Declining Physical Activity May Herald HF Hospitalization, Death

Batya Swift Yasgur MA, LSW

February 19, 2020

Decreases in physical activity (PA) in patients with heart failure (HF), as detected by implantable devices, may flag those at imminent risk for hospitalization or death, a new study suggests.

Investigators analyzed activity levels and changes over time in close to 29,000 patients with HF who had received an implantable cardioverter-defibrillator (ICD) or cardiac resynchronization therapy defibrillator (CRT-D).

Data on PA were derived from the devices, with the primary outcome measure being a composite of HF hospitalization and all-cause mortality.

PA levels in patients during the weeks prior to an event were compared with levels among patients who did not experience an event during the 5-year follow-up period, and also with patients who experienced an event after 1 year.

The group with the impending clinical event had the largest percentage of low-activity patients, defined as those doing less than half an hour of activity per day.

An even more powerful predictor of impending clinical event was the rate of change in activity, in which a 10-minute reduction in physical activity increased the risk for hospitalization or mortality fourfold.

"We found that device-measured physical activity and the trajectory of the physical activity in patients with heart failure were associated with heart failure hospitalization and mortality well in advance of these important cardiovascular events in the patient's life," lead author Jacob Kelly, MD, MHS, advanced heart failure cardiologist, Alaska Heart and Vascular Institute, Anchorage, told | Medscape Cardiology.

"The main take-home messages for practicing clinicians are that there are a lot of data available in patient devices," he said, and that "decrements [in device-measured PA] are a possible signal that could be targeted with a multitude of therapies, including cardiac rehabilitation, opportunities to titrate guideline-directed medical therapy for heart failure, or consider alternative therapies in the expanding device realm."

The study was published online February 5 in JACC: Heart Failure.

Not Commonly Used

"As a clinician, I have always been fascinated with physical activity and noticed sharp drop-offs in activity in patients with worsening heart failure that correspond with their disease severity and symptoms," Kelly said.

He observed that this trend was associated with hospital admissions and worsening symptoms and, together with his colleagues, developed the hypothesis for the current study.

"While interrogating implantable devices on patients with heart failure, I noticed that their accelerometer data were available and not commonly being used, and this prompted me to start clinically following it," Kelly said.

Coauthor Brett D. Atwater, MD, associate professor of medicine and director of clinical cardiac electrophysiology, Duke University Medical Center, Durham, North Carolina, told | Medscape Cardiology that it has "proven difficult to accurately predict who will have worsening of their [HF] symptoms and impending events." He wondered if the reduction in activities requiring physical exertion prior to hospital admission for HF could be used to predict hospitalization or death.

To investigate the question, the researchers linked daily activity data from a dataset consisting of Medicare beneficiaries who received Boston Scientific ICD or CRT-D implantations from 2007 to 2009.

The registry's remote data transmissions from 2007 to 2014 were then linked with Medicare claims from 2006 to 2014.

The researchers divided a total of 28,982 patients (mean age, 75.7; 74% male; 92% white) into three groups, obtaining both the PA value and the PA slope for each patient.      

  • Group 1: patients in whom death or HF hospitalization were imminent, using the PA value just prior to the event and the most recent PA slope value over the 8-week period prior to the event. The median survival time for these patients was 90 weeks. Two additional groups were then formed to compare with group 1:

  • Group 2: patients with PA and PA slope values at the 90th week of follow-up who did not experience an event during the entire 5-year follow-up period

  • Group 3: patients with PA and PA slope values at the 90th week of follow-up who experienced an event at least 1 year after the 90th week.

The researchers used a joint model — i.e., a submodel for PA and PA trajectory slope and a submodel for the event.

Women Undertreated

Of the entire sample, 14.1% experienced a composite end point event of HF hospitalization or death, and close to half (49.6%) experienced one by 5 years.

The median daily PA level across all PA records was 85 minutes (IQR, 53.1 - 124.7), the authors report.

Median Physical Activity Prior to HF Hospitalization or Death
Time Point Prior to Event Prior to HF Hospitalization, minutes (IQR) Prior to Death, minutes (IQR)
15 weeks 60 (34–97) 41 (20–72)
1 week 50 (IQR 26 – 85) 22 (8–50)

Using a half-hour mark as a measure of low exercise, the researchers found that the largest proportion of patients (37%) with an imminent event (group 1) were active for less than 30 minutes per day, compared with 6% of patients without an event (group 2) and 13% of patients with a future event (group 3).

One-third (32%) of patients with an imminent event (group 1) had a reduction of at least 30% in activity over an 8-week period, compared with 6% and 8% of patients without an event (group 2) and those with a future event (group 3), respectively.

A 1.13-fold increase in the hazard of the composite end point was found in the adjusted joint model for 75 minutes of daily PA, compared with 85 minutes (95% CI, 1.12 - 1.13).

Moreover, a 10-minute decrease in average daily PA (from 85 to 75 minutes) over an 8-week period was associated with a hazard ratio of 4.02 (95% CI, 3.82 - 4.22) for the composite end point.

The authors note that their findings were derived from a retrospective analysis of Medicare patients, and might not be generalizable to younger patients, non-Medicare patients, or patients without an implantable device.

Moreover, only 26% of patients included in the analyses were female, "which highlights the underutilization of ICD and CRT-D in women," they add.

Real-World Information

Commenting on the study for told | Medscape Cardiology, Emily P. Zeitler, MD MHS, assistant professor of medicine, Geisel School of Medicine, Hanover, New Hampshire, said that the "best part of the design was the sheer number of patients, which is a really wonderful consequence of these large databases generated from heart failure patients with implantable devices."

These devices have "extraordinary monitoring capabilities to store and analyze data and I'm hopeful that studies like this can get real-world information about our patients who are walking around with these data-generating machines in their bodies," said Zeitler, who is also assistant professor of healthcare policy, Dartmouth Institute, Dartmouth-Hitchcock Medical Center, and was not involved with the study.

Additionally, she was "struck" by the low level of physical activity in HF patients to begin with; the average activity was less than 6% of the entire day, or about 85 minutes, which is "abysmally low."

The high 5-year mortality rate "shows that half of these patients are dying, which goes to show we need to do a lot to improve our management of heart failure patients," she emphasized.

Zeitler cautioned that the study was observational and therefore could not determine causality. "The Holy Grail question to be addressed in clinical trials is whether, if you change physical activity, you can change mortality."

Kelly agreed that further work is needed to "determine if device-measured physical activity can be used to motivate patients with heart failure to become more physically active and whether this translates into improved outcomes."

Kelly receives research support from Boston Scientific and Abbott Medical, and honoraria from Novartis. Atwater receives research support from Boston Scientific and Abbott Medical, and honoraria from Biosense Webster, Abbott Medical, Medtronic, and Biotronik. The other authors' disclosures are listed on the original paper. Zeitler reports no relevant financial relationships.

JACC Heart Fail. Published online February 5, 2020. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.