Fig. 3
From: Neuropathy-specific alterations in a Mexican population of diabetic patients
Circulating levels of cytokines. (a) Box-plots of pro-inflammatory cytokine IL-6 levels (for non-zero values) showed considerable variability in all groups, although a significant increase was observed in diabetic patients. Data are represented as the median and the 25th–75th percentiles of the distribution. Whiskers comprise the 5th -95th percentiles of the distribution. The raw data is plotted to the right of box plots for all conditions: ND (N = 127), DNN (N = 16), DShP (N = 58), DSHP (N = 28) and DMN (N = 27). Different letters indicate statistical differences with a p < 0.05 and the alphabetical order points toward a decrease in magnitude (i.e. the “a” correspond to the highest value, the “b” to the second, and so on). Groups were compared by a Kruskal-Wallis non-parametric test, using a Mann-Whitney post test for a higher the statistical power with respect to Dunn test (b) A high percentage of measurements of IL-6 in in diabetic patients yielded null values. (c) Box-plots of IL-10 levels (non-zero values) were similar among groups; ND (N = 146), DNN (N = 16), DShP (N = 69), DSHP (N = 31) and DMN (N = 28). (d) A high percentage of measurements of IL-10 in diabetic patients also yielded null values. The percentages were analyzed using a Chi-squared test with Yate’s correction for continuity on 2 × 2 contingency tables