2.11. Statistical analysis

Statistical analysis was performed using a two-way repeated measures ANOVA (Figs. 1,​,2)2) or a one-way ANOVA (Fig. 3) with Holm-Sidak multiple comparisons test for post-hoc contrasts. Figure 4 data were not normally distributed so statistical analysis was performed using a Kruskal-Wallis one-way analysis of variance followed by Dunn’s multiple comparisons tests for post hoc contrasts. Data are presented as the mean ± standard error of the mean (Figs. 1,​,2)2) or scatter plots (Figs. 3,​,4),4), and differences are considered significant at a P value less than 0.05 (Prism 5, GraphPad Software, San Diego, CA).

Sera was collected from normal control subjects (Controls, n=20), early (1–12 months post injury) CRPS patients (Early CRPS, n=20), chronic (> 12 months post injury) CRPS patients (Chronic CRPS, n=20), and patients that initially had CRPS and pronociceptive serum effects in fracture mice, but were re-evaluated at least a year later and at that point had resolved CRPS and resolved pronociceptive serum effects (Resolved CRPS, n=15). Sera from all four groups were tested with IgM binding dot blot studies using 4 recombinant human proteins (keratin 16, histone 3.2, gamma actin, and alpha enolase). These potential autoantigens had been identified as promising candidates from our previous liquid chromatography-mass spectrometry studies in fracture mouse skin.[42] Early CRPS patient IgM binding was increased for all 4 candidate autoantigens, relative to control subject IgM immunoreativity (A-D). Chronic CRPS IgM binding was increased only for gamma actin proteins, relative to control IgM immunoreactivity (C). Interestingly, the 2 chronic CRPS patients with pronociceptive serum effects in fracture mice (Gi-1, WK-7, Table 3) had the highest IgM immunoreactivity levels to histone 3.2, gamma actin, and alpha enolase of any chronic CRPS patients tested (B-D, red symbols). Resolved CRPS patient IgM binding levels were similar to controls (A-D). A one-way repeated measures analysis of variance was performed followed by a Sidak correction test for post hoc contrasts. Data expressed as mean values ± SD. ***P<0.001, **P<0.01, and *P<0.05 vs Controls, ##P<0.01, and #P<0.05 vs Early CRPS.