The Association Between Fall Frequency, Injury Risk, and Characteristics of Falls in Older Residents of Long-Term Care: Do Recurrent Fallers Fall More Safely?

The Association Between Fall Frequency, Injury Risk, and Characteristics of Falls in Older... Abstract Background Although a fall is a necessary prerequisite to a fall-related injury, previous studies suggest that frequent fallers are at lower injury risk for a given fall. We tested the hypotheses that differences in protective responses or the circumstances of falls underlie differences in injury risk with fall frequency. Methods We analyzed video footage of 897 falls experienced by 220 long-term care residents (mean age 82 ± 9 years) to identify the cause of imbalance, activity leading to falling, direction of fall initiation, balance recovery and fall protective responses, and occurrence of impact to the head or hip. We further obtained injury information from the facilities’ fall registration. We used generalized estimating equation models to examine the association between quartiles of fall frequency, injury risk, and fall characteristics. Results Residents with the highest fall frequency group (Q4; ≥5.6 falls/year) were less likely to sustain an injury per fall. They were less likely to fall during walking and more likely to fall during stand-to-sit transfers. Residents in the lowest fall frequency group (Q1; <1.15 falls/year) were more likely to fall during walking, and walking was associated with an increased risk for injury. Conclusion When compared to less frequent fallers, more frequent fallers had a lower risk for injury per fall. This appeared to be explained by differences in the circumstances of falls, and not by protective responses. Injury prevention strategies in long-term care should target both frequent and infrequent fallers, as the latter are more mobile and apt to sustain injury. Accidental falls, Injury, Aged 65 and over, Long-term care, Frequent fallers About 60% of long-term care (LTC) residents falls at least once annually, and with an average frequency of 1.7 falls per year, most residents will fall recurrently (1–3). Falls are the number-one cause of injuries in adults over the age of 65, including 90% of hip fractures (4) and 70% of traumatic brain injuries (5). Nevertheless, only about 20%–30% of falls in older adults actually results in injury (6–9). An improved understanding of factors that separate injurious and noninjurious falls would serve to guide prevention efforts. A fall is clearly a prerequisite for experiencing a fall-related injury. Several studies have shown that the cumulative risk for injury is higher among older adults who fall recurrently compared to incidental fallers (10–13). Furthermore, falls and fall-related injuries have shared risk factors, suggesting that frequent fallers are more likely to experience an injury in the event of a fall. For example, female gender, slowed reaction time, impaired cognition, reduced muscle strength and the use of psychoactive medications are associated with both increased risk for falling and injury (7,11,14). Presumably, this is due to the fact that these factors also influence the characteristics (and severity) of falls. However, although counter intuitive, evidence suggests that a given fall is less likely to cause injury in a frequent faller when compared to a less frequent faller. For example, Baranzini and colleagues (15) reported that frequent fallers in LTC (≥4 falls/year) were less likely to sustain an injury in falling than less frequent fallers (odds ratio [OR] = 0.42, 95% confidence interval = 0.20–0.89). Furthermore, Tinetti (12) found that frequent fallers in LTC (≥4 falls/year or fell every 2 months) had a relative risk of 0.6 for experiencing an injury compared to those who fell less often. Nevitt and colleagues (11) observed similar trends among community-dwelling fallers but the differences were not statistically significant. While the reasons for these trends are unclear, several explanations have been offered. First, frequent fallers may use more effective protective responses to avoid injury during a fall (11). For example, falling on an outstretched arm decreases the risk for hip fractures (16,17) and body rotation during descent decreases the risk for head and hip impact (18,19). Second, the circumstances and dynamics of falls might be different for frequent fallers. For example, falls in frequent fallers may be from a lower height (eg, sitting vs standing) and involve a smaller change in potential energy and lower risk for injury. Third, individuals at greatest risk for injury (eg, due to osteoporosis or impaired protective responses) may demonstrate greater fear of falling, which often leads to activity restriction (eg, (20)). The activity restriction may lead to a reduction in falls, but the low tissue tolerance or impaired protective responses lead to an increased risk for injury in the event of a fall. The aim of the current study was to test the hypotheses that the frequency of falls in LTC residents associates with injury risk during falling, and with the characteristics of falls. To address this aim, we analyzed a unique video dataset of real-life falls sustained by older adults who lived in two LTC facilities. Methods Falls were recorded on video in common areas of two LTC facilities in Greater Vancouver, as described in Baranzini and colleagues (21). The research ethics boards at Simon Fraser University and Fraser Health Authority approved the study. At the time of admission to LTC, each resident or proxy provided written informed consent for video recording. For the current study we analyzed falls experience between April 20, 2007 and September 21, 2015 by residents from whom we had consent to access their medical records. Fall Frequency, Group Demographics and Consequences of Falls Each facility used a standardized fall incidence report from which we accessed the total number of falls experienced by each study participant and data on injuries due to these falls. The report documents the time and location of the fall, a narrative description of the fall and information on injuries associated with the fall. The admission date (and where relevant discharge date), as well as observational tests of physical and cognitive status, disease diagnoses and medication use, were obtained from the facility’s minimum data set (MDS) health records. Fall frequency was calculated as total number of falls divided by the length of stay. Characteristics of Falls and Protective Responses Each fall video was analyzed by a three-member research team who sought consensus on the characteristics of the fall using a validated questionnaire (22,23). In the current analysis, we focused on characteristics of fall initiation (nature of imbalance, activity at the time of the fall, initial fall direction), fall descent (presence of grasping or stepping responses for balance recovery, axial rotation during descent, upper limb fall arrest) and fall impact (sites of impact on the body). Statistics We first determined whether fall frequency was associated with fall-related injury rate using Spearman correlation. Second, we determined the association between fall frequency and characteristics of falls. Visualization of the data suggested a nonlinear association; hence, residents were categorized into four quartiles based on fall frequency. We performed chi-square and ANOVA tests to identify differences in group demographics. We subsequently used generalized estimating equation models to estimate OR. To reduce the possibility that residents who fell very frequently would unduly influence our findings, we employed a bootstrapping approach where we randomly sampled a maximum of 20 falls per resident to be included and averaged results over 1,000 repetitions. All statistics were performed in SPSS (IBM, version 23) and a p-value of .05 was considered statistically significant. Results Group Demographics We analyzed falls from 220 residents with a mean age of 82.1 (SD = 8.8) years. Descriptive characteristics are provided in Table 1. For 69% of participants (n = 152), we had MDS data within 6 months of the fall date. Participants scored an average of 2.9 (SD 1.7) out of 6 on the activities of daily living (ADL) self-performance hierarchy scale and 3.4 (SD 1.6) out of 6 on the cognitive performance scale, indicating the average resident required assistance in performing ADLs and had a moderate cognitive impairment (24,25). The 220 residents experienced a total of 3,573 falls; 897 were captured on video. The observation period ranged from 94 to 2,751 days, with a mean of 1,380 (SD = 704) days. Fall frequencies ranged from 0.15 to 39.2 falls per year (median 2.5, IQR 4.4; Figure 1). Fall frequencies were categorized into four equally sized quartiles with 1.15, 2.5, and 5.6 falls per year as cut points. We will refer to these groups as lowest (Q1), low–medium (Q2), medium–high (Q3), and highest (Q4) fall frequency groups. Table 1. Descriptive Characteristics of the Fall Frequency Groups Total Fall Frequency Group p-Value Lowest (Q1) Low–Medium Medium–High Highest (Q4) Number of residents 220 55 55 55 55 N/A Number of video-captured falls 897 76 129 193 499 N/A Number of video-captured falls with injury data# 762 70 115 144 433 N/A Number of sustained injuries 177 24 38 37 78 N/A Demographics and general health  Mean fall rate (number/year) 5.40 (6.79) 0.62 (0.27)± 1.74 (0.40)± 3.88 (0.96)± 13.16 (8.44)± <0.0001  Mean injurious fall rate (number/year) 0.28 (0.55) 0.15 (0.21)± 0.22 (0.31)§ 0.28 (0.39)& 0.54 (0.92)±,§,& <0.0001  Mean length of stay (years) 3.76 (1.94) 3.59 (1.93) 4.27 (1.70) 3.74 (2.02) 3.51 (2.01) 0.165  Age at time of fall (years) 82.1 (8.8) 83.3 (7.9) 79.9 (9.3) 80.6 (8.5) 82.6 (9.1) 0.780  Female gender (%) 62.5 78.6 63.4 56.1 57.1 0.224  BMI (kg/m2) 24.7 (4.9) 24.7 (4.7) 25.6 (5.0) 24.6 (4.6) 23.7 (5.4) 0.492  ADL at time of fall (score) 2.9 (1.7) 2.7 (2.0) 3.6 (1.7) 3.0 (1.5) 3.3 (1.4) 0.443  CPS at time of fall (score) 3.4 (1.6) 3.0 (1.3) 3.8 (1.8) 3.5 (1.4) 4.0 (1.4) 0.443 Disease diagnosis  Alzheimer’s disease (%) 29.6 25.0 24.4 24.4 42.9 0.206  Diabetes (%) 21.7 21.4 17.1 19.5 28.6 0.632  Cardiac arrhythmia (%) 4.6 3.6 2.4 4.9 7.1 0.882  Hypertension (%) 40.8 53.6 34.1 48.8 31.0 0.151  Stroke (%) 11.8 3.6 7.3 24.4 9.5 0.044  Parkinson’s disease (%) 2.6 3.6 0.0 0.0 7.1 0.107  COPD (%) 11.8 7.1 9.8 12.2 16.7 0.670 Use of medication  Number of medications 8.76 (3.99) 7.96 (4.25) 8.15 (3.46) 9.71 (3.30) 8.98 (4.77) 0.212  Antipsychotic (%) 41.4 35.7 24.4± 48.8 54.8± 0.025  Antianxiety (%) 21.1 3.6±§ 14.6 29.3± 31.0§ 0.011  Antidepressant (%) 48.0 32.1 48.8 53.7 52.4 0.303  Hypnotic (%) 15.8 17.9 12.2 17.1 16.7 0.915  Diuretic (%) 25.0 17.9 19.5 36.6 23.8 0.254  Analgesics (%) 52.6 46.4 46.3 63.4 52.4 0.402 Total Fall Frequency Group p-Value Lowest (Q1) Low–Medium Medium–High Highest (Q4) Number of residents 220 55 55 55 55 N/A Number of video-captured falls 897 76 129 193 499 N/A Number of video-captured falls with injury data# 762 70 115 144 433 N/A Number of sustained injuries 177 24 38 37 78 N/A Demographics and general health  Mean fall rate (number/year) 5.40 (6.79) 0.62 (0.27)± 1.74 (0.40)± 3.88 (0.96)± 13.16 (8.44)± <0.0001  Mean injurious fall rate (number/year) 0.28 (0.55) 0.15 (0.21)± 0.22 (0.31)§ 0.28 (0.39)& 0.54 (0.92)±,§,& <0.0001  Mean length of stay (years) 3.76 (1.94) 3.59 (1.93) 4.27 (1.70) 3.74 (2.02) 3.51 (2.01) 0.165  Age at time of fall (years) 82.1 (8.8) 83.3 (7.9) 79.9 (9.3) 80.6 (8.5) 82.6 (9.1) 0.780  Female gender (%) 62.5 78.6 63.4 56.1 57.1 0.224  BMI (kg/m2) 24.7 (4.9) 24.7 (4.7) 25.6 (5.0) 24.6 (4.6) 23.7 (5.4) 0.492  ADL at time of fall (score) 2.9 (1.7) 2.7 (2.0) 3.6 (1.7) 3.0 (1.5) 3.3 (1.4) 0.443  CPS at time of fall (score) 3.4 (1.6) 3.0 (1.3) 3.8 (1.8) 3.5 (1.4) 4.0 (1.4) 0.443 Disease diagnosis  Alzheimer’s disease (%) 29.6 25.0 24.4 24.4 42.9 0.206  Diabetes (%) 21.7 21.4 17.1 19.5 28.6 0.632  Cardiac arrhythmia (%) 4.6 3.6 2.4 4.9 7.1 0.882  Hypertension (%) 40.8 53.6 34.1 48.8 31.0 0.151  Stroke (%) 11.8 3.6 7.3 24.4 9.5 0.044  Parkinson’s disease (%) 2.6 3.6 0.0 0.0 7.1 0.107  COPD (%) 11.8 7.1 9.8 12.2 16.7 0.670 Use of medication  Number of medications 8.76 (3.99) 7.96 (4.25) 8.15 (3.46) 9.71 (3.30) 8.98 (4.77) 0.212  Antipsychotic (%) 41.4 35.7 24.4± 48.8 54.8± 0.025  Antianxiety (%) 21.1 3.6±§ 14.6 29.3± 31.0§ 0.011  Antidepressant (%) 48.0 32.1 48.8 53.7 52.4 0.303  Hypnotic (%) 15.8 17.9 12.2 17.1 16.7 0.915  Diuretic (%) 25.0 17.9 19.5 36.6 23.8 0.254  Analgesics (%) 52.6 46.4 46.3 63.4 52.4 0.402 Note: Values are based on the status at the time of the first fall and are reported as mean (standard deviation) or percentage. p-Values are based on Fisher’s exact test (categorical) or ANOVA (continuous). Bolded p-values indicate a p < .05. ±, § and & indicate Bonferroni-corrected significant differences in post hoc group-wise comparisons. #Falls where the incident report confirmed that either no injury occurred, or an injury occurred. View Large Table 1. Descriptive Characteristics of the Fall Frequency Groups Total Fall Frequency Group p-Value Lowest (Q1) Low–Medium Medium–High Highest (Q4) Number of residents 220 55 55 55 55 N/A Number of video-captured falls 897 76 129 193 499 N/A Number of video-captured falls with injury data# 762 70 115 144 433 N/A Number of sustained injuries 177 24 38 37 78 N/A Demographics and general health  Mean fall rate (number/year) 5.40 (6.79) 0.62 (0.27)± 1.74 (0.40)± 3.88 (0.96)± 13.16 (8.44)± <0.0001  Mean injurious fall rate (number/year) 0.28 (0.55) 0.15 (0.21)± 0.22 (0.31)§ 0.28 (0.39)& 0.54 (0.92)±,§,& <0.0001  Mean length of stay (years) 3.76 (1.94) 3.59 (1.93) 4.27 (1.70) 3.74 (2.02) 3.51 (2.01) 0.165  Age at time of fall (years) 82.1 (8.8) 83.3 (7.9) 79.9 (9.3) 80.6 (8.5) 82.6 (9.1) 0.780  Female gender (%) 62.5 78.6 63.4 56.1 57.1 0.224  BMI (kg/m2) 24.7 (4.9) 24.7 (4.7) 25.6 (5.0) 24.6 (4.6) 23.7 (5.4) 0.492  ADL at time of fall (score) 2.9 (1.7) 2.7 (2.0) 3.6 (1.7) 3.0 (1.5) 3.3 (1.4) 0.443  CPS at time of fall (score) 3.4 (1.6) 3.0 (1.3) 3.8 (1.8) 3.5 (1.4) 4.0 (1.4) 0.443 Disease diagnosis  Alzheimer’s disease (%) 29.6 25.0 24.4 24.4 42.9 0.206  Diabetes (%) 21.7 21.4 17.1 19.5 28.6 0.632  Cardiac arrhythmia (%) 4.6 3.6 2.4 4.9 7.1 0.882  Hypertension (%) 40.8 53.6 34.1 48.8 31.0 0.151  Stroke (%) 11.8 3.6 7.3 24.4 9.5 0.044  Parkinson’s disease (%) 2.6 3.6 0.0 0.0 7.1 0.107  COPD (%) 11.8 7.1 9.8 12.2 16.7 0.670 Use of medication  Number of medications 8.76 (3.99) 7.96 (4.25) 8.15 (3.46) 9.71 (3.30) 8.98 (4.77) 0.212  Antipsychotic (%) 41.4 35.7 24.4± 48.8 54.8± 0.025  Antianxiety (%) 21.1 3.6±§ 14.6 29.3± 31.0§ 0.011  Antidepressant (%) 48.0 32.1 48.8 53.7 52.4 0.303  Hypnotic (%) 15.8 17.9 12.2 17.1 16.7 0.915  Diuretic (%) 25.0 17.9 19.5 36.6 23.8 0.254  Analgesics (%) 52.6 46.4 46.3 63.4 52.4 0.402 Total Fall Frequency Group p-Value Lowest (Q1) Low–Medium Medium–High Highest (Q4) Number of residents 220 55 55 55 55 N/A Number of video-captured falls 897 76 129 193 499 N/A Number of video-captured falls with injury data# 762 70 115 144 433 N/A Number of sustained injuries 177 24 38 37 78 N/A Demographics and general health  Mean fall rate (number/year) 5.40 (6.79) 0.62 (0.27)± 1.74 (0.40)± 3.88 (0.96)± 13.16 (8.44)± <0.0001  Mean injurious fall rate (number/year) 0.28 (0.55) 0.15 (0.21)± 0.22 (0.31)§ 0.28 (0.39)& 0.54 (0.92)±,§,& <0.0001  Mean length of stay (years) 3.76 (1.94) 3.59 (1.93) 4.27 (1.70) 3.74 (2.02) 3.51 (2.01) 0.165  Age at time of fall (years) 82.1 (8.8) 83.3 (7.9) 79.9 (9.3) 80.6 (8.5) 82.6 (9.1) 0.780  Female gender (%) 62.5 78.6 63.4 56.1 57.1 0.224  BMI (kg/m2) 24.7 (4.9) 24.7 (4.7) 25.6 (5.0) 24.6 (4.6) 23.7 (5.4) 0.492  ADL at time of fall (score) 2.9 (1.7) 2.7 (2.0) 3.6 (1.7) 3.0 (1.5) 3.3 (1.4) 0.443  CPS at time of fall (score) 3.4 (1.6) 3.0 (1.3) 3.8 (1.8) 3.5 (1.4) 4.0 (1.4) 0.443 Disease diagnosis  Alzheimer’s disease (%) 29.6 25.0 24.4 24.4 42.9 0.206  Diabetes (%) 21.7 21.4 17.1 19.5 28.6 0.632  Cardiac arrhythmia (%) 4.6 3.6 2.4 4.9 7.1 0.882  Hypertension (%) 40.8 53.6 34.1 48.8 31.0 0.151  Stroke (%) 11.8 3.6 7.3 24.4 9.5 0.044  Parkinson’s disease (%) 2.6 3.6 0.0 0.0 7.1 0.107  COPD (%) 11.8 7.1 9.8 12.2 16.7 0.670 Use of medication  Number of medications 8.76 (3.99) 7.96 (4.25) 8.15 (3.46) 9.71 (3.30) 8.98 (4.77) 0.212  Antipsychotic (%) 41.4 35.7 24.4± 48.8 54.8± 0.025  Antianxiety (%) 21.1 3.6±§ 14.6 29.3± 31.0§ 0.011  Antidepressant (%) 48.0 32.1 48.8 53.7 52.4 0.303  Hypnotic (%) 15.8 17.9 12.2 17.1 16.7 0.915  Diuretic (%) 25.0 17.9 19.5 36.6 23.8 0.254  Analgesics (%) 52.6 46.4 46.3 63.4 52.4 0.402 Note: Values are based on the status at the time of the first fall and are reported as mean (standard deviation) or percentage. p-Values are based on Fisher’s exact test (categorical) or ANOVA (continuous). Bolded p-values indicate a p < .05. ±, § and & indicate Bonferroni-corrected significant differences in post hoc group-wise comparisons. #Falls where the incident report confirmed that either no injury occurred, or an injury occurred. View Large Figure 1. View largeDownload slide Fall frequency in quartiles. The black line indicates the fall frequency of each participant and gray bars indicate the quartiles; the numbers above the quartile bars provide the mean fall frequency among participants in that fall frequency group. Figure 1. View largeDownload slide Fall frequency in quartiles. The black line indicates the fall frequency of each participant and gray bars indicate the quartiles; the numbers above the quartile bars provide the mean fall frequency among participants in that fall frequency group. The four fall frequency groups did not differ significantly in age, gender, length of stay, BMI, ADL self-performance hierarchy, and cognitive performance scale scores (all p > .224; Table 1), but differed significantly in rates of falls and fall-related injuries, the prevalence of stroke, and the use of antipsychotic and antianxiety medications (all p < .044). Residents in the higher frequency groups fell more often and experienced cumulatively more injuries. Despite a significant main effect of group on the prevalence of stroke, this effect did not persist in post hoc comparisons. The use of antipsychotics was higher in the highest fall frequency group compared to the low-medium fall frequency group (54.8% vs 24.6%). The medium-high and highest fall frequency groups used more antianxiety medications compared to the lowest fall frequency group (respectively, 29.3% and 31.0% vs 3.6%). These differences between groups did not lead to confounding in subsequent analyses (tested as a <10% change in regression coefficients) hence uncorrected associations are reported. Consequences of Falls We were unable to verify whether or not an injury occurred in 140 of the 897 falls captured on video (15.5%), as no response was provided to the question “did an injury occur (YES/ NO)?” in the corresponding fall incident reports. We excluded these cases from our analysis of the consequences of falls. For the remaining 762 falls, injuries were reported in 23.2% of cases (n = 177). Injuries in order of most to least common were lacerations or abrasions (50.7% of cases), swelling or hematomas (29.4%), localized pain without superficial signs (14.0%), bone fracture (2.9%), and joint sprain or dislocation (2.9%). Anatomical sites of injury were the head (50.4%), elbow (13.6%), knee/shin (8.8%), hand/wrists (8.0%), shoulder/arm (7.2%), spine/back (5.6%), hip/thigh (4.8%), and pelvis/tailbone (0.8%). Fall frequency associated moderately with injury rate (r = .48, p < .001). Highest frequency fallers were less likely than lowest (OR = 0.44, 95% CI = 0.24–0.81) and low–medium frequency fallers (OR = 0.46, 95% CI = 0.28–0.78) to have an injury reported for any given fall (Table 2). Of the seven falls that resulted in hip fractures and five falls that resulted in wrist sprains, none occurred among the highest fall frequency group. Table 2. Odds Ratios for Fall Characteristics as a Function of Fall Incidence, Based on Comparison of the Low Fall Frequency Group (Q1) to All Other Groups (Q2–Q4) Characteristics of Falls Prevalence, % of All Falls Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 56.8 0.129 0.83 (0.45–1.55) 0.83 (0.45–1.53) 1.26 (0.71–2.23)  Stand–sit transfer 17.5 0.018 0.95 (0.38–2.41) 0.97 (0.38–2.49) 2.76 (1.29–5.88)*  Sit–stand transfer 10.7 0.714 1.37 (0.42–4.50) 1.99 (0.60–6.58) 1.56 (0.51–4.79)  Sitting 14.2 0.197 2.85 (0.96–8.41) 1.89 (0.66–5.42) 2.65 (0.97–7.29)  Standing 24.0 0.766 1.36 (0.70–2.65) 1.03 (0.53–2.04) 1.01 (0.55–1.85)  Walking 33.6 0.005 0.44 (0.23–0.84)* 0.59 (0.31–1.11) 0.35 (0.19–0.63)*  Initially forward 19.3 0.055 1.05 (0.51–2.17) 0.60 (0.29–1.23) 0.47 (0.23–0.96)  Initially backward 32.2 0.094 0.88 (0.40–1.94) 1.54 (0.72–3.27) 1.71 (0.87–3.36)  Initially sideward 29.1 0.727 1.11 (0.57–2.17) 1.12 (0.60–2.11) 0.84 (0.46–1.53)  Initially straight down 19.3 0.461 0.93 (0.42–2.06) 0.84 (0.39–1.79) 1.35 (0.68–2.67) Descent phase  Stepping 49.8 0.005 0.56 (0.29–1.08) 0.57 (0.30–1.07) 0.33 (0.18–0.61)*  Grasping 24.2 0.265 0.59 (0.31–1.12) 0.55 (0.29–1.02) 0.66 (0.37–1.17)  Turning during descent 36.7 0.500 0.91 (0.46–1.78) 0.77 (0.41–1.46) 0.68 (0.38–1.21)  Attempted upper limb arrest 88.0 0.416 0.54 (0.18–1.62) 0.95 (0.32–2.81) 0.61 (0.22–1.70) Impact phase  Head impact 33.6 0.399 0.75 (0.37–1.54) 0.93 (0.45–1.93) 0.61 (0.30–1.22)  Hip impact 40.8 0.052 1.55 (0.81–2.97) 1.35 (0.71–2.57) 0.93 (0.53–1.62)  Sustained injury 23.2 0.001 0.94 (0.48–1.85) 0.66 (0.34–1.32) 0.44 (0.24–0.81)* Characteristics of Falls Prevalence, % of All Falls Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 56.8 0.129 0.83 (0.45–1.55) 0.83 (0.45–1.53) 1.26 (0.71–2.23)  Stand–sit transfer 17.5 0.018 0.95 (0.38–2.41) 0.97 (0.38–2.49) 2.76 (1.29–5.88)*  Sit–stand transfer 10.7 0.714 1.37 (0.42–4.50) 1.99 (0.60–6.58) 1.56 (0.51–4.79)  Sitting 14.2 0.197 2.85 (0.96–8.41) 1.89 (0.66–5.42) 2.65 (0.97–7.29)  Standing 24.0 0.766 1.36 (0.70–2.65) 1.03 (0.53–2.04) 1.01 (0.55–1.85)  Walking 33.6 0.005 0.44 (0.23–0.84)* 0.59 (0.31–1.11) 0.35 (0.19–0.63)*  Initially forward 19.3 0.055 1.05 (0.51–2.17) 0.60 (0.29–1.23) 0.47 (0.23–0.96)  Initially backward 32.2 0.094 0.88 (0.40–1.94) 1.54 (0.72–3.27) 1.71 (0.87–3.36)  Initially sideward 29.1 0.727 1.11 (0.57–2.17) 1.12 (0.60–2.11) 0.84 (0.46–1.53)  Initially straight down 19.3 0.461 0.93 (0.42–2.06) 0.84 (0.39–1.79) 1.35 (0.68–2.67) Descent phase  Stepping 49.8 0.005 0.56 (0.29–1.08) 0.57 (0.30–1.07) 0.33 (0.18–0.61)*  Grasping 24.2 0.265 0.59 (0.31–1.12) 0.55 (0.29–1.02) 0.66 (0.37–1.17)  Turning during descent 36.7 0.500 0.91 (0.46–1.78) 0.77 (0.41–1.46) 0.68 (0.38–1.21)  Attempted upper limb arrest 88.0 0.416 0.54 (0.18–1.62) 0.95 (0.32–2.81) 0.61 (0.22–1.70) Impact phase  Head impact 33.6 0.399 0.75 (0.37–1.54) 0.93 (0.45–1.93) 0.61 (0.30–1.22)  Hip impact 40.8 0.052 1.55 (0.81–2.97) 1.35 (0.71–2.57) 0.93 (0.53–1.62)  Sustained injury 23.2 0.001 0.94 (0.48–1.85) 0.66 (0.34–1.32) 0.44 (0.24–0.81)* Note: Bolded p-values indicate a p < .05. *OR significantly different from 1 at p < .05. View Large Table 2. Odds Ratios for Fall Characteristics as a Function of Fall Incidence, Based on Comparison of the Low Fall Frequency Group (Q1) to All Other Groups (Q2–Q4) Characteristics of Falls Prevalence, % of All Falls Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 56.8 0.129 0.83 (0.45–1.55) 0.83 (0.45–1.53) 1.26 (0.71–2.23)  Stand–sit transfer 17.5 0.018 0.95 (0.38–2.41) 0.97 (0.38–2.49) 2.76 (1.29–5.88)*  Sit–stand transfer 10.7 0.714 1.37 (0.42–4.50) 1.99 (0.60–6.58) 1.56 (0.51–4.79)  Sitting 14.2 0.197 2.85 (0.96–8.41) 1.89 (0.66–5.42) 2.65 (0.97–7.29)  Standing 24.0 0.766 1.36 (0.70–2.65) 1.03 (0.53–2.04) 1.01 (0.55–1.85)  Walking 33.6 0.005 0.44 (0.23–0.84)* 0.59 (0.31–1.11) 0.35 (0.19–0.63)*  Initially forward 19.3 0.055 1.05 (0.51–2.17) 0.60 (0.29–1.23) 0.47 (0.23–0.96)  Initially backward 32.2 0.094 0.88 (0.40–1.94) 1.54 (0.72–3.27) 1.71 (0.87–3.36)  Initially sideward 29.1 0.727 1.11 (0.57–2.17) 1.12 (0.60–2.11) 0.84 (0.46–1.53)  Initially straight down 19.3 0.461 0.93 (0.42–2.06) 0.84 (0.39–1.79) 1.35 (0.68–2.67) Descent phase  Stepping 49.8 0.005 0.56 (0.29–1.08) 0.57 (0.30–1.07) 0.33 (0.18–0.61)*  Grasping 24.2 0.265 0.59 (0.31–1.12) 0.55 (0.29–1.02) 0.66 (0.37–1.17)  Turning during descent 36.7 0.500 0.91 (0.46–1.78) 0.77 (0.41–1.46) 0.68 (0.38–1.21)  Attempted upper limb arrest 88.0 0.416 0.54 (0.18–1.62) 0.95 (0.32–2.81) 0.61 (0.22–1.70) Impact phase  Head impact 33.6 0.399 0.75 (0.37–1.54) 0.93 (0.45–1.93) 0.61 (0.30–1.22)  Hip impact 40.8 0.052 1.55 (0.81–2.97) 1.35 (0.71–2.57) 0.93 (0.53–1.62)  Sustained injury 23.2 0.001 0.94 (0.48–1.85) 0.66 (0.34–1.32) 0.44 (0.24–0.81)* Characteristics of Falls Prevalence, % of All Falls Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 56.8 0.129 0.83 (0.45–1.55) 0.83 (0.45–1.53) 1.26 (0.71–2.23)  Stand–sit transfer 17.5 0.018 0.95 (0.38–2.41) 0.97 (0.38–2.49) 2.76 (1.29–5.88)*  Sit–stand transfer 10.7 0.714 1.37 (0.42–4.50) 1.99 (0.60–6.58) 1.56 (0.51–4.79)  Sitting 14.2 0.197 2.85 (0.96–8.41) 1.89 (0.66–5.42) 2.65 (0.97–7.29)  Standing 24.0 0.766 1.36 (0.70–2.65) 1.03 (0.53–2.04) 1.01 (0.55–1.85)  Walking 33.6 0.005 0.44 (0.23–0.84)* 0.59 (0.31–1.11) 0.35 (0.19–0.63)*  Initially forward 19.3 0.055 1.05 (0.51–2.17) 0.60 (0.29–1.23) 0.47 (0.23–0.96)  Initially backward 32.2 0.094 0.88 (0.40–1.94) 1.54 (0.72–3.27) 1.71 (0.87–3.36)  Initially sideward 29.1 0.727 1.11 (0.57–2.17) 1.12 (0.60–2.11) 0.84 (0.46–1.53)  Initially straight down 19.3 0.461 0.93 (0.42–2.06) 0.84 (0.39–1.79) 1.35 (0.68–2.67) Descent phase  Stepping 49.8 0.005 0.56 (0.29–1.08) 0.57 (0.30–1.07) 0.33 (0.18–0.61)*  Grasping 24.2 0.265 0.59 (0.31–1.12) 0.55 (0.29–1.02) 0.66 (0.37–1.17)  Turning during descent 36.7 0.500 0.91 (0.46–1.78) 0.77 (0.41–1.46) 0.68 (0.38–1.21)  Attempted upper limb arrest 88.0 0.416 0.54 (0.18–1.62) 0.95 (0.32–2.81) 0.61 (0.22–1.70) Impact phase  Head impact 33.6 0.399 0.75 (0.37–1.54) 0.93 (0.45–1.93) 0.61 (0.30–1.22)  Hip impact 40.8 0.052 1.55 (0.81–2.97) 1.35 (0.71–2.57) 0.93 (0.53–1.62)  Sustained injury 23.2 0.001 0.94 (0.48–1.85) 0.66 (0.34–1.32) 0.44 (0.24–0.81)* Note: Bolded p-values indicate a p < .05. *OR significantly different from 1 at p < .05. View Large Hip and head impacts were observed in 40.8% and 33.6% of falls, respectively. Neither hip nor head impact were significantly associated with fall frequency group (p = .052 and .399, respectively; Table 2). Characteristics of Falls and Protective Responses The most common cause of imbalance preceding a fall was incorrect weight transfer, accounting for 49.5% of falls. Other causes were loss of support from an external object (22.0%), trip or stumble (12.5%), hit or bump (8.8%), collapse (6.7%), and slip (0.5%). There was no difference in the cause of imbalance, or for intrinsic vs extrinsic causes, among the four fall frequency groups (all p > .129; Table 2). Most falls were initially directed backward (33.8%) or sideways (28.8%). Less often, falls were straight down (19.2%) and forward (18.2%). Although not statistically significant (p = .055), there was a trend toward highest frequency fallers falling less often forward compared to lowest and low–medium frequency fallers (Table 2). Analysis of only falls during walking revealed that highest frequency fallers fell more often backwards during walking compared to lowest frequency fallers (OR = 3.19, 95% CI = 1.09–9.35; Table 3). Table 3. Odds Ratios for Characteristics of Falls During Walking as a Function of Fall Incidence, Based on Comparison of the Low Fall Frequency Group (Q1) to All Other Groups (Q2–Q4) Characteristics of Falls During Walking Prevalence, % Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 55.0 0.352 0.33 (0.12–0.93) 0.64 (0.27–1.52) 1.09 (0.50–2.39)  Initially forward 30.0 0.292 1.30 (0.46–3.65) 0.71 (0.28–1.77) 0.41 (0.18–0.94)  Initially backward 20.2 0.028 0.52 (0.09–2.90) 1.82 (0.49–6.73) 3.19 (1.09–9.35)*  Initially sideward 43.0 0.124 1.33 (0.48–3.69) 1.42 (0.61–3.35) 1.20 (0.55–2.62)  Initially straight down 6.8 0.208 0.22 (0.03–2.05) 0.23 (0.04–1.18) 0.73 (0.21–2.55) Descent phase  Stepping 85.0 0.302 0.93 (0.17–5.07) 0.65 (0.16–2.73) 0.38 (0.10–1.44)  Grasping 34.5 0.178 0.71 (0.28–1.80) 0.49 (0.24–1.02) 0.92 (0.45–1.87)  Turning during descent 45.9 0.683 0.90 (0.32–2.52) 0.82 (0.33–2.02) 0.68 (0.31–1.51)  Attempted upper limb arrest 91.8 0.129 2.97 (0.33–26.67) 6.41 (0.72–56.72) 0.80 (0.21–2.99) Impact phase  Head impact 45.1 0.567 1.29 (0.52–3.23) 1.12 (0.46–2.75) 0.96 (0.41–2.28)  Hip impact 52.7 0.064 1.67 (0.62–4.52) 2.03 (0.94–4.38) 0.89 (0.46–1.74)  Sustained injury 32.8 0.020 0.79 (0.28–2.19) 0.61 (0.25–1.50) 0.34 (0.15–0.74)* Characteristics of Falls During Walking Prevalence, % Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 55.0 0.352 0.33 (0.12–0.93) 0.64 (0.27–1.52) 1.09 (0.50–2.39)  Initially forward 30.0 0.292 1.30 (0.46–3.65) 0.71 (0.28–1.77) 0.41 (0.18–0.94)  Initially backward 20.2 0.028 0.52 (0.09–2.90) 1.82 (0.49–6.73) 3.19 (1.09–9.35)*  Initially sideward 43.0 0.124 1.33 (0.48–3.69) 1.42 (0.61–3.35) 1.20 (0.55–2.62)  Initially straight down 6.8 0.208 0.22 (0.03–2.05) 0.23 (0.04–1.18) 0.73 (0.21–2.55) Descent phase  Stepping 85.0 0.302 0.93 (0.17–5.07) 0.65 (0.16–2.73) 0.38 (0.10–1.44)  Grasping 34.5 0.178 0.71 (0.28–1.80) 0.49 (0.24–1.02) 0.92 (0.45–1.87)  Turning during descent 45.9 0.683 0.90 (0.32–2.52) 0.82 (0.33–2.02) 0.68 (0.31–1.51)  Attempted upper limb arrest 91.8 0.129 2.97 (0.33–26.67) 6.41 (0.72–56.72) 0.80 (0.21–2.99) Impact phase  Head impact 45.1 0.567 1.29 (0.52–3.23) 1.12 (0.46–2.75) 0.96 (0.41–2.28)  Hip impact 52.7 0.064 1.67 (0.62–4.52) 2.03 (0.94–4.38) 0.89 (0.46–1.74)  Sustained injury 32.8 0.020 0.79 (0.28–2.19) 0.61 (0.25–1.50) 0.34 (0.15–0.74)* Note: Bolded p-values indicate a p < .05, underlined p-values highlight differences from Table 2 containing the analysis of all falls. *OR significantly different from 1 at p < .05. View Large Table 3. Odds Ratios for Characteristics of Falls During Walking as a Function of Fall Incidence, Based on Comparison of the Low Fall Frequency Group (Q1) to All Other Groups (Q2–Q4) Characteristics of Falls During Walking Prevalence, % Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 55.0 0.352 0.33 (0.12–0.93) 0.64 (0.27–1.52) 1.09 (0.50–2.39)  Initially forward 30.0 0.292 1.30 (0.46–3.65) 0.71 (0.28–1.77) 0.41 (0.18–0.94)  Initially backward 20.2 0.028 0.52 (0.09–2.90) 1.82 (0.49–6.73) 3.19 (1.09–9.35)*  Initially sideward 43.0 0.124 1.33 (0.48–3.69) 1.42 (0.61–3.35) 1.20 (0.55–2.62)  Initially straight down 6.8 0.208 0.22 (0.03–2.05) 0.23 (0.04–1.18) 0.73 (0.21–2.55) Descent phase  Stepping 85.0 0.302 0.93 (0.17–5.07) 0.65 (0.16–2.73) 0.38 (0.10–1.44)  Grasping 34.5 0.178 0.71 (0.28–1.80) 0.49 (0.24–1.02) 0.92 (0.45–1.87)  Turning during descent 45.9 0.683 0.90 (0.32–2.52) 0.82 (0.33–2.02) 0.68 (0.31–1.51)  Attempted upper limb arrest 91.8 0.129 2.97 (0.33–26.67) 6.41 (0.72–56.72) 0.80 (0.21–2.99) Impact phase  Head impact 45.1 0.567 1.29 (0.52–3.23) 1.12 (0.46–2.75) 0.96 (0.41–2.28)  Hip impact 52.7 0.064 1.67 (0.62–4.52) 2.03 (0.94–4.38) 0.89 (0.46–1.74)  Sustained injury 32.8 0.020 0.79 (0.28–2.19) 0.61 (0.25–1.50) 0.34 (0.15–0.74)* Characteristics of Falls During Walking Prevalence, % Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 55.0 0.352 0.33 (0.12–0.93) 0.64 (0.27–1.52) 1.09 (0.50–2.39)  Initially forward 30.0 0.292 1.30 (0.46–3.65) 0.71 (0.28–1.77) 0.41 (0.18–0.94)  Initially backward 20.2 0.028 0.52 (0.09–2.90) 1.82 (0.49–6.73) 3.19 (1.09–9.35)*  Initially sideward 43.0 0.124 1.33 (0.48–3.69) 1.42 (0.61–3.35) 1.20 (0.55–2.62)  Initially straight down 6.8 0.208 0.22 (0.03–2.05) 0.23 (0.04–1.18) 0.73 (0.21–2.55) Descent phase  Stepping 85.0 0.302 0.93 (0.17–5.07) 0.65 (0.16–2.73) 0.38 (0.10–1.44)  Grasping 34.5 0.178 0.71 (0.28–1.80) 0.49 (0.24–1.02) 0.92 (0.45–1.87)  Turning during descent 45.9 0.683 0.90 (0.32–2.52) 0.82 (0.33–2.02) 0.68 (0.31–1.51)  Attempted upper limb arrest 91.8 0.129 2.97 (0.33–26.67) 6.41 (0.72–56.72) 0.80 (0.21–2.99) Impact phase  Head impact 45.1 0.567 1.29 (0.52–3.23) 1.12 (0.46–2.75) 0.96 (0.41–2.28)  Hip impact 52.7 0.064 1.67 (0.62–4.52) 2.03 (0.94–4.38) 0.89 (0.46–1.74)  Sustained injury 32.8 0.020 0.79 (0.28–2.19) 0.61 (0.25–1.50) 0.34 (0.15–0.74)* Note: Bolded p-values indicate a p < .05, underlined p-values highlight differences from Table 2 containing the analysis of all falls. *OR significantly different from 1 at p < .05. View Large The most common activity at time of fall was walking (33.6%) or standing (24.0%). Less often falls occurred during stand-to-sit transfers (17.5%), sitting (14.2%), and sit-to-stand transfers (10.7%). Highest frequency fallers were more likely to fall during a stand-to-sit transition compared to lowest frequency fallers (OR = 2.76, 95% CI = 1.29–5.88; Table 2). Moreover, highest frequency fallers were less likely to fall during walking compared to lowest frequency fallers (OR = 0.35, 95% CI = 0.19–0.63). The association between injury risk and fall frequency group remained significant after correcting for the height of the fall (above, at, or lower than standing height; p < .027). Falls during walking had a higher risk for injury (OR = 2.2, 95% CI = 1.6–2.6 vs not walking). Nevertheless, the association between injury risk and fall frequency group remained significant after stratifying for activity at the time of falling (walking or transferring; p < .044). For falls during walking (N = 265), highest frequency fallers remained less likely to injure themselves than lower frequency fallers (OR = 0.34, 95% CI = 0.15–0.74 for Q4 vs Q1 and OR = 0.43, 95% CI = 0.18–1.00 for Q4 vs Q2; Table 3). Attempts to recover balance by grasping or body rotation during descent, and attempts to arrest the fall with the upper limbs occurred frequently; their prevalence was not associated with fall frequency (Table 1, all p > .265). Recovery steps were observed in 48.9% of falls and associated significantly with fall frequency (p = .005). Highest frequency fallers had lower odds of executing recovery steps during a fall compared to lowest frequency fallers (OR = 0.33, 95% CI= 0.18–0.61). Similar trends were observed for low–medium and medium–high frequency fallers compared to lowest frequency fallers. Recovery steps were less often performed during stand-to-sit falls (OR = 0.14, 95% CI = 0.08–0.26 vs not stand-to-sit fall) and more often during falls during walking (OR = 12.24, 95% CI = 6.13–24.45 vs not walking). Analysis of only falls during walking revealed no significant differences in the odds of executing recovery steps between fall frequency groups (Table 3). Discussion We extend the current literature on the relationships between fall frequency, fall characteristics, and risk for injury, by linking injury data to evidence of the circumstances of falls from unique video footage of real-life falls experienced by older adults in LTC. Our results support previous findings (10–13) that more frequent fallers are less likely than less frequent fallers to injure themselves in the event of a fall. Our findings do not support the notion that recurrent fallers more often exhibit protective responses during falls such as upper limb fall arrest, or rotation during descent. Instead, the difference in injury risk appears to be due in part to differences in the activities leading to falls. Falls during walking created higher risk for injury, and more frequent fallers were less likely to fall during walking. Recurrent fallers were instead more likely to fall while transferring, which tends to involve a lower change in potential energy. While fall frequency did not associate with odds of impact to the head (or to the hip), the lower energy content may have caused the severity of head impacts to be lower in higher frequency fallers. However, activity at the time of falling could not explain all of the observed trends. More frequent fallers had lower risk for injury per fall from falls during walking, perhaps due to their increased tendency to fall backward, which creates lower risk for head impact than falling forward (18). Moreover, the relation between injury risk and fall frequency persisted after correcting for fall height. This suggests that injury risk may be dominated by tissue tolerance to injury (as opposed to applied force), or depend on fall characteristics that were not examined in this study, and perhaps cannot be easily discerned from video. These may include differences in impact velocity, impact location, or sharing of impact energy between multiple body parts. Compared to lower frequency fallers, higher frequency fallers fell more often during transfers and less often during walking. An explanation for differences in activity at time of the fall could be shared group demographics that increase activity-specific fall risk, or decreased exposure to specific activities. In contrast to previous studies that reported that recurrent fallers are more often female, have slower reaction time, impaired cognition and lower muscle strength (7,11), we did not observe differences between fall frequency groups in gender, age or observational measures of physical and cognitive status. We did note a higher prevalence of antipsychotic and antianxiety medication in the highest frequency fallers. These observations may call for more sensitive tools of physical and cognitive capacity in LTC, and investigation of the effect of medication on fall rate. Altered habitual activity, such as less walking, could have decreased exposure and thereby reduced the number of falls during walking. Previous studies using self-reports indeed suggest that recurrent fallers are less physically active (26). Greater understanding is required of the quantity of daily activities in frequent fallers and of how different activity types associate with fall risk. Our study has important limitations. First, we video-captured only falls in communal areas of two LTC facilities (about 40% of falls (21)). We cannot assume that our results are applicable for falls in private areas (eg, bedrooms and bathrooms), or for falls in other populations (eg, community-dwelling) older adults. Second, we established our groups as quartiles of fall frequency as there are no established cut-offs for fall frequency among LTC residents. Our results remained consistent when we used tertiles and quintiles for analyses. Moreover, the median fall frequency of 2.5 falls per year was comparable to those previously reported for the LTC population (1,2,27). Third, we used the LTC facilities’ fall incident reports to identify injuries, and we did not differentiate severity of injury. The 23.2% injury rate we report is similar to values reported in other studies of falls in the LTC population (6–9). Finally, while we analyzed videos to determine the occurrence of impact to specific body parts, and again we did not estimate the severity of a given impact, nor did we consider the sequence of impacts to the various body parts, which are important outcomes for future research. In summary, we used footage of real-life falls to examine the association between fall frequency, injury risk, and fall circumstances in LTC. We found that, when compared to lower frequency fallers, falls in high frequency fallers resulted in a lower injury risk per fall, and occurred more often while transferring and less often while walking. Our results support the implementation of strategies to prevent falls and fall-related injuries in LTC (eg, exercise, medication review, hip protectors) for both frequent and infrequent fallers, as the latter are more apt to sustain injury during a fall. While only 5% of older people reside in LTC, falls in this population account for 20% of all fall-related injury deaths among older people (2,28). Thus, future studies should examine the effectiveness of targeted interventions in this most frail population. Funding This work was supported by grants from the Canadian Institute for Health Research (AMG-100487, TIR-103945, and TEI-138295; CIHR New Investigator Award to J.S.-G.), the AGE-WELL National Centres of Excellence (AW CRP 2015-WP5.2) and the Michael Smith Foundation for Health Research (Postdoctoral Fellowship Awards #16606 to K.S.S. and #16556 to Y.Y., and Research Scholar Award to J.S.-G.). References 1. Rapp K , Becker C , Cameron ID , König H-H , Büchele G . Epidemiology of falls in residential aged care: analysis of more than 70,000 falls from residents of bavarian nursing homes . J Am Med Dir Assoc . 2012 ; 13 : 187 .e1–6. doi: 10.1016/j.jamda.2011.06.011 Google Scholar CrossRef Search ADS PubMed 2. Rubenstein LZ , Josephson KR , Robbins AS . Falls in the nursing home . Ann Intern Med . 1994 ; 121 : 442 – 451 . doi:10.7326/0003- 4819-121-6-199409150-00009 Google Scholar CrossRef Search ADS PubMed 3. Rubenstein LZ . Falls in older people: epidemiology, risk factors and strategies for prevention . Age Ageing . 2006 ; 35 ( suppl 2 ): ii37 – ii41 . doi: 10.1093/ageing/afl084 Google Scholar CrossRef Search ADS PubMed 4. Grisso JA , Kelsey JL , Strom BL et al. Risk factors for falls as a cause of hip fracture in women. The Northeast Hip Fracture Study Group . N Engl J Med . 1991 ; 324 : 1326 – 1331 . doi: 10.1056/NEJM199105093241905 Google Scholar CrossRef Search ADS PubMed 5. Head Injuries in Canada: A Decade of Change (1994-1995 to 2003-2004). Canadian Institute for Health Information, 2006. https://secure.cihi.ca/free_products/ntr_head_injuries_2006_e.pdf. 6. Lloyd BD , Williamson DA , Singh NA et al. Recurrent and injurious falls in the year following hip fracture: a prospective study of incidence and risk factors from the Sarcopenia and Hip Fracture study . J Gerontol A Biol Sci Med Sci . 2009 ; 64 : 599 – 609 . doi: 10.1093/gerona/glp003 Google Scholar CrossRef Search ADS PubMed 7. Towne SD , Cho J , Smith ML , Ory MG . Factors associated with injurious falls in residential care facilities . J Aging Health . 2016 : 29 : 669 – 687 . doi: 10.1177/0898264316641083 Google Scholar CrossRef Search ADS PubMed 8. O’Loughlin JL , Robitaille Y , Boivin JF , Suissa S . Incidence of and risk factors for falls and injurious falls among the community-dwelling elderly . Am J Epidemiol . 1993 ; 137 : 342 – 354 . doi:10.1093/oxfordjournals.aje.a116681 Google Scholar CrossRef Search ADS PubMed 9. Koski K . Risk factors for major injurious falls among the home-dwelling elderly by functional abilities . Gerontology . 1998 ; 44 : 232 – 238 . doi:10.1159/ 000022017 Google Scholar CrossRef Search ADS PubMed 10. Poss JW , Hirdes JP . Very frequent fallers and future fall injury: continuous risk among community-dwelling home care recipients . J Aging Health . 2016 ; 28 : 587 – 599 . doi: 10.1177/0898264315599941 Google Scholar CrossRef Search ADS PubMed 11. Nevitt MC , Cummings SR , Hudes ES . Risk factors for injurious falls: a prospective study . J Gerontol . 1991 ; 46 : M164 – M170 . doi:10.1093/geronj/ 46.5.M164 Google Scholar CrossRef Search ADS PubMed 12. Tinetti ME . Factors associated with serious injury during falls by ambulatory nursing home residents . J Am Geriatr Soc . 1987 ; 35 : 644 – 648 . doi:10.1111/ j.1532-5415.1987.tb04341.x Google Scholar CrossRef Search ADS PubMed 13. Sambrook PN , Cameron ID , Chen JS et al. Influence of fall related factors and bone strength on fracture risk in the frail elderly . Osteoporos Int . 2007 ; 18 : 603 – 610 . doi: 10.1007/s00198-006-0290-z Google Scholar CrossRef Search ADS PubMed 14. Deandrea S , Bravi F , Turati F , Lucenteforte E , La Vecchia C , Negri E . Risk factors for falls in older people in nursing homes and hospitals. A systematic review and meta-analysis . Arch Gerontol Geriatr . 2013 ; 56 : 407 – 415 . doi: 10.1016/j.archger.2012.12.006 Google Scholar CrossRef Search ADS PubMed 15. Baranzini F , Diurni M , Ceccon F et al. Fall-related injuries in a nursing home setting: is polypharmacy a risk factor ? BMC Health Serv Res . 2009 ; 9 : 228 . doi: 10.1186/1472-6963-9-228 Google Scholar CrossRef Search ADS PubMed 16. Nevitt MC , Cummings SR , The Study of Osteoporotic Fractures Research Group . Type of fall and risk of hip and wrist fractures: the study of osteoporotic fractures . J Am Geriatr Soc . 1993 ; 41 : 1226 – 1234 . doi:10.1111/ j.1532-5415.1993.tb07307.x Google Scholar CrossRef Search ADS PubMed 17. Schwartz AV , Kelsey JL , Sidney S , Grisso JA . Characteristics of falls and risk of hip fracture in elderly men . Osteoporos Int . 1998 ; 8 : 240 – 246 . doi: 10.1007/s001980050060 Google Scholar CrossRef Search ADS PubMed 18. Schonnop R , Yang Y , Feldman F , Robinson E , Loughin M , Robinovitch SN . Prevalence of and factors associated with head impact during falls in older adults in long-term care . CMAJ . 2013 ; 185 : E803 – E810 . doi: 10.1503/cmaj.130498 Google Scholar CrossRef Search ADS PubMed 19. Yang Y , Mackey DC , Liu-Ambrose T , Feldman F , Robinovitch SN . Risk factors for hip impact during real-life falls captured on video in long-term care . Osteoporos Int . 2016 ; 27 : 537 – 547 . doi: 10.1007/s00198-015-3268-x Google Scholar CrossRef Search ADS PubMed 20. Yardley L , Beyer N , Hauer K , Kempen G , Piot-Ziegler C , Todd C . Development and initial validation of the Falls Efficacy Scale-International (FES-I) . Age Ageing . 2005 ; 34 : 614 – 619 . doi: 10.1093/ageing/afi196 Google Scholar CrossRef Search ADS PubMed 21. Robinovitch SN , Feldman F , Yang Y et al. Video capture of the circumstances of falls in elderly people residing in long-term care: an observational study . Lancet . 2013 ; 381 : 47 – 54 . doi: 10.1016/S0140-6736(12)61263-X Google Scholar CrossRef Search ADS PubMed 22. Yang Y , Schonnop R , Feldman F , Robinovitch SN . Development and validation of a questionnaire for analyzing real-life falls in long-term care captured on video . BMC Geriatr . 2013 ; 13 : 40 . doi: 10.1186/1471-2318-13-40 Google Scholar CrossRef Search ADS PubMed 23. Hayes WC , Myers ER , Morris JN , Gerhart TN , Yett HS , Lipsitz LA . Impact near the hip dominates fracture risk in elderly nursing home residents who fall . Calcif Tissue Int . 1993 ; 52 : 192 – 198 . doi:10.1007/BF00298717 Google Scholar CrossRef Search ADS PubMed 24. Morris JN , Fries BE , Mehr DR et al. MDS Cognitive Performance Scale . J Gerontol . 1994 ; 49 : M174 – M182 . Google Scholar CrossRef Search ADS PubMed 25. Morris JN , Fries BE , Morris SA . Scaling ADLs within the MDS . J Gerontol A Biol Sci Med Sci . 1999 ; 54 : M546 – M553 . doi:10.1093/geronj/49.4.M174 Google Scholar CrossRef Search ADS PubMed 26. Peeters GM , van Schoor NM , Pluijm SM , Deeg DJ , Lips P . Is there a U-shaped association between physical activity and falling in older persons ? Osteoporos Int . 2010 ; 21 : 1189 – 1195 . doi: 10.1007/s00198-009-1053-4 Google Scholar CrossRef Search ADS PubMed 27. Lord SR , March LM , Cameron ID et al. Differing risk factors for falls in nursing home and intermediate-care residents who can and cannot stand unaided . J Am Geriatr Soc . 2003 ; 51 : 1645 – 1650 . doi: 10.1046/ j.1532-5415.2003.51518.x Google Scholar CrossRef Search ADS PubMed 28. Health, United States, 2005: With Chartbook on Trends in the Health of Americans . Hyattsville, MD: National Center for Health Statistics ; 2005 . © The Author(s) 2017. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences Oxford University Press

The Association Between Fall Frequency, Injury Risk, and Characteristics of Falls in Older Residents of Long-Term Care: Do Recurrent Fallers Fall More Safely?

Loading next page...
 
/lp/ou_press/the-association-between-fall-frequency-injury-risk-and-characteristics-ktfP4hvvfL
Publisher
Oxford University Press
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
1079-5006
eISSN
1758-535X
D.O.I.
10.1093/gerona/glx196
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Although a fall is a necessary prerequisite to a fall-related injury, previous studies suggest that frequent fallers are at lower injury risk for a given fall. We tested the hypotheses that differences in protective responses or the circumstances of falls underlie differences in injury risk with fall frequency. Methods We analyzed video footage of 897 falls experienced by 220 long-term care residents (mean age 82 ± 9 years) to identify the cause of imbalance, activity leading to falling, direction of fall initiation, balance recovery and fall protective responses, and occurrence of impact to the head or hip. We further obtained injury information from the facilities’ fall registration. We used generalized estimating equation models to examine the association between quartiles of fall frequency, injury risk, and fall characteristics. Results Residents with the highest fall frequency group (Q4; ≥5.6 falls/year) were less likely to sustain an injury per fall. They were less likely to fall during walking and more likely to fall during stand-to-sit transfers. Residents in the lowest fall frequency group (Q1; <1.15 falls/year) were more likely to fall during walking, and walking was associated with an increased risk for injury. Conclusion When compared to less frequent fallers, more frequent fallers had a lower risk for injury per fall. This appeared to be explained by differences in the circumstances of falls, and not by protective responses. Injury prevention strategies in long-term care should target both frequent and infrequent fallers, as the latter are more mobile and apt to sustain injury. Accidental falls, Injury, Aged 65 and over, Long-term care, Frequent fallers About 60% of long-term care (LTC) residents falls at least once annually, and with an average frequency of 1.7 falls per year, most residents will fall recurrently (1–3). Falls are the number-one cause of injuries in adults over the age of 65, including 90% of hip fractures (4) and 70% of traumatic brain injuries (5). Nevertheless, only about 20%–30% of falls in older adults actually results in injury (6–9). An improved understanding of factors that separate injurious and noninjurious falls would serve to guide prevention efforts. A fall is clearly a prerequisite for experiencing a fall-related injury. Several studies have shown that the cumulative risk for injury is higher among older adults who fall recurrently compared to incidental fallers (10–13). Furthermore, falls and fall-related injuries have shared risk factors, suggesting that frequent fallers are more likely to experience an injury in the event of a fall. For example, female gender, slowed reaction time, impaired cognition, reduced muscle strength and the use of psychoactive medications are associated with both increased risk for falling and injury (7,11,14). Presumably, this is due to the fact that these factors also influence the characteristics (and severity) of falls. However, although counter intuitive, evidence suggests that a given fall is less likely to cause injury in a frequent faller when compared to a less frequent faller. For example, Baranzini and colleagues (15) reported that frequent fallers in LTC (≥4 falls/year) were less likely to sustain an injury in falling than less frequent fallers (odds ratio [OR] = 0.42, 95% confidence interval = 0.20–0.89). Furthermore, Tinetti (12) found that frequent fallers in LTC (≥4 falls/year or fell every 2 months) had a relative risk of 0.6 for experiencing an injury compared to those who fell less often. Nevitt and colleagues (11) observed similar trends among community-dwelling fallers but the differences were not statistically significant. While the reasons for these trends are unclear, several explanations have been offered. First, frequent fallers may use more effective protective responses to avoid injury during a fall (11). For example, falling on an outstretched arm decreases the risk for hip fractures (16,17) and body rotation during descent decreases the risk for head and hip impact (18,19). Second, the circumstances and dynamics of falls might be different for frequent fallers. For example, falls in frequent fallers may be from a lower height (eg, sitting vs standing) and involve a smaller change in potential energy and lower risk for injury. Third, individuals at greatest risk for injury (eg, due to osteoporosis or impaired protective responses) may demonstrate greater fear of falling, which often leads to activity restriction (eg, (20)). The activity restriction may lead to a reduction in falls, but the low tissue tolerance or impaired protective responses lead to an increased risk for injury in the event of a fall. The aim of the current study was to test the hypotheses that the frequency of falls in LTC residents associates with injury risk during falling, and with the characteristics of falls. To address this aim, we analyzed a unique video dataset of real-life falls sustained by older adults who lived in two LTC facilities. Methods Falls were recorded on video in common areas of two LTC facilities in Greater Vancouver, as described in Baranzini and colleagues (21). The research ethics boards at Simon Fraser University and Fraser Health Authority approved the study. At the time of admission to LTC, each resident or proxy provided written informed consent for video recording. For the current study we analyzed falls experience between April 20, 2007 and September 21, 2015 by residents from whom we had consent to access their medical records. Fall Frequency, Group Demographics and Consequences of Falls Each facility used a standardized fall incidence report from which we accessed the total number of falls experienced by each study participant and data on injuries due to these falls. The report documents the time and location of the fall, a narrative description of the fall and information on injuries associated with the fall. The admission date (and where relevant discharge date), as well as observational tests of physical and cognitive status, disease diagnoses and medication use, were obtained from the facility’s minimum data set (MDS) health records. Fall frequency was calculated as total number of falls divided by the length of stay. Characteristics of Falls and Protective Responses Each fall video was analyzed by a three-member research team who sought consensus on the characteristics of the fall using a validated questionnaire (22,23). In the current analysis, we focused on characteristics of fall initiation (nature of imbalance, activity at the time of the fall, initial fall direction), fall descent (presence of grasping or stepping responses for balance recovery, axial rotation during descent, upper limb fall arrest) and fall impact (sites of impact on the body). Statistics We first determined whether fall frequency was associated with fall-related injury rate using Spearman correlation. Second, we determined the association between fall frequency and characteristics of falls. Visualization of the data suggested a nonlinear association; hence, residents were categorized into four quartiles based on fall frequency. We performed chi-square and ANOVA tests to identify differences in group demographics. We subsequently used generalized estimating equation models to estimate OR. To reduce the possibility that residents who fell very frequently would unduly influence our findings, we employed a bootstrapping approach where we randomly sampled a maximum of 20 falls per resident to be included and averaged results over 1,000 repetitions. All statistics were performed in SPSS (IBM, version 23) and a p-value of .05 was considered statistically significant. Results Group Demographics We analyzed falls from 220 residents with a mean age of 82.1 (SD = 8.8) years. Descriptive characteristics are provided in Table 1. For 69% of participants (n = 152), we had MDS data within 6 months of the fall date. Participants scored an average of 2.9 (SD 1.7) out of 6 on the activities of daily living (ADL) self-performance hierarchy scale and 3.4 (SD 1.6) out of 6 on the cognitive performance scale, indicating the average resident required assistance in performing ADLs and had a moderate cognitive impairment (24,25). The 220 residents experienced a total of 3,573 falls; 897 were captured on video. The observation period ranged from 94 to 2,751 days, with a mean of 1,380 (SD = 704) days. Fall frequencies ranged from 0.15 to 39.2 falls per year (median 2.5, IQR 4.4; Figure 1). Fall frequencies were categorized into four equally sized quartiles with 1.15, 2.5, and 5.6 falls per year as cut points. We will refer to these groups as lowest (Q1), low–medium (Q2), medium–high (Q3), and highest (Q4) fall frequency groups. Table 1. Descriptive Characteristics of the Fall Frequency Groups Total Fall Frequency Group p-Value Lowest (Q1) Low–Medium Medium–High Highest (Q4) Number of residents 220 55 55 55 55 N/A Number of video-captured falls 897 76 129 193 499 N/A Number of video-captured falls with injury data# 762 70 115 144 433 N/A Number of sustained injuries 177 24 38 37 78 N/A Demographics and general health  Mean fall rate (number/year) 5.40 (6.79) 0.62 (0.27)± 1.74 (0.40)± 3.88 (0.96)± 13.16 (8.44)± <0.0001  Mean injurious fall rate (number/year) 0.28 (0.55) 0.15 (0.21)± 0.22 (0.31)§ 0.28 (0.39)& 0.54 (0.92)±,§,& <0.0001  Mean length of stay (years) 3.76 (1.94) 3.59 (1.93) 4.27 (1.70) 3.74 (2.02) 3.51 (2.01) 0.165  Age at time of fall (years) 82.1 (8.8) 83.3 (7.9) 79.9 (9.3) 80.6 (8.5) 82.6 (9.1) 0.780  Female gender (%) 62.5 78.6 63.4 56.1 57.1 0.224  BMI (kg/m2) 24.7 (4.9) 24.7 (4.7) 25.6 (5.0) 24.6 (4.6) 23.7 (5.4) 0.492  ADL at time of fall (score) 2.9 (1.7) 2.7 (2.0) 3.6 (1.7) 3.0 (1.5) 3.3 (1.4) 0.443  CPS at time of fall (score) 3.4 (1.6) 3.0 (1.3) 3.8 (1.8) 3.5 (1.4) 4.0 (1.4) 0.443 Disease diagnosis  Alzheimer’s disease (%) 29.6 25.0 24.4 24.4 42.9 0.206  Diabetes (%) 21.7 21.4 17.1 19.5 28.6 0.632  Cardiac arrhythmia (%) 4.6 3.6 2.4 4.9 7.1 0.882  Hypertension (%) 40.8 53.6 34.1 48.8 31.0 0.151  Stroke (%) 11.8 3.6 7.3 24.4 9.5 0.044  Parkinson’s disease (%) 2.6 3.6 0.0 0.0 7.1 0.107  COPD (%) 11.8 7.1 9.8 12.2 16.7 0.670 Use of medication  Number of medications 8.76 (3.99) 7.96 (4.25) 8.15 (3.46) 9.71 (3.30) 8.98 (4.77) 0.212  Antipsychotic (%) 41.4 35.7 24.4± 48.8 54.8± 0.025  Antianxiety (%) 21.1 3.6±§ 14.6 29.3± 31.0§ 0.011  Antidepressant (%) 48.0 32.1 48.8 53.7 52.4 0.303  Hypnotic (%) 15.8 17.9 12.2 17.1 16.7 0.915  Diuretic (%) 25.0 17.9 19.5 36.6 23.8 0.254  Analgesics (%) 52.6 46.4 46.3 63.4 52.4 0.402 Total Fall Frequency Group p-Value Lowest (Q1) Low–Medium Medium–High Highest (Q4) Number of residents 220 55 55 55 55 N/A Number of video-captured falls 897 76 129 193 499 N/A Number of video-captured falls with injury data# 762 70 115 144 433 N/A Number of sustained injuries 177 24 38 37 78 N/A Demographics and general health  Mean fall rate (number/year) 5.40 (6.79) 0.62 (0.27)± 1.74 (0.40)± 3.88 (0.96)± 13.16 (8.44)± <0.0001  Mean injurious fall rate (number/year) 0.28 (0.55) 0.15 (0.21)± 0.22 (0.31)§ 0.28 (0.39)& 0.54 (0.92)±,§,& <0.0001  Mean length of stay (years) 3.76 (1.94) 3.59 (1.93) 4.27 (1.70) 3.74 (2.02) 3.51 (2.01) 0.165  Age at time of fall (years) 82.1 (8.8) 83.3 (7.9) 79.9 (9.3) 80.6 (8.5) 82.6 (9.1) 0.780  Female gender (%) 62.5 78.6 63.4 56.1 57.1 0.224  BMI (kg/m2) 24.7 (4.9) 24.7 (4.7) 25.6 (5.0) 24.6 (4.6) 23.7 (5.4) 0.492  ADL at time of fall (score) 2.9 (1.7) 2.7 (2.0) 3.6 (1.7) 3.0 (1.5) 3.3 (1.4) 0.443  CPS at time of fall (score) 3.4 (1.6) 3.0 (1.3) 3.8 (1.8) 3.5 (1.4) 4.0 (1.4) 0.443 Disease diagnosis  Alzheimer’s disease (%) 29.6 25.0 24.4 24.4 42.9 0.206  Diabetes (%) 21.7 21.4 17.1 19.5 28.6 0.632  Cardiac arrhythmia (%) 4.6 3.6 2.4 4.9 7.1 0.882  Hypertension (%) 40.8 53.6 34.1 48.8 31.0 0.151  Stroke (%) 11.8 3.6 7.3 24.4 9.5 0.044  Parkinson’s disease (%) 2.6 3.6 0.0 0.0 7.1 0.107  COPD (%) 11.8 7.1 9.8 12.2 16.7 0.670 Use of medication  Number of medications 8.76 (3.99) 7.96 (4.25) 8.15 (3.46) 9.71 (3.30) 8.98 (4.77) 0.212  Antipsychotic (%) 41.4 35.7 24.4± 48.8 54.8± 0.025  Antianxiety (%) 21.1 3.6±§ 14.6 29.3± 31.0§ 0.011  Antidepressant (%) 48.0 32.1 48.8 53.7 52.4 0.303  Hypnotic (%) 15.8 17.9 12.2 17.1 16.7 0.915  Diuretic (%) 25.0 17.9 19.5 36.6 23.8 0.254  Analgesics (%) 52.6 46.4 46.3 63.4 52.4 0.402 Note: Values are based on the status at the time of the first fall and are reported as mean (standard deviation) or percentage. p-Values are based on Fisher’s exact test (categorical) or ANOVA (continuous). Bolded p-values indicate a p < .05. ±, § and & indicate Bonferroni-corrected significant differences in post hoc group-wise comparisons. #Falls where the incident report confirmed that either no injury occurred, or an injury occurred. View Large Table 1. Descriptive Characteristics of the Fall Frequency Groups Total Fall Frequency Group p-Value Lowest (Q1) Low–Medium Medium–High Highest (Q4) Number of residents 220 55 55 55 55 N/A Number of video-captured falls 897 76 129 193 499 N/A Number of video-captured falls with injury data# 762 70 115 144 433 N/A Number of sustained injuries 177 24 38 37 78 N/A Demographics and general health  Mean fall rate (number/year) 5.40 (6.79) 0.62 (0.27)± 1.74 (0.40)± 3.88 (0.96)± 13.16 (8.44)± <0.0001  Mean injurious fall rate (number/year) 0.28 (0.55) 0.15 (0.21)± 0.22 (0.31)§ 0.28 (0.39)& 0.54 (0.92)±,§,& <0.0001  Mean length of stay (years) 3.76 (1.94) 3.59 (1.93) 4.27 (1.70) 3.74 (2.02) 3.51 (2.01) 0.165  Age at time of fall (years) 82.1 (8.8) 83.3 (7.9) 79.9 (9.3) 80.6 (8.5) 82.6 (9.1) 0.780  Female gender (%) 62.5 78.6 63.4 56.1 57.1 0.224  BMI (kg/m2) 24.7 (4.9) 24.7 (4.7) 25.6 (5.0) 24.6 (4.6) 23.7 (5.4) 0.492  ADL at time of fall (score) 2.9 (1.7) 2.7 (2.0) 3.6 (1.7) 3.0 (1.5) 3.3 (1.4) 0.443  CPS at time of fall (score) 3.4 (1.6) 3.0 (1.3) 3.8 (1.8) 3.5 (1.4) 4.0 (1.4) 0.443 Disease diagnosis  Alzheimer’s disease (%) 29.6 25.0 24.4 24.4 42.9 0.206  Diabetes (%) 21.7 21.4 17.1 19.5 28.6 0.632  Cardiac arrhythmia (%) 4.6 3.6 2.4 4.9 7.1 0.882  Hypertension (%) 40.8 53.6 34.1 48.8 31.0 0.151  Stroke (%) 11.8 3.6 7.3 24.4 9.5 0.044  Parkinson’s disease (%) 2.6 3.6 0.0 0.0 7.1 0.107  COPD (%) 11.8 7.1 9.8 12.2 16.7 0.670 Use of medication  Number of medications 8.76 (3.99) 7.96 (4.25) 8.15 (3.46) 9.71 (3.30) 8.98 (4.77) 0.212  Antipsychotic (%) 41.4 35.7 24.4± 48.8 54.8± 0.025  Antianxiety (%) 21.1 3.6±§ 14.6 29.3± 31.0§ 0.011  Antidepressant (%) 48.0 32.1 48.8 53.7 52.4 0.303  Hypnotic (%) 15.8 17.9 12.2 17.1 16.7 0.915  Diuretic (%) 25.0 17.9 19.5 36.6 23.8 0.254  Analgesics (%) 52.6 46.4 46.3 63.4 52.4 0.402 Total Fall Frequency Group p-Value Lowest (Q1) Low–Medium Medium–High Highest (Q4) Number of residents 220 55 55 55 55 N/A Number of video-captured falls 897 76 129 193 499 N/A Number of video-captured falls with injury data# 762 70 115 144 433 N/A Number of sustained injuries 177 24 38 37 78 N/A Demographics and general health  Mean fall rate (number/year) 5.40 (6.79) 0.62 (0.27)± 1.74 (0.40)± 3.88 (0.96)± 13.16 (8.44)± <0.0001  Mean injurious fall rate (number/year) 0.28 (0.55) 0.15 (0.21)± 0.22 (0.31)§ 0.28 (0.39)& 0.54 (0.92)±,§,& <0.0001  Mean length of stay (years) 3.76 (1.94) 3.59 (1.93) 4.27 (1.70) 3.74 (2.02) 3.51 (2.01) 0.165  Age at time of fall (years) 82.1 (8.8) 83.3 (7.9) 79.9 (9.3) 80.6 (8.5) 82.6 (9.1) 0.780  Female gender (%) 62.5 78.6 63.4 56.1 57.1 0.224  BMI (kg/m2) 24.7 (4.9) 24.7 (4.7) 25.6 (5.0) 24.6 (4.6) 23.7 (5.4) 0.492  ADL at time of fall (score) 2.9 (1.7) 2.7 (2.0) 3.6 (1.7) 3.0 (1.5) 3.3 (1.4) 0.443  CPS at time of fall (score) 3.4 (1.6) 3.0 (1.3) 3.8 (1.8) 3.5 (1.4) 4.0 (1.4) 0.443 Disease diagnosis  Alzheimer’s disease (%) 29.6 25.0 24.4 24.4 42.9 0.206  Diabetes (%) 21.7 21.4 17.1 19.5 28.6 0.632  Cardiac arrhythmia (%) 4.6 3.6 2.4 4.9 7.1 0.882  Hypertension (%) 40.8 53.6 34.1 48.8 31.0 0.151  Stroke (%) 11.8 3.6 7.3 24.4 9.5 0.044  Parkinson’s disease (%) 2.6 3.6 0.0 0.0 7.1 0.107  COPD (%) 11.8 7.1 9.8 12.2 16.7 0.670 Use of medication  Number of medications 8.76 (3.99) 7.96 (4.25) 8.15 (3.46) 9.71 (3.30) 8.98 (4.77) 0.212  Antipsychotic (%) 41.4 35.7 24.4± 48.8 54.8± 0.025  Antianxiety (%) 21.1 3.6±§ 14.6 29.3± 31.0§ 0.011  Antidepressant (%) 48.0 32.1 48.8 53.7 52.4 0.303  Hypnotic (%) 15.8 17.9 12.2 17.1 16.7 0.915  Diuretic (%) 25.0 17.9 19.5 36.6 23.8 0.254  Analgesics (%) 52.6 46.4 46.3 63.4 52.4 0.402 Note: Values are based on the status at the time of the first fall and are reported as mean (standard deviation) or percentage. p-Values are based on Fisher’s exact test (categorical) or ANOVA (continuous). Bolded p-values indicate a p < .05. ±, § and & indicate Bonferroni-corrected significant differences in post hoc group-wise comparisons. #Falls where the incident report confirmed that either no injury occurred, or an injury occurred. View Large Figure 1. View largeDownload slide Fall frequency in quartiles. The black line indicates the fall frequency of each participant and gray bars indicate the quartiles; the numbers above the quartile bars provide the mean fall frequency among participants in that fall frequency group. Figure 1. View largeDownload slide Fall frequency in quartiles. The black line indicates the fall frequency of each participant and gray bars indicate the quartiles; the numbers above the quartile bars provide the mean fall frequency among participants in that fall frequency group. The four fall frequency groups did not differ significantly in age, gender, length of stay, BMI, ADL self-performance hierarchy, and cognitive performance scale scores (all p > .224; Table 1), but differed significantly in rates of falls and fall-related injuries, the prevalence of stroke, and the use of antipsychotic and antianxiety medications (all p < .044). Residents in the higher frequency groups fell more often and experienced cumulatively more injuries. Despite a significant main effect of group on the prevalence of stroke, this effect did not persist in post hoc comparisons. The use of antipsychotics was higher in the highest fall frequency group compared to the low-medium fall frequency group (54.8% vs 24.6%). The medium-high and highest fall frequency groups used more antianxiety medications compared to the lowest fall frequency group (respectively, 29.3% and 31.0% vs 3.6%). These differences between groups did not lead to confounding in subsequent analyses (tested as a <10% change in regression coefficients) hence uncorrected associations are reported. Consequences of Falls We were unable to verify whether or not an injury occurred in 140 of the 897 falls captured on video (15.5%), as no response was provided to the question “did an injury occur (YES/ NO)?” in the corresponding fall incident reports. We excluded these cases from our analysis of the consequences of falls. For the remaining 762 falls, injuries were reported in 23.2% of cases (n = 177). Injuries in order of most to least common were lacerations or abrasions (50.7% of cases), swelling or hematomas (29.4%), localized pain without superficial signs (14.0%), bone fracture (2.9%), and joint sprain or dislocation (2.9%). Anatomical sites of injury were the head (50.4%), elbow (13.6%), knee/shin (8.8%), hand/wrists (8.0%), shoulder/arm (7.2%), spine/back (5.6%), hip/thigh (4.8%), and pelvis/tailbone (0.8%). Fall frequency associated moderately with injury rate (r = .48, p < .001). Highest frequency fallers were less likely than lowest (OR = 0.44, 95% CI = 0.24–0.81) and low–medium frequency fallers (OR = 0.46, 95% CI = 0.28–0.78) to have an injury reported for any given fall (Table 2). Of the seven falls that resulted in hip fractures and five falls that resulted in wrist sprains, none occurred among the highest fall frequency group. Table 2. Odds Ratios for Fall Characteristics as a Function of Fall Incidence, Based on Comparison of the Low Fall Frequency Group (Q1) to All Other Groups (Q2–Q4) Characteristics of Falls Prevalence, % of All Falls Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 56.8 0.129 0.83 (0.45–1.55) 0.83 (0.45–1.53) 1.26 (0.71–2.23)  Stand–sit transfer 17.5 0.018 0.95 (0.38–2.41) 0.97 (0.38–2.49) 2.76 (1.29–5.88)*  Sit–stand transfer 10.7 0.714 1.37 (0.42–4.50) 1.99 (0.60–6.58) 1.56 (0.51–4.79)  Sitting 14.2 0.197 2.85 (0.96–8.41) 1.89 (0.66–5.42) 2.65 (0.97–7.29)  Standing 24.0 0.766 1.36 (0.70–2.65) 1.03 (0.53–2.04) 1.01 (0.55–1.85)  Walking 33.6 0.005 0.44 (0.23–0.84)* 0.59 (0.31–1.11) 0.35 (0.19–0.63)*  Initially forward 19.3 0.055 1.05 (0.51–2.17) 0.60 (0.29–1.23) 0.47 (0.23–0.96)  Initially backward 32.2 0.094 0.88 (0.40–1.94) 1.54 (0.72–3.27) 1.71 (0.87–3.36)  Initially sideward 29.1 0.727 1.11 (0.57–2.17) 1.12 (0.60–2.11) 0.84 (0.46–1.53)  Initially straight down 19.3 0.461 0.93 (0.42–2.06) 0.84 (0.39–1.79) 1.35 (0.68–2.67) Descent phase  Stepping 49.8 0.005 0.56 (0.29–1.08) 0.57 (0.30–1.07) 0.33 (0.18–0.61)*  Grasping 24.2 0.265 0.59 (0.31–1.12) 0.55 (0.29–1.02) 0.66 (0.37–1.17)  Turning during descent 36.7 0.500 0.91 (0.46–1.78) 0.77 (0.41–1.46) 0.68 (0.38–1.21)  Attempted upper limb arrest 88.0 0.416 0.54 (0.18–1.62) 0.95 (0.32–2.81) 0.61 (0.22–1.70) Impact phase  Head impact 33.6 0.399 0.75 (0.37–1.54) 0.93 (0.45–1.93) 0.61 (0.30–1.22)  Hip impact 40.8 0.052 1.55 (0.81–2.97) 1.35 (0.71–2.57) 0.93 (0.53–1.62)  Sustained injury 23.2 0.001 0.94 (0.48–1.85) 0.66 (0.34–1.32) 0.44 (0.24–0.81)* Characteristics of Falls Prevalence, % of All Falls Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 56.8 0.129 0.83 (0.45–1.55) 0.83 (0.45–1.53) 1.26 (0.71–2.23)  Stand–sit transfer 17.5 0.018 0.95 (0.38–2.41) 0.97 (0.38–2.49) 2.76 (1.29–5.88)*  Sit–stand transfer 10.7 0.714 1.37 (0.42–4.50) 1.99 (0.60–6.58) 1.56 (0.51–4.79)  Sitting 14.2 0.197 2.85 (0.96–8.41) 1.89 (0.66–5.42) 2.65 (0.97–7.29)  Standing 24.0 0.766 1.36 (0.70–2.65) 1.03 (0.53–2.04) 1.01 (0.55–1.85)  Walking 33.6 0.005 0.44 (0.23–0.84)* 0.59 (0.31–1.11) 0.35 (0.19–0.63)*  Initially forward 19.3 0.055 1.05 (0.51–2.17) 0.60 (0.29–1.23) 0.47 (0.23–0.96)  Initially backward 32.2 0.094 0.88 (0.40–1.94) 1.54 (0.72–3.27) 1.71 (0.87–3.36)  Initially sideward 29.1 0.727 1.11 (0.57–2.17) 1.12 (0.60–2.11) 0.84 (0.46–1.53)  Initially straight down 19.3 0.461 0.93 (0.42–2.06) 0.84 (0.39–1.79) 1.35 (0.68–2.67) Descent phase  Stepping 49.8 0.005 0.56 (0.29–1.08) 0.57 (0.30–1.07) 0.33 (0.18–0.61)*  Grasping 24.2 0.265 0.59 (0.31–1.12) 0.55 (0.29–1.02) 0.66 (0.37–1.17)  Turning during descent 36.7 0.500 0.91 (0.46–1.78) 0.77 (0.41–1.46) 0.68 (0.38–1.21)  Attempted upper limb arrest 88.0 0.416 0.54 (0.18–1.62) 0.95 (0.32–2.81) 0.61 (0.22–1.70) Impact phase  Head impact 33.6 0.399 0.75 (0.37–1.54) 0.93 (0.45–1.93) 0.61 (0.30–1.22)  Hip impact 40.8 0.052 1.55 (0.81–2.97) 1.35 (0.71–2.57) 0.93 (0.53–1.62)  Sustained injury 23.2 0.001 0.94 (0.48–1.85) 0.66 (0.34–1.32) 0.44 (0.24–0.81)* Note: Bolded p-values indicate a p < .05. *OR significantly different from 1 at p < .05. View Large Table 2. Odds Ratios for Fall Characteristics as a Function of Fall Incidence, Based on Comparison of the Low Fall Frequency Group (Q1) to All Other Groups (Q2–Q4) Characteristics of Falls Prevalence, % of All Falls Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 56.8 0.129 0.83 (0.45–1.55) 0.83 (0.45–1.53) 1.26 (0.71–2.23)  Stand–sit transfer 17.5 0.018 0.95 (0.38–2.41) 0.97 (0.38–2.49) 2.76 (1.29–5.88)*  Sit–stand transfer 10.7 0.714 1.37 (0.42–4.50) 1.99 (0.60–6.58) 1.56 (0.51–4.79)  Sitting 14.2 0.197 2.85 (0.96–8.41) 1.89 (0.66–5.42) 2.65 (0.97–7.29)  Standing 24.0 0.766 1.36 (0.70–2.65) 1.03 (0.53–2.04) 1.01 (0.55–1.85)  Walking 33.6 0.005 0.44 (0.23–0.84)* 0.59 (0.31–1.11) 0.35 (0.19–0.63)*  Initially forward 19.3 0.055 1.05 (0.51–2.17) 0.60 (0.29–1.23) 0.47 (0.23–0.96)  Initially backward 32.2 0.094 0.88 (0.40–1.94) 1.54 (0.72–3.27) 1.71 (0.87–3.36)  Initially sideward 29.1 0.727 1.11 (0.57–2.17) 1.12 (0.60–2.11) 0.84 (0.46–1.53)  Initially straight down 19.3 0.461 0.93 (0.42–2.06) 0.84 (0.39–1.79) 1.35 (0.68–2.67) Descent phase  Stepping 49.8 0.005 0.56 (0.29–1.08) 0.57 (0.30–1.07) 0.33 (0.18–0.61)*  Grasping 24.2 0.265 0.59 (0.31–1.12) 0.55 (0.29–1.02) 0.66 (0.37–1.17)  Turning during descent 36.7 0.500 0.91 (0.46–1.78) 0.77 (0.41–1.46) 0.68 (0.38–1.21)  Attempted upper limb arrest 88.0 0.416 0.54 (0.18–1.62) 0.95 (0.32–2.81) 0.61 (0.22–1.70) Impact phase  Head impact 33.6 0.399 0.75 (0.37–1.54) 0.93 (0.45–1.93) 0.61 (0.30–1.22)  Hip impact 40.8 0.052 1.55 (0.81–2.97) 1.35 (0.71–2.57) 0.93 (0.53–1.62)  Sustained injury 23.2 0.001 0.94 (0.48–1.85) 0.66 (0.34–1.32) 0.44 (0.24–0.81)* Characteristics of Falls Prevalence, % of All Falls Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 56.8 0.129 0.83 (0.45–1.55) 0.83 (0.45–1.53) 1.26 (0.71–2.23)  Stand–sit transfer 17.5 0.018 0.95 (0.38–2.41) 0.97 (0.38–2.49) 2.76 (1.29–5.88)*  Sit–stand transfer 10.7 0.714 1.37 (0.42–4.50) 1.99 (0.60–6.58) 1.56 (0.51–4.79)  Sitting 14.2 0.197 2.85 (0.96–8.41) 1.89 (0.66–5.42) 2.65 (0.97–7.29)  Standing 24.0 0.766 1.36 (0.70–2.65) 1.03 (0.53–2.04) 1.01 (0.55–1.85)  Walking 33.6 0.005 0.44 (0.23–0.84)* 0.59 (0.31–1.11) 0.35 (0.19–0.63)*  Initially forward 19.3 0.055 1.05 (0.51–2.17) 0.60 (0.29–1.23) 0.47 (0.23–0.96)  Initially backward 32.2 0.094 0.88 (0.40–1.94) 1.54 (0.72–3.27) 1.71 (0.87–3.36)  Initially sideward 29.1 0.727 1.11 (0.57–2.17) 1.12 (0.60–2.11) 0.84 (0.46–1.53)  Initially straight down 19.3 0.461 0.93 (0.42–2.06) 0.84 (0.39–1.79) 1.35 (0.68–2.67) Descent phase  Stepping 49.8 0.005 0.56 (0.29–1.08) 0.57 (0.30–1.07) 0.33 (0.18–0.61)*  Grasping 24.2 0.265 0.59 (0.31–1.12) 0.55 (0.29–1.02) 0.66 (0.37–1.17)  Turning during descent 36.7 0.500 0.91 (0.46–1.78) 0.77 (0.41–1.46) 0.68 (0.38–1.21)  Attempted upper limb arrest 88.0 0.416 0.54 (0.18–1.62) 0.95 (0.32–2.81) 0.61 (0.22–1.70) Impact phase  Head impact 33.6 0.399 0.75 (0.37–1.54) 0.93 (0.45–1.93) 0.61 (0.30–1.22)  Hip impact 40.8 0.052 1.55 (0.81–2.97) 1.35 (0.71–2.57) 0.93 (0.53–1.62)  Sustained injury 23.2 0.001 0.94 (0.48–1.85) 0.66 (0.34–1.32) 0.44 (0.24–0.81)* Note: Bolded p-values indicate a p < .05. *OR significantly different from 1 at p < .05. View Large Hip and head impacts were observed in 40.8% and 33.6% of falls, respectively. Neither hip nor head impact were significantly associated with fall frequency group (p = .052 and .399, respectively; Table 2). Characteristics of Falls and Protective Responses The most common cause of imbalance preceding a fall was incorrect weight transfer, accounting for 49.5% of falls. Other causes were loss of support from an external object (22.0%), trip or stumble (12.5%), hit or bump (8.8%), collapse (6.7%), and slip (0.5%). There was no difference in the cause of imbalance, or for intrinsic vs extrinsic causes, among the four fall frequency groups (all p > .129; Table 2). Most falls were initially directed backward (33.8%) or sideways (28.8%). Less often, falls were straight down (19.2%) and forward (18.2%). Although not statistically significant (p = .055), there was a trend toward highest frequency fallers falling less often forward compared to lowest and low–medium frequency fallers (Table 2). Analysis of only falls during walking revealed that highest frequency fallers fell more often backwards during walking compared to lowest frequency fallers (OR = 3.19, 95% CI = 1.09–9.35; Table 3). Table 3. Odds Ratios for Characteristics of Falls During Walking as a Function of Fall Incidence, Based on Comparison of the Low Fall Frequency Group (Q1) to All Other Groups (Q2–Q4) Characteristics of Falls During Walking Prevalence, % Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 55.0 0.352 0.33 (0.12–0.93) 0.64 (0.27–1.52) 1.09 (0.50–2.39)  Initially forward 30.0 0.292 1.30 (0.46–3.65) 0.71 (0.28–1.77) 0.41 (0.18–0.94)  Initially backward 20.2 0.028 0.52 (0.09–2.90) 1.82 (0.49–6.73) 3.19 (1.09–9.35)*  Initially sideward 43.0 0.124 1.33 (0.48–3.69) 1.42 (0.61–3.35) 1.20 (0.55–2.62)  Initially straight down 6.8 0.208 0.22 (0.03–2.05) 0.23 (0.04–1.18) 0.73 (0.21–2.55) Descent phase  Stepping 85.0 0.302 0.93 (0.17–5.07) 0.65 (0.16–2.73) 0.38 (0.10–1.44)  Grasping 34.5 0.178 0.71 (0.28–1.80) 0.49 (0.24–1.02) 0.92 (0.45–1.87)  Turning during descent 45.9 0.683 0.90 (0.32–2.52) 0.82 (0.33–2.02) 0.68 (0.31–1.51)  Attempted upper limb arrest 91.8 0.129 2.97 (0.33–26.67) 6.41 (0.72–56.72) 0.80 (0.21–2.99) Impact phase  Head impact 45.1 0.567 1.29 (0.52–3.23) 1.12 (0.46–2.75) 0.96 (0.41–2.28)  Hip impact 52.7 0.064 1.67 (0.62–4.52) 2.03 (0.94–4.38) 0.89 (0.46–1.74)  Sustained injury 32.8 0.020 0.79 (0.28–2.19) 0.61 (0.25–1.50) 0.34 (0.15–0.74)* Characteristics of Falls During Walking Prevalence, % Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 55.0 0.352 0.33 (0.12–0.93) 0.64 (0.27–1.52) 1.09 (0.50–2.39)  Initially forward 30.0 0.292 1.30 (0.46–3.65) 0.71 (0.28–1.77) 0.41 (0.18–0.94)  Initially backward 20.2 0.028 0.52 (0.09–2.90) 1.82 (0.49–6.73) 3.19 (1.09–9.35)*  Initially sideward 43.0 0.124 1.33 (0.48–3.69) 1.42 (0.61–3.35) 1.20 (0.55–2.62)  Initially straight down 6.8 0.208 0.22 (0.03–2.05) 0.23 (0.04–1.18) 0.73 (0.21–2.55) Descent phase  Stepping 85.0 0.302 0.93 (0.17–5.07) 0.65 (0.16–2.73) 0.38 (0.10–1.44)  Grasping 34.5 0.178 0.71 (0.28–1.80) 0.49 (0.24–1.02) 0.92 (0.45–1.87)  Turning during descent 45.9 0.683 0.90 (0.32–2.52) 0.82 (0.33–2.02) 0.68 (0.31–1.51)  Attempted upper limb arrest 91.8 0.129 2.97 (0.33–26.67) 6.41 (0.72–56.72) 0.80 (0.21–2.99) Impact phase  Head impact 45.1 0.567 1.29 (0.52–3.23) 1.12 (0.46–2.75) 0.96 (0.41–2.28)  Hip impact 52.7 0.064 1.67 (0.62–4.52) 2.03 (0.94–4.38) 0.89 (0.46–1.74)  Sustained injury 32.8 0.020 0.79 (0.28–2.19) 0.61 (0.25–1.50) 0.34 (0.15–0.74)* Note: Bolded p-values indicate a p < .05, underlined p-values highlight differences from Table 2 containing the analysis of all falls. *OR significantly different from 1 at p < .05. View Large Table 3. Odds Ratios for Characteristics of Falls During Walking as a Function of Fall Incidence, Based on Comparison of the Low Fall Frequency Group (Q1) to All Other Groups (Q2–Q4) Characteristics of Falls During Walking Prevalence, % Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 55.0 0.352 0.33 (0.12–0.93) 0.64 (0.27–1.52) 1.09 (0.50–2.39)  Initially forward 30.0 0.292 1.30 (0.46–3.65) 0.71 (0.28–1.77) 0.41 (0.18–0.94)  Initially backward 20.2 0.028 0.52 (0.09–2.90) 1.82 (0.49–6.73) 3.19 (1.09–9.35)*  Initially sideward 43.0 0.124 1.33 (0.48–3.69) 1.42 (0.61–3.35) 1.20 (0.55–2.62)  Initially straight down 6.8 0.208 0.22 (0.03–2.05) 0.23 (0.04–1.18) 0.73 (0.21–2.55) Descent phase  Stepping 85.0 0.302 0.93 (0.17–5.07) 0.65 (0.16–2.73) 0.38 (0.10–1.44)  Grasping 34.5 0.178 0.71 (0.28–1.80) 0.49 (0.24–1.02) 0.92 (0.45–1.87)  Turning during descent 45.9 0.683 0.90 (0.32–2.52) 0.82 (0.33–2.02) 0.68 (0.31–1.51)  Attempted upper limb arrest 91.8 0.129 2.97 (0.33–26.67) 6.41 (0.72–56.72) 0.80 (0.21–2.99) Impact phase  Head impact 45.1 0.567 1.29 (0.52–3.23) 1.12 (0.46–2.75) 0.96 (0.41–2.28)  Hip impact 52.7 0.064 1.67 (0.62–4.52) 2.03 (0.94–4.38) 0.89 (0.46–1.74)  Sustained injury 32.8 0.020 0.79 (0.28–2.19) 0.61 (0.25–1.50) 0.34 (0.15–0.74)* Characteristics of Falls During Walking Prevalence, % Main Group Effect, p-Value Fall Frequency Group, OR (95% CI) Low–Medium (Q2) Medium–High (Q3) Highest (Q4) Initiation phase  Intrinsic cause 55.0 0.352 0.33 (0.12–0.93) 0.64 (0.27–1.52) 1.09 (0.50–2.39)  Initially forward 30.0 0.292 1.30 (0.46–3.65) 0.71 (0.28–1.77) 0.41 (0.18–0.94)  Initially backward 20.2 0.028 0.52 (0.09–2.90) 1.82 (0.49–6.73) 3.19 (1.09–9.35)*  Initially sideward 43.0 0.124 1.33 (0.48–3.69) 1.42 (0.61–3.35) 1.20 (0.55–2.62)  Initially straight down 6.8 0.208 0.22 (0.03–2.05) 0.23 (0.04–1.18) 0.73 (0.21–2.55) Descent phase  Stepping 85.0 0.302 0.93 (0.17–5.07) 0.65 (0.16–2.73) 0.38 (0.10–1.44)  Grasping 34.5 0.178 0.71 (0.28–1.80) 0.49 (0.24–1.02) 0.92 (0.45–1.87)  Turning during descent 45.9 0.683 0.90 (0.32–2.52) 0.82 (0.33–2.02) 0.68 (0.31–1.51)  Attempted upper limb arrest 91.8 0.129 2.97 (0.33–26.67) 6.41 (0.72–56.72) 0.80 (0.21–2.99) Impact phase  Head impact 45.1 0.567 1.29 (0.52–3.23) 1.12 (0.46–2.75) 0.96 (0.41–2.28)  Hip impact 52.7 0.064 1.67 (0.62–4.52) 2.03 (0.94–4.38) 0.89 (0.46–1.74)  Sustained injury 32.8 0.020 0.79 (0.28–2.19) 0.61 (0.25–1.50) 0.34 (0.15–0.74)* Note: Bolded p-values indicate a p < .05, underlined p-values highlight differences from Table 2 containing the analysis of all falls. *OR significantly different from 1 at p < .05. View Large The most common activity at time of fall was walking (33.6%) or standing (24.0%). Less often falls occurred during stand-to-sit transfers (17.5%), sitting (14.2%), and sit-to-stand transfers (10.7%). Highest frequency fallers were more likely to fall during a stand-to-sit transition compared to lowest frequency fallers (OR = 2.76, 95% CI = 1.29–5.88; Table 2). Moreover, highest frequency fallers were less likely to fall during walking compared to lowest frequency fallers (OR = 0.35, 95% CI = 0.19–0.63). The association between injury risk and fall frequency group remained significant after correcting for the height of the fall (above, at, or lower than standing height; p < .027). Falls during walking had a higher risk for injury (OR = 2.2, 95% CI = 1.6–2.6 vs not walking). Nevertheless, the association between injury risk and fall frequency group remained significant after stratifying for activity at the time of falling (walking or transferring; p < .044). For falls during walking (N = 265), highest frequency fallers remained less likely to injure themselves than lower frequency fallers (OR = 0.34, 95% CI = 0.15–0.74 for Q4 vs Q1 and OR = 0.43, 95% CI = 0.18–1.00 for Q4 vs Q2; Table 3). Attempts to recover balance by grasping or body rotation during descent, and attempts to arrest the fall with the upper limbs occurred frequently; their prevalence was not associated with fall frequency (Table 1, all p > .265). Recovery steps were observed in 48.9% of falls and associated significantly with fall frequency (p = .005). Highest frequency fallers had lower odds of executing recovery steps during a fall compared to lowest frequency fallers (OR = 0.33, 95% CI= 0.18–0.61). Similar trends were observed for low–medium and medium–high frequency fallers compared to lowest frequency fallers. Recovery steps were less often performed during stand-to-sit falls (OR = 0.14, 95% CI = 0.08–0.26 vs not stand-to-sit fall) and more often during falls during walking (OR = 12.24, 95% CI = 6.13–24.45 vs not walking). Analysis of only falls during walking revealed no significant differences in the odds of executing recovery steps between fall frequency groups (Table 3). Discussion We extend the current literature on the relationships between fall frequency, fall characteristics, and risk for injury, by linking injury data to evidence of the circumstances of falls from unique video footage of real-life falls experienced by older adults in LTC. Our results support previous findings (10–13) that more frequent fallers are less likely than less frequent fallers to injure themselves in the event of a fall. Our findings do not support the notion that recurrent fallers more often exhibit protective responses during falls such as upper limb fall arrest, or rotation during descent. Instead, the difference in injury risk appears to be due in part to differences in the activities leading to falls. Falls during walking created higher risk for injury, and more frequent fallers were less likely to fall during walking. Recurrent fallers were instead more likely to fall while transferring, which tends to involve a lower change in potential energy. While fall frequency did not associate with odds of impact to the head (or to the hip), the lower energy content may have caused the severity of head impacts to be lower in higher frequency fallers. However, activity at the time of falling could not explain all of the observed trends. More frequent fallers had lower risk for injury per fall from falls during walking, perhaps due to their increased tendency to fall backward, which creates lower risk for head impact than falling forward (18). Moreover, the relation between injury risk and fall frequency persisted after correcting for fall height. This suggests that injury risk may be dominated by tissue tolerance to injury (as opposed to applied force), or depend on fall characteristics that were not examined in this study, and perhaps cannot be easily discerned from video. These may include differences in impact velocity, impact location, or sharing of impact energy between multiple body parts. Compared to lower frequency fallers, higher frequency fallers fell more often during transfers and less often during walking. An explanation for differences in activity at time of the fall could be shared group demographics that increase activity-specific fall risk, or decreased exposure to specific activities. In contrast to previous studies that reported that recurrent fallers are more often female, have slower reaction time, impaired cognition and lower muscle strength (7,11), we did not observe differences between fall frequency groups in gender, age or observational measures of physical and cognitive status. We did note a higher prevalence of antipsychotic and antianxiety medication in the highest frequency fallers. These observations may call for more sensitive tools of physical and cognitive capacity in LTC, and investigation of the effect of medication on fall rate. Altered habitual activity, such as less walking, could have decreased exposure and thereby reduced the number of falls during walking. Previous studies using self-reports indeed suggest that recurrent fallers are less physically active (26). Greater understanding is required of the quantity of daily activities in frequent fallers and of how different activity types associate with fall risk. Our study has important limitations. First, we video-captured only falls in communal areas of two LTC facilities (about 40% of falls (21)). We cannot assume that our results are applicable for falls in private areas (eg, bedrooms and bathrooms), or for falls in other populations (eg, community-dwelling) older adults. Second, we established our groups as quartiles of fall frequency as there are no established cut-offs for fall frequency among LTC residents. Our results remained consistent when we used tertiles and quintiles for analyses. Moreover, the median fall frequency of 2.5 falls per year was comparable to those previously reported for the LTC population (1,2,27). Third, we used the LTC facilities’ fall incident reports to identify injuries, and we did not differentiate severity of injury. The 23.2% injury rate we report is similar to values reported in other studies of falls in the LTC population (6–9). Finally, while we analyzed videos to determine the occurrence of impact to specific body parts, and again we did not estimate the severity of a given impact, nor did we consider the sequence of impacts to the various body parts, which are important outcomes for future research. In summary, we used footage of real-life falls to examine the association between fall frequency, injury risk, and fall circumstances in LTC. We found that, when compared to lower frequency fallers, falls in high frequency fallers resulted in a lower injury risk per fall, and occurred more often while transferring and less often while walking. Our results support the implementation of strategies to prevent falls and fall-related injuries in LTC (eg, exercise, medication review, hip protectors) for both frequent and infrequent fallers, as the latter are more apt to sustain injury during a fall. While only 5% of older people reside in LTC, falls in this population account for 20% of all fall-related injury deaths among older people (2,28). Thus, future studies should examine the effectiveness of targeted interventions in this most frail population. Funding This work was supported by grants from the Canadian Institute for Health Research (AMG-100487, TIR-103945, and TEI-138295; CIHR New Investigator Award to J.S.-G.), the AGE-WELL National Centres of Excellence (AW CRP 2015-WP5.2) and the Michael Smith Foundation for Health Research (Postdoctoral Fellowship Awards #16606 to K.S.S. and #16556 to Y.Y., and Research Scholar Award to J.S.-G.). References 1. Rapp K , Becker C , Cameron ID , König H-H , Büchele G . Epidemiology of falls in residential aged care: analysis of more than 70,000 falls from residents of bavarian nursing homes . J Am Med Dir Assoc . 2012 ; 13 : 187 .e1–6. doi: 10.1016/j.jamda.2011.06.011 Google Scholar CrossRef Search ADS PubMed 2. Rubenstein LZ , Josephson KR , Robbins AS . Falls in the nursing home . Ann Intern Med . 1994 ; 121 : 442 – 451 . doi:10.7326/0003- 4819-121-6-199409150-00009 Google Scholar CrossRef Search ADS PubMed 3. Rubenstein LZ . Falls in older people: epidemiology, risk factors and strategies for prevention . Age Ageing . 2006 ; 35 ( suppl 2 ): ii37 – ii41 . doi: 10.1093/ageing/afl084 Google Scholar CrossRef Search ADS PubMed 4. Grisso JA , Kelsey JL , Strom BL et al. Risk factors for falls as a cause of hip fracture in women. The Northeast Hip Fracture Study Group . N Engl J Med . 1991 ; 324 : 1326 – 1331 . doi: 10.1056/NEJM199105093241905 Google Scholar CrossRef Search ADS PubMed 5. Head Injuries in Canada: A Decade of Change (1994-1995 to 2003-2004). Canadian Institute for Health Information, 2006. https://secure.cihi.ca/free_products/ntr_head_injuries_2006_e.pdf. 6. Lloyd BD , Williamson DA , Singh NA et al. Recurrent and injurious falls in the year following hip fracture: a prospective study of incidence and risk factors from the Sarcopenia and Hip Fracture study . J Gerontol A Biol Sci Med Sci . 2009 ; 64 : 599 – 609 . doi: 10.1093/gerona/glp003 Google Scholar CrossRef Search ADS PubMed 7. Towne SD , Cho J , Smith ML , Ory MG . Factors associated with injurious falls in residential care facilities . J Aging Health . 2016 : 29 : 669 – 687 . doi: 10.1177/0898264316641083 Google Scholar CrossRef Search ADS PubMed 8. O’Loughlin JL , Robitaille Y , Boivin JF , Suissa S . Incidence of and risk factors for falls and injurious falls among the community-dwelling elderly . Am J Epidemiol . 1993 ; 137 : 342 – 354 . doi:10.1093/oxfordjournals.aje.a116681 Google Scholar CrossRef Search ADS PubMed 9. Koski K . Risk factors for major injurious falls among the home-dwelling elderly by functional abilities . Gerontology . 1998 ; 44 : 232 – 238 . doi:10.1159/ 000022017 Google Scholar CrossRef Search ADS PubMed 10. Poss JW , Hirdes JP . Very frequent fallers and future fall injury: continuous risk among community-dwelling home care recipients . J Aging Health . 2016 ; 28 : 587 – 599 . doi: 10.1177/0898264315599941 Google Scholar CrossRef Search ADS PubMed 11. Nevitt MC , Cummings SR , Hudes ES . Risk factors for injurious falls: a prospective study . J Gerontol . 1991 ; 46 : M164 – M170 . doi:10.1093/geronj/ 46.5.M164 Google Scholar CrossRef Search ADS PubMed 12. Tinetti ME . Factors associated with serious injury during falls by ambulatory nursing home residents . J Am Geriatr Soc . 1987 ; 35 : 644 – 648 . doi:10.1111/ j.1532-5415.1987.tb04341.x Google Scholar CrossRef Search ADS PubMed 13. Sambrook PN , Cameron ID , Chen JS et al. Influence of fall related factors and bone strength on fracture risk in the frail elderly . Osteoporos Int . 2007 ; 18 : 603 – 610 . doi: 10.1007/s00198-006-0290-z Google Scholar CrossRef Search ADS PubMed 14. Deandrea S , Bravi F , Turati F , Lucenteforte E , La Vecchia C , Negri E . Risk factors for falls in older people in nursing homes and hospitals. A systematic review and meta-analysis . Arch Gerontol Geriatr . 2013 ; 56 : 407 – 415 . doi: 10.1016/j.archger.2012.12.006 Google Scholar CrossRef Search ADS PubMed 15. Baranzini F , Diurni M , Ceccon F et al. Fall-related injuries in a nursing home setting: is polypharmacy a risk factor ? BMC Health Serv Res . 2009 ; 9 : 228 . doi: 10.1186/1472-6963-9-228 Google Scholar CrossRef Search ADS PubMed 16. Nevitt MC , Cummings SR , The Study of Osteoporotic Fractures Research Group . Type of fall and risk of hip and wrist fractures: the study of osteoporotic fractures . J Am Geriatr Soc . 1993 ; 41 : 1226 – 1234 . doi:10.1111/ j.1532-5415.1993.tb07307.x Google Scholar CrossRef Search ADS PubMed 17. Schwartz AV , Kelsey JL , Sidney S , Grisso JA . Characteristics of falls and risk of hip fracture in elderly men . Osteoporos Int . 1998 ; 8 : 240 – 246 . doi: 10.1007/s001980050060 Google Scholar CrossRef Search ADS PubMed 18. Schonnop R , Yang Y , Feldman F , Robinson E , Loughin M , Robinovitch SN . Prevalence of and factors associated with head impact during falls in older adults in long-term care . CMAJ . 2013 ; 185 : E803 – E810 . doi: 10.1503/cmaj.130498 Google Scholar CrossRef Search ADS PubMed 19. Yang Y , Mackey DC , Liu-Ambrose T , Feldman F , Robinovitch SN . Risk factors for hip impact during real-life falls captured on video in long-term care . Osteoporos Int . 2016 ; 27 : 537 – 547 . doi: 10.1007/s00198-015-3268-x Google Scholar CrossRef Search ADS PubMed 20. Yardley L , Beyer N , Hauer K , Kempen G , Piot-Ziegler C , Todd C . Development and initial validation of the Falls Efficacy Scale-International (FES-I) . Age Ageing . 2005 ; 34 : 614 – 619 . doi: 10.1093/ageing/afi196 Google Scholar CrossRef Search ADS PubMed 21. Robinovitch SN , Feldman F , Yang Y et al. Video capture of the circumstances of falls in elderly people residing in long-term care: an observational study . Lancet . 2013 ; 381 : 47 – 54 . doi: 10.1016/S0140-6736(12)61263-X Google Scholar CrossRef Search ADS PubMed 22. Yang Y , Schonnop R , Feldman F , Robinovitch SN . Development and validation of a questionnaire for analyzing real-life falls in long-term care captured on video . BMC Geriatr . 2013 ; 13 : 40 . doi: 10.1186/1471-2318-13-40 Google Scholar CrossRef Search ADS PubMed 23. Hayes WC , Myers ER , Morris JN , Gerhart TN , Yett HS , Lipsitz LA . Impact near the hip dominates fracture risk in elderly nursing home residents who fall . Calcif Tissue Int . 1993 ; 52 : 192 – 198 . doi:10.1007/BF00298717 Google Scholar CrossRef Search ADS PubMed 24. Morris JN , Fries BE , Mehr DR et al. MDS Cognitive Performance Scale . J Gerontol . 1994 ; 49 : M174 – M182 . Google Scholar CrossRef Search ADS PubMed 25. Morris JN , Fries BE , Morris SA . Scaling ADLs within the MDS . J Gerontol A Biol Sci Med Sci . 1999 ; 54 : M546 – M553 . doi:10.1093/geronj/49.4.M174 Google Scholar CrossRef Search ADS PubMed 26. Peeters GM , van Schoor NM , Pluijm SM , Deeg DJ , Lips P . Is there a U-shaped association between physical activity and falling in older persons ? Osteoporos Int . 2010 ; 21 : 1189 – 1195 . doi: 10.1007/s00198-009-1053-4 Google Scholar CrossRef Search ADS PubMed 27. Lord SR , March LM , Cameron ID et al. Differing risk factors for falls in nursing home and intermediate-care residents who can and cannot stand unaided . J Am Geriatr Soc . 2003 ; 51 : 1645 – 1650 . doi: 10.1046/ j.1532-5415.2003.51518.x Google Scholar CrossRef Search ADS PubMed 28. Health, United States, 2005: With Chartbook on Trends in the Health of Americans . Hyattsville, MD: National Center for Health Statistics ; 2005 . © The Author(s) 2017. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

The Journals of Gerontology Series A: Biomedical Sciences and Medical SciencesOxford University Press

Published: Oct 12, 2017

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off