Introduction
By 2040, 19 million adults over the age of 65 will have a history of cancer. Falls in older cancer survivors have serious implications including missed cancer treatments and being associated with increased morbidity, mortality, and a decreased quality of life.1–3 Many risk factors are thought to contribute to falls in cancer survivors, including age, frailty, fatigue, chemotherapy-induced peripheral neuropathy, and lab abnormalities in addition to impairments in functional status, balance, vestibular function, vision, and cognition.4–6 Despite identification of these factors, there continues to be uncertainty on how to best assess fall risk during cancer survivorship.7 Moreover, there is a tradeoff between fall risk assessments that include identifying a greater number of risk factors, which may enhance fall prediction accuracy, and clinical feasibility given limited time, space, and expertise.4 Therefore, there is a critical need for a screening test that is both accurate and clinically feasible in identifying cancer survivors at significant fall risk throughout survivorship.
The primary aim of this study was to determine the association between clinically feasible measures of physical function and recurrent falls in older cancer survivors treated with chemotherapy. Additionally, the study aimed to determine if the association was independent of age and Eastern Cooperative Oncology Group (ECOG) scores. Unipedal stance time (UST), five sit-to-stand time (5STS), and repetitions of sit-to-stand completed in thirty seconds (30STS) were selected as measures due to their associations with falls in older adults, their clinical feasibility, and their emphasis on different aspects of physical function.8–10 Recurrent falls were selected as the primary outcome as recurrent falls are less likely to happen by extrinsic chance (i.e., a fall on black ice) and have a stronger association to intrinsic physiologic impairments and conditions.9 However, due to the concern about any falls and injurious falls in this vulnerable population, the study also recorded these outcomes.3,11 The study’s a priori hypothesis was that shorter UST would be associated with recurrent falls, injurious falls, and all falls and be independent of age and functional status, measured by ECOG score. UST was chosen due to the emphasis on medial and lateral balance function and because it closely approximates the single leg requirements of gait.
Methods
Study Population
The analytical sample included male and female cancer survivors who received chemotherapy for their cancer treatment. Participants were recruited using a convenient sampling strategy from the University of Vermont Oncologic Rehabilitation Program between January 1, 2012 and June 1, 2022 and were excluded if they were actively receiving radiation or chemotherapy, had a neurologic disorder aside from suspected chemotherapy-induced peripheral neuropathy, or functional limitation rendering balance testing unsafe (e.g., amputation). Informed consent was obtained for all subjects and the study was approved by the University of Vermont Institutional Review Board.
Covariates
Relevant sociodemographic and medical history was included and obesity was defined by ≥30 kg/m2. An ECOG score was obtained at the time of testing. Chemotherapy-induced peripheral neuropathy (CIPN) was defined as the development of new distal extremity numbness and tingling during chemotherapy treatment which persisted until the time of participation in the study.
Main Independent Variables
Physical function was assessed by UST, 5STS, and 30STS, which have been previously validated in older adults and cancer survivors.8,9 For UST, participants underwent two trials with the best time selected and a maximum UST of 60 seconds. Participants completed one trial of 5STS followed by one trial of STS30 and participants had rest breaks between balance measures.
Outcomes
Fall history was collected through participant-reported falls during the past year. If the participant answered yes to having a fall, they were then asked if they had an injury during the fall and if they had fallen more than once during the past year. Falls were defined by any slip, trip, or fall in which the participant landed unintentionally on a floor, the ground, or a supporting surface at a lower level. A fall could not be the result of a medical event (e.g., seizure, fainting) or a push or shove.12
Statistical Analysis:
Descriptive statistics for demographics, past medical history, and cancer history were compared between recurrent fallers and non-recurrent fallers using t-tests and Wilcox rank sum tests for continuous normal and non-normal variables, respectively. Normality of variables was inferred from histograms. Fisher’s exact tests were used for categorical variables. Plots of the residuals from logistic regression models of UST, age, and ECOG score with falls outcomes approximated near-normal distributions. Multivariable logistic regression models were used to evaluate the association of UST with recurrent falls while adjusting for variables available to oncologists or supportive care providers, including age and ECOG score. Exploratory models were constructed for injurious falls and any fall. To assess the predicted probability of falls outcomes by UST we calculated the marginal means adjusting for all aforementioned covariates in the models for fall outcomes. To visually demonstrate this association, we plotted the predicted probability on the y axis and mean margins from a value 0 seconds of UST to 60 seconds on the x axis.13 Stata 17 was used for the analysis and statistical significance was defined as p<0.05 (two-tailed).
Results
The characteristics of the full cohort were grouped by the presence of recurrent falls and are presented in Table 1. Overall, the full cohort comprised functional, older cancer survivors with an average age of 66 years. Cancer survivors with and without recurrent falls were comparable in sociodemographics, functional status, cancer stage, and treatment history but recurrent fallers had a significantly lower UST compared to those without recurrent falls (p<0.05). The groups demonstrated similar performances on 5STS and 30STS (Table 1).
Table 1:
Demographic, cancer history, and balance assessment data grouped by recurrent fall status and p-values for comparison of no recurrent falls and recurrent falls using t-tests, Wilcox Rank Sum, and Fishers exact test for continuous and categorical variables.
| Full Cohort N=50 | No Recurrent Falls, N=41, 82% | Recurrent Falls N=9, 18% | P-value | |
|---|---|---|---|---|
|
| ||||
| Mean (SD), Median (IQR), Percent (N) | ||||
|
| ||||
| Age | 65.6 (11.5) | 66.4 (10.8) | 62.1 (14.8) | p= 0.32 |
|
| ||||
| Female | 88% (44) | 83% (34) | 67% (6) | p=0.36 |
|
| ||||
| Obesity | 22% (11) | 17% (7) | 44% (4) | p=0.09 |
|
| ||||
| Diabetes | 8% (4) | 7% (3) | 11% (1) | p=0.56 |
|
| ||||
| Cancer Type | p=0.01 | |||
| Breast | 62% (31) | 68% (28) | 33% (3) | |
| Gynecological | 4% (2) | 5% (2) | 0% (0) | |
| Hematological | 8% (4) | 5% (2) | 22% (2) | |
| Head Neck | 8% (4) | 10% (4) | 0% (0) | |
| GI | 4% (2) | 0% (0) | 22% (2) | |
| Other: Lung, Myeloma, Prostrate | 14% (7) | 12% (5) | 22% (2) | |
|
| ||||
| Cancer Stage | p=0.33 | |||
| I | 18% (9) | 20% (8) | 11% (1) | |
| II | 36% (18) | 33% (14) | 44% (4) | |
| III | 14% (7) | 17% (7) | 0% (0) | |
| IV | 14% (7) | 10% (4) | 33% (3) | |
| Unknown | 18% (9) | 20% (8) | 11% (1) | |
|
| ||||
| Surgery | 84% (42) | 88% (41) | 67% (6) | p=0.14 |
|
| ||||
| Radiation | 70% (35) | 71% (29) | 67% (6) | p=1.00 |
|
| ||||
| Years from Chemotherapy | 8.6 (5.2) | 8.3 (4.5) | 9.7 (7.4) | p=0.48 |
|
| ||||
| CIPN | 44% (22) | 39% (16) | 67% (6) | p=0.16 |
|
| ||||
| ECOG | p=0.37 | |||
| 0 | 66% (33) | 61% (25) | 89% (8) | |
| 1 | 28% (14) | 32% (13) | 11% (1) | |
| 2 | 6% (3) | 7% (3) | 0% (0) | |
|
| ||||
| 5STS time (s) | 10 (9,12) | 10 (9,12) | 10 (7,12) | p=0.73 |
|
| ||||
| 30STS Reps | 15 (13,19) | 15 (12,19) | 15 (13,18) | p=1.00 |
|
| ||||
| UST (s) | 56 (10,60) | 60 (16,60) | 12 (5,55) | p=0.04* |
p<0.05
GI: Gastrointestinal
CIPN: Chemotherapy induced peripheral neuropathy
Eastern Cooperative Oncology Group (ECOG) scores
Five sit-to-stand time (5STS)
Unipedal stance time (UST)
Repetitions of sit-to-stand completed in thirty seconds (30STS)
Multivariable logistic regression models with UST as the primary independent variable for recurrent falls, injurious falls, and any fall are represented in Figure 1 and in Supplementary Table 1. Recurrent falls were not significantly associated with increased age (for each increase in one year: odds ratio [OR]=0.914, [0.834, 1.00], p= 0.055). However, lower risk of reporting recurrent falls was significantly associated with an ECOG score of 0 compared to an ECOG score of 1 (OR= [17.7, 251.5], p=0.03) and longer UST (for each increase of one second: OR=921, [0.869, 0.976], p=0.005). Increased age and longer UST were significantly associated with fewer reports of injurious falls and any falls, but ECOG status was not significantly associated with injurious and any falls in either model (Supplementary Table 1). 5STS, 30STS, and UST for falls and injurious fall groups are shown in Supplementary Table 2.
Figure 1:

Predicted Probabilities of Recurrent Falls, Injurious Falls, and Any Falls with increased Unipedal Stance Time values from multivariable models.
Discussion
In this study of older cancer survivors treated with chemotherapy, longer UST was associated with decreased recurrent falls, injurious falls, and any falls. 5STS and 30STS performances were not associated with fall outcomes. In multivariable models, UST’s association with fall outcomes were independent of age and functional status measured by ECOG status. Interestingly, increased age was associated with decreased any falls and injurious falls, but not recurrent falls. Collectively, these findings suggest that UST may aid in fall risk assessment during cancer survivorship, particularly in aging, chemotherapy-treated cancer survivors.
The association between UST and falls in older cancer survivors is supported by recent literature.8 Previously, UST, also known as single leg stance time, was found to be feasible in a large population of early cancer survivors in which chemotherapy treatment was not explicitly detailed and recurrent and injurious falls were not evaluated.8 Our data showed no association between any fall outcome and sit-to-stand performances, regardless of technique used. In contrast, UST was significantly associated with all fall outcomes and this could be explained in part by the balance requirements emphasized by UST. UST recapitulates requirements of ambulation in the single leg stance phase of the gait cycle, specifically medial and lateral stability, which may be of relevance given that most falls occur with walking and many fall injuries occur laterally.5 In comparison, sit-to-stand function is performed bipedally and emphasizes anterior and posterior postural stability.10 Cautious interpretation of these results is called for given that this study could be underpowered. Regardless, the current study adds to the literature by reporting an association between UST and recurrent falls, injurious falls, and any fall in older cancer survivors.
The study’s findings that younger age within multivariable models is associated with a higher risk of injurious falls and any fall were unexpected. Previously, older age has been strongly associated with falls in cancer survivors; however, multivariable models evaluating the changing risk for increasing age were not found. ECOG score was largely uninformative of fall risk in multivariable models because ECOG score may incompletely capture functional status.14 Participants with an ECOG score of 0 had increased recurrent falls compared to those with an ECOG score of 1. An explanation could be that cancer survivors with an ECOG status of 1 are inactive for longer periods of the day and, therefore, encounter fewer opportunities to fall. While this is a pilot study and cautious interpretation is warranted, the finding that younger age was associated with higher risk of recurrent falls and ECOG status was not informative emphasizes the need for a physical function test to supplement fall risk assessment in functional, older cancer survivors.
This study has limitations consistent with a single site, retrospective study consisting of a relatively homogeneous participant cohort. Specifically, the participants were highly functional via ECOG status and this may have limited its ability to find an association with any falls or injurious falls, as most participants reported an ECOG score of 0 or 1. Future studies could consider measuring functional status in a more quantitative and evidence-based method, such as using a geriatric assessment.14 Although selection bias is probable from the mode of recruitment, our participant fall rate was 34%, similar to the 33%-50% commonly reported in the cancer survivor literature.11 Falls were retrospectively assessed, a method used extensively, but limited by participant recall.3
Our collective findings suggest that in older chemotherapy-treated cancer survivors, UST is associated with recurrent, injurious, and any falls and is independent of age and functional status measured by ECOG. Moving forward, research should use larger longitudinal studies to establish cutoffs for UST to better assess its potential to inform providers of older cancer survivors who require fall intervention strategies.
Supplementary Material
Acknowledgements and Funding Statement:
The work was supported in part by P30AG02482 to BLM
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Ethics and Integrity: The authors (BLM, KM, JM, NK, JAS, HC, MKH, RR, SGB MR, CR, JKR, NK) all report ethical research conduct in the proposed manuscript and confirm this data has not been published prior.
Conflict of Interest Disclosure and Competing Interest: BLM, KM, JM, NK, JAS, HC, MKH, RR, SGB, MR, CR, JKR, NK report no conflicts of interest with the current manuscript.
NK reports receiving consulting fees from Eisana Corporation as a medical consultant and honoraria for lectures from AANEM.
MKH reports receiving consulting fees from Argenx, Alexion, UCB Pharma, Janssen, and Immunovant and honoraria from Medscape, Springer Health, AANEM, Medlink Neurology, and Continuum Lifelong Learning in Neurology.
Patient Consent: Informed consent was obtained for all subjects and the study was approved by the University of Vermont Institutional Review Board.
Consent to Participate: Informed consent was obtained from all individual participants included in the study
Data Availability:
Data is available upon reasonable request.
References:
- 1.Toomey A, Friedman L. Mortality in cancer patients after a fall-related injury: The impact of cancer spread and type. Injury. 2014;45(11):1710–1716. doi: 10.1016/j.injury.2014.03.008 [DOI] [PubMed] [Google Scholar]
- 2.McNeish BL, Richardson JK, Bell SG, Whitney DG. Chemotherapy-induced peripheral neuropathy increases nontraumatic fracture risk in breast cancer survivors. JBMR Plus. 2021;5(8):e10519. doi: 10.1002/jbm4.10519 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sattar S, Alibhai SMH, Spoelstra SL, Puts MTE. The assessment, management, and reporting of falls, and the impact of falls on cancer treatment in community-dwelling older patients receiving cancer treatment: Results from a mixed-methods study. J Geriatr Oncol. 2019;10(1):98–104. doi: 10.1016/j.jgo.2018.08.006 [DOI] [PubMed] [Google Scholar]
- 4.Wildes TM, Maggiore RJ, Tew WP, et al. Factors associated with falls in older adults with cancer: a validated model from the Cancer and Aging Research Group. Support Care Cancer. 2018;26(10):3563–3570. doi: 10.1007/s00520-018-4212-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kolb NA, Smith AG, Singleton JR, et al. The Association of Chemotherapy-Induced Peripheral Neuropathy Symptoms and the Risk of Falling. JAMA Neurol. 2016;73(7):860–866. doi: 10.1001/jamaneurol.2016.0383 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.McNeish BL. Executive function is associated with balance and falls in older cancer survivors treated with chemotherapy: A cross-sectional study. J Geriatr Oncol. Published online 2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Huang MH, Hile E, Croarkin E, et al. Academy of Oncologic Physical Therapy EDGE Task Force: A Systematic Review of Measures of Balance in Adult Cancer Survivors. Rehabil Oncol. 2019;37(3):92. doi: 10.1097/01.REO.0000000000000177 [DOI] [Google Scholar]
- 8.Marker RJ, Kakar RS, Scorsone JJ, Peters JC, Purcell WT. Single-Leg Stance Times in a Diverse Group of Survivors of Cancer and the Relationship to History of Recent Falls. Rehabil Oncol. 2021;39(1):23. doi: 10.1097/01.REO.0000000000000243 [DOI] [Google Scholar]
- 9.Tiedemann A, Shimada H, Sherrington C, Murray S, Lord S. The comparative ability of eight functional mobility tests for predicting falls in community-dwelling older people. Age Ageing. 2008;37(4):430–435. doi: 10.1093/ageing/afn100 [DOI] [PubMed] [Google Scholar]
- 10.Lord SR, Murray SM, Chapman K, Munro B, Tiedemann A. Sit-to-Stand Performance Depends on Sensation, Speed, Balance, and Psychological Status in Addition to Strength in Older People. J Gerontol Ser A. 2002;57(8):M539–M543. doi: 10.1093/gerona/57.8.M539 [DOI] [PubMed] [Google Scholar]
- 11.Sattar S, Haase K, Kuster S, et al. Falls in older adults with cancer: an updated systematic review of prevalence, injurious falls, and impact on cancer treatment. Support Care Cancer. 2021;29(1):21–33. doi: 10.1007/s00520-020-05619-2 [DOI] [PubMed] [Google Scholar]
- 12.Tinetti ME, Speechley M, Ginter SF. Risk Factors for Falls among Elderly Persons Living in the Community. N Engl J Med. 1988;319(26):1701–1707. doi: 10.1056/NEJM198812293192604 [DOI] [PubMed] [Google Scholar]
- 13.Williams R Using the Margins Command to Estimate and Interpret Adjusted Predictions and Marginal Effects. Stata J. 2012;12(2):308–331. doi: 10.1177/1536867X1201200209 [DOI] [Google Scholar]
- 14.Dale W, Klepin HD, Williams GR, et al. Practical Assessment and Management of Vulnerabilities in Older Patients Receiving Systemic Cancer Therapy: ASCO Guideline Update. J Clin Oncol. 2023;41(26):4293–4312. doi: 10.1200/JC0.23.00933 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data is available upon reasonable request.
