Physical activity among Portuguese university students and its relation to knowledge and perceived barriers

Para citar este artículo utilice la siguiente referencia: Alves, R.; Precioso, J.; Becoña, E. (2021) Physical activity among Portuguese university students and its relation to knowledge and perceived barriers. Sportis Sci J, 7 (1), 25-42 https://doi.org/10.17979/sportis.2021.7.1.6924 http://revistas.udc.es/ 25 Physical activity among Portuguese university students and its relation to knowledge and perceived barriers La actividad física entre los universitarios portugueses y su relación con el conocimiento y las barreras percibidas


Introduction
The World Health Organization identified physical inactivity as the fourth major risk factor for global mortality (WHO, 2010) and other authors as the greatest public health problem of the 21 st Century (Trost, Blair, & Khan, 2014). In contrast, the practice of physical activity (PA) is considered a positive health measure.
Scientific research has shown that the regular practice of PA contributes to physical health (Füzéki, Engeroff, & Banzer, 2017;Lear et al., 2017;Pandey et al., 2017), as well as to the reduction of mortality and cardiovascular and chronic diseases (Ekelund et al., 2016;Lear et al., 2017). In the same way, it contributes to psychological and emotional well-being (Álvarez Rivera, Cuevas Ferrera, Lara Pot, & González Hernández, 2015;Joseph, Royse, Benitez, & Pekmezi, 2014;Wu, Tao, Zhang, Zhang, & Tao, 2015). For example, Wu et al. (2015) found that the daily practice of PA contributed to the reduction of depression, anxiety and sleep problems.
In recent years, several public health guidelines for the PA of the adult population were published, differing in the amount, intensity and frequency of PA (Haskell et al., 2007).
The most recent recommends at least 150 minutes of moderate PA per week or 75 minutes of vigorous PA per week, allowing the combination of these two intensities and in episodes of at least 10 minutes (WHO, 2010).
Despite the widespread recommendations and health benefits, in most developed countries there is a high proportion of adults who do not meet these recommendations for PA and show a low level of PA (Guthold, Stevens, Riley, & Bull, 2018), particularly in Portugal (Batista et al., 2012). The level of PA decreases from childhood to adolescence, as well as from adolescence to adulthood. University students, regardless of the country, have a very high prevalence of physical inactivity that is bigger (Arias-Palencia et al., 2015;Haase, Steptoe, Sallis, & Wardle, 2004;Keating, Guan, Piñero, & Bridges, 2005 (Guthold et al., 2018). This evidence becomes relevant when stating that the transition to adulthood, matching it with starting higher education, is a critical and the most vulnerable life period. Health habits acquired during this period tend to remain throughout life, including those related to physical exercise (Curry, Jenkins, & Weatherford, 2015;Filla, Hays, Gonzales, & Hakkak, 2013;Rodriguez Rodriguez, Santibañez Miranda, Montupin Rozas, Chávez Ramírez, & Solis Urra, 2016).
Several factors contribute to the decrease or absence of PA, which is why it is interesting to assess how predisposing factors include knowledge and attitudes towards PA (Ennis, 2010;Keating et al., 2009;WHO, 2012).
Given the above, this article aims to analyse the relationship between knowledge on the recommendations for the practice of PA, the perceived barriers towards PA and the practice of PA by students in higher education.

Population and sample
For the 2018/2019 academic year, 5447 students were registered in the 1 st and 3 rd year of integrated bachelors and masters degrees. Courses related to health sciences, undergraduate or postgraduate masters and those that did not have classes in the 1 st or 3 rd years were excluded from the sample. The minimum sample size needed for this study was 592 students (margin of error = 5%, confidence level = 99%, and response distribution = 50%). For this purpose, stratified probabilistic sampling of university students was performed according to the year of study and the scientific area. The different undergraduate and master's degrees were divided into scientific areas (as defined by the Foundation for Science and Technology): Human and Social Sciences, Law and Economic Sciences, Exact and Natural Sciences and Engineering Sciences.
In this cross-sectional study with a representative sample of university students (n = 840) in one university in Portugal, data were collected using a validated self-reported

Instruments
Currently, there are several scientific instruments to monitor the prevalence of PA among young adults. Therefore, the development of the instruments present in this investigation was carried out in three stages: scale construction (1 st stage); content validity (2 nd stage); psychometric validity (3 rd stage), according to the procedures defined by Bowling (1998).
For the scale construction (1 st stage), a systematic review of the literature was made to identify the questions and items commonly used to assess knowledge, attitudes and practices of PA in higher education (Alves, 2019). Based on this review, an analytical matrix was created for each of the dimensions to be analysed, and those with the same semantic similarities were eliminated.
For the content validity (2 nd stage) 10 PhD researchers from several Portuguese universities with recognised work in the area of Health Education in Higher Education were invited, and the feedback received from 5 of them was considered as well as all the proposed semantic changes. Similarly, the instrument was applied to 12 university students, using the method "thinking aloud" (Bowling, 1998;Keszei, Novak, & Streiner, 2010) to identify items that might be confusing, to exclude less relevant or redundant items, and to verify that precoded response options were sufficient. After the suggested redrafting, the preliminary version of the questionnaire survey was presented to a sample of 32 students, not included in the final sample.
The questionnaire included sociodemographic variables (sex, age, scientific area of study, academic year, weight and height (to calculate BMI), being in a love relationship, Para  professional situation e current residence) and specific questions to measure the following variables: -Prevalence of PA: the Godin Leisure-Time Exercise Questionnaire (Godin & Shephard, 1997) was used and one question about sedentary lifestyle ("On a normal day, how many hours do you spend seated?"); -PA knowledge: a 6-item scale with options for the answer -True, False, Don't Know (e.g., "The practice of an intense physical activity is necessary to achieve health benefits.") -Perceived barriers towards PA: a 5-item scale on a 5-point Likert scale (1 -I strongly disagree, 5 -I totally agree) concerning lack of energy, lack of motivation, lack of resources, lack of support and lack of time.

Procedure and Statistical analysis
The application of the instrument was carried out in classroom context and paper-andpencil format for all students in the sample, after giving informed consent. Activity counts were converted to metabolic equivalents (METS) by multiplying episodes of vigorous activity by 9, moderate activity by 6, and mild activity by 3. Total activity scores were calculated by adding the MET calculations for each PA intensity level and were classified in three types: Insufficiently Active/Sedentary; Moderately Active and Active.
For the knowledge scale, the number of correct responses was added to give an overall knowledge score, and the mean was calculated. This means that the higher the scale value, the higher the level of knowledge. The perceived barriers scale was subject to Cronbach's alpha analysis to analyse its reliability (Bowling, 1998), and a good reliability index was obtained (α = .742). In addition, inter-item correlations ranging from .644 to .269. To read the results of this scale, the higher the average of the scale, the more will the barriers be towards PA.

Results
The results show that 35.7% (n = 298) of the students surveyed are sedentary, 17.5% (n = 146) moderately active and 46.8% were classified as active. It was also found that, on average, on a typical day, university students spend 7.44 ± 2,656 hours seated. There were no differences between the practice of PA and the number of hours that the students declared to spend seated (F (2, 798) = 1.617, p = .199). Table 1 shows that among all the sociodemographic variables analysed the practice of physical exercise was only associated with the sex of the respondents (χ 2 (2) = 22.554, p = .000). This indicates that female students are more sedentary and less active compared to male students. Regarding the sedentary behaviour declared by the students, it was found that, on average, university students spend 7.44 ± 2.656 hours seated, according to their professional situation (t (795) = 3.693, p = .000) and scientific area (F (3, 799) = 12.380, p = .000). Thus, full-time and engineering students were the ones who declare to spend daily the most hours seated, compared with working students and students of other scientific areas, respectively.
The level of knowledge about PA was low, presenting, on average, 2.79 ± 0.090 correct answers, in a total of 6 possible correct answers. Table 2 shows statistically significant differences between knowledge about PA and the scientific area (F (3,827) = 3.060, p = .028), indicating that students in the Exact and Natural Sciences area have more knowledge compared to students of the Human and Social Sciences area. The scale of perceived barriers to the practice of PA obtained an average of 2.26 ± 0.037, in a range between 1 and 5, with the highest value corresponding to more barriers. The inactive students were those who revealed more barriers compared to students who practice PA (moderately and actively), with moderately active students showing more barriers compared to active students; female students and full-time students presented a higher score in the perceived barriers scale compared to male students and working students, respectively.
Regarding BMI, Hochberg's GT2 was not significant. There were no associations between the level of knowledge about the practice of PA and the perceived barriers (ρ = .023, p = .516). Table 4 presents the results of the ordinal logistic regression model for PA. The model is statistically significant (G 2 (2) = 107.057, p = .000). According to the model, as the perceived barriers increase, the probability of practising physical exercise decreases (βPerceived barriers = -.750, p = .000). Regarding sex, there is a higher probability of practising PA in male students, compared to female students (βMale = .578, p = .000).

Discussions
The practice of PA by university students was low, considering that 35.7% of respondents were sedentary or insufficiently active. In comparison with a study carried out worldwide, the results of this study show that there is a higher prevalence of inactivity than that seen in European countries in the general population (23.5%) (Guthold et al., 2018).
However, when considering the university population, the percentage of inactive students corroborates with other international studies in which physical inactivity varied between 23% and 39% in western countries (Haase et al., 2004) and between 21.9% and 80.6% in non-European countries (Pengpid et al., 2015). Despite the variations recorded, it should be noted that, on average, the prevalence of physical inactivity among university students is lower in European countries compared to other non-European countries.
Scientific research in this area has identified some variations regarding the relationship between PA and the sex of university students. In this study, there was a lower prevalence of PA in female students compared to male students (Rodriguez Rodriguez et al., 2016;Romaguera et al., 2011;Wu et al., 2015). However, other studies have reported that women tend to exercise more than men (Gnanendran, Pyne, Fallon, & Fricker, 2011;Morgan & Elizondo, 2016) or that there were no differences considering the sex variable (Arias-Palencia et al., 2015).
Social representations tend to relate BMI to the practice of PA (Vadeboncoeur, Townsend, & Foster, 2015); however, there is no evidence to prove it, as verified in the present study as well as in other studies (Steptoe et al., 1997).
A possible explanation for the low prevalence of PA could refer to the number of hours that students spend seated (Arias-Palencia et al., 2015); however, in this study, there was no relation between the number of hours that students spend seated daily and the practice of PA. The number of hours that students spend seated per day (about 7 hours), coincides with that shown in other studies (Farinola & Bazán, 2011).
Regarding the knowledge variable, students revealed a low level of knowledge about PA (Abula, Gröpel, Chen, & Beckmann, 2018;Keating et al., 2009;Martins et al., 2019), although the theory of planned behaviour (Ajzen, 1991) shows that knowledge has an impact on attitudes and that, in turn, it influences intentions in relation to behaviour. And, specifically, knowledge about PA and its health benefits direct people to the practice of PA (Haslem, 2014), as has been empirically verified in international research (Abula et al., 2018).
In the present study, the level of knowledge of the students is not correlated with the barriers to the practice of PA; neither was it included in the prediction model of the practice of PA. As in the studies carried out by collaborators (2009) andHaslem (2014), no significant correlation was found between knowledge about PA and the practice of PA.
However, improving and developing PA knowledge could be the first step towards establishing healthy behaviours in PA (Keating et al., 2009).
The perceived barriers were related to the practice of PA. Thus, the students who had a higher score of perceived barriers were those who presented themselves as being less active.  Samara, Nistrup, Al-Rammah, & Aro, 2015). In this sense, the investment in time management training seems to make sense. Moreover, it is considered that betting on intrinsic motivation can be a good way to improve adherence to PA (Chacón Cuberos et al., 2017). This is because the lack of motivation was also a prominent barrier (King, Vidourek, English, & Merianos, 2014).
Lack of energy is also often mentioned as a barrier to the practice of PA (Ramírez-Vélez et al., 2016;Samara et al., 2015), and it was also considered by the students surveyed.
Given this item, the students found that lack of energy could constitute a barrier to the practice of PA, preceded as already mentioned by lack of time and lack of motivation.
The support of parents and friends is considered as essential to encourage students to practice PA (King et al., 2014). The present study showed that students perceive that they have the support of family and friends to practice PA. Therefore, this was not considered a  Staten, 2008). However, they do not take advantage of these resources to increase the practice of PA (Keating et al., 2005). For this reason, it makes perfect sense that the item 'lack of resources' has not been evidenced by the respondents.

Conclusions
The increase in PA levels of university students is a social necessity, and interventions in this area should focus on female students and develop strategies to break the most frequent barriers.
Time management training, intracurricular PA integration or the creation of accredited programs can be analysed in future interventions. In universities in which PA has been incorporated into curricular study plans, statistically significant improvements have been shown in the level of PA practice and, consequently, an impact on students' health (Warren & Odenheimer Brin, 2017).
Higher education institutions should undertake a key role in promoting student health, by implementing actions to reduce their students' inactivity, developing the transmission of information about the health benefits related to the practice of PA, so that university students have the opportunity to make informed decisions about PA, develop more positive attitudes towards PA and choose a more active lifestyle.
Some limitations should be considered. Firstly, this study included only one university, so constraints should be considered in generalising the results to the entire university population in Portugal. Secondly, it should be noted that the results presented are based on data collected by self-report, so it may be distorted regarding the difficulty in distinguishing the different levels of activity or underestimating or overestimating the PA done. Thirdly, it should not be forgotten that the barriers to the practice of PA may not correspond to all the barriers considered by students and that they can be excuses and not exactly barriers. Finally, considering that social desirability is always present in this type of research methodology.
However, the application of a questionnaire in paper-pencil format and in the classroom can