Abstract

Several healthy, community-dwelling elderly people who live independently experience falls. However, their falls are not always attributed to decreased physical function. We aimed to examine the cause of falls in active elderly people. They were divided into fallers (n = 49) and non-fallers (n = 170) based on their fall experience during the previous year. The subjects participated in the following tests: transferring velocity during the sit-to-stand, one-leg stand time, functional reach, 10-m maximal walking time, muscle strength (foot grip, ankle plantar flexion, hip flexion, knee extension strength), systemic reaction time, agility stepping, and maximal side step length after completing the ADL questionnaire, fall risk assessment sheet, fall risk profile from Demura et al, and life-space assessment sheet. No significant differences were observed in physical functions, which were the basis of fall prevention (muscle strength, balance, and gait) and fall avoidance (systemic reactions, agility stepping, and maximal side step length) between both groups. A significant difference was observed in the ADL score but not between both fall risk scores, and the Life-Space Assessment scores. Moreover, no significant differences were observed between the high risk ratio for falls and nursing care based on each questionnaire. In conclusion, the differences in physical function and fall risk factors between the fallers and non-fallers in active, independent elderly people living in the community were not observed. The cause of these falls (22.4%) was not attributed to physical function decreases or other fall risk factors. These may mainly depend on accidents resulting from high activity.

1. Introduction

At least 20%-35% of elderly people over 65 years of age who are community-dwelling experience falls each year ^{1, 2, 3, 4, 5, 6, 7}. In addition, almost three (2.7) times as many females experience falls ⁴. If an elderly person experienced a fall once, the risk of recurrent falls increases as well ¹⁰. Because falls that do not result in fractures or any injury may still induce severe psychological consequences that lead to an accelerated decline in functional capacities ⁶, primary prevention is very important.

Several risk factors such as disease, medicine, changes in physical functions with age, and environment can relate to falls as reported by Suzuki ³¹ and several other previous studies ^{1, 8, 9, 32, 33, 34, 35, 36}. Moreover, Tinetti et al. ¹ reported that the incidence of falls was approximately 27% when a community-dwelling elderly person had one or fewer risk factors, but it increased to 78% when there were four or more risk factors involved. It was believed that the presence of more risk factors led to higher fall incidence ^{1, 12, 13}. Therefore, we need to adequately evaluate and understand fall risk factors to manage fall prevention.

Santos et al. ²⁹ calculated a cut-off value for screening for fall risks based on physical functioning of the active elderly. In addition, he reported that although the specificity was high, the sensitivity was very low. Namely, even elderly people with a low fall risk had the possibility to be at a high risk for falls. The above results from because the elderly who experienced fall exist in spite of having high physical functions. Considering the above statements, falls in the elderly do not always have to be caused by decreased physical functioning. According to Speechley and Tinetti ²⁸, falls of the active elderly frequently occur during displacing activities, particularly in stair climbing. Moreover, they reported that active elderly people fall because they are active and frequently encounter fall triggers more than the inactive elderly.

Although falls induced by physical functioning decreases and accidental falls are at opposite extremes, it can be very difficult to distinguish between the two situations. Robinovitch et al. ¹⁴ observed the fall incidences and characteristics of the elderly, living in long-term care facilities using digital video cameras installed in common areas (dining room, lounges, and hallways) over 39 months. They confirmed 227 falls from 130 individuals and reported that 41% of these falls occurred during incorrect weight shifting, followed by a trip or a stumble (21%), loss of support (11%), and collapse (11%). A report by Robinovitch et al. ¹⁴, which evaluated the details from falls, including scenes, causes, and characteristics, may be a useful method to distinguish falls that were caused by physical functioning decreases or accidents. However, because the activity range of the community-dwelling elderly can be very widespread, we may not be able to use Robinovitch’s method ¹⁴. Physical functioning and fall risk factors of the elderly with/without fall experience will need to be examined comprehensively and in detail.

This study aimed to examine the cause of falls in active, community-dwelling elderly people who live independently.

2. Methods

Two hundred and nineteen healthy elderly females who were >60 years of age (age, 76.9 +/- 6.1 yr; height, 147.2 +/- 5.8 cm; body-mass, 49.7 +/- 7.3 kg) living in Sabae-City, Fukui, Japan participated in this study. Their suitability was assessed through medical checkups prior to the study commencement. Written informed consent was obtained from all participants after the purpose and protocol of the study were fully explained. The study protocol was approved by the Ethics Committee on Human Experimentation of the Faculty of Education, Kanazawa University, Japan (authorization number: 19-18).

The subjects first completed the ADL questionnaire of the Ministry of Education, Culture, Sports, Science and Technology (MEXT) in Japan, the fall risk assessment sheet of Tokyo Metropolitan Institute of Gerontology (TMIG), the fall risk profile of Demura et al. ^{16, 17}, and the life-space assessment sheet. Furthermore, they participated in the following tests: transferring velocity during sit-to-stand (STS), one-leg stand time, functional reach, 10 m maximal walking time, muscle strength (hand grip, foot grip, ankle plantar flexion, hip flexion, and knee extension strength), systemic reaction time, and agility stepping and maximal side step length.

ADL: ADL was evaluated by an ADL questionnaire made up by MEXT ¹⁵. This contained 12 items regarding daily living activities such as STS, gait, obstacle avoidance, balance, and putting on/taking off clothes; a total score of each item can screen the physical functioning of each subject. Subjects with a total score of <24 may have a difficult time in participating in all physical function tests ¹⁵.

Fall risk: Fall risk was evaluated using the fall risk profile proposed by Demura et al. ^{16, 17} and the TMIG fall risk assessment sheet ¹⁸. The former consisted of questions regarding previous fall experience and 50 other fall risk assessment items representing the following five risk factors: potential for falling, physical function, disease and physical symptoms, environment, and behavior and character. Subjects responded to all items answering “yes” or “no.” The fall risk according to each subject and each risk factor was evaluated with the answer number indicating the risk of falling. Namely, subjects were judged to have a high fall risk when the responses were more than one for “potential for falls,” 13 for “physical function,” six in “disease and physical symptoms,” and three for both “behavior and character” and “environment.” Meanwhile, the TMIG fall risk assessment sheet evaluated the fall risk of the elderly using 15 items regarding eight factors such as gait, balance, muscle strength, disease, medicine, environment, audiovisual function, and fear of falling. Subjects answered all items with “yes” or ”no,” and they were determined to have high fall risk when the number of high-risk responses for falling was five and over.

Life-Space Assessment: The activity level of each subject was evaluated by the Life-Space Assessment (LSA) ^{20, 21, 27}, which can evaluate the distance and frequency of going outdoors and the degree of independence of subjects. This questionnaire divided the subjects’ living space into five levels from the bedroom to outside of town, and quantified their active mass by moving, frequency, and degree of independence. Moving was scored by the following five levels: moving to rooms without bedroom (1 point), moving within the house premises (2 points), moving to a neighboring house (3 points), moving around town (4 points), and moving unlimited distances (5 points). Moreover, the frequency of going outdoors was scored by the following four levels: everyday (4 points), 4–6 times per week (3 points), 1–3 times per week (2 points), and <1 time per week (1 point). The degree of independence was scored by the following three levels: moving independently (2 points), moving with assistive device (1.5 point), and moving dependently (1 point). The LSA score was defined as the sum of multiplying the frequency and degree of independence scores of each living space, and their full marks were evaluated by 120 points. A LSA score under 56 points was judged to be low and intervention regarding living space would be required.

Transferring velocity during STS movement: Transferring velocity during STS movement was measured according to the method used by Yamada and Demura ³⁷ using FITR Odyne Premium (Fitronic s.r.o., Slovakia). Subjects were instructed to adopt the appropriate sitting posture during the measurement; they maintained both lower limbs (with bare feet) shoulder-width apart, with the trunk in neutral flexion, ankles at neutral flexion, and arms crossed over the chest. The movements in the STS test were performed as quickly as possible from a sitting position on the instructor’s signal. The chair seat was set at knee height for each subject. Data were uploaded to a personal computer every 0.01 s. Peak and mean velocities from start to finish were calculated from the data obtained from the transfer velocity of the center of gravity during the STS test. The measurements were conducted twice and a mean value of the two trials was used as representative value.

Maximal Isometric knee extension, hip flexion, ankle plantar flexion, and foot grip strength: Maximal isometric knee extension and hip flexion strengths were measured using a hand-held dynamometer (Anima Co., Ltd.: μ-tas F-1). Subjects were instructed to extend or flex with maximal effort from a sitting posture with the knee or hip joint at 90 degrees. Maximal isometric ankle plantar flexion strength was measured using an ankle plantar flexion muscle strength device (Takei Co., Ltd.: T.K.K. 3360). Subjects were instructed to flex the ankle joint with maximal effort from a sitting posture with a fully extended knee joint and 90 degree flexed ankle joint. Moreover, isometric foot grip strength was measured using a foot grip strength device (Takei Co., Ltd.: T.K.K. 3362). Subjects were instructed to grip their foot with maximal effort against a lever connected to a strain gauge. All strengths were measured twice and their mean values were used as representative values.

Ten meter walking speed: Walking speed was measured while walking 10 m with maximal effort on a flat floor without obstacles according to the method used by Bohannon ²². In addition, while taking the acceleration and deceleration of walking into consideration, the subject actually walked 16 m to get the 10-m measurement.

Functional reach: Functional reach was measured according to the method used by Demura and Yamada ²³, which was proposed as a modified version of Duncan et al. ²⁴. Namely, each subject stood in line with a yardstick, which was fixed at dominant acromion height at the wall, and adjusted to the middle fingertip of the dominant hand at the yardstick’s starting point. Furthermore, a tester measured the distance where the subject can maximally extend the fingertip while holding an upright posture.

One-leg stand time: One-leg stand time with vision was measured as outlined in the method proposed by MEXT ¹⁵. In brief, the maximum time that the subjects could stand on one leg while holding their hands on their waist was measured. In addition, the measurement ended when the lifted leg contacted the floor, the supporting leg moved, one or both hands left the waist, or after 120 s had elapsed.

Systemic reaction time: Systemic reaction time was measured using a systemic reaction time device (Takei Co., Ltd.: T.K.K. 5408). Subjects were instructed to lift their feet from the floor as fast as possible with arms crossed over the chest in sitting posture after shining a red photic stimulation device, which was set in front of subjects. Time from shining the phonic simulation to lifting both feet from the floor was measured. This measurement was conducted five times, and the mean value was used as representative value.

Agility stepping: Agility stepping was measured as outlined in the method proposed by Yamaji and Demura ²⁶. Subjects were instructed to step counterclockwise with the right foot, which started from the right direction. After returning to the original position, the subject stepped clockwise with the left foot for alteration, which started from the left direction and back to the original position. This measurement was conducted twice, and the mean value was used as representative value.

Maximal side step length: Maximal side-step length was measured with a tape measure, which was set on the floor in front of each subject. Subjects maximally extended their leg with arms crossed over the chest, and then drew and matched the first leg to the second one while maintaining body stability with either supporting leg. The distance between the start line and the end of the little toe of the second step was measured. The measurement was conducted for right and left hand side once, and the mean value of both sides was used as representative value.

An independent t-test was used to examine the mean difference in age, physical characteristics, ADL, fall risk and LSA scores, and physical function tests between the fallers and non-fallers. The effect size of the mean difference was calculated by Cohen’s d ²⁵. The difference of ratio was tested for the high risk ratio of falls and nursing care based on ADLs and fall risk scores between fallers and non-fallers. p values of <0.05 indicated statistical significance.

3. Results

Table 1 shows the mean age and physical characteristics of the fallers and non-fallers. Their fall experience rate was 22.4% (non-faller: 170, faller: 49). Moreover, no significant difference was observed in age and physical characteristics between both groups. Table 2 shows the results of the independent t-test for the mean values of physical functioning tests (physical functioning as the basis of fall prevention) between both groups. A significant difference was observed only in the ADL score, which was higher in non-fallers. The effect size of the mean difference was moderate (d = 0.40). Table 3 shows the results of the independent t-test for the mean values of physical functioning tests (fall avoidance ability) between both groups. No significant difference was observed between both groups in all items. Table 4 shows the results of the independent t-test for the mean value of the LSA score between both groups. The LSA of both groups was higher than 56 points and no significant difference was observed. Table 5 and Table 6 show the results of the difference in the high risk ratio of falls and nursing care based on the ADL and fall risk questionnaire between the fallers and non-fallers. No significant difference was observed for any high risk ratios between both groups.

4. Discussion

The living space mobility (LSA: frequency of going outdoors, living space, and using assistive devices) of the community-dwelling elderly population in this study was very high regardless of fall experience, and was the same in both elderly groups with/without fall experience. Peel et al. ¹⁹ examined the relationships between LSA and physical functions (ADL, IADL, Short Physical Performance Battery (SPPB: balance, gait, and STS), social demographic statistics (sex, ethnic, age, residential area, and revenue), Geriatric Depression Scale (GDS), and Mini Mental State Examination for elderly people ≥65 years. They reported that the LSA was significantly related to all parameters except for the residential one; above all, a high correlation was observed between IADL and SPPB. Moreover, Baker et al. ²¹ surveyed LSA, SPPB, ADL, IADL, the physical component score (SF-12: self-reported health, mental health), GDS, and comorbidity to examine the reliability and validity of LSA using 306 elderly people. They reported that the reliability of the LSA was high and the correlation with all physical functions (SPPB, ADL, IADL, and SF-12) was also high. Namely, LSA was observed to closely relate to the basic ADLs such as gait and STS in addition to the instrumental ADLs (housework such as shopping, laundry and cleaning, financial control, medication management, and going out and getting on/in vehicle), which were of a higher order than basic ADLs. LSA scores in the present fallers and non-fallers were 69.1 ± 21.5 and 74.9 ± 22.0, respectively, and higher compared with that of Peel et al. (64.1 ± 24.9) and Baker et al. (62.9 ± 24.7). From the above findings, the present elderly were judged to be active, and had a high degree of independence regardless of fall experience.

Physical functions except for ADL (gait, balance, STS, and muscle strength) were also the same as those in the fallers and non-fallers (Table 2). Although a significant difference was observed only in ADL and the difference for the fallers were higher than the non-fallers, the former score was the same as 24 points corresponding to the criterion of screening ¹⁵. In addition, the high risk ratio based on the above stated criterion (Table 6) was also the same between both fallers and non-fallers. To establish guidelines for fall prevention of the elderly, the American Geriatric Society ¹¹ determined fall risk factors based on 16 previous studies and ranked them based on the relative risk ratio of fall incidence as follows: muscle weakness (4.4), history of falls (3.0), gait deficit (2.9), balance deficit (2.9), use assistive device (2.6), visual deficit (2.5), arthritis (2.4), impaired ADL (2.3), depression (2.2), cognitive impairment (1.8) and aged >80 years (1.7), with the values in parentheses indicating the relative risk ratio. Most of all, muscle strength, gait, and balance were the strongest risk factors, which have the largest effect on fall incidence; therefore, these factors were selected as the basic physical functions for fall prevention. Considering these facts, the falls of the elderly with very high activity level were assumed to not to be caused by a physical function decrease.

Meanwhile, systemic reaction, agility stepping, and maximal side-step length were also examined in addition to the stated physical functions in this study. These tests could evaluate the ability to avoid falls when an elderly person encountered a fall trigger. For example, when the person was going to stumble, the following abilities allow the person to avoid a fall: a rapid reaction (systemic reaction), quick step and base of support extension (agility stepping), and support for the unstable center of gravity (maximal side step length). Therefore, having the ability to manage these movements required a higher order of physical functioning as compared with the above stated physical functions. However, the difference regarding fall experience was not observed as well as the above stated fundamental physical functions for fall prevention. Santos et al. ²⁹ measured fall experience and the Berg Balance Scale (BBS) for 91 physically active elderly people and 96 physically inactive elderly people to examine the validity of the BBS for fall screening. As a result, the cut-off value for screening the physically inactive elderly people with/without fall experience was 49 points, and its sensitivity and specificity were very high (sensitivity: 91%, specificity: 92%). Meanwhile, sensitivity was low (0 %–15%) in the physically active elderly people, and specificity ranged from 82% to 100%. That is to say, because fall risk due to physical function decreases the increases in inactive people, screening fall experience may be possible for that population. However, for physically active elderly people, it may be highly possible that even if they had a low fall risk, but cannot do that they have high fall risk. The measurement items and analysis methods differed from our study, but the study by Santos et al. suggested that falls of the active elderly were not caused by physical function decreases. Because the present fallers had the sufficient fundamental physical functioning for fall prevention and fall avoidance, we believed that the number of the elderly who fell because of a physical function decrease was fairly small.

However, as stated in the introduction, several risk factors such as disease, medicine, and changes in physical functioning with aging and environment all relate to falls ^{1, 31, 32, 33, 34, 35, 36}. Moreover, more risk factors of the elderly leads to more falls ^{1, 12, 13}. The fall risk factors that the community-dwelling elderly in this study had were evaluated considering the above statements. However, no significant difference was observed between fallers and non-fallers in all fall risk factors (Table 5). From the present results, it is difficult to clarify the fall causes that were related to the occurrence of falls in the elderly. Speechley and Tinetti ²⁸ measured the fall history, the environment of fall causes, the degree of injury after a fall, the cognitive function, depression, the physical function (gait, balance), medical history, and physical complaints and disorders to classify the types of falls. They reported that 17% of active elderly people experienced a fall during the 1-year follow-up and several of their falls occurred during displacing ability, particularly on the stairs. Moreover, they reported that the active elderly frequently encountered fall triggers because of their higher level of activity. In addition, Freiberger and Menz ³⁰ reviewed 12 months of data to clarify the fall scenes, time, causes, injuries, and medical treatment associated with falls in 293 physically active elderly people. During the study period, there were 322 falls; the typical fall occurred outside home during leisure activities, at midday or in the afternoon. in particular, these falls occurred because the active elderly encountered fall triggers more often because of their higher activity. This is the antithesis of the fall caused by a physical function decrease. We inferred that several of the present fallers encountered fall triggers as well. Speechley and Tinetti ²⁸ also reported that 22% of falls by the active lderly resulted in a serious injury, but this occurred in only 6% of falls by the frail elderly, although 17% of the active elderly and 52% of the frail elderly were in a 1-year of follow-up. More active elderly people may experience serious injury with a fall, which could be a trigger for post-fall syndrome requiring primary nursing care. Although multifactorial intervention, such as an improvement of physical function, would be useful for fall prevention, most interventions depend on physical function decreases. Different preventive measures may be required for the active elderly. Therefore, further evaluation of the details of fall occurrence mechanism of the active elderly are needed.

5. Conclusion

There were no differences in the physical functions and fall risk factors (disease, medicine, and environment) of the active, independent, community-dwelling elderly individuals, irrespective of whether they had a fall experience in the previous one year. The falls in the active, community-dwelling elderly women were not attributed to the level of physical functions or other fall risk factors. In fact, they may be attributable to the higher number of fall triggers that they encounter because of their active lifestyle, which makes them more vulnerable to accidental falls.

Acknowledgements

The authors wish to thank the staff of the Division of Longevity and Welfare, Department of Health and Welfare, Sabae City Hall and participants of the lectures on nursing prevention.

References