This paper examined the prevalence and correlates of weekday sitting time among a large sample of European adults of the 32 Eurobarometer-participating countries. Population-based multi-country studies of sitting prevalence that use standardised instruments are scarce. Our study is the first, to our knowledge, to provide a comprehensive comparison of sitting prevalence across a large cross-section of European countries.
The key finding of our analysis of the Eurobarometer 64.3 was that IPAQ assessed usual weekday sitting time was geographically distributed, with countries within north-western Europe generally reporting higher sitting times than countries situated within south-eastern Europe (Figure 1). Within the scope of the present study it is possible only to speculate on the causes of this geographical pattern. However, an explanatory factor may be the unequal distribution of wealth between the more affluent north-western countries, when compared to less wealthy south-eastern Europe countries. Wealth inequalities could theoretically influence sitting time across all sitting domains (e.g. occupation, transport, leisure-time and household). Within the occupational setting, it is likely north-western countries have greater proportions of adults in white collar, office-based occupations (which require greater volumes of sitting), and greater exposure to technology at work (e.g. computers and other labour saving devices). Moreover, unequal wealth disruptions across Europe may result in greater proportions of passive transport (car use). For example, when compared to south-eastern Europe countries, north-western countries may be more likely to have building and transportation practices that require car use for most trips, . Within the household domain, it is also possible that those within north-western countries have greater access to labour-saving devices and technology (e.g. internet, electronic entrainment) . Although these data were collected in a similar time-frame (November-December 2005), it is possible that climatic factors may partially explain differences in sitting times across countries. For example, northern European countries tend to be in the highest quartile of sitting time, where the climate is colder than in southern European countries where fewer people belonged to the highest quartile. Furthermore, cultural differences in interpreting sitting questions and reporting biases may have contributed to some of the differences in sitting time between countries .
The present study showed that usual weekday sitting time among 32 European countries was 309 minutes/day. This daily sitting time was slightly lower than what was observed by the IPS which reported a mean sitting time of 346 minutes/day . When compared to population studies using objective assessment instruments, sitting time may be underestimated by self-report measurements [38, 39]. For example, sitting data from the IPS and Eurobarometer 64.3 indicated that adults self-reported 5–6 hrs of sitting per day. However, accelerometer data from several national population representative samples has shown daily sedentary time to be 8.8-11.2 h/day [13, 40–42]. Despite the disparity between objective and subjective tools, it is not clear whether the prevalence of sitting observed in this study are harmful to health because at present there is no consensus among researchers surrounding the dose–response relationship between sitting and detrimental health outcomes [43, 44]. In parallel with the wide variations in physiological adaptations to exercise training , it is likely that the dose and volume of sitting and associated health consequences differs from individual to individual [45, 46]. Moreover, the way sitting time is accrued within the context of a whole day may be important . Recently, controlled laboratory studies further examined the acute cardiometabolic effects of breaking up prolonged sedentary time [47, 48]. In these experiments, when prolonged sitting was displaced with light- or moderate-intensity walking, there was a significant lowering of postprandial glucose and insulin levels , and improved whole-body insulin sensitivity .
The adjusted analyses identified that being low-active, being in poor health and having high sitting usual activities were among the strongest correlates of reporting high sitting time. These findings are somewhat similar to other studies [28, 33, 34]. For education, our results and those from the IPS , suggest that high levels of education are associated with higher levels of sitting. This result may be explained by the fact that those with higher education levels may have occupations that require higher volumes of sitting. Although we assessed sitting time in a somewhat younger population than the IPS study (EB 64.3: ≥15 years vs. IPS: ≥18 years), both studies show a trend for younger people to report higher sitting times. Potential explanations may be that in contrast to older adults, a greater proportion of younger adults may be students, which may require high sitting volumes [28, 49, 50]. Moreover, despite younger adults having a greater amount of leisure time, time use studies have shown that 85-90% of their leisure time is spent sedentary [49, 50]. Last, younger adults may be less likely to household activities (which are often not sedentary) . These self-report data on sitting time and its relationship to age are to some extent inconsistent with studies using objective physical activity assessment tools. For example, in the U.S. population NHANES 2003–2004 study, accelerometer-defined sedentary time showed a linear trend, with sitting times increasing with increasing age . However, a potential explanation of these inconsistent findings may be the age differences of the participants assessed in the EB 64.3 (15-18 years) and NHANES 2003–2004 participants (>20 years). A recent review of population studies assessing accelerometer-defined sedentary time of Belgian adults and children (n = 2,083) showed that those aged between15-18 years had the highest levels of sedentary time when compared to other age groups . These conflicting findings highlight the need for further research that objectively assesses cross-country sedentary time within population-representative samples and among a wide variety of age groups.
The findings around sitting time and physical activity levels are consistent with previous research, with the IPS study showing a similar inverse relationship between sitting time and levels of physical activity . This finding differs from some small-scale research using objective assessments of physical activity and sedentary behaviour patterns. For example, when accelerometer-assessed physical activity and sedentary time were examined among participants in short-term exercise studies, those engaging in high levels of moderate-to-vigorous exercise may compensate for high activity levels by being more sedentary during non-exercise periods [52, 53], resulting in no net gain in energy expenditure . While the present study does not support this hypothesis, it may be possible that adults reporting high physical activity levels may under report their sitting time. The possible compensation for increased sitting time among adults with high levels of physical activity warrants further investigation in population-representative samples. Furthermore, objective assessments tools, such as accelerometers and inclinometers, should be used to examine relationships between sitting and physical activity behaviour patterns. However, population studies that implement objective physical activity assessment tools have significant cost and logistical issues .
A novel aspect of this study was the description of the high-sit/low-active and low-sit/high-active specific sub-groups within the Eurobarometer 64.3 sample, which reflects the combined risks of sitting and levels of physical activity. The cross-country distributions of these sub-groups suggested another potential geographical pattern. There was a greater tendency for adults from north-western European countries to be classified as high-sit/low-active. In contrast, those within south-eastern countries were more likely to be classified as low-sit/high-active. While this trend requires replication, this geographical pattern suggests that adults within north-western European countries may be at risk of health consequences associated with a combination of high volumes of sitting and low levels of physical activity. Another observation was that some countries that scored high on sitting prevalence (Netherlands, Denmark, West-Germany), scored relatively low in the high-sit/low-active category. This seems to indicate higher levels of physical activity in these countries, which might be partly due to the good active transportation infrastructure in these countries. In contrast, Great Britain which has poorer active transportation infrastructure scored much higher on the high-sit/low-active category. More educated adults were classified more frequently as high-sit/low-active, and less frequently as low-sit/high-active. These findings differ from previous research examining physical activity levels without relation to sitting time. Research has consistently shown a positive association between increased education levels and high physical activity levels [23, 55]. Although the analyses adjusted for usual activities, this finding may be explained by higher educated adults employed in occupations that require higher volumes of sitting. Future studies should continue to assess sitting and physical activity patterns concurrently. Research examining the prevalence and correlates of high-sit/low-active individuals may be used to target at risk populations in intervention studies.
Strengths of this study include the recruitment of a large sample of adults across a large number of European countries. This resulted in a reasonably heterogonous sample, making it possible to compare sitting data across various sociodemographic factors. A further strength was the use of a standardised sitting time assessment instrument, which makes it possible to compare the findings of the present study to future research. Limitations include the cross-sectional design, which makes it difficult to infer causality from the study findings. Moreover, given that the modest response rate of 54.6%, there may be limitations around the generalisability of these results. Further limitations included the use of self-report measures of sitting and physical activity, which may result in social desirability and recall biases . Furthermore, it is also possible that those who participated in the study may have different sitting patterns than non-participants. Therefore, these factors might have resulted in the under-reporting of sitting time, which may suggest our estimates of sitting time in Europe are on the conservative side. However, given the large sample size across a wide-variety of countries, objective sitting and physical activity assessments (e.g. accelerometers and inclinometers) were too difficult to implement due to high cost, logistical issues and participant burden. Furthermore, despite limitations around the validity of self-reported methods, among large samples, standardised self-report tools have a use for ranking individuals sitting time and physical activity levels .