratory outcomes, including 6-minute walk distance, gait speed, hand-grip strength, frailty status, quality of life, cognition, depression, and clinical events including falls and rehospitalizations, are shown in Table 2 and Figure S5.

There were 21 deaths in the intervention group and 16 deaths in the control group; the rates of death were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27) (Table 2). Among these deaths, 15 in the intervention group and 8 in the control group were from cardiovascular causes (Table S9). Serious and nonserious adverse events are summarized in Tables S10 and S11. Chest pain, hypertension, dizziness, hyperglycemia, and hypoglycemia were more common in the intervention group than in the control group, and falls and heart failure were more common in the control group.

Discussion

The REHAB-HF trial examined the effects of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains in frail, older patients who were hospitalized for acute decompensated heart failure. The intervention group had significantly greater improvement in physical function, as assessed by the score on the Short Physical Performance Battery at 3 months, than the control group. The results of the analyses of 6-minute walk distance, frailty status, quality of life, and depression also suggested clinical benefits of the intervention. Over the course of 6 months, the incidence of rehospitalization for any cause, rehospitalization for heart failure, and death was high in both groups.

Our trial was designed to address several critical evidence gaps regarding physical rehabilitation in patients with heart failure. Most previous trials excluded patients who had been hospitalized within the previous 6 weeks — a period during which the severity of physical dysfunction and the risk of clinical events are highest; those trials also involved few older, frail patients with multiple coexisting conditions in whom different approaches may be appropriate.3,10 In previous early trials of rehabilitation after heart failure, enrollment of the patients and initiation of the intervention began, on average, 7 weeks after hospital discharge; traditional endurance exercise training was commonly used24,25; the enrolled patients were younger and much less frail and diverse than those in our trial25; there was no control group24,26; and the trials were unblinded25,27 and often small, single-center trials.27 One of the largest of such trials involving recently hospitalized patients, EJECTION-HF (Exercise Joins Education: Combined Therapy to Improve Outcomes in Newly-Discharged Heart Failure), showed no benefit of the intervention over usual care with respect to 6-minute walk distance, rehospitalization, and death, but adherence to the intervention was low (43%).25

Physical dysfunction, frailty, and depression are often unrecognized clinically in older patients hospitalized for heart failure,8,28 are generally not addressed in clinical care pathways,10,29 and probably contribute to delayed, incomplete recovery and high rates of rehospitalization, death, and long-term loss of independence after hospital discharge.2,11,12,14 Physical-function impairments in the patients in the REHAB-HF trial were broader and more severe than those observed in patients with chronic heart failure.28 For example, the mean baseline 6-minute walk distance in the REHAB-HF trial was half that observed in the HF-ACTION (Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training) trial,30 and severe leg weakness prevented nearly one third of the patients from standing even once from a seated position without the use of their arms. The patients in our trial also had severe deficits in balance and mobility, and a history of falls and other geriatric conditions was common — findings that are not typically seen in patients with chronic heart failure and are not addressed by conventional cardiac rehabilitation. The initiation of standard endurance exercise training in frail, older patients without first addressing deficits in balance and mobility can limit efficacy31 and increase the risk of injuries and falls.32,33

The intervention-related benefits seen in the REHAB-HF trial generally exceeded previously reported values for the minimal clinically important difference. The mean difference between the groups in the Short Physical Performance Battery score (1.5 points) was three times as large as the reported minimal clinically important difference (0.5 points).21,34 All three of the components of the Short Physical Performance Battery — corresponding to balance, strength, and mobility — showed greater improvement in the intervention group than in the control group. The apparent benefits in 6-minute walk distance (34 m) and KCCQ score (7.1 points) were also larger than the reported minimal clinically important differences (30 m and 5 points, respectively).35,36 The suggested benefit for depression is of interest, since depression is common among patients with heart failure and is associated with frequent rehospitalization,37 and trials targeting depression in patients with heart failure have had neutral results.

The greater improvements in physical function relative to the control group were seen despite the receipt of routine physical or occupational therapy or traditional cardiac or pulmonary rehabilitation as part of usual care by 43% of the patients in the control group. At 6 months, 83% of the patients in the intervention group who were alive and were being followed by telephone contact reported regular home exercise, which suggested that behavioral change — a requisite for long-term adherence — may have occurred.

The intervention-related benefits may be related to both the severity of the baseline deficits and the robust, broad systemic effects of physical exercise, which favorably alters energy metabolism, oxidative stress, inflammation, tissue repair, growth-factor response, and regulatory pathways.38 Older patients with heart failure can have severe skeletal-muscle myopathy that contributes to physical dysfunction and abates with exercise.9

The number of deaths, including deaths from cardiovascular causes, was higher in the intervention group than in the control group, although the numbers and differences were small and may have been due to chance. A meta-analysis of trials of exercise-based rehabilitation for heart failure showed no significant effect on death from any cause during follow-up for up to 12 months.39 In the EJECTION-HF trial, fewer deaths were reported in the intervention group than in the control group at 12 months.25 Given the wide confidence interval for the rate ratio for death from any cause, we cannot rule out the possibility of an increase (or decrease) in risk with an early exercise regimen among some patients.

Our trial has other important limitations. First, the results did not show a beneficial effect on clinical events. However, a study that examined patient preferences in patients with heart failure indicated that improving physical function and maintaining independence are highly valued, independent of clinical events.40 Second, although the staff members who assessed the primary outcome were unaware of the trial-group assignments, it was not possible for patients to be unaware of the group to which they had been randomly assigned. Third, the benefits of the intervention over usual care may have been moderated owing to the usual-care exercise therapy received by the control group. Fourth, differences between the groups in the amount of caregiver attention could have influenced outcomes. Fifth, the long-term durability of the benefit of the intervention is uncertain. Finally, many patients were ineligible or unable or unwilling to participate, and some discontinued the intervention.

Among patients who were hospitalized for acute decompensated heart failure, a transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains and that began during, or early after, hospitalization and continued for 12 weeks after hospital discharge resulted in significantly greater improvement in physical function than usual care.

Notes

This article was published on May 16, 2021, at NEJM.org.

A data sharing statement provided by the authors is available with the full text of this article at NEJM.org.

Supported by research grants from the National Institutes of Health (R01AG045551, R01AG18915, P30AG021332, P30AG028716, and U24AG059624), the Kermit Glenn Phillips II Chair in Cardiovascular Medicine, and the Oristano Family Fund at Wake Forest School of Medicine.

Dr. Kitzman reports receiving steering committee fees and data and safety monitoring board fees from AbbVie, grant support, paid to Wake Forest Baptist Health, and advisory board fees from AstraZeneca and Bayer, advisory board fees from Boehringer Ingelheim and Corvia Medical, consulting fees from Keyto and Merck, and grant support, paid to Wake Forest Baptist Health, and consulting fees from Novartis; Dr. Whellan, receiving grant support, paid to Thomas Jefferson University, from Amgen, end-point committee fees from CVRx and FibroGen, steering committee fees from Cytokinetics, and grant support, paid to Thomas Jefferson University, and advisory board fees from Novo Nordisk; Dr. Duncan, owning stock in Care Directions; and Dr. Mentz, receiving grant support from American Regent and Cytokinetics, consulting fees from Amgen, Boehringer Ingelheim, and Relypsa, research support and honoraria from Abbott, American Regent, Amgen, Boehringer Ingelheim, Eli Lilly, Boston Scientific, Cytokinetics, FAST BioMedical, Gilead Sciences, Innolife, Medtronic, Relypsa, Respicardia, Roche, Sanofi, Vifor Pharma, and Windtree Therapeutics, grant support, research support, and honoraria from AstraZeneca, and research support, consulting fees, and honoraria from Bayer, Merck, and Novartis. No other potential conflict of interest relevant to this article was reported.

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

We thank Alain Bertoni, M.D., Peter Brubaker, Ph.D., and Jack Rejeski, Ph.D. for their critical review and editing of an earlier version of the manuscript; and Timothy Morgan, Ph.D. (deceased), for his biostatistical expertise.

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Authors: Dalane W. Kitzman, M.D., David J. Whellan, M.D., M.H.S., Pamela Duncan, P.T., Ph.D., Amy M. Pastva, P.T., Ph.D., Robert J. Mentz, M.D., Gordon R. Reeves, M.D., M.P.T., M. Benjamin Nelson, M.S. https://orcid.org/0000-0002-1701-6071, +5 , and Christopher M. O’Connor, M.D.Author Info & Affiliations

Published May 16, 2021

N Engl J Med 2021;385:203-216

DOI: 10.1056/NEJMoa2026141

VOL. 385 NO. 3

Copyright © 2021

• Abstract
• Methods
• Results
• Discussion
• Notes
• Supplementary Material
• References

• Information & Authors
• Metrics & Citations
• View Options
• References
• Media
• Tables
• Share

AbstractBackground

Older patients who are hospitalized for acute decompensated heart failure have high rates of physical frailty, poor quality of life, delayed recovery, and frequent rehospitalizations. Interventions to address physical frailty in this population are not well established.

Methods

We conducted a multicenter, randomized, controlled trial to eval‎uate a transitional, tailored, progressive rehabilitation intervention that included four physical-function domains (strength, balance, mobility, and endurance). The intervention was initiated during, or early after, hospitalization for heart failure and was continued after discharge for 36 outpatient sessions. The primary outcome was the score on the Short Physical Performance Battery (total scores range from 0 to 12, with lower scores indicating more severe physical dysfunction) at 3 months. The secondary outcome was the 6-month rate of rehospitalization for any cause.

Research Summary

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Results

A total of 349 patients underwent randomization; 175 were assigned to the rehabilitation intervention and 174 to usual care (control). At baseline, patients in each group had markedly impaired physical function, and 97% were frail or prefrail; the mean number of coexisting conditions was five in each group. Patient retention in the intervention group was 82%, and adherence to the intervention sessions was 67%. After adjustment for baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001). At 6 months, the rates of rehospitalization for any cause were 1.18 in the intervention group and 1.28 in the control group (rate ratio, 0.93; 95% CI, 0.66 to 1.19). There were 21 deaths (15 from cardiovascular causes) in the intervention group and 16 deaths (8 from cardiovascular causes) in the control group. The rates of death from any cause were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27).

Conclusions

In a diverse population of older patients who were hospitalized for acute decompensated heart failure, an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains resulted in greater improvement in physical function than usual care. (Funded by the National Institutes of Health and others; REHAB-HF ClinicalTrials.gov number, NCT02196038.)

Quick Take

Physical Rehabilitation for Heart Failure

1m 56s

Acute decompensated heart failure is the leading cause of hospitalization among older persons in the United States1 and is associated with poor health-related quality of life, frequent rehospitalizations, high mortality, and costs exceeding $39 billion per year.1,2 Most intervention trials in acute decompensated heart failure have had neutral results, which suggests that outcomes may be driven in part by mechanisms that have been overlooked.3-5

Among older patients with acute heart failure, physical function is markedly impaired, and frailty rates and the burden of coexisting conditions are high.5-8 Even among older patients with stable and well-compensated heart failure, severe impairments in physical function are often present owing to the combined effects of aging, cardiovascular dysfunction, and skeletal-muscle dysfunction.9,10 As patients with chronic heart failure transition to acute decompensated heart failure, physical function worsens further, and this decline is exacerbated by hospitalization and bed rest.8 These deficits often persist. Many patients never recover baseline function, lose independence, and have high risks of rehospitalization and death after discharge (sometimes referred to as “post-hospital syndrome”).4,5,11–14

However, management guidelines do not address physical dysfunction in patients hospitalized for heart failure,15 and previous exercise training trials excluded patients with heart failure who had recently been hospitalized.10,16 To address these issues, we conducted the Rehabilitation Therapy in Older Acute Heart Failure Patients (REHAB-HF) trial, a multicenter, randomized, single-blind, controlled trial of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains. We hypothesized that the intervention would improve physical function and reduce rates of rehospitalization for any cause at 6 months.

MethodsTrial Design and Oversight

Details of the trial design and intervention methods have been described previously.17,18 The organizational structure is shown in Figure S1 in the Supplementary Appendix, available with the full text of this article at NEJM.org. The steering committee designed the trial and oversaw operations. The protocol, which is available at NEJM.org, was approved by the institutional review board at each site. An independent data and safety monitoring committee eval‎uated patient safety. The authors vouch for the accuracy and completeness of the data and analyses and for the fidelity of the trial to the protocol.

Patients and Randomization

Patients were screened at the time of hospital admission and were enrolled before discharge. Patients were eligible for participation if they were 60 years of age or older, if they had been admitted for acute decompensated heart failure regardless of ejection fraction, if they could walk at least 4 m at enrollment (with or without the aid of an assistive device), if they were functionally independent before admission, and if they were expected to be discharged home. Full details of the inclusion and exclusion criteria are provided in the Supplementary Appendix.17 After eligible patients provided written informed consent and completed baseline testing, they were randomly assigned with equal probability to the rehabilitation intervention (intervention group) or to usual care (control group) by a centralized, Web-based system, with the use of block randomization. Randomization was stratified according to ejection fraction (<45% vs. ≥45%) and clinical site. The patients in both trial groups received usual care, as recommended by their medical providers, which could include inpatient or outpatient physical therapy and standard cardiac rehabilitation.

Trial Procedures

The trial intervention (for the intervention group) was an early, transitional, tailored, progressive physical rehabilitation program that had been developed for frail, older patients with acute decompensated heart failure.18,19 The intervention focused on four physical-function domains (strength, balance, mobility, and endurance) and progressed through four prespecified functional levels within each domain (Table S1). The progression of exercise intensity and the types of exercises at each session were individualized on the basis of the patient’s performance level within each domain.18 A key goal was to increase each patient’s endurance (duration of walking); doing this safely required first addressing deficits in balance, strength, and mobility.

The intervention was initiated in the hospital when feasible and was subsequently transitioned to an outpatient facility as soon as possible after discharge. If needed, home-based sessions were provided by interventionists until the patient was physically able to attend the facility-based outpatient sessions. Outpatient sessions were 60 minutes long, occurred 3 days per week for 12 weeks (or 36 sessions), and were conducted at a 1:1 interventionist–patient ratio. Outpatient sessions were complemented by home exercise (low-intensity walking, which was gradually increased to up to 30 minutes daily, and strengthening exercises) on nonprogram days. The home exercise component of the intervention was initiated only after a visit to the patient’s home by an interventionist to eval‎uate the home environment.18

A key goal of the intervention during the first 3 months (the outpatient phase) was to prepare the patient to transition to the independent maintenance phase (months 4 through 6). At the 3-month visit, patients were provided with individualized exercise prescriptions and were subsequently followed every 4 weeks by telephone contact. Patient retention in the intervention group and adherence to the intervention sessions were reviewed and discussed every 2 weeks by a dedicated committee in accordance with the recommendations of the National Institutes of Health Behavior Change Consortium Treatment Fidelity Workgroup.18 Additional details regarding the intervention are provided in the Supplementary Appendix.

Patients who had been randomly assigned to the control group received a telephone call every 2 weeks and had in-person clinic visits at 1 month and 3 months after discharge from the index hospitalization.17 Information regarding the occurrence of symptoms or clinical events and the receipt of rehabilitation therapy unrelated to the trial was collected. Patients received no specific recommendations with respect to exercise, but they were encouraged to adhere to prescribed usual-care therapy and follow-up appointments. Additional details regarding the control group are provided in the Supplementary Appendix.

Trial Outcomes

Outcome measures of physical and cognitive function were assessed by personnel who were unaware of the trial-group assignments. Physical function, quality of life, depression, and cognitive function were assessed at baseline in the hospital and at 3 months.17 Clinical events were ascertained throughout follow-up from monthly interviews and from review of medical records.

The primary outcome was the score on the Short Physical Performance Battery at 3 months. The Short Physical Performance Battery is a standardized, reproducible measure of global physical function that has been validated in frail, older persons and predicts a wide range of clinical outcomes.20-22 It has three components: a standing balance test, a gait-speed (4-m walk) test, and a strength test (as assessed by the time needed to rise from a chair five times). Each component is scored on a scale of 0 to 4; the sum of the scores ranges from 0 to 12, with lower scores indicating more severe physical dysfunction.

The secondary outcome was the rate of rehospitalization for any cause at 6 months, with rehospitalization defined as any hospital stay longer than 24 hours. The reasons for rehospitalization were categorized as noncardiovascular cause, heart failure, or another cardiovascular cause by an independent adjudicator who was unaware of the trial-group assignments.

Additional physical-function outcomes included 6-minute walk distance, frailty status (assessed according to modified Fried criteria6), hand-grip strength, and gait speed at 3 months. Quality of life was assessed at 3 months with the use of the Kansas City Cardiomyopathy Questionnaire (KCCQ) and the EQ-5D-5L (also known as the European Quality of Life 5-Dimension 5-Level questionnaire) visual-analogue scale. Other outcomes included the Geriatric Depression Scale–15 score and the Montreal Cognitive Assessment score.

Statistical Analysis

On the basis of the results from a pilot study,19 we estimated that 258 patients who could be eval‎uated for efficacy would provide the trial with 80% power to detect a 10% difference (equivalent to a difference of 0.6 points) between the intervention group and the control group in the score on the Short Physical Performance Battery at 3 months (the primary outcome); the enrollment of 334 patients who could be eval‎uated for efficacy would be needed to detect a 25% difference in the rate of rehospitalization at 6 months (the secondary outcome). We planned to enroll 360 patients in order to allow for approximately 7% of the patients to withdraw from the trial.

Baseline characteristics are presented as means and standard deviations for continuous variables and as counts and percentages for categorical variables. To account for deaths and loss to follow-up, joint models of continuous and survival outcomes were used to assess differences between the intervention group and the control group in the 3-month outcomes (including the primary outcome), with adjustment for baseline measures.23 Differences between the two groups in the rate of rehospitalization (for any cause and for heart failure) and in the number of days of rehospitalization were assessed with the use of joint models similar to those described above, with a Poisson distribution for clinical events based on counts and a negative binomial distribution for days of rehospitalization for any cause to account for overdispersion. Differences between the two groups in the rate of death and in the rate of combined rehospitalization for any cause and death were assessed with the use of generalized linear models, with a Poisson distribution. Differences in proportion-based (binary) clinical measures were analyzed with the use of logistic regression. All the models were adjusted for clinical site, ejection fraction category, age, and sex. The secondary outcome was adjusted for the baseline score on the Short Physical Performance Battery. The potential consistency of intervention effects among prespecified subgroups for the primary outcome was examined with the use of forest plots.

For the primary outcome, a two-tailed P value of less than 0.05 was considered to indicate statistical significance. For all other outcomes, effect-size estimates and 95% confidence intervals are reported without P values. The widths of the confidence intervals were not adjusted for multiple comparisons, so the intervals should not be used to infer definitive treatment effects for the secondary outcome or other outcomes. Analyses were performed with SAS Enterprise Guide, version 7.11, and SAS software, version 9.4.

ResultsTrial Population and Baseline Characteristics

The first patient was enrolled in September 2014, and the last patient was enrolled in September 2019. A total of 27,300 hospital admission records (which included multiple repeat admissions and thus did not represent unique patients) were electronically reviewed. Ultimately, 410 patients were screened for eligibility, and 349 were enrolled; 175 were randomly assigned to the intervention group and 174 to the control group. Data were available for analysis of the primary and secondary outcomes for 87% and 99% of the patients, respectively (Figure 1 and Fig. S2).

Figure 1

Screening and Randomization.

The demographic and clinical characteristics of the patients at baseline are shown in Table 1 and in Table S3. The mean (±SD) age was 72.7±8.1 years, 52% of the patients were women, and 49% were non-White (of whom 94% were Black). The cause of heart failure was ischemic heart disease in 35% of the patients, and 53% had preserved ejection fraction. Patients had high burdens of coexisting conditions, and the incidences of previous hospitalization for any cause and previous hospitalization for heart failure were high (approximately 45% and 25%, respectively). Most (97%) of the patients were assessed as frail or prefrail, according to the modified Fried criteria. Urinary incontinence, falls, and depression were common. The incidence of diabetes mellitus was higher in the intervention group than in the control group (58% vs. 47%). At baseline, the patients were assessed as having severely impaired physical function, poor quality of life, and at least mild cognitive dysfunction (Table 2).

Table 1

Demographic and Clinical Characteristics of the Patients at Baseline.

Table 2

Trial Outcomes.

Follow-up and Trial Outcomes

The last follow-up visit was in March 2020. Among the 175 patients who had been randomly assigned to the rehabilitation intervention, 12 died before completing the intervention, 14 were lost to follow-up for the analysis of the primary outcome, and 16 permanently discontinued the intervention but were included in the analyses of the primary and secondary outcomes. In an analysis that excluded patients who died, patient retention in the intervention group was 82%, and patients completed a mean (±SE) of 24.3±1.0 outpatient intervention sessions; adherence to the sessions was 67±3% (Fig. S3 and Table S4). After adjustment for sessions missed because of medical appointments and illness, adherence to the intervention was 78±3%. Patients generally progressed to higher functional levels in each domain during the course of the intervention (Figure 2). A key goal was to increase each patient’s exercise endurance (duration of walking); among patients who participated in the first and last sessions, the mean (±SD) endurance doubled from 10.7±5.9 minutes in the first session to 22.0±11.1 minutes in the last session. Additional data regarding exercise during the hospitalization, outpatient, and maintenance phases are provided in Tables S5 through S7; information regarding usual-care exercise therapy not associated with the trial is provided in Table S8.

Figure 2

Progression of Physical Function over Time in the Intervention Group, According to Exercise Domain.

After adjustment for the baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001) (Table 2 and Fig. S4). This effect appeared to be relatively uniform across a wide variety of prespecified subgroups (Figure 3). The results for each of the three components of the Short Physical Performance Battery are shown in Table 2. The benefit of the rehabilitation intervention persisted after post hoc adjustment for baseline imbalances in diagnoses of diabetes and peripheral vascular disease; the least-squares mean score was 8.3±0.2 in the intervention group and 6.8±0.2 in the control group (mean between-group difference, 1.5; 95% CI, 0.9 to 2.1).

Figure 3

Prespecified Subgroup Analysis of the Primary Outcome.

The secondary outcome, the rate of rehospitalization for any cause at 6 months, showed no appreciable difference between the intervention group and the control group, with rates of 1.18 and 1.28, respectively (rate ratio, 0.93; 95% CI, 0.66 to 1.19) (Table 2). The exploratory outcomes, including 6-minute walk distance, gait speed, hand-grip strength, frailty status, quality of life, cognition, depression, and clinical events including falls and rehospitalizations, are shown in Table 2 and Figure S5.

There were 21 deaths in the intervention group and 16 deaths in the control group; the rates of death were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27) (Table 2). Among these deaths, 15 in the intervention group and 8 in the control group were from cardiovascular causes (Table S9). Serious and nonserious adverse events are summarized in Tables S10 and S11. Chest pain, hypertension, dizziness, hyperglycemia, and hypoglycemia were more common in the intervention group than in the control group, and falls and heart failure were more common in the control group.

Discussion

The REHAB-HF trial examined the effects of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains in frail, older patients who were hospitalized for acute decompensated heart failure. The intervention group had significantly greater improvement in physical function, as assessed by the score on the Short Physical Performance Battery at 3 months, than the control group. The results of the analyses of 6-minute walk distance, frailty status, quality of life, and depression also suggested clinical benefits of the intervention. Over the course of 6 months, the incidence of rehospitalization for any cause, rehospitalization for heart failure, and death was high in both groups.

Our trial was designed to address several critical evidence gaps regarding physical rehabilitation in patients with heart failure. Most previous trials excluded patients who had been hospitalized within the previous 6 weeks — a period during which the severity of physical dysfunction and the risk of clinical events are highest; those trials also involved few older, frail patients with multiple coexisting conditions in whom different approaches may be appropriate.3,10 In previous early trials of rehabilitation after heart failure, enrollment of the patients and initiation of the intervention began, on average, 7 weeks after hospital discharge; traditional endurance exercise training was commonly used24,25; the enrolled patients were younger and much less frail and diverse than those in our trial25; there was no control group24,26; and the trials were unblinded25,27 and often small, single-center trials.27 One of the largest of such trials involving recently hospitalized patients, EJECTION-HF (Exercise Joins Education: Combined Therapy to Improve Outcomes in Newly-Discharged Heart Failure), showed no benefit of the intervention over usual care with respect to 6-minute walk distance, rehospitalization, and death, but adherence to the intervention was low (43%).25

Physical dysfunction, frailty, and depression are often unrecognized clinically in older patients hospitalized for heart failure,8,28 are generally not addressed in clinical care pathways,10,29 and probably contribute to delayed, incomplete recovery and high rates of rehospitalization, death, and long-term loss of independence after hospital discharge.2,11,12,14 Physical-function impairments in the patients in the REHAB-HF trial were broader and more severe than those observed in patients with chronic heart failure.28 For example, the mean baseline 6-minute walk distance in the REHAB-HF trial was half that observed in the HF-ACTION (Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training) trial,30 and severe leg weakness prevented nearly one third of the patients from standing even once from a seated position without the use of their arms. The patients in our trial also had severe deficits in balance and mobility, and a history of falls and other geriatric conditions was common — findings that are not typically seen in patients with chronic heart failure and are not addressed by conventional cardiac rehabilitation. The initiation of standard endurance exercise training in frail, older patients without first addressing deficits in balance and mobility can limit efficacy31 and increase the risk of injuries and falls.32,33

The intervention-related benefits seen in the REHAB-HF trial generally exceeded previously reported values for the minimal clinically important difference. The mean difference between the groups in the Short Physical Performance Battery score (1.5 points) was three times as large as the reported minimal clinically important difference (0.5 points).21,34 All three of the components of the Short Physical Performance Battery — corresponding to balance, strength, and mobility — showed greater improvement in the intervention group than in the control group. The apparent benefits in 6-minute walk distance (34 m) and KCCQ score (7.1 points) were also larger than the reported minimal clinically important differences (30 m and 5 points, respectively).35,36 The suggested benefit for depression is of interest, since depression is common among patients with heart failure and is associated with frequent rehospitalization,37 and trials targeting depression in patients with heart failure have had neutral results.

The greater improvements in physical function relative to the control group were seen despite the receipt of routine physical or occupational therapy or traditional cardiac or pulmonary rehabilitation as part of usual care by 43% of the patients in the control group. At 6 months, 83% of the patients in the intervention group who were alive and were being followed by telephone contact reported regular home exercise, which suggested that behavioral change — a requisite for long-term adherence — may have occurred.

The intervention-related benefits may be related to both the severity of the baseline deficits and the robust, broad systemic effects of physical exercise, which favorably alters energy metabolism, oxidative stress, inflammation, tissue repair, growth-factor response, and regulatory pathways.38 Older patients with heart failure can have severe skeletal-muscle myopathy that contributes to physical dysfunction and abates with exercise.9

The number of deaths, including deaths from cardiovascular causes, was higher in the intervention group than in the control group, although the numbers and differences were small and may have been due to chance. A meta-analysis of trials of exercise-based rehabilitation for heart failure showed no significant effect on death from any cause during follow-up for up to 12 months.39 In the EJECTION-HF trial, fewer deaths were reported in the intervention group than in the control group at 12 months.25 Given the wide confidence interval for the rate ratio for death from any cause, we cannot rule out the possibility of an increase (or decrease) in risk with an early exercise regimen among some patients.

Our trial has other important limitations. First, the results did not show a beneficial effect on clinical events. However, a study that examined patient preferences in patients with heart failure indicated that improving physical function and maintaining independence are highly valued, independent of clinical events.40 Second, although the staff members who assessed the primary outcome were unaware of the trial-group assignments, it was not possible for patients to be unaware of the group to which they had been randomly assigned. Third, the benefits of the intervention over usual care may have been moderated owing to the usual-care exercise therapy received by the control group. Fourth, differences between the groups in the amount of caregiver attention could have influenced outcomes. Fifth, the long-term durability of the benefit of the intervention is uncertain. Finally, many patients were ineligible or unable or unwilling to participate, and some discontinued the intervention.

Among patients who were hospitalized for acute decompensated heart failure, a transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains and that began during, or early after, hospitalization and continued for 12 weeks after hospital discharge resulted in significantly greater improvement in physical function than usual care.

Notes

This article was published on May 16, 2021, at NEJM.org.

A data sharing statement provided by the authors is available with the full text of this article at NEJM.org.

Supported by research grants from the National Institutes of Health (R01AG045551, R01AG18915, P30AG021332, P30AG028716, and U24AG059624), the Kermit Glenn Phillips II Chair in Cardiovascular Medicine, and the Oristano Family Fund at Wake Forest School of Medicine.

Dr. Kitzman reports receiving steering committee fees and data and safety monitoring board fees from AbbVie, grant support, paid to Wake Forest Baptist Health, and advisory board fees from AstraZeneca and Bayer, advisory board fees from Boehringer Ingelheim and Corvia Medical, consulting fees from Keyto and Merck, and grant support, paid to Wake Forest Baptist Health, and consulting fees from Novartis; Dr. Whellan, receiving grant support, paid to Thomas Jefferson University, from Amgen, end-point committee fees from CVRx and FibroGen, steering committee fees from Cytokinetics, and grant support, paid to Thomas Jefferson University, and advisory board fees from Novo Nordisk; Dr. Duncan, owning stock in Care Directions; and Dr. Mentz, receiving grant support from American Regent and Cytokinetics, consulting fees from Amgen, Boehringer Ingelheim, and Relypsa, research support and honoraria from Abbott, American Regent, Amgen, Boehringer Ingelheim, Eli Lilly, Boston Scientific, Cytokinetics, FAST BioMedical, Gilead Sciences, Innolife, Medtronic, Relypsa, Respicardia, Roche, Sanofi, Vifor Pharma, and Windtree Therapeutics, grant support, research support, and honoraria from AstraZeneca, and research support, consulting fees, and honoraria from Bayer, Merck, and Novartis. No other potential conflict of interest relevant to this article was reported.

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

We thank Alain Bertoni, M.D., Peter Brubaker, Ph.D., and Jack Rejeski, Ph.D. for their critical review and editing of an earlier version of the manuscript; and Timothy Morgan, Ph.D. (deceased), for his biostatistical expertise.

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Authors: Dalane W. Kitzman, M.D., David J. Whellan, M.D., M.H.S., Pamela Duncan, P.T., Ph.D., Amy M. Pastva, P.T., Ph.D., Robert J. Mentz, M.D., Gordon R. Reeves, M.D., M.P.T., M. Benjamin Nelson, M.S. https://orcid.org/0000-0002-1701-6071, +5 , and Christopher M. O’Connor, M.D.Author Info & Affiliations

Published May 16, 2021

N Engl J Med 2021;385:203-216

DOI: 10.1056/NEJMoa2026141

VOL. 385 NO. 3

Copyright © 2021

• Abstract
• Methods
• Results
• Discussion
• Notes
• Supplementary Material
• References

• Information & Authors
• Metrics & Citations
• View Options
• References
• Media
• Tables
• Share

AbstractBackground

Older patients who are hospitalized for acute decompensated heart failure have high rates of physical frailty, poor quality of life, delayed recovery, and frequent rehospitalizations. Interventions to address physical frailty in this population are not well established.

Methods

We conducted a multicenter, randomized, controlled trial to eval‎uate a transitional, tailored, progressive rehabilitation intervention that included four physical-function domains (strength, balance, mobility, and endurance). The intervention was initiated during, or early after, hospitalization for heart failure and was continued after discharge for 36 outpatient sessions. The primary outcome was the score on the Short Physical Performance Battery (total scores range from 0 to 12, with lower scores indicating more severe physical dysfunction) at 3 months. The secondary outcome was the 6-month rate of rehospitalization for any cause.

Research Summary

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Results

A total of 349 patients underwent randomization; 175 were assigned to the rehabilitation intervention and 174 to usual care (control). At baseline, patients in each group had markedly impaired physical function, and 97% were frail or prefrail; the mean number of coexisting conditions was five in each group. Patient retention in the intervention group was 82%, and adherence to the intervention sessions was 67%. After adjustment for baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001). At 6 months, the rates of rehospitalization for any cause were 1.18 in the intervention group and 1.28 in the control group (rate ratio, 0.93; 95% CI, 0.66 to 1.19). There were 21 deaths (15 from cardiovascular causes) in the intervention group and 16 deaths (8 from cardiovascular causes) in the control group. The rates of death from any cause were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27).

Conclusions

In a diverse population of older patients who were hospitalized for acute decompensated heart failure, an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains resulted in greater improvement in physical function than usual care. (Funded by the National Institutes of Health and others; REHAB-HF ClinicalTrials.gov number, NCT02196038.)

Quick Take

Physical Rehabilitation for Heart Failure

1m 56s

Acute decompensated heart failure is the leading cause of hospitalization among older persons in the United States1 and is associated with poor health-related quality of life, frequent rehospitalizations, high mortality, and costs exceeding $39 billion per year.1,2 Most intervention trials in acute decompensated heart failure have had neutral results, which suggests that outcomes may be driven in part by mechanisms that have been overlooked.3-5

Among older patients with acute heart failure, physical function is markedly impaired, and frailty rates and the burden of coexisting conditions are high.5-8 Even among older patients with stable and well-compensated heart failure, severe impairments in physical function are often present owing to the combined effects of aging, cardiovascular dysfunction, and skeletal-muscle dysfunction.9,10 As patients with chronic heart failure transition to acute decompensated heart failure, physical function worsens further, and this decline is exacerbated by hospitalization and bed rest.8 These deficits often persist. Many patients never recover baseline function, lose independence, and have high risks of rehospitalization and death after discharge (sometimes referred to as “post-hospital syndrome”).4,5,11–14

However, management guidelines do not address physical dysfunction in patients hospitalized for heart failure,15 and previous exercise training trials excluded patients with heart failure who had recently been hospitalized.10,16 To address these issues, we conducted the Rehabilitation Therapy in Older Acute Heart Failure Patients (REHAB-HF) trial, a multicenter, randomized, single-blind, controlled trial of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains. We hypothesized that the intervention would improve physical function and reduce rates of rehospitalization for any cause at 6 months.

MethodsTrial Design and Oversight

Details of the trial design and intervention methods have been described previously.17,18 The organizational structure is shown in Figure S1 in the Supplementary Appendix, available with the full text of this article at NEJM.org. The steering committee designed the trial and oversaw operations. The protocol, which is available at NEJM.org, was approved by the institutional review board at each site. An independent data and safety monitoring committee eval‎uated patient safety. The authors vouch for the accuracy and completeness of the data and analyses and for the fidelity of the trial to the protocol.

Patients and Randomization

Patients were screened at the time of hospital admission and were enrolled before discharge. Patients were eligible for participation if they were 60 years of age or older, if they had been admitted for acute decompensated heart failure regardless of ejection fraction, if they could walk at least 4 m at enrollment (with or without the aid of an assistive device), if they were functionally independent before admission, and if they were expected to be discharged home. Full details of the inclusion and exclusion criteria are provided in the Supplementary Appendix.17 After eligible patients provided written informed consent and completed baseline testing, they were randomly assigned with equal probability to the rehabilitation intervention (intervention group) or to usual care (control group) by a centralized, Web-based system, with the use of block randomization. Randomization was stratified according to ejection fraction (<45% vs. ≥45%) and clinical site. The patients in both trial groups received usual care, as recommended by their medical providers, which could include inpatient or outpatient physical therapy and standard cardiac rehabilitation.

Trial Procedures

The trial intervention (for the intervention group) was an early, transitional, tailored, progressive physical rehabilitation program that had been developed for frail, older patients with acute decompensated heart failure.18,19 The intervention focused on four physical-function domains (strength, balance, mobility, and endurance) and progressed through four prespecified functional levels within each domain (Table S1). The progression of exercise intensity and the types of exercises at each session were individualized on the basis of the patient’s performance level within each domain.18 A key goal was to increase each patient’s endurance (duration of walking); doing this safely required first addressing deficits in balance, strength, and mobility.

The intervention was initiated in the hospital when feasible and was subsequently transitioned to an outpatient facility as soon as possible after discharge. If needed, home-based sessions were provided by interventionists until the patient was physically able to attend the facility-based outpatient sessions. Outpatient sessions were 60 minutes long, occurred 3 days per week for 12 weeks (or 36 sessions), and were conducted at a 1:1 interventionist–patient ratio. Outpatient sessions were complemented by home exercise (low-intensity walking, which was gradually increased to up to 30 minutes daily, and strengthening exercises) on nonprogram days. The home exercise component of the intervention was initiated only after a visit to the patient’s home by an interventionist to eval‎uate the home environment.18

A key goal of the intervention during the first 3 months (the outpatient phase) was to prepare the patient to transition to the independent maintenance phase (months 4 through 6). At the 3-month visit, patients were provided with individualized exercise prescriptions and were subsequently followed every 4 weeks by telephone contact. Patient retention in the intervention group and adherence to the intervention sessions were reviewed and discussed every 2 weeks by a dedicated committee in accordance with the recommendations of the National Institutes of Health Behavior Change Consortium Treatment Fidelity Workgroup.18 Additional details regarding the intervention are provided in the Supplementary Appendix.

Patients who had been randomly assigned to the control group received a telephone call every 2 weeks and had in-person clinic visits at 1 month and 3 months after discharge from the index hospitalization.17 Information regarding the occurrence of symptoms or clinical events and the receipt of rehabilitation therapy unrelated to the trial was collected. Patients received no specific recommendations with respect to exercise, but they were encouraged to adhere to prescribed usual-care therapy and follow-up appointments. Additional details regarding the control group are provided in the Supplementary Appendix.

Trial Outcomes

Outcome measures of physical and cognitive function were assessed by personnel who were unaware of the trial-group assignments. Physical function, quality of life, depression, and cognitive function were assessed at baseline in the hospital and at 3 months.17 Clinical events were ascertained throughout follow-up from monthly interviews and from review of medical records.

The primary outcome was the score on the Short Physical Performance Battery at 3 months. The Short Physical Performance Battery is a standardized, reproducible measure of global physical function that has been validated in frail, older persons and predicts a wide range of clinical outcomes.20-22 It has three components: a standing balance test, a gait-speed (4-m walk) test, and a strength test (as assessed by the time needed to rise from a chair five times). Each component is scored on a scale of 0 to 4; the sum of the scores ranges from 0 to 12, with lower scores indicating more severe physical dysfunction.

The secondary outcome was the rate of rehospitalization for any cause at 6 months, with rehospitalization defined as any hospital stay longer than 24 hours. The reasons for rehospitalization were categorized as noncardiovascular cause, heart failure, or another cardiovascular cause by an independent adjudicator who was unaware of the trial-group assignments.

Additional physical-function outcomes included 6-minute walk distance, frailty status (assessed according to modified Fried criteria6), hand-grip strength, and gait speed at 3 months. Quality of life was assessed at 3 months with the use of the Kansas City Cardiomyopathy Questionnaire (KCCQ) and the EQ-5D-5L (also known as the European Quality of Life 5-Dimension 5-Level questionnaire) visual-analogue scale. Other outcomes included the Geriatric Depression Scale–15 score and the Montreal Cognitive Assessment score.

Statistical Analysis

On the basis of the results from a pilot study,19 we estimated that 258 patients who could be eval‎uated for efficacy would provide the trial with 80% power to detect a 10% difference (equivalent to a difference of 0.6 points) between the intervention group and the control group in the score on the Short Physical Performance Battery at 3 months (the primary outcome); the enrollment of 334 patients who could be eval‎uated for efficacy would be needed to detect a 25% difference in the rate of rehospitalization at 6 months (the secondary outcome). We planned to enroll 360 patients in order to allow for approximately 7% of the patients to withdraw from the trial.

Baseline characteristics are presented as means and standard deviations for continuous variables and as counts and percentages for categorical variables. To account for deaths and loss to follow-up, joint models of continuous and survival outcomes were used to assess differences between the intervention group and the control group in the 3-month outcomes (including the primary outcome), with adjustment for baseline measures.23 Differences between the two groups in the rate of rehospitalization (for any cause and for heart failure) and in the number of days of rehospitalization were assessed with the use of joint models similar to those described above, with a Poisson distribution for clinical events based on counts and a negative binomial distribution for days of rehospitalization for any cause to account for overdispersion. Differences between the two groups in the rate of death and in the rate of combined rehospitalization for any cause and death were assessed with the use of generalized linear models, with a Poisson distribution. Differences in proportion-based (binary) clinical measures were analyzed with the use of logistic regression. All the models were adjusted for clinical site, ejection fraction category, age, and sex. The secondary outcome was adjusted for the baseline score on the Short Physical Performance Battery. The potential consistency of intervention effects among prespecified subgroups for the primary outcome was examined with the use of forest plots.

For the primary outcome, a two-tailed P value of less than 0.05 was considered to indicate statistical significance. For all other outcomes, effect-size estimates and 95% confidence intervals are reported without P values. The widths of the confidence intervals were not adjusted for multiple comparisons, so the intervals should not be used to infer definitive treatment effects for the secondary outcome or other outcomes. Analyses were performed with SAS Enterprise Guide, version 7.11, and SAS software, version 9.4.

ResultsTrial Population and Baseline Characteristics

The first patient was enrolled in September 2014, and the last patient was enrolled in September 2019. A total of 27,300 hospital admission records (which included multiple repeat admissions and thus did not represent unique patients) were electronically reviewed. Ultimately, 410 patients were screened for eligibility, and 349 were enrolled; 175 were randomly assigned to the intervention group and 174 to the control group. Data were available for analysis of the primary and secondary outcomes for 87% and 99% of the patients, respectively (Figure 1 and Fig. S2).

Figure 1

Screening and Randomization.

The demographic and clinical characteristics of the patients at baseline are shown in Table 1 and in Table S3. The mean (±SD) age was 72.7±8.1 years, 52% of the patients were women, and 49% were non-White (of whom 94% were Black). The cause of heart failure was ischemic heart disease in 35% of the patients, and 53% had preserved ejection fraction. Patients had high burdens of coexisting conditions, and the incidences of previous hospitalization for any cause and previous hospitalization for heart failure were high (approximately 45% and 25%, respectively). Most (97%) of the patients were assessed as frail or prefrail, according to the modified Fried criteria. Urinary incontinence, falls, and depression were common. The incidence of diabetes mellitus was higher in the intervention group than in the control group (58% vs. 47%). At baseline, the patients were assessed as having severely impaired physical function, poor quality of life, and at least mild cognitive dysfunction (Table 2).

Table 1

Demographic and Clinical Characteristics of the Patients at Baseline.

Table 2

Trial Outcomes.

Follow-up and Trial Outcomes

The last follow-up visit was in March 2020. Among the 175 patients who had been randomly assigned to the rehabilitation intervention, 12 died before completing the intervention, 14 were lost to follow-up for the analysis of the primary outcome, and 16 permanently discontinued the intervention but were included in the analyses of the primary and secondary outcomes. In an analysis that excluded patients who died, patient retention in the intervention group was 82%, and patients completed a mean (±SE) of 24.3±1.0 outpatient intervention sessions; adherence to the sessions was 67±3% (Fig. S3 and Table S4). After adjustment for sessions missed because of medical appointments and illness, adherence to the intervention was 78±3%. Patients generally progressed to higher functional levels in each domain during the course of the intervention (Figure 2). A key goal was to increase each patient’s exercise endurance (duration of walking); among patients who participated in the first and last sessions, the mean (±SD) endurance doubled from 10.7±5.9 minutes in the first session to 22.0±11.1 minutes in the last session. Additional data regarding exercise during the hospitalization, outpatient, and maintenance phases are provided in Tables S5 through S7; information regarding usual-care exercise therapy not associated with the trial is provided in Table S8.

Figure 2

Progression of Physical Function over Time in the Intervention Group, According to Exercise Domain.

After adjustment for the baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001) (Table 2 and Fig. S4). This effect appeared to be relatively uniform across a wide variety of prespecified subgroups (Figure 3). The results for each of the three components of the Short Physical Performance Battery are shown in Table 2. The benefit of the rehabilitation intervention persisted after post hoc adjustment for baseline imbalances in diagnoses of diabetes and peripheral vascular disease; the least-squares mean score was 8.3±0.2 in the intervention group and 6.8±0.2 in the control group (mean between-group difference, 1.5; 95% CI, 0.9 to 2.1).

Figure 3

Prespecified Subgroup Analysis of the Primary Outcome.

The secondary outcome, the rate of rehospitalization for any cause at 6 months, showed no appreciable difference between the intervention group and the control group, with rates of 1.18 and 1.28, respectively (rate ratio, 0.93; 95% CI, 0.66 to 1.19) (Table 2). The exploratory outcomes, including 6-minute walk distance, gait speed, hand-grip strength, frailty status, quality of life, cognition, depression, and clinical events including falls and rehospitalizations, are shown in Table 2 and Figure S5.

There were 21 deaths in the intervention group and 16 deaths in the control group; the rates of death were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27) (Table 2). Among these deaths, 15 in the intervention group and 8 in the control group were from cardiovascular causes (Table S9). Serious and nonserious adverse events are summarized in Tables S10 and S11. Chest pain, hypertension, dizziness, hyperglycemia, and hypoglycemia were more common in the intervention group than in the control group, and falls and heart failure were more common in the control group.

Discussion

The REHAB-HF trial examined the effects of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains in frail, older patients who were hospitalized for acute decompensated heart failure. The intervention group had significantly greater improvement in physical function, as assessed by the score on the Short Physical Performance Battery at 3 months, than the control group. The results of the analyses of 6-minute walk distance, frailty status, quality of life, and depression also suggested clinical benefits of the intervention. Over the course of 6 months, the incidence of rehospitalization for any cause, rehospitalization for heart failure, and death was high in both groups.

Our trial was designed to address several critical evidence gaps regarding physical rehabilitation in patients with heart failure. Most previous trials excluded patients who had been hospitalized within the previous 6 weeks — a period during which the severity of physical dysfunction and the risk of clinical events are highest; those trials also involved few older, frail patients with multiple coexisting conditions in whom different approaches may be appropriate.3,10 In previous early trials of rehabilitation after heart failure, enrollment of the patients and initiation of the intervention began, on average, 7 weeks after hospital discharge; traditional endurance exercise training was commonly used24,25; the enrolled patients were younger and much less frail and diverse than those in our trial25; there was no control group24,26; and the trials were unblinded25,27 and often small, single-center trials.27 One of the largest of such trials involving recently hospitalized patients, EJECTION-HF (Exercise Joins Education: Combined Therapy to Improve Outcomes in Newly-Discharged Heart Failure), showed no benefit of the intervention over usual care with respect to 6-minute walk distance, rehospitalization, and death, but adherence to the intervention was low (43%).25

Physical dysfunction, frailty, and depression are often unrecognized clinically in older patients hospitalized for heart failure,8,28 are generally not addressed in clinical care pathways,10,29 and probably contribute to delayed, incomplete recovery and high rates of rehospitalization, death, and long-term loss of independence after hospital discharge.2,11,12,14 Physical-function impairments in the patients in the REHAB-HF trial were broader and more severe than those observed in patients with chronic heart failure.28 For example, the mean baseline 6-minute walk distance in the REHAB-HF trial was half that observed in the HF-ACTION (Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training) trial,30 and severe leg weakness prevented nearly one third of the patients from standing even once from a seated position without the use of their arms. The patients in our trial also had severe deficits in balance and mobility, and a history of falls and other geriatric conditions was common — findings that are not typically seen in patients with chronic heart failure and are not addressed by conventional cardiac rehabilitation. The initiation of standard endurance exercise training in frail, older patients without first addressing deficits in balance and mobility can limit efficacy31 and increase the risk of injuries and falls.32,33

The intervention-related benefits seen in the REHAB-HF trial generally exceeded previously reported values for the minimal clinically important difference. The mean difference between the groups in the Short Physical Performance Battery score (1.5 points) was three times as large as the reported minimal clinically important difference (0.5 points).21,34 All three of the components of the Short Physical Performance Battery — corresponding to balance, strength, and mobility — showed greater improvement in the intervention group than in the control group. The apparent benefits in 6-minute walk distance (34 m) and KCCQ score (7.1 points) were also larger than the reported minimal clinically important differences (30 m and 5 points, respectively).35,36 The suggested benefit for depression is of interest, since depression is common among patients with heart failure and is associated with frequent rehospitalization,37 and trials targeting depression in patients with heart failure have had neutral results.

The greater improvements in physical function relative to the control group were seen despite the receipt of routine physical or occupational therapy or traditional cardiac or pulmonary rehabilitation as part of usual care by 43% of the patients in the control group. At 6 months, 83% of the patients in the intervention group who were alive and were being followed by telephone contact reported regular home exercise, which suggested that behavioral change — a requisite for long-term adherence — may have occurred.

The intervention-related benefits may be related to both the severity of the baseline deficits and the robust, broad systemic effects of physical exercise, which favorably alters energy metabolism, oxidative stress, inflammation, tissue repair, growth-factor response, and regulatory pathways.38 Older patients with heart failure can have severe skeletal-muscle myopathy that contributes to physical dysfunction and abates with exercise.9

The number of deaths, including deaths from cardiovascular causes, was higher in the intervention group than in the control group, although the numbers and differences were small and may have been due to chance. A meta-analysis of trials of exercise-based rehabilitation for heart failure showed no significant effect on death from any cause during follow-up for up to 12 months.39 In the EJECTION-HF trial, fewer deaths were reported in the intervention group than in the control group at 12 months.25 Given the wide confidence interval for the rate ratio for death from any cause, we cannot rule out the possibility of an increase (or decrease) in risk with an early exercise regimen among some patients.

Our trial has other important limitations. First, the results did not show a beneficial effect on clinical events. However, a study that examined patient preferences in patients with heart failure indicated that improving physical function and maintaining independence are highly valued, independent of clinical events.40 Second, although the staff members who assessed the primary outcome were unaware of the trial-group assignments, it was not possible for patients to be unaware of the group to which they had been randomly assigned. Third, the benefits of the intervention over usual care may have been moderated owing to the usual-care exercise therapy received by the control group. Fourth, differences between the groups in the amount of caregiver attention could have influenced outcomes. Fifth, the long-term durability of the benefit of the intervention is uncertain. Finally, many patients were ineligible or unable or unwilling to participate, and some discontinued the intervention.

Among patients who were hospitalized for acute decompensated heart failure, a transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains and that began during, or early after, hospitalization and continued for 12 weeks after hospital discharge resulted in significantly greater improvement in physical function than usual care.

Notes

This article was published on May 16, 2021, at NEJM.org.

A data sharing statement provided by the authors is available with the full text of this article at NEJM.org.

Supported by research grants from the National Institutes of Health (R01AG045551, R01AG18915, P30AG021332, P30AG028716, and U24AG059624), the Kermit Glenn Phillips II Chair in Cardiovascular Medicine, and the Oristano Family Fund at Wake Forest School of Medicine.

Dr. Kitzman reports receiving steering committee fees and data and safety monitoring board fees from AbbVie, grant support, paid to Wake Forest Baptist Health, and advisory board fees from AstraZeneca and Bayer, advisory board fees from Boehringer Ingelheim and Corvia Medical, consulting fees from Keyto and Merck, and grant support, paid to Wake Forest Baptist Health, and consulting fees from Novartis; Dr. Whellan, receiving grant support, paid to Thomas Jefferson University, from Amgen, end-point committee fees from CVRx and FibroGen, steering committee fees from Cytokinetics, and grant support, paid to Thomas Jefferson University, and advisory board fees from Novo Nordisk; Dr. Duncan, owning stock in Care Directions; and Dr. Mentz, receiving grant support from American Regent and Cytokinetics, consulting fees from Amgen, Boehringer Ingelheim, and Relypsa, research support and honoraria from Abbott, American Regent, Amgen, Boehringer Ingelheim, Eli Lilly, Boston Scientific, Cytokinetics, FAST BioMedical, Gilead Sciences, Innolife, Medtronic, Relypsa, Respicardia, Roche, Sanofi, Vifor Pharma, and Windtree Therapeutics, grant support, research support, and honoraria from AstraZeneca, and research support, consulting fees, and honoraria from Bayer, Merck, and Novartis. No other potential conflict of interest relevant to this article was reported.

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

We thank Alain Bertoni, M.D., Peter Brubaker, Ph.D., and Jack Rejeski, Ph.D. for their critical review and editing of an earlier version of the manuscript; and Timothy Morgan, Ph.D. (deceased), for his biostatistical expertise.

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Authors: Dalane W. Kitzman, M.D., David J. Whellan, M.D., M.H.S., Pamela Duncan, P.T., Ph.D., Amy M. Pastva, P.T., Ph.D., Robert J. Mentz, M.D., Gordon R. Reeves, M.D., M.P.T., M. Benjamin Nelson, M.S. https://orcid.org/0000-0002-1701-6071, +5 , and Christopher M. O’Connor, M.D.Author Info & Affiliations

Published May 16, 2021

N Engl J Med 2021;385:203-216

DOI: 10.1056/NEJMoa2026141

VOL. 385 NO. 3

Copyright © 2021

• Abstract
• Methods
• Results
• Discussion
• Notes
• Supplementary Material
• References

• Information & Authors
• Metrics & Citations
• View Options
• References
• Media
• Tables
• Share

AbstractBackground

Older patients who are hospitalized for acute decompensated heart failure have high rates of physical frailty, poor quality of life, delayed recovery, and frequent rehospitalizations. Interventions to address physical frailty in this population are not well established.

Methods

We conducted a multicenter, randomized, controlled trial to eval‎uate a transitional, tailored, progressive rehabilitation intervention that included four physical-function domains (strength, balance, mobility, and endurance). The intervention was initiated during, or early after, hospitalization for heart failure and was continued after discharge for 36 outpatient sessions. The primary outcome was the score on the Short Physical Performance Battery (total scores range from 0 to 12, with lower scores indicating more severe physical dysfunction) at 3 months. The secondary outcome was the 6-month rate of rehospitalization for any cause.

Research Summary

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Results

A total of 349 patients underwent randomization; 175 were assigned to the rehabilitation intervention and 174 to usual care (control). At baseline, patients in each group had markedly impaired physical function, and 97% were frail or prefrail; the mean number of coexisting conditions was five in each group. Patient retention in the intervention group was 82%, and adherence to the intervention sessions was 67%. After adjustment for baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001). At 6 months, the rates of rehospitalization for any cause were 1.18 in the intervention group and 1.28 in the control group (rate ratio, 0.93; 95% CI, 0.66 to 1.19). There were 21 deaths (15 from cardiovascular causes) in the intervention group and 16 deaths (8 from cardiovascular causes) in the control group. The rates of death from any cause were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27).

Conclusions

In a diverse population of older patients who were hospitalized for acute decompensated heart failure, an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains resulted in greater improvement in physical function than usual care. (Funded by the National Institutes of Health and others; REHAB-HF ClinicalTrials.gov number, NCT02196038.)

Quick Take

Physical Rehabilitation for Heart Failure

1m 56s

Acute decompensated heart failure is the leading cause of hospitalization among older persons in the United States1 and is associated with poor health-related quality of life, frequent rehospitalizations, high mortality, and costs exceeding $39 billion per year.1,2 Most intervention trials in acute decompensated heart failure have had neutral results, which suggests that outcomes may be driven in part by mechanisms that have been overlooked.3-5

Among older patients with acute heart failure, physical function is markedly impaired, and frailty rates and the burden of coexisting conditions are high.5-8 Even among older patients with stable and well-compensated heart failure, severe impairments in physical function are often present owing to the combined effects of aging, cardiovascular dysfunction, and skeletal-muscle dysfunction.9,10 As patients with chronic heart failure transition to acute decompensated heart failure, physical function worsens further, and this decline is exacerbated by hospitalization and bed rest.8 These deficits often persist. Many patients never recover baseline function, lose independence, and have high risks of rehospitalization and death after discharge (sometimes referred to as “post-hospital syndrome”).4,5,11–14

However, management guidelines do not address physical dysfunction in patients hospitalized for heart failure,15 and previous exercise training trials excluded patients with heart failure who had recently been hospitalized.10,16 To address these issues, we conducted the Rehabilitation Therapy in Older Acute Heart Failure Patients (REHAB-HF) trial, a multicenter, randomized, single-blind, controlled trial of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains. We hypothesized that the intervention would improve physical function and reduce rates of rehospitalization for any cause at 6 months.

MethodsTrial Design and Oversight

Details of the trial design and intervention methods have been described previously.17,18 The organizational structure is shown in Figure S1 in the Supplementary Appendix, available with the full text of this article at NEJM.org. The steering committee designed the trial and oversaw operations. The protocol, which is available at NEJM.org, was approved by the institutional review board at each site. An independent data and safety monitoring committee eval‎uated patient safety. The authors vouch for the accuracy and completeness of the data and analyses and for the fidelity of the trial to the protocol.

Patients and Randomization

Patients were screened at the time of hospital admission and were enrolled before discharge. Patients were eligible for participation if they were 60 years of age or older, if they had been admitted for acute decompensated heart failure regardless of ejection fraction, if they could walk at least 4 m at enrollment (with or without the aid of an assistive device), if they were functionally independent before admission, and if they were expected to be discharged home. Full details of the inclusion and exclusion criteria are provided in the Supplementary Appendix.17 After eligible patients provided written informed consent and completed baseline testing, they were randomly assigned with equal probability to the rehabilitation intervention (intervention group) or to usual care (control group) by a centralized, Web-based system, with the use of block randomization. Randomization was stratified according to ejection fraction (<45% vs. ≥45%) and clinical site. The patients in both trial groups received usual care, as recommended by their medical providers, which could include inpatient or outpatient physical therapy and standard cardiac rehabilitation.

Trial Procedures

The trial intervention (for the intervention group) was an early, transitional, tailored, progressive physical rehabilitation program that had been developed for frail, older patients with acute decompensated heart failure.18,19 The intervention focused on four physical-function domains (strength, balance, mobility, and endurance) and progressed through four prespecified functional levels within each domain (Table S1). The progression of exercise intensity and the types of exercises at each session were individualized on the basis of the patient’s performance level within each domain.18 A key goal was to increase each patient’s endurance (duration of walking); doing this safely required first addressing deficits in balance, strength, and mobility.

The intervention was initiated in the hospital when feasible and was subsequently transitioned to an outpatient facility as soon as possible after discharge. If needed, home-based sessions were provided by interventionists until the patient was physically able to attend the facility-based outpatient sessions. Outpatient sessions were 60 minutes long, occurred 3 days per week for 12 weeks (or 36 sessions), and were conducted at a 1:1 interventionist–patient ratio. Outpatient sessions were complemented by home exercise (low-intensity walking, which was gradually increased to up to 30 minutes daily, and strengthening exercises) on nonprogram days. The home exercise component of the intervention was initiated only after a visit to the patient’s home by an interventionist to eval‎uate the home environment.18

A key goal of the intervention during the first 3 months (the outpatient phase) was to prepare the patient to transition to the independent maintenance phase (months 4 through 6). At the 3-month visit, patients were provided with individualized exercise prescriptions and were subsequently followed every 4 weeks by telephone contact. Patient retention in the intervention group and adherence to the intervention sessions were reviewed and discussed every 2 weeks by a dedicated committee in accordance with the recommendations of the National Institutes of Health Behavior Change Consortium Treatment Fidelity Workgroup.18 Additional details regarding the intervention are provided in the Supplementary Appendix.

Patients who had been randomly assigned to the control group received a telephone call every 2 weeks and had in-person clinic visits at 1 month and 3 months after discharge from the index hospitalization.17 Information regarding the occurrence of symptoms or clinical events and the receipt of rehabilitation therapy unrelated to the trial was collected. Patients received no specific recommendations with respect to exercise, but they were encouraged to adhere to prescribed usual-care therapy and follow-up appointments. Additional details regarding the control group are provided in the Supplementary Appendix.

Trial Outcomes

Outcome measures of physical and cognitive function were assessed by personnel who were unaware of the trial-group assignments. Physical function, quality of life, depression, and cognitive function were assessed at baseline in the hospital and at 3 months.17 Clinical events were ascertained throughout follow-up from monthly interviews and from review of medical records.

The primary outcome was the score on the Short Physical Performance Battery at 3 months. The Short Physical Performance Battery is a standardized, reproducible measure of global physical function that has been validated in frail, older persons and predicts a wide range of clinical outcomes.20-22 It has three components: a standing balance test, a gait-speed (4-m walk) test, and a strength test (as assessed by the time needed to rise from a chair five times). Each component is scored on a scale of 0 to 4; the sum of the scores ranges from 0 to 12, with lower scores indicating more severe physical dysfunction.

The secondary outcome was the rate of rehospitalization for any cause at 6 months, with rehospitalization defined as any hospital stay longer than 24 hours. The reasons for rehospitalization were categorized as noncardiovascular cause, heart failure, or another cardiovascular cause by an independent adjudicator who was unaware of the trial-group assignments.

Additional physical-function outcomes included 6-minute walk distance, frailty status (assessed according to modified Fried criteria6), hand-grip strength, and gait speed at 3 months. Quality of life was assessed at 3 months with the use of the Kansas City Cardiomyopathy Questionnaire (KCCQ) and the EQ-5D-5L (also known as the European Quality of Life 5-Dimension 5-Level questionnaire) visual-analogue scale. Other outcomes included the Geriatric Depression Scale–15 score and the Montreal Cognitive Assessment score.

Statistical Analysis

On the basis of the results from a pilot study,19 we estimated that 258 patients who could be eval‎uated for efficacy would provide the trial with 80% power to detect a 10% difference (equivalent to a difference of 0.6 points) between the intervention group and the control group in the score on the Short Physical Performance Battery at 3 months (the primary outcome); the enrollment of 334 patients who could be eval‎uated for efficacy would be needed to detect a 25% difference in the rate of rehospitalization at 6 months (the secondary outcome). We planned to enroll 360 patients in order to allow for approximately 7% of the patients to withdraw from the trial.

Baseline characteristics are presented as means and standard deviations for continuous variables and as counts and percentages for categorical variables. To account for deaths and loss to follow-up, joint models of continuous and survival outcomes were used to assess differences between the intervention group and the control group in the 3-month outcomes (including the primary outcome), with adjustment for baseline measures.23 Differences between the two groups in the rate of rehospitalization (for any cause and for heart failure) and in the number of days of rehospitalization were assessed with the use of joint models similar to those described above, with a Poisson distribution for clinical events based on counts and a negative binomial distribution for days of rehospitalization for any cause to account for overdispersion. Differences between the two groups in the rate of death and in the rate of combined rehospitalization for any cause and death were assessed with the use of generalized linear models, with a Poisson distribution. Differences in proportion-based (binary) clinical measures were analyzed with the use of logistic regression. All the models were adjusted for clinical site, ejection fraction category, age, and sex. The secondary outcome was adjusted for the baseline score on the Short Physical Performance Battery. The potential consistency of intervention effects among prespecified subgroups for the primary outcome was examined with the use of forest plots.

For the primary outcome, a two-tailed P value of less than 0.05 was considered to indicate statistical significance. For all other outcomes, effect-size estimates and 95% confidence intervals are reported without P values. The widths of the confidence intervals were not adjusted for multiple comparisons, so the intervals should not be used to infer definitive treatment effects for the secondary outcome or other outcomes. Analyses were performed with SAS Enterprise Guide, version 7.11, and SAS software, version 9.4.

ResultsTrial Population and Baseline Characteristics

The first patient was enrolled in September 2014, and the last patient was enrolled in September 2019. A total of 27,300 hospital admission records (which included multiple repeat admissions and thus did not represent unique patients) were electronically reviewed. Ultimately, 410 patients were screened for eligibility, and 349 were enrolled; 175 were randomly assigned to the intervention group and 174 to the control group. Data were available for analysis of the primary and secondary outcomes for 87% and 99% of the patients, respectively (Figure 1 and Fig. S2).

Figure 1

Screening and Randomization.

The demographic and clinical characteristics of the patients at baseline are shown in Table 1 and in Table S3. The mean (±SD) age was 72.7±8.1 years, 52% of the patients were women, and 49% were non-White (of whom 94% were Black). The cause of heart failure was ischemic heart disease in 35% of the patients, and 53% had preserved ejection fraction. Patients had high burdens of coexisting conditions, and the incidences of previous hospitalization for any cause and previous hospitalization for heart failure were high (approximately 45% and 25%, respectively). Most (97%) of the patients were assessed as frail or prefrail, according to the modified Fried criteria. Urinary incontinence, falls, and depression were common. The incidence of diabetes mellitus was higher in the intervention group than in the control group (58% vs. 47%). At baseline, the patients were assessed as having severely impaired physical function, poor quality of life, and at least mild cognitive dysfunction (Table 2).

Table 1

Demographic and Clinical Characteristics of the Patients at Baseline.

Table 2

Trial Outcomes.

Follow-up and Trial Outcomes

The last follow-up visit was in March 2020. Among the 175 patients who had been randomly assigned to the rehabilitation intervention, 12 died before completing the intervention, 14 were lost to follow-up for the analysis of the primary outcome, and 16 permanently discontinued the intervention but were included in the analyses of the primary and secondary outcomes. In an analysis that excluded patients who died, patient retention in the intervention group was 82%, and patients completed a mean (±SE) of 24.3±1.0 outpatient intervention sessions; adherence to the sessions was 67±3% (Fig. S3 and Table S4). After adjustment for sessions missed because of medical appointments and illness, adherence to the intervention was 78±3%. Patients generally progressed to higher functional levels in each domain during the course of the intervention (Figure 2). A key goal was to increase each patient’s exercise endurance (duration of walking); among patients who participated in the first and last sessions, the mean (±SD) endurance doubled from 10.7±5.9 minutes in the first session to 22.0±11.1 minutes in the last session. Additional data regarding exercise during the hospitalization, outpatient, and maintenance phases are provided in Tables S5 through S7; information regarding usual-care exercise therapy not associated with the trial is provided in Table S8.

Figure 2

Progression of Physical Function over Time in the Intervention Group, According to Exercise Domain.

After adjustment for the baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001) (Table 2 and Fig. S4). This effect appeared to be relatively uniform across a wide variety of prespecified subgroups (Figure 3). The results for each of the three components of the Short Physical Performance Battery are shown in Table 2. The benefit of the rehabilitation intervention persisted after post hoc adjustment for baseline imbalances in diagnoses of diabetes and peripheral vascular disease; the least-squares mean score was 8.3±0.2 in the intervention group and 6.8±0.2 in the control group (mean between-group difference, 1.5; 95% CI, 0.9 to 2.1).

Figure 3

Prespecified Subgroup Analysis of the Primary Outcome.

The secondary outcome, the rate of rehospitalization for any cause at 6 months, showed no appreciable difference between the intervention group and the control group, with rates of 1.18 and 1.28, respectively (rate ratio, 0.93; 95% CI, 0.66 to 1.19) (Table 2). The exploratory outcomes, including 6-minute walk distance, gait speed, hand-grip strength, frailty status, quality of life, cognition, depression, and clinical events including falls and rehospitalizations, are shown in Table 2 and Figure S5.

There were 21 deaths in the intervention group and 16 deaths in the control group; the rates of death were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27) (Table 2). Among these deaths, 15 in the intervention group and 8 in the control group were from cardiovascular causes (Table S9). Serious and nonserious adverse events are summarized in Tables S10 and S11. Chest pain, hypertension, dizziness, hyperglycemia, and hypoglycemia were more common in the intervention group than in the control group, and falls and heart failure were more common in the control group.

Discussion

The REHAB-HF trial examined the effects of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains in frail, older patients who were hospitalized for acute decompensated heart failure. The intervention group had significantly greater improvement in physical function, as assessed by the score on the Short Physical Performance Battery at 3 months, than the control group. The results of the analyses of 6-minute walk distance, frailty status, quality of life, and depression also suggested clinical benefits of the intervention. Over the course of 6 months, the incidence of rehospitalization for any cause, rehospitalization for heart failure, and death was high in both groups.

Our trial was designed to address several critical evidence gaps regarding physical rehabilitation in patients with heart failure. Most previous trials excluded patients who had been hospitalized within the previous 6 weeks — a period during which the severity of physical dysfunction and the risk of clinical events are highest; those trials also involved few older, frail patients with multiple coexisting conditions in whom different approaches may be appropriate.3,10 In previous early trials of rehabilitation after heart failure, enrollment of the patients and initiation of the intervention began, on average, 7 weeks after hospital discharge; traditional endurance exercise training was commonly used24,25; the enrolled patients were younger and much less frail and diverse than those in our trial25; there was no control group24,26; and the trials were unblinded25,27 and often small, single-center trials.27 One of the largest of such trials involving recently hospitalized patients, EJECTION-HF (Exercise Joins Education: Combined Therapy to Improve Outcomes in Newly-Discharged Heart Failure), showed no benefit of the intervention over usual care with respect to 6-minute walk distance, rehospitalization, and death, but adherence to the intervention was low (43%).25

Physical dysfunction, frailty, and depression are often unrecognized clinically in older patients hospitalized for heart failure,8,28 are generally not addressed in clinical care pathways,10,29 and probably contribute to delayed, incomplete recovery and high rates of rehospitalization, death, and long-term loss of independence after hospital discharge.2,11,12,14 Physical-function impairments in the patients in the REHAB-HF trial were broader and more severe than those observed in patients with chronic heart failure.28 For example, the mean baseline 6-minute walk distance in the REHAB-HF trial was half that observed in the HF-ACTION (Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training) trial,30 and severe leg weakness prevented nearly one third of the patients from standing even once from a seated position without the use of their arms. The patients in our trial also had severe deficits in balance and mobility, and a history of falls and other geriatric conditions was common — findings that are not typically seen in patients with chronic heart failure and are not addressed by conventional cardiac rehabilitation. The initiation of standard endurance exercise training in frail, older patients without first addressing deficits in balance and mobility can limit efficacy31 and increase the risk of injuries and falls.32,33

The intervention-related benefits seen in the REHAB-HF trial generally exceeded previously reported values for the minimal clinically important difference. The mean difference between the groups in the Short Physical Performance Battery score (1.5 points) was three times as large as the reported minimal clinically important difference (0.5 points).21,34 All three of the components of the Short Physical Performance Battery — corresponding to balance, strength, and mobility — showed greater improvement in the intervention group than in the control group. The apparent benefits in 6-minute walk distance (34 m) and KCCQ score (7.1 points) were also larger than the reported minimal clinically important differences (30 m and 5 points, respectively).35,36 The suggested benefit for depression is of interest, since depression is common among patients with heart failure and is associated with frequent rehospitalization,37 and trials targeting depression in patients with heart failure have had neutral results.

The greater improvements in physical function relative to the control group were seen despite the receipt of routine physical or occupational therapy or traditional cardiac or pulmonary rehabilitation as part of usual care by 43% of the patients in the control group. At 6 months, 83% of the patients in the intervention group who were alive and were being followed by telephone contact reported regular home exercise, which suggested that behavioral change — a requisite for long-term adherence — may have occurred.

The intervention-related benefits may be related to both the severity of the baseline deficits and the robust, broad systemic effects of physical exercise, which favorably alters energy metabolism, oxidative stress, inflammation, tissue repair, growth-factor response, and regulatory pathways.38 Older patients with heart failure can have severe skeletal-muscle myopathy that contributes to physical dysfunction and abates with exercise.9

The number of deaths, including deaths from cardiovascular causes, was higher in the intervention group than in the control group, although the numbers and differences were small and may have been due to chance. A meta-analysis of trials of exercise-based rehabilitation for heart failure showed no significant effect on death from any cause during follow-up for up to 12 months.39 In the EJECTION-HF trial, fewer deaths were reported in the intervention group than in the control group at 12 months.25 Given the wide confidence interval for the rate ratio for death from any cause, we cannot rule out the possibility of an increase (or decrease) in risk with an early exercise regimen among some patients.

Our trial has other important limitations. First, the results did not show a beneficial effect on clinical events. However, a study that examined patient preferences in patients with heart failure indicated that improving physical function and maintaining independence are highly valued, independent of clinical events.40 Second, although the staff members who assessed the primary outcome were unaware of the trial-group assignments, it was not possible for patients to be unaware of the group to which they had been randomly assigned. Third, the benefits of the intervention over usual care may have been moderated owing to the usual-care exercise therapy received by the control group. Fourth, differences between the groups in the amount of caregiver attention could have influenced outcomes. Fifth, the long-term durability of the benefit of the intervention is uncertain. Finally, many patients were ineligible or unable or unwilling to participate, and some discontinued the intervention.

Among patients who were hospitalized for acute decompensated heart failure, a transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains and that began during, or early after, hospitalization and continued for 12 weeks after hospital discharge resulted in significantly greater improvement in physical function than usual care.

Notes

This article was published on May 16, 2021, at NEJM.org.

A data sharing statement provided by the authors is available with the full text of this article at NEJM.org.

Supported by research grants from the National Institutes of Health (R01AG045551, R01AG18915, P30AG021332, P30AG028716, and U24AG059624), the Kermit Glenn Phillips II Chair in Cardiovascular Medicine, and the Oristano Family Fund at Wake Forest School of Medicine.

Dr. Kitzman reports receiving steering committee fees and data and safety monitoring board fees from AbbVie, grant support, paid to Wake Forest Baptist Health, and advisory board fees from AstraZeneca and Bayer, advisory board fees from Boehringer Ingelheim and Corvia Medical, consulting fees from Keyto and Merck, and grant support, paid to Wake Forest Baptist Health, and consulting fees from Novartis; Dr. Whellan, receiving grant support, paid to Thomas Jefferson University, from Amgen, end-point committee fees from CVRx and FibroGen, steering committee fees from Cytokinetics, and grant support, paid to Thomas Jefferson University, and advisory board fees from Novo Nordisk; Dr. Duncan, owning stock in Care Directions; and Dr. Mentz, receiving grant support from American Regent and Cytokinetics, consulting fees from Amgen, Boehringer Ingelheim, and Relypsa, research support and honoraria from Abbott, American Regent, Amgen, Boehringer Ingelheim, Eli Lilly, Boston Scientific, Cytokinetics, FAST BioMedical, Gilead Sciences, Innolife, Medtronic, Relypsa, Respicardia, Roche, Sanofi, Vifor Pharma, and Windtree Therapeutics, grant support, research support, and honoraria from AstraZeneca, and research support, consulting fees, and honoraria from Bayer, Merck, and Novartis. No other potential conflict of interest relevant to this article was reported.

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

We thank Alain Bertoni, M.D., Peter Brubaker, Ph.D., and Jack Rejeski, Ph.D. for their critical review and editing of an earlier version of the manuscript; and Timothy Morgan, Ph.D. (deceased), for his biostatistical expertise.

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Authors: Dalane W. Kitzman, M.D., David J. Whellan, M.D., M.H.S., Pamela Duncan, P.T., Ph.D., Amy M. Pastva, P.T., Ph.D., Robert J. Mentz, M.D., Gordon R. Reeves, M.D., M.P.T., M. Benjamin Nelson, M.S. https://orcid.org/0000-0002-1701-6071, +5 , and Christopher M. O’Connor, M.D.Author Info & Affiliations

Published May 16, 2021

N Engl J Med 2021;385:203-216

DOI: 10.1056/NEJMoa2026141

VOL. 385 NO. 3

Copyright © 2021

• Abstract
• Methods
• Results
• Discussion
• Notes
• Supplementary Material
• References

• Information & Authors
• Metrics & Citations
• View Options
• References
• Media
• Tables
• Share

AbstractBackground

Older patients who are hospitalized for acute decompensated heart failure have high rates of physical frailty, poor quality of life, delayed recovery, and frequent rehospitalizations. Interventions to address physical frailty in this population are not well established.

Methods

We conducted a multicenter, randomized, controlled trial to eval‎uate a transitional, tailored, progressive rehabilitation intervention that included four physical-function domains (strength, balance, mobility, and endurance). The intervention was initiated during, or early after, hospitalization for heart failure and was continued after discharge for 36 outpatient sessions. The primary outcome was the score on the Short Physical Performance Battery (total scores range from 0 to 12, with lower scores indicating more severe physical dysfunction) at 3 months. The secondary outcome was the 6-month rate of rehospitalization for any cause.

Research Summary

Physical Rehabilitation for Older Patients Hospitalized for Heart Failure

Results

A total of 349 patients underwent randomization; 175 were assigned to the rehabilitation intervention and 174 to usual care (control). At baseline, patients in each group had markedly impaired physical function, and 97% were frail or prefrail; the mean number of coexisting conditions was five in each group. Patient retention in the intervention group was 82%, and adherence to the intervention sessions was 67%. After adjustment for baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001). At 6 months, the rates of rehospitalization for any cause were 1.18 in the intervention group and 1.28 in the control group (rate ratio, 0.93; 95% CI, 0.66 to 1.19). There were 21 deaths (15 from cardiovascular causes) in the intervention group and 16 deaths (8 from cardiovascular causes) in the control group. The rates of death from any cause were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27).

Conclusions

In a diverse population of older patients who were hospitalized for acute decompensated heart failure, an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains resulted in greater improvement in physical function than usual care. (Funded by the National Institutes of Health and others; REHAB-HF ClinicalTrials.gov number, NCT02196038.)

Quick Take

Physical Rehabilitation for Heart Failure

1m 56s

Acute decompensated heart failure is the leading cause of hospitalization among older persons in the United States1 and is associated with poor health-related quality of life, frequent rehospitalizations, high mortality, and costs exceeding $39 billion per year.1,2 Most intervention trials in acute decompensated heart failure have had neutral results, which suggests that outcomes may be driven in part by mechanisms that have been overlooked.3-5

Among older patients with acute heart failure, physical function is markedly impaired, and frailty rates and the burden of coexisting conditions are high.5-8 Even among older patients with stable and well-compensated heart failure, severe impairments in physical function are often present owing to the combined effects of aging, cardiovascular dysfunction, and skeletal-muscle dysfunction.9,10 As patients with chronic heart failure transition to acute decompensated heart failure, physical function worsens further, and this decline is exacerbated by hospitalization and bed rest.8 These deficits often persist. Many patients never recover baseline function, lose independence, and have high risks of rehospitalization and death after discharge (sometimes referred to as “post-hospital syndrome”).4,5,11–14

However, management guidelines do not address physical dysfunction in patients hospitalized for heart failure,15 and previous exercise training trials excluded patients with heart failure who had recently been hospitalized.10,16 To address these issues, we conducted the Rehabilitation Therapy in Older Acute Heart Failure Patients (REHAB-HF) trial, a multicenter, randomized, single-blind, controlled trial of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains. We hypothesized that the intervention would improve physical function and reduce rates of rehospitalization for any cause at 6 months.

MethodsTrial Design and Oversight

Details of the trial design and intervention methods have been described previously.17,18 The organizational structure is shown in Figure S1 in the Supplementary Appendix, available with the full text of this article at NEJM.org. The steering committee designed the trial and oversaw operations. The protocol, which is available at NEJM.org, was approved by the institutional review board at each site. An independent data and safety monitoring committee eval‎uated patient safety. The authors vouch for the accuracy and completeness of the data and analyses and for the fidelity of the trial to the protocol.

Patients and Randomization

Patients were screened at the time of hospital admission and were enrolled before discharge. Patients were eligible for participation if they were 60 years of age or older, if they had been admitted for acute decompensated heart failure regardless of ejection fraction, if they could walk at least 4 m at enrollment (with or without the aid of an assistive device), if they were functionally independent before admission, and if they were expected to be discharged home. Full details of the inclusion and exclusion criteria are provided in the Supplementary Appendix.17 After eligible patients provided written informed consent and completed baseline testing, they were randomly assigned with equal probability to the rehabilitation intervention (intervention group) or to usual care (control group) by a centralized, Web-based system, with the use of block randomization. Randomization was stratified according to ejection fraction (<45% vs. ≥45%) and clinical site. The patients in both trial groups received usual care, as recommended by their medical providers, which could include inpatient or outpatient physical therapy and standard cardiac rehabilitation.

Trial Procedures

The trial intervention (for the intervention group) was an early, transitional, tailored, progressive physical rehabilitation program that had been developed for frail, older patients with acute decompensated heart failure.18,19 The intervention focused on four physical-function domains (strength, balance, mobility, and endurance) and progressed through four prespecified functional levels within each domain (Table S1). The progression of exercise intensity and the types of exercises at each session were individualized on the basis of the patient’s performance level within each domain.18 A key goal was to increase each patient’s endurance (duration of walking); doing this safely required first addressing deficits in balance, strength, and mobility.

The intervention was initiated in the hospital when feasible and was subsequently transitioned to an outpatient facility as soon as possible after discharge. If needed, home-based sessions were provided by interventionists until the patient was physically able to attend the facility-based outpatient sessions. Outpatient sessions were 60 minutes long, occurred 3 days per week for 12 weeks (or 36 sessions), and were conducted at a 1:1 interventionist–patient ratio. Outpatient sessions were complemented by home exercise (low-intensity walking, which was gradually increased to up to 30 minutes daily, and strengthening exercises) on nonprogram days. The home exercise component of the intervention was initiated only after a visit to the patient’s home by an interventionist to eval‎uate the home environment.18

A key goal of the intervention during the first 3 months (the outpatient phase) was to prepare the patient to transition to the independent maintenance phase (months 4 through 6). At the 3-month visit, patients were provided with individualized exercise prescriptions and were subsequently followed every 4 weeks by telephone contact. Patient retention in the intervention group and adherence to the intervention sessions were reviewed and discussed every 2 weeks by a dedicated committee in accordance with the recommendations of the National Institutes of Health Behavior Change Consortium Treatment Fidelity Workgroup.18 Additional details regarding the intervention are provided in the Supplementary Appendix.

Patients who had been randomly assigned to the control group received a telephone call every 2 weeks and had in-person clinic visits at 1 month and 3 months after discharge from the index hospitalization.17 Information regarding the occurrence of symptoms or clinical events and the receipt of rehabilitation therapy unrelated to the trial was collected. Patients received no specific recommendations with respect to exercise, but they were encouraged to adhere to prescribed usual-care therapy and follow-up appointments. Additional details regarding the control group are provided in the Supplementary Appendix.

Trial Outcomes

Outcome measures of physical and cognitive function were assessed by personnel who were unaware of the trial-group assignments. Physical function, quality of life, depression, and cognitive function were assessed at baseline in the hospital and at 3 months.17 Clinical events were ascertained throughout follow-up from monthly interviews and from review of medical records.

The primary outcome was the score on the Short Physical Performance Battery at 3 months. The Short Physical Performance Battery is a standardized, reproducible measure of global physical function that has been validated in frail, older persons and predicts a wide range of clinical outcomes.20-22 It has three components: a standing balance test, a gait-speed (4-m walk) test, and a strength test (as assessed by the time needed to rise from a chair five times). Each component is scored on a scale of 0 to 4; the sum of the scores ranges from 0 to 12, with lower scores indicating more severe physical dysfunction.

The secondary outcome was the rate of rehospitalization for any cause at 6 months, with rehospitalization defined as any hospital stay longer than 24 hours. The reasons for rehospitalization were categorized as noncardiovascular cause, heart failure, or another cardiovascular cause by an independent adjudicator who was unaware of the trial-group assignments.

Additional physical-function outcomes included 6-minute walk distance, frailty status (assessed according to modified Fried criteria6), hand-grip strength, and gait speed at 3 months. Quality of life was assessed at 3 months with the use of the Kansas City Cardiomyopathy Questionnaire (KCCQ) and the EQ-5D-5L (also known as the European Quality of Life 5-Dimension 5-Level questionnaire) visual-analogue scale. Other outcomes included the Geriatric Depression Scale–15 score and the Montreal Cognitive Assessment score.

Statistical Analysis

On the basis of the results from a pilot study,19 we estimated that 258 patients who could be eval‎uated for efficacy would provide the trial with 80% power to detect a 10% difference (equivalent to a difference of 0.6 points) between the intervention group and the control group in the score on the Short Physical Performance Battery at 3 months (the primary outcome); the enrollment of 334 patients who could be eval‎uated for efficacy would be needed to detect a 25% difference in the rate of rehospitalization at 6 months (the secondary outcome). We planned to enroll 360 patients in order to allow for approximately 7% of the patients to withdraw from the trial.

Baseline characteristics are presented as means and standard deviations for continuous variables and as counts and percentages for categorical variables. To account for deaths and loss to follow-up, joint models of continuous and survival outcomes were used to assess differences between the intervention group and the control group in the 3-month outcomes (including the primary outcome), with adjustment for baseline measures.23 Differences between the two groups in the rate of rehospitalization (for any cause and for heart failure) and in the number of days of rehospitalization were assessed with the use of joint models similar to those described above, with a Poisson distribution for clinical events based on counts and a negative binomial distribution for days of rehospitalization for any cause to account for overdispersion. Differences between the two groups in the rate of death and in the rate of combined rehospitalization for any cause and death were assessed with the use of generalized linear models, with a Poisson distribution. Differences in proportion-based (binary) clinical measures were analyzed with the use of logistic regression. All the models were adjusted for clinical site, ejection fraction category, age, and sex. The secondary outcome was adjusted for the baseline score on the Short Physical Performance Battery. The potential consistency of intervention effects among prespecified subgroups for the primary outcome was examined with the use of forest plots.

For the primary outcome, a two-tailed P value of less than 0.05 was considered to indicate statistical significance. For all other outcomes, effect-size estimates and 95% confidence intervals are reported without P values. The widths of the confidence intervals were not adjusted for multiple comparisons, so the intervals should not be used to infer definitive treatment effects for the secondary outcome or other outcomes. Analyses were performed with SAS Enterprise Guide, version 7.11, and SAS software, version 9.4.

ResultsTrial Population and Baseline Characteristics

The first patient was enrolled in September 2014, and the last patient was enrolled in September 2019. A total of 27,300 hospital admission records (which included multiple repeat admissions and thus did not represent unique patients) were electronically reviewed. Ultimately, 410 patients were screened for eligibility, and 349 were enrolled; 175 were randomly assigned to the intervention group and 174 to the control group. Data were available for analysis of the primary and secondary outcomes for 87% and 99% of the patients, respectively (Figure 1 and Fig. S2).

Figure 1

Screening and Randomization.

The demographic and clinical characteristics of the patients at baseline are shown in Table 1 and in Table S3. The mean (±SD) age was 72.7±8.1 years, 52% of the patients were women, and 49% were non-White (of whom 94% were Black). The cause of heart failure was ischemic heart disease in 35% of the patients, and 53% had preserved ejection fraction. Patients had high burdens of coexisting conditions, and the incidences of previous hospitalization for any cause and previous hospitalization for heart failure were high (approximately 45% and 25%, respectively). Most (97%) of the patients were assessed as frail or prefrail, according to the modified Fried criteria. Urinary incontinence, falls, and depression were common. The incidence of diabetes mellitus was higher in the intervention group than in the control group (58% vs. 47%). At baseline, the patients were assessed as having severely impaired physical function, poor quality of life, and at least mild cognitive dysfunction (Table 2).

Table 1

Demographic and Clinical Characteristics of the Patients at Baseline.

Table 2

Trial Outcomes.

Follow-up and Trial Outcomes

The last follow-up visit was in March 2020. Among the 175 patients who had been randomly assigned to the rehabilitation intervention, 12 died before completing the intervention, 14 were lost to follow-up for the analysis of the primary outcome, and 16 permanently discontinued the intervention but were included in the analyses of the primary and secondary outcomes. In an analysis that excluded patients who died, patient retention in the intervention group was 82%, and patients completed a mean (±SE) of 24.3±1.0 outpatient intervention sessions; adherence to the sessions was 67±3% (Fig. S3 and Table S4). After adjustment for sessions missed because of medical appointments and illness, adherence to the intervention was 78±3%. Patients generally progressed to higher functional levels in each domain during the course of the intervention (Figure 2). A key goal was to increase each patient’s exercise endurance (duration of walking); among patients who participated in the first and last sessions, the mean (±SD) endurance doubled from 10.7±5.9 minutes in the first session to 22.0±11.1 minutes in the last session. Additional data regarding exercise during the hospitalization, outpatient, and maintenance phases are provided in Tables S5 through S7; information regarding usual-care exercise therapy not associated with the trial is provided in Table S8.

Figure 2

Progression of Physical Function over Time in the Intervention Group, According to Exercise Domain.

After adjustment for the baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001) (Table 2 and Fig. S4). This effect appeared to be relatively uniform across a wide variety of prespecified subgroups (Figure 3). The results for each of the three components of the Short Physical Performance Battery are shown in Table 2. The benefit of the rehabilitation intervention persisted after post hoc adjustment for baseline imbalances in diagnoses of diabetes and peripheral vascular disease; the least-squares mean score was 8.3±0.2 in the intervention group and 6.8±0.2 in the control group (mean between-group difference, 1.5; 95% CI, 0.9 to 2.1).

Figure 3

Prespecified Subgroup Analysis of the Primary Outcome.

The secondary outcome, the rate of rehospitalization for any cause at 6 months, showed no appreciable difference between the intervention group and the control group, with rates of 1.18 and 1.28, respectively (rate ratio, 0.93; 95% CI, 0.66 to 1.19) (Table 2). The exploratory outcomes, including 6-minute walk distance, gait speed, hand-grip strength, frailty status, quality of life, cognition, depression, and clinical events including falls and rehospitalizations, are shown in Table 2 and Figure S5.

There were 21 deaths in the intervention group and 16 deaths in the control group; the rates of death were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27) (Table 2). Among these deaths, 15 in the intervention group and 8 in the control group were from cardiovascular causes (Table S9). Serious and nonserious adverse events are summarized in Tables S10 and S11. Chest pain, hypertension, dizziness, hyperglycemia, and hypoglycemia were more common in the intervention group than in the control group, and falls and heart failure were more common in the control group.

Discussion

The REHAB-HF trial examined the effects of an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains in frail, older patients who were hospitalized for acute decompensated heart failure. The intervention group had significantly greater improvement in physical function, as assessed by the score on the Short Physical Performance Battery at 3 months, than the control group. The results of the analyses of 6-minute walk distance, frailty status, quality of life, and depression also suggested clinical benefits of the intervention. Over the course of 6 months, the incidence of rehospitalization for any cause, rehospitalization for heart failure, and death was high in both groups.

Our trial was designed to address several critical evidence gaps regarding physical rehabilitation in patients with heart failure. Most previous trials excluded patients who had been hospitalized within the previous 6 weeks — a period during which the severity of physical dysfunction and the risk of clinical events are highest; those trials also involved few older, frail patients with multiple coexisting conditions in whom different approaches may be appropriate.3,10 In previous early trials of rehabilitation after heart failure, enrollment of the patients and initiation of the intervention began, on average, 7 weeks after hospital discharge; traditional endurance exercise training was commonly used24,25; the enrolled patients were younger and much less frail and diverse than those in our trial25; there was no control group24,26; and the trials were unblinded25,27 and often small, single-center trials.27 One of the largest of such trials involving recently hospitalized patients, EJECTION-HF (Exercise Joins Education: Combined Therapy to Improve Outcomes in Newly-Discharged Heart Failure), showed no benefit of the intervention over usual care with respect to 6-minute walk distance, rehospitalization, and death, but adherence to the intervention was low (43%).25

Physical dysfunction, frailty, and depression are often unrecognized clinically in older patients hospitalized for heart failure,8,28 are generally not addressed in clinical care pathways,10,29 and probably contribute to delayed, incomplete recovery and high rates of rehospitalization, death, and long-term loss of independence after hospital discharge.2,11,12,14 Physical-function impairments in the patients in the REHAB-HF trial were broader and more severe than those observed in patients with chronic heart failure.28 For example, the mean baseline 6-minute walk distance in the REHAB-HF trial was half that observed in the HF-ACTION (Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training) trial,30 and severe leg weakness prevented nearly one third of the patients from standing even once from a seated position without the use of their arms. The patients in our trial also had severe deficits in balance and mobility, and a history of falls and other geriatric conditions was common — findings that are not typically seen in patients with chronic heart failure and are not addressed by conventional cardiac rehabilitation. The initiation of standard endurance exercise training in frail, older patients without first addressing deficits in balance and mobility can limit efficacy31 and increase the risk of injuries and falls.32,33

The intervention-related benefits seen in the REHAB-HF trial generally exceeded previously reported values for the minimal clinically important difference. The mean difference between the groups in the Short Physical Performance Battery score (1.5 points) was three times as large as the reported minimal clinically important difference (0.5 points).21,34 All three of the components of the Short Physical Performance Battery — corresponding to balance, strength, and mobility — showed greater improvement in the intervention group than in the control group. The apparent benefits in 6-minute walk distance (34 m) and KCCQ score (7.1 points) were also larger than the reported minimal clinically important differences (30 m and 5 points, respectively).35,36 The suggested benefit for depression is of interest, since depression is common among patients with heart failure and is associated with frequent rehospitalization,37 and trials targeting depression in patients with heart failure have had neutral results.

The greater improvements in physical function relative to the control group were seen despite the receipt of routine physical or occupational therapy or traditional cardiac or pulmonary rehabilitation as part of usual care by 43% of the patients in the control group. At 6 months, 83% of the patients in the intervention group who were alive and were being followed by telephone contact reported regular home exercise, which suggested that behavioral change — a requisite for long-term adherence — may have occurred.

The intervention-related benefits may be related to both the severity of the baseline deficits and the robust, broad systemic effects of physical exercise, which favorably alters energy metabolism, oxidative stress, inflammation, tissue repair, growth-factor response, and regulatory pathways.38 Older patients with heart failure can have severe skeletal-muscle myopathy that contributes to physical dysfunction and abates with exercise.9

The number of deaths, including deaths from cardiovascular causes, was higher in the intervention group than in the control group, although the numbers and differences were small and may have been due to chance. A meta-analysis of trials of exercise-based rehabilitation for heart failure showed no significant effect on death from any cause during follow-up for up to 12 months.39 In the EJECTION-HF trial, fewer deaths were reported in the intervention group than in the control group at 12 months.25 Given the wide confidence interval for the rate ratio for death from any cause, we cannot rule out the possibility of an increase (or decrease) in risk with an early exercise regimen among some patients.

Our trial has other important limitations. First, the results did not show a beneficial effect on clinical events. However, a study that examined patient preferences in patients with heart failure indicated that improving physical function and maintaining independence are highly valued, independent of clinical events.40 Second, although the staff members who assessed the primary outcome were unaware of the trial-group assignments, it was not possible for patients to be unaware of the group to which they had been randomly assigned. Third, the benefits of the intervention over usual care may have been moderated owing to the usual-care exercise therapy received by the control group. Fourth, differences between the groups in the amount of caregiver attention could have influenced outcomes. Fifth, the long-term durability of the benefit of the intervention is uncertain. Finally, many patients were ineligible or unable or unwilling to participate, and some discontinued the intervention.

Among patients who were hospitalized for acute decompensated heart failure, a transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains and that began during, or early after, hospitalization and continued for 12 weeks after hospital discharge resulted in significantly greater improvement in physical function than usual care.

Notes

This article was published on May 16, 2021, at NEJM.org.

A data sharing statement provided by the authors is available with the full text of this article at NEJM.org.

Supported by research grants from the National Institutes of Health (R01AG045551, R01AG18915, P30AG021332, P30AG028716, and U24AG059624), the Kermit Glenn Phillips II Chair in Cardiovascular Medicine, and the Oristano Family Fund at Wake Forest School of Medicine.

Dr. Kitzman reports receiving steering committee fees and data and safety monitoring board fees from AbbVie, grant support, paid to Wake Forest Baptist Health, and advisory board fees from AstraZeneca and Bayer, advisory board fees from Boehringer Ingelheim and Corvia Medical, consulting fees from Keyto and Merck, and grant support, paid to Wake Forest Baptist Health, and consulting fees from Novartis; Dr. Whellan, receiving grant support, paid to Thomas Jefferson University, from Amgen, end-point committee fees from CVRx and FibroGen, steering committee fees from Cytokinetics, and grant support, paid to Thomas Jefferson University, and advisory board fees from Novo Nordisk; Dr. Duncan, owning stock in Care Directions; and Dr. Mentz, receiving grant support from American Regent and Cytokinetics, consulting fees from Amgen, Boehringer Ingelheim, and Relypsa, research support and honoraria from Abbott, American Regent, Amgen, Boehringer Ingelheim, Eli Lilly, Boston Scientific, Cytokinetics, FAST BioMedical, Gilead Sciences, Innolife, Medtronic, Relypsa, Respicardia, Roche, Sanofi, Vifor Pharma, and Windtree Therapeutics, grant support, research support, and honoraria from AstraZeneca, and research support, consulting fees, and honoraria from Bayer, Merck, and Novartis. No other potential conflict of interest relevant to this article was reported.

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

We thank Alain Bertoni, M.D., Peter Brubaker, Ph.D., and Jack Rejeski, Ph.D. for their critical review and editing of an earlier version of the manuscript; and Timothy Morgan, Ph.D. (deceased), for his biostatistical expertise.