The Inventory of Physical Activity Barriers for Community-Dwelling Adults 50 Years and Older: Development and Preliminary Validation?

Abstract

Background:

Healthcare providers, including physical therapists, need to identify the reasons for insufficient physical activity (PA) to assist the 56–73% of community-dwelling adults 50 years and older who are not performing the recommended 150 minutes of moderate-to-vigorous PA. Currently, there is no feasible, multifactorial tool to assess PA barriers among this population. Without a tool, healthcare providers must either rely on self-generated questions or collate results from multiple assessments to identify PA barriers related to personal, social, and environmental factors, which can be time-consuming and incomplete.

Purpose:

To develop the Inventory of Physical Activity Barriers (IPAB), an assessment tool that examines personal, social, and environmental PA barriers.

Method:

We developed and psychometrically evaluated the IPAB using a three-phase process. For phase 1, we used a deductive method to develop the initial scale. During phase 2, we refined the scale and explored its psychometric properties by collecting cross-sectional pilot data on community-dwelling adults 50 years and older. We used descriptive statistics, item-scale correlations, construct validity via Mann Whitney U Test, and internal consistency via Cronbach alpha to analyze the data from phase 2. After identifying the scale’s potential for being valid and reliable, we implemented phase 3, a modified Delphi technique.

Results:

Using item-scale correlations, descriptive statistics, and consensus among PA experts, we refined the initial scale from 172 items to 40 items. The 40-item IPAB demonstrated good construct validity (determined by the scale’s ability to differentiate between individuals who did and did not meet 150 minutes/week of moderate-to-vigorous PA; p=.01) and internal consistency (Cronbach alpha of .97).

Conclusion:

Our preliminary results suggest that the IPAB is valid and reliable. Using the IPAB, healthcare providers will be able to identify patients’ PA barriers and thus develop individualized PA prescriptions, an evidence-based method of increasing PA.

INTRODUCTION

Non-communicable diseases such as diabetes, chronic lung disease, and cardiovascular disease are leading causes of morbidity and mortality.¹ Among adults, a primary risk factor for non-communicable diseases is insufficient physical activity (PA).^2,3 Physical activity is defined as any bodily movements produced by skeletal muscles that result in energy expenditure.⁴ It is important to note that PA is different from ‘exercise.’ Exercise is a subset of PA, and it is defined as a planned, structured, and repeated behavior aimed to maintain or improve components of physical fitness.⁴ Insufficient PA is defined as not meeting the recommended amount of PA.⁵ Among adults 50 years and older, the recommended amount of PA is a minimum of 150 minutes of moderate-to-vigorous PA per week.⁶ The prevalence of insufficient PA and non-communicable diseases increases with aging.⁷ Approximately 27–35% of adults 50 years and older report no leisure-time PA⁷ in the last month and 56–73% of adults 65 years and older do not meet the recommended levels of aerobic PA.⁸ The association between insufficient PA, non-communicable diseases, aging, and its health-related sequelae such as the risk of injury, frailty, falls, and disability, highlight the need for healthcare providers to promote PA.⁹ Among healthcare providers, a profession that has the needed educational background and expertise to be key PA promoters are physical therapists.¹⁰ As adults above the age of 50 tend to have a greater number of PA barriers, including pain, injury, or health status, the need for a physical therapist’s expertise in PA prescription is even greater.^11,12

It is important to note that providing general PA recommendations such as “you would benefit from a walking program,” is ineffective.¹³ Instead healthcare providers, such as physical therapists, need to individualize PA prescriptions and include intervention or strategies that address the patient’s PA barriers.^14,15 However, currently, there are no clinically feasible multifactorial PA barrier scales for community-dwelling adults above the age of 50 without mobility impairments. Therefore, physical therapists who evaluate barriers to PA must either rely on self-generated questions, incomplete scales,¹⁶ scales validated on specific populations, such as individuals with mobility impairments,^17–19 or use multiple scales to comprehend the personal, social, institutional, and community factors that impact adherence to PA. Each type of factor could be further split into multiple assessments. For example, to examine personal barriers, one must combine tools such as the Physical Activity Enjoyment Scale,²⁰ Outcome Expectation Scale,²¹ Index of Self-Regulation,²² and Multidimensional Self-Efficacy for Exercise Scale.²³ Thus, there is a need for a feasible, multifactorial PA barrier assessment tool for patients 50 years and older. The tool can be for both middle-aged and older adults, because middle-aged and older adults have been identified as having similar PA barriers.^12,24 However, because their PA barriers are different than those among younger adults,^25,26 it is important to develop the scale based on barriers identified among adults 50 years and older. Using a validated PA barrier assessment tool, one can use the results to individualize PA prescriptions, an evidence-based approach to increasing PA adherence.²⁷

To ensure that a barrier assessment tool examines the multiple factors associated with not meeting the recommended levels of PA, it is important to use a conceptual model that encompasses the complexity of insufficient PA. One such model is the Social-Ecological Model.²⁸ The Social-Ecological Model aims to elucidate the complex interplay between individual, social, institutional, community, and policy domains that impact behavior. In the model, the individual domain, also referred to as the intrapersonal domain, is defined by individual characteristics including knowledge, attitudes, behavior, self-concept, and skills.²⁸ The social domain, also referred to as the interpersonal domain, is defined as either formal or informal social networks or support systems, including family, work, workout partners, and friends.²⁸ The institutional construct is defined as social institutions with organizational characteristics and formal or informal rules and regulations for operation.²⁸ And the community construct is defined as the relationship among organizations, institutions, and informal networks within defined boundaries.²⁸ Using a well-established theoretical model also ensures that the scale’s foundation is based on clear constructs and definitions.²⁹

The purpose of our study was to use the Social-Ecological Model to develop a PA barrier assessment tool called the Inventory of Physical Activity Barriers (IPAB). We achieved this objective by following a three-phase process: 1) Develop initial items for the IPAB, 2) Collect pilot data to refine the IPAB based on input from scale respondents (community-dwelling adults 50 years and older) and evaluate the IPAB’s potential to become a valid and reliable method of assessing PA barriers, and 3) Incorporate expert opinions of diverse stakeholders, including healthcare professionals, public health officials, and PA researchers.

MATERIALS AND METHODS

Design and Overall Procedure

Our Institution’s Internal Review Board-approved study was conducted from June until November 2019. All participants completed the Internal Review Board-approved consenting process prior to study participation. The study was broken into three phases. During the first phase we developed the IPAB using a deductive method. The second phase was a pilot study during which we collected cross-sectional data and the third phase was a modified Delphi study used to evaluate the instrument’s content validity.

Phase 1: Initial Scale Development

The purpose of the first phase was to generate items that captured the dimensions of the theoretical constructs pertaining to personal, social, and environmental PA barriers.²⁹ To ensure scale quality, the scale development process was based on a social-ecological framework and the recommended deductive method of item generation, including an extensive literature review for pre-existing scales and systematic reviews looking at PA barriers among adults 50 years and older.^30–33

The social-ecological model was used to develop the following MESH terms “(((((((“scale”[All Fields] OR “questionnaire”[All Fields]) OR “inventory”[All Fields]) AND “barrier”[All Fields]) OR “personal barrier”[All Fields]) OR “social barrier”[All Fields]) OR “environmental barrier”[All Fields]) AND (“physical activity”[All Fields])) OR (“exercise”[All Fields]))))))).” The construct of environmental was selected over the constructs of institutional, community, and policy, because an initial literature search using the three constructs led to few relevant articles. Additionally, by grouping them into the category of environmental, we were able to simplify our scale and increase ease of use. All results were data-mined for any additional articles that may have been missed during the literature search. Using our systematic search strategy we identified the 17 previously validated scales seen in Table 1.

Table 1.

Validated scales used for the initial item pool

Personal barriers:  • Self-Efficacy for Exercise 34  • Barriers Self-Efficacy Scale 35  • Exercise Self-Efficacy Scale 36  • Multidimensional Outcome Expectations for Exercise Scale 37  • Outcome Expectations for Exercise Scale 21  • Physical Activity Enjoyment Scale 20  • Index of Self-Regulation 22 Social barriers:  • Family Support for Exercise Habits Scale 38  • Social Support Assessment Scale 39  • Social Support for Exercise 40  • Physical Activity Social Support 39  • Multidimensional Scale for Assessing Positive and Negative Social Influences on Physical Activity 41 Environmental:  • Neighborhood Environment Walkability Scale 42 Multi-factorial scale with poor psychometric properties:  • Exercise Benefits/Barriers Scale 43 Multi-factorial scales that are validated on populations with mobility impairments:  • Barriers to Health Promotion Activities for Disabled Persons Scale 17  • Barriers to Physical Activity Questionnaire for People with 19

After combining items from previously validated scales and systematic reviews we created a 172-item pool. We reviewed the item pool and removed items that had the potential of being agreed upon by all participants regardless of their PA level, were absolutes, or did not fit into a domain within the social-ecological model.²⁹ We also modified items that contained verbs written in an incorrect tense, contained ambiguity, had double negatives, or consisted of complex sentences.²⁹ The item review resulted in the first draft of the IPAB and contained 148 items. Using the social-ecological model and the above-mentioned operational definitions, we classified each item into one of the following subscales: personal, social, or environmental. The scale-including the items, subscale classifications, instructions, and formatting- were finalized after feedback was provided by a medical doctor, an exercise physiologist, and a physical therapist who had experience working with patients 50 years and older, were aware of the recommended PA guidelines, and incorporated them into clinical practice.

Phase 2: Cross-Sectional Pilot Data

The second phase was conducted to examine the quality of the items, refine the scale, and to evaluate the IPAB’s potential of becoming a valid and reliable PA barrier scale.

Design and procedure:

In this cross-sectional study, participants completed a demographic questionnaire, the Physical Activity Vital Sign (PAVS)⁴⁴ and the 148-item IPAB. Data was collected using both an electronic and a pen-and-paper version secondary to previous research identifying that validating both a pen-and-paper and an electronic version increases clinical implementation and utility.⁴⁵ All electronic data was collected, recorded, and stored using Research Electronic Data Capture (Vanderbilt University, Nashville, TN) and pen-and-paper versions of surveys were stored in a locked cabinet located in a locked office and did not contain identifiable information. Due to the end goal of performing cross-validation on an electronic and a pen-and-paper version of the scale, the items of the electronic scale were not randomized.

Participants:

Eligible participants were community-dwelling adults 50 years and older and were able to leave the house independently or with an assistive device. They were excluded if they reported not being fluent in English. Participants were recruited using a flyer that was distributed at grocery stores, libraries, senior centers, centers on aging, and via social media. We recruited a total of 39 community-dwelling adults age 50 years and older. Out of the 39 participants, 24 completed the electronic version and 15 completed the pen-and-paper version of the scale. All scales returned were at least 95% complete with no individual item being missed more than once, thus data were assumed to be missing at random and were analyzed without additional action, meaning they were treated as missing and not included in the analysis.

Measures:

Demographic Questionnaire:

The questionnaire solicited information about age, gender, race, ethnicity, marital status, living environment, annual household income, mobility status, and retrospective 12-month history of falls. Mobility status was examined by asking participants if they could walk 6 blocks or about a half a mile without external assistance from another person.

Physical Activity Vital Sign:

Participants were asked the following two questions: 1) “On average, how many days per week do you engage in moderate to strenuous physical activity (like a brisk walk)?” and 2) “On average, how many minutes per day do you exercise at this level?”. The score is calculated by multiplying the answer to questions one and two.⁴⁶ The scale has been validated among adults 18 years and older.⁴⁴ The scale is able to identify those who meet the 150 minutes of moderate-vigorous physical activity (κ = 0.55, P < .001).⁴⁴ Compared to accelerometer-measured PA, the PAVS is a reliable and valid measure that identifies individuals who do or do not meet the weekly recommended 150 minutes of moderate to vigorous level of PA (κ=.12, P<0.05).⁴⁷ As there is no gold-standard for assessing PA barriers and the amount of PA is negatively associated with the number of PA barriers, we used PAVS as an indirect indicator for construct validity.^48,49

IPAB:

The 148-item version of the IPAB was used to assess participants’ perception of PA barriers. Participants were instructed to read each item and circle the response that best matches their perception of how much each item impacts their PA. The response options were “Not at All” (1), “Somewhat” (2), and “A Lot” (3). The IPAB was administered with a comment box after each item and at the end of the scale. The comment box encouraged participants to provide feedback about the clarity and quality of the individual items and the IPAB as a whole. The scale’s average score was obtained by adding up the individual item scores and dividing them by 148. For additional insight about the three social-ecological domains, we calculated the average score for the items classified as personal, social, and environmental PA barriers.

Analysis:

Data analysis was performed using IBM SPSS Statistics for Windows, version 21 (IBM Corp, Armonk, NY). Items were identified as poor performers when study participants provided written comments about the items being unclear or vague. Factor analysis, the gold standard of examining a scale’s dimensionality, was not completed secondary to having insufficient data. Instead, we grouped items based on the type of barrier they addressed. Within each group, the items with a higher item-scale correlation, a wider variance or spread, and a mean near the center of the range were kept in the scale. If two items had similar item-scale correlation, variance, and mean, we kept both in the scale.²⁹ Descriptive statistics and tests of normality (Kolmogorov-Smirnov) were computed for all variables. Since our IPAB results were not normally distributed, we performed the Mann-Whitney U test to identify the scale’s ability to differentiate between active and inactive individuals. Active and inactive individuals were categorized based on if they met or did not meet the 150 minutes of weekly moderate-to-vigorous PA as determined per their PAVS scores. Internal consistency of the entire scale and each subscale (personal, social, and environmental) was assessed via Cronbach Alpha. To compare electronic and pen-and-paper version we calculated internal consistency of each administration form (via Cronbach alpha coefficient) as well as the mean differences and its effect size (via Mann-Whitney U test and Cohen’s d). For Cronbach alpha, we classified a value of 0.70–0.90 as acceptable and >0.90 as strong.⁵⁰ For Cohen’s d, we classified a value of 0.2 as a small effect, 0.5 as a medium effect, 0.8 as large, and 1.3 as very large.⁵¹ The significance level for the Mann-Whitney U test, a two-sided test, was set at p<0.05.

Phase 3: Modified Delphi Study

Design and procedure:

To further refine the scale, we implemented a modified Delphi technique. The modified Delphi technique uses a systematic method with three distinct rounds to collect professional opinions and consensus from a PA expert panel.

Participants:

A panel of 30 experts was identified based on their participation in professional organizations, such as the American College of Sports Medicine and American Physical Therapy Association, and their knowledge as illustrated by previous research publications related to PA barriers and adherence. A recruitment e-mail was sent to the 30 experts and nine agreed to participate, including physical therapists, an occupational therapist, exercise physiologists, a medical doctor, and public health officials. Due to the recommendations to keep the number of modified Delphi participants between 8–12, we did not perform additional recruitment.⁵²

Method:

Our modified Delphi study included three phases that were based on the recommendations provided by De Meyrick.⁵³

Round 1: Participants were instructed to provide general feedback about the scale items’ clarity, quality, and about potential items that should be added. The results were used to refine the scale, and the refined scale was then used for rounds two and three.

Round 2: Participants were asked to rate the importance of including each item in the final scale using a 5-point Likert Scale (‘1’ indicating strongly disagree to ‘5’ strongly agree). Consensus for including an item in the final scale was determined if ≥70% of the participants rated an item as strongly agree (5) or agree (4). Consensus for excluding an item in the final scale was determined if ≥ 70% of the participants rated an item as strongly disagree (1) or disagree (2).

Round 3: For each Delphi participant, a personalized questionnaire was developed. The questionnaire included the group’s mean and standard deviation for each item, along with any items that the participant rated greater than one standard deviation from the group’s mean rating. Using this information and their professional opinion, each expert was asked to either agree with the group or provide justification(s) about why they did not agree.

RESULTS

Phase 1: Initial Scale Development

After the review of the 148-item pool, a total of 137 items were selected for the initial IPAB which consisted of three subscales: personal (73 items), social (32 items), and environmental (32 items). Out of the 11 items eliminated, five items were eliminated due to being ambiguous, three were eliminated due to the items not applying to individuals who are already active but do not meet the recommended levels of PA, and three items were eliminated due to being confusing or too complex. Out of the remaining 137 items, 68 items were modified to prevent them from being double-barrelled, having incorrect tenses, or having double negatives. The three clinicians who reviewed the initial scale identified that the four response options (“strongly agree” [1], “disagree” [2], “agree” [3], and “strongly agree” [4]) were too complex for responders and did not correspond to all items. Therefore, we changed our response options to: “Not at All” (1), “Somewhat” (2), and “A Lot” (3). We also received feedback about the directions for the IPAB, this led to shorter and clearer directions for the respondents.

Phase 2: Cross-Sectional Pilot Data

Thirty-nine participants completed the IPAB (mean age was 58.2 [7.0] years), with 24 completing the electronic version (mean age was 58.4 [7.1] years) and 15 completing the pen-and-paper version (mean age was 58.0 [7.0] years). The overall sample was comprised of mostly female (67%), non-Hispanic white (87%), people who were married (92%), had at least a college degree (85%), and reported a gross income of greater than $150,000 (51%). See Table 2 for additional detail. Demographic variables between participants who completed the pen-and-paper version versus those that completed the electronic version were not statistically different.

Table 2.

Participant Characteristics

Characteristics	N (%) Mean (SD) (n=39)
Age, mean yrs (SD)	58.2 (7.0)
Female	26 (67)
Race/Ethnicity  Non-Hispanic White  Black or African-American  Asian  Native Hawaiian/Pacific Islander  Some other Race/Ethnicity	34 (87) 1 (3) 2 (5) 1 (3) 1 (3)
Education  High School Graduate or GED  Some College  Vocational Training (beyond High School)  College Degree  Graduate College	1 (3) 3 (8) 1 (3) 9 (23) 24 (62)
Income  $20,000–44,999  $45,000–139,999  $140,000–149,999  >$150,000	1 (3) 12 (31) 6 (15) 20 (51)
Married	36 (92)
Able to Walk 6 Blocks without Help from Another Person	39 (100)
1-year Retrospective Fall History	14 (36)
Perform 150 Minutes/Week of Moderate-to-Vigorous Physical Activity	24 (62)
Body Mass Index, mean kg/m2 (SD)	28.6 (5.4)

IPAB descriptive statistics:

For the combined sample, the scores on the IPAB ranged from 1.00 to 2.10 out of a possible range of 1.00 to 3.00, with a mean of 1.29 and a standard deviation of 0.27.

Item refinement:

Two rounds of item refinement were conducted. During the first round, we reworded all items that participants identified as being confusing or vague. During the second round, the 137 items were grouped based on the type of barrier each item was addressing, resulting in 50 groups of items. For example, one group was “usefulness and benefit.” This group contained items such as “My physical activity is limited because I don’t find it useful,” “My physical activity is limited because I have been told that physical activity will not help me,” and “I don’t believe that there are benefits to regular PA.” After analyzing items in each group, 81 items were kept secondary to high item-score correlation, variance, and a mean close to the central tendency. The 81 items contained 50 items examining personal barriers, 14 items examining social barriers, and 17 items examining environmental barriers.

Reliability:

After item refinement, Cronbach’s alpha was used to assess internal consistency for the total score and each subscale. Both the scale’s overall internal consistency (α=.97) and the personal subscale’s internal consistency were identified as strong (α=.96), the social subscale was identified as acceptable (α=.75), and the environmental subscale also as acceptable (α=.89). The reliability analyzes were completed on a combined data set, using both pen-and-paper and electronic format.

Construct validity:

We examined the scale’s construct validity by comparing two contrasting groups: those who reported performing ≥150 minutes of moderate-to-vigorous PA per week were categorized into the “met” group (n=24; mean age was 55.8 [5.2] years), and those that did not were categorized into the “not met” group (n=15; mean age was 62.0 [9.6]). There was a significant difference in IPAB scores among participants who met the PA recommendations and those who did not (p=.01). The construct validity analyzes were completed on a combined data set, using both pen-and-paper and electronic format. For more information on the scale’s psychometric properties and descriptive statistics see Table 3.

Table 3.

Descriptive statistics and psychometric properties of the IPAB

Construct	Min-Max Value*	Mean (SD)*	Standard Error of Measure	Cronbach α	Ability to differentiate between meeting/ not meeting PA (p-value)**
Personal	1.00–2.24	1.35 (.33)	.05	.96	<.001
Social	1.00–1.79	1.21 (.21)	.03	.75	<.001
Environmental	1.00–1.94	1.19 (.23)	.04	.89	<.001
Entire Scale (40 items)	1.00–2.10	1.29 (.27)	.04	.97	.01

^*Scores are calculated by taking an average of the item scores for the entire scale and each sub-scale (personal, social, and environmental). Each item’s response option were “Not at All” (1), “Somewhat” (2), and “A Lot” (3).

^**The Physical Activity Vital Sign was used to determine if individuals meet or do not meet the recommended 150 minutes of moderate-to-vigorous weekly physical activity. Those who perform less than 150 minutes were classified as not meeting and those who perform at least 150 minutes were identified as meeting the recommended levels.

Comparison of electronic and pen-and-paper version:

The internal consistency for each administration was similar (electronic α=.97 and pen-and-paper α=.96). There was no statistically significant difference between the electronic mean (SD) IPAB score (1.31 [0 .21]) and the pen-and-paper mean IPAB score (1.21 [0.23], p=.43).

Phase 3: Modified Delphi Study

Nine experts completed the modified Delphi study. During the first round, they provided valuable feedback related to needing a common stem for all items, changing the scale from a total of three choices (“Not at All” [1], “Somewhat” [2], and “A Lot” [3]) to five (“Never” 1), “Rarely” [2], “Sometimes” [3], “Often” [4], and “Always” [5]) and changing the wording on several items to decrease double negatives and improve clarity. Also, during this round, when at least 70% of the participants identified an item as being too specific, too vague, or unclear, the item was removed from the scale. This process further refined the IPAB into a 67-item scale which was then implemented for the second round of feedback. During the second round, participants met consensus on 85% of items, with 40 items identified for inclusion and 16 items identified for elimination. After eliminating 16 items, participants met a 100% consensus on removing an additional 11 items and keeping 40 items. The modified Delphi process (outlined in Figure 1) resulted in the final 40-item version of the IPAB. The final scale consists of 27 items examining personal barriers, 3 items examining social barriers, and 10 items examining environmental barriers. For a digital copy of the scale, see Supplemental Digital Content 1. After the review of the 148-item pool, a total of 137 items were selected for the initial IPAB which consisted of three subscales: personal (73 items), social (32 items), and environmental (32 items). Out of the 11 items eliminated, five items were eliminated due to being ambiguous, three were eliminated due to the items not applying to individuals who are already active but do not meet the recommended levels of PA, and three items were eliminated due to being confusing or too complex. Out of the remaining 137 items, 68 items were modified to prevent them from being double-barrelled, having incorrect tenses, or having double negatives. The three clinicians who reviewed the initial scale identified that the four response options (“strongly agree” [1], “disagree” [2], “agree” [3], and “strongly agree” [4]) were too complex for responders and did not correspond to all items. Therefore, we changed our response options to: “Not at All” (1), “Somewhat” (2), and “A Lot” (3). We also received feedback about the directions for the IPAB, this led to shorter and clearer directions for the respondents.

Figure 1.

Item Refinement Process

Phase 2: Cross-Sectional Pilot Data

Thirty-nine participants completed the IPAB (mean age was 58.2 [7.0] years), with 24 completing the electronic version (mean age was 58.4 [7.1] years) and 15 completing the pen-and-paper version (mean age was 58.0 [7.0] years). The overall sample was comprised of mostly female (67%), non-Hispanic white (87%), people who were married (92%), had at least a college degree (85%), and reported a gross income of greater than $150,000 (51%). See Table 2 for additional detail. Demographic variables between participants who completed the pen-and-paper version versus those that completed the electronic version were not statistically different.

IPAB descriptive statistics:

For the combined sample, the scores on the IPAB ranged from 1.00 to 2.10 out of a possible range of 1.00 to 3.00, with a mean of 1.29 and a standard deviation of 0.27.

Item refinement:

Two rounds of item refinement were conducted. During the first round, we reworded all items that participants identified as being confusing or vague. During the second round, the 137 items were grouped based on the type of barrier each item was addressing, resulting in 50 groups of items. For example, one group was “usefulness and benefit.” This group contained items such as “My physical activity is limited because I don’t find it useful,” “My physical activity is limited because I have been told that physical activity will not help me,” and “I don’t believe that there are benefits to regular PA.” After analyzing items in each group, 81 items were kept secondary to high item-score correlation, variance, and a mean close to the central tendency. The 81 items contained 50 items examining personal barriers, 14 items examining social barriers, and 17 items examining environmental barriers.

Reliability:

After item refinement, Cronbach’s alpha was used to assess internal consistency for the total score and each subscale. Both the scale’s overall internal consistency (α=0.97) and the personal subscale’s internal consistency were identified as strong (α=.96), the social subscale was identified as acceptable (α=.75), and the environmental subscale also as acceptable (α=.89). The reliability analyzes were completed on a combined data set, using both pen-and-paper and electronic format.

Construct validity:

We examined the scale’s construct validity by comparing two contrasting groups: those who reported performing ≥150 minutes of moderate-to-vigorous PA per week were categorized into the “met” group (n=24; mean age was 55.8 [5.2] years), and those that did not were categorized into the “not met” group (n=15; mean age was 62.01 [9.6]). There was a significant difference in IPAB scores among participants who met the PA recommendations and those who did not (p=.01). The construct validity analyzes were completed on a combined data set, using both pen-and-paper and electronic format. For more information on the scale’s psychometric properties and descriptive statistics see Table 3.

Comparison of electronic and pen-and-paper version:

The internal consistency for each administration was similar (electronic α=.97 and pen-and-paper α=.96). There was no statistically significant difference between the electronic mean (SD) IPAB score (1.31 [0 .21]) and the pen-and-paper mean IPAB score (1.21 [0.23], p=.43).

Phase 3: Modified Delphi Study

Nine experts completed the modified Delphi study. During the first round, they provided valuable feedback related to needing a common stem for all items, changing the scale from a total of three choices (“Not at All” [1], “Somewhat” [2], and “A Lot” [3]) to five (“Never” 1), “Rarely” [2], “Sometimes” [3], “Often” [4], and “Always” [5]) and changing the wording on several items to decrease double negatives and improve clarity. Also, during this round, when at least 70% of the participants identified an item as being too specific, too vague, or unclear, the item was removed from the scale. This process further refined the IPAB into a 67-item scale which was then implemented for the second round of feedback. During the second round, participants met consensus on 85% of items, with 40 items identified for inclusion and 16 items identified for elimination. After eliminating 16 items, participants met a 100% consensus on removing an additional 11 items and keeping 40 items. The modified Delphi process (outlined in Figure 1) resulted in the final 40-item version of the IPAB. The final scale consists of 27 items examining personal barriers, 3 items examining social barriers, and 10 items examining environmental barriers. For a digital copy of the scale, see Supplemental Digital Content 1. Please note that a thicker line was used to separate the personal, social, and environmental barriers.

DISCUSSION

We used a three-phase study to develop the IPAB, a PA barrier scale that examines personal, social, and environmental barriers, and thus provides a more feasible method of evaluating PA barrier(s). The IPAB is the first documented instrument that has attempted to capture the multi-factorial nature of PA barriers in community-dwelling adults 50 years and older. The constructs of the scale and the generated items reflect the domains of the Social-Ecological Model,²⁸ which were used during phase 1 to develop the IPAB’s content. In phase 2, quantitative data was used to demonstrate that the scale has the potential of being a valid and reliable scale. These initial two steps are important because for a clinical outcome measure to be useful, the instrument must be reliable, valid, and be able to examine the multiple domains that impact an individual’s level of physical activity, including personal, social, and environmental domains. Additionally, if the scale was identified as not having the potential to be valid, it would be inappropriate to continue the development process. To further refine the scale to a more feasible format, we completed phase 3, a systematic consensus method using both qualitative and quantitative data. Each item’s relevance and importance were demonstrated by a high level of agreement among experts in the field. This agreement also provides support for the content validity of the IPAB. Our iterative process led to the development of a theoretically strong scale with preliminary data indicating that the scale is highly reliable and valid. Initial findings also demonstrate the potential to use electronic or pen-and-paper IPAB versions interchangeably, but a more rigorous study examining the reliability of alternative modes is needed before recommending this approach.

Clinicians can use our newly developed multi-factorial PA barrier scale to identify why their patients are not meeting the recommended levels of PA. The self-administered scale takes about five minutes to complete and provides clinicians with an in-depth understanding of their patients’ PA barriers. Using the results, clinicians can guide their conversation to address the barriers and develop an individualized PA prescription. For example, if the barriers are the cost of going to a fitness center and the fear of being injured, clinicians can provide and educate patients about a safe PA prescription that does not require a membership to a fitness center. When prescribing PA, it is important to remember how it differs from exercise. Physical activity is any bodily movements produced by skeletal muscles that result in energy expenditure. A subset of PA is exercise, a planned, structured, and repeated behavior aimed to maintain or improve components of physical fitness.⁴ By providing individualized PA plans that incorporate solutions to PA barriers, physical therapists are more likely to increase patients’ levels of PA and thus reduce the burden of illness, enhance their patients’ quality of life and promote healthy aging.¹⁴

A major strength of this study is that the IPAB is the first multifactorial PA barrier assessment tool for community-dwelling adults age 50 years and older. Another strength was the IPAB’s development process, including item development and scale finalization. Our item-development process built on previously validated scales identified via a systematic search using the social-ecological model and therefore allowed for a list of items that were used to examine personal, social, and environmental barriers. While finalizing the IPAB, we used multiple rounds of feedback from an interprofessional group of experts. Using an interprofessional group of experts allowed for the inclusion of different perspectives and therefore we are certain that the removal of 118 items did not compromise the IPAB’s ability to examine barriers from the personal, social, and environmental domains of the social-ecological model. Developing the scale with feedback from other healthcare professionals also provides the potential of implementing the scale across a variety of healthcare professions, not just physical therapy.

Several study limitations must be considered before implementing the IPAB. The primary limitation is the size of our phase 2 sample. Due to feasibility reasons, we needed to determine the potential of the scale prior to finalizing the scale and initiating a well-powered validation study. The small sample size resulted in the inability to complete a factor analysis, an important component of scale validation. Our second limitation is the homogeneity of our phase 2 sample, which consisted of individuals who were primarily identified as women, non-Hispanic white, well-educated, married, and reported a higher-income level. The lack of diversity limits the generalizability of our results. Therefore, a large-scale validation study consisting of a more representative sample needs to be completed before determining the final scale’s validity and reliability. The large-scale validation study should include assessments of validity and reliability, including test-retest reliability, alternative form reliability, and factor analysis.

CONCLUSION

Our study used the well-established Social-Ecological model to develop the first multifactorial PA barrier inventory for adults 50 years and older (the IPAB). Our preliminary data identified that the scale appears to be reliable, valid, and has the potential to be administered in pen-and-paper or electronic formats. Clinicians can use our scale to guide their conversation about PA barriers and develop an individualized PA intervention. When implementing the scale, it is important to remember that the validation of the scale was conducted on a small homogenous sample, limiting the IPAB’s generalizability.