h group and age strata as fixed factors. Since there was evidence of difference in variability in Glu and Glx between the HC and FM groups, we preformed an additional analysis using weighted least squares, with weights equaling the inverse of the corresponding estimated group variances. We next correlated pressure pain thresholds with Glu bcr-abl and Glx levels from the posterior insula as these levels were found to be elevated in the FM participants. Pearson,s correlations were calculated on the combined group of FM and HC participants. Separate multiple linear regression models were constructed with Glu or Glx as dependent variables, and group medium pressure threshold, and the age strata as independent variables. Significance was set at a p value of 0. 05.
Results FM Patients have Elevated Glu and Glx Levels in the Posterior Insula As evidenced by Table 1 and Figure 2A, individuals with FM displayed elevated levels of both Pelitinib Glu and Glx within the posterior insula. Glu and Glx remained significantly elevated in similar analyses that used weighted least squares. 18 of the 19 FM patients had Glu and Glx levels that were higher than the healthy control mean. FM patients also had a trend towards higher NAA in the posterior insula however this did not meet statistical significance. There were no differences between FM and HC groups in the other major metabolites within the posterior insula. These data suggest a relatively specific elevation of Glu and Glx in right posterior insula for the FM group. As evidenced by Table 1 and Figure 2B, there were no significant group differences in Glu and Glx or other metabolites within the anterior insula.
These data suggest that the elevated levels of Glu are specific for the posterior insula and do not extend into the anterior regions. Glu and Glx Levels are Negatively Correlated with Pressure Pain Thresholds Significant negative correlations between pressure pain thresholds and posterior insula Glu and Glx levels were observed when both groups were combined. A scatter plot of posterior insula Glx values versus medium pressure pain thresholds is illustrated in Figure 3. These data suggest that, regardless of whether an individual is a FM patient or control, individuals with higher levels of Glu and/or Glx also have enhanced sensitivity to experimentally induced pressure pain.
Since group status and pressure pain thresholds were both related to Glu and Glx levels in the posterior insula, we constructed separate linear regression models with Harris et al. Page 4 Arthritis Rheum. Author manuscript, available in PMC 2010 October 1. either Glu or Glx as dependent variable, and group and medium pressure pain threshold as independent variables. Since the FM and HC were age matched, we further used age as a stratum variable in the regression model. This is akin to the stratified analysis traditionally carried out in case control designs. As evidenced by Table 3, both group and pressure pain threshold were significant predictors of Glx and these factors trended towards significance for Glu. For both Glu and Glx, FM patients exhibited higher values than the controls. For example, the FM patients on an average had 1. 16 units higher Glx values compared to the healthy controls, for a fixed pressure pain level and age stratum.