The aims of this study were to objectively investigate the time spent in SB and PA in a lifespan perspective by investigating accelerometer data in different age groups (preschoolers, primary schoolchildren, secondary schoolchildren and adult) and to calculate the balance between time spent in LIPA and SB.
The first finding of the present study was that Belgian preschoolers, primary and secondary school children, and adults (3–65 years old) spent on average 55% of their waking time (7.3 hours) in SB (range 51%-59% for different age groups), which is equal to the prevalence rate found in the US population (6–85 years old) . The average time spent in LIPA (39% or 2.9 hours a day) and MVPA (6% or 45 minutes a day) found in the Belgian sample was difficult to compare with other accelerometer studies as different age ranges, cut-off points or data reduction rules were used. Nevertheless, our results confirmed that sedentary time constitutes the bulk of waking hours in Western countries, with the remainder of time disproportionately spent in LIPA and MVPA. Sitting time is therefore an important target behavior in health behavior changes programs.
A second important finding of our study was that differences in sedentary time and PA levels between age groups were gender dependent. Our results revealed a linear relationship in men, namely higher percentages in SB (and lower percentages in PA) with increasing age, resulting in a maximum amount of 59% of SB in adults. However, in women a reversed U curve was found, identifying secondary schoolchildren as the most sedentary group with 61% of their waking time being sedentary. This is in line with the findings of Matthews et al.  who reported a peak of sedentary time (60%) in US adolescents (16–19 years old). However, in contrast with the current study, Matthews et al. found also a reversed U-curve in adolescent boys, who spent more time in SB compared to adults under 50 years old.
Due to the cross-sectional study design it is unclear whether the age-related gender effects are influenced by developmental changes or whether cohort effects are present. Hypothesizing that cohort effects are minimal between two consecutive age groups, it could be assumed that secondary schoolchildren (when compared to primary schoolchildren) replace LIPA by SB, whereas percentage of time spent in MVPA is still very low (6%). In adulthood (when compared to secondary schoolchildren), women reverse the pattern by decreasing their sedentary time and increase their PA levels, but only at light intensity. The latter showed that high prevalence rates of sedentary time found in secondary schoolgirls are partly restored by itself during adulthood. This remarkable substitution of SB into LIPA in adult women is probably associated with an increase in time spent doing household chores and child care, activities mostly done at light intensity . In contrast, our data reveal that adult men become another important risk group as their amount of SB increases with increasing age. They probably replace their sports activities (MVPA) during adolescence by SB in adulthood. Large longitudinal studies with objective assessments, supplemented by context-specific self-reports, are needed to confirm these hypotheses and exclude potential cohort effects.
The finding that Belgian secondary schoolgirls spent more time in SB compared to secondary schoolboys is in accordance with the accelerometer data found in US youth , but in contrast to a previous European study using self-report data in adolescents . This contradiction stresses the importance of accurate methods to measure SB as it seems that self-reported screen time behaviors do not capture overall sedentary time, especially in adolescent girls [27, 70, 71].
A third important finding of our study is the low prevalence rate of participants with a positive LIPA-SB balance, namely 18% of the total sample. Based on the studies of Healy et al. , Hamilton et al.  suggested that a positive balance between time spent in LIPA and SB might be health beneficial in adults To our knowledge, no prevalence data for this balance have been reported in the literature yet. Our results showed that the smallest proportion (6%) of a positive LIPA-SB balance was found in secondary schoolgirls. Recent research has proven the independent effect of SB on health in adults , , ,, , , , ,  which means that, to profit from the health benefits, people should 1) fulfill the PA guidelines and 2) limit their time spent in SB. In our Belgian sample, 36% fulfilled the PA guidelines, but only 10% combined sufficient PA with a positive LIPA-SB balance. In secondary schoolgirls, only 1% combined a positive LIPA-SB balance with sufficient MVPA each day. Also remarkable is the high proportion of adult men (43%) who are combining sufficient MVPA with a negative LIPA-SB balance. Those men are probably not aware of their increased health risk caused by high levels of sedentary time, as SB was not included in the PA recommendations promoted during the last decennia. Although the present study has revealed that there are groups of ‘active men’ that also have a negative LIPA-SB balance, it should be further investigated if this LIPA-SB balance is indeed a good health indicator in itself. Also the evidence for the negative health impact of SB and the positive health impact of LIPA is consistent in adults but until now mixed results are found in youth . So more research on the health impact of SB, LIPA and balance LIPA-SB in the younger age groups is needed.
In PA literature, adolescent girls have been recognized as an important target group for interventions . Based on the results of the current study, we recommend that next to promoting MVPA, the reduction in SB in this target group is very important as well. As adolescent girls were not identified as an at risk group for SB in studies using self-reports, it is possible that existing questionnaires insufficiently capture those types of sedentary activities adolescent girls engage in. Qualitative studies with adolescent girls can provide more in depth insight into the SB contexts.
Interventions aimed at decreasing SB and increasing PA, should therefore not only focus on MVPA but also on LIPA. The highly negative correlations between SB and LIPA found in our study (r=−0.94) and in the study of Healy et al.  (r=−0.96) indicate that replacing SB by LIPA is a good alternative to decrease time spent in SB in people who are not able or willing to increase their MVPA levels. Further studies should also determine which types of sedentary behaviors are most common in each age group and which behaviors are most feasible to replace by which types of LIPA.
Strengths and limitations
This study objectively assessed SB in a large sample and is one of the first to report prevalence rates about the LIPA-SB balance as recommended by Hamilton et al. . The results of our study can probably be generalized to the Belgian population, due to the sample size, different age groups and the probability sampling method.
However, this study also has several limitations. First, the study design was cross-sectional which comprises that age-effects could be confounded by cohort effects.
Secondly, accelerometer data were collected from different PA studies for this secondary data-analysis. One of the consequences was that a 15 seconds epoch time was used in the youngest age group, whereas a 1 minute epoch time was used in the other age groups. Data of the preschoolers were transformed to cpm, however Corder et al.  showed that a different epoch length affects the estimation of time spent in different activity categories. Another consequence was that we used the same PA guideline for preschoolers  as for primary and secondary schoolchildren (60 minutes MVPA/day). However, recent guidelines for preschoolers developed in the UK  and Australia  do not longer take into account the intensity of PA but recommend 3 hours a day of total PA.
A third limitation is that by using accelerometry only the amount of SB and PA could be collected but nothing is known about the behavioral context. Therefore, questionnaires that assess a wide range of sedentary and PA behaviors, remain important to gain additional information that is necessary to develop interventions.
Further, cut-offs were used to categorise activities as SB, LIPA or MVPA. In the literature, the 100cpm cut-off for SB is very common [29–34, 55–57]. However, it has been shown that a hip-mounted accelerometer, measuring accelerations in the vertical plane, cannot correctly make the distinction between sitting and standing still . According to the definition of The Sedentary Behavior Research Network , standing still is not categorised as SB but as LIPA because standing requires muscle contractions. Therefore, the amount of time spent in SB measured with accelerometers could somewhat be overestimated by incorporating periods of standing still, and as a consequence LIPA somewhat underestimated. Future studies could overcome this bias by taking into account posture, assessed for example by an inclinometer .
Also the cut-offs for LIPA and MVPA are often subject of discussion in the literature and differ between age groups. Therefore it is difficult to compare our PA results with other accelerometer studies using different cut-offs.
A final limitation of the study is that we did not take into account the bouts of both SB and PA. For SB, it has been shown  that not only the amount of SB (minutes a day) but also the breaks (prolonged sitting versus regular breaks) have an influence on health parameters. It was beyond the scope of this paper to make differences between “prolongers” and “breakers” but future research should take this into account. With respect to PA behavior, the guidelines of MVPA for adults include that minutes can be accumulated throughout the day, but in blocks of minimum 10 minutes. The bouts of PA were not analysed in our study and therefore the percentage of participants fulfilling the guidelines could even be lower than the percentage reported here. In the study of Hagstromer et al.  the proportion of the adult sample that fulfilled the PA guidelines assessed with an accelerometer decreased from 52% to 1% when only blocks of at least 10 minutes MVPA were included.