Moreover, a growing body of research in both oncology and non-oncology settings suggests that physical activity data can be used to predict numerous health outcomes, including symptoms of treatment toxicity.2,3 However, these studies are often conducted with middle-aged adults with earlier-stage disease. Therefore, the feasibility of collecting physical activity data from older adults, as well as the generalizability of these associations to older adults is unclear. Given the rapid uptake of this technology in health care settings, we conducted a prospective cohort study to address these gaps in the literature.
A total of 47 older adults starting active treatment for metastatic prostate cancer were recruited (median age=75). Most participants had a higher level of education (80.9% with at least some college/university) and were treated with chemotherapy (40.4%) or an androgen receptor axis-targeted therapy (38.3%). Participants in our study also had a low level of comorbidities (Cumulative Illness Rating Scale-Geriatric [CIRS-G] median=1.6), high physical function (93.4% with Eastern Cooperative Oncology Group Performance Status [ECOG PS] 1 or 2), and some were frail (42.5% with a score of >3 on Vulnerable Elders Survey-13 [VES-13]).
Participants were asked to report their step count and symptoms each day (i.e., daily physical activity monitoring) for one treatment cycle (3-4 weeks) during the study period (January 2020 to December 2021). The response rate (adherence to daily reporting) was 90.5% and the retention rate (study completion) was 94%, suggesting a high level of feasibility and acceptability. Moreover, in semi-structured interviews, participants reported that they enjoyed participating in daily step count and symptom monitoring. Many reported that it motivated them to be more active and allowed them to become more aware of their physical activity and symptom trajectories. These findings suggest that physical activity monitoring may be useful in the context of health promotion (i.e., to encourage older adults to be more physically active and engaged in their health).
Nevertheless, some challenges associated with physical activity monitoring were discussed in interviews. These included the logistics of using a device (e.g., remembering to charge the battery, remembering to wear the device, etc.), as well as its inability to track other forms of physical activities that don’t involve step counts (e.g., biking). Therefore, the usage patterns of older adults will need to be considered for the effective implementation of wearable devices in geriatric healthcare.
When analyzing the predictive validity (i.e., positive predictive value [PPV] and sensitivity) of step counts on the emergence of overall symptoms, reasonable validity estimates were found (sensitivity=81.8%, PPV=73.0% for any moderate-to-severe symptoms; sensitivity=76.9%, PPV=27.0% for any severe symptoms). However, the predictive validity of step counts on the emergence of pain was low (sensitivity=77.8%, PPV=37.8% for moderate-to-severe pain; sensitivity=100%, PPV=13.5% for severe pain). Validity estimates are further described in Table 1 below. In addition, associations between declines in step count and symptoms of treatment toxicity were not found in logistic regression models, even after adjustment for potential confounders (aOR=2.24, 95% CI=0.45-11.19 for any moderate-to-severe symptom; aOR=1.10, 95% CI=0.23-5.36 for moderate-to-severe pain).
Table 1. Predictive validity of a step count decline on the emergence of symptoms within 24 hours.
Note. CI = confidence interval
Taken together, a decline in physical activity may be better used as a general prediction of health rather than pain specifically, but false positives are possible. Moreover, associations between physical activity and symptoms may not be as generalizable to older adults, although our models were limited by a small sample size. Following participants over a longer duration of time (i.e., beyond one treatment cycle) may also be warranted to identify associations with health outcomes.
Overall, physical activity monitoring was found to be feasible and acceptable among older adults undergoing active treatment for metastatic prostate cancer. However, it may be more useful as a health-promoting intervention, rather than as a means to predict treatment toxicity. Further details on our study can be found in the Journal of Geriatric Oncology.4
Written by: Gregory Feng, MPH, Milothy Parthipan, MScPT(c), Henriette Breunis, CCRP, Shabbir M. H. Alibhai, MD, MSc
Department of Supportive Care, Princess Margaret Cancer Centre, University Health Network, Toronto, Canada
References:
- Wang J, Du Y,Coleman D, et al. Mobile and connected health technology needs for older adults aging in place: cross-sectional survey study. JMIR Aging 2019;2(1): e13864.
- Bennett AV, Reeve BB, Basch EM, et al. Evaluation of pedometry as a patient-centered outcome in patients undergoing hematopoietic cell transplant (HCT): a comparison of pedometry and patient reports of symptoms, health, and quality of life. Qual Life Res 2016;25(3):535–46.
- Purswani JM, Dicker AP, Champ CE, Cantor M, Ohri N. Big data from small devices: the future of smartphones in oncology. Semin Radiat Oncol 2019;29(4): 338–47.
- Feng G, Parthipan M, Breunis H, et al. Daily physical activity monitoring in older adults with metastatic prostate cancer on active treatment: Feasibility and associations with toxicity. JGO 2023;14(7): 101576.