Healthy lifestyles are an obstacle to diseases and include preventive health, good nutrition, weight control, leisure, regular physical activity, periods of rest and relaxation, the capacity to face adverse conditions or situations and establishment of affective relationships of solidarity and citizens, adopting a posture of being in the world with the purpose of living with quality [1
The study of healthy lifestyles can be marked up to the present time by three great periods: a first period beginning in the nineteenth century until the middle of the twentieth century; a second period, which ends the second half of the twentieth century and a third period that focuses on the present [2
]. The same authors state that during the first period, the studies emphasize healthy lifestyles, dependent on a sociological vision and individual factors, of the individuals who belonged to a certain social stratum and could keep that. In the second period, there were studies that reported an incorporation of healthy lifestyles in the field of health and studies with isolated conducts, apparently salutary or conducive to a healthy living style. In the third period, there was a development up to the present day, in which work reflects healthy behaviours integrated in different contexts, or even the relationship of different dimensions with healthy lifestyles, such as motivation or basic psychological needs, seeking to perceive self-determination for a particular practice and the adoption of healthy behaviours.
It should be noted that, only in the second and third chronological periods of the study of lifestyles, did this construct begin to be evaluated, according to valid and reliable instruments for this purpose. Examples of these instruments are those that were developed and applied in the works of several authors [3
The EVS instrument is derived from The Health Behavior in School Children (HBSC) [6
]. This instrument initially evaluated health behaviours in the school population, which determines sociodemographic variables, healthy lifestyle variables, daily life variables and psychosocial health variables. This instrument reflects an inventory of health behaviours in students [3
], with a factorial distribution of 29 items [7
This instrument was later adapted and used in Spain [7
]. The Healthy Lifestyle Questionnaire (EVS) proposes an instrument of 30 items, distributed by six factors that determine behaviours, more specifically, eating habits and the subfactors: balanced diet and respect for mealtime, resting habits, tobacco and alcohol consumption and other drugs consumption. In the same line of research [14
], the relationship between physical activity and sport on behaviours belonging to healthy lifestyles was analyzed, using a sample of 402 students of Physical Education, from 14 to 18 years of age. The results showed that this practice positively and significantly predicts the balanced diet and respect for mealtime and, without showing statistical significance, resting habits. Some authors [15
] demonstrated the relationship between the practice of physical-sports activities and the most self-determined motivation [17
], with the consequence of maintaining healthy lifestyles [18
]. It is also important to note that some authors [20
] indicated that tobacco consumption was a significantly predisposing factor for alcohol consumption.
The preliminary evaluation of EVS for the Portuguese language [21
] was developed with a sample of Portuguese secondary school students, where goodness indexes of the confirmatory factorial analysis were acceptable. The structure of the EVS was composed of 26 items distributed by five factors: tobacco and alcohol consumption, other drugs consumption, respect for mealtimes, and maintenance of a balanced diet. The resting habits factor was eliminated because its items presented a factorial weight of 40.
Later, other authors [22
] sought to adapt to the context of the Portuguese veteran sport of EVS. The overall results of the model indicated a good fit, as well as a good validity concurrent with the motivation evidenced for the practice. The structure of the EVS was composed of 20 items distributed by three factors: tobacco consumption, resting habits and eating habits. This version of EVS did not include all the original dimensions of the questionnaire, specifically alcohol consumption and consumption of other drugs, since they were not considered adequate for the evaluation of the collective veteran sports. There were also problems with the factorial load of some items related to the dimensions of balanced diet and respect for the mealtimes, and the grouping of the same ones was chosen, being denominated by the factor of eating habits.
The Spanish validation of EVS [2
] was developed with a sample of 812 participants between the ages of 14 and 88 years. The overall results of the model indicated an optimal fit, as well as a good concurrency against the motivational orientation based on the theory of self-determination. The structure of the EVS was composed of 12 items divided by four factors: tobacco consumption, resting habits, respect for mealtime, and maintenance of a balanced diet. Currently, the healthy lifestyle construct has a close relationship with the modern food pyramid and is investigated according to multivariate models, which involve quality of life or absence of disease.
The present study aims to validate the Healthy Lifestyle Questionnaire (EVS II) for the Portuguese language.
The use of this questionnaire will allow a more comprehensive characterization of the healthy habits and lifestyles of athletes, regarding the validations of the EVS [21
], supporting translation of the Spanish version [2
] to Portuguese. The Healthy Lifestyles Questionnaire (EVS II) itself presents as a new and a more appropriate instrument, which has an adequate number of items per factor.
3.1. Confirmatory Factor Analysis
In the present work, we have chosen to eliminate six items, since they did not meet the factorial load equal to or greater than 0.40, as proposed by the author [31
]. In the balanced diet factor, item 12 (Like sweets, cakes, ... at most once or twice a week) was eliminated, in the factor of tobacco consumption, item 5 [I consider that tobacco helps to relate to me, (alcohol makes it better)] and item 16 (I have the feeling that I’m always drinking more alcohol) were eliminated, in the dimension of other drugs consumption, item 3 (I usually take some drug) was eliminated, and in the resting habits dimension, item 22 (I usually take a nap for approximately 30 min) was eliminated.
Confirmatory factor analysis to evaluate the six-factor model of the Healthy Lifestyle Questionnaire-EVS II, showed that the 24 items were grouped into six factors, respectively: balanced diet (5 items), respect for mealtime (5 items), tobacco consumption (4 items), alcohol consumption (3 items), other drugs consumption (4 items) and resting habits (3 items).
Likewise, the standardized factor loads were all statistically significant (p
< 0.01), so it can be concluded that the model presented in Figure 1
, at the analytical level, presents satisfactory results.
After a first analysis, the general results of the model indicated a reasonable fit of the adapted version of the Healthy Lifestyle Questionnaire (EVS II), which was composed of 24 items: χ2 = 305.925, p = 0.000, df = 120.017, χ2/df = 2.549, NFI = 0.909, TLI = 0.918, CFI = 0.944, GFI = 0.944, AGFI = 0.909, SRMR = 0.048, and RMSEA = 0.060. With these results, the structural model reveals a satisfactory overall fit, having models with satisfactory adjustment in previous versions, although with a smaller number of analysis dimensions than the EVS II.
3.2. Analysis of Internal Consistency and Convergent Validity
shows the internal consistency values of EVS II. The internal consistency of each of the factors resulting from the factorial analysis (McDonald’s Omega ω), presented the following results: 0.81 balanced diet, 0.90 respect for mealtime, 0.96 tobacco consumption, 0.76, alcohol consumption, 0.78 consumption of other drugs and 0.74 resting habits. The mean extracted variance and the composite reliability for each factor were 0.82 and 0.53 in the balanced diet, 0.91 and 0.71 for respect for mealtime, 0.96 and 0.85 for tobacco consumption, 0.79 and 0.51 for alcohol consumption, 0.78 and 0.51 in the consumption of other drugs, and 0.77 and 0.54 in resting habits, fulfilling all the factors evaluated in the assumptions [30
To verify whether the number of factors is reasonable based on the specific measurement model, according to the equations proposed by the author [30
], OmegaHS values were obtained to balance diet by 0.27, respect for mealtime 0.36, tobacco consumption 0.49, alcohol consumption 0.35, other drugs consumption 0.27 and resting habits 0.34, these values being typical or relatively large.
From a descriptive point of view (Table 2
), the values of healthy lifestyles were obtained in this group of veteran athletes, with higher means in balanced eating behaviours (
= 3.63 ± 0.80), respect for mealtime (
= 3.64 ± 0.89), and lower means in smoking average (
= 1.57 ± 1.02), alcohol consumption (
= 1.52 ± 0.68) and consumption of other drugs (
= 1.55 ± 0.74). In the motivation variables, the veteran athletes demonstrated a high autonomic motivation (
= 5.43 ± 0.89), and reduced values of controlled motivation (
= 1.99 ± 1.07) and amotivation (
= 2.02 ± 1.20).
The evaluation of concurrent validity through a bivariate correlation analysis, most of the correlations between the variables of the EVS II and the BRSQ were significant and in the expected direction. Autonomic motivation is positively correlated with a balanced diet, respect for mealtime and resting habits. It assumes a negative correlation like tobacco consumption, alcohol consumption and consumption of other drugs. Controlled motivation and amotivation assumed correlations in the opposite direction to autonomous motivation, given the different variables of healthy lifestyles.
The main objective of the present study was to broaden previous research on healthy lifestyles, particularly associated with the practice of veteran athletes, through the validation of the Healthy Lifestyle Questionnaire—EVS II, for the Portuguese competitive sport context.
According to that, each new application of a measuring instrument represents a contribution to improving the theoretical value of the research domain [27
]. This study extends this core of knowledge, confirming the validity of the EVS II instrument in a research context, as well as through improved knowledge of how to help sports and exercise psychologists, to understand healthy practices and health indicators in veteran athletes.
Measuring the internal consistency of each of the calculated factors, using McDonald’s omega, we obtained values greater than or equal to 0.70 in both measures, as proposed by the authors [30
We estimated the composite reliability, the mean variance extracted and OmegaHS for each factor, and we observed that the values obtained correspond to the indicators proposed [30
], to conclude that a substantial amount of the variance is captured by the construct, where the composite reliability must present a minimum value of 0.70, the mean variance extracted is greater than 0.50, and OmegaHS sets to typical centered values 0.20 to 0.30 and relatively large values > 0.30.
Confirmatory factor analysis showed that the 24 items were grouped into six factors, respectively: balanced diet (5 items), respect for mealtimes (5 items), tobacco consumption (4 items), alcohol consumption (3 items), consumption of other drugs (4 items) and resting habits (3 items). With these results, the structural model reveals a satisfactory overall fit, having models with satisfactory adjustment in previous versions, although with smaller number of analysis dimensions than the EVS II.
The results obtained through the psychometric quality indexes revealed [28
] an acceptable fit in χ2
, and in the value of χ2/df
, NFI, GFI and RMSEA. They showed good fit in the AGFI and SRMR indexes. Despite the values of TLI and CFI being very close to those indicated [30
], they did not fail to comply with the values proposed by some authors [32
These results are consistent with previous research using EVS [2
] and confirm the importance of each of the six dimensions in understanding the healthy lifestyles of athletes. If we observe the results of the preliminary validation of EVS [21
], and the work of other authors [2
], with those obtained in our study, both presented good psychometric properties, based on what the literature advises [28
], with EVS appearing in these four studies as a reliable tool for assessing healthy lifestyles. However, we emphasize that the validation of the EVS II is the adapted version that best respects the initial model of the questionnaire presented by another author [6
] with six factors extracted.
The previously validated versions found problems with some items with a factorial load lower than 0.40 [31
], leading to the elimination of some factors extracted, such as the resting habits, in the preliminary research [21
], or agglutination of items of the dimensions of balanced diet and respect for mealtime, which in the Portuguese research [22
] gave rise to the size of eating habits. In the present research, we have also opted to eliminate some items, since they did not meet the factorial load equal to or greater than 0.40, as proposed by the author [31
], and six items were eliminated.
This instrument has the potential to become more refined as new contributions on healthy lifestyles arise, where in later studies it will be interesting and very convenient for these items to be remeasured and tested in order to obtain other models of valid equations, such as that proposed by the authors [28
In the descriptive analysis, the results showed that the participants in the study tend to value the items of the questionnaire, which in fact seems to be demonstrated by the moderate and high averages in the dimensions of balanced feeding, respect for mealtime and resting habits, as well as reduced means in the dimensions of tobacco consumption, alcohol consumption and consumption of other drugs. This thus shows the theoretical importance underlying the construction of healthy lifestyles. The same descriptive trend was obtained in some previous studies [2
]. Furthermore, the results of these studies supported concurrent validity through the analysis of bivariate correlations. In the present study, most of the correlations between the EVS II and BRSQ variables showed significant associations, emphasizing the validity in this work, particularly with the motivation continuum [17
]. The authors [34
], in a meta-analysis with most of the non-experimental studies, determined that there is a strong relationship between the self-determination theory [17
] and positive health behaviours. Analogous observations were obtained in several studies [14
In future applications, it would be interesting to observe additional samples with athletes from other countries, and data could also be collected to compare the validity of the scale in different cultural contexts. We also point out the interest of the possible inclusion of the practical factor of physical exercise in the questionnaire, which in future studies can be measured in the same way as other research has already done, with other specific instruments. It may be interesting to apply studies that are based on the trans-theoretical model of motivation, or on the theory of planned behaviour or even others that are based equally on the theory of self-determination [17
] and to evaluate the adoption of healthy lifestyles in different strata of the population that show the practice of physical or sports activity.