Introduction

The importance of physical activity (PA) on health is well established. PA can protect against health problems such as cardiovascular disease ¹, diabetes ², obesity ³, osteoarthritis ⁴, mortality ⁵, cancers ⁶, metabolic syndrome ⁷, and mental problems ⁸. Therefore promoting engagement in regular PA is considered an important step for public health, preventing a wide range of chronic diseases ^2,9.

Despite evidence of the health benefits of PA, many people do not exercise enough to benefit their health ^10,11. Furthermore, PA levels tend to decrease with age ¹⁰, which means that older ages are at risk. Knowing that the percentage of the Portuguese population engaged in regular PA is among the lowest in the European Union ^10,11,12, it is important to understand the related factors of PA. The study of correlates of PA in Portugal is particularly important because in the past decades Portugal has undergone a demographic transition, going from a predominately rural country to one where most of the population lives in urban areas. Therefore, it is important to understand the habits of the people that are now living in urban areas and the correlates of PA participation.

A better understanding of the contributing factors of PA is critical in designing effective intervention programs. Moreover, increasing the PA levels of Portuguese people is also important for the prevention of overweight and obesity, since the rate of overweight and obesity among the Portuguese population is amongst the highest in the European Union ^13,14.

In spite of a growing emphasis on the importance of PA in promoting health, factors related to PA among Portuguese adults remains unclear. PA is a multi-faceted behavior influenced by environmental, psychological, social and demographic variables. The study on non-modifiable socio-demographic variables, such as age and gender, suggest that some groups of adults are more likely to have lower PA levels and are in more need of interventions or that the interventions must be more tailored towards the specific characteristics of these groups ^15,16. Other socio-demographic variables, such as, socioeconomic status (SES) have shown mixed findings with regard to PA ^17,18,19,20. Since quantitative research that explores the correlates of PA is still rather limited among Portuguese adults, this study aimed to identify the socio-demographic correlates that explain the recommended PA level among Portuguese adults.

Methods

Results

The general sample characteristics are presented in Table 1. The proportion of the total subjects who achieved the recommended PA levels in their LT (MPA ≥ 5*30 min/wk and/or VPA ≥ 3*20 min/wk, or both, achieving at least 10 MET-minute.hr.wk^-1) was 31.8%. Most of the respondents were 41-50 years of age (52.5%), which is understandable, because many had children attending elementary and secondary schools; 32.2%, 36.8%, and 30.9% were classified as having a lower, middle and high SES, respectively; 55.4% reported lower education levels, 21.6% had graduated from high school, and 23% had a university degree.

Table 2 shows the comparison of socio-demographic characteristics and correlates of LTPA between the two groups with different levels of LTPA according to the PA recommendations. The subjects who met PA recommended guidelines in leisure time accounted for a large percentage in the highest SES for men (χ²(2) = 13.238, p = 0.001) and middle SES for women (χ²(2) = 8.350, p = 0.015). In addition, among women, those who met the PA recommendations were likely to have a lower education level (χ²(2) = 7.389, p = 0.025). There were no significant differences related to age for either genders.

Table 3 shows the results of the unadjusted logistic regression analysis. For men, those with high SES (p < 0.001) had a significantly higher tendency of meeting PA recommendations. For women, middle SES (p = 0.009), and middle level of education (p = 0.007) had a 45% and a 47% greater chance of meeting PA recommendations during leisure time, respectively.

Table 4 presents the results of the adjusted regression analysis. For men, having a high SES remained significant after adjusting variables, a 93% greater chance of being physically active at the recommended level than those with low SES (p = 0.001). For women, those with middle education levels had a 41% greater chance of being physically active at the recommended level than those with lower education levels (p = 0.032), and having a middle educational level increased the likelihood of meeting PA guidelines by 34% more than those with lower educational levels.

Discussion

The current study examined the link between socio-demographic factors and engagement in the recommended PA level for Portuguese adults. From the results, three main findings were identified. First, age was not significantly related with PA, which differed from what was found in previous studies ^10,16,17. High SES for men and middle SES for women were linked to meeting PA recommendations, confirming that SES is a significant correlate of PA ^17,19. Third, educational levels were only related with meeting the PA recommendation for women. Though not related to the main goal of the study, it was also found that 31.8% of the study participants were physically active and achieved the PA recommended levels during leisure time, similar to findings observed in developed countries ^16,28,29, but higher than in some developing countries ^23,24. This difference may reflect the availability of and accessibility to PA recreational activities in developed countries, since LTPA is not common in developing countries ³⁰.

From the preliminary analysis of data, men are not more active than women, unlike other studies which showed that women were less likely to engage in PA that meets the guidelines ^16,28,29. This result is interesting and suggests that the difference between men and women observed in other studies, may be related to PA in the workplace and while commuting.

In contrast to other studies, which found that participation in PA decreases with age ^16,17,31, a significant association was not found between age and recommended PA levels. Knowing that certain age groups could skew the results, those who were over 60 years of age were removed from the sample (since they were more likely to have already retired). This was necessary because there have been mixed reports about the relationship between age and LTPA, with some studies finding an increase of LTPA after the aforementioned age ^12,23, whereas others show a decrease ^32,33. The absence of the association between age and meeting recommended PA levels could reflect the types of employment the subjects had and the context in which they lived. It is known that people who live in urban areas face considerable economic pressure, and have time constraints, regardless of age. Although differences among age groups were not found, some differences could have been noticed due to the fact that age is related to different stages of life. Usually, young adults (31-40) have good biological function and their physical performance is not limited ³⁴. Therefore, it was expected they would also have higher levels of PA. Most middle age adults are experiencing the demands of a growing family as well as an increase in body fat, and are more focused on their jobs. These facts could lead to a decrease in the levels of PA. Some years later, during later middle-age, women reach the menopause, men reduce their output of sex hormones and, for men and women career opportunities have commonly peaked. This could also lead to a decline of PA.

Data from this study indicates that, among men, low SES respondents were less likely to meet the PA recommendations in leisure time. This finding is consistent with a number of studies that indicated that the highest prevalence of physical inactivity is among lower SES adults ^17,19.

However, this is contrary to findings among young Portuguese individuals, where the influence of SES on adolescents’ self-reported PA was not found ¹⁸. Our findings support the hypothesis that higher SES adults are more likely to follow preventive programs and health-promoting behavior, whether due to greater motivation or access to resources. On the other hand, women with middle SES were more likely to meet the recommendation for PA in leisure time than those from low and high SES. This confirms that people with low SES have a limited ability to control their PA in the face of inaccessible environments ^19,35. With regard to women with high SES, one can only speculate that they generally have high control, high daily demands and long work hours, which might therefore reduce their time available for LTPA.

Educational levels were not associated with meeting PA recommendations for men, but were for women. This result is not consistent with previous investigations which showed that participants who achieved a higher educational level showed lower prevalence of a sedentary lifestyle ^12,17. Nevertheless, results from Japan are similar to this Portuguese study for men and women ²⁹, despite the variance in cultural practices. Among women, the results showed that the most educated were more likely to attain the recommended PA levels. However, the same is not seen among men. Because this factor has a different effect on men and women, attention should be given to developing intervention programs which aim to increase PA participation. These factors confirm Varo et al. ¹², which stated that the correlation between PA and education level is notentirely understood.

Although environmental correlations were not explored in this study, the literature has shown that factors such as neighborhood crime and safety rates, aesthetics, and accessibility of recreational facilities are related to LTPA ^15,36. These findings highlight the importance of supportive environments for LTPA and emphasize the need to incorporate factors at multiple levels of influence in interventions to promote PA ³⁷. It is expected that the intervention would be different in rural and urban areas, however, although environments differ between rural and urban areas, the correlates of LTPA are similar among both urban and rural populations ³⁸. Furthermore, unfortunately the instrument used for data collection did not allow us to differentiate week day and weekend PA, though the potential for variability does exist. On weekends, men and women spend less time on low-intensity PA, and men spend more time on moderate PA than they do on weekdays ³⁹. PA patterns on weekends are different than on weekdays and might be related with walking to work, which is associated with overall higher levels of PA among adults ⁴⁰.

The current investigation had a number of limitations. The analysis was cross-sectional, thereby making it impossible to determine cause and effect. PA was self-reported rather than objectively measured, which means the results could be subject to bias. In addition, the study focused only on LTPA and did not include activities at work, in the home or while commuting. Individuals who are inactive during leisure time may be more active in other contexts. The use of only four socio-demographic variables is also a limitation, because there are other variables that could help to characterize the study population, such as, nationality, ethnicity, skin colour, marital status and neighborhood location. Nonetheless, the four variables used are the most important for the Portuguese context. Finally, this study was conducted in an urban area, and thus the results cannot be extrapolated to other populations, particularly adults from rural areas.

Conclusion

In summary, significant correlates of attaining the recommended levels of PA were observed. For men, having a high SES was associated with meeting the recommended level of PA in leisure time. For women, having a middle SES was related to achieving the recommended PA level during leisure time. Moreover, men and women had different patterns of SES by education level. The findings of this study contribute additional evidence to the literature on the multivariate factors associated with PA behavior. Future studies should require that more correlates be observed in the literature, the potential of PA, and, concomitantly, that specific domains of PA, such as commuting, activity at work and housework be analyzed.