Abstract
Injurious falls pose a significant threat to the safety of stroke patients, particularly among older adults. While the influence of activities of daily living (ADL) on falls is acknowledged, the precise connection between ADL ability and fall-related injuries in older stroke patients undergoing rehabilitation, particularly those with varying mobility levels, remains unclear. This multicenter cross-sectional study in China recruited 741 stroke patients aged 65 years and above, categorized into bedridden, domestic, and community groups based on their mobility levels using the Longshi Scale. ADL ability was assessed using the Barthel Index. Logistic regression models, generalized additive models, smoothed curve-fitting, and threshold effect analysis were employed to explore the relationship between ADL ability and injurious falls across the three mobility groups. Results revealed an inverted U-shaped relationship between ADL ability and injurious falls among patients in the domestic group (p = 0.011). Below the inflection point of 35 on the Barthel Index, the likelihood of injurious falls increased by 14% with each unit increase in ADL ability (OR = 1.14, 95% CI 1.010–1.29, p = 0.0331), while above the inflection point, it decreased by 3% per unit increase (OR = 0.97, 95% CI 0.95–0.99, p = 0.0013). However, no significant association between ADL ability and injurious falls was observed in either the bedridden or community groups (p > 0.05). These findings suggest that only older stroke patients capable of engaging in activities at home demonstrate a correlation between ADL ability and injurious falls. The identified inverted U-shaped relationship may aid in identifying fall injury risk in this population.
Keywords: Activities of daily living, Injurious falls, Longshi scale, Barthel index, Older patients, Stroke
Subject terms: Diseases, Medical research
Introduction
Injurious falls are among the reasons for mortality and morbidity in older adults1. The risk of falls is higher among survivors of stroke, persisting even years after the stroke occurrence2,3. During inpatient rehabilitation, stroke patients have a fall rate of up to 48%, with nearly one-third of these falls resulting in potentially severe injuries4. Fall-related injuries encompass abrasions, contusions, sprains, fractures, and, in severe cases, fatalities, with these consequences intricately linked to clinical deterioration, increased caregiving dependency, and additional economic burdens5,6. Therefore, preventing falls and fall-related injuries is crucial for older stroke patients in the rehabilitation department.
Several studies have indicated that injurious falls were associated with a loss of independence on activities of daily living (ADL)7–9. However, the precise correlation between ADL ability and the risk of fall-related injuries remains unclear among older stroke patients with different ranges of mobility. Their mobility range, determined by their ADL ability and functional level, potentially restrict the locations where injurious falls may occur. Additionally, stroke patients with different functional abilities have distinctive social roles and experience diverse daily living environments, which may influence their functional performance and, consequently, their risk of injurious falls10,11. Therefore, a stratified analysis of these individuals based on their range of activities allows for a more comprehensive and accurate understanding of the relationship between their ADL abilities and the occurrence of injurious falls. Our research team has developed the Longshi Scale (LS), a novel pictorial-based self-care assessment tool that categorizes patients into bedridden, domestic, and community groups based on the range of activities they can perform12. In previous research, we observed that stroke patients in different LS subgroups demonstrate varying fall risk levels13. However, the correlation between ADL ability and the risk of injurious falls remains unclear within these three subgroups, necessitating further in-depth research.
The purpose of this study was to investigate the relationship between ADL ability and the risk of occurrence of injurious falls in older stroke patients based on the population classification criteria of LS and to provide a basis for more accurately assessing and predicting the occurrence of injurious falls.
Materials and methods
Study design and setting
This was a multicenter, cross-sectional study. The data was gathered from a clinical study “the Promotion of Application Research of Longshi Scale” which explored the application value of LS nationwide14. Participant data were collected from the rehabilitation departments of 103 hospitals across 23 provinces in China from December 2018 to May 2020 using a convenience sampling approach. The study was approved by the ethics committee of the Shenzhen Second People’s Hospital (approval number No. 20180926006) and registered with the China Clinical Trials Registry (ChiCTR-2000034067). This study was performed in accordance with the Declaration of Helsinki. All patients or their guardians provided written informed consent and all experiments were performed in accordance with relevant guidelines and regulations.
Participants
Initially, the study recruited 15,205 inpatients from the rehabilitation departments. Among them, 741 older stroke patients were selected for further analysis. The inclusion criteria were: (1) diagnosed with stroke according to the 10th edition of the International Classification of Disease15; (2) aged 65 years or older; (3) had a history of falls (post-stroke) within the past 3 months of the survey date (during hospitalization or before). The exclusion criteria were: (1) had cognitive impairments, Mini-mental State Examination (MMSE) ≤ 24; (2) participated in other clinical studies simultaneously. All patients or their guardians provided informed consent forms.
Data collection
Demographic and clinical data were collected via face-to-face questionnaires. The sociodemographic characteristics of the participants included sex, age, illness duration, living situation, annual household income, and psychotropic substance use status. Psychotropic drugs include antipsychotics, antidepressants, antimanic, and anxiolytics16. The living situation was categorized as “living alone” or “not living alone.” For participants with communication disabilities such as aphasia, their general information and fall events were reported by family members or caregivers who had frequent or recent contact with them.
A fall was defined as “a sudden, unintended, uncontrolled downward displacement of a patient’s body to the ground or a lower level”17. The occurrence of falls was reported using a single item: “Did you experience a fall in the last 3 months, and if so, did the fall occur after your stroke (either during hospitalization or before)? What was the result?” The consequences of a fall include no injury; moderate injuries, including bruises, abrasions, sprains, and minor lacerations; serious injuries, including fractures, joint dislocations, or lacerations requiring sutures18. In our study, “no injury” was classified as non-injurious falls and injurious falls included “moderate injuries” and “serious injuries”.
Measurement
When conducting questionnaires, all participants were assigned to a bedridden, domestic, or community group based on the LS, and the patient’s ability to perform ADL was assessed using the BI. The smartphone application Quicker Recovery Line (QRL), an online rehabilitation management system, was used to conduct the two assessments.
The LS, a pictorial scale, categorizes individuals into bedridden, domestic, and community groups depending on their ability to move out of bed, move outdoors, and return indoors (Fig. 1). This scale was recommended as a national standard in 201819. The bedridden group included those who could not get out of bed independently and could only move around. The domestic group was defined as a person who could get down or transfer to a wheelchair and could move independently in the home environment (including a wheelchair) but could not transfer outdoors. A community group was defined as a person who could get down or transfer to a wheelchair and engage in outdoor activities (including wheelchair use) (Fig. 2).
Fig. 1.
Flow chart for the assessment of the Longshi Scale.
Fig. 2.
The Longshi Scale.
The Barthel Index (BI) is a scale used to assess ADL ability, covering various aspects of daily life activities, and providing a comprehensive assessment of an individual’s overall functional level20,21. It consists of 10 items, including controlling the bowel and bladder, grooming, using the toilet, feeding, transferring from a wheelchair to the bed and vice versa, walking on a level surface, dressing, ascending and descending stairs, and bathing. Each item is scored on a scale of 0, 5, 10, or 15, reflecting the level of independence in ADL. The total score ranged from 0 to 100, with higher scores indicating greater independence. Individuals with scores below 20 were considered completely disabled and fully dependent on others. Scores between 20 and 40 indicated severe functional impairment and a significant need for assistance. Scores between 40 and 60 indicated moderate functional impairment and the need for assistance in ADL ability. For individuals with mild functional impairment, a score of 60 or higher indicated basic self-care ability22.
All data collection via questionnaires and patient evaluations, including the LS and BI, were performed by healthcare professionals, including physicians, nurses, and therapists who received online training and understood the purpose of the study.
Statistical analysis
A database was established using Excel 2013 software, and statistical analyses were performed using EmpowerStats and R software (http://www.empowerstats.com, X&Y Solutions, Inc, Boston, MA, USA). Descriptive statistics for categorical data were presented as frequencies and percentages, and continuous variables were described using means ± standard deviations (SD). For normally distributed continuous data with equal variance, t-tests were used for between-group comparisons. Nonparametric tests were used when the data did not follow a normal distribution. Multiple group comparisons of categorical data were conducted using Pearson’s chi-squared test or Fisher’s precision probability test. Binary logistic regression analysis was employed to identify the risk factors associated with injurious falls, including sex, age, BI, illness duration, living situation, annual family income, psychotropic substance status, and LS grouping. Sensitivity analysis was employed by handling the continuous variable BI score as a categorical variable (four subgroups) to assess the robustness of the findings. Generalized additive models (GAM) were used to evaluate the nonlinear relationship between the BI scores and the occurrence of injurious falls. Curve relationship plots were generated, and two-piece linear regression models were used to explore the potential saturation or threshold effect23. p < 0.05 was considered statistically significant.
Results
Characteristics of participants
Among the 741 enrolled older stroke inpatients (≥ 65 years) in this study, 334 were male (45.07%), and 407 were female (54.93%). All participants experienced at least one fall event, with 74.76% of patients experiencing injurious falls. Based on LS classification, 359 participants belonged to the bedridden group, 259 to the domestic group, and 123 to the community group. The mean scores of BI for patients in the bedridden group were 23.39 (SD: 18.98), for those in the domestic group were 64.61(SD: 17.76), and for those in the community group were 88.07 (SD: 13.69). The differences in BI scores among the 3 groups were statistically significant (p < 0.001).
Falls were described as “non-injurious falls” or “injurious falls” based on whether participants sustained injuries from falls, and the risk of injurious falls was higher in the bedridden group (53.61%) than in domestic (33.39%) and community groups (13.00%). Significant differences (p < 0.05) were observed between the group of “non-injurious falls” and “injurious falls” regarding sex, age, BI, illness duration, living situation, annual family income, and LS grouping. However, there were no statistically significant differences (p > 0.05) in the use of psychotropic medications (Table 1).
Table 1.
Baseline characteristics of participants.
| Variables | Non-injurious falls(n = 187) | Injurious falls(n = 554) | p-value |
|---|---|---|---|
| Sex n (%) | |||
| Male | 100 (53.48%) | 234 (42.31%) | 0.008 |
| Female | 87 (46.52%) | 319 (57.69%) | |
| Age (mean ± SD), year | 75.19 ± 7.56 | 78.07 ± 8.39 | < 0.001 |
| Barthel Index (mean ± SD) | 59.39 ± 30.73 | 44.93 ± 29.89 | 0.001 |
| Illness duration, median (Q1, Q3), year | 1.01 (0.21, 3.42) | 0.26 (0.04, 2.25) | < 0.001 |
| Living situation n (%) | |||
| Not living alone | 151 (80.75%) | 484 (87.36%) | 0.025 |
| Living alone | 36 (19.25%) | 70 (12.64%) | |
| Annual family income n (%), RMB | |||
| < 50, 000 | 109 (58.29%) | 199 (35.92%) | < 0.001 |
| 50,000–100,000 | 47 (25.13%) | 152 (27.44%) | |
| 100,000–150,000 | 18 (9.63%) | 141 (25.45%) | |
| 150,000–200,000 | 7 (3.74%) | 47 (8.48%) | |
| ≥ 200,000 | 6 (3.21%) | 15 (2.71%) | |
| Psychotropic substances status n (%) | |||
| Using | 166 (88.77%) | 516 (93.14%) | 0.056 |
| Not using | 21 (11.23%) | 38 (6.86%) | |
| Longshi Scale grouping n (%) | |||
| Bedridden group | 62 (33.16%) | 297 (53.61%) | < 0.001 |
| Domestic group | 74 (39.57%) | 185 (33.39%) | |
| Community group | 51 (27.27%) | 72 (13.00%) | |
Covariate unadjusted analyses using a binary logistic regression model
Female sex, BI, living alone, LS grouping, and annual family income (≤ 200,000 RMB) were significant risk factors for injurious falls among older patients (p < 0.05). Illness duration, annual family income (> 200,000 RMB), and psychotropic substance use did not significantly affect the occurrence of injurious falls (p > 0.05) (Table 2).
Table 2.
Univariate results by binary logistic regression.
| Variables | Statistics | OR (95% CI) | p-value |
|---|---|---|---|
| Sex n (%) | |||
| Male | 334 (45.14%) | 1.0 | – |
| Female | 406 (54.86%) | 1.57 (1.12, 2.19) | 0.0082 |
| Barthel Index (mean ± SD) | 48.58 ± 30.73 | 0.98 (0.98, 0.99) | < 0.0001 |
| Illness duration, (mean ± SD), year | 2.83 ± 5.53 | 0.97 (0.95, 1.00) | 0.0600 |
| Living situation n (%) | |||
| Not living alone | 635 (85.70%) | 1.0 | – |
| Living alone | 106 (14.30%) | 0.61 (0.39, 0.94) | 0.0265 |
| Annual family income n (%), RMB | |||
| < 50,000 | 308 (41.57%) | 1.0 | – |
| 50,000–100,000 | 199 (26.86%) | 1.77 (1.19, 2.65) | 0.0053 |
| 100,000–150,000 | 159 (21.46%) | 4.29 (2.49, 7.39) | < 0.0001 |
| 150,000–200,000 | 54 (7.29%) | 3.68 (1.61, 8.41) | 0.0020 |
| > 200,000 | 21 (2.83%) | 1.37 (0.52, 3.63) | 0.5275 |
| Psychotropic substances status n (%) | |||
| Using | 682 (92.04%) | 1.0 | – |
| Not using | 59 (7.96%) | 0.58 (0.33, 1.02) | 0.0587 |
| Longshi Scale grouping n (%) | |||
| Bedridden group | 359 (48.45%) | 1.0 | – |
| Domestic group | 259 (34.95%) | 0.52 (0.36, 0.77) | 0.0009 |
| Community group | 123 (16.60%) | 0.29 (0.19, 0.46) | < 0.0001 |
Multivariable analyses using the linear binomial logistic models
Patients were stratified into 3 groups (domestic, bedridden, and community) based on LS. Multivariate analysis using binary logistic regression was conducted to investigate the relationship between BI scores and injurious falls in each group.
In the domestic group, in both the Original model (OR = 0.98, 95% CI 0.97–1.00, p = 0.0540) and Model I (adjusted for age and sex) (OR = 0.98, 95% CI 0.97–1.00, p = 0.0620), the BI was not statistically significant about the occurrence of injurious falls. However, in Model II, after adjusting for age, sex, BI, illness duration, living situation, annual family income, and psychotropic medication status, a negative correlation was observed between BI and injurious falls. For every 1-unit increase in BI score, the risk of injurious falls decreased by 2% (OR = 0.98, 95% CI 0.96–1.00, p = 0.0173) (Table 3).
Table 3.
Multivariate results by binary logistic regression.
| Exposure | Original model (OR, 95% CI, p) | Model I (OR, 95% CI, p) | Model II (OR, 95% CI, p) | |
|---|---|---|---|---|
| Domestic group | Barthel Index score |
0.98 (0.97, 1.00) 0.0540 |
0.98 (0.97, 1.00) 0.0620 |
0.98 (0.96, 1.00) 0.0173 |
| Barthel Index score (quartile) | ||||
| Q1 | 1.0 | 1.0 | 1.0 | |
| Q2 |
0.72 (0.35, 1.47) 0.3624 |
0.74 (0.36, 1.52) 0.4060 |
0.49 (0.22, 1.09) 0.0816 |
|
| Q3 |
0.49 (0.24, 1.02) 0.0560 |
0.51 (0.24, 1.07) 0.0734 |
0.37 (0.16, 0.87) 0.0223 |
|
| Q4 |
0.46 (0.16, 1.37) 0.1659 |
0.49 (0.16, 1.46) 0.1984 |
0.29 (0.09, 0.94) 0.0392 |
|
| p-value | 0.0425 | 0.0578 | 0.0125 | |
| Bedridden group | Barthel Index score |
1.00 (0.98, 1.01) 0.7592 |
1.00 (0.99, 1.02) 0.6711 |
1.00 (0.98, 1.02) 0.9464 |
| Barthel Index score (quartile) | ||||
| Q1 | 1.0 | 1.0 | 1.0 | |
| Q2 |
1.17 (0.64, 2.14) 0.6000 |
1.44 (0.77, 2.69) 0.2506 |
1.11 (0.57, 2.15) 0.7543 |
|
| Q3 |
0.60 (0.27, 1.37) 0.2281 |
0.85 (0.36, 2.01) 0.7082 |
0.90 (0.36, 2.24) 0.8144 |
|
| Q4 | – | – | – | |
| p-value | 0.4925 | 0.8697 | 0.9217 | |
| Community group | Barthel Index score |
0.98 (0.95, 1.01) 0.2136 |
0.98 (0.95, 1.01) 0.2097 |
0.98 (0.95, 1.02) 0.3885 |
| Barthel Index score (quartile) | ||||
| Q1 | 1.0 | 1.0 | 1.0 | |
| Q2 |
0.00 (0.00, Inf) 0.9884 |
0.00 (0.00, Inf) 0.9884 |
0.00 (0.00, Inf) 0.9922 |
|
| Q3 |
0.00 (0.00, Inf) 0.9878 |
0.00 (0.00, Inf) 0.9878 |
0.00 (0.00, Inf) 0.9921 |
|
| Q4 |
0.00 (0.00, Inf) 0.9879 |
0.00 (0.00, Inf) 0.9878 |
0.00 (0.00, Inf) 0.9921 |
|
| p-value | 0.0932 | 0.0904 | 0.2569 | |
Original model: no covariates were adjusted.
Model I: we only adjusted age and sex.
Model II: we adjusted age, sex, Barthel Index, illness duration, living situation, annual family income, and psychotropic substances status.
OR odds ratio, CI confidence interval, Inf infimum.
In the bedridden group, BI was not significantly associated with injurious falls in the Original model (OR = 1.00, 95% CI 0.98–1.01, p = 0.7592), Model I (OR = 1.00, 95% CI 0.99–1.02, p = 0.6711), or Model II (OR = 1.00, 95% CI 0.98–1.02, p = 0.9464). Similarly, in the community group, BI was also not significantly associated with injurious falls in the Original model (OR = 0.98, 95% CI 0.95–1.01, p = 0.2136), Model I (OR = 0.98, 95% CI 0.95–1.01, p = 0.2097), or Model II (OR = 0.98, 95% CI 0.95–1.02, p = 0.3885) (Table 3).
Sensitivity analysis
A series of sensitivity analyses were conducted to validate the robustness of our findings. We converted the BI from continuous to categorical variables (based on quartiles) and then returned the categorically converted BI scores into the model. The results showed that the trends in effect sizes across groups were unequally spaced after converting BI into a categorical variable, suggesting a possible nonlinear relationship between BI and the occurrence of injurious falls in the domestic group (p = 0.0125) (Table 3). In both the bedridden and community groups, sensitivity analysis demonstrated no significant linear relationship between the BI and the occurrence of injurious falls (p > 0.05) (Table 3).
The nonlinearity addressed by the GAM
In the domestic group, after adjusting for age, sex, illness duration, living situation, annual family income, and psychotropic medication status, curve fitting revealed a reverse U-shaped relationship between BI and the occurrence of injurious falls (Fig. 3). Building on the curve-fitting analysis, further threshold effect analysis showed that the likelihood ratio test between Models 1 and 2 had a statistically significant difference (p = 0.011), which further confirmed the curvilinear relationship between BI and the occurrence of injurious falls. Based on the analysis of Model 2, it was observed that before the inflection point of 35, each unit increase in BI was associated with a 14% increased risk of falling (OR = 1.14, 95% CI 1.010–1.29, p = 0.0331). After the inflection point of 35, each unit increase in BI was associated with a 3% decreased risk of injurious falls (OR = 0.97, 95% CI 0.95–0.99, p = 0.0013) (Table 4).
Fig. 3.
Curve-fitting association between Barthel Index and injurious falls in domestic group.
Table 4.
The threshold association between Barthel Index and injurious falls.
| Outcome | OR | 95% CI | p-value |
|---|---|---|---|
| Domestic group | |||
| Model 1 | |||
| Fitting model by standard linear regression | 0.98 | 0.96, 1.00 | 0.0173 |
| Model 2 | |||
| Fitting model by two-piecewise linear regression | |||
| Inflection point | 35 | ||
| < 35 | 1.14 | 1.01, 1.29 | 0.0331 |
| > 35 | 0.97 | 0.95, 0.99 | 0.0013 |
| p for log likelihood ratio test | 0.011 | ||
| Bedridden group | |||
| Model 1 | |||
| Fitting model by standard linear regression | 1.00 | 0.98, 1.02 | 0.9464 |
| Model 2 | |||
| Fitting model by two-piecewise linear regression | |||
| Inflection point | 45 | ||
| < 45 | 0.99 | 0.97, 1.01 | 0.5921 |
| > 45 | 0.97 | 0.94, 1.15 | 0.4342 |
| p for log likelihood ratio test | 0.397 | ||
| Community group | |||
| Model 1 | |||
| Fitting model by standard linear regression | 0.98 | 0.95 ~ 1.00 | 0.3885 |
| Model 2 | |||
| Fitting model by two-piecewise linear regression | |||
| Inflection point | 65 | ||
| < 65 | 1.04 | 0.00, inf | 0.9893 |
| > 65 | 0.97 | 0.97, 1.06 | 0.6333 |
| p for log likelihood ratio test | 0.031 | ||
We adjusted age, sex, Barthel Index, illness duration, living situation, annual family income, and psychotropic substances status.
OR odds ratio, CI confidence interval, Inf infimum.
However, in the bedridden group, no significant nonlinear association between BI and injurious falls was found (p > 0.05). The statistical results indicated an inflection point at 45; to the left of this inflection point, the OR was 0.99 (95% CI 0.97–1.01, p = 0.5921), and to the right, the OR was 0.97 (95% CI 0.94–1.15, p = 0.4342) (Table 4, Fig. 4). Similarly, in the community group, there was also no significant nonlinear association between the both. (p > 0.05). The statistical results indicated an inflection point at 65; to the left of this inflection point, the OR was 1.04 (95% CI 0.00–inf., p = 0.9893), and to the right, the OR was 0.97 (95% CI 0.97–1.06, p = 0.6333) (Table 4, Fig. 5).
Fig. 4.
Curve-fitting association between Barthel Index and injurious falls in bedridden group.
Fig. 5.
Curve-fitting association between Barthel Index and injurious falls in community group.
Discussion
In this cross-sectional study, we found a high risk of injurious falls (74.76%) among older stroke inpatients. The result highlights the importance of remaining vigilant and proactive in preventing such adverse events. Our findings also revealed that the risk of injurious falls is higher among bedridden older stroke patients (53.61%) compared to those restricted to indoor environments and those who can mobilize in the community. This can be attributed to the severe impairment in ADL and long-term bedridden status commonly observed in this group, which makes bedridden patients more prone to muscle weakness, osteoporosis, and other medical conditions, thereby increasing the risk of fractures and other injuries following falls24.
Our findings revealed that the relationship between BI and injurious falls follows an inverted U-shaped curve among domestic older stroke patients, with an inflection point at a BI score of 35. In individuals with BI scores < 35, the injurious fall risk increased rapidly as their BI score increased. Conversely, for those with a BI score of > 35, the risk decreased slowly as the BI score increased. This may be explained by the fact that stroke patients with a BI score of < 35 have severe functional impairment and high levels of dependence. As their ability to perform ADL improve, they may engage in more rehabilitative exercises or self-care activities, such as walking, transfers, and dressing25. Additionally, those who reside in home environments often have more opportunities to participate in ADL, such as personal grooming and cooking26. Owing to significant functional limitations in balance, muscle strength, and motor coordination, the risk of injury following a fall increases when completing those activities.
In contrast, patients with a BI score > 35 tend to have moderate-to-mild dysfunction, indicating a higher level of ability and functioning in daily living. This implies that they have a better ability to walk, maintain balance, and navigate stairs, which reduces their probability of falling and getting injured. A similar trend was observed by Kato et al.27, who examined the relationship between transfer ability and fall risk using the Functional Independence Measure (FIM) scale. They found that patients with stroke with moderate transfer ability (FIM score of 4–6) had a higher fall rate than those with high or low transfer ability (FIM scores of 1 or 7). This finding further supports our observed curvilinear relationship between BI and the risk of injurious falls, although their focus was primarily on fall risk, whereas our study specifically examined injurious falls. Furthermore, as patients’ functional abilities improve, they increasingly engage in activities within the home and community environments and their family members can play an important role in assisting with caregiving tasks, providing vigilance, and reminding them about falls prevention activities28.
Our findings also revealed that no association was found between BI and injurious fall risk in bedridden or community group stroke patients. The occurrence of injuries in individuals bedridden is more likely associated with their movements in bed, such as rolling, turning, or attempting to leave the bed, actions that may not be directly influenced by ADL. Additionally, previous studies indicated that the height of the bed and the falling posture from beds may be related to the severity of fall-related injuries29,30. Therefore, fall-related injury in bedridden individuals may be more related to bed-specific behaviors and environmental factors, which may not have a direct association with the ADL included in BI, such as walking. In the community group, stroke patients have better functional and transfer abilities to engage in social and outdoor activities, leading to higher BI scores. Previous studies have indicated that stroke patients in the community group, assessed using LS, generally exhibit a BI score of 80 or higher, with scores indicating a notable tendency for clustering19,31. This suggests that ADL may not have a significant correlation with the risk of fall-related injuries in a community population.
To the best of our knowledge, this is the first study to examine the association between BI and injurious falls in older stroke patients. Our research findings may assist healthcare personnel in promptly and accurately identifying the stroke patient population at a high risk of fall-related injuries. However, this study had a few limitations. First, this was a cross-sectional study and can only be considered as an initial analysis, necessitating further research to provide confirmatory evidence. In addition, information on injurious falls was collected via questionnaires based on self-reporting or proxy-reporting, which may have introduced a recall bias regarding fall-related injuries. Another limitation is that we did not collect specific information on the location in which the injurious falls occurred. This lack of locational context limits our ability to fully understand the contributing factors to these falls. Finally, although a multivariate analysis was conducted to adjust for potential confounding factors, unmeasured confounders may remain. Further research is needed to incorporate more impact factors related to injurious falls to explore the relationship.
Conclusion
Our study demonstrates that among older stroke patients, only those with sufficient mobility to leave the bed but unable to reach the outdoors exhibit a correlation between ADL and fall-related injuries which further clarifies the association between both. Moreover, we recommend that healthcare professionals pay particular attention to older domestic stroke patients with BI scores of approximately 35 and enhance fall risk screening and education for this subgroup, as it may contribute to reducing the occurrence of injurious falls.
Acknowledgements
We would like to thank all the researchers and participants involved in this study. Furthermore, we would like to thank the Yilanda, Network Technology co., LTD, for developing Quicker Recovery Line platform.
Author contributions
Q.F.Z., J.Y., J.J.L. and Y.L.W. contributed to the conception and design of the study, supervised the surveys and edited the manuscript. D.X.L. and M.C.Z. coordinated the data collection and conducted the statistical analyses. D.R.H., Y.Q.H., L.Z., and M.H.K. provided methodological input. All authors read and approved the final manuscript.
Funding
This study was supported by the National Key R&D Program of China (grant number 2020YFC2008700), and the Sanming Project of Medicine in Shenzhen (grant number SZSM202111010).
Data availability
The datasets generated during and analysed during the current study are available from the corresponding author on reasonable request.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
These authors contributed equally: Qingfang Zhang, Jie Yan, Jianjun Long and Yulong Wang.
Contributor Information
Jianjun Long, Email: longjianjun@szu.edu.cn.
Yulong Wang, Email: ylwang668@163.com.
References
- 1.Karinkanta, S., Kannus, P., Uusi-Rasi, K., Heinonen, A. & Sievänen, H. Combined resistance and balance-jumping exercise reduces older women’s injurious falls and fractures: 5-year follow-up study. Age Ageing44, 784–789 (2015). 10.1093/ageing/afv064 [DOI] [PubMed] [Google Scholar]
- 2.Quigley, P. A. Redesigned fall and injury management of patients with stroke. Stroke47, e92-94 (2016). 10.1161/STROKEAHA.116.012094 [DOI] [PubMed] [Google Scholar]
- 3.Whitney, D. G., Dutt-Mazumder, A., Peterson, M. D. & Krishnan, C. Fall risk in stroke survivors: Effects of stroke plus dementia and reduced motor functional capacity. J. Neurol. Sci.401, 95–100 (2019). 10.1016/j.jns.2019.04.035 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Breisinger, T. P., Skidmore, E. R., Niyonkuru, C., Terhorst, L. & Campbell, G. B. The stroke assessment of fall risk (SAFR): Predictive validity in inpatient stroke rehabilitation. Clin. Rehab.28, 1218–1224 (2014). 10.1177/0269215514534276 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Haddad, Y. K., Bergen, G. & Florence, C. S. Estimating the economic burden related to older adult falls by state. J. Public Health Manag. Pract.25, E17-e24 (2019). 10.1097/PHH.0000000000000816 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Del Din, S. et al. Falls risk in relation to activity exposure in high-risk older adults. J. Gerontol. Ser. A Biol. Sci. Med. Sci. 75, 1198–1205 (2020). [DOI] [PMC free article] [PubMed]
- 7.Ek, S. et al. Predicting first-time injurious falls in older men and women living in the community: Development of the first injurious fall screening tool. J. Am. Med. Direct. Assoc.20, 1163-1168.e1163 (2019). 10.1016/j.jamda.2019.02.023 [DOI] [PubMed] [Google Scholar]
- 8.Dever Fitzgerald, T. et al. The impact of fall risk assessment on nurse fears, patient falls, and functional ability in long-term care. Disability Rehab. 38, 1041–1052 (2016). [DOI] [PubMed]
- 9.Kobayashi, K. et al. Association between fall-related serious injury and activity during fall in an acute care hospital. PloS One.18, e0288320 (2023). 10.1371/journal.pone.0288320 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Gao, Y. et al. Disability assessment in stroke: Relationship among the pictorial-based Longshi Scale, the Barthel Index, and the modified Rankin Scale. Clin. Rehab.35, 606–613 (2021). 10.1177/0269215520975922 [DOI] [PubMed] [Google Scholar]
- 11.Schmid, A. A. et al. Circumstances and consequences of falls among people with chronic stroke. J. Rehab. Res. Dev.50, 1277–1286 (2013). 10.1682/JRRD.2012.11.0215 [DOI] [PubMed] [Google Scholar]
- 12.Zhao, J. et al. A novel Longshi Scale measured activity of daily living disability in elderly patients affected by neurological diseases: A multi-center cross-sectional study in China. BMC Geriatr.21, 348 (2021). 10.1186/s12877-021-02296-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Li, D., Zha, F. & Wang, Y. Association between falling and activities of daily living measured by the Longshi Scale in patients poststroke: A cross-sectional study. J. Nurs. Care Qual.38, E25-e31 (2023). 10.1097/NCQ.0000000000000679 [DOI] [PubMed] [Google Scholar]
- 14.Wang, Y. et al. User testing of the psychometric properties of pictorial-based disability assessment Longshi Scale by healthcare professionals and non-professionals: A Chinese study in Shenzhen. Clin. Rehab.33, 1479–1491 (2019). 10.1177/0269215519846543 [DOI] [PubMed] [Google Scholar]
- 15.Kokotailo, R. A. & Hill, M. D. Coding of stroke and stroke risk factors using international classification of diseases, revisions 9 and 10. Stroke.36, 1776–1781 (2005). 10.1161/01.STR.0000174293.17959.a1 [DOI] [PubMed] [Google Scholar]
- 16.Laberge, S. & Crizzle, A. M. A literature review of psychotropic medications and alcohol as risk factors for falls in community dwelling older adults. Clin. Drug Investig.39, 117–139 (2019). 10.1007/s40261-018-0721-6 [DOI] [PubMed] [Google Scholar]
- 17.Ang, G. C., Low, S. L. & How, C. H. Approach to falls among the elderly in the community. Singapore Med. J.61, 116–121 (2020). 10.11622/smedj.2020029 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Burns, Z. et al. Classification of injurious fall severity in hospitalized adults. J. Gerontol. Ser. A Biol. Sci. Med. Sci. 75, e138–e144 (2020). [DOI] [PubMed]
- 19.Liu, X. et al. Functional independence and disability evaluation in stroke patients: Optimal cutoff scores for a pictorial-based Longshi Scale, Barthel Index, and Modified Rankin Scale. Front. Neurol.13, 710852 (2022). 10.3389/fneur.2022.710852 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Mahoney, F. I. & Barthel, D. W. Functional evaluation: The Barthel index. Md State Med. J.14, 61–65 (1965). [PubMed] [Google Scholar]
- 21.Martínez-Velilla, N. et al. Effect of exercise intervention on functional decline in very elderly patients during acute hospitalization: A randomized clinical trial. JAMA Int. Med.179, 28–36 (2019). 10.1001/jamainternmed.2018.4869 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Al-Khawaja, I., Wade, D. T. & Turner, F. The Barthel Index and its relationship to nursing dependency in rehabilitation. Clin. Rehab.11, 335–337 (1997). 10.1177/026921559701100411 [DOI] [PubMed] [Google Scholar]
- 23.Lin, L., Chen, C. Z., & XD, Y. The analysis of threshold effect using Empower Stats software. Zhonghua Liu Xing Bing Xue Za Zhi. 34, 1139–1141 (2013). [PubMed]
- 24.Eimori, K. et al. Disrupted bone metabolism in long-term bedridden patients. PloS One.11, e0156991 (2016). 10.1371/journal.pone.0156991 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Liu, M. et al. The bidirectional relationship between homebound status and falls among older adults: A longitudinal study. BMC Geriatr.22, 923 (2022). 10.1186/s12877-022-03588-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Currin, M. L., Comans, T. A., Heathcote, K. & Haines, T. P. Staying safe at home. Home environmental audit recommendations and uptake in an older population at high risk of falling. Australas. J. Ageing.31, 90–95 (2012). 10.1111/j.1741-6612.2011.00545.x [DOI] [PubMed] [Google Scholar]
- 27.Kato, Y. et al. Stroke patients with nearly independent transfer ability are at high risk of falling. J. Stroke Cereb. Dis.31, 106169 (2022). 10.1016/j.jstrokecerebrovasdis.2021.106169 [DOI] [PubMed] [Google Scholar]
- 28.Almeida, A. S., Paguia, A. & Neves, A. P. Nursing interventions to empower family caregivers to manage the risk of falling in older adults: A scoping review. Int. J. Environ. Res. Public Health.21, 1 (2024). 10.3390/ijerph21030246 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Tzeng, H. M. & Yin, C. Y. Innovation in patient safety: A new task design in reducing patient falls. J. Nurs. Care Qual.23, 34–42 (2008). 10.1097/01.NCQ.0000303803.07457.e5 [DOI] [PubMed] [Google Scholar]
- 30.Bertocci, G. E. et al. Using test dummy experiments to investigate pediatric injury risk in simulated short-distance falls. Arch. Pediatr. Adolesc. Med.157, 480–486 (2003). 10.1001/archpedi.157.5.480 [DOI] [PubMed] [Google Scholar]
- 31.Zhou, M. et al. Stroke outcome assessment: Optimizing cutoff scores for the Longshi Scale, modified Rankin Scale and Barthel Index. PloS One.16, e0251103 (2021). 10.1371/journal.pone.0251103 [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.
Data Availability Statement
The datasets generated during and analysed during the current study are available from the corresponding author on reasonable request.