Risk factors of repeated falls in the community dwelling old people

Article information

J Exerc Rehabil Vol. 15, No. 2, 275-281, April, 2019
Publication date (electronic) : 2019 April 26
doi : https://doi.org/10.12965/jer.1938086.043
1Department of Nursing, Chungbuk National University, Cheongju, Korea
2Department of Physical Therapy, Sahmyook University, Seoul, Korea
3College of Nursing, Institute of Health Science, Gyeongsang National University, Jinju, Korea
*Corresponding author: Mi Yang Jeon, https://orcid.org/0000-0002-5058-9912, College of Nursing, Gyeongsang National University, 63 Gajwa-gil 29beon-gil, Jinju 52828, Korea, E-mail: myjeon68@gnu.ac.kr
Received 2019 February 20; Accepted 2019 March 23.

Abstract

This study aimed to provide evidence for the development of an algorithm to identify older adults with a high risk for repeated falls, along with strategies to prevent repeated falls, by analyzing the known physical, psychological, and environmental factors related to falls in older adults. One hundred fifty-seven community-dwelling older adults aged 65 years or older who experienced a fall within the past year were enrolled in this study. Participants’ physical, psychological, environmental, and fall prevention-related characteristics were surveyed using structured questionnaires to identify the risk factors for repeated falls. The use of antidepressants, depression score, and compliance with fall prevention behaviors were found to differ significantly between the two groups, and the use of antidepressants and depression were found to be significant predictors of repeated falls. Depression should be considered as a major variable when developing an algorithm to identify the risk of repeated falls among older adults living at home. Also, the practice of fall prevention behaviors was higher in the repeated-falls group, likely due to that group’s efforts to prevent additional falls.

INTRODUCTION

Falls refer to accidents wherein people fall to the floor or ground due to an unintentional postural change. With decreasing muscle mass due to aging, older adults experience a deterioration of their overall motor capacity, including muscle strength, power, flexibility, agility, and endurance (Spirduso et al., 2005). Such decline of physical functions increases the risk for falls among older adults.

One study reported that 17.2% of older adults above the age of 65 years in Korea have experienced a fall injury in the past year (Ministry of Health, Welfare and Family Affairs, 2009). In particular, about 20% of older adults aged 70 years or older and 35% of older adults aged 75 years or older experience fall accidents (Ministry of Health, Welfare and Family Affairs, 2009), and 62% of trauma inpatients aged 65 years or older and 81% of trauma inpatients aged 85 years or older are admitted to the hospital for a fall injury (Korea Centers for Disease Control and Prevention, 2005). In the United States, over 30% of older adults aged 65 years or older experience a fall. About 40% of all deaths caused by an accident involve a fall (World Health Organization, 2007), and falls account for 85% of hospital admissions and 40% of nursing home admissions among older adults (Donaldson et al., 2005). About 50% of older adults who experience a fall face various other problems; about 10%–15% of them require medical treatment for severe injury, 0.2%–1% experience femoral fracture, 5% experience nonfemoral fracture, and another 5% encounter restrictions in activity due to soft tissue damage (Sattin, 1992; Sjögren et al., 1991).

As stated above, falls have serious health repercussions for older adults, including morbidity, mortality, immobility, hospitalization, and early admission to a long-term care facility (Bulat et al., 2008; Rubenstein and Josephson, 2006). Furthermore, older adults who experience a fall not only suffer from pain and fracture but also develop a fear of falling, which limits their physical activity, thereby deteriorating their physical capacity and ultimately increasing their vulnerability to falls (Jeon et al., 2017).

Korean studies report that about 44.2%–53.8% of those who experienced a fall were involved in at least two fall incidents (Jeon et al., 2017; Park et al., 2008; Yoo, 2011). In Canada, 33% of community-dwelling older adults experience a fall, and 42.4% of them experience repeated falls (Fletcher et al., 2009). Repeated falls not only lead to a more dependent lifestyle in older adults but also undermine their quality of life and increases society’s healthcare costs, calling for strategies to prevent repeated falls among this population.

Factors related to repeated falls in older adults should be identified to develop preventive strategies or programs. However, Korean studies on the topic are limited to those that examined the physical factors related to repeated falls, including studies that compared the physical fitness of older community-dwelling women who have experienced a fall and those who experience repeated falls (Jeon et al., 2017; Park et al., 2008) and a study that analyzed multilateral risk factors for repeated falls using a functional state assessment scale (Fletcher et al., 2009). However, falls are a complex health problem that should be examined in relation to physical, psychological, and environmental factors.

Therefore, this study aimed to provide evidence for developing an algorithm for identifying older adults with a high risk for repeated falls and strategies to prevent repeated falls by analyzing the known physical, psychological, and environmental factors related to falls in older adults.

MATERIALS AND METHODS

Research design

This is a descriptive survey study analyzing the risk factors of repeated falls in community-dwelling older adults (approval number: CIRB-2012-Q02).

Participants

One hundred fifty-seven community-dwelling older adults aged 65 years or older who experienced a fall within the past year were enrolled in this study. The inclusion criteria were older adults without disorientation, capable of understanding and answering the given questions. The required sample size for logistic regression analysis was 140, as 14 independent variables were chosen and 10 participants are needed per predictor. Considering a 20% dropout rate, 164 participants were recruited, but after excluding eight participants for omitting responses, 157 participants were included in the final analysis.

Method

Participants’ physical (existing disease, drugs currently being taken, visual impairment, hearing impairment, dizziness, ambulatory discomfort, physical activity), psychological (fear of fall, depression), environmental (cohabiting family, type of residence, home environment safety), and fall prevention-related characteristics (knowledge of falls, compliance with fall prevention behaviors) were surveyed using structured questionnaires.

Measurement of physical activity

The Korean Version of Physical Activity Scale for the Elderly (K-PASE), developed through a process of translation and back translation and modifications of the original scale developed by Washburn et al. (Choe et al., 2010) was used. Each item of this 10-item scale is weighted according to the type of physical activity (leisure activity, household activity, work-related activity) and activity frequency to analyze the risk. Items 1–6 (weekly and daily assessments of leisure activities) comprise two subitems to measure the content and duration of the activity. Items 7–9 assess household activities, and only item 9 comprises four subitems. Item 10 assesses work-related activity, and those who answer “yes” are instructed to answer the five additional items. The degree of physical activity was computed based on the PASE scoring method suggested by Washburn et al. (Choe et al., 2010). In other words, leisure activities were computed by converting the activity frequencies using the “time per day” conversion table, and household activities were computed in percentages based on the degree of participation in the past week. Contribution to the total physical activity score was calculated by multiplying the activity frequency by the PASE weighted value. The total score ranges from 0–360, with a higher score indicating more physical activity. The test-retest reliability at the time of the development was 0.75.

Measurement of fear of falling

Fear of falling was measured by the question “Are you afraid of falling?” which was rated on a 4-point scale, wherein 1 indicates “not at all,” 2 indicates “slightly,” 3 indicates “very,” and 4 indicates “very much.” A higher score indicates greater fear of falling.

Measurement of depression

Depression was measured using the Korean short form of Geriatric Depression Scale, developed by Song (Kee, 1996) by translating and standardizing the Geriatric Depression Scale: Short form developed by Sheikh and Yesavage (Kee, 1996). Each item of this 15-item scale is answered with a yes or no. “Yes” and “no” to a positive item was given a score of 0 and 1, respectively, and “yes” and “no” to a negative item was given a score of 1 and 0, respectively. Score 0–4 indicates normal, 5–9 indicates mild depression, and 10–15 indicates severe depression, and a higher score indicates more severe depression. The Cronbach α at the time of development was 0.89 and that in this study was 0.79.

Measurement of home environment safety

Home environment safety was measured using the Korean version (developed by Gu et al., 2006) (Hyeon et al., 2010) of the checklist for risk of fall at home in the Merck manual of geriatrics developed by Beers et al. (Hyeon et al., 2010). This scale comprises 25 items, including living room, furniture, lighting, bathroom, kitchen, stairs, and mobility, and each item is answered with a yes (1 point) or no (0 points). The total score ranges from 0–25, and a higher score indicates a safer home environment. The Cronbach α of the scale in this study was 0.77.

Measurement of knowledge of falls

The scale developed by Kim to measure nurses’ knowledge of falls was modified by Kim to assess older adult inpatients (Hwang and Shin, 2013), and this scale was again modified by Hwang (Hwang and Shin, 2013) to assess community-residing older adults, which was used for this study. The scale comprises 18 items pertaining to the definition of fall, physical risk factors, and drugs that increase the risk for falls. Each item is answered with a “yes,” “no,” or “I don’t know.” Correct answers were given 1 point, while wrong answers and “I don’t know” were given 0 point. The total score ranges from 0–18, and a higher score indicates a greater knowledge about falls. The Kuder–Richardson Formula 20 (K-R 20) in Hwang and Shin (2013) was 0.86, and that in this study was 0.76. Measurement of compliance with fall prevention behavior

Fall prevention behaviors were measured using the scale developed by Gu et al. (2006) This scale comprises 14 items, including wearing safe shoes, practicing safe walking habits, limiting alcohol intake, and taking drugs. Each item is answered with a “yes” or “no,” which is given a score of 2 and 1, respectively. The total score ranges from 14–28, with a higher score indicating higher compliance with fall prevention behaviors. The reliability (Cronbach α) of the scale in the study of Hyeon et al. (2010) was 0.78 and that in this study was 0.73.

Data analysis

Data were analyzed using IBM SPSS Statistics ver. 21.0 (IBM Co., Armonk, NY, USA). The general, physical, psychological, and environmental characteristics of the single-fall and repeated-fall groups were presented as mean and standard deviation as well as frequencies and percentages. The differences between the two groups were analyzed using the χ2-test, Fisher exact test, and independent t-test. The risk factors for repeated falls were analyzed by logistic regression, and the goodness of fit of the model was tested using the Hosmer and Lemeshow test. The reliability of each scale was measured by KR-20 or Cronbach alpha, and statistical significance was set at P<0.05.

RESULTS

Comparison of the demographic characteristics of the single-fall group and repeated-falls group

A total of 157 participants were included in this study. One hundred twelve (71.3%) were assigned to the single-fall group, while 45 (28.7%) were assigned to the repeated-falls group. The mean ages were 78.19 years and 77.87 years, respectively, showing no significant difference (t=0.27, P=0.766). The percentage of women in each group was 89.3% and 84.4%, respectively, also showing no significant difference (χ2=0.71, P=0.423). Most of the participants were not educated (χ2=1.02, P=0.832), widowed (χ2=2.61, P=0.451), and had poor economic status (χ2=0.83, P=0.843), and there were no significant differences between the two groups (Table 1).

Comparison of general characteristics (n=157)

Comparison of the physical characteristics of both groups

Regarding disease prevention in the single-fall group and repeated-falls group, the prevalence of hypertension was 73.9% and 68.2% (χ2=0.51, P=0.551), respectively, and the prevalence of diabetes mellitus was 37.8% and 28.9% (χ2=1.12, P=0.356), respectively. The prevalence of stroke was 11.8% and 8.9% (χ2= 0.28, P=0.779), respectively, and the prevalence of arthritis was 77.7% and 77.8% (χ2=0.00, P>0.999), respectively. The prevalence of urinary incontinence was 4.5% and 13.3% (χ2=3.74, P= 0.080), respectively, showing no significant differences between the two groups. Only one person in the repeated-falls group was diagnosed with alcoholism (Table 2). Regarding the drugs being taken, the use of antidepressants significantly differed between the two groups, with 0.9% in the single-fall group and 8.9% in the repeated-falls group (χ2=6.66, P=0.024); there were no significant differences in the use of diuretics (χ2=0.61, P=0.557), cardiotonics (χ2=0.34, P=0.730), sedatives (χ2=1.39, P=0.355), sleeping pills (χ2=0.03, P>0.999), use of sedatives (χ2=0.06, P>0.999), arthritis medication (χ2=2.46, P=0.140), and diabetes medication (χ2=0.64, P=0.511). There were also no significant differences in the prevalence of visual impairment (χ2=0.06, P=0.839), hearing impairment (χ2=3.61, P=0.076), dizziness (χ2=2.28, P=0.156), and physical activities (t=0.14, P=0.888) (Table 2).

Comparison of physical, psychological, ecological and fall related characteristics (n=157)

Comparison of the psychological characteristics of both groups

The two groups significantly differed in their depression scores (t=−3.25, P=0.001) but not in their fear of falling (t=−1.40, P= 0.165) (Table 3).

Factors influencing the repeated falls

Comparison of the environmental characteristics of both groups

Regarding cohabitation, most participants lived alone, with no significant difference between the two groups (χ2=0.81, P= 0.985). Home environment safety (t=−0.54, P=0.592) was also not significantly different between the two groups (Table 2).

Comparison of the fall prevention-related characteristics of both groups

The two groups did not significantly differ in the knowledge of falls (t=0.38, P=0.707) but did significantly differ in compliance with fall prevention behaviors (t=−2.85, P=0.005) (Table 2).

Actors affecting falls in both groups

To identify the risk factors for repeated falls, the use of antidepressants, depression score, and compliance with fall prevention behaviors, which were found to differ significantly between the two groups, were analyzed by logistic regression. Older adults taking antidepressants were 5.40 times more likely to experience repeated falls, and those with a high depression score were 1.17 times more likely to experience repeated falls. The model was found to be significant in the Nagelkerke test (χ2=22.25, P< 0.001) with an explanatory power of 11.7% (Table 3).

DISCUSSION

Older adults are at a higher risk for falls due to their reduced sensory functions, musculoskeletal problems, such as reduced lower limb muscle strength and articular contracture, and deteriorated cognitive ability (McGibbon et al., 2001). Those who experience a fall limit their activities due to a fear of falling again (Jeon et al., 2017). Falling is a safety accident in which various physical, psychological, and environmental factors are at play; thus, a comprehensive assessment of these factors is needed to identify the risk factors for falling. Hence, this study divided community- dwelling older adults into those who have only fallen once and those who have experience repeated falls in order to compare their physical characteristics (underlying disease, drugs they take, visual impairment, hearing impairment, dizziness, ambulatory discomfort, physical activities), psychological characteristics (fear of falling, depression), environmental characteristics (cohabiting family member, form of residence, home environment safety), and fall prevention-related characteristics (knowledge of falls, compliance with fall prevention behavior).

In our study, there were no significant differences in the demographic characteristics (age, sex, marital status, education level, economic status) between the single-fall and repeated-falls groups. These results are in line with the previous findings (Jeon et al., 2017; Park et al., 2008; Yoo, 2011). However, according to Woolcott et al. (2009), the risk for falls increases with advancing age. This difference may be attributed to the fact that the mean age of participants in our study and previous studies (Jeon et al., 2017; Park et al., 2008; Yoo, 2011) was more than 75 years. Although a previous study (Yoo, 2011) reported that older adults living alone show an increased incidence of falls, there was no significant difference in the relation to cohabitation in our study. This may be because a high percentage of participants lived alone (more than 50%) in both the single-fall and repeated-falls groups in our study. Based on these results, older adults aged 75 years or older or those living alone should be weighted additionally when developing an algorithm to identify those with a high risk for repeated falls.

In our study, there was no significant difference in the prevalence of comorbidities between the single-fall group and the repeated-falls group but the two groups significantly differed in their use of antidepressants. This result is in keeping with a previous finding that drug administration is associated with falls (Chen et al., 2014; Woolcott et al., 2009). A meta-analysis of antipsychotic drugs that increase the risk for falls (Woolcott et al., 2009) reported that antipsychotics, antidepressants, and benzodiazepines increase the risk for falls (Chen et al., 2014). In our study, older adults taking antidepressants were 5.40 times more likely to experience repeated falls (95% confidence interval [CI], 1.19–24.60). Antidepressants may increase the risk for falls because the elevated serotonin concentration in the brain not only decreases depression but also induces sedation. Furthermore, antidepressants may have elevated the risk for repeated falls due to their side effects, including extrapyramidal symptoms, such as tremor and ataxia, postural hypotension, anxiety, enuresis, and sleep disturbance (Chen et al., 2014).

In our study, the single-fall group and repeated-falls group did not show a significant difference in their fear of falling score but did have a significant difference in their depression score. This result is consistent with a previous finding that single-fall and repeated-falls groups have significantly different depression scores. In addition, we found that a higher depression score is associated with 1.17 time higher risk for repeated falls (95% CI, 1.07–1.27). Based on these results, depression and the use of antidepressants should be considered as major variables when developing an algorithm to identify older adults with a high risk for repeated falls.

In our study, the practice of fall prevention behaviors was higher in the repeated-falls group than in the single-fall group. This might be a result of the repeated-falls group’s effort to prevent additional falls after having experienced several falls.

This study assessed the physical (health status, physical activity, compliance with fall prevention behaviors), psychological (fear of falling, depression), and environmental factors (cohabiting family members, type of residence, and home environment safety) of older adults who had experienced a fall once and those who had experienced multiple falls in order to provide data for developing a program to prevent repeated falls in community-dwelling older adults. The use of antidepressants and depression were found to be significant predictors of repeated falls. Based on these results, depression should be considered as a major variable when developing an algorithm for identifying the risks for repeated falls among older adults living at home.

Notes

CONFLICT OF INTEREST

No potential conflict of interest relevant to this article was reported.

ACKNOWLEDGMENTS

This work was supported by the intramural research grant of Chungbuk National University in 2015.

References

Bulat, T., Castle, S.C., Rutledge, M., & Quigley, P. Clinical practice algorithms: medication management to reduce fall risk in the elderly—part 2, summary algorithm. J Am Acad Nurse Pract, (2008). 20, 1–4.
Chen, Y., Zhu, L.L., & Zhou, Q. Effects of drug pharmacokinetic/pharmacodynamic properties, characteristics of medication use, and relevant pharmacological interventions on fall risk in elderly patients. Ther Clin Risk Manag, (2014). 10, 437–448.
Choe, M.A., Kim, J., Jeon, M.Y., & Chae, Y.R. Evaluation of the Korean version of physical activity scale for the elderly (K-PASE). Korean J Women Health Nurs, (2010). 16, 47–59.
Donaldson, M.G., Khan, K.M., Davis, J.C., Salter, A.E., Buchanan, J., McKnight, D., Janssen, P.A., Bell, M., & McKay, H.A. Emergency department fall-related presentations do not trigger fall risk assessment: a gap in care of high-risk outpatient fallers. Arch Gerontol Geriatr, (2005). 41, 311–317.
Fletcher, P.C., Berg, K., Dalby, D.M., & Hirdes, J.P. Risk factors for falling among community-based seniors. J Patient Saf, (2009). 5, 61–66.
Gu, M.O., Jeon, M.Y., & Eun, Y. The development & effect of a tailored fall prevention exercise for older adults. Taehan Kanho Hakhoe Chi, (2006). 36, 341–352.
Hwang, S.Y., & Shin, S.J. Predictors of fall prevention behaviors in elderly inpatients. Korean J Health Promot, (2013). 13, 76–85.
Hyeon, I.S., Park, M., Park, K.M., & Kim, C.N. The effects of a fall prevention program on the low-income elderly at risk of falls. J Korean Acad Community Health Nurs, (2010). 21, 200–209.
Jeon, M., Gu, M.O., & Yim, J. Comparison of walking, muscle strength, balance, and fear of falling between repeated fall group, one-time fall group, and nonfall group of the elderly receiving home care service. Asian Nurs Res (Korean Soc Nurs Sci), (2017). 11, 290–296.
Kee, B.S. A preliminary study for the standardization of geriatric depression scale short form-Korea version. J Korean Neuropsychiar Assoc, (1996). 35, 298–307.
Korea Centers for Disease Control and Prevention. Epidemiologic characteristics of injury inpatients in Korea: The results of (Korea hospital discharge survey) 2004–2005. Cheongju (Korea): Korea Centers for Disease Control and Prevention; (2005.
McGibbon, C.A., Krebs, D.E., & Puniello, M.S. Mechanical energy analysis identifies compensatory strategies in disabled elders’ gait. J Biomech, (2001). 34, 481–490.
Ministry of Health Welfare and Family Affairs. 2008 National Elderly Survey: The living condition and welfare needs for older adults. Seoul: Ministry of Health, Welfare and Family Affairs; (2009.
Park, H.S., Chang, R., & Park, K.Y. Comparative study on fall related characteristics between single and recurrent falls in community-dwelling older women. Korean J Adult Nurs, (2008). 20, 905–916.
Rubenstein, L.Z., & Josephson, K.R. Falls and their prevention in elderly people: what does the evidence show? Med Clin North Am, (2006). 90, 807–824.
Sattin, R.W. Falls among older persons: a public health perspective. Annu Rev Public Health, (1992). 13, 489–508.
Sjögren, H., & Björnstig, U. Trauma in the elderly: the impact on the health care system. Scand J Prim Health Care, (1991). 9, 203–207.
Spirduso, W.W., Francis, K., Eakin, T., & Stanford, C. Quantification of manual force control and tremor. J Mot Behav, (2005). 37, 197–210.
Woolcott, J.C., Richardson, K.J., Wiens, M.O., Patel, B., Marin, J., Khan, K.M., & Marra, C.A. Meta-analysis of the impact of 9 medication classes on falls in elderly persons. Arch Intern Med, (2009). 169, 1952–1960.
World Health Organization. WHO global report on falls prevention in older age. Geneva (Switzerland): World Health Organization; (2007.
Yoo, I.Y. Analysis of multi-variate recurrent fall risk factors in elderly people using Residential Assessment Instrument-Home Care: comparisons between single and recurrent fallers. J Korean Acad Nurs, (2011). 41, 119–128.

Article information Continued

Table 1

Comparison of general characteristics (n=157)

Characteristic One-time fall group (n=112) Repeated falls group (n=45) χ2 or t P-value
Age (yr) 78.19±5.66 77.87±7.11 0.27 0.766

Gender
 Male 12 (10.7) 7 (15.6) 0.71 0.423
 Female 100 (89.3) 38 (84.4)

Marital status
 Married 31 (27.7) 12 (26.7) 2.61 0.451
 Divorce 4 (3.6) 0 (0)
 Bereaved 75 (67.0) 33 (73.3)
 Single 2 (1.8) 0 (0)

Education
 Uneducated 74 (66.1) 30 (66.7) 1.02 0.832
 Elementary school 27 (24.1) 11 (24.4)
 Middle school 5 (4.5) 3 (6.7)
 Above high school 6 (5.4) 1 (2.2)

Living standard
 Very bad 16 (14.3) 6 (13.3) 0.83 0.843
 Bad 53 (47.3) 20 (44.4)
 Moderate 38 (33.9) 18 (40.0)
 Good 5 (4.5) 1 (2.2)

Values are presented as mean±standard deviation or number (%).

Table 2

Comparison of physical, psychological, ecological and fall related characteristics (n=157)

Characteristic One-time fall group (n=112) Repeated falls group (n=45) χ2 or t P-value
Physical factors
 Diagnosis
  Hypertension
   No 29 (26.1) 14 (31.8) 0.51 0.551
   Yes 82 (73.9) 30 (68.2)
  Diabetes
   No 69 (62.2) 32 (71.1) 1.12 0.356
   Yes 42 (37.8) 13 (28.9)
  Stoke
   No 97 (88.2) 41 (91.1) 0.28 0.779
   Yes 13 (11.8) 4 (8.9)
  Arthritis
   No 25 (22.3) 10 (22.2) 0.00 >0.999
   Yes 87 (77.7) 35 (77.8)
  Urinary, no 105 (95.5) 39 (86.7) 3.74 0.080
  Incontinence, yes 5 (4.5) 6 (13.3)
  Alcoholism
   No 110 (100) 44 (97.8) 2.46 0.290
   Yes 0 (0) 1 (2.2)
 Medication
  Antidiuretic
   No 100 (89.3) 42 (93.3) 0.61 0.557
   Yes 12 (10.7) 3 (6.7)
  Cardiotonic
   No 105 (93.8) 41 (91.1) 0.34 0.730
   Yes 7 (6.3) 4 (8.9)
  Sedatives
   No 103 (97.3) 42 (93.3) 1.39 0.355
   Yes 3 (2.7) 3 (6.7)
  Hypnotic
   No 109 (97.3) 44 (97.8) 0.03 >0.999
   Yes 3 (2.7) 1 (2.2)
  Analgesic
   No 98 (87.5) 40 (88.9) 0.06 >0.999
   Yes 14 (12.5) 5 (11.1)
  Antidepressant
   No 111 (99.1) 41 (91.1) 6.66 0.024
   Yes 1 (0.9) 4 (8.9)
  Arthritis drug
   No 77 (68.8) 25 (55.6) 2.46 0.140
   Yes 35 (31.3) 20 (44.4)
  Diabetes drug
   No 91 (81.3) 34 (75.6) 0.64 0.511
   Yes 21 (18.8) 11 (24.4)
 Visual disturbance
  No 27 (24.1) 10 (22.2) 0.06 0.839
  Yes 85 (75.9) 35 (77.8)
 Hearing defect
  No 46 (41.1) 26 (57.8) 3.61 0.076
  Yes 66 (58.9) 19 (42.2)
 Dizziness
  No 14 (12.5) 2 (4.4) 2.28 0.156
  Yes 98 (87.5) 43 (95.6)
 Physical activity 56.17±44.54 55.08±42.05 0.14 0.888

Psychological factors
 Fear of fall 2.86±0.75 3.04±0.80 −1.40 0.165
 Depression 7.00±3.50 9.00±3.44 −3.25 0.001

Environmental factors
 Living
  Alone 63 (56.3) 25 (55.6) 0.81 0.985
  With spouse 29 (25.9) 11 (24.4)
  With offspring 16 (14.3) 6 (13.3)
  With family 1 (0.9) 1 (2.2)
  Other 3 (2.7) 2 (4.4)
 House
  Apartment 33 (29.5) 10 (22.2) 5.64 0.139
  Villa, multi-family 9 (8.0) 4 (8.9)
  Single house 70 (62.5) 29 (64.4)
  Other 0 (0) 2 (4.4)
 Safety of home environment 15.66±5.56 16.11±4.38 −0.54 0.592

Fall prevention factors
 Fall prevention activity compliance 11.46±2.81 12.49±1.63 −2.85 0.005
 Knowledge of fall 8.99±3.10 8.71±4.57 0.38 0.707

Values are presented as number (%) or mean±standard deviation.

Table 3

Factors influencing the repeated falls

Variable B SE Wald P-value Odds ratio 95% Cl
Intercept −5.36 1.24 18.70 <0.001
Antidepressant 1.69 0.77 4.76 0.029 5.40 1.187–24.601
Fall prevention activity compliance 0.11 0.07 2.52 0.112 1.12 0.974–1.283
Depression 0.15 0.04 13.15 <0.001 1.17 1.073–1.267

SE, standard error; CI, confidence interval.

χ2=22.25, P<0.001, Nagelkerke R 2=0.117.