Literature Evaluation Table

Characteristics and fall experiences of older adults with and without fear of falling outdoors

Tracy Chippendale and Chang Dae Lee

Department of Occupational Therapy, New York University, Steinhardt School of Culture, Education, and Human Development, New York, NY, USA

ARTICLE HISTORY Received 15 November 2016 Accepted 17 March 2017

ABSTRACT Objective: Using a theoretical model that combines an ecological perspective and Bandura's theory of self-efficacy as a guide, we sought to compare experiences and characteristics of community dwelling older adults with and without concern about falling outdoors. Method: A survey of randomly selected community dwelling older adults across NYC (N = 120) was conducted using the outdoor falls questionnaire. Descriptive quantitative analyses of participant characteristics were conducted for all participants and for those with and without concern about falling outside. Conventional content analysis using two coders was employed to examine outdoor fall experiences for each group. A mixed methods matrix was used to integrate qualitative and quantitative findings. Results: Some participant characteristics were more common among those with a concern about falling outside such as decreased functional status, female gender, and number of prior outdoor falls. As per descriptions of outdoor fall experiences, participants with concern were more likely to report a fall while climbing stairs or stepping up a curb, describe an intrinsic factor as a cause of their fall, use an injury prevention strategy during the fall, sustain a moderate to severe injury, seek medical attention, have had an ambulance called, require help to get up, and describe implementation of a behavioral change after the fall. Conclusions: Differences exist in participant characteristics and outdoor fall experiences of those with and without concern about falling outside. The proposed model can be used to understand fear of falling outdoors and can help to inform the target population and content of intervention programs.

KEYWORDS Fear of falling; outdoor falls; mixed methods

Introduction

Falls are a significant public health concern due to their high prevalence rates and serious consequences for the health and well-being of older adults (Center for Disease Control, 2016). Although risk of falls increases with age, there is evidence to suggest adults in late mid-life are also at risk (Verma, Willetts, Marucci-Wellman, Lombardi, & Courtney, 2016). Outdoor falls are just as common as indoor falls and are just as likely to result in serious injuries. However, they are associated with different risk factors. Examples of risk factors for indoor falls include slow gait speed, impaired balance and strength, impaired cognition, multiple medication, and limitation in activities of daily living to name a few. Conversely, male gen- der, younger age, having a fast gait speed, and participating in more leisure time physical activity are associated with risk for outdoor falls (Kelsey, Proctor-Gray, Hannan, & Li, 2012; Li et al., 2006). Thus, characteristics and profiles of people at risk for outdoor falls differ significantly from those at risk for indoor falls. Fear of falling (FOF), one of the fall-related psy- chological concerns (Hull, Kneebone, & Farquharson, 2013), can also increase risk for falls (Friedman, Munoz, West, Rubin, & Fried, 2002). Noteworthy is that older adults may experience a FOF regardless of whether or not they have experienced a fall (Denkinger, Lukas, Nikolaus, & Hauer, 2015). Prevalence rates for FOF among community dwelling older adults are var- ied, but have been found to be between 33% and 46% among older adults who have not sustained a fall and

between 21% and 85% among older adults who have fallen (Kumar, Carpenter, Morris, Iliffe, & Kendrick, 2014).

In addition to an increased risk for falls, fear or concern about falling can result in a number of other negative conse- quences that have both immediate and long term effects. Immediate effects may include activity restriction (Deshpande et al., 2008), and gait changes (Chamberlin, Fulwider, Sanders, & Medeiros, 2005; Reelick, Iersel, Kessels, Marcel, & Rikkert, 2009) whereas long term effects include a decline in function and social isolation (Deshpande et al., 2008; Liu, 2014). Given the serious physical, psychological, and social consequences, studies that explore factors that contribute to FOF are war- ranted to help inform interventions and clinical practice. Some factors associated with FOF in general (indoors and out- doors) have been found consistently across studies such as female gender, impaired physical function, and use of a walk- ing aid (Denkinger et al., 2015). Less common or consistent factors include visual impairments/self-perception of vision (Donoghue et al., 2014; Liu, 2014), declining muscle mass and strength (Oh, Hong, Lee, & Han, 2015; Trombetti et al., 2016), use of multiple medications, (Kumar et al., 2014), the neigh- borhood environment (Chippendale & Boltz, 2014), hearing impairment (Malini, Lourenco, & Lopes 2015), education level (Kumar et al., 2014), and mental health (e.g. depression and anxiety) (Hull et al., 2013; Liu, 2014; Malini et al., 2016). In sum- mary, impaired physical function can be both the cause and consequence of FOF.

CONTACT Tracy Chippendale [email protected]

© 2017 Informa UK Limited, trading as Taylor & Francis Group

AGING & MENTAL HEALTH, 2018 VOL. 22, NO. 6, 849–855 https://doi.org/10.1080/13607863.2017.1309639

Although factors associated with FOF have been explored for all fall locations, examination of factors associated with con- cern about falling outdoors specifically have not been studied, and may be different given the differences in profiles/charac- teristics of indoor and outdoor fallers. Furthermore, outdoor fall experiences of older adults with and without concern about falling have yet to be examined, and may help to inform clini- cal practice. With the rapidly growing older adult population and trend towards urban living, the fit between communities and the changing motor abilities of older adults is increasingly important. To our knowledge, there is no theoretical model that has been developed to explain FOF. Bandura's theory of self-efficacy has been proposed in previous studies (Jorstad, Hauer, Becker, & Lamb, 2005). However, we posit that the eco- logical perspective (National Cancer Institute, 2005) combined with Bandura's theory (Bandura, 1977) can best be used to guide studies on FOF given multiple sources/levels of influence, including intrinsic factors (Denkinger et al., 2015), the physical and social environment (Chippendale & Boltz, 2014), and psy- chological factors. This combined model promotes a more holistic and comprehensive conceptualization of FOF. The eco- logical perspective, which we propose as the overarching framework for understanding FOF, suggests that health prob- lems and behaviors are affected by intrapersonal/individual (e.g. genetic/constitutional factors, knowledge and beliefs), interpersonal (e.g. peers, family members), and community level factors (e.g. policy, neighborhood features). Furthermore, people are shaped by and can shape the social environment. Bandura asserts that expectations of personal efficacy are based on information from performance accomplishments, vicarious experience, verbal persuasion, and emotional arousal, but that personal accomplishments are more influential than vicarious experiences (Bandura, 1977). Bandura's theory can help to explain and unpack the role of intrapersonal level fac- tors (i.e. vicarious experience of family/friends regarding out- door falls) and some interpersonal factors, specifically beliefs about personal efficacy pertaining to outdoor falls. Bandura's theory suggests inclusion of the role of personal experiences and successes regarding outdoor falls, depressive symptoms, and fall experiences of family/close friends (in this order of importance) in considering factors associated with FOF out- doors (see Diagram 1). Our proposed model reflects the

hierarchy of importance of factors associated with self-efficacy for outdoor activity with the most important factor listed at the top of the circle. The conceptual model also highlights and visually represents the idea that the ecological perspective (large circle in the diagram) is the overarching framework for understanding FOF outdoors and that Bandura's theory of self- efficacy contributes in part to our understanding of this health- related concern.

Given identified gaps in knowledge about FOF outdoors, we used Bandura's theory and the ecological perspective as a guiding framework (see Diagram 1) to examine differences in characteristics and fall experiences of those with and without concern about falling outdoors. Based on assertions from this combined framework, we hypothesized that (1) the neighbor- hood environment will impact concern about falling outdoors; (2) participants with concern about outdoor falls will identity more potential hazards and use more strategies for outdoor fall prevention; (3) fear or concern about falling outdoors will be more common among participants who are female, older, have greater depressive symptoms, and functional impairment; and (4) personal experience of outdoor falls will have greater relevance that vicarious experiences of friends/family.

Methods

Design

A survey using random digit telephone dialing was conducted across the five boroughs of New York City. In order to have a representative sample, in our case of community dwelling adults 55 and older residing in NYC, a sample size of 100 partic- ipants was needed for a sampling error of +/¡ 10% with a 95% CI (Dillman, 2000). Random numbers were generated for a combination of landlines and cell phones, and a web-based application with screening for business and non-working num- bers was used to create the sample of random numbers. The study was approved by the university's IRB and the survey was conducted by the PI and by trained research assistants. An IRB approved script, which included a description of the study pur- pose, time commitments, incentives, etc., was used to present the study details. Qualitative and quantitative data were ana- lyzed separately and a mixed methods matrix was used to interpret and integrate the findings (O'Cathain, Murphy, & Nich- oll, 2010). However, results of quantitative and qualitative anal- ysis are reported in separate tables for clarity of presentation.

Participants

Inclusion criteria were: age 55 and older, community dwelling (not living in a long term care facility), English-speaking, able to ambulate outdoors independently with or without an assis- tive device, and residing in NYC. Given that socioeconomic inequities can cause some people to age at a faster rate (Hos- pers, Smulders, Maier, Deeg, & Muller, 2015; Sood et al., 2015), we chose to include adults 55–64 years of age to account for biological not just chronological age. Furthermore, there is mounting evidence that injurious falls are also problematic for adults in late mid-life (Verma et al., 2016).

Measures

The outdoor falls questionnaire (OFQ) was used as the survey instrument. It includes a series of Likert style questions

ECOLOGICAL CONSIDERATIONS

Neighborhood Environment

Knowledge of outdoor risk factors

Intrinsic factors (e.g. age, gender,

func�onal status)

SELF-EFFICACY

Personal experience of outdoor falls

+

Depressive symptoms

+

Experiences of family/friends re: outdoor falls

FEA R O

F FA LLIN

G

Diagram 1. Framework for conceptualizing factors associated with fear of fall- ing outdoors.

850 T. CHIPPENDALE AND C. D. LEE

pertaining to demographic characteristics associated with out- door fall risk (e.g. gender, age, neighborhood of residence, race/ethnicity, education level, type and amount of physical activity, use of an assistive device, and work status), ADLs/IADLs requiring assistance, concern about falls during outdoor activ- ity, depressive symptoms (as per the PHQ-2), personal fall his- tory and that of friends/family members, perceived outdoor fall risks, and strategies used in outdoor fall prevention. The final question is an open ended question about most recent out- door fall. More specifically, there is one multiple choice ques- tion about ADLs/IADLs requiring assistance, where participants can choose from a list of options (bathing, dressing, shopping, cooking, and laundry). Fall history is obtained by asking how many times they have fallen indoors and outdoors on two Lik- ert scale questions, as well as one question with a yes/no response about whether a spouse, partner, or close friend has fallen outdoors. Concern about falling outdoors is measured using two questions from the falls efficacy scale international (Yardley et al., 2005). The first asks about level of concern when walking around the neighborhood and the second, about level of concern going to the store. Responses are recorded on a 4-point Likert scale ranging from ‘Not at all con- cerned’ (score = 0) to ‘Very concerned’ (score = 3). Additional subscales or domains of the OFQ include perceived outdoor fall risks (e.g. parking lots/garages, stairs rated as ‘does not affect my risk’, ‘increases my risk a little’, or ‘definitely increases my risk’) and strategies used for outdoor fall prevention, where participants are asked to respond ‘I don't use’ (score = 0), ‘I use sometimes’ (score = 1), or ‘I use most of the time’(score = 2) to a number of items such as holding rails on stairs, avoiding cell phone use while walking, etc. The final open ended question asks participants ‘Please describe what happened during your most recent outdoor fall’. Examples of follow up probes include ‘Why do you think you fell’ and ‘Did you injure yourself’. The OFQ has been shown to have good face and content validity, and as per analysis of data from this survey, internal consis- tency ranges from 0.7 to 0.9 for each domain. Internal consis- tency values (i.e. Chronbach's alphas) for this study are consistent with values obtained in a prior pilot study. Details about the development of the OFQ and additional examples of questions were published previously (Chippendale, 2015).

Analysis

Descriptive analyses of participant characteristics were con- ducted using SPSS version 21 for the entire sample and then separately for two groups, i.e. those with (score > 0) and with- out concern (score = 0) about falling outdoors. Variables cho- sen for quantitative analysis were informed by past study findings as well as the ecological perspective and Bandura's theory of self-efficacy. Therefore, at the individual level, we included age, gender, use of an assistive device, assistance needed with ADL/IADL, education level, depressive symp- toms, perceived outdoor fall risks, strategies used for preven- tion, and fall history. At the intrapersonal level, and consistent with Bandura's concept of vicarious experience, we consid- ered whether or not the participant had a close friend or fam- ily member experience a fall. At the community level, we examined neighborhood of residence and description of fall experiences, including those that were caused by environ- mental factors. To assess representativeness of the sample, we compared demographic characteristics for the sample with population level data for older adults in NYC. We also

examined differences between groups, i.e. those with and without concern about falling outdoors, for quantitative varia- bles using chi-square tests of independence and independent samples t-tests as appropriate at a 0.05 significance level. According to a power analysis using a medium effect with power set at 0.8, 51 participants were needed in each group in order to detect significant differences between those with and without concern about falling outdoors.

Qualitative analysis was conducted using conventional content analysis (Hsieh & Shannon, 2005). Each transcription of participant's description of their most recent fall was read through from beginning to end. Then, each was read through carefully and significant statements pertaining to their out- door fall experience were highlighted. A key word or phrase that captured the experience or description of the fall was assigned to each significant statement. After open coding of several transcripts, preliminary codes were decided on. The remainder of the transcripts was read through and new codes added as needed if the existing codes did not fit the data. In order to enhance the reliability of the analysis and to reduce bias, two coders independently read through the transcripts, extracted significant statements, and coded the statements. Discussion until consensus occurred between the two coders with regard to significant statements and their assigned codes. Following completion of coding, each coder grouped existing codes into themes for each group of participants (i.e. those with and those without FOF) as appropriate. Once again, discussion until consensus occurred between the two coders with regard to the themes for each group of partici- pants. We also incorporated counting of codes to ascertain if certain fall experiences were more common among one group as compared to the other. We coded injury severity as follows: minor injuries were those that entailed scrapes, bruises or temporary soreness. Moderate to severe injuries were those that resulted in prolonger pain or time for recov- ery, sprains, stitches, fractures, or surgery. We chose to exam- ine moderate and severe injuries together given that thy both negatively impact function and quality of life.

A mixed methods methodology allowed for study of the FOF outdoors using different methods in order to gain a more complete picture (O'Cathain et al., 2010). Using a mixed meth- ods matrix, the investigators simultaneously examined and interpreted findings from both qualitative and quantitative analysis. Specifically, hypothesis 1 was examined by compar- ing neighborhood of residence (urban versus suburban) using chi-square tests of independence for those with and those without concern about falling outdoors. Hypothesis 1 was also examined using qualitative analysis regarding identified environmental causes of the outdoor fall. For hypothesis 2, we examined differences in number of identified risks and strategies used for prevention using independent samples t- tests, a statistical test that was also used to compare partici- pant characteristics consistent with hypothesis 3 (i.e. age, gender, depressive symptoms, level of function). The mixed matrix provides a more complete picture of participant's level of function. Finally, we compared effect sizes for differences between groups in number of outdoor falls versus having had a friend or family member fall outdoors to test hypothesis 4.

Results

A total of N = 120 older adults completed the survey. Among people for whom we could determine eligibility, 58% agreed

AGING & MENTAL HEALTH 851

to participate. Reasons given for non-participation were not being willing to share information over the phone or being too busy. Descriptive statistics for the entire sample are pre- sented in Table 1. Similarities are noted between demo- graphic characteristics for our sample and for older adults in NYC as a whole. We had 45% males and 55% females in our sample which is reasonably consistent with census data for NYC (41.2% male and 58.8% female). Although we had a diverse sample, as compared to NYC population data, white participants were slightly over-represented (50% versus 43.5% for NYC) and Asians underrepresented (2.5% versus 11.7% for NYC as a whole). Percentage of African American participants (22.5%) was similar to the older adult population in NYC (21.8%) (NYC Department for the Aging, 2016). Note- worthy is that our sample consisted of people age 55 plus whereas statistics for NYC are based on age 60 and older.

In relation to hypothesis 1, which stated that the neighbor- hood environment would affect FOF, we find a non-significant difference in neighborhood of residence between those with and without concern about falling outdoors (p = 0.5). Number of perceived risks and strategies used for prevention were dif- ferent between groups (ps < 0.001), which is consistent with our proposed model and second hypothesis that stated that participants with concern about falling outdoors would iden- tify more hazards and use more strategies. In relation to hypothesis 3 (i.e. concern would be more common among women, older participants, those with depressive symptoms, and those with functional impairments), there were significant differences between groups with regard to gender and age (ps = 0.01). In addition to differences in participant demo- graphic characteristics (age and gender), significant

differences were also found between the two groups with regard to use of an assistive device, an indication of functional status which also supports hypothesis 3. Differences in depressive symptoms between the two groups approached significance (p = 0.06). In support of hypothesis 4, effect size was larger for differences in number of outdoor falls (Cohen's d = 0.5) as compared to having had a friend or family member experience a fall outside (Cramer's v = 0.1).

Table 2 presents the results of the analysis for qualitative data. Italicized font was used to indicate subthemes that were unique to one participant group. In addition, some themes and subthemes were much more common among one group as compared to the other, with percentages of the sample reporting each theme/subtheme noted in parenthesis in the table. In relation to hypothesis 1, which states that the neigh- borhood environment will impact concern about falling, we see that falls on curbs and stairs are much more common among those with a FOF. For hypothesis 2, qualitative responses indicate that behavioral strategies were more com- monly used post fall among those with concern about falling outdoors. Consistent with the quantitative results, qualitative findings also support hypothesis 3 regarding functional status of participants, specifically, only those with concern about outdoor falls reported intrinsic factors as the cause of their fall.

Discussion

Prevalence of FOF outdoors was 40% among those who had not sustained a fall outdoors and 58.8% among those who had. This is consistent with the rates from prior studies, which have been reported in the 21%–46% range for community-

Table 1. Participant characteristics.

Total sample Concern No concern N = 120 N = 56 N = 64 p

Age M(SD) 66.7 (8.1) 68.6 (8.4) 64.5 (7.1) 0.01 Gender N (%) 0.01 Male 54 (45) 22 (34.4) 32 (57.1) Female 66 (55) 42 (65.6) 24 (42.9)

Race/ethnicity N (%) 0.9 Caucasian 61 (50.8) 31 (48.5) 30 (53.6) Asian 3 (2.5) 1 (1.6) 2 (3.6) African American 27 (22.5) 16 (25) 11 (19.6) Caribbean American 7 (5.8) 4 (6.3) 3 (5.4) Hispanic 8 (6.7) 5 (7.8) 3 (5.4) Other/mixed race 14 (11.7) 7 (10.9) 7 (12.5)

Neighborhood N (%) 0.5 Urban 85 (70.8) 42 (65.6) 43 (76.8) Suburban 32 (26.7) 20 (31.3) 12 (21.4) Other 3 (2.5) 2 (3.2) 1 (1.8)

Education N (%) 0.2 < High school 6 (5) 4 (6.3) 2 (3.6) High school 20 (16.7) 11 (17.2) 9 (16.1) Some college 39 (32.5) 26 (40.6) 13 (23.2) College degree 30 (25) 13 (20.3) 17 (30.4) Graduate degree 25 (20.8) 10 (15.6) 15 (26.8) College graduate 45.8% 35.9% 57.2% 0.02

Number of ADLs/IADLs requiring assistance M(SD) 0.2 (0.7) 0.3 (0.9) 0.1 (0.4) 0.1 Walking as primary source of PA 75.8% 79.7% 71.4% 0.3 Daily PA participation 43.3% 35.9% 51.8% 0.08 Use of an assistive device 15% 25% 3.6% 0.02 Perceived risks (OFQ subscale) M(SD) 14.4 (8.9) 18.8 (7.9) 9.7 (7.5) <0.001 Strategies used (OFQ subscale) M(SD) 21.8 (5.8) 23.8 (5.5) 19.6 (5.5) <0.001 Depressive symptoms (PHQ-2) M (SD) 0.9 (1.1) 1.0 (1.2) 0.6 (1.0) 0.06 Friend/family member fell outdoors 61.7% 62.5% 60.7% 0.6 Number of outdoor falls 0.01 Never 28.3% 20.3% 37.5% Once 23.3% 23.4% 23.2% 2-3 times 35.5% 37.5% 33.9% 4 or more 12.4% 18.8% 5.4%

Note: p values as per chi-square tests of independence for categorical variables and independent samples t-tests for continuous variables.

852 T. CHIPPENDALE AND C. D. LEE

dwelling seniors without a history of falls, and 21%–85% among fallers (Kumar et al., 2014). Also consistent with previ- ous studies are our findings regarding female gender, use of an assistive device, fall history, and depressive symptoms which were more common among those with concern about falling outdoors. Prior studies have examined characteristics of people who have concern about falling in general, but not about outdoors falls specifically. Our findings reveal consis- tencies with these previous studies. However, we add new knowledge regarding additional factors that are consistent with outdoor activity and our proposed framework incorpo- rating the ecological perspective and Bandura's theory of self- efficacy, e.g. the neighborhood environment and vicarious experience of family/friends regarding outdoor falls.

The impact of the neighborhood environment on FOF has been noted in a previous study (Chippendale & Boltz, 2014) and is consistent with a community level factor according to the proposed framework (see Diagram 1). In the current study, although not statistically significant, suburban rather than urban place of residence was found to be more common among those with a concern about falling, which supports our initial hypothesis 1, i.e. that the neighborhood environ- ment will have an impact FOF. This may be due to characteris- tics of the physical/social environment (e.g. fewer sidewalks, poorer street/sidewalk maintenance) or characteristics of the people who reside in these settings. Walkability audits for NYC reveal lower levels of walkability in Queens and Staten Island, boroughs that have more suburban communities (NYC Department of Health and Mental Hygiene, 2014). In future studies, collection of specific walkability data is warranted. Noteworthy is that there is also a growing body of evidence that suggests differences between urban, suburban, and rural residents with regard to health behaviors. Urban dwellers have been found to have riskier profiles of health behavior (House et al., 2000; Robinson, Wyatt, Dubbert, May, & Sims, 2016), which may translate into less concern about safety risks including falls. Risky behaviors are consistent with the intra- personal (e.g. beliefs) and interpersonal (e.g. effects of peers)

levels of influence on health in the proposed framework. Also noteworthy, and in support of hypothesis 1, is the differences in frequency of qualitative reports of some environmental fall hazards. For example, those with concern about falls outdoors were more likely to identify steps or curbs as the cause of their fall.

Participants who had concern about falling had a higher OFQ subscale score for strategies used to prevent falls. This is consistent with the qualitative findings with regard to descrip- tion of use of behavioral strategies post-fall. This consistency between qualitative and quantitative data provides triangula- tion of findings. Identification of a greater number of per- ceived risks and use of prevention strategies among those with concern about falling outdoors makes logical sense and is an expected finding consistent with our proposed frame- work and hypothesis 2 (see Diagram 1). Knowledge, in this case of outdoor fall risks, is an important intrapersonal level factor associated with health behaviors, i.e. use of outdoor fall prevention strategies. Examples of strategies commonly used by participants were visual scanning, use of rails on stairs, and choice of footwear. Given these results, teaching behavioral strategies and action planning for outdoor fall prevention may be most warranted among those at high risk for outdoor falls, but with low levels of concern about falling outside. In the future, an investigation of the level of concern associated with activity restriction and gait changes is warranted. Although high levels of fear and concern can have negative consequences (Deshpande et al., 2008; Liu, 2014), some con- cern may be productive and result in older adults being more alert to environmental hazards and lead to increased use of fall prevention strategies (Twibell, Siela, Sproat, & Coers, 2015). Given that prior studies have focused on hospitalized older adults (Twibell et al., 2015), further study is warranted among community-dwelling well-elders to establish the link between level of concern about falling outdoors and safe strategy use.

Integrating the findings from qualitative and quantitative analysis, we can conjecture that physical abilities were more

Table 2. Outdoor fall experiences comparing participants with and without concern about falling outdoors.

Qualitative themes and subthemes Qualitative themes and subthemes Concern N = 64 No concern N = 56

Context Context With others (8%)/alone With others (9%)/alone Temporal (time of day/year) Temporal (time of day/year) At work (4%) At work (3%)

Cause Cause Activity-related, e.g. opening/closing a door (8%), running, climbing stairs or a curb (18%)

Activity-related, e.g. opening or closing a door (6%), running, climbing stairs or stepping up a curb (8.5%)

Environmental conditions only (e.g. wet, slippery, or uneven surface) (38%)

Environmental conditions alone (e.g. wet, slippery, or uneven surface) (37%)

Behavioral only (e.g. rushing, not paying attention, choice of footwear) (2%)

Behavioral cause alone (e.g. rushing, not paying attention, choice of footwear) (14%)

Combination of factors (e.g. environmental & behavioral) (38%) Combination of factors (e.g. environmental & behavioral) (37%) Intrinsic factor (loss of balance, brain ‘foggy’, knee buckled) (6%) Hit (by car, bike, or cart) (3%)

Mechanics Mechanics Fell forward/backward/sideways (34%/18%/8%) Fell forward/backward/sideways (31%/20%/11.4%) Holding rail (4%) Use of injury prevention strategy during the fall (to protect self or others) (10%)

Sequela Sequela Injuries: mild (30%), moderate to severe (40%) Injuries: mild (40%), moderate to severe (25.7%) Medical attention (hospital, clinic, primary care doctor) (32%) Medical attention (hospital, clinic, primary care doctor) (17%) Ambulance called (12%) Help to get up (14%) Help to get up (5.7%) Behavioral response (e.g. avoid certain areas, use of caution) (26%) Behavioral response (e.g. avoid certain areas, use of caution) (8.6%) Emotional response (6%) Emotional Response (3%)

Note: Italicized font indicates themes and subthemes unique to this group. Percentages indicate the percent of the sample reporting this theme or subtheme.

AGING & MENTAL HEALTH 853

impaired in the group with concern about falling, which sup- ports our initial hypothesis 3. This conjecture is supported by the fact that more participants in the group with concern about falling were using assistive devices for walking, which indicates probable impairments in balance and/or physical function (Bateni & Maki, 2005). Furthermore, only the group with concern about falling reported they fell due to an intrin- sic factor (e.g. loss of balance, knee buckled). Potential rea- sons why participants with concern about falling were more likely to report having had a fall while climbing stairs or step- ping up a curb is that this activity requires more physical demands than other daily activities such as walking on level surfaces. This activity, however, is inevitable in daily life, it is not like other activities (e.g. running, carrying boxes, and engaging in vigorous physical activities) which people can avoid if necessary or for which they can enlist the help of others. Another possible explanation is that a fall from heights may be more fear inducing than a fall on a level surface due to the fall experience itself or the severity of injuries that can ensue. Our results are consistent with past findings as func- tional status, an intrapersonal factor in the proposed frame- work, has been associated with FOF in a number of prior studies (Denkinger et al., 2015).

Consistent with Bandura's theory of self-efficacy, vicarious experience, in this case a fall among a close friend or family member, appeared to be less relevant than the participant's personal experience of outdoor falls (Cramer's v = 0.1 a small effect for friend/family member fell versus Cohen's d = 0.5 a medium effect for number of outdoor falls). These results sup- port our initial hypothesis 4. Noteworthy are the differences between those with and without concern about falling out- doors with regard to number of previous outdoor falls. Similar percentages were found between groups with regard to having had one fall, but multiple falls were more common in the group with concern about falling outdoors. According to Bandura's theory, performance accomplishments and emotional arousal are important factors associated with self-efficacy, therefore it is not surprising that those who had poor performance in the past as reflected in multiple outdoor falls would express greater levels of concern regarding outdoor falls. Furthermore, those with concern about falling outdoors had a greater mean score for depressive symptoms, and the difference between groups approached significance (p = 0.06, d = 0.4).

A comparison of fall experiences of those with and without concern about falling outdoors has not been explored in pre- vious studies. Our findings reveal that those who attribute their fall to an intrinsic factor, sustain a moderate to severe injury, were struck or hit (by a bike, car, or cart), required any medical attention, had an ambulance called, fell on a curb or stairs, and required help to get up were more likely to report concern about outdoor falls. Those without concern were more likely to attribute the cause of the fall to their own behavior (e.g. rushing or not paying attention), i.e. something that is within their control. Although prior studies have revealed intrinsic factors, such as self-rated health, as signifi- cant predictors of FOF, to our knowledge this is the first study to identify differences between those with and without con- cern about falling with regard to attribution of the fall to an intrinsic rather than behavioral factor. This lends support to our proposed framework that incorporates Bandura's theory of self-efficacy given that performance accomplishments, or lack thereof (e.g. a loss of balance during a functional activity), are an important component of perceived self-efficacy.

Clinical implications include informing the target popula- tion for interventions that address FOF outdoors, but also the content of intervention programs. Given that assistance needed to get up after a fall is more common among those with concern about falling, teaching strategies to get up safely from the ground outdoors may be warranted. Also, knowledge and skills in safely navigating curbs and stairs, despite physical limitations, are also indicated. Since fear or concern about falling outdoors was more common among those who attributed their fall to an intrinsic factor, in addition to training in use of outdoor fall prevention strategies (e.g. avoiding cell phone use, taking an alternate route), training to increase physical abilities and balance or ways to compensate for these limitations may also be warranted. Noteworthy is that there are existing programs that have been shown to be effective for reducing FOF and fall risk, and that promote self- efficacy such as Matter of Balance (Tennstedt et al., 1998) and Stepping on (Clemson et al., 2004). However, they have not specifically targeted people who have concern or are at risk for outdoor falls. Therefore, programs specific to this popula- tion are still needed.

According to findings from both current and previous studies, environmental factors (e.g. slippery, uneven surfaces) can have a significant impact on outdoor falls and FOF out- side (Li et al., 2006), therefore efforts to reduce environmental hazards are needed at a national level for both existing com- munities and for new community/housing developments. Examples include regular repair of indentations and cracks, curbs and sidewalks made of non-slippery materials even when wet, and for new constructions, ground level entrances with lightweight or automatic opening doors to name a few.

Due to the cross-sectional design, it is not possible to ascertain when strategies for prevention identified in partici- pant's quantitative responses were adopted (i.e. before or after a fall). Longitudinal studies of outdoor falls are thus war- ranted. Our study was conducted among NYC residents only, which may impact generalizability of finding to other loca- tions. Also, we relied on self-report data which may have introduced bias, including possible difficulty recalling a past fall event. Use of performance based measures for strategy use and prospective designs are suggestions for future studies.

In conclusion, differences in participant characteristics and fall experiences were found between people with and without FOF outdoors. A model that combines the ecological perspec- tive with Bandura's theory of self-efficacy appears to provide an effective guide for understanding factors associated with FOF outdoors. Our results support our initial hypotheses which were based upon this framework. Specifically, FOF was impacted by multiple levels of influence including intrinsic factors (e.g. age, gender, functional status), the neighborhood environment (urban/suburban, stairs, and curbs), as well as factors associated with self-efficacy (e.g. personal experience of outdoor falls, ability to get up after a fall, etc.). Results can help to inform programs that address outdoor falls and FOF outside.

Acknowledgments

The authors would like to thank the graduate research assistants who contributed to data collection for this study and Dr Jim Hinojosa for his feedback on the conceptual model.

854 T. CHIPPENDALE AND C. D. LEE

Disclosure statement

No potential conflict of interest was reported by the authors.

References

Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191–215.

Bateni, H., & Maki, B.E. (2005). Assistive devices for balance and mobility: Benefits, demands, and adverse consequences. Archives of Physical Medicine and Rehabilitation, 86(1), 134–145. doi:10.1016/j. apmr.2004.04.023

Center for Disease Control. (2016). Older adult falls. Retrieved from http:// www.cdc.gov/homeandrecreationalsafety/falls/

Chamberlin, M.E., Fulwider, B.D., Sanders, S.L., & Medeiros, J.M. (2005). Does fear of falling influence spatial and temporal gait parameters in elderly persons beyond changes associated with normal aging?. Jour- nal of Gerontology, 60A, 1163–1167.

Chippendale, T. (2015). Development and validity of the outdoor falls questionnaire (OFQ). International Journal of Rehabilitation Research, 38, 263–269.

Chippendale, T., & Boltz, M. (2014). The neighborhood environment: Per- ceived fall risk, resources and strategies for fall prevention. The Geron- tologist, 55, 575–583.

Clemson, L., Cumming, R.C., Kendig, H., Swann, M., Heard, R., & Taylor, K. (2004). The effectiveness of a community-based program for reducing the incidence of falls in the elderly: A randomized Trial. Journal of the American Geriatrics Society, 52, 1487–1494.

Denkinger, M.D., Lukas, A., Nikolaus, T., & Hauer, K. (2015). Factors associ- ated with fear of falling and associated activity restriction in commu- nity-dwelling older adults: A systematic review. American Journal of Geriatric Psychiatry, 23, 72–85.

Deshpande, N., Metter, E.J., Lauretani, F., Bandinelli, S., Guralnik, J., & Fer- rucci, L. (2008). Activity restriction induced by fear of falling and objec- tive and subjective measures of physical function: A prospective cohort study. Journal of the American Geriatrics Society, 56, 615–620. doi:10.1111/j.1532-5415.2007.01639.x

Dillman, D.A. (2000). Mail and Internet surveys. New York, NY: John Wiley & Sons.

Donoghue, O.A., Ryan, H., Duggan, E., Finucane, C., Saav, G. Cronin, & Kenny, R.A. (2014). Relationship between fear of falling and mobility varies with visual function among older adults. Geriatrics and Gerontol- ogy International, 14, 827–836. doi:10.1111/ggi.12174

Friedman, S.M., Munoz, B., West, S.K., Rubin, G.S., & Fried, L.P. (2002). Falls and fear of falling: Which comes first? A longitudinal prediction model suggests strategies for primary and secondary prevention. Journal of the American Geriatrics Society, 50, 1329–1335.

Hospers, G.P., Smulders, Y.M., Maier, A.B., Deeg, D.J., & Muller, M. (2015). Relation between blood pressure and mortality risk in an older popu- lation: Role of chronological and biological age. Journal of Internal Medicine, 277, 488–497. doi:10.111/joim.12284

House, J.S., Lepkowski, J.M., Williams, D.R., Mero, R.P., Lantz, P.M., Robert, S. A., & Chen, J. (2000). Excess mortality among urban residents: How much, for whom, and why?. American Journal of Public Health, 90, 1898–1904.

Hsieh, H.-F., & Shannon, S.E. (2005). Three approaches to qualitative con- tent analysis. Qualitative Health Research, 15, 1277–1288. doi:10.1177/ 1049732305276687

Hull, S.L., Kneebone, I.I., & Farquharson, L. (2013). Anxiety, depression, and fall-related psychological concerns in community-dwelling older peo- ple. American Journal of Geriatric Psychiatry, 21, 1287–1291. doi:10.1016/j.jagp.2013.01.038

Jorstad, E.C., Hauer, K., Becker, C., & Lamb, S.E. (2005). Measuring the psy- chological outcomes of falling: A systematic review. Journal of the American Geriatrics Society, 53, 501–510.

Kelsey, J.L., Proctor-Gray, E., Hannan, M.T., & Li, W. (2012). Heterogeneity of falls among older adults: Implications for public health prevention. American Journal of Public Health, 102, 2149–2156.

Kumar, A., Carpenter, H., Morris, R., Iliffe, S., & Kendrick, D. (2014). Which factors are associated with fear of falling in community-dwelling older people?. Age and Ageing, 43, 76–84. doi:10.1093/ageing/aft154

Li, W., Keegan, T.H.M., Sternfeld, B., Sidney, S., Quesenberry, C.P., & Kelsey, J.L. (2006). Outdoor falls among middle-aged and older adults: A neglected public health problem. American Journal of Public Health, 96, 1192–1199.

Liu, J.Y.W. (2014). Fear of falling in robust community dwelling older peo- ple: Results of a cross-sectional study. Journal of Clinical Nursing, 24, 393–405. doi:10.1111/jocn.12613

Malini, F.M., Lourenco, R.A., & Lopes, C.S. (2015). Prevalence of fear of fall- ing in older adults, and its associations with clinical, functional and psychosocial factors: The frailty in Brazilian older people-Rio de Janeiro study. Geriatrics and Gerontology International, 16, 336–344. doi:10.1111/ggi.12477

National Cancer Institute. (2005). Theory at a glance: A guide for health promotion practice. Retrieved from http://www.sbccimplementation kits.org/demandrmnch/wp-content/uploads/2014/02/Theory-at-a- Glance-A-Guide-For-Health-Promotion-Practice.pdf

NYC Department for the Aging. (2016). Profile of older New Yorkers. Retrieved from http://www.nyc.gov/html/dfta/downloads/pdf/demo graphic/profileçolderNYRSç0911çnewsummariesçupdate.pdf

NYC Department of Health and Mental Hygiene. (2014). Epi data brief: Neighborhood walkability and physical activity in New York City. Retrieved from https://www1.nyc.gov/assets/doh/downloads/pdf/epi/ databrief42.pdf

O'Cathain, A., Murphy, E., & Nicholl, J. (2010). Three techniques for inte- grating data into mixed methods studies. BMJ, 341, 1147–1150. doi:10.1136/bmj.c4587

Oh, E., Hong, G.-R.S., Lee, S., & Han, S. (2015). Fear of falling and its predictors among community-living older adults in Korea. Aging & Mental Health. Advanced online publication. doi:10.1080/13607863.2015.1099034

Reelick, M.F., Iersel, M.B., Kessels, R.P.C., Marcel, G.M., & Rikkert, O. (2009). The influence of fear of falling on gait and balance in older people. Age and Ageing, 38, 4350440. doi:10.1093/ageing/afp066

Robinson, J.C., Wyatt, S.B., Dubbert, P.M., May, W., & Sims, M. (2016). The impact of the neighborhood environment on physical activity in the Jackson Heart Study. Preventive Medicine, 90, 216–222. doi:10.1016/j. ypmed.2016.07.025

Sood, S., Gallagher, I.J., Lunnon, K., Rullman, E., Keohane, A., Crossland, H., & Timmons, J.A. (2015). A novel multi-tissue diagnostic of healthy age- ing relates to cognitive health status. Genome Biology, 16. Retrieved from https://genomebiology.biomedcentral.com/articles/10.1186/ s13059-015-0750-x doi:10.1186/s13059-015-0750-x

Tennstedt, S., Howland, J., Lachman, M., Peterson, E., Kasten, L., & Jette, A. (1998). A randomized, controlled trial of a group intervention to reduce fear of falling and associated activity restriction in older adults. Journals of Gerontology Series B, 53, 384–392.

Trombetti, A., Reid, K.F., Hars, M., Herrman, F.R., Pasha, E., Phillips, E.M., & Fielding, R.A. (2016). Age-associated declines in muscle mass, strength, power, and physical performance: Impact on fear of falling and quality of life. Osteoporosis International, 27, 463–471. doi:10.1007/s00198- 015-3236-5

Twibell, R.S., Siela, D., Sproat, T., & Coers, G. (2015). Perceptions related to falls and fall prevention among hospitalized adults. American Journal of Critical Care, 24, e78–e85. doi:http://dx.doi.org/10.4037/ajcc2015375

Verma, S.K., Willetts, J.L., Marucci-Wellman, H.R., Lombardi, D.A., & Court- ney, T.K. (2016). Falls and fall-related injuries among community- dwelling adults in the United States. PLos ONE, 11(3), e0150939. doi:10.1371/journal.pone.0150939

Yardley, L., Beyer, N., Hauer, K., Kempen, G., Piot-Ziegler, C., & Todd, C. (2005). Development and initial validation of the falls efficacy scale- international (FES-I). Age and Ageing, 34, 614–619. doi:10.1093/ageing/ afil96

AGING & MENTAL HEALTH 855

Copyright of Aging & Mental Health is the property of Routledge and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.