1 Falls and Fear of Falling: Which Comes First? A Longitudinal Prediction Model Suggests Strategies for Primary and Secondary Prevention Susan M. Friedman, MD, MPH,* Beatriz Munoz, MS, Sheila K. West, PhD, Gary S. Rubin, PhD, and Linda P. Fried, MD, MPH OBJECTIVES: Previous cross-sectional studies have shown a correlation between falls and fear of falling, but it is unclear which comes first. Our objectives were to determine the temporal relationship between falls and fear of falling, and to see whether these two outcomes share predictors. DESIGN: A 20-month, population-based, prospective, observational study. SETTING: Salisbury, Maryland. Each evaluation consisted of a home-administered questionnaire, followed by a 4- to 5-hour clinic evaluation. PARTICIPANTS: The 2,212 participants in the Salisbury Eye Evaluation project who had baseline and 20-month follow-up clinic evaluations. At baseline, subjects were aged 65 to 84 and community dwelling and had a Mini- Mental State Examination score of 18 or higher. MEASUREMENTS: Demographics, visual function, comorbidities, neuropsychiatric status, medication use, and physical performance based measures were assessed. Stepwise logistic regression analyses were performed to evaluate independent predictors of falls and fear of falling at the follow-up evaluation, first predicting incident outcomes and then predicting fall or fear-of-falling status at 20 months with baseline falling and fear of falling as predictors. RESULTS: Falls at baseline were an independent predictor of developing fear of falling 20 months later (odds ratio (OR) 1.75; P.0005), and fear of falling at baseline From the *University of Rochester School of Medicine and Dentistry, Rochester, New York; Dana Center for Preventive Ophthalmology and Departments of Medicine and Epidemiology, The Johns Hopkins University, Baltimore, Maryland; and The Institute of Ophthalmology, London, United Kingdom. Presented in part at the American Geriatrics Society National Meeting, May This research was supported by National Institute on Aging Grant AG Dr. Friedman was supported by a Pfizer/American Geriatrics Society Postdoctoral Fellowship and by a K23 career development award, #1K23AG Address correspondence to Susan M. Friedman, MD, MPH, Highland Hospital, 1000 South Avenue, Box 58, Rochester, NY was a predictor of falling at 20 months (OR 1.79; P.0005). Women with a history of stroke were at risk of falls and fear of falling at follow-up. In addition, Parkinson s disease, comorbidity, and white race predicted falls, whereas General Health Questionnaire score, age, and taking four or more medications predicted fear of falling. CONCLUSION: Individuals who develop one of these outcomes are at risk for developing the other, with a resulting spiraling risk of falls, fear of falling, and functional decline. Because falls and fear of falling share predictors, individuals who are at a high risk of developing these endpoints can be identified. J Am Geriatr Soc 50: , Key words: falls; fear of falling; older adults; predictors; prevention Falls and fear of falling are common syndromes with potentially serious outcomes in older communitydwelling adults. The correlation between falls and fear of falling has been well demonstrated, 1 6 but the temporal relationship between these two syndromes in a populationbased sample is less clear. Any of the following could explain the association between falls and fear of falling: falls may cause fear of falling, fear of falling may cause falls, and the two outcomes may be related to shared risk factors and not themselves causally related. It is certainly plausible that each syndrome may lead to the other. Namely, an individual who falls may subsequently develop fear of further falls. This has been demonstrated as an immediate consequence of falls 7 and in studies that have examined this restrospectively, 1 3,5,6 but the long-term prospective risk is less well characterized. Moreover, it is possible that individuals who fall and have subsequent fear of falling had experienced fear before the fall as well. This distinction underlies the concept of primary prevention, in which the onset of fear of falling is prevented, versus secondary prevention, in which fear of falling is prevented from progressing. JAGS 50: , by the American Geriatrics Society /02/$15.00
2 1330 FRIEDMAN ET AL. AUGUST 2002 VOL. 50, NO. 8 JAGS Conversely, fear of falling may in turn cause falls. This could potentially be mediated through gait changes or through activity restriction and deconditioning. The data supporting this direction of association are mixed. Maki et al. 8 demonstrated that fear of falling was associated with multiple gait changes, namely, decreased stride length, decreased speed, increased double stance time, decreased clinical gait scores, and increased stride width. However, each of these gait changes should be stabilizing, thereby decreasing fall risk, and the only one of these gait changes that was predictive of falls at follow-up was increased stride width. Several studies have shown activity restriction secondary to fear of falling, 1,9 which could in turn lead to deconditioning and increased risk of falling. Alternatively, a reduction in activity might lead to fewer opportunities to fall. Two recent prospective studies have suggested that fear of falling causes falls. Cumming et al. 10 showed that patients with poor Falls Efficacy Scale (FES) scores recruited from two hospitals were twice as likely to fall in the subsequent 12 months than were those with better FES scores. Another study, using a recruited population, showed that home-dwelling individuals who expressed frequent fear of falling were twice as likely as those who did not to experience multiple falls in the subsequent year. 11 What remains to be shown is whether fear of falling predicts onset of falling in a general older community-based population. A final explanation for the cross-sectional correlation of falls and fear of falling is that these outcomes are not directly related but result from common underlying risk factors, including balance disturbance, 6,7,9,12,13 cognitive status, 7,12 and vision problems, 1,14,15 that increase the risk of both outcomes. If falls lead to fear of falling and vice versa, then development of one may start a cascade effect, with a resulting escalation of risk for functional decline and other adverse outcomes. Furthermore, if these syndromes have shared risk factors, then individuals with these characteristics would be at particularly high risk of developing this vicious cycle, and this group would be important to target for prevention strategies to reduce falls and fear of falling. This study uses longitudinal data from the Salisbury Eye Evaluation (SEE) study, a population-based, prospective, observational study in Salisbury, Maryland, to answer the following questions. Is fear of falling an independent predictor for becoming a faller, and, conversely, is a history of falls an independent predictor of the onset of fear of falling? What are the independent risk factors for the new onset of falls and for the new onset of fear of falling in individuals who do not report these syndromes at baseline? Do these risk profiles differ? Are there shared predictors for developing these two outcomes that could be targeted for primary prevention? What are the predictors of falls and fear of falling when baseline falls and fear of falling are included in the model? Are there similar risk profiles for both outcomes that could be targeted for secondary prevention? METHODS Study Design A major goal of the SEE study was to determine the effect of visual impairment on function and falls in older adults. Participants were recruited from age-stratified random samples of Health Care Financing Administration (HCFA; now called Center for Medicare and Medicaid Services) Medicare eligibility lists, with oversampling of African Americans and individuals aged 75 to 84 so as to maximize the power of the study to draw conclusions about these subgroups. Further details of the recruitment of this cohort have been reported previously. 16 Participants Two thousand five hundred twenty men and women aged 65 to 84 with a Mini-Mental State Examination (MMSE) score of 18 or higher were enrolled in the SEE study. Individuals with lower MMSE scores were not included because of concerns of recall and participation in testing. The original cohort was evaluated between September 1993 and September Of the original cohort, 147 died (5.8%) and 133 did not return for follow-up (5.3%). This left 2,220 (88.9%) who returned for a second evaluation 20 months after their first evaluation, of whom 2,212 (99.6%) had complete falls and fear of falling information. The 308 nonresponders were older (74.6 vs 72.6) and had a lower MMSE score (a median of 27 vs 28), a higher General Health Questionnaire (GHQ) score (median of 1 vs 0), and lower knee strength (median of 18.2 vs 20.0 kg) than the 2,212 subjects who were evaluated for this analysis. They were also more likely to report falls in the previous year (34.3% vs 27.8%) and to report fear of falling (29.3% vs 20.8%). Data Collection Each evaluation consisted of a home-administered questionnaire, followed by a 4- to 5-hour clinic evaluation, in which subjects underwent standardized vision and function testing and answered questions about functional status, falls, and balance. Demographics Age was recorded as the age at the time of the first clinic visit, as indicated by HCFA Medicare data. Other demographic characteristics, including gender, race, and education, were determined via self-report. Vision Assessment The vision variables assessed for this analysis were visual acuity, contrast sensitivity, stereoacuity, and visual fields. Visual acuity was tested using binocular habitual correction with Early Treatment of Diabetic Retinopathy Study (ETDRS) charts. ETDRS charts measure acuity on a log of the minimum angle of resolution (logmar) scale, where a logmar value of 0.0 corresponds to 20/20 Snellen, legal blindness (20/200) has a value of 1.0, and a change of 0.1 units represents a change of one line in acuity. A 0.3 increase in ETDRS acuity means a doubling of the angular size of the letters and is generally considered clinically significant. Contrast sensitivity was measured at 1 meter under controlled room illumination, using the Pelli-Robson chart. The chart has 16 groups of three letters each, of
3 JAGS AUGUST 2002 VOL. 50, NO. 8 FALLS AND FEAR OF FALLING 1331 constant size but decreasing contrast. Results were recorded as the number of letters read correctly. Normal contrast sensitivity is about 36 letters, and a change of six letters represents a doubling of the contrast required to see the targets. Stereoacuity was tested using the Randot Circles test, with a graded disparity ranging from 450 arcsec to 17 arcsec at a distance of 36 cm. The disparity represents the angular displacement of an object in depth. Ten patterns requiring increasing depth perception were used, and the number read correctly was recorded (0 10). Participants who were able to see depth in six or more of the stereoacuity patterns would be considered to have excellent stereovision. Those who could not see depth in any of the patterns were classed as stereo blind. Visual fields were tested with the 81-point screening test of the Humphrey Field Analyzer. The number of points missed was transformed by the square root to normalize the distribution. A score of 5 or less would be considered normal for our older population. Further details of the vision testing have been previously described. 17 Comorbidities Fifteen chronic conditions (arthritis, history of hip fracture, stroke, Parkinson s disease, vertigo within the previous 2 years, back problems, history of myocardial infarction, angina pectoris, congestive heart failure, claudication, hypertension, diabetes mellitus, emphysema, asthma, and cancer) were determined by self-report. The first five of these chronic conditions were used as individual predictors, because they have been shown in previous studies to be predictors of falls in older adults. 9,18,19 The remaining 10 were totaled and used as a continuous measure of comorbidity, in a comorbidity index that could range from Subjects were also asked whether they experienced foot pain on most days for at least 1 month during the previous year. Neuropsychiatric Status Cognitive status was evaluated via the MMSE. 20 Scores range from 0 to 30, with lower scores indicating greater cognitive impairment. Psychiatric status was measured with the GHQ, a 28-question screen for current psychiatric disturbance that asks questions about depression, anxiety, social impairment, and hypochondriasis. 21 A higher score represents a greater number of psychiatric symptoms. Medications Medications were recorded at the time of the home interview and were defined as any oral, intravenous, or intramuscular medication other than vitamins or electrolytes. The number of medications was then totaled. Number of medications was dichotomized to those taking four or more versus fewer than four medications because this division had previously been shown to be associated with fall risk. 13 Sedatives were defined as any benzodiazepines, phenothiazines, or antidepressants, as defined in a previous study. 7 Physical Performance-Based Testing Knee extensor and hip flexor strength were measured using a hand-held dynamometer, measuring to the nearest 0.1 kg resistance, and using the average of two trials on the weaker leg. Vibratory sensation was measured with a tuning fork on the great toe of each foot. Sensation was reported as normal if the subject felt the vibration in both toes and accurately identified when it stopped and abnormal otherwise. Gait speed was measured by timing how long it took to traverse a straight 4-meter course at usual walking pace. Balance was assessed with 30-second timed stands. For this analysis, balance was dichotomized, with good balance defined as being able to perform a side-byside stand for 30 seconds, with eyes open, arms folded across chest, and feet close to each other but not touching and poor balance as being unable to do so. Falls and Fear of Falling At each visit, subjects were asked about their fall and fearof-falling status. They were asked the question Have you fallen within the past 12 months? Falling includes unintentionally coming to rest on the ground or other level such as a chair. Fear of falling status was determined with two questions. First, they were asked Apart from being in a high place, in the past 12 months, have you been worried or afraid that you might fall? If they responded positively, they were then asked, Do you ever limit your activities, for example, what you do or where you go, because you are afraid of falling? Data Analysis All 2,212 subjects who completed their follow-up evaluation were included in the analysis. The proportion reporting falls and fear of falling at follow-up were calculated according to baseline fall and fear-of-falling status and depicted graphically. Chi-square analysis was performed to determine odds ratios (ORs) of developing the two outcomes according to baseline status. Stepwise logistic regression analyses were then performed to evaluate independent predictors of falls and fear of falling at the follow-up evaluation, first predicting incident outcomes and then predicting fall or fear-of-falling status at 20 months with baseline falling and fear of falling as predictors. All predictor variables were entered into the model, with an inclusion criterion of P.05. Regressions were then rerun, using only the variables that had been significant, to maximize the number of observations that were used for each regression. Results are reported using ORs and 95% confidence intervals (CIs). RESULTS The population was 58.6% female and 25.9% African American. Subjects had a median age of 72.6 and MMSE of 28 (Table 1). Overall, 615 subjects (27.8%) reported having fallen in the previous 12 months at their baseline visit; 459 (20.8%) expressed fear of falling at baseline, of whom 212 (46.2%) had cut back on activities because of fear. Figures 1 and 2 show the risk of reporting falls and of reporting fear of falling, respectively, at the follow-up visit, based on fall and fear-of-falling status at baseline. Of those who were nonfallers at baseline, those with fear were more likely to report falls at follow-up than those without fear at baseline (32.6% (combining groups 1 and 2 with fear of falling) vs 17.9%; OR 2.22, 95% CI
4 1332 FRIEDMAN ET AL. AUGUST 2002 VOL. 50, NO. 8 JAGS Table 1. Characteristics of Salisbury Eye Evaluation Population (N 2,212) Characteristic Prevalence Demographics Age, median (range) 72.6 ( ) Female, % 58.6 African American, % 25.9 Chronic conditions Arthritis, % 52.8 History of hip fracture, % 2.5 History of stroke, % 8.1 Parkinson s disease, % 1.0 Vertigo, % 5.7 Comorbidity index, median (range) 1 (0 5) Foot pain, % 21.3 Mini-Mental State Examination score, median (range)* 28 (18 30) General Health Questionnaire score, median (range) 0 (0 24) Number of medications, median (range) 3 (0 6) Use of sedatives, % 11.1 Performance-based measures Knee strength, kg, median (range) 20.0 ( ) Hip flexor strength, kg, median (range) 16.0 ( ) Vibratory sensation normal, % 97.7 Gait speed, m/sec, median (range) 0.86 ( ) Balance poor, % 6.8 History of falls, % 27.8 Report fear of falling, % 20.8 Restrict activities because of fear of falling, % 9.6 Vision Visual acuity, logmar, median (range) 0.02 ( ) Contrast sensitivity, # letters, median (range) 35 (0 44) Stereoacuity, arcsec, median (range) 1.8 ( ) Visual fields, points missed, median (range) 17 (0 81) * Potential range of test Potential range of test ; P.0001) (Figure 1). Of those with no fear of falling at baseline, subjects who were fallers at baseline were twice as likely to report fear at follow-up than were nonfallers (20.6% vs 11.6%; OR 1.97, 95% CI Figure 2. Percentage of subjects reporting fear of falling at follow-up, by baseline reports of falls, fear of falling, and activity status. 1 Fear of falling, restrict activities. 2 Fear of falling, no restriction in activities. 3 No fear of falling ; P.0001) (Figure 2). Conversely, individuals who expressed fear of falling at baseline (Figure 2, groups 1 and 2) were very likely to express fear of falling at follow-up, regardless of falling status. On the 1,466 subjects who did not express fear of falling at baseline and for whom we had complete data for all the variables considered in the analysis, we performed a stepwise logistic regression to determine predictors for developing fear of falling at follow-up. The independent predictors for the new onset of fear were being female, older age, taking four or more medications, worse GHQ score, and being a faller at baseline (P.05 for each). Because there were missing values when all the variables were used, thereby reducing the number of subjects included in the regression, the model was run using only the five variables that were significant. In this regression (n 1,753), similar results were obtained, with each OR within 0.3 of the previous regression, and are presented in Table 2. When falls at baseline were removed from the model as a predictor, no other characteristics entered the model (data not shown). Of the entire group, 1,597 were not fallers at baseline, and similar analysis was performed to determine predictors for becoming a faller at follow-up. The independent predictors for the new onset of falling were white race, female gender, history of stroke, sedative use, and fear of falling at baseline. When the model was rerun using only the five variables that were significant (n 1,597), ORs were similar, with each OR within 0.3 of the previous regression, and all remained significant. These results are presented in Table 3. When baseline fear of falling was tri- Table 2. Predictors of Developing Fear of Falling in Individuals Without Fear of Falling at Baseline (n 1,735) Predictor Odds Ratio (95% Confidence Interval) Figure 1. Percentage of subjects reporting falls at follow-up, by baseline report of falls, fear of falling, and activity status. 1 Fear of falling, restrict activities. 2 Fear of falling, no restriction in activities. 3 No fear of falling. Age (per additional year) 1.04 ( ) Female gender 2.53 ( ) 4 medications 2.18 ( ) General Health Questionnaire score (per additional point)* 1.10 ( ) Falls at baseline 1.75 ( ) * Potential range of test 0 28.
5 JAGS AUGUST 2002 VOL. 50, NO. 8 FALLS AND FEAR OF FALLING 1333 Table 3. Predictors of Incident Falls at the 20-Month Follow-Up in Individuals Without a History of Falls at Baseline (n 1,597) Predictor Odds Ratio (95% Confidence Interval) White 1.83 ( ) Female 1.62 ( ) History of stroke 1.90 ( ) Sedative use 1.50 ( ) Fear of falling at baseline 1.79 ( ) chotomized to subjects without fear, those with fear and no restriction of activities, and those with fear and restriction of activities, race, gender, sedative use, and stroke history remained predictors, but only those subjects who cut back on activities were significantly more likely to become fallers, with an OR of 2.51 (95% CI ; P.0005). Subjects without activity restriction did not have a higher risk of becoming fallers. When fear of falling was taken out of the model as a predictor, vertigo became a predictor of new onset of falls at 20 months, with an OR of 1.66 (95% CI ; P.05) (data not shown). In the next analyses, we sought to predict reporting falls and reporting fear of falling at 20 months. This approach allowed us to model previous history of falls and fear of falling as independent predictors of both outcomes (Table 4). In addition, other predictors were included. Two predictors were common to both outcomes, namely, female gender and history of stroke. Being a faller at baseline was a modest predictor of reporting fear of falling at the follow-up visit (OR 1.58; P.001). Conversely, reporting fear of falling at baseline was an independent predictor of being a faller at follow-up (OR 1.78; P.005). When individuals who expressed fear at baseline were further divided into those who did not cut back activities and those who did, only those who did cut back on activities had a significantly higher likelihood of being a faller (OR 2.10; P.0001). DISCUSSION In this population-based prospective study of 2,212 older adults aged 65 to 84 at baseline, falls at baseline were an independent predictor for the onset of fear of falling after 20 months, and fear of falling at baseline independently predicted becoming a faller. Because each is a predictor of the other, an individual who develops one of these outcomes is at greater risk for developing the other. These data provide evidence of a spiraling effect of increasing falls, fear, and functional decline. We found that people who reported having fallen at baseline but did not report fear were still at higher risk of expressing fear of falling than were nonfallers 20 months later. This suggests that fear of falling is not just an acute outcome that results from a fall. Rather, it is likely a recognition of being at risk, both of falling and of the adverse outcomes that can result from falls. Moreover, once fear develops, especially if activities are limited, it is highly likely to persist, regardless of whether there is a fall. In fact, individuals who expressed fear of falling at baseline were more than five times as likely to express fear at follow-up, even after adjusting for falls and other predictors. Individuals who limit activities because of fear of falling are at particularly high risk of becoming fallers. This suggests, as has previously been postulated, 22 that decreasing activity may lead to a decline in function, which leads to a higher risk of subsequent falls. In addition, because individuals who cut back on activities have a higher prevalence of risk factors for falls at baseline, their physiological status alone puts them at higher risk of becoming fallers at follow-up. The U.S. Preventive Services Task Force identifies targets for prevention using the criteria of conditions that are common, associated with morbidity, in which a high-risk group can be identified, and where interventions can de- Table 4. Independent Predictors of Falls and Fear of Falling at Follow-Up, Including Falls and Fear of Falling at Baseline As Predictors Falls Predictors Fear of Falling Predictors Predictor Odds Ratio (95% Confidence Interval) Female 1.53 ( ) 2.00 ( ) History of stroke 1.61 ( ) 1.54 ( ) Parkinson s disease 4.18 ( ) Comorbidity index (per point) 1.17 ( ) White 1.56 ( ) General Health Questionnaire score (per point)* 1.07 ( ) Age (per year) 1.04 ( ) Four medications 1.68 ( ) Fear of falling at baseline 1.78 ( ) 5.40 ( ) Falls at baseline 2.51 ( ) 1.58 ( ) Note: Outcomes of falls and fear of falling modeled separately, via stepwise logistic regression, with n 2,211 for each regression. Significant odds ratios listed. Sedative use, arthritis, walking speed, contrast sensitivity, visual acuity, visual fields, vibratory sensation, vertigo, and Mini-Mental State Examination were entered into the model but were not significant predictors. * Potential range of test 0 28.
6 1334 FRIEDMAN ET AL. AUGUST 2002 VOL. 50, NO. 8 JAGS crease the occurrence of the target condition. 23 Falls and fear of falling potentially fulfill these criteria as targets for prevention. They are common syndromes among older adults living in the community and have several potentially serious adverse consequences, including functional decline, 10,24 26 decreased quality of life, 10,27 and increased institutionalization. 10,28 Much work has been done to identify those at high risk of falling 7,9,13,15,18,29 and to develop interventions to reduce the incidence of falls, but the literature provides less information on how to target those at high risk of developing fear of falling and on strategies for reducing the incidence of fear of falling. In addition, the temporal relationship between falls and fear of falling has not previously been well characterized in a population-based sample. This paper examines those temporal relationships and identifies risk factors common to both outcomes. By identifying those at highest risk, individuals can then be targeted for prevention. We have identified predictors that are useful for primary prevention strategies in individuals who have not yet developed the target condition. Of the predictors for becoming a faller (white race, female gender, history of stroke, sedative use, and fear of falling), only sedative use and fear of falling are potentially modifiable. Based on these predictors, once an individual is identified as being at high risk of becoming a faller, interventions could be put into place to reduce fear of falling 34,35 and to eliminate sedative use. Also, because vertigo is a risk factor for the onset of falls when fear was removed from the model, treatment of this condition when possible may also be beneficial. The predictors of developing fear of falling (older age, female sex, 4 medications, worse GHQ score, history of falls) have more-obvious implications for primary prevention strategies. Efforts to reduce the number of medications, evaluation and treatment for psychiatric disturbances, and efforts to reduce further fall risk may in turn reduce the risk of developing fear of falling. The regressions to evaluate risk of falls and fear of falling at follow-up (Table 4) show shared risk factors that help to identify, prospectively, individuals who are at highest risk of these two outcomes. Women with a history of stroke have a higher risk of each outcome and might therefore be particularly important to target for prevention strategies. In addition, Parkinson s disease, comorbidity, and white race predict falls at follow-up, and higher GHQ score, age, and being on four or more medications predict fear of falling at follow-up. These risk factors have implications for secondary prevention. The approach of multiple risk factor intervention has been used successfully for fall prevention 32 and may be appropriate for preventing fear of falling. In this population-based study of older adults, 27.8% reported a history of falls in the previous year. These results are similar to several previous community-based studies of falls that have shown that approximately one-third of community-dwelling older adults fall each year, 7,18,29 but the frequency of falls may have been underestimated with our method of ascertainment. It has previously been shown 36 that falls may be forgotten when asked about retrospectively, even in the setting of a study specifically evaluating falls. The prevalence of fear of falling at baseline was 20.8%, which is somewhat lower than the range of 29% to 43% seen in the few other population-based samples that have addressed fear of falling. 5,6,12 There may be several explanations for this finding. First, the age of our study population spanned two decades and did not include individuals aged 85 and older. Because fear of falling is more prevalent with increasing age, this may have had an effect on our findings relative to previously reported populations. 5 Second, the way the question was worded ( apart from being in a high place... ) may have identified true fear of falling rather than including individuals with acrophobia. Finally, the prevalence of fear was determined for all subjects, not just those who had fallen previously. 12 Previous studies using volunteers showed an even higher prevalence of fear of falling. 1,2,4 Fear of falling was assessed as a dichotomous response to a single question. This approach to determining fear of falling has been shown to be less sensitive than other measures, such as the FES 5,37,38 or the Activities-Specific Balance Confidence Scale. 37 As a result, the associations may have appeared less robust than they would have if a more detailed scale had been used. Nevertheless, the fact that these relationships remained significant despite a relatively insensitive measure suggests that the associations are important. From a clinical standpoint, asking a single yes/no question, such as are you afraid of falling? or have you fallen within the past year? is a useful starting point for the assessment of risk. In summary, falls and fear of falling are common, serious, potentially preventable conditions in older community-dwelling adults. Because each is a risk factor for the other, an individual who has one of these factors is at risk for developing the other. This in turn may set off a vicious cycle of falls, fear of falling, and the many adverse outcomes that can result, such as functional decline, a decrease in quality of life, and institutionalization. Further studies are needed to determine whether efforts at primary prevention based on the risk factors for developing falls and fear of falling, as well as secondary prevention efforts for fear of falling, can be useful in reducing these adverse outcomes. ACKNOWLEDGMENTS The authors would like to thank Brian Heppard, Robert McCann, Annie Medina-Walpole, Stephen Ryan, and Paula Taylor for their input on earlier drafts of this article. REFERENCES 1. Howland J, Lachman ME, Peterson EW et al. Covariates of fear of falling and associated activity curtailment. Gerontologist 1998;38: Howland J, Walker Peterson E, Levin WC et al. Fear of falling among community-dwelling elderly. J Aging Health 1993;5: Lachman ME, Howland J, Tennestedt S et al. Fear of falling and activity restriction: The survey of activities and fear of falling in the elderly (SAFE). J Gerontol B Psychol Sci Soc Sci 1998;53B:P43 P Fessel KD, Nevitt MC. Correlates of fear of falling and activity limitation among persons with rheumatoid arthritis. Arthritis Care Res 1997;10: Tinetti ME, Mendes de Leon CF, Doucette JT et al. Fear of falling and fallrelated efficacy in relationship to functioning among community-living elders. J Gerontol 1994;49:M140 M Arfken CL, Lach HW, Birge SJ et al. The prevalence and correlates of fear of falling in elderly persons living in the community. Am J Public Health 1994; 84:
7 JAGS AUGUST 2002 VOL. 50, NO. 8 FALLS AND FEAR OF FALLING Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among elderly persons living in the community. N Engl J Med 1988;319: Maki BE. Gait changes in older adults: Predictors of falls or indicators of fear? J Am Geriatr Soc 1997;45: Nevitt MC, Cummings SR, Kidd S et al. Risk factors for recurrent nonsyncopal falls. JAMA 1989;261: Cumming RG, Salkeld G, Thomas M et al. Prospective study of the impact of fear of falling on activities of daily living, SF-36 scores, and nursing home admission. J Gerontol A Biol Sci Med Sci 2000;55A:M299 M Luukinen H, Koski K, Kivela SL et al. Social status, life changes, housing conditions, health, functional abilities and life-style as risk factors for recurrent falls among the home-dwelling elderly. Public Health 1996;110: Vellas BJ, Wayne SJ, Romero LJ et al. Fear of falling and restriction of mobility in elderly fallers. Age Ageing 1997;26: Campbell AJ, Borrie MJ, Spears GF. Risk factors for falls in a communitybased prospective study of people 70 years and older. J Gerontol 1989;44: M112 M Tinetti ME, Inouye SK, Gill TM et al. Shared risk factors for falls, incontinence, and functional dependence. Unifying the approach to geriatric syndromes. JAMA 1995;273: Lord SR, Clark RD, Webster IW. Physiological factors associated with falls in an elderly population. J Am Geriatr Soc 1991;39: West SK, Munoz B, Rubin GS et al. Function and visual impairment in a population-based study of older adults. Invest Ophthalmol Vis Sci 1997;38: Rubin GS, West SK, Munoz B et al. A comprehensive assessment of visual impairment in a population of older Americans. Invest Ophthalmol Vis Sci 1997;38: Prudham D, Grimley Evans J. Factors associated with falls in the elderly: A community study. Age Ageing 1981;10: Dunn JE, Rudberg MA, Furner SE et al. Mortality, disability, and falls in older persons: The role of underlying disease and disability. Am J Public Health 1992;82: Folstein MF, Folstein SE, McHugh PR. Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12: Goldberg DP, Hillier VF. A scaled version of the General Health Questionnaire. Psychol Med 1979;9: King MB, Tinetti ME. Falls in community-dwelling older persons. J Am Geriatr Soc 1995;43: Department of Health and Human Services. Report of the U.S. Preventive Services Task Force, Guide to Clinical Preventive Services, 2nd Ed. Washington, DC: Department of Health and Human Services, Franzoni S, Rozzini R, Boffelli S et al. Fear of falling in nursing home patients. Gerontology 1994;40: Kiel DP, O Sullivan P, Teno JM et al. Healthcare utilization and functional status in the aged following a fall. Med Care 1991;29: Mendes de Leon CF, Seeman TE, Baker DI et al. Self-efficacy, physical decline, and change in functioning in community-living elders: A prospective study. J Gerontol B Psychol Sci Soc Sci 1996;51B:S183 S Salkeld G, Cameron ID, Cumming RG et al. Quality of life related to fear of falling and hip fracture in older women: A time trade off study. BMJ 2000; 320: Wolinski FD, Johnson RJ, Fitzgerald JF. Falling, health status and the use of health services by older adults. Med Care 1992;30: Graafmans WC, Ooms ME, Hofstee HMA et al. Falls in the elderly: A prospective study of risk factors and risk profiles. Am J Epidemiol 1996;143: Lord SR, Ward JA, Williams P et al. The effect of a 12-month exercise trial on balance, strength, and falls in older women: A randomized controlled trial. J Am Geriatr Soc 1995;43: Province MA, Hadley EC, Hornbrook MC et al. The effects of exercise on falls in elderly patients: A preplanned meta-analysis of the FICSIT trials. JAMA 1995;273: Tinetti ME, Baker DI, McAvay G et al. A multifactorial intervention to reduce the risk of falling among elderly people living in the community. N Engl J Med 1994;331: Wolf SL, Barnhart HX, Kutner NG et al. Reducing frailty and falls in older persons: An investigation of Tai Chi and computerized balance training. J Am Geriatr Soc 1996;44: Tennestedt S, Howland J, Lachman M et al. A randomized, controlled trial of a group intervention to reduce fear of falling and associated activity restriction in older adults. J Gerontol B Psychol Sci Soc Sci 1998;53B:P384 P 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: Cummings SR, Nevitt MC, Kidd S. Forgetting falls: The limited accuracy of recall of falls in the elderly. J Am Geriatr Soc 1988;36: Myers AM, Powell LE, Maki BE et al. Psychological indicators of balance confidence: Relationship to actual and perceived abilities. J Gerontol A Biol Sci Med Sci 1996;51A:M37 M Tinetti ME, Richman D, Powell L. Falls efficacy as a measure of fear of falling. J Gerontol 1990;45:P239 P243.