Prism software version 4

Experimental procedures and data analysis were performed as previously described (33 (link), 125 (link)). C. elegans wild-type Bristol strain N2 worms were grown at 22°C on nematode growth medium (NGM) agar plates using E. coli OP50 as the nutrient. Untreated or treated bacteria (10⁹ CFU mL⁻¹) were spread onto NGM solidified agar plates before incubation at 37°C overnight. The plates were cooled to room temperature for 4 h, and 20 to 30 L4-synchronized worms were plated and incubated at 22°C in a humid environment to prevent plate drying. Worm survival was scored daily for 32 days using an Axiovert S100 optical microscope (Zeiss, Oberkochen, Germany) equipped with a digital camera (DXM 1200F; Nikon Instruments, Melville, NY). Four independent experiments per condition were performed, and all worms from each condition were used for the survival assay. The Kaplan-Meier method was used to calculate the nematode survival, and the significance of survival differences was tested using a log-rank test (Prism software, version 4.0; GraphPad Software, San Diego, CA). As a control to ascertain similar growth on NGM plates between treated and untreated bacteria, the NGM agar plates were entirely scraped every 5 days for enumeration on LB agar plates.

Based on findings of a pilot-study, we made a sample size estimation for the correlation between both measurements (blood glucose and interstitial glucose during exercise) which gave a number of seven patients (correlation of r = 0.91 with a power of 0.95).
Distribution was tested via D’Agostino and Pearson omnibus normality tests. Descriptive statistics is given as mean ± standard deviation (SD). Comparison of means from CGM vs. BG was performed with paired t-tests, The Bland-Altman method (bias and 95% limits of agreement) and Pearson correlation. The CGM accuracy in CON vs. HIIE was compared with paired t-test of mean difference (MD) and mean absolute relative differences (MARD) of CGM and BG between exercise conditions per patient (CON L vs. HIIE L, CON M vs. HIIE M and CON H vs. HIIE H). Individual paired interstitial glucose and reference BG values are given using the Clarke Error Grid analysis [30 (link)]. Five zones characterize errors of varying levels of clinical significance, among which values in zones A and B are defined as “clinically acceptable”, whereas values in zones C, D, or E are considered potentially unsafe and could lead to clinically significant errors. Statistical analyses were performed with Microsoft Excel (Microsoft Corporation 2007, Washington, DC, USA) and standard software package Prism Software version 4.0 (GraphPad, La Jolla, CA, USA).

Data are expressed as the mean±SEM. The statistical analysis of differences between two groups was assessed with the unpaired t-test, and the differences among more than three groups were assessed by ANOVA and multiple comparison tests with Prism Software version 4.0 (GraphPad Software, San Diego California USA). P values <0.05 were considered to be statistically significant.