This is the first study to examine the relationships between the type of sustained MVPA defined by the Physical Activity Guidelines for Americans and sitting, using objective measures. The results of this study support the hypothesis that exercise and sedentary behaviors (e.g., sitting) are independent classes of behavior [3
]. These findings indicate that time spent sitting is not related to time spent in guideline defined sustained MVPA and that, between and within individuals, variations in this MVPA are not significantly associated with sitting, standing, or intermittent stepping time. Our study sample included a relatively large number of women meeting and exceeding the current activity recommendations for MVPA, with the mean of the entire sample averaging 146 min of MVPA/week. In spite of this enhanced time spent in MVPA (as compared to those in the general population), women in our study still spent the majority of their waking day (~63% of their waking hours) sitting. Further, time spent in sustained MVPA comprised only a small fraction of their waking time each day (~2%) and the time spent in MVPA appears to displace only a very small and non-significant amount of time spent in the domains of sleeping and sitting.
We are unaware of any other study that used direct and objective assessment measures from accelerometry and inclinometry to determine the relationship between sedentary behaviors and sustained MVPA (as defined by consensus recommendations of at least 10 minute bouts). However, there are studies that examined the relationship between total accumulated time spent in MVPA (irrespective of bout duration) and total time the body is accelerating below 100 counts per minute [22
]. While those studies generally show a strong inverse relationship of total low-intensity physical activity with sedentary behavior, they also consistently found low to moderate correlations (r= −0.27 to −0.66) of sedentary behavior with total accumulated MVPA. However, prior work has not specifically focused on the relationship between sustained MVPA performed in at least 10 minute bouts and sedentary time, as we do in the present study. Importantly, in the US population, on average, only about 30% of weekly MVPA is performed in bouts that are of at least 10 min in duration [33
]. Thus, previous studies reporting the relationship between sedentary duration and total accumulated MVPA were most likely examining predominantly intermittent MVPA that is not counted toward the weekly goal of 150 min, as defined by the current guidelines. This would include activities like taking out the trash or walking from the parking lot to the store [1
]. Because those previous studies using accelerometry did not specifically assess activity duration as defined by the guidelines, one cannot draw conclusions about the relationship between guideline defined sustained MVPA and sedentary behavior.
While hip worn accelerometry provides valuable objective information beyond self-report, an advantage of accelerometers with inclinometry, as in the present study, is that they extend the insights to posture (sitting). A recent study using direct observation, in combination with the activPAL and Actigraph monitors, has shown that while the activPAL agrees well with the direct observation of sitting (r=0.94), the hip worn accelerometer (Actigraph), with a 100 counts/min cutpoint, correlates less well with direct observation of sitting (r=0.39). Furthermore, the hip-worn accelerometer was not able to detect a change in sitting behavior caused by a sitting reduction intervention, while the activPAL captured this change [25
]. Therefore, the present study is the first, to our knowledge, that used a validated objective measure of sitting to determine if sitting is a distinct behavior or if it is related to the amount of guideline defined sustained MVPA one performs.
There are also many studies that used self-report to examine the association between sedentary behavior and total sustained MVPA. Generally, these studies did not examine sitting over the entire day, but rather focused on either specific sedentary behaviors, like TV watching, or on assessing specific periods of the day, such as leisure time sitting only [8
]. In a recent study that attempted to assess the relationship between total daily sitting and physical activity in 20 nations using the IPAQ, Bauman et al. reported that there was an inverse relationship between self-reported total MVPA and self-reported sitting [6
]. They noted that this finding disagreed with most other studies and speculated that this was due to the IPAQ targeting all domains of physical activity instead of just specific periods of the day like leisure or occupational sitting only. These results are difficult to compare to our current findings given the self-report nature of the measure in that study and the fact that sitting may be under-reported and physical activity over-reported. Proper et al. also used the IPAQ to assess sitting and sustained total MVPA in 1048 Australian adults. They reported an inverse relationship between physical activity (i.e., putatively sustained MVPA) and total daily sitting time but not between MVPA and leisure sitting duration [7
]. A point of concern in both studies is that the IPAQ sitting question (i.e., how much one typically sits in the previous seven days) has not been validated against an objective measure of posture like the activPAL or against direct observation. Further, it only weakly agrees with the Actigraph sedentary cutpoint (<100 cts/min) method, which itself is not a strong measure of sitting behavior [25
Our data show that there was no difference in sitting time between those who met or exceeded the 150 min/week recommendation (far more than the general American population [36
]) and the most inactive women (Figure A). In addition, there was not even a significant trend, with the women who spent 450–600 min/week in sustained MVPA still sitting ~550 min each day (Figure ). Thus, the results of the present study clearly show that regardless of the amount of sustained MVPA accumulated, with some women accumulating 4-times the 150 min/week benchmark (Figure ), more time spent in sustained MVPA is not associated with reduced sitting time.
Many people sit so much (~60-80 hrs/week) that it is impractical to replace significant amounts of total daily sitting time with the type of sustained MVPA recommended for the public. We found that variations in sustained MVPA do not significantly correlate with sitting time, and thus other approaches are necessary for reducing sedentary behavior. This has been alluded to in prior studies that implemented interventions aimed at increasing MVPA but which saw no concurrent decrease in self-reported sitting time [37
]. However, our data do show that sitting time was significantly and inversely related to intermittent and predominantly low-intensity upright activities. This has relevance to previous research indicating that even LIPA is apparently important for human health [15
] because of potent molecular responses related to low-intensity contractile activity locally and specifically caused by postural skeletal muscle [17
There are several strengths and limitations to the current study that should be noted. This is the largest study to date to utilize valid and objective measures (inclinometry) of sitting time and guideline defined sustained MVPA in a cohort of middle and older-aged adult women. In addition, we were well-positioned to examine our study hypothesis as our sample included many women meeting and exceeding the federal physical activity recommendations of 150 min/week of sustained MVPA. Consistent with the Physical Activity Guidelines for Americans, the MVPA bouts of at least 10 minutes were quantified. The determinants, or at least the physiology, of this type of sustained activity is different than the more intermittent and short bouts of activity. With respect to limitations, this was a cross-sectional study of women only and we cannot determine causal relationships between sitting time and time spent in sustained MVPA, nor can we generalize to men or other populations. It is also plausible that there are other unmeasured factors that could influence the relationships we observed in this study.