To compare the cumulative (3-day) effect of prolonged sitting on metabolic responses during a mixed meal tolerance test (MTT), with sitting that is regularly interrupted with brief bouts of light-intensity walking. Overweight/obese adults (n=19) were recruited for a randomized, 3-day, outpatient, cross-over trial involving: (1) 7-h days of uninterrupted sitting (SIT); and (2) 7-h days of sitting with light-intensity activity breaks [BREAKS; 2-min of treadmill walking (3.2 km/h) every 20 min (total: 17 breaks/day)]. On days 1 and 3, participants underwent a MTT (75 g of carbohydrate, 50 g of fat) and the incremental area under the curve (iAUC) was calculated from hourly blood samples. Generalized estimating equation (GEE) models were adjusted for gender, body mass index (BMI), energy intake, treatment order and pre-prandial values to determine effects of time, condition and time × condition. The glucose iAUC was 1.3±0.5 and 1.5±0.5 mmol·h·l−1 (mean differences ± S.E.M.) higher in SIT compared with BREAKS on days 1 and 3 respectively (condition effect: P=0.001), with no effect of time (P=0.48) or time × condition (P=0.8). The insulin iAUC was also higher on both days in SIT (day 1: ∆151±73, day 3: ∆91±73 pmol·h·l−1, P=0.01), with no effect of time (P=0.52) or time × condition (P=0.71). There was no between-treatment difference in triglycerides (triacylglycerols) iAUC. There were significant between-condition effects but no temporal change in metabolic responses to MTT, indicating that breaking up of sitting over 3 days sustains, but does not enhance, the lowering of postprandial glucose and insulin.

CLINICAL PERSPECTIVES

  • Recent experimental studies have demonstrated that regularly breaking up a single bout of prolonged sitting, even with short-duration light-intensity physical activity breaks, acutely lowers postprandial hyperglycaemia [6,7]. However, this effect has not been tested over multiple days in experimental studies.

  • We reports novel findings from a randomized, controlled laboratory trial in which regularly breaking up prolonged sitting over 3 days sustains, but does not enhance, the lowering of postprandial glucose and insulin.

  • Given that postprandial hyperglycaemia (together with related hyperinsulinaemia) have been implicated in the development of chronic metabolic diseases such as Type 2 diabetes and cardiovascular disease, breaking up sitting time with short bouts of activity might be a long-term potential strategy to help reduce disease risk, particularly for at-risk populations (e.g. the obese, the elderly) or when more intense forms of physical activity cannot be tolerated.

INTRODUCTION

For most adults in developed countries, a substantial portion of the waking day is spent sedentary (sitting, reclining or lying down with low energy expenditure) [1]. The adverse consequences of high durations spent sedentary are potentially manifold, as observational studies indicate that sitting time is adversely associated with cardio-metabolic risk biomarkers, Type 2 diabetes, cardiovascular disease, some cancers and premature mortality [2,3]. Furthermore, these associations generally remain after accounting for total or leisure-time moderate-to-vigorous physical activity (MVPA), suggesting that sedentary time can have detrimental health effects independent of MVPA.

In addition to total sitting time, the manner in which sedentary time is accumulated may also influence disease risk. Observational studies indicate that frequent breaks in sitting time (transitions between sitting to standing/movement) are associated with a more favourable cardio-metabolic risk profile than an equivalent amount of sitting time accumulated in longer uninterrupted bouts [4,5]. This is further supported by clinical trial evidence showing that regularly interrupting prolonged sitting with brief (∼2-min) bouts of walking acutely lowered postprandial plasma glucose and insulin concentrations when compared with uninterrupted sitting [6,7]. Interrupting sitting with light-intensity walking bouts was as effective in reducing postprandial responses as were breaks of moderate-intensity walking [7]. Furthermore, these regular short activity breaks from sitting time were more successful in decreasing postprandial responses, compared with an equivalent duration of sitting preceded by a single bout of continuous physical activity [6]. Although the underlying mechanisms were not assessed in these studies, it is well known that muscle contractions can acutely stimulate glucose uptake via insulin-independent cellular mechanisms [8]. Therefore, it is possible that interrupting sitting time facilitates glucose removal from the circulation through intermittent muscle contractions.

It remains to be experimentally determined whether breaking up sitting time has cumulative effects when performed over a number of days. Previous studies involving varying durations (3–56 days) of bed rest, an extreme model of imposed physical inactivity, have shown substantial deterioration in insulin sensitivity and associated elevations in post meal glucose and insulin concentrations, with metabolic derangements observed as early as after 3 days [911]. To our knowledge, there is only one experimental study reporting on the effects of longer-term exposure (>1 day) to increases in sedentary behaviour [12]. In the free-living environment, Lyden et al. [12] showed that increased sitting time and avoidance of exercise for 7 days were associated with increased postprandial insulin concentrations, when compared with an equivalent duration involving usual activity. The present study also found that elevated insulin levels were associated with decreases in light-intensity activity and longer bouts of uninterrupted sitting time (time >30 min and >60 min). However, the effect of interrupting sitting with frequent light-intensity walking bouts over several days is yet to be investigated experimentally. We hypothesized that consecutive days of prolonged sitting would impair glycaemic control and that breaking up sitting with short bouts of activity could potentially mitigate these adverse effects.

In an outpatient laboratory-based study involving overweight/obese sedentary adults, we compared the cumulative (3-day) effects of uninterrupted sitting on postprandial glycaemic control to a mixed meal tolerance test (MTT), with sitting interrupted by brief (2-min) bouts of light-intensity walking.

MATERIALS AND METHODS

Study overview

This randomized cross-over trial was undertaken at the Baker IDI Heart and Diabetes Institute between December 2010 and May 2012. Participants completed, in random order, three consecutive days of prolonged sitting (SIT) and three consecutive days of sitting with intermittent bouts of light-intensity walking (BREAKS), separated by a minimum 12-day washout period (Figure 1). To eliminate potential bias, condition order was randomly assigned by a third party using computer-generated random numbers (block randomization and balanced block sizes), stratified by gender. Study personnel were blinded to the condition order until the initial informed consent appointment.

Protocol diagram

Figure 1
Protocol diagram

Participants completed, in random order, three consecutive days of the prolonged sitting protocol (shown below) and three consecutive days of sitting plus intermittent walking breaks, separated by a minimum 12-day washout. On assessment days, participants underwent a 4-h MTT following a 1-h steady state period. Hourly bloods were collected 2 min prior to walking bouts scheduled on the hour. Participants received a snack (open circle) at the completion of the MTT.

Figure 1
Protocol diagram

Participants completed, in random order, three consecutive days of the prolonged sitting protocol (shown below) and three consecutive days of sitting plus intermittent walking breaks, separated by a minimum 12-day washout. On assessment days, participants underwent a 4-h MTT following a 1-h steady state period. Hourly bloods were collected 2 min prior to walking bouts scheduled on the hour. Participants received a snack (open circle) at the completion of the MTT.

Days 1 and 3 of each experimental condition were designated ‘assessment days’ and involved measurements of plasma glucose, insulin and triglyceride (triacylglycerol) concentrations during a MTT. Primary outcomes were changes in area under the curve (AUC) for plasma glucose and insulin, with triglyceride AUC being a secondary end-point. Based on estimates of population variability (mean post-intervention to baseline glucose AUC ratio 1.1, S.D.: 0.01; mean insulin AUC ratio 1.6, S.D.: 0.3) from our own laboratory and a conservative correlation of 0.50 for repeated measures, it was estimated that 15 participants would provide 80% power (two-sided, 5% level) to detect the smallest clinically meaningful change in the outcome variables. To compensate for potential withdrawals, 24 participants were recruited.

Participants

Sedentary overweight/obese [body mass index (BMI)=25–45 kg/m2)] adults (aged 45–75 years) were recruited from local community advertisements. Participant characteristics are presented in Table 1. All were sedentary (self-reported sitting time > 5 h/day), non-diabetic, non-smoking, not taking glucose- or lipid-lowering medications or on anti-coagulant therapy and not meeting current physical activity guidelines [13]. Of the women who participated, six were post-menopausal and two were peri-menopausal. For peri-menopausal women, the menstrual cycle phase could not be determined during the experimental conditions. Written informed consent was obtained from each participant following an explanation of the experimental procedures and the risks involved. The study was carried out in accordance with the principles of the Declaration of Helsinki (2008) and was approved by the Alfred Human Ethics Committee.

Table 1
Participant characteristics

Data are expressed as means (S.E.M.) or number (%) where specified.

Demographics
Age (years) 56.7 (1.5) 
Male number (%) 11 (58) 
Occupation  
Retired 8 (42%) 
Unemployed 2 (11%) 
Housewife 2 (11%) 
Working part-time 2 (11%) 
Clerical/administrative work 2 (11%) 
Public service/health work 3 (11%) 
Anthropometrics  
BMI (kg/m232.7 (1.0) 
Waist circumference (cm) 108.3 (2.7) 
Demographics
Age (years) 56.7 (1.5) 
Male number (%) 11 (58) 
Occupation  
Retired 8 (42%) 
Unemployed 2 (11%) 
Housewife 2 (11%) 
Working part-time 2 (11%) 
Clerical/administrative work 2 (11%) 
Public service/health work 3 (11%) 
Anthropometrics  
BMI (kg/m232.7 (1.0) 
Waist circumference (cm) 108.3 (2.7) 

Study protocol

Figure 1 shows the study protocol. Participants were asked to refrain from exercise, caffeine or alcohol for 48 h prior to each experimental condition. In the week preceding the experimental conditions, they wore Actigraph GTM1 accelerometers during waking hours for objective measurements of physical activity. The 1-min epoch activity data were processed to derive average sedentary [<100 counts per minute (cpm)], light-intensity activity (100–1951 cpm) and MVPA (1952+ cpm) time on valid days [14]. Weighed/measured food records and Foodworks nutritional software (Xyris) were used to quantify total energy and macronutrient intakes 48 h prior to experimental conditions.

To eliminate any diet-induced variability in study outcomes, dietary intakes were strictly controlled starting from the evening meal prior to day 1 to breakfast on day 3. Meal plans were individualized to meet daily estimated energy requirements (Schofield equation, 1.5 physical activity factor) [15] and a target macronutrient profile of 12–15% energy from protein, 55–58% energy from carbohydrate and 29–31% energy from fat. Meals were standardized across conditions and were carefully prepared by research staff in the laboratory kitchen. To accommodate dietary preferences, participants were able to select from a range of meal options, including bran-based cereal or muesli and toast at breakfast and a meat and salad bread roll at lunch. An evening meal pack, consisting of a commercially available drink, snack and microwave meal, was also provided for participants to prepare at home.

On experimental days, participants were instructed to minimize movement and to drive or be driven to the research laboratory. Participants arrived between 08:00 and 09:00, after a 12-h overnight fast. They then remained seated for the initial 1 h to achieve a steady-state before completing one of the following protocols over the next 6 h:

  • 1) SIT: on all days, participants sat upright in a comfortable chair and had access to a television, DVDs and reading materials (newspapers and magazines). They were required to minimize excessive movement, but were allowed to visit the toilet when necessary. At the completion of the experimental protocol, participants were instructed to return home and limit movements to only activities that were necessary for daily living.

  • 2) BREAKS: the experimental protocol was the same as for SIT, with the exception that participants rose from the seated position every 20 min and completed a 2-min bout of light-intensity walking on a motorized treadmill (level surface) at 3.2 km/h. After each walking bout, they rated their perceived exertion using the standard 6–20 Borg scale to ensure that walking bouts were classified as being light-intensity (acceptable range 6–11).

Meal tolerance test

On assessment days, an indwelling catheter was inserted into an antecubital vein for the hourly collection of blood. Following the steady-state period, participants completed a 4-h MTT. The mixed meal was provided as a 200 ml of strawberry-flavoured drink, containing 75 g carbohydrate (100% corn maltodextrin) and 50 g fat (Calogen, Nutricia). The MTT was used for the simultaneous measurement of postprandial glucose, insulin and triglyceride responses [16] and as MTTs show strong agreement with oral glucose tolerance test-derived model-based estimates of insulin sensitivity [17,18]. Blood samples were collected each hour immediately prior to the scheduled walking bout.

Assessment of metabolic outcomes and model-based estimates of insulin sensitivity

Plasma glucose, insulin and triglyceride concentrations were analysed with a commercial analytical system (Architect ci16200, Abbott Diagnostics). Glucose and triglycerides were measured on the day of testing using hexokinase [intra-assay coefficient of variation (CV)=1.1%] and glycerol phosphate oxidase methods (intra-assay CV=1.3%) respectively. Insulin samples were collected and stored at −80°C for later testing using a chemiluminescent microparticle immunoassay (intra-assay CV=2.9–4.3%).

Metabolic outcomes from the MTT were assessed using net incremental AUC (iAUC) and total AUC (tAUC) using the trapezoidal method, where net iAUC is calculated from basal levels and tAUC is taken from a plasma concentration of zero. The HOMA-IR (homeostasis model assessment of insulin resistance) index was calculated using fasting plasma glucose and insulin concentrations {[glucose (mmol/l) × insulin (pmol/ml)]/135} [19]. Basal insulin secretion was assessed by HOMA-β% (homeostasis model assessment of β-cell function), which was calculated as [fasting plasma insulin (pmol/l) × 3.33]/[fasting plasma glucose (mmol/l) − 3.5]. Skeletal muscle insulin sensitivity index (MISI) was assessed from postprandial changes in glucose and insulin concentrations during the MTT. According to Abdul-Ghani et al. [20], skeletal muscle insulin sensitivity is reflected by the rate of decline (dG/dt) in postprandial glucose concentrations (mg/l) divided by the mean plasma insulin concentration (microunits/ml), where dG/dt was calculated as the slope of least-squares fit from peak glucose concentration to nadir. This model correlates strongly with muscle insulin sensitivity determined from a hyperinsulinaemic euglycaemic clamp [20].

Statistical analysis

This study was analysed on a per-protocol basis. Randomized participants who did not complete the first experimental condition were excluded from the analysis (Figure 2). Reasons for exclusion were: unable to comply with experimental protocol (MTT or treadmill walking, n=3), participant withdrawal (n=1) and diagnosis of type 2 diabetes post-screening (n=1), an exclusion criterion.

Participant flow diagram

Figure 2
Participant flow diagram
Figure 2
Participant flow diagram

Generalized estimating equations (GEEs) with exchangeable working correlation were used to evaluate the differential effects of time, experimental condition and the differences in time course between groups (time × condition interaction) using Stata statistical analysis software (Version 12, StataCorp LP). All models were adjusted for potential covariates explaining residual outcome variance (age, gender and BMI), pre-prandial values (iAUC only) and period effects (treatment order). Due to between-condition discrepancies at baseline, models were also adjusted for the experimental change in energy intake and time spent doing light-intensity physical activity.

RESULTS

Potential confounding diet and physical activity variables

According to weighed/measured food records, energy intakes before both experimental conditions were lower than energy intakes during the controlled-feeding period (Table 2). However, a lower energy intake during the pre-experimental period for BREAKS resulted in a greater experimental increase when compared with the SIT condition [+2682±602 kJ (mean ± S.E.M.) compared with +1800±592 kJ, P=0.02]. Other dietary changes induced by controlled feeding were increases in carbohydrate intake (P<0.001 for both conditions) and decreases in protein (P<0.001 for both conditions) and fat (P<0.05 for both conditions) intakes.

Table 2
Dietary and activity variables during the pre-experimental period and the change observed with implementation of experimental protocols

Data are expressed as means (S.E.M.)

SITBREAKS
Dietary intakes   
Energy (kJ/day)   
Pre-experimental period* 9360 (556) 8478 (561) § 
Experimental change +1800 (592) +2682 (602) § 
Carbohydrate (% of total kJ)   
Pre-experimental period 49.1 (1.8) 51.6 (1.6) 
Experimental change +8.3 (1.8) +5.7 (1.6) 
Protein (% of total kJ)   
Pre-experimental period 17.0 (0.6) 17.1(0.6) 
Experimental change −3.3 (0.6) −3.3 (0.6) 
Fat (% of total kJ)   
Pre-experimental period 34.3 (1.6) 31.6 (1.3) 
Experimental change −4.4 (1.8) −1.8 (1.5) 
Accelerometer data   
Daily wear time (min)   
Pre-experimental period 821 (22) 797 (25) 
Experimental change −101 (26) −54 (22) 
Sedentary (min/day)   
Pre-experimental period 538 (14) 540 (18) 
Experimental change +36 (19) +17 (16) 
Light-intensity (min/day)   
Pre-experimental period 254 (16) 229 (17)§ 
Experimental change −124 (18) −59 (17) 
Moderate-vigorous (min/day)   
Pre-experimental period 29 (6) 28 (7) 
Experimental change −13 (4) −12 (5) 
SITBREAKS
Dietary intakes   
Energy (kJ/day)   
Pre-experimental period* 9360 (556) 8478 (561) § 
Experimental change +1800 (592) +2682 (602) § 
Carbohydrate (% of total kJ)   
Pre-experimental period 49.1 (1.8) 51.6 (1.6) 
Experimental change +8.3 (1.8) +5.7 (1.6) 
Protein (% of total kJ)   
Pre-experimental period 17.0 (0.6) 17.1(0.6) 
Experimental change −3.3 (0.6) −3.3 (0.6) 
Fat (% of total kJ)   
Pre-experimental period 34.3 (1.6) 31.6 (1.3) 
Experimental change −4.4 (1.8) −1.8 (1.5) 
Accelerometer data   
Daily wear time (min)   
Pre-experimental period 821 (22) 797 (25) 
Experimental change −101 (26) −54 (22) 
Sedentary (min/day)   
Pre-experimental period 538 (14) 540 (18) 
Experimental change +36 (19) +17 (16) 
Light-intensity (min/day)   
Pre-experimental period 254 (16) 229 (17)§ 
Experimental change −124 (18) −59 (17) 
Moderate-vigorous (min/day)   
Pre-experimental period 29 (6) 28 (7) 
Experimental change −13 (4) −12 (5) 

*Dietary intakes were assessed from weighed/measured food records during the 48-h period before the trial

†Defined as the change from the pre-experimental period

‡Data from the week preceding the experimental condition. No differences were observed during the habitual and restricted periods.

Paired Students t test revealed significant between-condition differences: §P<0.05, ║P<0.01.

In the week preceding experimental conditions, monitor-wear time, sedentary time and time spent in MVPA were not different between conditions. However, time spent in light-intensity activities was greater for SIT than BREAKS, indicating a greater experimental decrease in light-intensity physical activity in SIT when compared with BREAKS (−124±18 min compared with −59±17 min, P<0.001). From habitual levels, participants decreased the time spent in light-intensity physical activity (P<0.01 for both conditions) and MVPA (P<0.05 for both conditions), whereas sedentary time was not different.

Postprandial responses to the MTT

Mean fasting glucose levels were similar for both conditions on the respective assessment days, with both conditions demonstrating a small decrease over time (P<0.001). After adjusting for potential confounders, the iAUC for glucose, which calculates postprandial glycaemic response from fasting levels, was significantly lower during BREAKS (P=0.001; Figures 3a and 3b), with iAUC being 32% and 31% lower than the prolonged sitting condition on days 1 and 3 respectively. Similarly, tAUC for glucose, which reflects the total exposure to plasma glucose concentrations, was significantly lower in BREAKS than SIT (P=0.005; Table 3). Plasma glucose concentrations at 2 h and dG/dt were also significantly lower in BREAKS than SIT (P=0.012 and P=0.037 respectively; Table 3). For all measures of postprandial glycaemia, there was no effect of time or between-condition differences in the time course (time × condition effect) across the 3-day protocol.

Postprandial responses to the MTT

Figure 3
Postprandial responses to the MTT

(a) Unadjusted postprandial glucose responses to the MTT on day 1 and 3 of the experimental protocols. Circles=BREAKS; squares=SIT; closed marker + unbroken line=day 1; open marker + broken line=day 3. (b) Estimated marginal means for glucose iAUC (adjusted for age, sex, BMI, treatment order, fasting glucose, pre-condition light-intensity physical activity time and the experimental change in energy intake) and error bars are S.E.M. Circles=BREAKS; squares=SIT. (c) Unadjusted postprandial insulin responses to the MTT on day 1 and 3 of experimental protocols. Circles=BREAKS; squares=SIT; closed marker + unbroken line=day 1; open marker + broken line=day 3. (d) Estimated marginal means for insulin iAUC (adjusted for age, sex, BMI, treatment order, fasting insulin, pre-condition light-intensity physical activity time and the experimental change in energy intake) and error bars are S.E.M. Circles=BREAKS; squares=SIT. (e) Unadjusted postprandial triglyceride responses to the MTT on day 1 and 3 of experimental protocols. Circles=BREAKS; squares=SIT; closed marker + unbroken line=day 1; open marker + broken line=day 3. (f) Estimated marginal means for triglyceride iAUC (adjusted for age, sex, BMI, treatment order, fasting triglycerides, pre-condition light-intensity physical activity time and the experimental change in energy intake) and error bars S.E.M. Circles=BREAKS; squares=SIT.

Figure 3
Postprandial responses to the MTT

(a) Unadjusted postprandial glucose responses to the MTT on day 1 and 3 of the experimental protocols. Circles=BREAKS; squares=SIT; closed marker + unbroken line=day 1; open marker + broken line=day 3. (b) Estimated marginal means for glucose iAUC (adjusted for age, sex, BMI, treatment order, fasting glucose, pre-condition light-intensity physical activity time and the experimental change in energy intake) and error bars are S.E.M. Circles=BREAKS; squares=SIT. (c) Unadjusted postprandial insulin responses to the MTT on day 1 and 3 of experimental protocols. Circles=BREAKS; squares=SIT; closed marker + unbroken line=day 1; open marker + broken line=day 3. (d) Estimated marginal means for insulin iAUC (adjusted for age, sex, BMI, treatment order, fasting insulin, pre-condition light-intensity physical activity time and the experimental change in energy intake) and error bars are S.E.M. Circles=BREAKS; squares=SIT. (e) Unadjusted postprandial triglyceride responses to the MTT on day 1 and 3 of experimental protocols. Circles=BREAKS; squares=SIT; closed marker + unbroken line=day 1; open marker + broken line=day 3. (f) Estimated marginal means for triglyceride iAUC (adjusted for age, sex, BMI, treatment order, fasting triglycerides, pre-condition light-intensity physical activity time and the experimental change in energy intake) and error bars S.E.M. Circles=BREAKS; squares=SIT.

Table 3
Basal and postprandial responses to the MTT and model-based estimates of insulin sensitivity and β-cell function

GEE models were adjusted for age, sex, BMI, treatment order, pre-condition light-intensity physical activity time and the experimental change in energy intake. Data are expressed as estimated marginal means (S.E.M.). Abbreviation: AUCins/glu, ratio of the tAUCs for insulin and glucose.

SITBREAKSP
Day 1Day 3Day 1Day 3ConditionTimeCondition × Time
        
Weight (kg) 94.6 (14.5) 94.6 (14.4) 94.5 (14.4) 94.5 (2.9) 0.922 0.971 1.000 
Fasting glucose (mmol/l) 5.1 (0.5) 5.0 (0.4) 5.2 (0.4) 5.0 (0.4) 0.955 <0.001 0.425 
2-h glucose (mmol/l) 8.1 (2.8) 7.9 (3.0) 7.1 (2.6) 7.2 (2.6) 0.012 0.849 0.739 
dG/dt (mg·l−1·h−183.4 (8.1) 85.0 (8.1) 68.7 (8.1) 70.2 (8.1) 0.037 0.808 0.150 
Glucose tAUC (mmol·h·l−130.4 (1.2) 29.8 (1.2) 29.0 (1.2) 28.5 (1.2) 0.005 0.148 0.793 
Fasting insulin (pmol/l) 71.4 (5.4) 75.2 (5.7) 67.9 (5.1) 71.4 (5.4) 0.326 0.246 0.740 
Insulin tAUC (pmol·h·l−11474 (162) 1484 (163) 1301 (143) 1309 (144) 0.009 0.859 0.879 
AUCins/glu 48.8 (4.8) 50.0 (4.9) 44.3 (4.4) 45.3 (4.5) 0.037 0.526 0.983 
Fasting triglycerides (mmol/l) 1.5 (0.1) 1.8 (0.1) 1.5 (0.1) 1.8 (0.1) 0.828 <0.001 0.384 
Triglyceride tAUC (mmol·h·l−18.2 (0.7) 10.2 (0.8) 8.1 (0.7) 10.1 (0.8) 0.786 <0.001 0.522 
MISI 2.41 (0.32) 2.40 (0.32) 2.17 (0.32) 2.16 (0.32) 0.276 0.969 0.353 
HOMA-IR 2.73 (0.21) 2.77 (0.22) 2.61 (0.21) 2.65 (0.21) 0.404 0.744 0.691 
HOMA-β% 150 (12) 179 (14) 138 (11) 166 (13) 0.121 <0.001 0.955 
SITBREAKSP
Day 1Day 3Day 1Day 3ConditionTimeCondition × Time
        
Weight (kg) 94.6 (14.5) 94.6 (14.4) 94.5 (14.4) 94.5 (2.9) 0.922 0.971 1.000 
Fasting glucose (mmol/l) 5.1 (0.5) 5.0 (0.4) 5.2 (0.4) 5.0 (0.4) 0.955 <0.001 0.425 
2-h glucose (mmol/l) 8.1 (2.8) 7.9 (3.0) 7.1 (2.6) 7.2 (2.6) 0.012 0.849 0.739 
dG/dt (mg·l−1·h−183.4 (8.1) 85.0 (8.1) 68.7 (8.1) 70.2 (8.1) 0.037 0.808 0.150 
Glucose tAUC (mmol·h·l−130.4 (1.2) 29.8 (1.2) 29.0 (1.2) 28.5 (1.2) 0.005 0.148 0.793 
Fasting insulin (pmol/l) 71.4 (5.4) 75.2 (5.7) 67.9 (5.1) 71.4 (5.4) 0.326 0.246 0.740 
Insulin tAUC (pmol·h·l−11474 (162) 1484 (163) 1301 (143) 1309 (144) 0.009 0.859 0.879 
AUCins/glu 48.8 (4.8) 50.0 (4.9) 44.3 (4.4) 45.3 (4.5) 0.037 0.526 0.983 
Fasting triglycerides (mmol/l) 1.5 (0.1) 1.8 (0.1) 1.5 (0.1) 1.8 (0.1) 0.828 <0.001 0.384 
Triglyceride tAUC (mmol·h·l−18.2 (0.7) 10.2 (0.8) 8.1 (0.7) 10.1 (0.8) 0.786 <0.001 0.522 
MISI 2.41 (0.32) 2.40 (0.32) 2.17 (0.32) 2.16 (0.32) 0.276 0.969 0.353 
HOMA-IR 2.73 (0.21) 2.77 (0.22) 2.61 (0.21) 2.65 (0.21) 0.404 0.744 0.691 
HOMA-β% 150 (12) 179 (14) 138 (11) 166 (13) 0.121 <0.001 0.955 

For fasting insulin, there was no effect of time or difference between conditions. On days 1 and 3, iAUC for insulin was 15% lower in BREAKS than the SIT condition (P=0.01; Figures 3c and 3d), whereas absolute exposure to plasma insulin (tAUC) was 12% lower on both days in the BREAKS condition than SIT (P=0.009; Table 3). However, there was no effect of time or difference in the temporal change in postprandial insulinaemia. The AUC ratio, which evaluates the overall insulin response to variations in postprandial glycaemia, was significantly lower in BREAKS compared with SIT, suggesting less insulin release with breaks in prolonged sitting.

Fasting triglycerides and postprandial triglyceride responses increased in both conditions across time [P<0.001 for fasting triglycerides, P<0.001 for tAUC (Table 3) and P=0.049 for iAUC (Figures 3e and 3f)], with no significant differences between conditions.

Model-based assessments of β-cell function, hepatic insulin sensitivity and peripheral insulin sensitivity

There was no difference between conditions or in the temporal change in HOMA-IR or MISI. However, there was a significant increase in HOMA-β% over time, but no effect of condition or between-group difference in the time course across the 3-day protocol.

DISCUSSION

Compared with uninterrupted sitting, short bouts of light-intensity walking every 20 min attenuated the glycaemic response to the MTT, as demonstrated by 31–32% reduction in the incremental glucose response on days 1 and 3. The strongest between-condition difference was observed between 1 and 3 h, as plasma glucose concentrations continued to rise after 60 min in the sitting condition and were markedly higher than BREAKS at 2 h. By means of extrapolation based on classifications of hyperglycaemia [21], the magnitude of the reduction at 2 h circumvented plasma glucose concentrations rising to levels consistent with impaired glucose tolerance, a condition that progresses to Type 2 diabetes at a rate of 6% per year [22]. Despite significant between-condition effects, no temporal change in the relative hyperglycaemia was observed across the 3-day experimental period, suggesting that there was no metabolic maladaption to repeated days of prolonged sitting. Although observational studies indicate that more frequent breaks in sedentary time are associated with more favourable cardio-metabolic biomarker profiles [2,3], there are no prospective or clinical trials that provide clear evidence of a possible cause-and-effect relationship. Nevertheless, dietary factors and interventions that lower postprandial glycaemic responses have been associated with significant reductions in diabetes and cardiovascular risk in long-term clinical [23] and prospective observational [24] studies, suggesting that lowering postprandial glycaemia per se may reduce disease risk. Therefore, the relative reduction in postprandial glycaemia seen in our study provides a putative mechanism through which breaking up sitting time via short bouts of activity could reduce the long-term risk of developing diabetes and cardiovascular disease.

Our findings concur with those of previous trials that have shown that increasing physical activity during the postprandial period attenuates the glycaemic response to meals. However, previous trials vary greatly with respect to the timing (relative to meal), frequency (single compared with intermittent), intensity and duration of the physical activity bouts. Nygaard et al. [25] demonstrated that 40 min of slow post-meal walking produced a similar blood-glucose-lowering effect as 30 min of cycling at 70% of maximal heart rate [26] and 45 min of cycling at 57% of maximal oxygen consumption [27], indicating that the blunting effect was not directly proportional to the physical activity intensity level. Presumably, postprandial physical activity duration is another factor likely to affect tissue glucose uptake and utilization. However, Lunde et al. [28] showed that doubling the postprandial physical activity time from 20 to 40 min did not double the reduction in glycaemia (31% compared with 39% reduction in iAUC respectively when compared with sitting). Takaishi et al. [29] showed that a single 6-min bout of stair climbing–descending during the absorptive postprandial phase was sufficient to lower the glycaemic response when compared with sitting. However, when the 6-min bout was implemented before the blood glucose peak, there was a further rise in postprandial glycaemia. In a study similar to our own, Peddie et al. [6] demonstrated that interrupting sitting time with frequent short (<2-min) activity breaks reduced the day-long glycaemic response by 37% when compared with a single 30-min bout of continuous physical activity that preceded prolonged sitting. Importantly, those findings [6] highlight that equivalent amounts of physical activity, accumulated as a single continuous bout or as several short bouts over the course of the day, have distinct effects on blood glucose metabolism.

Clearly, more research is needed on the dose-response of higher-frequency and longer-duration physical activity, including activity at the lower end of the intensity continuum. In our study, we acknowledge that for those who are unfit and overweight/obese or for the elderly 3.2 km/h is close to the historical threshold of three METs, which defines moderate-intensity and increases the propensity for some exercise-like responses. However, our previous study found walking at 5.6–6.2 km/h was no more efficacious than breaks that were 3.2 km/h [7]. Current public health guidelines state that 30 min of moderate-intensity physical activity can be accumulated throughout the day in bouts >10 min [30]. Although the total accumulated time was similar (34 min) in our study, the relative reduction in postprandial hyperglycaemia with shorter and less intense bouts makes this a potentially feasible strategy for those who cannot (or behaviourally will not) tolerate exercise at a higher intensity. Furthermore, as light-intensity physical activity manifests in several typical domestic and occupational tasks, such as standing, light yard/house work (e.g. cleaning dishes, ironing) and casual walking, this makes it a potentially feasible strategy and may be incorporated into a variety of settings (office, home, etc.).

The sustained rise in the glucose levels with prolonged sitting is most probably related to the redundancy of skeletal muscle contractions, as breaks in sitting would increase muscle glucose uptake by virtue of associated contractile activity. Muscle contractions stimulate blood glucose clearance, independent of insulin, but can also act synergistically with insulin to enhance muscle glucose uptake [8]. These effects may be mediated by metabolic factors within the cell [i.e. AMPK (AMP-activated protein kinase)- and calcium-activated proteins] [31], as well as haemodynamic changes (i.e. blood volume increment, tissue perfusion and capillary permeability) [3234] associated with the muscular activity. The metabolic benefit of breaking up sitting in our study is highlighted by the flattening of the glucose curve, evidenced by the lower dG/dt and the lower insulin to glucose ratio, indicating greater economy in insulin secretion. In contrast, the higher insulin to glucose ratio in SIT suggests that glucose clearance is largely regulated by plasma insulin. This can present a ‘challenge’ to glucose homoeostasis between meals when carbohydrate absorption from the gastrointestinal tract decline and persistently high insulin levels cause a rapid decline in blood glucose concentrations. Although not specifically measured, a higher dG/dt may promote reactive hypoglycaemia, counter-regulatory hormone secretion and elevated non-esterified fatty acid (NEFA) concentrations [35]. Therefore, the negative metabolic consequences of prolonged sitting may include a complex interplay of several factors involving: (i) elevated counter-regulatory hormones and NEFAs, (ii) a redundancy of haemodynamic changes, and (iii) a dependency on insulin-mediated glucose uptake, which in some individuals may be impaired.

Contrary to our proposed hypothesis, we saw no effect on model-based assessments of hepatic and peripheral insulin sensitivity following 3 days of BREAKS when compared with SIT. One possible explanation is the significant heterogeneity in baseline measures of insulin sensitivity. The present study included older overweight/obese individuals, based on observational research suggesting that this demographic is at greatest risk of developing Type 2 diabetes. However, it should be noted that this group included both mildly overweight and morbidly obese individuals who are likely to vary in their treatment response. In BREAKS, we observed an inverse association between baseline insulin sensitivity and the day 1–3 change in MISI [r=−0.57, P=0.011 (results not shown)], indicating that, over time, breaking up sitting had a greater affect in individuals with poorer muscle insulin sensitivity. This is consistent with evidence suggesting that the metabolic benefit of exercise interventions is greater in Type 2 diabetic patients than normal controls [36] and more evident in diabetic patients with greater body weight, fat accumulation and poorer diabetic control [37]. However, no relationship (positive or otherwise) was observed between baseline insulin sensitivity and the temporal change in MISI in the sedentary condition. This might be because we recruited sedentary individuals and, as such, the accumulated sitting time during the experimental conditions was not significantly different from objective measurements taken at baseline. This is contrast with Lyden et al. [12], who, in physically active individuals, showed that reducing leisure time physical activity and increasing sitting time for 7 days decreased insulin action. Previous bed-rest studies that show worsening insulin sensitivity have failed to report the level of experimental change from habitual levels of physical activity/sedentary behaviours [11,3840]. As these studies have primarily involved trained and healthy individuals, who are likely to be more physically active than the current cohort, a greater effect on metabolic variables and measures of insulin sensitivity would be expected, especially with such an extreme model of physical inactivity.

It is also possible that changes in insulin sensitivity are not seen either until an individual shows clinically significant weight gain or there is significant caloric excess. One of the tenets of the inactivity physiology paradigm is that energy intakes are not normally reduced to match the low energy expenditure associated with prolonged sitting, resulting in a situation of energy surplus [41]. In our study, baseline weights were precisely maintained across protocols as energy intakes approximated the low level of energy expenditure. However, as energy intakes (and meals) were standardized across conditions, small between-condition differences in the energy expenditure were created throughout the day with the addition of the short walking bouts. In an individual weighing 90.5 kg, it is estimated that treadmill walking at 3.2 km/h would create a difference in energy expenditure of only 38 kJ per bout [42]. We suspect that the net sum of the breaks (∼640 kJ/day) was not sufficient in magnitude and/or duration to significantly affect insulin sensitivity. In line with this notion, Stephens et al. [43] showed that, when a day of prolonged sitting was accompanied by a 4184 kJ/day energy surplus, insulin action determined from glucose infusion technique was 26% lower compared with when energy intakes approximated the energy expenditure of prolonged sitting. Furthermore, compared with a day of minimal sitting (involving standing and ambulation, ∼45 min/h), whole-body insulin action was not significantly different from prolonged sitting when both conditions were in approximate energy balance (+678 kJ/day compared with −126 kJ/day respectively). However, there was a trend towards improved insulin sensitivity in the minimal sitting condition, suggesting that factors other than energy surplus might be involved in the decline in insulin action.

Interestingly, in both of our experimental conditions there were changes over time in certain outcome variables (fasting glucose, fasting triglycerides, postprandial triglycerides and HOMA-β%), suggesting that these variables do not reflect the dynamic circumstances associated with breaking up sitting time. Another aspect of the experimental protocol could also account for these changes; however, a number of potentially relevant variables were not measured (e.g., NEFA and markers of de novo lipogensis). Thus, we can only speculate as to the underlying mechanisms. The strictly controlled diet, specifically the increase in the relative contribution of carbohydrates to total energy, may be relevant. High-carbohydrate diets can increase fasting and postprandial triglycerides via hepatic de novo lipogenesis [4446], an effect that can occur within 3 days of feeding [45]. However, we cannot exclude the possibility that the small reduction in baseline levels of physical activity may have contributed to a reduction in lipoprotein lipase activity and peripheral clearance of triglycerides [41]. HOMA-β% also increased in both conditions over time, characterized by a reduction in fasting glucose. Reasons for the change in HOMA-β% remain unknown, but may relate to a role of postprandial hypertriglycaeridemia on β-cell function [47]. However, HOMA-β% was assessed in the basal state and dynamic measures of β-cell function may be more appropriate in the context of the current experimental design.

Strengths of our study include: the strict supervision of experimental conditions; the statistical adjustment for baseline discrepancies in diet and physical activity; and controlled feeding during experimental conditions, which avoided the confounding influence of large changes in energy balance. However, there are limitations. First, the simultaneous implementation of diet and activity protocols meant that we could not dissociate time-specific effects of diet from that of the experimental conditions. Secondly, we suspect that the small sample size and/or the inherent natural variation in the sample population hindered our ability to detect between-treatment differences in insulin sensitivity. Other considerations include the use of surrogate measures of insulin sensitivity and the basal-state assessment of β-cell function.

In conclusion, our findings are consistent with our previous single-day study and the theoretical suggestion of metabolic activation with intermittent bouts of gentle walking during prolonged sitting. Although there were no cumulative short-term effects on metabolic outcomes, the between-condition differences further highlight the need to examine the chronic effects of repetitive (daily) exposure to breaking up prolonged sitting on postprandial hyperglycaemia and hyperinsulinaemia, as well as the potential benefit to certain patient populations (e.g. Type 2 diabetes, morbidly obese) and the mechanisms and pathways that connect acute observations to a reduction in the development of chronic disease.

AUTHOR CONTRIBUTION

David Dunstan, Bronwyn Kingwell, Robyn Larsen, Ester Cerin, Jonathan Shaw, Genevieve Healy, Marc Hamilton and Neville Owen were involved in the concept and design of the study. Robyn Larsen, Carol Robinson and Louise Hammond were involved in participant recruitment, data collection and ensured that trial was conducted according to the specified protocol. Ester Cerin was responsible for the statistical analysis of the data. Robyn Larsen wrote the manuscript and editing of the manuscript for important intellectual content was done by Bronwyn Kingwell, Neville Owen, Ester Cerin and David Dunstan. Final manuscript revisions were made by Jonathan Shaw, Genevieve Healy and Marc Hamilton. David Dunstan obtained funding and takes full responsibility for the work as a whole, including the study design, access to data and the decision to submit and publish the manuscript.

We thank Miriam Clayfield (research nurse) and Anne-Sofie Laursen (research assistant) for assistance with participant recruitment and setting up clinical aspects of the study and the staff at the Alfred Heart Centre for providing technical assistance. Part of this work was presented at the 2013 American College of Sports Medicine Annual Scientific Meeting in Indianapolis, IN, U.S.A.

FUNDING

This work was supported by the National Health and Medical Research Council [grant numbers NHMRC #540107, NHMRC #569940, NHMRC #1099454 (to B.A.K.), NHMRC #586623 (to J.E.S.) and NHMRC #1003960 (to N.O.)]; the Australian Research Council Future Research Fellowship [grant number FT100100918 (to D.W.D.)]; the Australian Heart Foundation [grant number #PH 12B 7054 (to G.N.H.)]; the Coca-Cola and the Edward G. Schleider Foundation (to M.T.H.); and the Victorian State Government and a NHMRC Centre of Research Excellence Grant [grant number NHMRC #1041056].

Abbreviations

     
  • AUC

    area under the curve

  •  
  • BMI

    body mass index

  •  
  • cpm

    counts per minute

  •  
  • dG/dt

    rate of decline in plasma glucose concentration

  •  
  • GEE

    generalized estimating equation

  •  
  • HOMA-IR

    homeostasis model assessment of insulin resistance

  •  
  • HOMA-β%

    homeostasis model assessment of β-cell function

  •  
  • iAUC

    incremental area under the curve

  •  
  • MISI

    muscle insulin sensitivity index

  •  
  • MTT

    meal tolerance test

  •  
  • MVPA

    moderate-to-vigorous physical activity

  •  
  • NEFA

    non-esterified fatty acid

  •  
  • tAUC

    total AUC

References

References
1
Owen
 
N.
Salmon
 
J.
Koohsari
 
M.J.
Turrell
 
G.
Giles-Corti
 
B.
 
Sedentary behaviour and health: mapping environmental and social contexts to underpin chronic disease prevention
Br. J. Sports Med.
2014
, vol. 
48
 (pg. 
174
-
177
)
[PubMed]
2
Grontved
 
A.
Hu
 
F.B.
 
Television viewing and risk of type 2 diabetes, cardiovascular disease, and all-cause mortality: a meta-analysis
JAMA
2011
, vol. 
305
 (pg. 
2448
-
2455
)
[PubMed]
3
Dunstan
 
D.W.
Barr
 
E.L.
Healy
 
G.N.
Salmon
 
J.
Shaw
 
J.E.
Balkau
 
B.
Magliano
 
D.J.
Cameron
 
A.J.
Zimmet
 
P.Z.
Owen
 
N.
 
Television viewing time and mortality: the Australian Diabetes, Obesity and Lifestyle Study (AusDiab)
Circulation
2010
, vol. 
121
 (pg. 
384
-
391
)
[PubMed]
4
Healy
 
G.N.
Dunstan
 
D.W.
Salmon
 
J.
Cerin
 
E.
Shaw
 
J.E.
Zimmet
 
P.Z.
Owen
 
N.
 
Breaks in sedentary time: beneficial associations with metabolic risk
Diabetes Care
2008
, vol. 
31
 (pg. 
661
-
666
)
[PubMed]
5
Healy
 
G.N.
Matthews
 
C.E.
Dunstan
 
D.W.
Winkler
 
E.A.
Owen
 
N.
 
Sedentary time and cardio-metabolic biomarkers in US adults: NHANES
Eur. Heart J.
2011
, vol. 
32
 (pg. 
2003
(pg. 
590
-
06
)-
597
)
[PubMed]
6
Peddie
 
M.C.
Bone
 
J.L.
Rehrer
 
N.J.
Skeaff
 
C.M.
Gray
 
A.R.
Perry
 
T.L.
 
Breaking prolonged sitting reduces postprandial glycemia in healthy, normal-weight adults: a randomized crossover trial
Am. J. Clin. Nutr.
2013
, vol. 
98
 (pg. 
358
-
366
)
[PubMed]
7
Dunstan
 
D.W.
Kingwell
 
B.A.
Larsen
 
R.
Healy
 
G.N.
Cerin
 
E.
Hamilton
 
M.T.
Shaw
 
J.E.
Bertovic
 
D.A.
Zimmet
 
P.Z.
Salmon
 
J.
Owen
 
N.
 
Breaking up prolonged sitting reduces postprandial glucose and insulin responses
Diabetes Care
2012
, vol. 
35
 (pg. 
976
-
983
)
[PubMed]
8
Nesher
 
R.
Karl
 
I.E.
Kipnis
 
D.M.
 
Dissociation of effects of insulin and contraction on glucose transport in rat epitrochlearis muscle
Am. J. Physiol.
1985
, vol. 
249
 (pg. 
C226
-
C232
)
[PubMed]
9
Yanagibori
 
R.
Suzuki
 
Y.
Kawakubo
 
K.
Makita
 
Y.
Gunji
 
A.
 
Carbohydrate and lipid metabolism after 20 days of bed rest
Acta. Physiol. Scand. Suppl.
1994
, vol. 
616
 (pg. 
51
-
57
)
[PubMed]
10
Smorawinski
 
J.
Nazar
 
K.
Kaciuba-Uscilko
 
H.
Kaminska
 
E.
Cybulski
 
G.
Kodrzycka
 
A.
Bicz
 
B.
Greenleaf
 
J.E.
 
Effects of 3-day bed rest on physiological responses to graded exercise in athletes and sedentary men
J. Appl. Physiol.
2001
, vol. 
91
 (pg. 
249
-
257
)
[PubMed]
11
Lipman
 
R.
Raskin
 
P.
Love
 
T.
Triebwasser
 
J.
Lecocq
 
F.
Schnure
 
J.
 
Glucose intolerance during decreased physical activity in man
Diabetes
1972
, vol. 
21
 (pg. 
101
-
107
)
[PubMed]
12
Lyden
 
K.
Keadle
 
S.K.
Staudenmayer
 
J.
Braun
 
B.
Freedson
 
P.S.
 
Discrete features of sedentary behavior impact cardiometabolic risk factors
Med. Sci. Sports Exerc.
2014
 
doi:10.1249/MSS.0000000000000499
[PubMed]
13
Haskell
 
W.L.
Lee
 
I.M.
Pate
 
R.R.
Powell
 
K.E.
Blair
 
S.N.
Franklin
 
B.A.
Macera
 
C.A.
Heath
 
G.W.
Thompson
 
P.D.
Bauman
 
A.
, et al 
Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association
Circulation
2007
, vol. 
116
 (pg. 
1081
-
1093
)
[PubMed]
14
Freedson
 
P.S.
Melanson
 
E.
Sirard
 
J.
 
Calibration of the Computer Science and Applications, Inc. accelerometer
Med. Sci. Sports Exerc.
1998
, vol. 
30
 (pg. 
777
-
781
)
[PubMed]
15
Schofield
 
W.N.
 
Predicting basal metabolic rate, new standards and review of previous work
Hum. Nutr. Clin. Nutr.
1985
, vol. 
39
 
Suppl. 1
(pg. 
5
-
41
)
[PubMed]
16
Mohanlal
 
N.
Holman
 
R.R.
 
A standardized triglyceride and carbohydrate challenge: the oral triglyceride tolerance test
Diabetes Care
2004
, vol. 
27
 (pg. 
89
-
94
)
[PubMed]
17
Maki
 
K.C.
McKenney
 
J.M.
Farmer
 
M.V.
Reeves
 
M.S.
Dicklin
 
M.R.
 
Indices of insulin sensitivity and secretion from a standard liquid meal test in subjects with type 2 diabetes, impaired or normal fasting glucose
Nutr. J.
2009
, vol. 
8
 pg. 
22
 
[PubMed]
18
Brodovicz
 
K.G.
Girman
 
C.J.
Simonis-Bik
 
A.M.
Rijkelijkhuizen
 
J.M.
Zelis
 
M.
Bunck
 
M.C.
Mari
 
A.
Nijpels
 
G.
Eekhoff
 
E.M.
Dekker
 
J.M.
 
Postprandial metabolic responses to mixed versus liquid meal tests in healthy men and men with type 2 diabetes
Diabetes Res. Clin. Pract.
2011
, vol. 
94
 (pg. 
449
-
455
)
[PubMed]
19
Matthews
 
D.R.
Hosker
 
J.P.
Rudenski
 
A.S.
Naylor
 
B.A.
Treacher
 
D.F.
Turner
 
R.C.
 
Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man
Diabetologia
1985
, vol. 
28
 (pg. 
412
-
419
)
[PubMed]
20
Abdul-Ghani
 
M.A.
Matsuda
 
M.
Balas
 
B.
DeFronzo
 
R.A.
 
Muscle and liver insulin resistance indexes derived from the oral glucose tolerance test
Diabetes Care
2007
, vol. 
30
 (pg. 
89
-
94
)
[PubMed]
21
Expert Committee on the Diagnosis and Classification of Diabetes Mellitus
Report of the expert committee on the diagnosis and classification of diabetes mellitus
Diabetes Care
2003
, vol. 
26
 
Suppl. 1
(pg. 
S5
-
S20
)
[PubMed]
22
Edelstein
 
S.L.
Knowler
 
W.C.
Bain
 
R.P.
Andres
 
R.
Barrett-Connor
 
E.L.
Dowse
 
G.K.
Haffner
 
S.M.
Pettitt
 
D.J.
Sorkin
 
J.D.
Muller
 
D.C.
, et al 
Predictors of progression from impaired glucose tolerance to NIDDM: an analysis of six prospective studies
Diabetes
1997
, vol. 
46
 (pg. 
701
-
710
)
[PubMed]
23
Chiasson
 
J.L.
Josse
 
R.G.
Gomis
 
R.
Hanefeld
 
M.
Karasik
 
A.
Laakso
 
M.
Group
 
S.-N.T.R.
 
Acarbose for prevention of type 2 diabetes mellitus: the STOP-NIDDM randomised trial
Lancet
2002
, vol. 
359
 (pg. 
2072
-
2077
)
[PubMed]
24
Barclay
 
A.W.
Petocz
 
P.
McMillan-Price
 
J.
Flood
 
V.M.
Prvan
 
T.
Mitchell
 
P.
Brand-Miller
 
J.C.
 
Glycemic index, glycemic load, and chronic disease risk–a meta-analysis of observational studies
Am. J. Clin. Nutr.
2008
, vol. 
87
 (pg. 
627
-
637
)
[PubMed]
25
Nygaard
 
H.
Tomten
 
S.E.
Hostmark
 
A.T.
 
Slow postmeal walking reduces postprandial glycemia in middle-aged women
Appl. Physiol. Nutr. Metab.
2009
, vol. 
34
 (pg. 
1087
-
1092
)
[PubMed]
26
Hostmark
 
A.T.
Ekeland
 
G.S.
Beckstrom
 
A.C.
Meen
 
H.D.
 
Postprandial light physical activity blunts the blood glucose increase
Prev. Med.
2006
, vol. 
42
 (pg. 
369
-
371
)
[PubMed]
27
Nelson
 
J.D.
Poussier
 
P.
Marliss
 
E.B.
Albisser
 
A.M.
Zinman
 
B.
 
Metabolic response of normal man and insulin-infused diabetics to postprandial exercise
Am. J. Physiol.
1982
, vol. 
242
 (pg. 
E309
-
E316
)
[PubMed]
28
Lunde
 
M.S.
Hjellset
 
V.T.
Hostmark
 
A.T.
 
Slow post meal walking reduces the blood glucose response: an exploratory study in female Pakistani immigrants
J. Immigr. Minor. Health.
2012
, vol. 
14
 (pg. 
816
-
822
)
[PubMed]
29
Takaishi
 
T.
Imaeda
 
K.
Tanaka
 
T.
Moritani
 
T.
Hayashi
 
T.
 
A short bout of stair climbing-descending exercise attenuates postprandial hyperglycemia in middle-aged males with impaired glucose tolerance
Appl. Physiol. Nutr. Metab.
2012
, vol. 
37
 (pg. 
193
-
196
)
[PubMed]
30
Pate
 
R.R.
Pratt
 
M.
Blair
 
S.N.
Haskell
 
W.L.
Macera
 
C.A.
Bouchard
 
C.
Buchner
 
D.
Ettinger
 
W.
Heath
 
G.W.
King
 
A.C.
, et al 
Physical activity and public health. A recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine
JAMA
1995
, vol. 
273
 (pg. 
402
-
407
)
[PubMed]
31
Santos
 
J.M.
Ribeiro
 
S.B.
Gaya
 
A.R.
Appell
 
H.J.
Duarte
 
J.A.
 
Skeletal muscle pathways of contraction-enhanced glucose uptake
Int. J. Sports Med.
2008
, vol. 
29
 (pg. 
785
-
794
)
[PubMed]
32
Inyard
 
A.C.
Chong
 
D.G.
Klibanov
 
A.L.
Barrett
 
E.J.
 
Muscle contraction, but not insulin, increases microvascular blood volume in the presence of free fatty acid-induced insulin resistance
Diabetes
2009
, vol. 
58
 (pg. 
2457
-
2463
)
[PubMed]
33
Hespel
 
P.
Vergauwen
 
L.
Vandenberghe
 
K.
Richter
 
E.A.
 
Important role of insulin and flow in stimulating glucose uptake in contracting skeletal muscle
Diabetes
1995
, vol. 
44
 (pg. 
210
-
215
)
[PubMed]
34
Holmang
 
A.
Mimura
 
K.
Lonnroth
 
P.
 
Involuntary leg movements affect interstitial nutrient gradients and blood flow in rat skeletal muscle
J. Appl. Physiol.
2002
, vol. 
92
 (pg. 
982
-
988
)
[PubMed]
35
Ludwig
 
D.S.
 
The glycemic index: physiological mechanisms relating to obesity, diabetes, and cardiovascular disease
JAMA
2002
, vol. 
287
 (pg. 
2414
-
2423
)
[PubMed]
36
Thorburn
 
A.W.
Gumbiner
 
B.
Bulacan
 
F.
Brechtel
 
G.
Henry
 
R.R.
 
Multiple defects in muscle glycogen synthase activity contribute to reduced glycogen synthesis in non-insulin dependent diabetes mellitus
J. Clin. Invest.
1991
, vol. 
87
 (pg. 
489
-
495
)
[PubMed]
37
Oguri
 
M.
Adachi
 
H.
Ohno
 
T.
Oshima
 
S.
Kurabayashi
 
M.
 
Effect of a single bout of moderate exercise on glucose uptake in type 2 diabetes mellitus
J. Cardiol.
2009
, vol. 
53
 (pg. 
8
-
14
)
[PubMed]
38
Mikines
 
K.J.
Richter
 
E.A.
Dela
 
F.
Galbo
 
H.
 
Seven days of bed rest decrease insulin action on glucose uptake in leg and whole body
Appl. Physiol.
1991
, vol. 
70
 (pg. 
1245
-
1254
)
39
Alibegovic
 
A.C.
Højbjerre
 
L.
Sonne
 
M.P.
van Hall
 
G.
Stallknecht
 
B.
Dela
 
F.
Vaag
 
A.
 
Impact of 9 days of bed rest on hepatic and peripheral insulin action, insulin secretion, and whole-body lipolysis in healthy young male offspring of patients with type 2 diabetes
Diabetes
2009
, vol. 
58
 (pg. 
2749
-
2756
)
[PubMed]
40
Smorawiński
 
J.
Kaciuba-Uściłko
 
H.
Nazar
 
K.
Kubala
 
P.
Kamińska
 
E.
Ziemba
 
A.W.
Adrian
 
J.
Greenleaf
 
J.E.
 
Effects of three-day bed rest on metabolic, hormonal and circulatory responses to an oral glucose load in endurance or strength trained athletes and untrained subjects
J. Physiol. Pharmacol.
2000
, vol. 
51
 (pg. 
279
-
289
)
[PubMed]
41
Hamilton
 
M.T.
Hamilton
 
D.G.
Zderic
 
T.W.
 
Role of low energy expenditure and sitting in obesity, metabolic syndrome, type 2 diabetes, and cardiovascular disease
Diabetes
2007
, vol. 
56
 (pg. 
2655
-
2667
)
[PubMed]
42
Levine
 
J.A.
Schleusner
 
S.J.
Jensen
 
M.D.
 
Energy expenditure of nonexercise activity
Am. J. Clin. Nutr.
2000
, vol. 
72
 (pg. 
1451
-
1454
)
[PubMed]
43
Stephens
 
B.R.
Granados
 
K.
Zderic
 
T.W.
Hamilton
 
M.T.
Braun
 
B.
 
Effects of 1 day of inactivity on insulin action in healthy men and women: interaction with energy intake
Metabolism
2011
, vol. 
60
 (pg. 
941
-
949
)
[PubMed]
44
Schwarz
 
J.M.
Linfoot
 
P.
Dare
 
D.
Aghajanian
 
K.
 
Hepatic de novo lipogenesis in normoinsulinemic and hyperinsulinemic subjects consuming high-fat, low-carbohydrate and low-fat, high-carbohydrate isoenergetic diets
Am. J. Clin. Nutr.
2003
, vol. 
77
 (pg. 
43
-
50
)
[PubMed]
45
Chong
 
M.F.
Hodson
 
L.
Bickerton
 
A.S.
Roberts
 
R.
Neville
 
M.
Karpe
 
F.
Frayn
 
K.N.
Fielding
 
B.A.
 
Parallel activation of de novo lipogenesis and stearoyl-CoA desaturase activity after 3 d of high-carbohydrate feeding
Am. J. Clin. Nutr.
2008
, vol. 
87
 (pg. 
817
-
823
)
[PubMed]
46
Hudgins
 
L.C.
Hellerstein
 
M.K.
Seidman
 
C.E.
Neese
 
R.A.
Tremaroli
 
J.D.
Hirsch
 
J.
 
Relationship between carbohydrate-induced hypertriglyceridemia and fatty acid synthesis in lean and obese subjects
J. Lipid Res.
2000
, vol. 
41
 (pg. 
595
-
604
)
[PubMed]
47
Lopez
 
S.
Bermudez
 
B.
Pacheco
 
Y.M.
Villar
 
J.
Abia
 
R.
Muriana
 
F.J.
 
Distinctive postprandial modulation of beta cell function and insulin sensitivity by dietary fats: monounsaturated compared with saturated fatty acids
Am. J. Clin. Nutr.
2008
, vol. 
88
 (pg. 
638
-
644
)
[PubMed]

Author notes

Trial Registration Number: This trial was registered with the ANZCTR clinical trial registry ACTRN12610000657022 (http://www.anzctr.org.au).

Trial Registration Date: 11/08/2010