Prevalence of leisure-time sedentary behaviour and sociodemographic correlates: a cross-sectional study in Spanish adults

Lifestyle has been identified as a determining factor for mortality and morbidity [1, 2] and several studies have shown that physical activity (PA) is protective against both physical and mental illness, and perhaps even against death [3–6]. Conversely, physical inactivity has been linked to higher mortality [6–9] and a poor quality of life [4, 10, 11].

In an effort to stress the health benefits of PA, several institutions have issued guidelines indicating minimum recommended PA levels in terms of its frequency, duration and intensity. The most accepted guidelines have been those jointly drawn up by the Centers for Disease Control and Prevention and the American College of Sports Medicine (ACSM) in 1995 [7]. These recommendations were later updated in 2007 by the ACSM and American Heart Association (AHA) [12]. In general terms, these recommendations suggest the need for at least 30 minutes of moderate-intensity PA five days per week or a minimum of three days of intense PA performed at least 20 minutes each day, or a combination of both. Despite the evident benefits of regular PA, a large proportion of the population undertakes no PA whatsoever or does not meet the minimum recommendations described above [4, 13, 14].

Physical inactivity is a major source of concern in western societies. Population studies conducted in different countries [15] have reported a mean physical inactivity prevalence of 17.7% if considering total PA, that is, if including all domains of PA (work or daily occupation, leisure-time, and transportation, i.e., walking or cycling to get from one place to another). However, if only the leisure-time PA (LTPA) domain is considered, the prevalence of physical inactivity is higher. Thus, last Eurobarometer survey “Sport and Physical Activity” in 2013 [16], found that 59% of Europeans never exercise or practice any type of sport, or only do so seldom. This percentage was still higher in the southern countries of the European Union.

These data point to a clear need to define intervention strategies designed to encourage PA and, especially LTPA. Furthermore, among the different domains of total PA, LTPA (understood as PA or sport unrelated to work or transport [4]), has been positively linked to different health indicators [17–20]. Accordingly, efforts targeted at promoting PA should be in large measure directed at this component of total PA [21, 22].

For designing and implementing interventions to promote LTPA, a first step is to identify which population groups are more sedentary during their leisure-time. Therefore, it is necessary to understand the influence of sociodemographic correlates on physical inactivity during leisure-time. This information would help define targeted strategies for LTPA promotion in risk groups.

The aims of the present study were to determine the prevalence of leisure-time sedentarism in a representative sample of adults living in the Madrid region (Spain) and to examine its possible associations with several sociodemographic correlates and other variables: place of birth, sex, marital status, education level, employment and economic status and parents’ model of PA.

The findings of this study indicate that the majority (76.3%) of adults living in Madrid are physically inactive or sedentary during their leisure-time. This proportion is slightly higher than that reported by Meseguer et al. [14]. According to the latter authors, 71.2% of adults from Madrid do not fulfil the minimum LTPA recommendations for health. A slight reduction in LTPA (especially low to moderate intensity) was observed among Madrid’s inhabitants from 1995 to 2008. During this period, the percentage of adults not meeting PA recommendations rose from 71.3% to 72.9% [30].

Anyway, these comparisons should be made with caution and viewed more as a way of putting results into context rather than a precise comparison between them. These studies have used different methods to assess PA and different definitions of sedentarism, limiting the direct comparison of the results. According to European Union data collected in 1997 [31], percentages for leisure-time sedentary behaviour ranged from 43.3% (Sweden) to 87.8% (Portugal). For Spain, this figure was 71.0%, which is close to the prevalence rates described in previous studies for Madrid [14, 30].

If we consider the total PA (i.e., sum of PA undertaken at work or daily occupation, during leisure-time and when travelling to and from places), the prevalence of sedentarism is considerably lower. Several studies have shown that PA performed in domains outside leisure-time is an important contribution to fulfilling PA recommendations [32–35]. Guthold et al. [15] examined total PA data corresponding to 51 countries taking part in the World Health Survey, 2002-2003, and observed a mean physical inactivity prevalence of 17.7% (15.2% men, 19.8% women). These same authors estimated the physical inactivity rate in Spain at 27.5% for men and 32.9% for women. Although still high, this prevalence of sedentarism, defined as not meeting PA recommendations considering total PA, is far from the 71-76% inactivity in leisure-time observed for adults in Madrid both in our study and others [14, 30, 31].

These figures indicate a clear need to specifically encourage LTPA targeted at reducing the prevalence of sedentarism among adults. Specially, since some studies suggest that LTPA, rather than PA performed in other domains (active transport, work or habitual occupation), is positively associated with health indicators such as a reduced obesity and cardiovascular risk, and improved health perception [17–20].

The correlates observed in the present study between the prevalence of leisure-time sedentarism and several sociodemographic variables, are in large measure consistent with the trends described in the literature, except for those related to education level or employment status. In this regard, a recent systematic review [36] identified clear differences in LTPA levels according to socioeconomic position (SEP). The latter was defined by indicators such as income, education level and social class (based on occupational class). Most of the studies reviewed by these authors indicate that subjects with a higher SEP were more physically active during their leisure-time than those with a lower SEP.

Considering only the income indicator, most investigations have also detected a positive correlation of this indicator with LTPA (59% for total LTPA and 75% for vigorous LTPA) [36]. These findings are in line with those of the present study, given that we identified the group of participants with a high or middle to high economic status (based on own or family incomes) as showing a higher prevalence of a moderate to high LTPA level. In addition, these data were further supported by a greater tendency towards sedentarism observed among the participants with a middle economic status and, possibly, among those who failed to provide information about their income level. Indeed, some studies have correlated a lack of a reply to such a question with a low economic status [37, 38].

With regard to the level of education, this indicator has been positively correlated with LTPA [16, 35, 36, 39–42]. A commonly provided explanation is that people with a higher education level are more likely to understand the health benefits of regular PA, translating to higher LTPA levels. However, our results suggest that the likelihood of being sedentary during leisure-time is not related with education level. Although such lack of association is not in accordance with the majority of studies in the field, a recent meta-analysis by Beenackers et al. [36] found that LTPA was not associated with educational level in 27% of the studies reviewed. Further, a study conducted in a representative Spanish Mediterranean cohort showed that LTPA was unaffected by educational status in either gender [43].

When analysing the association between employment status and LTPA levels in our cohort, we found that those subjects whose habitual occupation was housework, were the only ones who had a lower likelihood of being sedentary during leisure-time. These findings are partly in line with those of Khaing et al. [35]. These authors noted that people that dedicated most of their time to housework and retired subjects were the population subsets with the lowest probability of not meeting PA recommendations, considering total PA. As PA was examined across each domain, retired people and students showed the greatest LTPA levels. However, persons doing housework as their main activity, showed a LTPA energy expenditure that was among the lowest compared to the remaining groups, indicating that PA performed in the home was the greatest contributor to the total PA level.

Sex and age are the variables that are most consistently associated with the PA behaviour of adults. Accordingly, men are usually more active during leisure-time than women and age is usually inversely related to LTPA level [16, 42, 44]. The results of the present study also confirm these trends, indicating that men and the younger study participants (18-30 years) are the two subsets of adults in which greater percentages exist of individuals performing a moderate to high level of LTPA. Similarly, women and participants aged 41 to 50 years emerged as the groups with a greater likelihood of being sedentary in their leisure-time. In addition, as in prior studies [14, 45], our results indicate that the difference between men and women becomes more evident as the intensity of LTPA increases (data not shown).

The relationship between marital status and PA established in the literature has been inconsistent [42]. Thus, while some authors have observed no link between these factors [46, 47], others have established that married compared to single persons, are more likely to be physically inactive in their leisure-time [31, 48, 49]. As in these latter studies, our results also indicate that married people, as well as separated/divorced individuals, are more likely to adopt a sedentary behaviour during their leisure-time than singles. It could be that the family, social and work responsibilities of married subjects, are the underlying cause of this difference. However, the relationships we found between marital status and PA were overall weak and disappeared in the more adjusted regression Model 3.

Recent reviews have shown that a parental model of PA can have a huge impact on the PA behaviour of children [50, 51]. Several studies have noted that when mothers and fathers undertake regular PA, their children are more likely to do the same [52–54]. Although we examined an adult population, the results obtained reveal that subjects whose mother and/or father performed or had in the past performed regular PA, formed a group that showed a higher prevalence of high LTPA levels. In contrast, participants who lacked such a model were more likely to show sedentary behaviour during their leisure-time.

The limitations of our study that could have influenced its results are first and foremost its cross-sectional design, which prevents extracting causal relationships among the variables analyzed. The lack of data on comorbities is another limitation, because ACSM recommendations on PA mainly apply to healthy subjects. In addition, we cannot rule out overestimation of PA levels by the study participants, since these were self-reported rather than obtained through objective methods [55]. However, a methodological strength, as well as a novelty, of the present study is the use of GPAQ as a measure instrument in a wide and representative cohort of adult population from the Madrid region. Indeed, this instrument was specifically designed to assess PA patterns in their different domains, and has proven valid and reliable for PA assessment [28].

The results of our study suggest that three out of every four adults living in Madrid show sedentary behaviour in their leisure-time. This high prevalence of sedentarism identifies a need for public health policies designed to promote LTPA. Our results also identify the following groups of people as more likely to adopt a leisure-time sedentary behaviour than their respective reference groups: those aged 41 to 50 years, individuals with middle and probably, with lower economic status, women, and those who lack a parental LTPA model. This type of information could be useful to decide which population groups would perhaps benefit most from intervention strategies designed to increase the LTPA of Madrid’s inhabitants. On the other hand, more research is needed using objective assessment of PA (e.g., accelerometry). Future strategies aiming at promoting PA should also take into account several important environment-related variables that were not analysed here, such as urban design, transport system or availability of recreation areas and green spaces [56, 57].