The present investigation examined the role of treatment of the interruptions in activity on the evaluation of PA bouts with regard to accelerometer-based PA assessment in middle- to older-aged individuals. The main finding of the present investigation is that the frequency and duration of the PA bouts were significantly lower and shorter when a one- or two-minute interruption was allowed, compared with that when no interruptions were allowed (Table 3), whereas the frequency and duration of the continuous MVPA bouts was significantly associated with the non-continuous MVPA bouts including a one- or two-minute break. As a result, the number of days accumulating at least 30-min of MVPA bouts allowing a one- or two-minute interruption was significantly higher compared with the number of days accumulating at least 30-min of continuous MVPA bouts without an interruption. Although several studies examined the effects of the treatment of interruptions on the accelerometer’s out puts [8, 9], the present investigation firstly showed that both of the daily frequency and the daily duration of MVPA bouts significantly differed according to the choice of the treatment of interruptions. Furthermore the frequency and duration of MVPA bouts gradually increased from when allowing no interruptions to when including 1-min or 2-min interruptions, additionally these significant difference were also found between that when allowing 1-min or 2-min interruptions and when allowing 2-min interruptions. These results indicate that, in the accelerometer’s outputs analysis, the treatment of the interruptions affects the estimation of the MVPA bouts under free-living conditions in middle-aged to older adults. Both the frequency and the duration of the MVPA bouts would be significantly higher depending on the allowing duration of interruptions. Thus, caution should be exercised regarding the accelerometer’s data setting when discussing the MVPA bouts reported in previous publications.
One of the original findings of the present investigation is that, with regard to the accelerometer-based MVPA assessment, allowing one- or two-minute interruptions results in a larger number of MVPA bouts compared with that when not allowing any interruptions under free-living conditions in middle aged to older adults. These findings are supported by the previous findings, where two studies demonstrated that allowing short interruptions increased the frequency of the MVPA bouts compared with that not allowing any interruptions [8, 9]. Miller et al. reported that the frequency of MVPA bouts was 0.5 ± 0.8 bouts/day for the continuous MVPA bouts, 0.6 ± 0.8 bouts/day for the non-continuous MVPA bouts allowing a one-minute interruption and 0.6 ± 0.9 bouts/day for the non-continuous MVPA bouts allowing a two-minute interruption [8]. Similarly, Masse al. reported that the frequency of MVPA bouts was 3.0 ± 6.8 bouts/day, 4.1 ± 7.8 bouts/day and 4.9 ± 7.9 bouts/day for the same measurements [9]. In the present investigation, the frequency of the MVPA bouts differed from 0.60 ± 0.68 to 1.02 ± 0.88 bouts/day. Based on the these findings, a range of 0.1 to 1.9 bouts/day of MVPA might be detected due to the treatment of the interruptions, and these differences resulting from the different criteria would be increased based on whether a one-minute or two-minute interruption was allowed. As a result, the duration of MVPA bouts allowing a two-minute break (21.28 ± 22.37 min/day) was almost double the duration of MVPA bouts without interruptions (11.43 ± 15.23 min/day). Furthermore, the duration of the MVPA bouts allowing a one-minute break differed significantly compared with the MVPA bouts allowing a two-minute break, the same as was noted in previous investigations [9]. These finding clearly indicate that the treatment of the interruption has a significant impact on the estimation of MVPA bouts under free-living conditions, and these differences may be also found between the MVPA bouts allowing a one-minute break and the MVPA bouts allowing a two-minute break.
Based on these findings, the large variability in the reported MVPA bouts has likely been due to the choice of the treatment of the interruptions. The duration of MVPA bouts allowing a one-minute break was 25.8 ± 23.4 min/day in normal weight subjects [5]. Furthermore, the duration of MVPA bouts allowing a two-minute break was 9 ± 13 [20], 9.1 ± 0.5 and 6.6 ± 0.5 min/day in males and females from the NHANES 2003-2006 study [21], 13 to 16 min/day in Swedish adults and 6 to 10 min/day in US adults [14]. In contrast, the duration of the continuous MVPA not including interruptions has been much shorter. Davis et al reported that the continuous MVPA bouts lasting longer than 10-min were 0.4 to 0.6 min/day in British, Italian and French volunteers, whereas the duration of the sporadic MVPA was 19.9 to 39.4 min/day [15].
There is no doubt that the treatment of the interruption has a significant impact on the estimation of MVPA bouts under free-living conditions, because these differences are dependent on the data treatment process. For example, if individuals repeated a five-minute walking bout following a one minute break, the MVPA bouts would be described as lasting zero minutes, 11 minutes and 17 minutes when the analysis allows no interruptions, a one-minute interruption and a two-minute interruption, respectively. Furthermore, the usual habitual physical activity mainly consists of intermittent activities [22]. In developed countries, purposeful walking may be frequently stopped by traffic signals and/or crowds, so the METs value will be below 3 METs. Thus, the magnitude of the impact of the interruption treatment on the MVPA bout estimation would be increased in the individuals accumulating MVPA bouts from habitual PA (walking to work, shopping, gardening, etc.) rather than for the individuals participating in purposeful continuous PA (jogging, running, sports activities, etc.).
It should be noted that the present investigation does not indicate that not allowing an interruption is the best procedure for the MVPA bouts estimation. Masse et al. [9] suggested that it appeared reasonable to allow a one- or two-minute interruption anytime during the bout, because extracting MVPA bouts has been used to determine whether the participant met the current physical activity recommendations. We agree that this may be more reasonable.
However, in addition to the treatment of interruptions, the choice of the epoch length is an important contributor to the MVPA bouts estimation [6]. Furthermore, not only waist acceleration signals but also physiological stress, such as the oxygen uptake and/or heart rate response, and direct observation should be considered to define the optimum procedure for the data treatment of the accelerometer outputs for the MVPA bouts estimations.
There are several limitations associated with the present investigation that should be considered when interpreting the results. First, the type (uni-axial) and position (waist) of the accelerometer used in the present investigation does not allow for the collection of upper body activity, and thus may underestimate the total PA. Second, the participants evaluated in the present investigation were primarily middle-aged, non-active males and females. Furthermore, the participants all lived in urban areas, and buses and trains were their primary means of transportation.