nueva página del texto (beta)
Inglés (pdf)
Artículo en XML
Referencias del artículo
Traducción automática
Enviar artículo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
INTRODUCTION
Comcáac ethnic group, better known in Mexico as Seri Indians (Alaez et al., 2002), is a non-agricultural group of people that originally lived under a traditional lifestyle. In the past, their food was mainly obtained by fishery with harpoons, complemented with collection of fruit from desert plants and hunting animals. They used to be a semi-nomadic group whose sedentarisation was influenced greatly by establishment of a fishery cooperative in 1939. Nowadays, this group is settled in communities of Punta Chueca and El Desemboque, located in northwestern Mexico, and has experienced a certain degree of modernization and acculturation; fishery is done using boats with outboard motors and food is mainly obtained from local and surrounding village’s shops. Commercial fishery, craft sale and ecotourism and hunting projects are their principal commercial activities (Luque & Robles, 2006; Noriega-Orozco, 2007; Rentería-Valencia, 2007). This changes might be associated to a decrease in physical activity (PA) and consequently, to increasing obesity and diabetes problems (Hu, 2011; Villela & Palinkas, 2000). There is strong evidence that physical inactivity is a risk factor for hypertension, diabetes, heart disease and some types of cancer (Bassett, Wyatt, Thompson, Peters & Hill, 2010; Ono et al., 2007). Furthermore, PA is well known as an important component of a healthy lifestyle to prevent obesity and a variety of chronic diseases (Assah et al., 2011; Hong, Trang, Van der Ploeg, Hardy & Dibley, 2012; Neilson, Robson, Friedenreich & Csizmadi, 2008; Siebeling, Wiebers, Beem, Puhan & Ter Riet, 2012; Tamioka, Iwamoto, Saeki & Okamoto, 2011). According to World Health Organization, almost half of adult population in developing countries, such as Mexico, doesn't perform enough PA to achieve its health benefits (Macfarlane, Chan & Cerin, 2010; Medina, Barquera & Janssen, 2013).
Physical activity questionnaires (PAQ) are the most common methods used to evaluate PA in population studies (Kriska et al., 1990), because of its low cost and practicality (Haskell & Kiernan, 2000; Helmerhorst, Brage, Warren, Besson & Ekelund, 2012; Hong et al., 2012), a greater flexibility in behavioral description of PA patterns and because it can precisely classify people (Assah et al., 2011; Ono et al., 2007).
Precise measurement of PA in epidemiological studies still represents a great challenge in terms of viability and validity (Assah et al., 2011). Since PA recalling is a complex cognitive process that can generate mistakes in interpretation of questions due to cultural differences in activities and terminologies, it is essential to ensure a safe and valid adaptation (Macfarlane et al., 2010).
Some studies have evaluated reproducibility of physical activity questionnaires to ensure they are useful tools in epidemiological studies. Kriska et al. (1990) evaluated reproducibility of a PAQ in the U.S. Pima Indian community. Further Nang et al. (2011) evaluated reproducibility of two PAQ: International Physical Activity Questionnaire (IPAQ) and Singapore Prospective Study Program Physical Activity Questionnaire (SP2PAQ). Likewise, Bae, Cho & Son (2015), similar to other studies, applied test-retest method using Korean Version of Neighborhood Physical Activity Questionnaire.
Other studies have evaluated physical activity questionnaires reproducibility in Mexican population. Medina et al. (2013) evaluated correlation of IPAQ in Mexicans and Mexican-Americans living in United States. Moreover, Vega-López, Chavez, Farr & Ainsworth (2014) evaluated two questionnaires: Stanford Brief Activity Survey (SBAS) and Rapid Assessment of Physical Activity (RAPA). Since, there is not a PAQ adapted for Seri people, the objectives of this study were to adapt and evaluate reproducibility of a questionnaire to estimate PA patterns, with the purpose of using it to study association of PA with obesity, diabetes, insulin resistance and metabolic syndrome in future investigations in this population.
Subjects
Men and non-pregnant women within 20 or more years old, members of Comcáac community from the State of Sonora, were invited to participate in this study through home visits. They received an explanation of the objectives of the study and were extended the invitation. Protocol was approved by ethics committee of Research Center for Food and Development (Centro de Investigación en Alimentación y Desarrollo, A.C.). All participants signed informed consent form.
Adaptation
Questionnaire used in this study was originally proposed by Kriska et al. (1990), which was adapted by Esparza et al. (2000) for Pima population of Sonora, Mexico. Questionnaire process of adaptation was carried out using information obtained through 15 interviews to people from Comcáac community (9 women and 6 men) who were invited to participate through a health promoter. Participants were first visited at their home and then interviewed about their leisure (defined as those activities not related with monetary compensation) and occupational activities (defined as those activities related with monetary compensation) performed as part of their daily life. Home activities were also considered as occupational activities (Esparza et al., 2000). All answers and observations were written down in a log book. Afterwards, those activities that were consistent among participants were identified and used for adaptation of questionnaire by including them in PAQ (Esparza et al., 2000). Additionally, an open question that considers any new important activity in their usual physical activity pattern was also considered in PAQ.
Reproducibility
Adapted questionnaire comprises two sections to register activities, one for previous-year and another for previous-week activities. In previous-year section, participants were asked about number of months peryear, average times per month and average hours per time their activities were performed. In previousweek section, proposed questionnaire evaluates PA by enquiring each participant whether previous-year reported activities were also performed during the last previous-week (days per week and hours per day time). Physical activity questionnaires were applied twice, with a margin of 3 - 5 weeks in between each one by properly trained personnel following a preestablished protocol calling test-retest method (Chun, 2012; Craig et al., 2003; Kriska et al., 1990).
Statistical analysis
To evaluate reproducibility of adapted PAQ, activities obtained from the two questionnaires were analyzed using Spearman coefficient ρ and intraclass correlation coefficient ICC, for both previous-year and previous-week sections of questionnaires. Evaluated variables were analyzed as hours-per-week and metabolic equivalents-hour-week (METs) either as total leisure and occupational activities or as group of activities (leisure activities: sport; occupational activities: home related, craftwork manufacture related, fishery related).
RESULTS
Subjects
Questionnaire was initially applied to 35 subjects (17 women and 18 men). Home visits were carried out with assistance of community's health promoter. However, due to reasons external to the study, only 26 participants, 11 men and 15 women, completed their second PAQ after 3 to 5 weeks apart from first application. Time to complete PAQ interview took approximately 30 min.
Average age was 38.2 years, with a minimum of 20.6 years and maximum of 60.9 years. Anthropometric characteristics such as weight, height, body mass index (BMI) and waist circumference are shown by sex in table 1.
Table 1 Anthropometric characteristics of participants by sex.
| Men (n = 11) | Women (n = 15) | |
|---|---|---|
| Age (year) | 38.3 ± 8.6 | 38.2 ± 10.6 |
| Weight (kg) | 80.6 ± 13.9 | 66.5 ± 10.1 |
| Height (cm) | 176.4 ± 4.7 | 162.4 ± 7.4 |
| BMI (kg/m2) | 25.8 ± 3.6 | 25.4 ± 5.2 |
| Waist circumference (cm) | 89.3 ± 8.2 | 88.1 ± 11.5 |
Data shown as mean ± SD.
Source: Author’s own elaboration.
Adaptation
Leisure activities such as watching television, attending church, walking or practicing sports, among others, as well as occupational activities related to craftwork manufacture and fishery activities were identified during process of adaptation. Regarding home activities, we found cooking, sweeping and mopping the floor, doing dishes, and laundry. List of activities by category (leisure or occupational) is shown in table 2.
Table 2 Lists of leisure and occupational groups of activities.
| Activities (METs) | ||
|---|---|---|
| Sports group - leisure | Washing by hand (4.00) | Counter (1.50) |
| Walking (3.80) | Sweeping the floor (4.00) | Arranging product (1.80) |
| Jogging (6.50) | Mopping the floor (4.00) | Carrying boxes (2.00) |
| Soccer (7.00) | Watering plants (1.50) | Fishery group |
| Other activities | Wood cutting (6.50) | Pen shell fishery (12.00)* |
| Watching television (1.30) | Cleaninng the bathroom (2.50) | Crab fishery (4.00) |
| Attending church (1.80) | Craftwork group | Net fishery (4.30) |
| Hunting (6.50) | Limberbush baskets (2.00) | Other activities |
| Driving (2.80) | Necklaces and bracelets (1.80) | Student (1.30) |
| Home group -occupational | Carving of figure (3.00) | Monitor wildlife (5.30) |
| Cooking (2.80) | Hand sewing (1.80) | Hunting guide (5.30) |
| Washing dishes (2.50) | Store group | Musician (2.00) |
METs: metabolic equivalents. *Diving.
Source: Author’s own elaboration.
In addition to home related activities, two main groups of occupational activities were identified: a) those related to craftwork manufacture, which includes activities related to production of limberbush baskets, bracelets and necklaces, and figure carvings made of ironwood or stone; b) those groups related to fishery activities, mainly pen shell, crab and fish.
Interviews with key people as part of adaptation process was a fundamental step to identify in detail community's main traditional activities in order to list them in adapted questionnaire. This process allowed subsequent questionnaire applications focusing on activities related to their cultural context. Identification and description of activities and sub-activities were essential for subsequent assignment of their metabolic equivalents (METs) and to classify them as sedentary, light, moderate or intense. Therefore, METs values that best represent each of identified activities (sub-activities) were assigned based on Ainsworth's compendium. For activities not listed in the compendium, values were assigned by looking at similar activities. A MET represents oxygen consumption rate, approximately3.5 ml/kg • min, for an average adult who is sitting quietly, whereas 6-MET requires 6 times resting energy expenditure (Ainsworth et al., 2011).
Reproducibility
Twenty-six subjects who were originally interviewed also completed their second questionnaire. Associations of total previous-year and previous-week leisure and occupational activities are shown on table 3.
Table 3 Correlation for total hours/week between the first and the second PAQ applications (Spearman coefficient ρ and intraclass correlation coefficient ICC).
| ρ | ICC | |
|---|---|---|
| Leisure activities (hours/week) | ||
| Previous year | 0.81 (0.82)* | 0.66 (0.60)* |
| Previous week | 0.79 (0.76)* | 0.75 (0.70)* |
| Occupational activities (hours/week) | ||
| Previous year | 0.58 | 0.49 |
| Previous week | 0.54 | 0.63 |
All values are significant (p ≤ 0.05). ()* Correlation after walking activity is excluded. PAQ: Physical Activity Questionnaire.
Source: Author’s own elaboration.
Reported values represent Spearman coefficient ρand intraclass correlation coefficient ICC between activities reported from first and second PAQ applied and expressed in hours per week (h/wk). First and second measurements of total previous-year leisure activities were highly correlated. Spearman correlation were almost the same when "walking" was excluded from total leisure activity (ρ= 0.81 including walking vs. ρ= 0.82 excluding walking; p < 0.05 for both cases). Moreover, first and second measurements of total previous-week leisure activities showed higher correlation when "walking" was included, in comparison to that found when it was excluded (ρ= 0.79 and ρ= 0.76, respectively; p < 0.05 for both cases). Regarding occupational activities (table 3), correlation was higher for that reported during previous-year in comparison to that reported during previous-week, with Spearman ρ values of 0.58 and 0.54, respectively; p < 0.05 for both cases.
Table 4 shows results of Spearman coefficient ρ and intraclass correlation coefficient ICC analysis by groups of activities (leisure and occupational) and by sub-groups of activities.
Table 4 Correlation for hours/week between the first and the second PAQ application for group of activities (Spearman coefficient ρ and intraclass correlation coefficient ICC).
| Previous year | Previous week | |||
|---|---|---|---|---|
| ρ | ICC | ρ | ICC | |
| Leisure activities (hours/week) | ||||
| Sports | 0.61 | 0.68 | 0.54 | 0.53 |
| Occupational activities (hours/week) | ||||
| Home | 0.90 | 0.83 | 0.83 | 0.75 |
| Craftwork Manufacture | 0.97 | 0.65 | 0.74 | 0.57 |
| **Fishery | 0.93 | 0.85 | 0.78 | 0.58 |
| Pen shell | 0.89 | 0.80 | 0.79 | 0.93 |
| Crab | 0.90 | 0.89 | 0.50 | 0.26 |
All values are significant (p ≤ 0.05). **Fishery of pen shell and crab together.
Source: Author’s own elaboration.
Activities corresponding to the same type were grouped in a representative category (sports, home related, craftwork manufacture and fishery related). Results showed stronger correlations in previous-year activities group in comparison to that reported during previous-week. Spearman ρ correlation for previous-year “home related” activities were 0.90 and 0.83 for that of previous-week. Additionally, Spearman ρ correlation for previous-year sub-group of “craftwork manufacture” activities were 0.97 and 0.74 for that of previous-week. Similar to previousyear, Spearman ρ correlation for “fishery” group activities were 0.93 and 0.78 for that of previous-week. Results showed high reproducibility of adapted PAQ.
Correlation values above described were obtained expressing physical activity as hours per week. However, similar correlations values were observed when physical activity was expressed as METs-hour-week (data not shown). Previous-year Spearman ρ correlations for leisure activities including and excluding “walking” were 0.83 and 0.89, respectively. Correlations for previous-week of leisure activities including and excluding “walking” were 0.81 and 0.84, respectively. Regarding occupational activities, a Spearman ρ correlation value of 0.70 for previous-year and of 0.51 for previous-week were found. In craftwork manufacture and fishery related activities cases, Spearman ρ correlations were 0.97 vs. 0.73 and 0.94 vs. 0.74, for previous- year and previous-week, respectively.
DISCUSSION
During adaptation process, common occupational activities among subjects were identified. Occupational activities were related to craftwork manufacture, home activities, and fishery. Each group of activities is further integrated by a series of sub-groups of activities that were important to know, break-down, and describe. This adaptation process made possible to place items of questionnaire in cultural context and identify relevant activities of this indigenous population (São-João et al., 2013).
An important recommendation for those who want to use a PAQ is to adapt it in order to ask people about activities they are used to do and also to be able to assign metabolic equivalents to activities as accurate as possible. This means that adapted questionnaires for a particular community may not be generalized to other populations. Some words and meanings may be interpreted differently among subjects, age groups, and cultures. Furthermore, physical activity patterns may be different among communities due to different factors that should be taken into account before PAQs are applied.
In studies where reproducibility of PAQ was evaluated, similar correlations have been found. Kriska et al. (1990) found Spearman correlation values of ρ= 0.92 and ρ= 0.62 for previous-year and previous-week leisure activities in 21 to 36 years old (n = 29) age group. In addition, in 37 to 59 years old (n = 17) age group, correlation values of ρ= 0.88 and ρ= 0.77 were found for previous-year and previous-week leisure activities. In a global analysis, they found that correlation for leisure activities was stronger when “walking” was excluded. Furthermore, correlation of previous-year occupational activities was ρ= 0.88. Correlation results for test-retest method in Pima Indians were similar to those found in our study; correlation was greater for previous-year activities than that reported during previous-week.
In the study by Nang et al. (2011), correlations were found for moderate activities of ρ= 0.58 and of ρ=0.38 for vigorous activities in IPAQ. Additionally, correlations of SP2PAQ were ρ= 0.55 and ρ= 0.75 for moderate and vigorous activities, respectively. Authors conclude that IPAQ had limited accuracy for distinguishing physical activity levels of individuals. Also they said that reproducibility of SP2PAQ is reasonably good and which can be used in large epidemiological studies.
In the reproducibility study of Korean Version of Neighborhood Physical Activity Questionnaire was assessed using Spearman correlation. Questionnaire was applied twice, one week apart. Considering total number of hours per week, values of ρ= 0.57 and ρ=0.66 for moderate and vigorous intensity activities were found. Degree of agreement was defined as slight agreement (ρ ≤ 0.2), fair (0.2 - 0.4), moderate (0.4 - 0.6), substantial (0.6 - 0.8), and almost perfect (ρ ≥0.8) (Bae et al., 2015).
Regarding studies in Mexican population, Medina et al. (2013) found a correlation of 0.55 and Vega- López, et al. (2014) found correlations of 0.61 and 0.65 for the SBAS and RAPA, respectively. Authors concluded that their results were modestly correlated. It is important to mention that questionnaires used by these studies are short versions compared to ours. High-quality result found in this study, is in part due to previous training of interviewer. This consisted of advice to understand questionnaire in detail, which was previously applied to four volunteers. This training process allowed interviewer to be familiarized with questionnaire format, answer doubts about used technique to complete interview and to understand handling and processing of numerical data.
CONCLUSION
This study achieved description of activities inhabitants of Seri community perform as part of their everyday life in their socio-cultural context. It was also possible to recognize reported activities frequency and duration and to categorize them as leisure and occupational, both for women and men.
Additionally, questionnaire reproducibility was evaluated through test-retest method, and according to Spearman correlation values and confirmed by those found with intraclass correlation coefficients analysis, it is possible to establish that physical activity questionnaire
is a useful tool to estimate physical activity patterns in Comcáac community of the State of Sonora, and it can be used to study association among physical activity, obesity and health.
