Graph prism v 7

Graph Prism v.7 was used for statistical analysis. The Passing–Bablok regression analysis was utilised to determine the correlation between DBS elute and plasma Nf-L levels measured by MSD. Comparison of protein expression values from DBS/Np-DPS elutes and plasma was performed using Pearson’s correlation analysis, while group analysis was obtained using the multiple comparisons Kruskal-Wallis test. Adjustment for age was also undertaken with 1) univariate general linear model with covariate of age (SPSS v25) and 2) calculating the expected Nf-L level in each control adding the amount of Nf-L that would be expected factoring in additional 9 years of age, using the function reported by Siller et al., where a 2.2% increase of Nf-L is reported for each additional year of age. Together with the adjustment for age performed in the group analysis, this method provides an additional way to normalize for age differences between the groups based on age-related increase of blood Nf-L levels. Evaluation of diagnostic performance was undertaken using a Receiving Operator Characteristic (ROC) analysis in Graph Prism v.7, comparing plasma and DBS elutes from ALS patients and control individuals.

Data were presented as mean ± standard deviation and compared between baseline and after 1 year of follow-up by using the Wilcoxon Signed-Rank test or paired T-test, as appropriate. Comparisons between sex were analysed by T-test or non-parametric Wilcoxon test, as appropriate. The correlation matrix for miR-21/ROS/HNE and other collected variables at follow-up was evaluated. Logistic regression models to evaluate the association between R group and each glycaemic parameter were performed and receiver operating characteristic (ROC) curves were drawn. For each model odds ratio (OR) with 95% CI (confidence interval) and area under the curve (AUC) were calculated. To evaluate the association between ROS/miR-21 values and glycaemic parameters, linear regression models were implemented. All statistical analyses were performed with SAS Software 9.4 and graphs were drawn by GraphPrism v7. P-values < 0.05 were considered statistically significant for all analyses.