A systematic review and meta-analysis of the effects of interrupting prolonged sitting with physical activity breaks on blood glucose, insulin, triacylglycerol

INTRODUCTION: Sedentary behaviour has been associated with all-cause mortality independent of leisure time physical activity (PA), although such findings have been disputed. The primary aim was to systematically review and meta-analyse experimental trials breaking up prolonged sitting with intermittent bouts of PA throughout the day (INT) compared with sitting (SIT) on glycaemia, insulinaemia and triacylglycerol (TAG). A secondary aim was comparing the effects of INT against continuous exercise (EX), on glucose, insulin and TAG. METHODS: PRISMA recommendations were followed. PROSPERO Registration: CRD42017080982 PubMed, OvidSP: Journals@Ovid and PsychINFO, Science Direct, SPORTDiscus were systematically searched on 04/03/2017. Eligibility Criteria: control trials comparing INT vs SIT or INT vs one bout of EX prior to sitting, in participants aged 18 or above, healthy, or with type 2 diabetes, but not with other health conditions such as chronic obstructive pulmonary disease or peripheral arterial disease. Risk of Bias (RoB) assessed with Cochrane RoB tool RESULTS: INT vs SIT: 16 trial comparisons, 618 participants for TAG, 30 trial comparisons ,1036 participants for glucose, and 896 participants,24 trial comparisons for insulin, were meta-analysed. INT vs EX: 6 trial comparisons,62 men, 75 women for TAG, 7 studies,90 men and 60 women for insulin, whereas 160 men and women, 8 studies for glucose, were meta-analysed. Synthesis: INT vs SIT: For glucose, SMD of -0.60 [-0.81, -0.39] in favour of INT. For insulin, SMD of -0.69 [-0.98, -0.41] in favour of INT. For TAG, SMD of - 0.27 [-0.44, -0.09] in favour of INT. BMI was associated with glucose responses (β= -.064, 95% CI: -0.123, -0.005, p= .034), but not insulin (β= -.066, -0.152, 0.02, p=.122), nor TAG (β= .010, -.048, .067, p= .722) INT vs EX: For glucose, SMD was -0.24 [-0.56, 0.08] in favour of INT, but not statistically significant (p=0.15). For TAG, SMD of 0.21 [-0.16, 0.57] in favour of EX vs INT, and not statistically significant (p=0.27). SMD for insulin was -0.08 [-0.31, 0.16] in favour of INT, and not statistically significant (p=0.53). When energy expenditure was matched, there was an effect in favour of INT, -0.36 [-0.65, -0.06] (p=0.02) on glucose, but not insulin, -0.16 [-0.40, 0.08] (p=0.20). CONCLUSION: Most trials did not report randomisation or allocation methods, nor attempted blinding, nor handling of data attrition. There was possible publication bias for TAG. BMI moderates glycaemic response to PA breaks, but insufficient studies reported cardiorespiratory fitness or quantitative PA status, to allow regression. PA breaks in sitting moderately attenuate post-prandial glucose, insulin, and TAG, with greater glycaemic attenuation in people with higher BMI, but any differences compared to one continuous bout of exercise are small, for glucose when exercise protocols are energy matched, or non-existent for insulin and TAG. Therefore, PA breaks can attenuate standard markers of metabolic health