Examining the Validity and Reliability of the ABC-6 in Underserved Older Adults

Abstract

Losing confidence in the ability to maintain balance can be more debilitating than a fall; therefore, the accurate measurement of balance confidence is critical. The purpose of this study was to examine the validity and reliability of the ABC-6, a shortened version of the Activities-specific Balance Confidence scale (ABC), among underserved older adults. The final sample included 251 older adults (M age = 71.2 years, SD = 8.9; 72.1% African Americans, 62.5% low-income, 61% low-education). Participants completed assessments of multiple falls risk factors, physical activity, and balance confidence. The ABC-6 had excellent internal consistency reliability, substantial intraclass correlations, and significant moderate to large correlations with physical activity, mobility, balance, and total falls risk. It also demonstrated the ability to discriminate between fallers and nonfallers, and it was a significant predictor of total falls risk. The ABC-6 was a valid and reliable measure of balance confidence among underserved older adults.

Keywords

falls prevention balance confidence aging

Nearly 40 million, or approximately 13%, of the U.S. population are over the age of 65 years (U.S. Census Bureau, 2010). On average, one out of every three older adults will fall at least one time annually (Centers for Disease Control and Prevention, 2011), and fall rates steadily increase with age with the largest risk occurring over the age of 80 years (American Geriatrics Society [AGS], 2001). However, the prevalence data for falls may be largely underestimated because it is usually only falls that resulted in hospitalization that are reported and documented (Powell & Myers, 1995), and currently it is estimated that only 2.5% of falls lead to hospitalization (Rubenstein, 2006).

Falls that require hospitalization usually result in fractures or soft tissue injuries (Powell & Myers, 1995), but sometimes the psychological issues that lead to diminished confidence and activity restriction (Yardley & Smith, 2002) can be more debilitating than the physical results of falls. Poor balance confidence can lead to activity restriction, which can then lead to decreased muscle strength, loss of independence, and functional decline (Cumming, Salkeld, Thomas, & Szonyi, 2000; Lach, 2002; Quigley, Hann, & Evitt, 2003; Yardley & Smith, 2002); therefore, creating a never-ending cycle making falls more likely. For this reason, it is important to understand falls risk factors.

Individuals who are at risk for falls must be identified (Schepens, Goldberg, & Wallace, 2010). The U.S. Public Health Service estimated that approximately two- thirds of falls involving older adults are preventable, and simply identifying risk factors and changing one’s environment can help reduce the number of falls that occur in this population (Rubenstein, 2006). The AGS (2001) classified the causes of falls as intrinsic (e.g., disease or health-related), extrinsic (e.g., medication-related), or environmental (e.g., lack of bathroom safety equipment). The AGS identified the most common risk factors for falls and the risk ratio (RR) or odds ratio (OR) related to each. The factors included muscle weakness (RR/OR = 4.4), history of falls (RR/OR = 3.0), gait deficit (RR/OR = 2.9), balance deficit (RR/OR = 2.9), use of assistive device(s) (RR/OR = 2.6), visual deficit (RR/OR = 2.5), arthritis (RR/OR = 2.4), impaired activities of daily living (ADL; RR/OR = 2.3), depression (RR/OR = 2.2), cognitive impairment (RR/OR = 1.8), and being over the age of 80 (RR/OR = 1.7). Falls usually result from a combination of risk factors, and as the number of risk factors increases, the likelihood of falling also increases (AGS). Tinetti, Speechley, and Ginter (1988) reported that 27% of older adults with no or one risk factor were likely to suffer a fall compared to 78% of older adults with four or more risk factors.

In addition to the physical risk factors, there are psychological issues related to falls that may elevate risk for falling and therefore, may be important clinical endpoints for falls prevention programs (Jorstad, Hauer, Becker, & Lamb, 2005; Moore & Ellis, 2008). According to the review by Moore and Ellis, the most commonly studied psychological issues related to falls are fear of falling (i.e., persistent concern about falling that may cause older adults to limit physical activities that they may be able to perform in an attempt to avoid a future fall; Tinetti & Powell, 1993), falls-related self-efficacy or falls efficacy (i.e., individual’s belief in his or her ability to carry out ADL without falling; Tinetti, Richman, & Powell, 1990), and balance confidence (i.e, situation-specific form of self-efficacy that is a person’s estimate of their ability to maintain balance while performing ADL; Powell & Myers, 1995). Although these constructs are similar in nature, they are in fact unique and care should be taken to ensure that appropriate measurement techniques are employed (Moore & Ellis, 2008; Moore et al., 2011). Furthermore, research is needed to determine the psychological issues that are important to identify within a falls risk screening. Moore and colleagues compared the measurement properties of four fall-related psychological instruments that assessed falls efficacy, fear of falling, balance confidence, and consequences of falling within a community-based falls risk screening context. Their findings showed that the balance confidence instrument was the only one to distinguish between fallers and nonfallers and to predict total falls risk among a sample of racially and socioeconomically diverse older adults. The researchers concluded that the assessment of balance confidence with the Activities-specific Balance Confidence scale (ABC; Powell & Myers, 1995) may be the better instrument to use within a community-based falls risk screening, but that shorter variations of the instrument should be tested and considered for quicker and easier use within that specific context. Therefore, further psychometric testing of the ABC and its modified versions to measure balance confidence within a community-based falls risk screening context is warranted.

The ABC 6-item scale (ABC-6) is a simplified version of the ABC that includes the six most challenging activities from the original ABC. It was developed to be a shorter, quicker version of the balance confidence scale, making it easier to administer in applied settings (Peretz, Herman, Hausdorff, & Giladi, 2006). Whereas the original ABC can require up to 20 min for administration, in applied settings a quicker assessment of balance confidence, such as the ABC-6, is preferred (Peretz et al., 2006). The ABC-6 also provides a more true representation of balance confidence by eliminating the questions in the original ABC that could potentially inflate overall confidence levels (Schepens et al., 2010).

During the development of the ABC-6, 70 participants with high-level gait disorders, 19 participants with Parkinson’s disease, and 68 healthy participants were administered the original ABC (Peretz et al., 2006). The six questions with the lowest scores were taken from the questionnaire and comparisons were made between the ABC and the ABC-6 to test for reliability and validity. The results showed that the internal consistency of the ABC-6 was comparable to the longer version (ABC). Schepens and colleagues (2010) also examined the ABC-6 for validity and reliability among older adults. With a sample of 35 adults aged 60 years and over, the researchers administered the ABC and ABC-6 1 month apart and found that the scores for the ABC and the scores for the ABC-6 were highly correlated. Furthermore, with good to excellent test-retest reliability, the ABC-6 was ruled to be a valid and reliable instrument to assess balance confidence (Schepens et al., 2010). Schepens and colleagues also found that the ABC-6 differentiated between fallers and nonfallers, which the ABC did not, and the ABC-6 showed a strong correlation with the number of falls experienced by the participants that suggested it was a better predictor of falls than the ABC. The validity of the ABC-6 is further supported by Lohnes and Earhart (2010) who showed moderate to large correlations between the ABC-6 and the Berg balance scale and timed-up-and-go test in 89 persons with Parkinson’s disease. While researchers have examined the ABC-6 in comparison to the ABC, the use of the ABC-6 is still relatively new and leaves room for further research to be done to examine the validity and reliability of this instrument in applied contexts.

Falls are a common problem for many older adults, and the consequences of falls including a loss of confidence in one’s ability to maintain balance can sometimes be more detrimental than a fall itself. Moreover, poor balance confidence can result in an older adult adopting a sedentary lifestyle or restricting activities, leading to functional decline, a loss of muscle mass, and an increased risk of falling. For this reason it is important to identify the best measurement instrument of balance confidence that can be used in a falls risk screening setting. Therefore, the purpose of this study was to examine the validity and reliability of the ABC-6 in underserved older adults within a community-based falls risk screening context. This research further validates the ABC-6 and builds upon previous research by: (a) incorporating a larger sample size, (b) using underserved older adults who are primarily minority, lower education, and lower income without documented movement disorders, (c) examining the ABC-6 in comparison to an overall falls risk score, and (d) testing it in an applied setting with time constraints. Specifically, we hypothesized that reliability of the ABC-6 would be comparable to the ABC by exhibiting (a) acceptable internal consistency reliability and (b) substantial intraclass correlations (ICC). We also hypothesized that the ABC and ABC-6 would be valid measures of balance confidence by (a) demonstrating moderate to large correlations with physical activity, mobility, balance, and total falls risk scores; (b) discriminating between fallers and nonfallers with fallers reporting significantly lower scores on the ABC and ABC-6 than nonfallers; and (c) evaluating which of the two instruments would be the stronger predictor of total falls risk.

Method

Participants

Over a 2-year period, 321 older adults volunteered for a falls risk screening. Falls risk screenings were completed at 13 sites during Year 1 and 8 sites during Year 2 (7 of these sites were repeats from Year 1). Each participant signed an informed consent form approved by the Georgia State University Institutional Review Board. Volunteers were eligible for the falls risk screenings if they were: (a) 50 years of age or older, (b) ambulatory, and (c) demonstrated the ability to comprehend and follow instructions. Participant recruitment was performed using passive techniques such as flyers advertising the upcoming screenings and by word of mouth. For the purpose of this study, in the event that participants completed two falls risk screenings (i.e., 1 each year of the study), only their 1st year data were used.

Procedures

The community-based falls risk screenings lasted approximately 3 hr at each site. Participants signed up for a 30-min testing block to visit the following four stations where the tests for the comprehensive falls risk screening instrument were administered by trained student research assistants: (a) demographics and home environment, (b) medication review, (c) physical functioning (i.e., mobility and balance), and (d) vision. After participants visited all four stations, a total falls risk score was calculated. During the time period between stations and the time required for calculating the total falls risk score (approximately 15 min), the participants completed physical activity and balance confidence questionnaires during an interview with a trained student research assistant. Once the questionnaires were completed, a trained student research assistant reviewed the total falls risk score and several recommendations for falls prevention with each participant.

Measures

Demographic Information

Information about participants’ age, gender, income, education, and race was collected using a questionnaire developed for the falls risk screening study. The demographic information was collapsed into the following groups: (a) income = low (< US$15,672 annually), medium (US$15,684-US$22,032 annually), and high (≥ US$25,000 annually), (b) education = low (high school degree or less), medium (some college or associate’s degree), and high (bachelor’s degree or more), and (c) race = Caucasian, African American, and other (i.e., American Indian/Alaskan Native, Hispanic/Latino, or Asian). Income levels for older adult federal housing assistance, as well as Healthy People 2010 (U.S. Department of Health and Human Services, 2010) were used as guidelines for income and education group formation.

Comprehensive Falls Risk Screening instrument

The Comprehensive Falls Risk Screening Instrument (CFRSI) was used to assess falls risk (Fabre et al., 2010). The CFRSI is a collection of questions and tests that are grouped to calculate the five following falls risk subscales: (a) history, (b) physical, (c) medication, (d) vision, and (e) home environment. The subscale scores are averaged to produce a total falls risk score. The subscales are weighted according to the AGS (2001) guidelines. The falls risk subscales and the total falls risk scores are converted to a 0- to 100-point scale with higher scores indicating greater falls risk. The CFRSI was validated with samples of community-dwelling older adults (Fabre et al., 2010; Moore et al., 2011).

History subscale

Self-reported age (RR/OR = 1.7), history of falls (RR/OR = 3.0; number of falls within the past 12 months and within the past 3 years), assistive device usage (RR/OR = 2.6), and diagnosis of arthritis (RR/OR = 2.4; AGS 2001) were used to calculate the history risk score. History subscale scores are higher for participants who report an age over 80, a history of falls, a diagnosis of arthritis, and/or usage of an assistive device.

Physical subscale

Based on the AGS (2001) report that falls risk was related to muscle weakness (RR = 4.4), gait deficits (RR = 2.9), and balance deficits (RR = 2.9), a physical functioning subscale was created to include mobility and balance. The mobility and balance tests used to calculate the physical falls risk score were the Expanded Timed Get Up and Go Test (ETGUG; Wall, Bell, Campbell, & Davis, 2000) and the Functional Reach Test (FRT; Duncan, Weiner, Chandler, & Studenski, 1990).

The ETGUG (Wall et al., 2000) measured functional mobility. This test requires the participant to stand up from a chair without use of his or her arms, walk 10 meters around a cone as fast as they can, but at a pace that feels comfortable and safe, and return to the chair in a seated position. The ETGUG score is the time (in seconds) taken to complete the task. A higher ETGUG score (meaning reduced functional mobility) contributes to a greater physical subscale score.

The FRT (Duncan et al., 1990) measured standing balance. This test requires the participant to reach forward as far as possible with the dominant arm along a measurement tape that is fixed to a wall, without taking a step. The distance reached was measured in inches between the starting position of the middle finger and the final position of the middle finger after reaching forward. Shorter distances reached (meaning reduced standing balance) contributes to a greater physical subscale score.

Medication subscale

Falls risk was based on the following prescription medications: psychotropics (OR = 1.7), Class 1a anti-arrhythmics (OR = 1.6), digoxin (OR = 1.2), and diuretics (OR = 1.1; AGS, 2001; Leipzig, Cumming, & Tinetti, 1999). The participants also reported any medication side effects, if they used multiple pharmacists, and the frequency of pharmacy consult. Medication subscale scores are greater for participants who report taking four or more prescription medications, experience side effects, do not fill prescriptions at the same pharmacy, and/or do not have a pharmacist review their current medications.

Vision subscale

Participants responded to questions about the use of corrective lenses, lens use compliance, and if they had a visual screen within the previous 12 months, and they also completed a visual acuity test. The visual acuity test required the participant to read the Snellen eye chart from 20 feet with their corrective lenses (if applicable). Vision subscale scores are higher for participants who have not had a vision test in the last 12 months, do not wear their corrective lenses as prescribed, and/or have Snellen eye scores greater than 20/20.

Environment subscale

Information about the home environment was assessed by a 12-item Home Safety Checklist (Centers for Disease Control and Prevention National Center for Injury Prevention and Control, 2004). Yes or no responses were obtained to questions such as “Are stairways well lit with lights at the top and bottom of the stairs?” and “Are your steps, landings, and floors clear of clutter?” More “no” responses indicate a higher environment subscale score because the checklist identified possible home hazards.

Physical Activity Scale for the Elderly

The Physical Activity Scale for the Elderly (PASE; Washburn, Smith, Jette, & Janney, 1993) is a self-report measure of physical activity. Participants recall the frequency (days/week) and duration (hours) of various physical activities such as strength and endurance, sport, occupational, family care, household, yard work, and gardening activities that were performed over the past 7 days. Scores on the PASE can be between 0 to 400 (or more) with higher scores indicating greater physical activity participation (Washburn et al., 1993). The PASE is a valid and reliable measure of physical activity for independent-living older adults (Moore et al., 2008; Washburn et al., 1993).

Activities-specific Balance Confidence Scale

Balance confidence was measured with the ABC scale (Powell & Myers, 1995). Sixteen ADL are rated on a 0 to 100% scale (i.e., 0% = no confidence to 100% = complete confidence) and are averaged to produce a total ABC score that is between 0 to 100 with higher scores indicating greater balance confidence (Powell & Myers, 1995). The ABC is a valid and reliable measure of balance confidence among independent-living older adults (Moore et al., 2011; Myers, Powell, Maki, Holliday, & Brawley, 1996; Powell & Myers, 1995).

Activities-specific Balance Confidence Scale-6

The ABC-6 score (Peretz et al., 2006) was calculated from the participant responses on the ABC. The six items were averaged to produce a total balance confidence score that was between 0 and 100 with higher scores reflecting greater balance confidence. The ABC-6 has good to very good internal consistency, validity, and reliability in comparison to the ABC (Peretz et al., 2006; Schepens et al., 2010).

Statistical Analyses

Tests for normality and outliers were conducted. To determine reliability of the ABC-6 in comparison to the ABC, Cronbach’s α was calculated to assess internal consistency and ICC were used to assess the relationships between the two instruments. Cronbach’s α was categorized as excellent (>.9), good (>.8), acceptable (>.7), questionable (>.6), poor (>.5), or unacceptable (<.5; George and Mallery, 2003). ICC were considered to have a poor (.01), slight (.01—.20), fair (.21—.40), moderate (.41—.60), substantial (.61—.80), or almost perfect (.81—1.00) agreement (Landis & Koch, 1977).

The construct validity of the ABC and the ABC-6 were evaluated using Pearson correlation coefficient with 95% confidence intervals (CI) with physical activity (PASE), mobility (ETGUG), balance (FRT), and total falls risk (CFRSI). Correlations between .10 and .29 were considered small, correlations between .30 and .49 were considered moderate, and correlations > .50 were considered large (Cohen, 1988). Separate ANOVAs with Bonferroni corrections for alpha were used to discriminate between fallers and nonfallers. The ABC and the ABC-6 were evaluated against the total falls risk score (CFRSI) using stepwise multiple regression. Effect sizes (Cohen’s d), observed power, and 95% CI were calculated and classified according to threshold values of Cohen where appropriate. Statistical calculations were significant at alpha level of p < .05 unless indicated otherwise.

Results

Over a 2-year period, 321 older adults from 14 community organizations volunteered for a falls risk screening. There were 40 repeat participants from Year 1, 9 who did not meet the inclusion criteria, 8 participants who did not complete some of the falls tests, and 13 who were missing the ABC and the PASE. The final sample for analyses included 251 participants (M age = 71.2 years, SD = 8.9). Of the final sample, 76.1 % of the participants were female, 72.1% were African American, 62.5% were classified as low-income, and 61% were classified as low-education. Additional participant characteristics are reported in Table 1. All outcome measures were approximately normally distributed (see Table 2).

Table 1.

Frequencies of Participant Characteristics.

Characteristic	n	%
Gender
Male	60	23.9
Female	191	76.1
Marital status
Single	63	25.1
Married	51	20.3
Widowed	79	31.5
Divorced	57	22.7
Did not answer	1	0.4
Education
≤ High school	153	61.0
Some college/associates	52	20.7
≥ Bachelors	46	18.3
Annual income
Low (< US$15,672 annually)	157	62.5
Medium (US$15,684-US$22,032 annually)	27	10.8
High (≥ US$25,000 per year)	41	16.3
Did not answer	26	10.4
Race
White or Caucasian	60	23.9
Black or African American	181	72.1
Other	9	3.6
Did not answer	1	0.4
Fallen in past 12 months
Yes	83	33.1
No	167	66.5
Did not answer	1	0.4
Use of assistive device
Yes	94	37.5
No	157	62.5
Arthritis
Yes	160	63.7
No	91	36.3
≥ 4 Prescription medications
Yes	137	54.6
No	114	45.4

Table 2.

Mean Performance Scores on Outcome Variables.

				Skewness		Kurtosis
	M	SD	Range	Statistic	SE	Statistic	SE
PASE	95.9	63.4	0.0-391.3	1.67	.156	3.79	.310
ETGUG	25.7	14.9	10.0-96.8	2.32	.154	6.49	.306
FRT	9.4	3.4	1.0-32.0	1.18	.154	6.99	.306
CFRSI	40.5	11.4	10.9-70.8	−.04	.154	−.19	.306
ABC	71.1	24.3	0.0-100.0	−.79	.154	−.16	.306
ABC-6	56.9	30.0	0.0-100.0	−.24	.154	−1.12	.307

Note. PASE = physical activity scale for the elderly. ETGUG = expanded time get up and go. FRT = functional reach test. CFRSI = comprehensive falls risk screening instrument. ABC = activities-specific balance confidence scale. ABC-6 = activities-specific balance confidence 6-item scale.

Excellent internal consistency reliability was shown by a Cronbach’s α coefficient of .95 and .90 for the ABC and ABC-6, respectively (George & Mallery, 2003). With an ICC of .81 (p < .01), the ABC and ABC-6 have a substantial to almost perfect level of agreement (Landis & Koch, 1977).

Significant moderate to large correlations were found between the ABC and ABC-6 with physical activity, mobility, balance, and total falls risk scores. Specifically, the ABC had large correlations with the ETGUG (r = −.53, p < .01, CI = −.62 - −.44) and total falls risk (r = −.64, p < .01, CI = −.70 - −.56). The ABC also had moderate correlations with the FRT (r = .42, p < .01, CI = .32-.52) and the PASE (r = .38, p < .01, CI = .26-.48). The ABC-6 had large correlations with the ETGUG (r = −.56, p < .01, CI = −.64 - −.47) and total falls risk (r = −.66, p < .01, CI = −.72 - −.58). The ABC-6 also had moderate correlations with the FRT (r = .45, p < .01, CI = .35-.54) and the PASE (r = .44, p < .01, CI = .33-.53).

Separate ANOVAs with Bonferroni corrections for alpha (p < .025) revealed that both the ABC, F (1, 251) = 15.77, p < .01, η_p² = .06, observed power = .98, d = .51, and the ABC-6, F (1, 250) = 11.20, p = .001, η_p² = .04, observed power .92, d = .44, discriminated between fallers and nonfallers, with fallers reporting significantly lower scores (M_ABC = 62.7, CI = 57.6-67.8; M_ABC-6 = 48.1, CI = 41.7-54.4) than nonfallers (M_ABC = 75.3, CI = 71.7-78.9; M_ABC-6 = 61.3, CI = 56.8-65.7) on both scales. Finally, the ABC-6 was the only variable selected to predict total falls risk in the stepwise multiple regression analysis (β = −.66, p < .01), accounting for 42.9% of the variance in total falls risk.

Discussion

The purpose of this study was to examine the validity and reliability of the ABC-6 in underserved older adults within a community-based falls risk screening context. Overall, the results indicated that the ABC-6 was a reliable and valid measure of balance confidence comparable to the original ABC. Furthermore, these findings suggest that the ABC-6 is a suitable measure of balance confidence for use among underserved older adults within community-based settings.

The ABC-6 was found to be a reliable measure of balance confidence in comparison to the original ABC, which confirmed the first study hypothesis. The excellent internal consistency found in this study is consistent with results reported by previous investigators (Peretz et al., 2006; Powell & Myers, 1995). In addition, these results showed a substantial to almost perfect agreement between the ABC-6 and ABC, which is also consistent with findings reported by Peretz and colleagues and Schepens and colleagues (2010). Therefore, these findings confirm the reliability of the ABC-6 and extend the literature regarding the reliability among underserved older adults in applied settings.

The second study hypothesis was also confirmed by multiple analyses that support the construct validity of the ABC and ABC-6. First, there were moderate to large correlations between the ABC and ABC-6 and the constructs of physical activity, mobility, balance, and falls risk. These results are consistent with previous research (Moore et al., 2011; Myers, Fletcher, Myers, & Sherk, 1998; Schepens et al., 2010) in that higher balance confidence is associated with better mobility and balance, and greater physical activity participation. Furthermore, these findings extend the literature regarding the validity of the ABC-6 because this is the first study to compare it against the PASE and the CFRSI total falls risk score. The finding that the ABC-6 measure of balance confidence is associated with overall falls risk in a similar manner as the ABC (Moore et al., 2011) has important implications for falls prevention. As noted earlier, psychological issues such as balance confidence may be to blame for activity restriction (Yardley & Smith, 2002), which can then lead to decreased muscle strength, loss of independence, and functional decline (Cumming et al., 2000; Lach, 2002; Quigley et al., 2003; Yardley & Smith, 2002), and actually increases the risk of a future fall. Although this study design did not allow for tests of cause-and-effect to determine if reduced balance confidence marks the beginning of this dangerous cycle, it does appear to be imperative for health care professionals and researchers to assess balance confidence during falls risk screenings because of its important relationship with individual falls risk factors and overall falls risk. Moreover, these findings establish the importance of the relationship between balance confidence and falls risk among underserved older adults and within a community-based context.

Further evidence of the construct validity of the ABC and ABC-6 was provided by the ability of both instruments to discriminate between fallers and nonfallers, with fallers reporting significantly lower balance confidence than nonfallers. For the ABC, these results are consistent with previous research (Lajoie & Gallagher, 2004; Moore et al., 2011; Myers et al., 1996; Powell & Myers, 1995); however, they do contradict the findings reported by Schepens et al. (2010). While Schepens and colleagues found that the ABC-6 discriminated between fallers and nonfallers, the ABC did not. They noted that the discrepancy in the findings between their study and other studies may have been caused by their inclusion of community-dwelling older adults. However, the present study, as well as previous research (Moore et al., 2011; Myers et al., 1996) also used community-dwelling older adults and found that fallers reported significantly lower scores on the ABC in comparison to nonfallers. Overall, these results provide additional evidence of the construct validity of the ABC and ABC-6, and extend the literature by suggesting that both instruments are effective in identifying fall-related psychological differences between fallers and nonfallers among underserved older adults in community-based settings.

Finally, results revealed that the ABC-6 was superior to the ABC in its ability to predict total falls risk. Balance confidence, as measured by the ABC-6, was the only variable selected in the stepwise multiple regression and explained nearly 43% of the variance in the total falls risk score. This is a significant amount of explained variance considering that overall falls risk is influenced by a variety of intrinsic, extrinsic, and environmental factors beyond the psychological risk factor of balance confidence (AGS, 2001). Moore et al. (2011) previously examined the original ABC in comparison to three other fall-related psychological instruments and their ability to predict the CFRSI total falls risk score. They reported that the ABC was the only significant predictor of falls risk and it explained about 25% of the variance. Moore et al. concluded that the ABC may be the better instrument to select to measure fall-related psychological difficulties within a community-based context. Therefore, the findings from the current study suggest that researchers may want to consider using the shorter, ABC-6 for the assessment of balance confidence at community-based falls risk screenings when time constraints are more likely. These findings also provide additional evidence for the construct validity of the ABC-6 and further extend the research because of its ability to predict total falls risk among underserved older adults.

There are several strengths of this study that extend the previous research on the evaluation of the psychometric properties of the ABC-6. First, there have been repeated calls for the inclusion of hard-to-reach and underserved populations in health-related research (Hendrickson, 2005), as well as interest in testing the reliability and validity of fall-related psychological instruments among diverse samples of older adults (Moore & Ellis, 2008). Although the study sample was comprised of mostly female participants, and a much larger percentage than the 50.8% in the general population (U.S. Census Bureau, 2010), this is comparable to previous research examining the ABC-6 (Schepens et al., 2010), and it is not uncommon within health-related research (Backer, Gregory, Jaen, & Crabtree, 2006). However, it is important to note that the larger falls risk study from which these data were drawn intentionally targeted community centers and senior residences with African American, low-income, and low-education older adults, who are often considered a hard-to-reach population. Indeed, this sample had a greater representation of African Americans and fewer people with at least a bachelor’s degree compared to 12.6% and 27.9%, respectively found in the general population (U.S. Census Bureau, 2010). In addition, 62.5% of this sample had an income less than US$15,672 annually, much less than the average annual income of US$31,408 for individuals of similar age (U.S. Census Bureau, 2010). Although generalizability may be limited, the sample demographics are a strength of the research because the findings demonstrate that the ABC-6 can successfully be used as a measure of balance confidence among a broad range of older adults and specifically among older adults who are not well-represented in this type of research.

Second, recent research findings indicate that African Americans and low-income older adults have greater overall falls risk in comparison to their counterparts (Ellis, Kosma, Fabre, Moore, & Wood, 2012) and the ABC-6 scores from this study provide additional evidence to support this conclusion. Specifically, the average ABC-6 score in this sample was 56.9; whereas the average ABC-6 score was 74.4 for Schepens and colleagues (2010) who surveyed a primarily healthy sample. Peretz and colleagues (2006) also surveyed a primarily White, healthy group of older adults and found a mean ABC-6 score to be 92.7. This research highlights the role of balance confidence as a contributing factor to falls in the hard-to-reach low-income, African American population.

Finally, another important contribution of this study was that it was the first to show the relationship between the ABC-6 and physical activity. Prior to this study, Schepens and colleagues (2010) compared the ABC-6 to mobility and balance, but there had been no supporting evidence showing that one’s level of physical activity may have a significant correlation with the shortened version of the ABC that targets the most challenging ADL of the original. Moreover, this study was also the first to evaluate the ABC-6 among independent-living older adults based on its ability to predict falls risk using the CFRSI (Fabre et al., 2010). This is a significant finding considering the CFRSI was designed and validated as an instrument to measure overall falls risk based on multiple risk factors (Fabre et al., 2010). As shown by the variance accounted for by the ABC-6 in reference to the overall falls risk, the study findings confirm that balance confidence is an important falls risk factor to assess.

There were some limitations to this investigation. First, the study participants were volunteers. This may have created selection bias resulting in the sample being mostly comprised of older adults who are concerned about their falls risk. Therefore, the participants in this study may not represent all community-dwelling older adults. However, most studies are comprised of volunteers and may include a similar limitation. It is also important to note that 33% of the study participants reported a fall during the year prior to the falls risk screening. This is consistent with population estimates that approximately one third of older adults over the age of 65 will experience at least one fall annually (Centers for Disease Control and Prevention, 2011) making it possible that the study sample is representative of the larger older adult population. Still, future investigators should attempt to recruit community members outside of community-based organizations.

Another limitation of this study is that interrater reliability was not evaluated. Although the student research assistants were trained before the falls risk screenings with a standardized approach, anytime there are multiple administrators, interrater reliability can be compromised. However, it should be noted that one of the main objectives of this study was to test the validity and reliability of the ABC-6 in a community-based setting and the use of multiple administrators is similar to the way this screening instrument would be used in applied settings. Future researchers should consider providing evidence of interrater reliability in community-based settings to further validate the ABC-6.

In addition, participants only completed the ABC and these scores were used to calculate the ABC-6 results. It is possible that participants might rate their balance confidence differently if they were only asked to rate their level of confidence on the ABC-6 questions. Therefore, future psychometric evaluations of the ABC-6 should be based on these six questions alone and within prospective designs to determine if the ABC-6 predicts actual falls. Finally, in addition to the ABC, many of the study outcomes were derived from self-reported information. While it is not uncommon for similar studies to also be largely based on self-reported information, it is possible that the accuracy of the results was affected by recall issues and social desirability bias, with participants more likely to overestimate balance confidence, physical activity participation, and so forth. Future studies should consider administering a mini-mental state exam to potentially recognize any mental issues that may affect one’s ability to recall events, and questionnaires should also be administered in a private, one-on-one interview session to reduce social desirability bias.

In summary, the study findings provide additional evidence of the reliability and validity of the ABC-6 as a psychological measure of balance confidence among underserved older adults. Moreover, these findings suggest that the ABC-6 may be the better instrument to choose for community-based falls risk screenings based on its ability to not only discriminate between fallers and nonfallers, but also explain the most variance in total falls risk. The ABC-6 may also be a more practical balance confidence assessment tool than the ABC in settings where balance confidence, as well as other falls risk factors, must be measured accurately, but quickly.

Footnotes

Acknowledgements

We would like to thank the thesis committee members, Drs. Elisabeth Burgess and Tai Wang, for their assistance and contributions to the research.

Author’s Note

This research is based on Antonius Skipper’s master’s thesis completed in June, 2012 under the direction of Dr. Rebecca Ellis.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: One year of this research was funded by a Research Initiation Grant from the University Research Services and Administration at Georgia State University (PI = Rebecca Ellis).

Author Biographies

Antonius Skipper, MS, is a Gerontologist and doctoral student in social work at Louisiana State University. He is a recipient of the Huel D. Perkins Diversity Fellowship. His research interests include health-related behaviors of underserved older adults and religiosity and coping in older African Americans.

Rebecca Ellis, PhD, is an Associate Professor in the Department of Kinesiology and Health at Georgia State University where she directs the Exercise Psychology doctoral concentration. Her research interests include understanding physical activity motivation and behavior change and its effects on health-related outcomes including falls risk factors and quality of life.

References

American Geriatrics Society, British Geriatrics Society, and American Academy of Orthopedic Surgeons Panel on Fall Prevention (2001). Guideline for the prevention of falls in older adults. Journal of the American Geriatrics Society, 49, 664-672.

Backer

Gregory

Jaen

Crabtree

(2006). A closer look at adult female healthcare maintenance visits. Health Services Research, 38, 355-360.

Centers for Disease Control and Prevention National Center for Injury Prevention and Control (2004). Home safety checklist. 2004 home safety council. Retrieved from www.cdc.gov/injury

Centers for Disease Control and Prevention (2011). Injury prevention and control: Home and recreational safety: 2010. Retrieved from http://www.cdc.gov/HomeandRecreationalSafety/Falls/adultfalls.html

Cohen

(1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

Cumming

Salkeld

Thomas

Szonyi

(2000). Prospective study of the impact of fear of falling on activities of daily living, sf-36 scores, and nursing home admission. Journal of Gerontology: Medical Sciences, 55A, M299-M305.

Duncan

Weiner

Chandler

Studenski

(1990). Functional reach: A new clinical measure of balance. Journal of Gerontology: Medical Sciences, 45, M192-M197.

Ellis

Kosma

Fabre

Moore

Wood

(2012). Proximal determinants of falls risk among independent-living older adults. Research on Aging. Advance online publication. doi:10.1177/0164027512446940

Fabre

Ellis

Kosma

Moore

McCarter

Wood

(2010). Development and validation of the comprehensive falls risk screening instrument. Physical and Occupational Therapy in Geriatrics, 28, 181-194.

10.

George

Mallery

(2003). SPSS for windows step by step: A simple guide and reference. 11.0 update (4th ed.). Boston, MA: Allyn & Bacon.

11.

Hendrickson

(2005). Reaching an underserved population with a randomly assigned home safety intervention. Injury Prevention, 11, 313-317.

12.

Jorstad

Hauer

Becker

Lamb

(2005). Measuring the psychological outcomes of falling: A systematic review. Journal of American Geriatrics Society, 53, 501-510.

13.

Lach

(2002). Fear of falling: An emerging public health problem. Generations: Journal of the American Society on Aging, 26, 33-37.

14.

Lajoie

Gallagher

(2004). Predicting falls within the elderly community, comparison of postural sway, reaction time, the Berg balance scale and the Activities-specific Balance Confidence (ABC) scale for comparing fallers and non-fallers. Archives of Gerontology and Geriatrics, 38, 11-26.

15.

Landis

Koch

(1977). The measurement of observer agreement for categorical data. Biometrics, 33, 159-174.

16.

Leipzig

Cumming

Tinetti

(1999). Drugs and falls in older people: A systematic review and meta-analysis: II. Cardiac and analgesic drugs. Journal of the American Geriatrics Society, 47, 40-50.

17.

Lohnes

Earhart

(2010). External validation of abbreviated versions of the Activities-specific Balance Confidence scale in Parkinson’s disease. Movement Disorders, 25, 485-489.

18.

Moore

Ellis

(2008). Measurement of fall-related psychological constructs: A review of the research literature. Aging and Mental Health, 12, 684-699.

19.

Moore

Ellis

Allen

P. D.

Cherry

K. E.

Monroe

P. A.

O’Neil

C. E.

Wood

(2008). Construct validation of physical activity surveys in culturally diverse older adults: A comparison of four commonly used questionnaires. Research Quarterly for Exercise and Sport, 79, 42-50.

20.

Moore

Ellis

Kosma

Fabre

McCarter

Wood

(2011). Comparison of the validity of four fall-related psychological measures in a community-based falls risk screening. Research Quarterly for Exercise and Sport, 82, 545-554.

21.

Myers

Powell

Maki

Holliday

Brawley

(1996). Psychological indicators of balance confidence: Relationship to actual and perceived abilities. Journals of Gerontology: Series A: Biological Sciences and Medical Sciences, 51, M37-M43.

22.

Myers

Fletcher

Myers

Sherk

(1998). Discriminative and evaluative properties of the Actvities-specific Balance Confidence (ABC) Scale. Journal of Gerontology: Medical Sciences, 53A, M287-M294.

23.

Peretz

Herman

Hausdorff

Giladi

(2006). Assessing fear of falling: Can a short version of the Activities-specific Balance Confidence scale be useful? Movement Disorders, 21, 2101-2105.

24.

Powell

Myers

(1995). The activities-specific balance confidence (ABC) scale. Journal of Gerontology: Medical Science, 50, M28-M34.

25.

Quigley

Hann

Evitt

(2003). When your patient fears falling. Federal Practitioner, 20, 21-28.

26.

Rubenstein

(2006). Falls in older people: Epidemiology, risk factors and strategies for prevention. Age & Ageing, 35, 37-41.

27.

Schepens

Goldberg

Wallace

(2010). The short version of the activities-specific balance confidence (ABC) scale: Its validity, reliability, and relationship to balance impairment and falls in older adults. Archives of Gerontology & Geriatrics, 51, 9-12.

28.

Tinetti

Powell

(1993). Fear of falling and low self-efficacy: A cause of dependence in elderly persons. Journals of Gerontology, 48, 35-38.

29.

Tinetti

Richman

Powell

(1990). Falls efficacy as a measure of fear of falling. Journals of Gerontology, 45, 239-243.

30.

Tinetti

Speechley

Ginter

(1988). Risk factors for falls among elderly persons living in the community. The New England Journal of Medicine, 319, 1701-1707.

31.

U. S. Census Bureau (2010). U.S.A. quickfacts from the U.S. Census Burean: 2010. Retrieved from http://quickfacts.census.gov/qfd/states/00000.html

32.

U. S. Department of Health and Human Services (2010). Healthy people 2010: Final review. Retrieved from http://www.cdc.gov/nchs/healthy_people/hp2010/hp2010_final_review.htm

33.

Wall

Bell

Campbell

Davis

(2000). The timed get-up-and-go test revisited: Measurement of the component tasks. Journal of Rehabilitation Research and Development, 37, 109-114.

34.

Washburn

Smith

Jette

Janney

(1993). The physical-activity scale for the elderly (PASE)—development and evaluation. Journal of Clinical Epidemiology, 46, 153-162.

35.

Yardley

Smith

(2002). A prospective study of the relationship between feared consequences of falling and avoidance of activity in community-living older people. The Gerontologist, 42, 17-23.