In older adults, physical activity is a necessary action to provide and maintain health.1 A substantial amount of empirical evidence has demonstrated that the health benefits associated with participation in physical activity can maintain functional independence and ultimately improve quality of life.2-.3 In addition, participation in physical activity improves self-esteem, mental alertness and social interaction, and decreases levels of depression.4-5 Physical activity is also a major independent modifiable risk factor that has a protective effect against the onset of cardiovascular disease, ischemic stroke, type 2 diabetes, and cancer.6-7 However, a preliminary report in 2007 by the National Statistical Office of Thailand showed that more than 58% of Thai elderly engaged in physical activity less often than recommended for good health.8 Therefore, it is imperative that physical activity among the aged population is evaluated.

Get Help With Your Essay
If you need assistance with writing your essay, our professional essay writing service is here to help!
Essay Writing Service

A review of physical activity studies reveals that the physical environment is significantly associated with physical activity participation. According to Bandura, physical environment plays either the role of facilitator or obstacle in human motivation through the interaction between cognitive functions and environment.9-10 Both natural features and human constructs related to environment may affect physical activity engagement.11 Physical environment positively influences physical activity within older people.12-18 While the significance of physical environment is well recognized, an accurate measure is needed to identify the specific characteristics of the physical environment for physical activity, with respect to older people. If there is low error in the assessment, the explanation of the relationship between physical environment and physical activity will be raised.
The concept of behavior setting helps to elucidate the influence of physical environment on physical activity.19 Behavior occurs within a physical and social context. So, the behavior of older people will be influenced by environment the neighborhood, community, or home environment. The home environment is a primary setting for the performance of daily activities among older people.20 In Thailand, most older people stay and participate in physical activity in their home. Although previous studies have assessed only convenience of facilities and access to equipment at home related to physical activity, the characteristics of home environment is the least studied potential determinant of participation in physical activity.11-12 Existing investigations use either neighborhood-focused scales to determine physical environment 14,21-22, or have applied both neighborhood and availability of facilities in the community.18,23 A review of relevant research indicates that aspects of the physical environment, such as safety, traffic volume, street lighting, unattended dogs, having a sidewalk, and accessibility to public recreation places influence physical activity,14-18 whereas other evidence suggests equivocal results.21-22,24 At least one potential reason exists for this inconsistency which may be related to the physical environment measurement, particularly with older adults.
Most physical environment questionnaires were developed in western countries. These measurements have involved different neighborhood and community settings; assessed in different dimensions such as convenience, safety, accessibility, and facility. Even though the physical environment questionnaires have been validated in western countries, they have not been tested in others settings. Due to differences of geographic features, culture, and patterns of living, a need exists to assess the physical environment of Thai elderly to confirm reliability and validity in this cultural context. Furthermore, the relationship of the physical environment of home and physical activity of older Thai people is still unknown. Understanding features of the physical environment related to physical activity may provide an effective implementation design to motivate older Thai people to participate in physical activity. For these reasons, a modified instrument, “the Thai Environmental Support for Physical Activity for Older Thai people (TESPA)” for assessing the physical environment including home, neighborhood, and community environment needs to be modified and validated for the older Thai population.
This study addressed this shortcoming by modifying and confirming a three-factor structure for physical environment in a sample of older adults in Thailand. We hypothesized that a correlated three-factor structure for the TESPA would fit the data well and that these factors would possess good internal consistency. Additionally, we hypothesized that higher scores on the subscales of the TESPA would be significantly and positively correlated with higher levels of physical activity. Moreover, our study purpose was to prepare physical environment questionnaire for a large descriptive study related to physical activity in older Thai people.
Objective
The purpose of this project was modification and assessment of the TESPA scale in older Thai people to provide a reliable and valid measure that is culturally congruent and useful for future research.
Measures
Demographics. -A personal data sheet was used to obtain demographic and socioeconomic data including age, gender, income, marital status, education level, area of living, the length of living in their residence and medical history.
The Chula Mental Test (CMT), an interviewing questionnaire developed by Jittapunkul, and colleagues was administered to determine the cognitive function of older Thai people who has difficulties in reading and writing.25 The CMT consists of 13 items related to cognitive function. Scales are coded on a dichotomous score of 0 (incorrect) and 1 (correct); items 5 and 12 have two sub-scales, and items 3 and 13 have three sub-scales each. Total scores indicate the cognitive function and range from 0-19. Scores 0-4 illustrate severe cognitive impairment, scores 5-9 depict moderate cognitive impairment, scores 10-14 reveal mild cognitive impairment, and scores 15-19 demonstrate normal cognitive function.
Physical activity. -Physical activity was assessed by using the International Physical Activity Questionnaire – Long form (IPAQ-L). The IPAQ-L was developed by the International Consensus Group for the Development of an International Physical Activity Questionnaire at the WHO in 1998.26 The IPAQ-L includes 5 parts: work-related activities, transport-related activities, domestic activities, and time spent sitting during the previous 7 days. In addition, the IPAQ-L assesses the frequency, intensity and duration of all daily physical activity. In summary, total physical activity equals the MET score, which is the sum of minutes spent in each domain multiplied by the MET value.26 We used a cut-off point 600 METs, as recommended by the Centers for Disease Control and Prevention of the United States of America and the American College of Sports Medicine.27 Threshold values for the IPAQ-L in the present study were categories insufficiently active ( 600 METs-min-week)
We translated and adapted the IPAQ-L to fit the habits of older Thai people; content validity was determined by three experts. The content validity index of the IPAQ-L was .96. Numerous studies testing the test-retest reliability of the IPAQ-L revealed results ranging from 0.63 to 0.91, which indicated good repeatability.26,29 In the present study, the stability of the IPAQ-L questionnaire was reported to be 0.77 in 30 Thai elderly.
Methods
A cross-sectional design was used in the current study. The process of modification of the questionnaire included two phases (see Figure 1). In phase I, the instrument was modified as follows: 1) questionnaire improvement stage and 2) quantification stage. In phase II, instrument assessment was conducted to examine construct validity by confirmatory factor analysis and the known-group method.
Insert figure 1 here
Phase 1: Modification of instrument
Stage 1: Questionnaire improvement stage consisted of measurement review, translation, instrument refinement and item construction.
Measurement review and translation procedure:
Physical environment is defined as older people’s perception of the physical environment to facilitate or hinder physical activity engagement in three settings: home, neighborhood, and community. We reviewed the relevant research regarding measurement of physical environment from the literature. The existing measurement was chosen based on definition and psychometric properties of measurement. After obtaining written consent from the author, the original questionnaire was translated into Thai versions by the researcher and an independent translator according to the translation-back translation method.28 The Thai version was evaluated by three Thai/English bilingual people. The questionnaire was translated back into English by two Thai-English independent translators who each had taught English to graduate students for more than 20 years. We then compared both versions in the original language, conducted checks with the translators, discussed the differences, and produced a final consensus version.
1.2 Instrument refinement and item construction:
We modified the translated instruments to achieve a closer cultural fit for older Thai people. During October to December 2007, a preliminary study was conducted with ten elders: five who lived in a municipal area and five in a non-municipal area. The participants were selected from a broad range of backgrounds: five elders had elementary education and had worked in the agricultural sector, three elders had secondary education and had worked as small businesses owners, and two elders were retired and held bachelor’s degrees. Open-ended interviews were applied to ensure that instrument content and language were suitable for Thai elderly. They were interviewed in their home or a temple in their village. The participants were encouraged to share their opinions regarding the relevancy of items and appropriateness to the culture of older Thai people. Additionally, participants were encouraged to think of additional items that potentially could be used in the questionnaire. Each participant was interviewed twice for 30 minutes for each time, or until no new data occurred. In the preliminary study, no participant refused to participate.
Interviews were audio taped and transcribed verbatim. Categories and coding were derived from data sources and previous related studies. Statements by participants about the physical environment were delineated and identified as a content domain. Representative phrases and terms were marked to be used as potential items so that the language of the participants could be preserved. Additional items then were discussed with two Thai experts and an American expert in geriatric nursing. As a consequence, the measurement was modified.
Stage 2: Quantification stage
This stage involved the validity and reliability of the psychometric properties of the modified measurement. Content validity of the scale was evaluated by three geriatric experts including one physician in geriatric physical activity, one expert in geriatric community nursing, and one expert in geriatric nursing. The experts were asked to rate the level of relevancy between the items and the definition of the concepts as represented. A four-point Likert-type scale ranging from 4 (strongly relevant) to 1 (strongly irrelevant) was used to rate each item.
Data collection started in October 2008 after obtaining approval from the Institutional Review Board at Chulalongkorn University, Thailand. Both written and verbal informed consent was obtained in Thai on the same date as the data collection. The informed consent form explained the purpose of the study, benefits, risks, the types of questionnaires and tasks to be completed, and the length of time needed to complete the interview. In particular, it explained about risk prevention and treatment when the risk may occur during the interview or when collection of data is taking place.
Permission was obtained from participants prior to data collection. At the setting, the participants were informed about the purpose of the study and their right to refuse participation. If participants chose not to answer the questionnaire, they could withdraw from the study at any time without penalty. They were also notified that their relationship with the health care team would not be affected. Their names were not used; instead, a code number was used to ensure confidentiality. There was no harm to the participants in this study.
In addition, to assess the feasibility of using psychometric properties, the modified measurement was determined in the pilot study. The consent was obtained from the directors of primary care units, one in an urban area and the second in a rural area. A purposive sample of 15 older people from each setting was recruited in the pilot study; no respondent refused or dropped out in this stage. The participants were older people, with a mean age of 70+ 4.19 years. Most participants were female (76.7%), married (53.3%), had elementary education (80%), were employed (62.6%), with a household income of less than 5,000 Baht per month (approximately US$147) (76.7%). A substantial proportion (63.3%) lived in urban areas and had lived on average for 44.6 years in their residence. Of the sample, more than half reported sufficient physical activity level (50%), whereas 20% had a low physical activity level. A total of 23.3 % reported having no current health problems, while 16.7 % had hypertension. The most frequent type of physical activity reported was household related activity, followed by leisure time activity, transportation related activity, and occupational activity. Stability of reliability was obtained in two weeks, whereas internal consistency was assessed at baseline.
Phase 2: Assessing the Instrument
Construct validity of the modified scale was determined in the main study. Multi-stage random sampling was employed to obtain a sample of 336 elderly (aged 60 years and older) residing in 12 villages from six provinces of Thailand, who were not part of the preliminary study or the pilot study. In each village, 28 participants were selected by a systematic sampling technique from a name list obtained from the village’s primary care unit. A simple random technique was applied and only one member in each family was included in the study. The participants were determined to be eligible to participate in the study if they scored >15 on the CMT, were able to ambulate without assistive devices, and were willing to participate in the present study.
Data were collected from November 2008 to April 2009. An authorization letter was sent to officers of the primary care unit in all 12 settings to ask for their consent. After obtaining their consent, the public health nurses of the primary care units were asked by the researcher to make appointments with participants. When verbal agreement was obtained, the participants were asked to sign a consent form. The modified questionnaire was used to conduct interviews lasting between 15 to 20 minutes; each participant received a handkerchief in appreciation for their participation.
A total of 336 questionnaires were selected for accuracy of data entry. Statistical analysis showed that two cases with a single or more than one missing value on community environment were deleted, leaving 334 cases for analysis. According to IPAQ Research Committee guidelines, the physical activity scores were processed to reduce data comparability. Ten cases were excluded by the truncation process due to the total duration value being more than 3 hours per activity, and four cases were eliminated because of multivariate outliers. Therefore, 320 cases remained for analysis.
Most of the participants in the main study were female, married, had elementary education, and a household income less than 5,000 Baht per month (approximately US$147). Approximately 70% of the participants reported at least one health problem. Shared living was the most frequent living arrangement reported (92.8%) with an average of 4.11 persons per household. A substantial proportion (61.9%) lived in an urban area and stayed in their own residence.
Data analysis
Descriptive data are presented as mean + SD. The internal reliability of the scale was based on an alpha coefficient greater than or equal to .70; 30 stability of the scale was analyzed by product correlation coefficient. Reliability of each item, overall reliability, and construct validity of the scale were determined using structural equation modeling (SEM).31 The known-group technique was conducted using the multivariate analysis of variance, to compare the physical environment of those who reported sufficient physical activity and those who did not. Statistical significance for analyses except SEM was defined as pResults
Phase 1: Modification of instrument
Stage 1: Questionnaire improvement stage
Following determination of the validity and reliability of the instrument, results were used to make modifications. The following procedures were undertaken.
1.1 Measurement review and translation procedure:
Findings from prior studies demonstrated that most physical environment measurements were designed to identify specific characteristics of the built environment (e.g. distance between destinations, presence of sidewalk), and showed moderate to high reliability. Blocks of questions appeared to have different reliability among urban and rural respondents. Only one measurement was constructed for older people.12 A meta-analysis study by Duncan and colleagues concluded that the environmental characteristics in measurement presented in physical activity include facilities, sidewalks, shops and services in walking distance, heavy traffic, high crime, street lighting, and unattended dogs.32
The Social-physical Environmental Supports for Physical Activity Questionnaire (ESPA) is a measurement that designed to capture and assess the supporting social and physical environment for physical activity typically performed by all age.33 The ESPA was selected to collect data in this study because it is closely congruent with the Thai context, and indicates both neighborhood environment and community environment. In addition, the coefficient differences between urban respondents and rural respondents of this scale were small when compared with the other questionnaires.34 Moreover, the previous validity and reliability value of ESPA was acceptable.34 The ESPA is composed of two domains: the social and the physical environment domains. The physical environment is composed of 20 items: 10 neighborhood items (access, characteristics, barriers, use), and 10 community items (access, and barriers). A Likert-scale was used to assess neighborhood items, except for an item on public recreation facilities which allowed response options as 1(yes) or 0 (no). The community items have response options of 1(yes) or 2(no), with score form items on recreation facilities, whereas a community item used a Likert Scale. The higher the summary score the stronger the physical environment. The ESPA questionnaire was translated into Thai.
Instrument refinement and item construction:
All participants recommended that some items of the ESPA questionnaire be deleted and many remarked that the questionnaire format be reviewed. Most participants suggested that 21 physical environment items be eliminated – including public swimming pools, sidewalks, parks, walking trails, bike paths, recreational centers, shopping malls (sometimes used for physical activity or walking programs), and being a private member of a recreational facility . These items reflected physical activities and sites that were uncommon in Thailand and thus were not deemed to be valid it this cultural context. As well, five pages of the scale and various types of choices took up too much time and were difficult to answer; for example:
“In general, would you say that motorized traffic in your neighborhood is…. Heavy, Moderate or Light.”, ” When walking at night, would you describe the STREET lighting in your neighborhood as…..Very good, Good, Fair, Poor or Very poor” and “How safe are the public recreational facilities in your community? would you say… Very safe, Somewhat safe, Somewhat unsafe or Not safe at al”.
Additionally, the literal translation of the word “physical activity” into Thai git-ja-gam-taang-gaa or กิจกรรมà¸-างกาย (git-ja-gam = activity; taang-gaai= physical) was unfamiliar to older people. Most participants thought this term was difficult to understand and felt that it was not applicable to them. The elders preferred the term kleuan-wai-ok-raeng or เคลื่อนไหà¸à¸­à¸­à¸à¹à¸£à¸‡ (kleuan-wai= movement; ok-raeng= expend energy) instead to define physical activity. The term “kleuan-wai-ok-raeng” was therefore applied to this study.
The interview findings illustrated that the statements contribute to physical environment among participants was delineated and identified as a content domain. Home was most frequently mentioned as a favorite place for engagement in physical activity. The majority of participants expressed that they generally preferred engaging in physical activity at home because of safety concerns and convenience. Representative phrases and terms were marked to be used as potential items so that the language of the participants could be preserved. As a result, the home environment subscale was formulated, and it was consisted of safety and convenience both inside and around home.
Based upon the experts’ suggestions, 21 items that were considered irrelevant on the ESPA were eliminated and four additional items related to home environment were added to the ESPA. Therefore, the modified scale was called “Thai Environment Support for Physical Activity in older Thai people (TESPA)”. A Likert scale was used to assess physical environment for physical activity. Possible responses were 1(strongly disagree), 2(disagree), 3(neither agree nor disagree), 4(agree), and 5(strongly agree). The possible ranges of scores for home, neighborhood, and community sub-scale were 4-20, 3-15, and 3 – 15 respectively. The total score was calculated by computing the numerical ratings for each answer. The possible scores ranged from 10 to 50. Higher scores indicated a higher level of perceived physical environment supportive of physical activity.
In short, the TESPA was conceptualized to include three subscales, with the six items serving as supportive neighborhood and community environments, and the four new items as supportive of home environment. Neighborhood environment was defined as the area around their home to which they could walk within 10 minutes.23 A supportive neighborhood environment referred to older people’s perceptions of support including: characteristics, access, and barriers of physical activity; a pleasant neighborhood for walking; low traffic volume; and lack of unattended dogs in their neighborhood.23 Secondary, community environment is defined as the area contained within a 20-minute drive from the respondent’s home.23 Community environment support refers to older people’s perceptions of convenience and safety of physical activity in their community: convenience of facilities of the Primary Care Unit; access to parks, playgrounds, and sports fields; and the safety of public recreation facilities.23 Finally, home environment support is defined as participants’ perceptions of convenience and safety regarding their home environment, both in and around their home, in relation to their participation in physical activity.
Stage 2: Quantification stage
The Content Validity Index of the TESPA questionnaire was 0.92. The reliability coefficients of TESPA scale was 0.73 and the test-retest reliability of scale was .76. The results of the pilot study demonstrated that respondents took between 15 to 20 minutes to complete the questionnaire. The measurements were culturally appropriate for older Thai people and the procedures were followed without any difficulty.
Phase 2: Assessing the Instrument
The physical activity score ranged from 0 to 2203.50 MET-minute/week with a median of 849.25 (SD = 438.63), and the interquartile rang of 670.13 MET-minute score. The skewness coefficient (.26) and the kurtosis statistic (-.52) indicated that the majority of the subjects reported a moderate physical activity score and a close proximity to a normal distribution. Also, 65% of the participants of this study were sufficiently active and 35% were not. The total sum scores of physical environment ranged from 19.00 to 50.00, with a mean of 34.87 (SD = 6.47). The skewness value (.26) and the kurtosis value (-.49) indicated that the majority of the respondents had moderate physical environment scores and the variance was distributed normally.
Construct validation was confirmed by confirmatory factor analysis, and the known-group technique. Based on confirmatory factor analysis, the findings demonstrated that the construct of TESPA was composed of three underlying subscales: home, neighborhood, and community environment. The correlation among items ranged from .02 to .77 and the total scale could explain 61.01% of the variance of physical environment. The home, neighborhood, and community subscale could account for 29.4%, 19.7%, and 12.0% of the variance, respectively.
The next analysis tested the three-factor model. The measurement model testing was designed to estimate which ten items were used as indicators for the model. Although the original model was statistically significant, the model was not consistent with the data /df= 4.59 and RMSEA more than 0.05 (= 151.54, df= 33; pInsert figure 2 here
Table 1 illustrates the loading with t-values and squared multiple correlation coefficients among each observed variables for TESPA scale. The results revealed that all indicators of the TESPA measurement had significant low to high parameter estimates, which were related to their specific constructs and validated the relationships among observed variables and their constructs. The squared multiple correlations for observed variables of the latent variables were ranged from 0.02 to 0.92. The R2 of item 2, 3, 4, 5, and 6 were acceptable indicators, except for the item 1, 7, 8, 9, and 10 which were less than .40.
Insert Table 1 here
The known-group technique is an examination of relationships based on theoretical prediction. 35 Table 2 shows the mean values for each of the TESPA subscales and physical activity subgroups for the complete sample. Those who demonstrated sufficient physical activity had significantly higher physical environment scores along all three subscales and total score. In addition, each subscale was positively correlated with participation in physical activity including neighborhood (r=.30, pInsert Table 2 here
Discussion:
Testing of the TESPA measurement model in the current study provided additional evidence for the validity and reliability. The findings are discussed in the following section.
The content validity and construct validity of the newly designed TESPA scale were accepted. Continued support for the construct validity of the scale was also provided through confirmatory factor analysis and the known-group method. The TESPA measurement model demonstrated that all sub-scales of the measurement had significant low to high parameter estimates, which were related to their specific constructs and validated the relationships among observed variables and their constructs. Within the known-group method, the findings demonstrated that physical environment was significantly correlated with physical activity. This finding indicated that older people with high perceptions of safety, convenience, and accessibility in their home, neighborhood, and community environment were more likely to participate in physical activity.
Similarly, active older people who met guideline recommendations for good health scored significantly higher in each of three factors – supportive home, neighborhood, and community environment – than those who did not. Physical environments affect the participation in behavior by means of the interaction between cognitive functions and environment through human motivation.9-10 Bandura argued that a better environment provides a great opportunity to perform a behavior.10 According to the empirical data in the present study, a plausible explanation for this result contributed to characteristics of the Thai elderly. The majority of the elders sampled have lived in their home for a median 30 years; consequently, they were familiar with the physical characteristics of their environment in and outside the home. It is possible that the friendly environment contributed to motivate them for physical activity engagement, while unfriendly environments discouraged activity.
For reliability, although the R2 for item 1, 7, 8, 9 and 10 indicated that these items should be considered irrelevant for the TESPA scale, the measurement model was a good fit with the empirical data. Not only was this scale firstly modified and validated in older Thai adults, but also normative data for comparison in the elderly or other study were not available. Approximately 61 % of the variance in the TESPA scale was explained by the 10 items, whereas 39 % of the variance in this scale remains unexplained. Corresponding with the SCT approach, Bandura argued that nearly all aspects of the physical environment can influence physical activity.36 Owing to the fact that the TESPA measurement includes only safety, convenience, and accessibility aspects of physical environment, it is possible that other aspects may contribute more to physical activity than these aspects. Other aspects of the physical environment may still influence engaging in physical activity; a need is indicated for work to identify these aspects yet unidentified. Therefore, continued evaluation of the psychometric properties both in the other sample and new additional items are necessary to confirm this study.
Limitations and recommendations:
This study was limited by homogeneity of the sample. The majority of participants was female, married, low socioeconomic status, and lived in their home. Continued testing of the TESPA scale, particularly with socio-economically diverse older adults, is needed to insure that this measure is consistent. Additionally, using objective measurement should be considered, to further add to the validity of the findings and confirm the subjective report. Further, the item related to community environment should be explored via a qualitative study to explore and understand with the physical environment of Thai elderly.
Nevertheless, based on factors of the TESPA, manipulation of these factors with cognitive behav