The prevalence and risk factors of falling and fear of falling among lower extremity amputees

1031

The Prevalence and Risk Factors of Falling and Fear of Falling Among Lower Extremity Amputees William C. Miller, PhD, MSc, OT, Mark Speechley, PhD, Barry Deathe, MD ABSTRACT. Miller WC, Speechley M, Deathe B. The prevalence of risk factors of falling and fear of falling among lower extremity amputees. Arch Phys Med Rehabil 2001;82: 1031-7. Objectives: To estimate the falling experience and fear of falling status and to describe characteristics associated with falling and fear of falling. Design: Population-based consecutive sample survey and chart review. Setting: Two Canadian, regional, university-affiliated outpatient amputee clinics. Participants: The sample (n ⫽ 435; mean age, 62 ⫾ 15.7yr) of community-living participants was mostly male (71%), had unilateral (below knee 73%; above knee 27%) amputations primarily for vascular (53% vs 47% nonvascular) reasons. Intervention: Review patient charts and survey questionnaires to determine sociodemographic information (eg, social support), information about the amputation (eg, cause, level, problems), physical health and function (eg, pain, limitations, comorbidity), and psychologic state (depression, adaptation). Main Outcome Measures: Occurrence of a fall in the past 12 months and presence of a fear of falling. Results: Exactly 52.4% subjects reported falling in the past year, whereas 49.2% reported a fear of falling. Logistic regression analyses revealed falling was related to having an above knee amputation (odds ratio [OR] ⫽ 2.78; 95% confidence interval [CI] ⫽ 1.71– 4.51), back (OR ⫽ 1.96; 95% CI ⫽ 1.08 –3.54) and joint (OR ⫽ 1.67; 95% CI ⫽ 1.01–2.74) pain, and multiple stump and prosthesis problems (OR ⫽ 3.09; 95% CI ⫽ 1.58 – 6.04). Having had the amputation ⱖ 4 years in the past was protective (OR ⫽ .53; 95% CI ⫽ .29 –.89). Factors related to an increase risk of fear of falling included having to concentrate on each step while walking (OR ⫽ 4.06; 95% CI ⫽ 2.46 – 6.71) and having a fall in the past 12 months (OR ⫽ 1.62; 95% CI ⫽ 1.04 –2.54), whereas being male (OR ⫽ 0.35; 95% CI ⫽ .21–.57) and having good to excellent perceived health (OR ⫽ .35; 95% CI ⫽ .21–.58) were protective. Conclusions: Falling and fear of falling are pervasive among amputees. Comprehensive and ongoing intervention and education should be considered. Research is required to assess the consequences of falling and fear of falling. Key Words: Accidental falls; Amputees; Rehabilitation. © 2001 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation

From the Faculty of Medicine, School of Rehabilitation Sciences, University of British Columbia, Vancouver, BC (Miller); and Departments of Epidemiology & Biostatistics (Speechley) and Physical Medicine and Rehabilitation (Deathe), University of Western Ontario, London, Ont, Canada. Accepted in revised form September 28, 2000. No commercial party having a direct finincial interest in the results of the reserch supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to William Miller, PhD, Faculty of Medicine, School of Rehabilitation Sciences, T325-2211 Wesbrook Mall, University of British Columbia, Vancouver, BC V6T 2B5 Canada, e-mail: [email protected]. 0003-9993/01/8208-6196$35.00/0 doi:10.1053/apmr.2001.24295

ALLS AND FEAR OF FALLING are significant health F problems that are of interest to health professionals and researchers because they may indicate a decline in function that is potentially modifiable.1 The investigation of falls has received considerable attention by researchers. Early studies determined the incidence and risk factors for fall and fall-related injury among the community- and institutional-living elderly. More recently, investigators have begun to examine the consequences of falls and fear of falling on mobility related and social activity.1-4 To date, the elderly population has been the primary focus of fall and fear of falling research, yet this area potentially has interesting implications for the rehabilitation and postrehabilitation care of the prosthetic amputee population. Although the amputee population would appear to be at obvious risk for falls given their altered balance, strength, and gait pattern, limited information regarding falls and fear of falling was found in a recent literature review. Studies indicate that the annual incidence rate of falls among elderly (ⱖ65yr) living in the community is approximately 30%5-12 and more than 50% among individuals living in institutions.13,14 Falls are associated with functional limitation15 and are likely indicative of declines in independence, decreased mobility, and self-imposed restrictions in activity.1,16 Aside from the potential for physical injury associated with falling, psychologic sequelae also exist, of which fear of falling is the most common. Although physical injury after a fall is estimated to be 5%,12,13 fear of falling among the elderly ranges from 20% to 46% among nonfallers and 40% to 73% among individuals who have fallen recently.9,17,18 The outcome of the reported fear is that individuals tend to avoid participation in activity.12,17,18 In addition, the psychologic consequence of falling influences performance of activities and self-imposed restriction can lead to deterioration in balance,17 muscle endurance, strength, flexibility, and coordination.19 Therefore, a cycle between dehabilitation and psychologic sequelae appears to exist. To date little effort has been devoted to studying falls and fall-related sequelae among the amputee population. A 25-year (1974 –1999) MEDLINE review of the literature revealed few published studies. The relationship between trauma to the stump as a result of falling, and the need for stump revision has received some attention,20,21 however, no citations were found regarding the rate of falls or proportion of fear of falling among community-dwelling amputees. The risk factors for falling suggest that amputees are vulnerable. Sattin,22 in his review of the literature on falls, identified intrinsic and extrinsic risks that are associated with falling. Although we are all susceptible to the extrinsic or environmental risks, most amputees are also predisposed to the intrinsic risk factors. These risk factors include, age, chronic disease, gait and balance instability, decreased vision, altered mental status, and medication use. The amputee population, especially older amputees who have lost their limb as a result of chronic vascular disease, experience at least 1 or more of these risk factors. This leads us to believe that risk of falling should be very salient among this group. It also seems plausible that fear of falling may be important among amputees. In a survey investigating which psychosocial Arch Phys Med Rehabil Vol 82, August 2001

1032

FALLING AND FEAR OF FALLING AMONG AMPUTEES, Miller

factors were important among inpatient amputees, 58% of the subjects reported they had a fear of falling, the most common concern among all subjects.23 Furthermore, in his seminal article about the management of lower extremity amputees, Russek24 identified fear of falling as a negative factor that was believed to limit prosthetic function. Thus, although the evidence is scarce and requires reasoning by analogy, it seems that falling and related sequelae are important issues to study among the amputee population. Our study sought to determine the prevalence of and identify factors associated with having had a fall in the past 12 months and fear of falling among a sample of community-living individuals who have had a lower extremity amputation. METHODS Design and Sampling The study design was a survey of community-living individuals with a unilateral below knee (BK) or above knee (AK) lower extremity amputation who attended 1 of 2 regional university-associated amputee clinics in 1998. Participants who had a hip disarticulation or partial foot amputation were not included in the sample. Further, only individuals who had completed their initial prosthetic rehabilitation a minimum 6 months earlier and were still wearing their prostheses on a daily basis were included in the analyses of this study. Measurement Patient charts from the 2 clinics were reviewed to determine the subjects’ gender, date of birth, amputation level and cause, and date of most recent major surgical amputation. Further detailed information was collected using a survey questionnaire that was designed and implemented using the techniques recommended by Dillman.25 A total of 3 reminder letters, 1 that included a second questionnaire, were sent to potential respondents. All questions used were taken from previous published studies. Sociodemographic information, in terms of marital status, highest education level, employment, and income status was collected from the questionnaire. To determine social support, we used the 6-item version of the Interpersonal Support Evaluation List.26,27 The summed scores range from 0 to 18, with higher scores indicating stronger instrumental and emotional support. A number of variables related to the participants’ amputation were collected, such as the number of years since the individual had his/her most recent major amputation. If individuals had a revision within the same level of amputation, the original date of amputation was recorded. Individuals who had a Symes (n ⫽ 23) amputation were classified at the BK level and those with a through knee amputation (n ⫽ 5) were classified as an AK amputee. The cause of the amputation was classified as vascular or nonvascular, depending on the original reason for amputation given in the patient chart. The unweighted sum of stump and/or prosthetic problems such as experiencing pain, phantom pain, stump ulcers, or swelling in the amputated limb was calculated into a single variable. Participants were asked if they usually used a mobility device (eg, a cane, walker, crutches, wheelchair) inside or outside the home. An affirmative response, regardless of the type, was recorded as needing a mobility device. Finally, the ability of individual to walk without having to concentrate on each step was recorded. This nominal variable was taken from the Gauthier-Gagnon and Grise´28 Prosthetic Profile of the Amputee Questionnaire. Nine physical health and function variables were collected. Limitations on activities of daily living (ADLs) were deterArch Phys Med Rehabil Vol 82, August 2001

mined using the Postal Barthel Index.29 This 7-item index was modified to include 2 items on independence in donning and doffing the prosthesis. A limitation was defined as requiring human assistance when performing any basic self-care related activity. Comorbidity was established by asking respondents whether a doctor had told them that they had 1 or more of 19 common health ailments over the past 12 months. An unweighted summary score was created for analyses. Medication use related to the list of comorbidities was also recorded and summed to provide a count of the number of prescribed medications each subject was using. The presence or absence of joint or severe back pain was established by asking the participant if they had experienced pain in the past 12 months. Problems with the respondent’s good leg were identified by asking if the individual experienced pain, decreased sensation, swelling, or open wounds on their good leg or foot. To determine body mass index (BMI), an individual’s weight in kilograms was divided by the square of their height in meters.30 Self-assessed health over the past year was measured by having participants rank their health on a 5-point Likert scale, ranging from poor to excellent. Several factors assessing psychologic adjustment of the participant were measured. Depression was determined using the Center for Epidemiological Studies–Depression (CES-D) scale.31 Scores range from 0 to 60, with higher scores indicating a higher level of depressive symptoms. Adaptation to amputation and to the prosthesis was determined using a measure developed by Grise´ et al.32 Using a 5-point Likert scale, participants rate their level of adaptation, ranging from not at all to completely adapted. Fall and Fear of Falling Variables To ascertain fall status, participants were asked “Have you fallen in the past 12 months?” A fall was defined as landing on the floor or the ground unintentionally.10,33 Answers were recorded as either a yes or no. In addition, a series of descriptive questions related to their fall, such as a resultant injury or need to seek medical attention, were asked of those who indicated they had fallen. Fear of falling was discerned by asking individuals “Are you afraid of falling?”—an approach used in other studies.17,33 Analyses Before conducting analyses of the data, we adjusted several of the variables to facilitate the presentation and understanding of the results. The values for continuous variables were divided into groups based on a meaningful cutoff point as established in the literature, such as a score of 15 or higher on the CES-D scale,31 or based on quartile, tertile, or median distribution. Because of the binary nature of the outcome variables, logistic regression was used as the primary form of analyses in the study. Univariate logistic regression was conducted to produce odds ratios (ORs) and 95% confidence intervals (CIs) to identify variables that would be considered as candidates for multivariable modeling. All variables found to be statistically associated (p ⬍ .20) at the univariate level and variables that were considered clinically important34 (amputation cause and level) were included in multivariable modeling to ensure that borderline relationships would not be overlooked. A priori clinical opinion was that amputation cause and level would be important factors for both outcomes, therefore these 2 variables were forced into the final models if they were not statistically important. Otherwise the final models presented include all variables found to be statistically related at p less than 0.1. The proportion of missing variables for the present study was low (⬍5%); however, to maximize cases available for

1033

FALLING AND FEAR OF FALLING AMONG AMPUTEES, Miller

multivariable modeling, missing values were imputed using the methods advocated by Schafer,35 using the multiple imputation software NORM.a All data entry and analyses were conducted using the SPSS for Windows, version 8.b RESULTS Sample Of the 563 eligible subjects, 435 (77%) responded to the study survey. The average age ⫾ standard deviation of the primarily male (71%) sample was 62 ⫾ 15.7 years. Most participants were married (62%), not working (75%), and had less than a grade 12 education (58%). Most also had a BK amputation (73%); amputation was slightly more commonly performed for dysvascular (53%) than for nonvascular (47%) reasons. The average time since amputation was 14.7 ⫾ 16.7 years. No differences were noted for nonresponders based on age, gender, amputation cause, level, or the number of years since amputation. Prevalence of Falling and Fear of Falling Of the total sample, 228 (52.4%) reported having fallen in the previous 12 months; of these, 171 (75.0%) reported falling 2 or more times. Injury as a result of the fall occurred among 92 (40.4%) of fallers, with 44 (19.3%) seeking medical attention due to their fall. A total of 214 (49.2%) subjects expressed a fear of falling, and 163 (76.2%) of these indicated that they avoided activities as a result of their fear. Of those who had fallen, 55% reported a fear of falling, compared with 43% among nonfallers. Covariates of Falling and Fear of Falling In table 1, we present various sociodemographic, amputation, function, physical, and psychologic characteristics and their bivariate relationship with fallers versus nonfallers. Characteristics related (p ⬍ .05) to having fallen in the past 12 months included those with AK amputation, stump or prosthetic problems, multiple comorbidities, back pain, and multiple problems with their contralateral or “good” leg. Those with a strong social support system, good to excellent self-assessed health, and having had their amputation a moderate amount of time seemed to be protected from experiencing a fall. Fear of falling was associated at p less than 0.2 with all variables except amputation level, BMI, and back pain at the univariate level (table 2). Furthermore, most covariates were associated with fear of falling at p less than .05, as indicated by the exclusion of a value of 1 in the CIs around the OR. Multivariable Analysis: Falling in the Past 12 Months In total, 22 variables were considered in the initial logistic regression model, with falling as the dependent variable. Of these, 8 variables, primarily health-related factors, remained in the model (table 3). The presence of back (OR ⫽ 1.96; 95% CI ⫽ 1.08 –3.54) and joint pain (OR ⫽ 1.67; 95% CI ⫽ 1.01–2.74), being a unilateral AK amputee (OR ⫽ 2.78; 95% CI ⫽ 1.71– 4.51), and having 4 or more problems related to the amputated stump or prosthesis (OR ⫽ 3.09; 95% CI ⫽ 1.58 – 6.04) appeared to present the largest level of risk associated with having fallen in the past 12 months. BMI score (OR ⫽ 1.04; 95% CI ⫽ 1.00 –1.09) was very close to being considered statistically significant, but fell just short of the acceptable a priori statistical cutoff point. Having had the amputation 4 or more years in the past appeared to provide a protective effect on fall occurrence.

Table 1: Characteristics of Subjects Who Did and Did Not Fall in the Past 12 Months Characteristics

Sociodemographic Age (yr) 23–55 56–70 ⱖ71 Men Married Social support (⬎ median) Education (⬎ grade 11) Not working Income (Can $) 0–19,999 20,000–39,999 ⱖ40,000 Amputation Time since amputation (yr) 0–3 4–13 14⫹ Cause (nonvascular) Level (AK) Concentrating on walking Problems with stump/ prosthesis 0 1 2–3 4⫹ Mobility device use Physical Health and Function Comorbidity None 1 2⫹ Back pain Joint pain ADL limitation(s) BMI (score ⬎ 27) Problems with good leg 0 1–2 3–4 Health (good to excellent) No. of medications 0 1 2⫹ Alcohol intake None Occasional Daily Psychologic Depression (CES-D score ⬎ 15) Adaptation to prosthesis (quite/extremely well) Adaptation to amputation (quite/extremely well) Fear of falling

Fall (n ⫽ 228)

No Fall (n ⫽ 207)

OR

95% CI

79 74 75 159 146 104 94 165

65 64 78 148 124 115 90 159

1.00 .95 .79 .92 1.19 .67 .91 .79

.60–1.52 .50–1.25 .61–1.39 .81–1.76 .46–.98 .62–1.33 .51–1.22

90 77 61

79 71 57

1.00 .95 .94

.61–11.48 .59–1.50

81 70 77 114 77 102

54 83 70 91 39 76

1.00 .56 .25–.90 .73 .46–1.18 1.27 .87–1.86 2.20 1.41–3.42 1.40 .95–2.05

23 42 49 93 99

39 38 51 50 90

1.00 1.87 .95–3.69 1.63 .85–3.11 3.15 1.70–5.86 1.00 .68–1.46

42 45 157 56 88 57 122

48 29 114 24 41 58 99

1.00 .74 .40–1.37 1.57 .97–2.54 2.48 1.47–4.18 2.54 1.65–3.93 .86 .56–1.31 1.26 .86–1.83

82 77 69 151

97 61 49 150

1.00 1.49 .96–2.33 1.67 1.04–2.66 .75 .49–1.12

72 39 117

81 38 88

1.00 1.16 1.50

.67–2.00 .98–2.28

142 46 37

134 36 40

1.00 1.21 1.02

.73–1.98 .61–1.69

63

43

1.46

.93–2.27

175

167

.79

.49–1.26

184 126

171 88

.88 .54–1.43 1.67 1.14–2.44

Arch Phys Med Rehabil Vol 82, August 2001

1034

FALLING AND FEAR OF FALLING AMONG AMPUTEES, Miller

Table 2: Characteristics of Subjects Who Did and Did Not Report a Fear of Falling Characteristics

Sociodemographic Age (yr) 23–55 56–70 ⱖ71 Men Married Social support (⬎15) Education (⬎ grade 12) Not working Income (Can $) 0–19,999 20,000–39,999 ⱖ 40,000 Amputation Time since amputation (yr) 0–3 4–13 14⫹ Cause (nonvascular) Level (AK) Concentrating on walking Problems with stump/prosthesis 0 1 2–3 4⫹ Mobility device use Physical Health and Function Comorbidity None 1 2⫹ Back pain Joint pain ADL limitation(s) Fall in past 12 months Injury from fall BMI (score ⬎ 27) Problems with good leg 0 1–2 3–4 Health (good to excellent) No. of medications 0 1 2⫹ Alcohol intake None Occasional Daily Psychologic Depression (CES-D score ⬎ 15) Adaptation to prosthesis (quite/extremely well) Adaptation to amputation (quite/extremely well)

Fear (n ⫽ 214)

No Fear (n ⫽ 221)

63 71 80 126 121 93 80 144

81 67 73 181 149 126 104 180

1.00 1.36 1.41 .32 .63 .58 .67 .47

.85–2.18 .89–2.23 .20–.49 .43–.93 .40–.85 .46–.98 .30–.73

102 66 46

67 82 72

1.00 .53 .42

.34–.83 .26–.68

OR

95% CI

79 71 64 89 55 129

56 82 83 116 61 49

1.00 .61 .55 .64 .91 5.33

24 30 76 84 121

38 50 74 59 68

1.17 1.00 .95 1.63 2.25 2.93

24 27 146 40 78 72 126 56 106

38 47 125 40 51 43 102 30 115

1.00 .69 1.40 1.04 1.91 2.10 1.67 2.56 .91

.37–1.29 .86–2.25 .64–1.69 1.26–2.91 1.36–3.25 1.14–2.44 1.38–3.69 .62–1.32

68 78 68 118

111 60 50 183

1.00 2.12 2.22 .26

1.35–3.33 1.38–3.57 .16–.40

65 30 119

88 47 86

1.00 .86 1.87

.49–1.51 1.23–2.86

146 34 34

130 48 43

1.00 .63 .070

.38–1.04 .42–1.17

71

35

2.64

1.67–4.18

147

195

.29

.18–.48

161

194

.42

.25–.70

Arch Phys Med Rehabil Vol 82, August 2001

.39–.98 .34–.88 .44–.94 .59–1.39 3.50–8.10 1.07–1.27 .48–1.88 .89–2.97 1.22–4.15 1.98–4.33

Table 3: Risk Factors for Presence of a Fall in the Past 12 Months Risk Factor

OR

95% CI

AK amputation Nonvascular amputation Back pain (% yes) Joint pain (% yes) Comorbidity None 1 2⫹ Stump/prosthetic problems 0 1 2–3 4⫹ BMI Years since amputation 0–3 4–13 14⫹

2.78 1.40 1.96 1.67

1.71–4.51 .83–2.37 1.08–3.54 1.01–2.74

1.00 .64 1.27

.33–1.26 .73–2.21

1.00 1.84 1.40 3.09 1.04

.89–3.77 .73–2.68 1.58–6.04 1.00–1.09

1.00 .53 .53

.32–.89 .29–.98

NOTE. Goodness of fit ⫽ 7.47; df ⫽ 8; p ⫽ .47; % correct average ⫽ 66.4; sensitivity ⫽ 69.3; specificity ⫽ 63.3.

Multivariable Analysis: Fear of Falling BMI and back pain were the only covariates at the univariate level (table 2) not included in multivariable modeling for fear of falling. Of the 26 factors originally considered in the model, 8 variables remained in the final reduced model (table 4). Being male (OR ⫽ .35; 95% CI ⫽ .21–.57), and having good or better perceived health (OR ⫽ .35; 95% CI ⫽ .21–.58) both appeared to protect against having a fear of falling. Individuals who had to concentrate on each step they took (OR ⫽ 4.06; 95% CI ⫽ 2.46 – 6.71) and who had fallen in the past 12 months (OR ⫽ 1.62; 95% CI ⫽ 1.04 –2.54) had a higher risk of having a fear of falling. Two additional factors—not working and use of a mobility device— did not reach the conventional level of statistical significance but were associated at p less than 0.1. DISCUSSION The prevalence of having a fall in the past 12 months was a common occurrence among our sample. Slightly over half of the respondents indicated that they had fallen at least once. Although the literature on falls among the amputee population is virtually nonexistent, having had an amputation has been reported as an increased risk for falling in acute-care hospitals.36 Compared with other populations studied, falling among community-living amputees is at the high end of the reported

Table 4: Risk Factors for Fear of Falling Risk Factor

OR

95% CI

Gender (% male) Not working AK amputation Nonvascular amputation Mobility device use Concentrating on walking Fall in the past 12 months Good to excellent perceived health

.35 .63 .66 1.22 1.61 4.06 1.62 .35

.21–.57 .38–1.05 .39–1.11 .75–1.99 .97–2.67 2.46–6.71 1.04–2.54 .21–.58

NOTE. Goodness of fit ⫽ 11.0; df ⫽ 8; p ⫽ .20; % correct average ⫽ 73%; sensitivity ⫽ 72%; specificity ⫽ 73%.

FALLING AND FEAR OF FALLING AMONG AMPUTEES, Miller

range. Although our sample reported a fall prevalence of 52.4%, previous studies report a fall frequency of 30% to 40% among community-living elderly individuals7,11 and 50% among institutionalized elderly.13,14 The occurrence of injury among the fallers (40.4%) and the need to seek medical attention as a result of a serious injury (19.3%) are within the reported range (18%–24%) of other studies.10,12 Fear of falling was also highly prevalent among our sample; 49.2% reported they experienced this fear. As shown in both univariate and multivariable analysis, the fear of falling appears to be more of a problem among the fallers (55%) in our sample, compared with nonfallers (43%). The reported range of this fear for community-living elderly is 25% to 50%1,37,38 and is similar to individuals who have rheumatoid arthritis.39 The amputation and physical health and function factors seemed to predominate the final model predicting falls. As anticipated, we found that having an AK amputation was independently associated with an increase risk of falls. This finding seems logical given that the knee joint is a major determinant of gait40 and the loss of the knee joint is associated with decreased proprioception and increased postural sway. Having a vascular amputation was not an independent risk for falling, despite evidence suggesting higher sway among vascular amputees when compared with traumatic amputees and nonamputee controls.41 Finding that individuals who had their amputation more recently had a higher risk of falling compared with those who have had their prosthesis for a longer period of time was expected. Aside from having an increased accommodation and more proficient use of their prosthesis, it is likely that these individuals are better able to assess and avoid risky environments that are associated with falls. Although the influence of the environment has been suggested as a potential causative factor of falling, assessing environmental hazards was not feasible given our study design. We also found that joint and back pain was associated with an increased risk for falling of between 1.67 and 1.96 times. It is plausible that these factors could lead to a fall because they are both commonly associated with the amputee population because of, among other things, an asymmetric gait. However, it is also possible that a fall may cause joint and back pain. The precise temporal nature of this relationship cannot be determined, using our cross-sectional design. On the other hand, in at least 1 study9 investigating incidence of multiple falls among older persons, having a history of falling with injury was independently important in multivariable modeling. Therefore, musculoskeletal pain may be predictive of falls in the amputee population. A plausible alternative explanation is that the pain variables may be acting as a proxy for analgesic use. Others12,17 have found that medication use was associated with falls among the elderly. Interestingly, medication use is not important even at the univariate level in our study. Perhaps combining all of the prescribed medications into an unweighted summary variable may explain the null finding. Distinguishing between classes of medications, particularly sedatives and analgesics, may shed additional light on this issue and is worthy of consideration in future studies. The findings in the literature related to alcohol use appear to be equivocal. O’Loughlin et al10 found that daily alcohol use was protective of falls in their study, however, similar to our study, Tinetti et al12 reported no relationship at the univariate or multivariable level for alcohol use. Although all of the studies used self-report methods to collect information about alcohol use, variable accuracy of self-report across studies may explain these inconsistent findings.

1035

A number of factors distinguished fear of falling in the final multivariate model. We found that being male and having a good to excellent perceived health were protective factors, while having had a fall in the past 12 months and having to concentrate on walking were risk factors. In general, most studies investigating fear of falling find that men report less fear.2,39,42,43 Gender seems to be the only factor that is consistently associated with fear of falling across studies. Common sense suggests that having a good or better perceived health status would be associated with less fear of falling. Individuals with better health are more likely to be stronger, vibrant, and more active. Furthermore, these individuals are also more likely to have greater confidence in their balance and ability to perform activities. Confidence does appear to be critical because individuals who have to concentrate on each step while walking have a fourfold risk of fear of falling even after controlling for other covariates. Other proxy measures of confidence have been investigated. For example, several authors2,17,37 found that using a mobility device was important at the univariate level but not at the multivariate level. Using a mobility device entered the final reduced model in our study; however, it just failed to reach statistical significance. The association between fear and confidence has received increased attention. In fact, measures of efficacy with balance have been used more in studies to act as proxy measures for fear of falling. Considerable debate surrounds the importance of having an antecedent fall on fear of falling. Among those studies examining factors associated with fear of falling, several researchers17,39,44 have found having an antecedent fall was not important, while others2,37,38 reported findings similar to ours. These findings likely reflect the complex and circular relationships that exists between these 2 factors. That is, while falling may lead to a fear of falling, fear of falling does not necessarily protect an individual from having a fall. A number of studies have used the global measure of fear of falling that we used in our study. Recently, researchers3,33,45 have suggested that general measures are susceptible to several limitations. One such limitation is the belief that a global question reflects a more general state of anxiety that is not specific enough for falling. Furthermore, it has been suggested that the stigma associated with being afraid may be a reason why gender differences are such a prevailing factor. To overcome these limitations, more refined measures of fear of falling based on the theory of self-efficacy or confidence46,47 have been developed. We are not alone in finding that those with a fear of falling avoid activities.39 Activity avoidance is of concern because it may potentially lead to a decreased quality of life. Furthermore, while reducing activity can be adaptive when that activity places an individual at increased risk of falling, over-restriction can become maladaptive. Inactivity can lead to further debilitation related to reduced strength, endurance, and balance, and may increase the potential for further health problems including falls. To date, the influence of falling and fear of falling as risk factors for inactivity such as prosthetic mobility and participation in social activities among individuals with a lower extremity amputation has not been investigated. Further study to assess the relevance of this relationship is required to determine if inactivity is a consequence of falling and fear of falling. Many prosthetic rehabilitation programs, such as the 2 clinics from which the study sample was drawn, conduct education on fall routines. In our case, the clinic uses a simulated fall situation to inform and train patients and their spouses to expect and solve problems in case there is a fall. Furthermore, a home visit is conducted in most cases to reduce environmenArch Phys Med Rehabil Vol 82, August 2001

1036

FALLING AND FEAR OF FALLING AMONG AMPUTEES, Miller

tal hazards that may exist. Despite this preparation, falling and fear of falling continue to be a pervasive problem that may be of greater concern among clinics that do not conduct fall education. Although the effectiveness of fall education programs cannot be answered by the results of this study, and is worthy of research, we suggest that additional education during initial rehabilitation and primer sessions may need to be considered. Study Limitations The results of the study were not obtained prospectively, therefore, claims of causal relationships must be made with caution. This is especially true for assessing the risks of falling. While it seems reasonable to accept claims of a temporal relationship with a fear of falling because fear is measured based on current status, we are less certain when the fall occurred in relation to the other factors. For example, problems with the stump or prosthesis could have occurred as a result of or as an antecedent to falling. For this reason, fear of falling was not considered as a risk for falling in our analyses, despite suggestions that such a relationship might be salient. Recall bias may also have been a factor in ascertaining the occurrence of a fall in the past 12 months. We chose to review the past 12 months because Cummings et al48 suggested that recall over this period of time appears to be better than over shorter periods. We could have, however, selected a longer period of time (last 5yr), because some investigators2 have suggested that this time period may be just as influential. We chose a year because it fit with our design. Longer periods may have selected out large portions of our sample. With the understanding that falling is an important issue among lower extremity amputees, research using prospective designs to determine the fall incidence is recommended. Further, a prospective study could investigate the circumstances of the falls and would be able to assess the direction of the fear of falling relationship. CONCLUSION We have presented the results from the first known study of falling and fear of falling among individuals with lower extremity amputations. Both were found to be prevalent and thus important and are worthy of further study. However, falling and fear of falling may be risk factors for inactivity among individuals with lower extremity amputations. Although the relationship with activity remains unclear at present, we are currently analyzing data that will provide a preliminary understanding of the potential consequence of falling and fear of falling. Falling and fear of falling may provide clinicians with a good marker to monitor physiologic and psychologic change as well as prosthetic use and usefulness. Increased surveillance may provide insight for the clinician so to indicate decreased prosthetic function and limitation in daily activities. Interventions directed at falling and fear of falling should be considered within the initial training and the principles reviewed and reinforced on follow-up visits. Further research is required to confirm the results of this study and to investigate the usefulness of interventions aimed at reducing falling and fear of falling among the amputee population. Acknowledgments: The authors thank Dr. Mark Bayley for his assistance with subject recruitment; the medical records staff at the 2 participating hospitals; and Susan Brown and Cathy Kingdon, who were instrumental in data collection. Arch Phys Med Rehabil Vol 82, August 2001

References 1. Tinetti ME, Mendes de Leon CF, Doucette JT, Baker DI. Fear of falling and fall-related efficacy in relationship to functioning among community-living elders. J Gerontol 1994;49:M140-7. 2. Howland J, Lachman ME, Walker JE, Peterson E, Cote J, Kasten L, et al. Covariates of fear of falling and associated activity curtailment. Gerontologist 1998;38:549-55. 3. Tinetti ME, Powell L. Fear of falling and low self-efficacy: a cause of dependence in elderly persons. J Gerontol 1993;48:35-8. 4. Vellas BJ, Wayne SJ, Romero LJ, Baumgartner RN, Garry PJ. Fear of falling and restriction of mobility in elderly fallers. Age Ageing 1997;26:189-93. 5. Blake AJ, Morgan CA, Bendall MJ, Dallosso H, Ebrahim SB, Arie TH, et al. Falls by elderly people at home: prevalence and associated factors. Age Ageing 1988;17:365-72. 6. Campbell AJ, Borrie M, Spears GF, Jackson SL, Brown JS, Fitzgerald JL. Circumstances and consequences of falls experience by a community population 70 years and over during a prospective study. Age Ageing 1990;19:136-41. 7. Campbell AJ, Reinken J, Allan BC, Martinez GS. Falls in old age: a study of frequency and related clinical factors. Age Ageing 1981;10:264-70. 8. Cwikel J. Falls among elderly people living at home: medical and social factors in a national sample. Isr J Med Sci 1992;28:446-53. 9. Nevitt MC, Cummings SR, Kidd S, Black D. Risk factors for recurrent nonsyncopal falls: a prospective study. JAMA 1989;261: 2663-8. 10. O’Loughlin JL, Robitaille Y, Boivin J, Suissa S. Incidence of and risk factors for falls and injurious falls among the communitydwelling elderly. Am J Epidemiol 1993;137:342-54. 11. Prudham D, Evans JG. Factors associated with falls in the elderly: a community study. Age Ageing 1981;10:141-6. 12. Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among elderly persons living in the community. N Engl J Med 1988;319: 1701-7. 13. Tinetti ME. Factors associated with serious injury during falls by ambulatory nursing home residents. J Am Geriatr Soc 1987;35: 644-8. 14. Tinetti ME, Speechley M. Prevention of falls among the elderly. N Engl J Med 1989;320:1055-9. 15. Speechley M, Tinetti ME. Falls and injuries in frail and vigorous community elderly persons. J Am Geriatr Soc 1991;39:46-52. 16. Nevitt MC. Falls in older persons: risk factors and prevention. In: Berg RL, Cassells JS, editors. The second fifty years: promoting health and preventing disability. Washington (DC): National Academy Pr; 1990. p 263-90. 17. Maki BE, Holliday PJ, Topper AK. Fear of falling and postural performance in the elderly. J Gerontol 1991;46:M123-21. 18. Walker JE, Howland J. Falls and fear of falling among elderly persons living in the community: occupational therapy interventions. Am J Occup Ther 1991;45:119-22. 19. Myers A, Gonda G. Research on physical activity in the elderly: practical implications for program planning. Can J Aging 1991; 5:175-87. 20. Aligne C, Farcot M, Favre JP, Alnashawati G, De Simone F, Barral X. Primary closure of below-knee amputation stumps: a prospective study of sixty-two cases. Ann Vasc Surg 1990;4: 143-6. 21. Behar TA, Burnham SJ, Johnson G. Major stump trauma following below-knee amputation. Outcome and recommendations for therapy. J Cariovasc Surg 1991;32:753-5. 22. Sattin RW. Falls among older persons: a public health perspective. Ann Rev Public Health 1992;13:489-508. 23. MacBride A, Rogers J, Whylie B, Freeman SJ. Psychosocial factors in the rehabilitation of elderly amputees. Psychosomatics 1980;21:258-65. 24. Russek AS. Management of lower extremity amputees. Arch Phys Med Rehabil 1961;42:687-703. 25. Dillman DA. Mail and telephone surveys: the total design method. New York: Wiley; 1978. 26. Cohen S, Mermelstein R, Kamarck T, Hoberman HN. Measuring the functional components of social support. In: Sarason I, Sara-

FALLING AND FEAR OF FALLING AMONG AMPUTEES, Miller

27. 28. 29. 30. 31. 32.

33. 34. 35. 36. 37. 38.

son B, editors. Social support: theory, research, and applications. Dordrecht (Netherlands): Martinus Nijhoff; 1985. p 73-94. Schulz R, Williamson G. A 2-year longitudinal study of depression among Alzheimer’s caregivers. Psychol Aging 1991;6:56978. Gauthier-Gagnon C, Grise´ MC. Prosthetic Profile of the Amputee Questionnaire: validity and reliability. Arch Phys Med Rehabil 1994;75:1309-14. Gompertz P, Pound P, Ebrahim S. A postal version of the Postal Barthel Index. Clin Rehabil 1994;8:233-9. Stavig GR, Leonard AR, Ingra A, Felten P. Indices of relative body weight and ideal weight charts. J Chron Dis 1984;37:255-62. Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas 1977;1: 385-401. Grise´ MC, Gauthier-Gagnon C, Martineau GG. Prosthetic profile of people with lower extremity amputation: conception and design of follow-up questionnaire. Arch Phys Med Rehabil 1993;74:86270. Tinetti ME, Richman D, Powell L. Falls efficacy as a measure of fear of falling. J Gerontol 1990;45:P239-43. Hosmer DW, Lemeshow S. Applied logistic regression. New York: Wiley; 1989. Schafer JL. Analysis of incomplete multivariate data. New York: Chapman & Hall; 1997. Valhov D, Myers AH, al-Ibrahim MS. Epidemiology of falls among patients in a rehabilitation hospital. Arch Phys Med Rehabil 1990;7:8-12. Arfken CL, Lach HW, Birge SJ, Miller JP. The prevalence and correlates of fear of falling in elderly persons living in the community. Am J Public Health 1994;84:565-70. Howland J, Peterson EW, Levin WC, Fried L, Pordon D, Bak S. Fear of falling among the community-dwelling elderly. J Aging Health 1993;5:229-43.

1037

39. Fessel KD, Nevitt MC. Correlates of fear of falling and activity limitation among persons with rheumatoid arthritis. Arthritis Care Res 1997;10:222-8. 40. Inman VT, Ralston HJ, Todd F. Human walking. Berkeley (CA): Williams & Wilkins; 1981. 41. Hermodsson Y, Ekdahl C, Persson BM, Roxendal G. Standing balance in trans-tibial amputees following vascular disease or trauma: a comparative study with health subjects. Prosthet Orthot Int 1994;18:150-8. 42. Maki BE. Gait changes in older adults: predictors of falls or indicators of fear? J Am Geriatr Soc 1997;45:313-20. 43. Vellas B, Bocquet H, de Pemile F, Albarde J. Prospective study of restriction of activity in old people after falls. Age Ageing 1987; 16:189-93. 44. Lawrence RH, Tennstedt SL, Kasten LE, Shih J, Howland J, Jette AM. Intensity and correlates of fear of falling and hurting oneself in the next year. J Aging Health 1998;10:267-86. 45. Powell L, Myers A. The activities-specific balance confidence (ABC) scale. J Gerontol 1995;50:M28-34. 46. Bandura A. Self-efficacy mechanism in human agency. Am Psychol 1982;37:122-47. 47. Bandura A. Self-efficacy: toward a unifying theory of behavior change. Psychol Rev 1997;84:191-215. 48. Cummings CR, Nevitt MC, Kidd S. Forgetting falls: the limited accuracy of recall of falls in the elderly. J Am Geriatr Soc 1988;36:613-6. Suppliers a. Schafer JL. NORM: multiple imputation of incomplete multivariate data under a normal model [computer program for Windows 95/98/ NT]. Version 2; 1999. Available from: http://www.stat.psu.edu/⬃/ jls/misoftwa.html#top. Accessed May 26, 2001. b. SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.

Arch Phys Med Rehabil Vol 82, August 2001