Validation of the Questionnaire “Stage of Change in Behavior of Scholars in Eating Habits and Physical Activity” (CEHAF)

To validate the questionnaire “Stage of Change in Behavior of Scholars in Eating Habits and Physical Activity" (CEHAF). It is based on the Transtheoretical Model, and it intends to be a useful tool for developing strategies against childhood obesity. It measures knowledge, attitudes, and habits towards nutrition and physical activity. Observational, crosssectional, randomized, and multi-center study in a sample of 1671 Mexican school-age children, from 8 to 12 years old, both genders. The CEHAF questionnaire is self-administered and consists of 35 questions, including 18 multiple choices, on habits and attitudes towards alimentation and physical activity. Reliability analysis through Cronbach’s alpha resulted in an alpha coefficient of 0.845. The exploratory factor analysis revealed three theoretically-congruent factors. The final version of the CEHAF questionnaire was conformed of 14 items. The psychometric characteristics of the CEHAF questionnaire indicate that it is a valid and reliable tool for assessing the stage of change in schoolage population.

About 70-80% of children and adolescents diagnosed with obesity will become obese adults (Binkiewicz-Glińska et al., 2012;Cowie, 2014;Wong & Leatherdale, 2009).They have an increased risk of presenting cardiovascular and endocrine disorders such as metabolic syndrome, fasting insulin resistance, myocardial infarction, and stroke.Other conditions may involve nephropathy, fatty liver, and postural deformities.Obese girls often present polycystic ovarian syndrome and precocious puberty (Binkiewicz-Glińska et al., 2012).Consequences also imply psychological problems such as eating disorders, depression, stress, anxiety about self-appearance, and low self-esteem (Binkiewicz-Glińska et al., 2012;Wang et al., 2013).Obese children are at more risk of being isolated, intimidated, insulted, or aggressed by their classmates with normal physical complexion, which can lead to low academic performance, absentism, and a general poor adaptation capacity at school (García, 2010).
Nowadays, children have become more sedentary by spending more hours watching television, playing video games, or surfing on the internet, and less hours participating in physical activities such as walking or sports (Cowie, 2014;Meléndez, Cañez, & Frías, 2010).Television remains as the main source of sedentary lifestyle, therefore, the American Academy of Pediatrics has recommended a maximum of 2 hours of television per day for children and adolescents (Pérez Rodríguez-Martín, Ruiz, Nieto, & Campoy, 2010).Furthermore, the World Health Organization (OMS) and the National Commission of Physical Culture and Sport (CONADE) in Mexico, both recommend at least 60 minutes of moderate to vigorous physical activity (PA) for this specific population (World Health Organization, 2010; Comisión Nacional de Cultura Física y Deporte, 2010).
Schools are the key places for receiving education about healthy lifestyles, this due the presence of qualified personnel and because of the long time that children spend at school (Pérez-Morales et al., 2009).Interventions at school-age are crucial, since habits and attitudes acquired at this age can prevail throughout life (Escalante-Guerrero, de la Roca-Chiapas, & Macías-Cervantes, 2012).In Mexico, the school interventions focus at increasing PA, water, fruit, vegetables, legumes, and whole grains intake, substituting whole milk for low fat milk, and decreasing the sugar and greasy foods intake (Villalobos, Campos Nonato, Camarillo, & Enríquez Santamaría, 2012).Nevertheless, school interventions have not yielded the expected results.A meta-analysis carried out by Pérez-Morales et al.., where 10 randomized controlled trials from 2006 to 2009 were analyzed, reflected a lack of effect of these interventions in the reduction of BMI (Pérez-Morales et al., 2009).
This could be translated as a poor adherence to nutrition and PA recommendations, which implies changing cognitive, emotional, and behavioral habits (Villalobos et al., 2012).
It is necessary to develop comprehensive strategies with measurable impact that address knowledge, attitudes, and habits towards nutrition and PA (Escalante-Guerrero et al., 2012).For these purposes, the trans-theoretical model of health behavior change (TTM) can result helpful, since it evaluates the readiness to change a behavior (Greene et al., 1999).The TTM involves a progression process through a series of five temporal stages that compose the behavior change (Di Noia & Prochaska, 2010): precontemplation (no intention of changing in the next 6 months), contemplation (intention to change within the next 6 months), preparation (intending to change in the next month), action (recent change in the past 6 months), and maintenance (maintaining change from 6 months to about 5 years) (Greene et al., 1999;Prochaska & Velicer, 1997).
Questionnaires and scales about the stage of change of an individual can be useful to assess the attitudinal factors and the willingness to change the lifestyle (Escalante-Guerrero et al., 2012).Several instruments for health-related issues have been developed worldwide based on the TTM, showing adequate psychometric properties (Andrés, Saldaña, & Gómez-Benito, 2011;Blaney et al., 2012;Henry, Smith, & Ahmad, 2013;Kushida & Murayama, 2012;Sarbandi, Niknami, Hidarnia, Hajizadeh, & Montazeri, 2013;Vereecken et al., 2009).In Mexico, literature suggests that just a few surveys and questionnaires based on the TTM have been created or adapted, from which most involve just PA and are directed to an adult population (Esparza del Villara et al., 2013;Villalobos et al., 2012).Ochoa-Meza, Sierra, Pérez-Rodrigo and Aranceta-Bartrina (2014)

Ethic clearance
The study was approved by the University of Guanajuato Psychology Department ethics committee.

Instrument
The "Stage of Change in Behavior of Scholars in Eating habits and Physical Activity" (CEHAF) is a self-administered questionnaire.It consists of 35 questions, including 18 multiple choices, on habits and attitudes towards alimentation and PA.For the purposes of this project, breakfast contents and food eaten as snacks were considered the main habits to change, since breakfast is the meal skipped more often (Prochaska, 2008).PA was defined as any movement of the musculoskeletal system to increase energy expenditure by individual activities, recreational or sports, considering as an indicator the number of minutes per week allocated to any of these practices (Johnson et al., 2008).Individuals were classified as sedentary if they do not perform any type of recreational PA, exercise or sport, during their free time and/ or assign more than 2 hours a day for watching television, using the computer, or playing video games.

Sample and procedure
This was an observational, cross sectional study.The participants were 1671 school-age children from several primary schools across Mexico (49.2% male, 50.8% female).They were studying from 4th through 6th grade.The ages of the participants were between 8 and 12 years old, since evidence shows that children at these ages have acquired greater autonomy to internalize knowledge and change attitudes towards food and PA (Daniels & Greer, 2008).The CEHAF questionnaire was applied collectively during school hours in the presence of the class teacher and the researchers, who clarified emerging doubts during the process.

Analyses
For the scale validation, the Reyes Lagunes and García y Barragán methodological and statistical proposal (Reyes Lagunes & García y Barragán, 2008) was implemented, which involves frequency analyses, item discrimination analyses through t-tests, reliability analyses through Cronbach's alpha, and an exploratory factor analysis to obtain evidence of construct validity.SPSS v.21 statistical software was used to perform these analyses.

Results
For the items associated with eating habits, a frequency analysis was implemented with the purpose of verifying the correct data collection and also to corroborate that all response options resulted appealing for the participants.A new variable was created with the sums of the scores of every item, and the values for percentiles 25 and 75 were obtained.With these values, a new dichotomous variable was created and utilized as grouping variable for executing t-tests for independent samples in order to verify that the items discriminated between high and low scores.An internal consistency analysis using Cronbach's alpha was implemented to determine the scale's reliability.Three items were deleted for showing low item-total correlations.This changed the alpha coefficient from 0.671 to 0.845, with 15 items remaining (table 1).Source: own source.

TABLE 1 Items remaining after Cronbach's alpha analysis
Following this, an exploratory factor analysis with oblique rotation was executed to test construct validity.Three factors were obtained, which together explained 52.498% of the variance.Since items 9 and 10 had very low factor loadings, these were omitted from the final structure (table 2).A bivariate Pearson's correlation analysis was implemented between the three factors.A moderate association was found, highlighting, with a correlation value of 0.549, the relation between the factors 1 and 3 (table 3).

TABLE 3 Correlations between factors
Note:Pearson's correlation analysis.** Correlation is significant at a significance level of 0.01.Source: own source.
Table 4 shows that means for factors 1 and 2 were below the theoretical mean (which corresponds to the value 3).In contrast, factor 3 had a mean higher than the theoretical mean.Cronbach's alpha analyses indicated that only factors one and two presented an adequate reliability.

Discussion
In the present study, the psychometric properties of questionnaire "Stage of Change in Behavior of Scholars in Eating Habits and Physical Activity" were investigated and validated.An alpha coefficient of 0.845 was obtained for the whole scale, which indicates a high internal consistency.Concerning the analysis of the resulting factor structure, theoretical congruence was found between the factors and the items that conformed them.The first factor relates to the attempts at eliminating or reducing fat and sugars in the daily diet.The second factor consisted of three items associated to physical activity, opposed to the sedentary lifestyle caused by spending time in front of electronic media.Finally, the third factor gathers items which relate to healthy habits (both physical activation and alimentation).
Since the mean scores of factors 1 and 2 were below the theoretical mean, this could indicate that there is only an intention to improve on these health habits, but no actual actions to change them yet.Factor 3 had a mean higher than the theoretical, which could mean that the participants have less than six months performing these behaviors associated with healthy habits.The alpha coefficient of factor two points out that it does not have a sufficient reliability to use it as an individual scale, however, it can still be used as part of the whole scale.
In conclusion, the psychometric properties of the CEHAF questionnaire indicate that it is a valid and reliable instrument for assessing the stage of change in school-age children.It can be utilized in school interventions to dissuade sedentary lifestyle and inadequate nutrition habits, or maintaining the current conditions if these are found appropriate.This instrument proved to comprehensively cover the fundamental aspects that involve the issue of childhood obesity: Diet, PA, and habits regarding the willingness to change (see the full questionnaire in Appendix).

TABLE 2
Resulting factor structureNote:F = Factor.Extraction Method: Principal Component Analysis.Rotation Method: Oblimin with Kaiser Normalization.The rotation converged in 6 iterations.
Source: own source.