Background
Adolescents have particularly high levels of sedentary behavior with breaks in sedentary time markedly decreasing as they age, making them vulnerable to possible adverse health effects [
11]. Current literature investigating the effects of sedentary behavior on cardiometabolic markers in adolescents is limited with mixed results. A crossover trial of non-overweight adolescents fed a healthy meal while undergoing three experimental conditions to assess the impact of reduced sitting time on cardiometabolic outcomes produced no measurable changes in cardiometabolic markers [
46]. An experimental trial involving 10 to 14 year old children who consumed a high fat meal, interrupting sitting time with bouts of moderate to vigorous physical activity (MVPA) resulted in a reduction in triglyceride concentrations [
43]. Another trial reported that interrupting sitting with MVPA reduced insulin and free fatty acid concentrations in children aged 7–11 years [
6]. Few studies have investigated the impact of breaking up sitting with light physical activity (LPA) breaks on cardiometabolic outcomes. One trial assessed the impact of a LPA treatment compared to a sitting day in 10 to 18 year olds and reported observing significant treatment by time interactions in HDL cholesterol and insulin, however the treatment effect was not significant [
48]. Collectively these studies suggest that reductions in sitting time are associated with positive health outcomes in adolescents but more research is needed.
Adolescent spent large portions of their day sitting in controlled environments such as schools [
9]. Thus, experimental studies conducted in school setting may be more generalizable to large proportions of the adolescent population. Adolescents have fewer breaks in sedentary time during the school day than in any other period of the week [
30]. To date, experimental studies have been conducted under controlled conditions, often testing the effect of a high-fat or high-carbohydrate meal on energy expenditure. As such, they are not indicative of a balanced dietary intake. There are no experimental studies which replicate a school day.
There is also growing interest in the effects of prolonged sitting on cognitive outcomes [
54]. Particular interest has been directed toward working memory, a core ‘executive function’ for activating, maintaining and manipulating information in the mind given its relation to academic achievement [
1], learning, reasoning, and cognitive control [
28]. Initial evidence suggests that breaking up sedentary time may improve cognitive outcomes in children [
16]. However there is a dearth of literature assessing the effect of reducing and breaking up bouts of sitting time in the school setting. An understanding of the effects of sitting time on both cognitive and health outcomes in the classroom setting could inform future classroom interventions with potential positive long term effects for children.
This study aimed to examine the acute effects of reducing total and prolonged bouts of sitting time compared to a traditional sitting time during an adolescents’ school day on cognitive function and cardiometabolic outcomes in a laboratory setting. It was hypothesised that there would be more favorable changes in cognitive function and cardiometabolic health outcomes following a simulated “reduced sitting” school day when compared to a simulated typical day and that participants would not compensate by altering their energy expenditure or intake. This intervention is particularly novel in that it replicated a typical school day and thus holds potential for translation to the school setting in future interventions.
Results
The rate and sequence of participation throughout the study is shown in Fig.
1. Between May and September 2014, 19 potential participants were assessed for eligibility. Eighteen met the inclusion criteria, attended the familiarisation visit (Visit 1) and consented to participate. Seven participants were classified as overweight or obese using body fat reference curves for children [
34]. Participant characteristics at baseline are shown in Table
1. Protocol compliance for both conditions was assessed by direct observation and written records, participants did not deviate from the protocol in either condition.
Table 1
Characteristics of Study Participants
Female (% number) | 7 | 5 | 2 |
Male (% number) | 11 | 9 | 2 |
Age (years) | 13.5 ± 0.9 | 13.4 ± 1.0 | 13.7 ± 0.8 |
Height (cm) | 161.7 ± 10.0 | 162.3 ± 11.0 | 159.6 ± 5.9 |
Weight (kg) | 55.8 ± 14.3 | 55.5 ± 15.5 | 56.8 ± 11.1 |
BMI (kg/m2) | 21.1 ± 3.9 | 20.8 ± 3.8 | 22.3 ± 4.7 |
Body Fat (%) | 23.0 ± 7.6 | 21.5 ± 5.6 | 28.1 ± 12.1 |
Overweight/obese (%) | 7 (38.9%) | 6 | 1 |
Blood Pressure- systolic/diastolic (mm Hg) | 115.2/69.7 ± 13.8/7.6 | 114.1/71.4± 13.44/7.1 | 118.8/63.8± 16.4/6.9 |
Resting Heart Rate (bpm | 78.1 ± 13.6 | 77.2 ± 8.2 | 82.5 ± 37.3 |
Each individual consumed the same food items during both conditions (average energy intake 4762.6 kJ ± 1596.8 kJ).
Data from four participants were incomplete as the nurse/phlebotomist was unable to draw blood from these participants during Visit 2 and additional data were missing due to processing problems at the pathology lab. As a result, cardiometabolic data were available for 14 participants for the apoB/apoA-1 ratio, 13 complete sets of data were available for total cholesterol, LDL cholesterol, HDL cholesterol, Non-HDLC, triglycerides and glucose and 10 complete sets of data were available for IL6. In addition one outlier was removed during the analysis for LDL cholesterol and IL6 (see Table
2).
Table 2
Condition and effects of experimental conditions (reduced sitting versus typical) on cardiometabolic outcomes (mean +/− SD) of total study participants (n = 18)
Figural Intersection task (Cognitive function)d
| 6.06 ± 0.10 | 5.89 ± 1.71 | −0.17 ± 0.15 | 5.71 ± 1.53 | 6.18 ± 1.47 | 0.47 ± 1.231 | 0.64 ± 0.15 | 0.149 | 0.54 |
ApolipoproteinA1/ ApolipoproteinB (g/L)b
| 0.52 ± 0.20 | 0.51 ± 0.20 | −0.01 ± 0.03 | 0.52 ± 0.16 | 0.49 ± 0.15 | −0.03 ± 0.03 | −0.02 ± 0.03 | 0.027 | −0.67 |
Interleukin-6 (fg/ml)c
| 453.71 ± 433.60 | 1010.53 ± 683.35 | 556.82 ± 776.23 | 1945.58 ± 2597.66 | 2556.97 ± 4185.09 | 611.39 ± 2410.06 | −54.56 ± 2750.40 | 0.227 | 0.03 |
Total cholesterol (mmol/L) a
| 4.36 ± 1.08 | 4.34 ± 1.08 | −0.01 ± 0.23 | 4.40 ± 0.82 | 4.20 ± 0.84 | −0.20 ± 0.29 | −0.19 ± 0.27 | 0.283 | −0.71 |
Total cholesterol/HDL ratio (mmol/L)a
| 3.14 ± 1.19 | 3.32 ± 1.29 | 0.18 ± 0.23 | 3.00 ± 0.90 | 3.44 ± 1.04 | 0.44 ± 0.68 | 0.25 ± 0.53 | 0.251 | 0.51 |
HDL-cholesterol (mmol/L) a
| 1.46 ± 0.41 | 1.39 ± 0.39 | −0.07 ± 0.09 | 1.52 ± 0.36 | 1.22 ± 0.44 | −0.30 ± 0.50 | −0.23 ± 0.50 | 0.117 | −0.66 |
LDL-cholesterol (mmol/L) a
| 2.46 ± 1.07 | 2.34 ± 1.13 | −0.12 ± 0.19 | 2.52 ± 0.83 | 2.44 ± 0.95 | −0.08 ± 0.47 | 0.04 ± 0.43 | 0.065 | 0.12 |
Non-HDLC (mmol/L) a
| 2.82 ± 1.08 | 2.90 ± 1.07 | 0.08 ± 0.18 | 2.86 ± 0.85 | 2.95 ± 0.91 | 0.09 ± 0.39 | −0.01 ± 0.39 | 0.901 | 0.03 |
Triglycerides (mmol/L) a
| 0.77 ± 0.33 | 1.21 ± 0.73 | 0.44 ± 0.45 | 0.72 ± 0.21 | 1.12 ± 0.49 | 0.40 ± 0.45 | 0.05 ± 0.19 | 0.718 | 0.09 |
Fasting glucose (mmol/L) a
| 5.23 ± 0.39 | 4.92 ± 0.51 | −0.31 ± 0.32 | 5.22 ± 0.48 | 4.95 ± 0.59 | −0.27 ± 0.58 | 0.04 ± 0.77 | 0.813 | −0.09 |
S-insulin (mU/L) a
| 14.22 ± 12.70 | 28.90 ± 33.78 | 14.68 ± 21.74 | 10.90 ± 3.24 | 24.92 ± 16.78 | 14.02 ± 16.47 | 0.65 ± 12.25 | 0.913 | 0.03 |
Systolic BP (mmHg)* | 115.15 ± 13.76 | 109.78 ± 11.13 | −5.39 ± 10.826 | 117 ± 13.75 | 112.72 ± 13.57 | −4.28 ± 15.56 | 1.11 ± 13.40 | 0.801 | 0.08 |
Diastolic BP (mmHg)* | 69.67 ± 7.57 | 65.61 ± 8.29 | −4.06 ± 2.6 | 72.28 ± 10.94 | 70.78 ± 10.27 | −1.50 ± 13.64 | 2.56 ± 12.40 | 0.490 | 0.26 |
A comparison of demographic data for participants who did not have complete sets of blood data compared to those that had complete sets of blood data (Table
1) found there was no significant difference in the scores for age (
M ± SD) = 13.4 ± 1.0vs (
M ± SD) = 13.7 ± 0.8,
t (18) = 0.16,
p = 0.25, BMI (
M ± SD) = 20.8 ± 3.8 vs (
M ± SD) = 22.3
± 4.7, t (18) = 0.68,
p = 0.9, and body fat percentage (
M ± SD) = 21.5
± 5.6) vs (
M ± SD) = 28.1
± 12.1,
t (18) = 0.71,
p = 0.18.
Cognitive function
Differences in mental attention capacity for pre and post typical and reduced-sitting school day are shown in Table
2. Mental attention capacity declined slightly during the typical school day, but increased in the reduced sitting day. The pre to post difference between the two conditions, while not statistically significant, nonetheless had a medium effect size (
d = 0.54).
The cardiometabolic health outcomes for the total sample pre and post typical and reduced sitting school day are also shown in Table
2. The pre to post difference between the two conditions was statistically significant for apoB/apoA-1 ratio and there was a medium effect size in the hypothesized direction for this outcome (
d = −0.67) and for total cholesterol (
d = −0.71). There were also medium effect sizes in the non-hypothesized direction for HDL cholesterol; and total cholesterol/HDL ratio. The differences between the reduced and typical school day for glucose, insulin, blood pressure, body fat and BMI were small and not statistically significant (see Table
2).
Monitoring of energy intake and expenditure
Participant mean energy intake whilst undergoing both trial protocols were 5337.68 ± 1778.83 kJ, mean carbohydrates were 186.77 ± 52.88 g and mean total fat 38.14 ± 18.85 g.
During the 48 h period following the protocol, there was no difference in the total energy expenditure (using Sensewear devices) between the ‘typical’ (M = 10,134.8 kJ ± 48.7 kJ) and ‘reduced sitting’ school day (M = 11,276.4 kJ ± 3458.5 kJ) in the 48 h after each condition (t [
14] = −1.06,
P = 0.31). There was no difference in energy intake (kilojoules) (using food diaries) between the ‘typical’ (M = 15,037.9 kJ ± 7021.5 kJ) and ‘reduced sitting’ school day (M = 14,942.6 kJ ± 3820.0 kJ) in the 48 h after each condition (t [
10] = −0.06,
P = 0.95), indicating that there was no compensation.
Discussion
This is the first study to examine the difference between adolescent cardiometabolic and cognitive outcomes during a simulated ‘typical’ and ‘reduced’ sitting school day. The only cardio-metabolic outcome exhibiting a significant improvement in pre to post change was apoB/apoA-1 ratio. Several outcomes demonstrated favourable improvements with medium effect sizes for: total cholesterol, HDL cholesterol, total cholesterol/HDL cholesterol ratio and the apoB/apoA-1 ratio. There were no differences for glucose, insulin, blood pressure, body fat or BMI. A medium effect size indicating improvements in outcomes were also observed for cognitive function.
Previous experimental studies investigating the impact of reducing sedentary behavior on cardiometabolic outcomes in adolescents have produced mixed results [
6,
43,
46]. Saunders et al. [
46] there were no significant differences in participants insulin, glucose, triglycerides, HDL and LDL cholesterol AUC. Ross et al. [
43] only assessed post prandial triglycerides with no significant differences in area under the concentration-time curve, however the findings indicated for 8 of the 12 participants triglyceride concentrations remained high during the day spent in 6 h of sitting. Belcher et al. [
6] found significantly lower insulin AUC, C-peptide AUC, glucose AUC and free fatty acid concentrations as a result interrupting 3 h of sitting every 30 min with MVPA. The results of the current study may have differed from previous studies for several reasons. First, previous studies used a different dose of physical activity to break up sitting time [
6,
43,
46]. For example Saunders et al. trial included three conditions: 8 unbroken hours of sitting, 8 h of sitting broken every 20 min with 2 min of light walking and finally 8 h of sitting broken every 20 min with 2 min of light walking as well as 2 × 20 min periods of MVPA [
46]. Belcher et al. interrupted 3 h of sitting with 3 min of MVPA every 30 min [
6], while Ross et al. broke up 6 h of sitting every 30 min with 4 min of moderate physical activity [
43]. Differences in the dose and type of physical activity and duration of sedentary time may impact lipoprotein lipase (LPL) activity influencing triglyceride and glucose uptake to differing extents, as reduced stimulation of weight bearing muscles results in reduced uptake of these markers [
23]. Second the current study measured cardio-metabolic biomarkers at the beginning and end of the simulated school day, whereas previous studies [
6,
43,
46] measured postprandial responses at regular intervals across the experimental period. In addition, the post-test blood measures in the current study were not true fasting samples due to the duration between the midday meal and the final blood draw. Whilst evidence indicates that a fasting duration of 3 h is sufficient to achieve comparable blood glucose levels between fasting and non-fasting states in adults [
36], the study design limited the opportunity for an equivalent fasting state for the post-intervention phase. In the study by Ross et al. [
43], participants consumed a high fat diet prior to the experimental protocol whilst in the other studies participants consumed a balanced diet. Ross et al. [
43], blood measures were taken using finger puncture whilst the other studies used venous blood. When comparing the findings mean values for adolescent cholesterol are higher using finger prick as opposed to venous blood [
4].
There is a growing body of evidence indicating that apoB/apoA-1 is a powerful biomarker of future cardiovascular disease as it impacts lipid metabolism, and is a marker of inflammation possessing both anti-oxidant and anti-inflammatory effects. In the current study, differences between the conditions resulted in a medium effect size (
d = 0.54) which were statistically significant for the apoB/apoA-1 ratio [
53,
55]. Although to date there have been no experimental studies investigating the link between interruptions to sedentary time and apoB/apoA-1 ratios, a study of 18 healthy 19 to 23 year olds demonstrated statistically significant reductions in apoB/apoA-1 ratios after replacing sitting time with walking or standing [
18]. As suggested by Hamilton et al. (2007), non-exercise activity thermogenesis (NEAT) is a large part of total energy expenditure. Muscular contractions that occur when standing or in very light physical activity may inhibit unhealthy molecular signals contributing to metabolic diseases [
24].
Keeping in mind the findings in the current study reflect the transition from a fasting to a postprandial state, the reduced sitting day condition had a medium effect on total cholesterol (d = −0.71), HDL cholesterol (
d = −0.66) and total cholesterol/HDL ratio (
d = 0.51). Studies show that high total blood cholesterol increases the risk of coronary heart disease and some types of stroke [
38]. Early identification and control of high cholesterol in youth has been found to reduce the risk of cardiovascular disease into adulthood [
21]. While there are no other experimental adolescent studies reporting an effect for reduced sitting time on total cholesterol, a 9-month prospective uncontrolled trial of overweight and obese office workers observed significant reductions in total cholesterol following the introduction of treadmill work stations [
27].
HDL cholesterol has a positive effect on health as it transports lipids away from the heart back to the liver [
31]. Prior to the commencement of the study we hypothesised that the intervention would result in improvements in HDL cholesterol [
33]. Medium effects sizes for HDL cholesterol (d = −0.66) and total cholesterol/HDL ratio (d = 0.51) in the current study were not in the anticipated direction, however, they are consistent with previous evidence. Two recent reviews of adult studies designed to assess the effects of exercise interventions on cholesterol levels and lipid profiles suggest increases in aerobic exercise (through increases in intensity or duration) and calorific expenditure resulted in improved HDL cholesterol levels [
31,
51]. As HDL seems to be related to MVPA and the current study aimed to increase LPA this may explain the lack of effect on HDL cholesterol in this instance. Another potential explanation is that postprandial changes in HDL cholesterol are often inversely associated with changes in triglyceride concentrations [
14] (Table
2). In addition, a systematic review of adult interventions designed to clarify the role of LPA on cardiovascular risk factors and markers in adults determined that all studies except one found no significant changes in HDL cholesterol [
5]. Although many of the studies were of low or fair quality, recognising a need for further studies.
The total cholesterol/HDL ratio is a good predictor of cardiovascular health [
35]. Lower ratios of total cholesterol/HDL indicate lower risk of heart disease, thus research often aims to reduce this risk factor [
26,
39]. The findings in the current study, however, were in the non-hypothesised direction. As previously explained, given that total cholesterol decreased and HDL cholesterol also decreased it is not surprising that the total cholesterol/HDL cholesterol ratio would increase.
Whilst evidence indicates that physical activity is positively associated with academic performance among adolescents [
20], there is a dearth of research investigating the association between sedentary behaviour and cognitive functioning in adolescents [
11]. Recent findings indicate that reducing sedentary time produces positive academic outcomes in adolescents [
15] and classrooms with stand-biased desks may have beneficial effects on academic engagement [
16]. Medium effect sizes for cognition in the current study indicate that mental attention capacity (a causal component underlying the development of working memory) declined after the typical school day and increased after the reduced sitting day. The size of the difference between the typical and reduced sitting day for mental-attentional capacity has important implications for reductions in sitting through the school setting as it is equivalent to 6 months’ improvement in effective mental-attentional capacity [
41]. Rather than sustained, structural changes in cognitive function (i.e., changes in the underlying capacity constraints of working memory), these changes are likely functional (e.g., temporary increases in the effective capacity of working memory, due to increased focus, sustained attention, less distraction from task-irrelevant aspects of a situation). Functional changes in cognition are nevertheless interesting and important for development and academic reasons. For instance, research on Cognitive Load Theory has demonstrated the learning advantages of optimising the focus and deployment of working memory (i.e., increasing attentional capacity focused on learning-relevant information and processes) [
29]. As such, even temporary improvements in functional working memory capacity can yield academic and developmental benefits in those situations in which it is optimised. This suggests the potential for acute enhancements to cognitive function (but perhaps also sustained benefits with a longer intervention) as a result of a reduced sitting day in the school environment. It also suggests the possibility of broader academic benefits, given research establishing a link between mental-attentional (working memory) capacity and non-verbal/fluid intelligence and academic ability (e.g., math, reading) [
25]. It is therefore possible that simple adjustments to the school classroom and outdoor environment can have acute (and possibly longer-term) effects on cognitive function in young people.
Reducing sitting time by approximately 50% during the school day in the current study did not result in a compensatory reduction in energy expenditure and increase in energy intake among the participants in the 48-h period post intervention. These findings are consistent with another within person comparisons investigating compensatory reductions in energy expenditure in adolescents. Saunders and colleagues [
45] exposed children and adolescents to three different experimental conditions involving prolonged and interrupted breaks in sedentary time and found that the participants did not compensate in the 24 h post experiment by increasing food intake or physical activity levels when sedentary time was reduced or interrupted. While these findings support the notion that compensation does not occur over the short term, longer experimental trials are needed to further investigate these findings.
The current study has some strengths and limitations which should be acknowledged. The study findings may be limited by its sample size, potentially affecting the opportunity to detect significant associations for some outcomes. Secondly, the study design did not incorporate multiple blood draws potentially improving the presentation of blood outcomes and comparisons with other studies. Further, data indicating participant maturation was not collected; pubertal maturation has biochemical and physiological effects during adolescents. Finally, a large number of statistical tests were performed without adjustment for multiple comparisons; the results should be interpreted in this context. There were several strengths in this study. The design of this study is a strength in terms of ecological validity [
46] as is the first randomized controlled trial investigating the acute effects of reducing adolescent sitting in a setting which replicates the school environment. Further, it considered the impact of sitting time on adolescent cognition, not previously reported. Finally it assessed novel cardiometabolic markers (apoB/apoA-1 ratio and IL-6), not previously considered in this context.