Abstract
Background
falls are common in older adults, and any fall from standing height onto a rigid surface has the potential to cause a serious brain injury or bone fracture. Safe strategies for falling in humans have traditionally been difficult to study.
Objective
to determine whether specific ‘safe landing’ strategies (body rotation during descent, and upper limb bracing) separate injurious and non-injurious falls in seniors.
Design
observational cohort study.
Setting
two long-term care homes in Vancouver BC.
Methods
videos of 2,388 falls experienced by 658 participants (mean age 84.0 years; SD 8.1) were analysed with a structured questionnaire. General estimating equations were used to examine how safe landing strategies associated with documented injuries.
Results
injuries occurred in 38% of falls, and 4% of falls caused injuries treated in hospitals. 32% of injuries were to the head. Rotation during descent was common and protective against injury. In 43% of falls initially directed forward, participants rotated to land sideways, which reduced their odds for head injury 2-fold. Upper limb bracing was used in 58% of falls, but rather than protective, bracing was associated with an increased odds for injury, possibly because it occurred more often in the demanding scenario of forward landings.
Conclusions
the risk for injury during falls in long-term care was reduced by rotation during descent, but not by upper limb bracing. Our results expand our understanding of human postural responses to falls, and point towards novel strategies to prevent fall-related injuries.
Keywords: falls, injury, protective responses, older adults, video capture, older people
Key Points
Among 2,388 falls in long-term care captured on video, 38% caused injuries and 4% caused injuries treated in hospitals.
Rotation during descent to change the direction of the fall was common and protective.
The patterns of rotation allowed for the avoidance of forward landings, which created the highest risk for head injury.
Falls that involved upper limb bracing were more likely to result in injury.
Introduction
Falls are common in older adults. Approximately 30% of adults over age 65 years living independently, and 60% of older adults residing in long-term care (LTC) will fall at least once per year, and many will fall repeatedly [1, 2]. Fortunately, most falls in older adults do not cause serious physical injuries. In the LTC setting, ~20–30% of falls cause some sort of injury, and 5% result in injuries treated in hospitals [1, 3–5].
The reasons why most falls are non-injurious are not well understood. Any fall from standing height onto a rigid surface has the potential to cause serious injury, even in young healthy adults [6, 7]. The energy available is more than sufficient to cause head impacts that lead to skull fracture or brain injury [8–11], or pelvis impacts that cause hip fracture [7, 12, 13].
Mechanisms must be acting during falls to absorb and distribute the impact energy in a way that prevents contact forces (and tissue stresses and strains) from exceeding injury thresholds. The nature of safe landing responses has been investigated in cats and other animals [14–16]. However, only a small number of studies have employed methods for eliciting falls in humans, with no restraint harness, and little chance for balance recovery [17–19]. Furthermore, participants in these studies have been restricted to young, healthy individuals.
The limited evidence suggests that two mechanisms used by humans for safe landing during falls are upper limb bracing and body rotation during descent. Upper limb bracing involves impacting the ground (or an object in the environment) with the hand(s) and/or forearm(s), to arrest downward movement of the torso, and thereby prevent impact and injury to the head [6, 20, 21]. Young adults who experienced unexpected falls in the lab environment consistently exhibited upper limbs bracing [17, 18]. A recent analysis of falls in babies captured on video shows that upper limb bracing emerges early in childhood, and tends to be tailored to the context of the fall [22].
Body rotation during descent involves altering the direction of a fall. Young adults are able to rotate the body after being unexpectedly released into a fall [23]. Judo practitioners are taught to rotate backward during the descent stage of a (forward or sideways) fall, and safely absorb energy through impact to the thigh, buttock(s) and torso [24, 25].
There is scant evidence on the prevalence and effectiveness of rotation or bracing in preventing injuries during falls in older adults. In a study of falls and fractures in community-dwelling older women, approximately 46% of women self-reported falling on their hand/wrist, and those who landed on their hand had a decreased risk for hip fracture [26–28] and an increased risk for distal radius fracture [27, 29]. It is unclear whether upper limb bracing persists as a protective response during falls in the more frail LTC population. Furthermore, no study has reported on whether rotation during descent is common and protective for falls in older adults.
We addressed these knowledge gaps by linking documented injuries to video evidence on the movement patterns of older adults during falls in LTC, and testing whether the odds for injury associate with upper limb bracing and body rotation during descent.
Methods
Participants and video footage of falls
Study participants were residents of two LTC homes in the Greater Vancouver area (Good Samaritan Delta View having 312 beds, and New Vista having 236 beds), involved in an observational study where video footage of falls and corresponding fall incident reports were shared as secondary data with our research team. See Appendix A for additional details on video capture. The research protocol was approved by the Office of Research Ethics at Simon Fraser University (protocol number H21–00741).
Between January 2010 and March 2020, a total of 2388 falls experienced by 658 participants over the age of 65 years were captured on video and analysed. For each fall, information was acquired from the fall incident report on the types and locations of injuries caused by the fall, and the participant age, sex, height and weight. Participants had a mean age of 84.0 years (SD = 8.1 years), and 56.4% were female. We captured only one fall on video for 43% of participants (n = 280), two to five falls for 41% (n = 270) and more than five falls for 16% (n = 108).
A subset of participants (260 of 658 total) provided written informed consent for us to access information on physical and cognitive status, disease diagnoses and medications via the Minimum Data Set (MDS). See Appendix A for additional details on MDS data. Among these 260 participants, 68% had moderate to severe cognitive impairment, 57% were not independent in performing activities of daily living, 41% were taking antipsychotic medications and 48% were on antidepressants (Table 1).
Table 1.
Characteristics of the 260 participants who provided consent to access medical records
| All participants with falls on video (n = 260) | Participants without injuries in any fall on video (n = 86) | Participants with injuries in all falls on video (n = 62) | Participants with and without injuries from falls on video (n = 112) | P* | |
|---|---|---|---|---|---|
| Demographics and health status | |||||
| Age, mean (SD) | 84.1 (7.8) | 84.1 (7.8) | 85.1 (8.1) | 83.7 (7.4) | 0.544 |
| Female, n (%) | 149 (57.3) | 46 (53.5) | 34 (54.8) | 69 (61.6) | 0.469 |
| Height (cm), mean (SD) | 163.2 (10.8) | 163.4 (9.4) | 164.9 (11.6) | 162.0 (11.2) | 0.246 |
| Body mass (kg), mean (SD) | 62.5 (16.0) | 63.7 (15.8)B | 67.9 (17.8)B | 58.4 (13.9)A | <0.001 |
| BMI (kg/m2), mean (SD)a | 23.4 (5.2) | 23.7 (5.3)AB | 25.0 (5.9)B | 22.2 (4.5)A | 0.003 |
| Dependent in ADL performance, n (%)b | 149 (57.3) | 52 (60.5) | 31 (50.0) | 66 (58.9) | 0.402 |
| Moderate to severe cognitive impairment, n (%)c | 176 (67.7) | 61 (71.8)B | 33 (53.2)A | 81 (72.3)B | 0.022 |
| Comorbidities, n (%) | |||||
| Diabetes | 60 (23.1) | 17 (19.8) | 17 (27.4) | 26 (23.2) | 0.551 |
| Cardiac dysrhythmia | 15 (5.8) | 4 (4.7) | 6 (9.7) | 5 (4.5) | 0.318 |
| Congestive heart failure | 20 (7.7) | 8 (9.3) | 6 (9.7) | 6 (5.4) | 0.468 |
| Hypertension | 129 (49.6) | 38 (44.2) | 39 (62.9) | 52 (46.4) | 0.054 |
| Hypotension | 12 (4.6) | 2 (2.3) | 5 (8.1) | 5 (4.5) | 0.259 |
| Alzheimer’s disease | 66 (25.3) | 24 (27.9)B | 8 (12.9)A | 34 (30.4)B | 0.033 |
| Stroke | 39 (15.0) | 18 (20.9) | 7 (11.3) | 14 (12.5) | 0.166 |
| Parkinson’s disease | 10 (3.8) | 4 (4.7) | 2 (3.2) | 4 (3.6) | 0.888 |
| Emphysema/COPD | 30 (11.5) | 11 (12.8) | 6 (9.7) | 12 (11.6) | 0.842 |
| Use of medications, n (%) | |||||
| Antipsychotics | 106 (40.8) | 33 (38.4) | 22 (35.5) | 51 (45.5) | 0.372 |
| Antianxiety agents | 48 (18.5) | 17 (19.8) | 8 (12.9) | 23 (20.5) | 0.430 |
| Antidepressants | 125 (48.1) | 40 (46.5) | 32 (51.6) | 53 (47.3) | 0.810 |
| Hypnotics | 54 (20.8) | 21 (24.4) | 13 (21.0) | 20 (17.8) | 0.529 |
| Diuretics | 51 (19.6) | 18 (20.9) | 17 (27.4) | 16 (14.3) | 0.105 |
| Analgesics | 129 (49.6) | 38 (44.2) | 37 (59.7) | 54 (48.2) | 0.164 |
Data represent participant status at the time of the first fall recorded on video.
*Statistical comparisons between participants who fell and injured versus fell and did not injure are based on Chi-square for categorical variables, and one-way ANOVA for age. Values shown in bold indicate statistically significant factors (P < 0.05). Groups with the same superscript (A, B) are statistically similar.
BMI = body mass index (kg/m2).
Based on the ‘Activities of Daily Living’ score from Minimal Data Set; see text for explanation.
Based on the ‘Cognitive Performance Scale’ score from Minimal Data Set; see text for explanation.
Video analysis
Videos were analysed by teams of three expert raters (research assistants trained by one of the authors [S.N.R.]), who worked together to review video footage for each fall, and reach consensus on the best answer to each question in a structured, validated questionnaire [30]. See Appendix A for additional details on rater training. The questionnaire classified the initial direction of the fall at the onset of steady downward movement (as backward, forward, sideways or straight down), the configuration of the body at landing from the fall based on the aspect of the body that came into primary contact with the floor (posterior = backward, anterior = forward, lateral = sideways), and upper limb fall arrest (yes or no) defined as a deliberate attempt to arrest the fall involving contact of the hand(s) or elbow(s) with the external environment (e.g. the ground, furniture or walls). We classified body rotation during descent based on differences between the initial fall direction and the landing direction. In analysing the effect on injury of body rotation during descent, we excluded cases where the initial fall direction was classified as ‘straight down’ (n = 391).
Injury data
For each of the 2,388 falls, injuries were determined from fall incident reports and review of medical records for the 7-day period after the fall. Some falls caused injuries to more than one body part, which we treated as separate injuries. If more than one type of injury was described for a specific body part, we recorded it as a single injury, classified as most severe based on the following order: fracture, dislocation, laceration, bruising and pain. We defined ‘serious’ injuries as cases that led to a hospital visit or stay, or a formal diagnostic procedure such as an X-ray or sutures [31]. All injuries that were not classified as serious were classified as ‘minor’. We defined ‘head injuries’ as injuries occurring to the head or face, regardless of whether these were serious or minor.
Statistical analysis
We used binary logistic general estimating equations (GEE) to determine how the odds for injury associated with the characteristics of falls observed on video. An advantage of the GEE method is that it accounts for potential correlation between repeated falls in a given participant, by including participant ID as a random factor, and fall number as a within-subject factor in the model. We ran separate models comparing falls involving no injury to falls that resulted in (i) any type of injury, (ii) serious injury and (iii) head injury. Age and sex were included as covariates, since these are established risk factors for fall-related injury [5, 27, 32–35]. We report odds ratios and 95% confidence intervals. We also report results from Chi-square and ANOVA comparing the characteristics of participants who experienced injurious versus non-injurious falls. All statistical analyses were conducted with SPSS Statistics (Version 27, IBM Corp, Armonk, NY,USA) using a significance level of α = 0.05.
Results
Injury patterns
Among the 2,388 falls we analysed, 912 falls (38.2%) caused a total of 1,223 injuries, including 99 falls (4.1%) that caused 104 serious injuries. Injury was documented to one body part in 678 falls, two body parts in 167 falls and three or more body parts in 67 falls.
The most common sites of injury (Table 2) were the head (n = 390), the torso/upper arm (n = 220) and elbow/forearm (n = 174). Serious injuries occurred most often to the hip/pelvis (n = 45) and the head (n = 35). Of the 53 falls that caused fractures, hip fractures accounted for 31 cases and upper limb fractures accounted for 11.
Table 2.
Distribution in anatomical locations and types of injuries, n = 1,223 injuries from 2,388 falls by 658 participants
| Injury location (n, (% of falls where injury was observed)) | ||||||||
|---|---|---|---|---|---|---|---|---|
| Head/face | Torso/upper arm | Elbow/forearm | Knee/shin | Hip/pelvis | Hand/wrist | Ankle | Unspecified | |
| Injury type | ||||||||
| Pain, with or without palpation | 35 (1.5) |
93 (3.9) |
9 (0.4) |
37 (1.5) |
61 (2.6) |
9 (0.4) |
2 (0.1) |
12 (0.5) |
| Bruise/redness/swelling | 184 (7.7) |
64 (2.7) |
45 (1.9) |
62 (2.6) |
35 (1.5) |
53 (2.2) |
13 (0.5) |
1 (<0.1) |
| Cut/laceration | 168 (7.0) |
51 (2.1) |
115 (4.8) |
49 (2.1) |
4 (0.2) |
56 (2.3) |
0 (0.0) |
1 (<0.1) |
| Sprain/strain/dislocation | 0 (0.0) |
1 (<0.1) |
0 (0.0) |
2 (0.1) |
1 (<0.1) |
2 (0.1) |
2 (0.1) |
0 (0.0) |
| Fracture | 3 (0.1) |
11 (0.5) |
4 (0.2) |
0 (0.0) |
33 (1.4) |
2 (0.1) |
0 (0.0) |
0 (0.0) |
| Injury severity | ||||||||
| Minor | 355 (14.9) |
207 (8.7) |
169 (7.1) |
149 (6.2) |
89 (3.7) |
120 (5.0) |
16 (0.7) |
14 (0.6) |
| Serious | 35 (1.5) |
13 (0.5) |
5 (0.2) |
2 (0.1) |
45 (1.9) |
3 (0.1) |
1 (<0.1) |
0 (0.0) |
| Total | 390 (16.3) |
220 (9.2) |
174 (7.3) |
151 (6.3) |
134 (5.6) |
123 (5.2) |
17 (0.7) |
14 (0.6) |
For three injuries (one to the knee/shin, one to the elbow/forearm and one to the hand/wrist), the injury type was unspecified.
Resident characteristics and injury risk
Among the 260 participants who fell and provided consent for access to medical records, the baseline characteristics of participants who fell and experienced injury were, in general, similar to those who fell and were not injured (Table 1). The exceptions were substantial cognitive impairment and Alzheimer’s disease, which were less common in participants who experienced injury when falling. There were no differences in baseline age, sex, body mass, height or BMI between individuals who fell and always experienced injury, and those who fell and never injured.
Rotation during descent and injury risk
The odds for injury associated with initial fall direction, landing direction and rotation during descent (Tables 3 and 4). The odds were higher in falls initially directed forward than backward for any injury (OR = 3.05; 95% CI = 2.36–3.95), serious injury (4.33; 2.43–7.70) and head injury (4.06; 2.98–5.54). The odds were also higher in falls initially directed forward than sideways for any injury (1.99; 1.57–2.54) and head injury (2.38; 1.72–3.30), but not for serious injury (1.29; 0.82–2.02). Falls initially directed sideways had higher odds than backward for any injury (1.51; 1.21–1.88), serious injury (3.26; 1.98–5.36) and head injury (1.61; 1.21–2.13). The trends were similar but the differences were generally larger for the direction at landing from the fall.
Table 3.
Associations between fall direction and risk for injury (n = 2,388 falls)
| Any injury versus no injury | Serious injury versus no injury | Head injury versus no injury | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Fall characteristic | Number of falls (row %) | Odds ratio (95% CI) | Number of falls (row %) | Odds ratio (95% CI) | Number of falls (row %) | Odds ratio (95% CI) | |||
| Any injury | No injury | Serious injury | No injury | Head injury | No injury | ||||
| Fell backward versus |
253 (31.4) |
553 (68.6) |
19 (3.3) |
553 (96.7) |
99 (15.2) |
553 (84.8) |
|||
| Fell sideways | 291 (41.7) |
407 (58.3) |
1.51
(1.21–1.88) * |
47 (10.4) |
407 (89.6) |
3.26
(1.98–5.36) * ab |
125 (23.5) |
407 (76.5) |
1.61
(1.21–2.13) * b |
| Fell forward | 286 (58.0) |
207 (42.0) |
3.05
(2.36–3.95) * |
31 (13.0) |
207 (87.0) |
4.33
(2.43–7.70) * |
151 (42.2) |
207 (57.8) |
4.06
(2.98–5.54) * |
| Fell sideways versus |
291 (41.7) |
407 (58.3) |
47 (10.4) |
407 (89.6) |
125 (23.5) |
407 (76.5) |
|||
| Fell forward | 286 (58.0) |
207 (42.0) |
1.99
(1.57–2.54) * b |
31 (13.0) |
207 (87.0) | 1.29 (0.82–2.02)b |
151 (42.2) |
207 (57.8) |
2.38
(1.72–3.30) * b |
| Landed backward versus |
377 (29.7) |
893 (70.3) |
27 (2.9) |
893 (97.1) |
147 (14.1) |
893 (85.9) |
|||
| Landed sideways | 376 (44.0) |
479 (56.0) |
1.83
(1.52–2.22) * |
56 (10.5) |
479 (89.5) |
3.86
(2.56–5.82) * ab |
138 (22.4) |
479 (77.6) |
1.74
(1.36–2.23) * a |
| Landed forward | 159 (60.5) |
104 (39.5) |
3.72
(2.88–4.80) * |
16 (13.3) |
104 (86.7) |
5.77
(3.13–10.63) * a |
105 (50.2) |
104 (49.8) |
6.25
(4.54–8.60) * |
| Landed sideways versus |
376 (44.0) |
479 (56.0) |
56 (10.5) |
479 (89.5) |
138 (22.4) |
479 (77.6) |
|||
| Landed forward | 159 (60.5) |
104 (39.5) |
1.99
(1.50–2.64) * a |
16 (13.3) |
104 (86.7) |
1.31 (0.75–2.29)b |
105 (50.2) |
104 (49.8) |
3.61
(2.61–4.98) * a |
*Statistically significant at P ≤ 0.05.
Age was significant as a covariate.
Sex was significant as a covariate.
Table 4.
Associations between rotation during descent, upper limb protective responses and risk and risk for injury (n = 2,388 falls)
| Any injury versus no injury | Serious injury versus no injury | Head injury versus no injury | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Fall characteristic | Number of falls (row %) | Odds ratio (95% CI) | Number of falls (row %) | Odds ratio (95% CI) | Number of falls (row %) | Odds ratio (95% CI) | |||
| Any injury | No injury | Serious injury | No injury | Head injury | No injury | ||||
| Strictly forward versus |
133 (63.9) |
75 (36.1) |
14 (15.7) |
75 (84.3) |
89 (54.3) |
75 (45.7) |
|||
| Rotated from forward to sideways | 123 (57.5) |
91 (42.5) |
0.75 (0.49–1.15)a |
14 (13.3) |
91 (86.7) |
0.88 (0.42–1.84) |
53 (36.8) |
91 (63.2) |
0.50
(0.32–0.79) * a |
| Rotated from forward to backward | 30 (42.3) |
41 (57.7) |
0.36
(0.21–0.63) * a |
3 (6.8) |
41 (93.2) |
0.38 (0.13–1.13) |
9 (18.0) |
41 (82.0) |
0.15
(0.07–0.34) * a |
| Strictly sideways versus |
182 (45.1) |
222 (54.9) |
35 (13.6) |
222 (86.4) |
68 (23.5) |
222 (76.5) |
|||
| Rotated from sideways to forward | 17 (48.6) |
18 (51.4) |
1.14 (0.56–2.29)a |
2 (10.0) |
18 (90.0) |
0.62 (0.14–7.79)ab |
15 (45.5) |
18 (54.5) |
2.85
(1.34–6.05) * b |
| Rotated from sideways to backward | 92 (35.5) |
167 (64.5) |
0.65
(0.45–0.92) * a |
10 (5.6) |
167 (94.4) |
0.37
(0.20–0.69) * ab |
42 (20.1) |
167 (79.9) |
0.75 (0.49–1.15)b |
| Strictly backward versus |
193 (30.5) |
440 (69.5) |
13 (2.9) |
440 (97.1) |
82 (15.7) |
440 (84.3) |
|||
| Rotated from backward to sideways | 60 (35.3) |
110 (64.7) |
1.22 (0.85–1.76) |
6 (5.2) |
110 (94.8) |
1.83 (0.74–4.54) |
17 (13.4) |
110 (86.6) |
0.81 (0.47–1.39)b |
| Rotated from backward to forward | 0 (0.0) |
3 (100.0) |
. . . | 0 (0.0) |
3 (100.0) |
. . . | 0 (0.0) |
3 (100.0) |
. . . |
| No upper limb arrest versus |
357 (36.0) |
636 (64.0) |
32 (4.8) |
636 (95.2) |
146 (18.7) |
636 (81.3) |
|||
| Upper limb arrest | 555 (39.8) |
840 (60.2) |
1.22
(1.04–1.43) * |
67 (7.4) |
840 (92.6) |
1.82
(1.21–2.74) * ab |
244 (22.1) |
840 (77.9) |
1.36
(1.09–1.70) * b |
*Statistically significant at P ≤ 0.05.
Age was significant as a covariate.
Sex was significant as a covariate.
Odds ratio could not be calculated.
Body rotation during descent occurred in 57% of falls initially directed forward, 42% of falls initially directed sideways and 21% of falls initially directed backward (Table 4). Among falls initially directed forward, 43% resulted in sideways landings and 14% resulted in backwards landings. Rotating from forward to land sideways decreased the odds for head injury (0.50; 0.32–0.79), and rotating to land backward decreased the odds for head injury (0.15; 0.07–0.34) and any injury (0.36; 0.21–0.63). Among falls initially directed sideways, 37% resulted in backward landings and 5% resulted in forward landings. Rotating from sideways to land backward decreased the odds of any injury (0.65; 0.45–0.92) and serious injury (0.37; 0.20–0.69), while rotating to land forward increased the odds for head injury (2.85; 1.34–6.05). For falls initially directed backward, 21% resulted in sideways landings and 0.3% involved forward landings. Rotation during falls initially directed backward had no effect on the odds for injury.
Upper limb protective responses and injury risk
Upper limb bracing was observed in 58% of falls (Table 4), including 79% of forward landings, 71% of sideways landings and 46% of backward landings. A combination of upper limb bracing and rotation during descent was observed in 39% of falls. Upper limb bracing was associated with an increased odds for any injury (1.22; 1.04–1.43), serious injury (1.82; 1.21–2.74) and head injury (1.36; 1.09–1.70).
Age, sex and injury risk
Age and sex were significant as covariates (P < 0.05) in several analyses (Tables 3 and 4). There were no differences between men and women in initial fall direction, landing direction or tendency to rotate during descent (P > 0.08 by Chi square), or in the portion of falls that involved upper limb bracing (57% versus 61%; P = 0.06). Furthermore, there was no difference between men and women in the portion of falls that caused injury (40% versus 36%; P = 0.08 by Chi square). However, women were more likely than men to experience serious injury (P = 0.0005) and head injury (P = 0.002) during falls. Women experienced 57% of all falls, 58% of all injuries, 72% of serious injuries and 64% of head injuries.
Age did not associate with tendency to rotate during descent (P = 0.16). However, those who used upper limb bracing tended to be younger (mean difference = 1.9 years; 95% CI = 1.2–2.5). When compared with falls that involved no injury, the mean age was older for those who fell and experienced serious injury (mean difference = 2.2 years; 95% CI = 0.6–3.8) or head injury (0.9; 0.04–1.8), but not any injury (0.6; –0.05 to 1.3).
Discussion
While any fall from standing height onto a rigid surface has the potential to cause catastrophic injury, our results agree with others [1, 3–5, 31] in showing that most falls by older adults in LTC did not cause injury, and only 4% caused injuries requiring treatment in hospitals. By linking documented injuries to video footage of real-life falls experienced by older adults in LTC, we provide objective evidence on how older adults use protective responses to avoid injury during falls.
We observed common patterns of rotation during falls that reduced risk for injury. The unintentional nature of imbalance episodes makes it difficult if not impossible to dictate the initial direction of a fall. However, our results show that older adults can and often do change the direction of a fall after fall initiation, to achieve a safer landing configuration. The patterns of rotation reflected the avoidance of forward landings, which created highest risk for head injury, and favoured backward landings, which were safest for all injury outcomes.
The high prevalence of rotation during descent is relevant to how we measure and classify falls. Previous studies have classified falls with a single fall direction [26, 27, 35], leaving ambiguity on whether the data reported reflect the initial fall direction or the landing configuration. Our results highlight the importance of acquiring details on both the initial fall direction and the landing configuration to provide a complete picture of the fall.
The observation that rotation was protective against injury has implications for exercise programmes for fall and injury prevention. The patterns of rotation we observed are similar to those used in the ‘ukemi’ falling technique in judo [24, 25]. The evidence of a natural tendency for older adults in LTC to rotate backward during falls lends support and external validity to exercise-based programmes that have emerged for training older adults in judo falling techniques [36, 37].
We did not find that upper limb bracing reduced the odds for injury during falls. Indeed, falls that involved bracing were more likely to result in any injury, serious injury and head injury. We see several possible explanations for these surprising results.
First, the diminished muscle strength, flexibility and reaction time of older adults in LTC may have rendering upper limb bracing ineffective in preventing impact to the head, which was the most common site of injury. Upper limb bracing is a physically challenging task even for healthy young adults [21, 38–40]. Successful bracing requires quick recognition of falling, and rapid movement of the upper limb(s) to contact the ground (or environment) in a biomechanically effective position for halting downward movement of the torso and head. Successful bracing also requires that the muscles are activated so the limb does not collapse at impact, but the joints are not so rigid that the impact results in fracture of the wrist or elbow [21, 41].
Second, upper limb bracing was especially common in the demanding scenario of forward landings (where it was observed in 79% of cases, compared with 49% in backward landings), where the upper limbs bore primary responsibility for preventing head impact, with little assistance from other contacting body parts. In backward or sideways landings, participants tended to land with the pelvis impacting the ground, absorbing energy near-simultaneously with the upper limbs.
Third, bracing may have emerged more often, or as a last resort, when alternative or complementary safe landing strategies were not used. Bracing was more common in falls that did not involve rotation during descent. Bracing was also less common in falls where the upper limbs were occupied in grasping weight-bearing objects such as tables or walkers, which was previously found to be protective [42, 43]. The tendency to rotate during falls, or retain grasp of weight-bearing objects, may arise in older adults who recognize, perhaps based on experience, a loss in their capacity to arrest falls through upper limb bracing.
While bracing did not prevent injuries, it persisted as a response that was observed in the majority of falls. This raises the question of what approaches can be taken to maintain and enhance upper limb bracing as protective for older adults. Exercise programmes should target the strength of triceps, pectoralis and deltoid muscles (e.g., through wall-press exercises) and the stability and ranges of motion of the shoulder, elbow and wrist [20, 44]. Perturbation training, which has focused to date on balance recovery [45, 46], should incorporate safe methods for practicing safe landing during a fall [47]. Technology may have a role in enhancing upper limb bracing of falls. While rigid wrist guards have a role in protecting the wrist, they do little to absorb the energy of the fall [48]. Potential solutions include soft bionics that span the shoulder, elbow and wrist, or forearm cuffs that cushion falls, perhaps through inflatable airbags.
Our study has important limitations. First, we analysed falls by older adults occurring in common areas (dining rooms, lounges, hallways) of two LTC homes, where there was a high prevalence among participants of physical and/or cognitive impairment. The falling patterns may be different in healthier populations of older adults, or for falls in different environments (e.g., outdoors, stairways, bedrooms or bathrooms). At the same time, we found that participants who fell and experienced injury were similar to those who fell without injury in demographic and health status, disease diagnoses and use of medications. The notable exception was cognitive impairment, which was more common in those who experienced non-injurious falls. This observation agrees with results from a previous study, which found that, while cognitively impaired residents of LTC fell more often, they had a lower risk for head impact during falls, probably due to differences in the circumstances of falls [49]. Second, we did not examine how injury risk during falls depended on tissue resistance to trauma (e.g. bone mass). Among the 39% of participants who provided consent to access medical records, we found few differences in the clinical status of those who fell and injured versus those who fell and did not injure. Third, we expect that fall incident reports underestimated the neural consequences of falls in LTC. Of the 35 serious head injuries, 5 were diagnosed TBIs or skull fractures, but no concussions were noted, which may reflect the challenges and need for improved guidelines in detecting and distinguishing the neural consequences of concussions from any baseline dementia in older adults in the LTC environment.
Conclusions
The odds for injury from falls in LTC were reduced by common patterns of rotation during descent, to convert forward falls to sideways landings, and sideways falls to backward landings. In contrast, upper limb bracing was associated with increased risk for injury. Rotation during descent may emerge as a compensatory mechanism following loss in the capacity of older adults to successfully stop falls through upper limb bracing. By providing evidence on the strategies used by older adults to avoid injury during falls, our results may inform innovation in fall injury prevention through exercise and protective clothing.
Supplementary Material
Acknowledgements
We thank the Good Samaritan Delta View Life Enrichment and Habilitation Centres, and the New Vista Society Centre, for their partnership in this research.
Contributor Information
Stephen N Robinovitch, Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, BC, Canada.
Aleksandra Dojnov, Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, BC, Canada; Institute of Biomedical Engineering, University of Toronto, Toronto, ON, Canada.
Vicki Komisar, Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, BC, Canada; School of Engineering, The University of British Columbia, Kelowna, BC, Canada.
Yijian Yang, Department of Sports Science and Physical Education, The Chinese University of Hong Kong, Hong Kong, China.
Nataliya Shishov, Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, BC, Canada.
Ying Yu, Department of Statistics and Actuarial Science, Simon Fraser University, Burnaby, BC, Canada.
Ian Bercovitz, Department of Statistics and Actuarial Science, Simon Fraser University, Burnaby, BC, Canada.
Michael D Cusimano, Division of Neurosurgery, Li Ka Shing Knowledge Institute, Injury Prevention Research Office, St. Michael’s Hospital – Unity Health, Toronto, ON, Canada.
Clemens Becker, Robert Bosch Hospital, Stuttgart, Germany; Unit of Digital Geriatric Medicine, University Hospital, Heidelberg, Germany.
Dawn C Mackey, Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, BC, Canada.
Helen Chong, Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, BC, Canada.
Declaration of Conflicts of Interest
None.
Declaration of Sources of Funding
This work was supported by operating grants from the Canadian Institutes of Health Research (CIHR; funding reference numbers AMG-100487, TIR-103945 and TEI-138295), and the AGE-WELL National Centre for Excellence (AW CRP 2015-WP5.2, AW CRP-2020-04). V.K. and Y.Ya. were supported by Michael Smith Foundation for Health Research (MSFHR) Postdoctoral Awards. V.K. was also supported by an AGE-WELL NCE Postdoctoral Award. N.S. was supported by a Simon Fraser University Graduate Dean’s Entrance Scholarship. None of the funding agencies influenced the experimental design; the collection, analysis or interpretation of data; the writing of the manuscript or the decision to submit the manuscript for publication.
References
- 1. Büchele G, Becker C, Cameron ID, König H-H, Robinovitch S, Rapp K. Predictors of serious consequences of falls in residential aged care: analysis of more than 70,000 falls from residents of Bavarian nursing homes. J Am Med Dir Assoc 2014; 15: 559–63. [DOI] [PubMed] [Google Scholar]
- 2. Rubenstein LZ, Josephson KR, Robbins AS. Falls in the nursing home. Ann Intern Med 1994; 121: 442–51. [DOI] [PubMed] [Google Scholar]
- 3. Krueger PD, Brazil K, Lohfeld LH. Risk factors for falls and injuries in a long-term care facility in Ontario. Can J Public Health 92: 117–20. 10.1007/BF03404944. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Nurmi I, Luthje P. Incidence and costs of falls and fall injuries among elderly in institutional care. Scand J Prim Health Care 2002; 20: 118–22. [PubMed] [Google Scholar]
- 5. Sadigh S, Reimers A, Andersson R, Laflamme L. Falls and fall-related injuries among the elderly: a survey of residential-care facilities in a Swedish municipality. J Community Health 2004; 29: 129–40. [DOI] [PubMed] [Google Scholar]
- 6. Robinovitch SN, Chiu J. Surface stiffness affects impact force during a fall on the outstretched hand. J Orthop Res 1998; 16: 309–13. [DOI] [PubMed] [Google Scholar]
- 7. Robinovitch SN, Hayes WC, McMahon TA. Prediction of femoral impact forces in falls on the hip. J Biomech Eng 1991; 113: 366–74. [DOI] [PubMed] [Google Scholar]
- 8. Cusimano MD, Saarela O, Hart K, Zhang S, McFaull SR. A population-based study of fall-related traumatic brain injury identified in older adults in hospital emergency departments. Neurosurg Focus 2020; 49: E20. 10.3171/2020.7.FOCUS20520. [DOI] [PubMed] [Google Scholar]
- 9. Fu TS, Jing R, Fu WW, Cusimano MD. Epidemiological trends of traumatic brain injury identified in the emergency department in a publicly-insured population, 2002-2010. PLoS One 2016; 11: e0145469. 10.1371/journal.pone.0145469. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Harvey LA, Close JC. Traumatic brain injury in older adults: characteristics, causes and consequences. Injury 2012; 43: 1821–6. [DOI] [PubMed] [Google Scholar]
- 11. Lecky FE, Otesile O, Marincowitz Cet al. The burden of traumatic brain injury from low-energy falls among patients from 18 countries in the CENTER-TBI Registry: a comparative cohort study. PLoS Med 2021; 18: e1003761. 10.1371/journal.pmed.1003761. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Courtney AC, Wachtel EF, Myers ER, Hayes WC. Age-related reductions in the strength of the femur tested in a fall-loading configuration. J Bone Joint Surg Am 1995; 77: 387–95. [DOI] [PubMed] [Google Scholar]
- 13. Cummings SR, Nevitt MC. A hypothesis: the causes of hip fractures. J Gerontol 1989; 44: M107–11. [DOI] [PubMed] [Google Scholar]
- 14. Jusufi A, Kawano DT, Libby T, Full RJ. Righting and turning in mid-air using appendage inertia: reptile tails, analytical models and bio-inspired robots. Bioinspir Biomim 2010; 5: 045001. 10.1088/1748-3182/5/4/045001. [DOI] [PubMed] [Google Scholar]
- 15. Mesquita AR, Pego JM, Summavielle T, Maciel P, Almeida OF, Sousa N. Neurodevelopment milestone abnormalities in rats exposed to stress in early life. Neuroscience 2007; 147: 1022–33. [DOI] [PubMed] [Google Scholar]
- 16. Sechzer JA, Folstein SE, Geiger EH, Mervis RF, Meehan SM. Development and maturation of postural reflexes in normal kittens. Exp Neurol 1984; 86: 493–505. [DOI] [PubMed] [Google Scholar]
- 17. Feldman F, Robinovitch SN. Reducing hip fracture risk during sideways falls: evidence in young adults of the protective effects of impact to the hands and stepping. J Biomech 2007; 40: 2612–8. [DOI] [PubMed] [Google Scholar]
- 18. Hsiao ET, Robinovitch SN. Common protective movements govern unexpected falls from standing height. J Biomech 1998; 31: 1–9. 10.1016/S0021-9290(97)00114-0. [DOI] [PubMed] [Google Scholar]
- 19. Moon Y, Sosnoff JJ. Safe landing strategies during a fall: systematic review and meta-analysis. Arch Phys Med Rehabil 2017; 98: 783–94. [DOI] [PubMed] [Google Scholar]
- 20. Arnold CM, Walker-Johnston J, Lanovaz JL, Lattimer LJ. Does fall arrest strategy training added to a fall prevention programme improve balance, strength, and agility in older women? A pilot study. Physiother Can 2017; 69: 323–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. DeGoede KM, Ashton-Miller JA. Biomechanical simulations of forward fall arrests: effects of upper extremity arrest strategy, gender and aging-related declines in muscle strength. J Biomech 2003; 36: 413–20. [DOI] [PubMed] [Google Scholar]
- 22. Han D, Adolph KE. The impact of errors in infant development: falling like a baby. Dev Sci 2021; 24: e13069. 10.1111/desc.13069. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Robinovitch SN, Inkster L, Maurer J, Warnick B. Strategies for avoiding hip impact during sideways falls. J Bone Miner Res 2003; 18: 1267–73. [DOI] [PubMed] [Google Scholar]
- 24. Koshida S, Ishii T, Matsuda T, Hashimoto T. Kinematics of judo breakfall for osoto-gari: considerations for head injury prevention. J Sports Sci 2017; 35: 1059–65. [DOI] [PubMed] [Google Scholar]
- 25. Zijden AM, Groen BE, Tanck E, Nienhuis B, Verdonschot N, Weerdesteyn V. Can martial arts techniques reduce fall severity? An in vivo study of femoral loading configurations in sideways falls. J Biomech 2012; 45: 1650–5. [DOI] [PubMed] [Google Scholar]
- 26. Greenspan SL, Myers ER, Maitland LAet al. Fall direction, bone mineral density, and function: risk factors for hip fracture in frail nursing home elderly. Am J Med 1998; 104: 539–45. [DOI] [PubMed] [Google Scholar]
- 27. Nevitt MC, Cummings RS, Study of Osteoporotic Fractures Research Group . Type of fall and risk of hip and wrist fractures: the study of osteoporotic fractures. JAGS 1993; 41: 1226–34. [DOI] [PubMed] [Google Scholar]
- 28. Parkkari J, Kannus P, Palvanen Met al. Majority of hip fractures occur as a result of a fall and impact on the greater trochanter of the femur: a prospective controlled hip fracture study with 206 consecutive patients. Calcif Tissue Int 1999; 65: 183–7. [DOI] [PubMed] [Google Scholar]
- 29. Palvanen M, Kannus P, Parkkari Jet al. The injury mechanisms of osteoporotic upper extremity fractures among older adults: a controlled study of 287 consecutive patients and their 108 controls. Osteoporos Int 2000; 11: 822–31. [DOI] [PubMed] [Google Scholar]
- 30. 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; 1: 40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Mackey DC, Lachance CC, Wang PTet al. The Flooring for Injury Prevention (FLIP) Study of compliant flooring for the prevention of fall-related injuries in long-term care: a randomized trial. PLoS Med 2019; 16: e1002843. 10.1371/journal.pmed.1002843. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Gryfe CI, Amies A, Ashley MJ. A longitudinal study of falls in an elderly population: 1. Incidence and morbidity. Age Ageing 1977; 6: 201–10. [DOI] [PubMed] [Google Scholar]
- 33. Stevens JA, Sogolow ED. Gender differences for non-fatal unintentional fall related injuries among older adults. Inj Prev 2005; 11: 115–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Tinetti ME, Doucette J, Claus E, Marottoli R. Risk factors for serious injury during falls by older persons in the community. J Am Geriatr Soc 1995; 43: 1214–21. [DOI] [PubMed] [Google Scholar]
- 35. Vellas BJ, Wayne SJ, Garry PJ, Baumgartner RN. A two-year longitudinal study of falls in 482 community-dwelling elderly adults. J Gerontol A Biol Sci Med Sci 1998; 53: M264–74. [DOI] [PubMed] [Google Scholar]
- 36. Groen BE, Smulders E, Kam D, Duysens J, Weerdesteyn V. Martial arts fall training to prevent hip fractures in the elderly. Osteoporos Int 2010; 21: 215–21. [DOI] [PubMed] [Google Scholar]
- 37. Kalina RM, Barczyński BJ, Jagiełło Wet al. Teaching of safe falling as most effective element of personal injury prevention in people regardless of gender, age and type of body build - the use of advanced information technologies to monitor the effects of education. Archiv Budo 2008; 4: 82–90. [Google Scholar]
- 38. Lattimer LJ, Lanovaz JL, Farthing JP, Madill S, Kim S, Arnold C. Upper limb and trunk muscle activation during an unexpected descent on the outstretched hands in young and older women. J Electromyogr Kinesiol 2016; 30: 231–7. [DOI] [PubMed] [Google Scholar]
- 39. Robinovitch SN, Normandin SC, Stotz P, Maurer JD. Time requirement for young and elderly women to move into a position for breaking a fall with outstretched hands. J Gerontol A Biol Sci Med Sci 2005; 60: 1553–7. [DOI] [PubMed] [Google Scholar]
- 40. Sran MM, Stotz PJ, Normandin SC, Robinovitch SN. Age differences in energy absorption in the upper extremity during a descent movement: implications for arresting a fall. J Gerontol A Biol Sci Med Sci 2010; 65: 312–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Chiu J, Robinovitch SN. Prediction of upper extremity impact forces during falls on the outstretched hand. J Biomech 1998; 31: 1169–76. [DOI] [PubMed] [Google Scholar]
- 42. Komisar V, Shishov N, Yang Y, Robinovitch SN. Effect of holding objects on the occurrence of head impact in falls by older adults: evidence from real-life falls in long-term care. J Gerontol A Biol Sci Med Sci 2021; 76: 1463–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Yang Y, Komisar V, Shishov Net al. The effect of fall biomechanics on risk for hip fracture in older adults: a cohort study of video-captured falls in long-term care. J Bone Miner Res 2020; 35: 1914–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Yang Y, Schooten KS, Komisar Vet al. Effects of the mobility-fit physical activity program on strength and mobility in older adults in assisted living: a feasibility study. Int J Environ Res Public Health 2022; 19: 5453. 10.3390/ijerph19095453. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Liu X, Bhatt T, Wang Y, Wang S, Lee A, Pai YC. The retention of fall-resisting behavior derived from treadmill slip-perturbation training in community-dwelling older adults. Geroscience 2020; 43: 913–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Mansfield A, Peters AL, Liu BA, Maki BE. A perturbation-based balance training program for older adults: study protocol for a randomised controlled trial. BMC Geriatr 2007; 31: 12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Moon Y, Bishnoi A, Sun R, Shin JC, Sosnoff JJ. Preliminary investigation of teaching older adults the tuck-and-roll strategy: can older adults learn to fall with reduced impact severity. J Biomech 2019; 83: 291–7. [DOI] [PubMed] [Google Scholar]
- 48. Hwang IK, Kim KJ. Shock-absorbing effects of various padding conditions in improving efficacy of wrist guards. J Sports Sci Med 2004; 3: 23–9. [PMC free article] [PubMed] [Google Scholar]
- 49. Yang Y, Mackey DC, Liu-Ambrose T, Leung PM, Feldman F, Robinovitch SN. Clinical risk factors for head impact during falls in older adults: a prospective cohort study in long-term care. J Head Trauma Rehabil 2017; 32: 168–77. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.