Corrigendum to “The effect of preprandial versus postprandial physical activity on glycaemia: Meta-analysis of human intervention studies”. [Diabetes Res. Clin. Pract. 210 (2024) 111638] (Diabetes Research and Clinical Practice (2024) 210, (S0168822724001281), (10.1016/j.diabres.2024.111638))

The authors regret that the original published article [1] contains an error in the calculation of the standard errors, which result in small changes of the pooled effect estimates from the random-effects models. This Corrigendum displays the corrected information ahead. Importantly, these corrections do not alter the overall interpretation of the results or the conclusion of the systematic review. - The seventh sentence of the Abstract should instead state the following:For the RCTs, glucose area under the curve during a glucose tolerance test was slightly, but not significantly lower in preprandial PA vs postprandial PA (-0.36 [95%CI: -0.79, 0.07] mmol/L, I ² = 0.0%). - The first sentence of the subsection ‘3.2 Postprandial outcomes in acute response studies’ of the Results should instead state the following:All postprandial outcomes (as provided in Table 3A and Fig. 2A-D) showed either no differences or followed the expected physiological patterns and were supported by different qualitative analyses [26, 36, 37, 41, 42, 45-50]. - The third sentence of the subsection ‘3.4 Long term outcomes from RCTs over multiple weeks’ of the Results should instead state the following:Glucose AUC during the OGTT was slightly but not significantly lower in preprandial PA (-0.36 [-0.79, 0.07] mmol/L, Fig. 2I). - The seventh to ninth sentence of the second paragraph of the the subsection ‘3.4 Long term outcomes from RCTs over multiple weeks’ of the Results should instead state the following:When pooled, estimated glucose levels were not-significantly higher in preprandial PA (0.15 [-0.14, 0.43] and 0.28 [-0.12, 0.69] mmol/L, respectively, Fig. 2J). Insulin response during the test did not differ in one trial [40]. Furthermore, fasting glucose, extracted from five studies, was also slightly but not significantly higher in preprandial PA (0. 14 [-0.09, 0.36] mmol/L, Fig. 2K). - The subsection ‘3.6 Publication bias’ of the Results should instead state the following:As we only included a few studies per outcome, funnel plots were not usable to assess publication bias (Appendix S4). However, Egger's test confirmed the absence of asymmetry in most cases (p> 0.05). Only the outcomes of mean glucose over 22-24h and glucose value before physical performance test showed statistical significant results on the Egger's test (p = 0.02162 and p = 0.0183, respectively). - Table 3 has been amended to provide the correct pooled estimates and can be viewed below.Table 3. Resume of the analysed outcome measures.[Figure presented] - Figure 2 has been amended and the correct version can be viewed ahead. Furthermore, we noticed that part I of figure 2 was not correctly inserted in the original publication. The correct figure can be found below. [Figure presented] [Figure presented] [Figure presented] [Figure presented] [Figure presented] [Figure presented] [Figure presented] [Figure presented] [Figure presented] [Figure presented] [Figure presented] - Last, Appendix S3 and Appendix S4 have been amended to include the correct funnel plots and Egger's test p-values. The corrected version of supplementary material can be viewed attached alongside this Corrigendum.The authors would like to apologise for any inconvenience caused. References 1. Slebe R, Wenker E, Schoonmade LJ, Bouman EJ, Blondin DP, Campbell DJT, et al. The effect of preprandial versus postprandial physical activity on glycaemia: Meta-analysis of human intervention studies. Diabetes Res Clin Pract. 2024;210:111638.