Reproducibility of immunological tests used to assess multiple chemical sensitivity syndrome

Whether persons with multiple chemical sensitivity syndrome (MCS) have immunological abnormalities is unknown. To assess the reliability of selected immunological tests that have been hypothesized to be associated with MCS, replicate blood samples from 19 healthy volunteers, 15 persons diagnosed with MCS, and 11 persons diagnosed with autoimmune disease were analyzed in five laboratories for expression of four T-cell surface activation markers (CD25, CD26, CD38, and HLA-DR) and in four laboratories for autoantibodies (to smooth muscle, thyroid antigens, and myelin). For T-cell activation markers, the intralaboratory reproducibility was very good, with 90% of the replicates analyzed in the same laboratory differing by < or = 3%. Interlaboratory differences were statistically significant for all T-cell subsets except CD4+ cells, ranging from minor to eightfold for CD25+ subsets. Within laboratories, the date of analysis was significantly associated with the values for all cellular activation markers. Although reproducibility of autoantibodies could not be precisely assessed due to the rarity of abnormal results, there were inconsistencies across laboratories. The effect of shipping on all measurements, while sometimes statistically significant, was very small. These results support the reliability of fresh and shipped samples for detecting large (but perhaps not small) differences between groups of donors in the T-cell subsets tested. When comparing markers that are not well standardized, it may be important to distribute samples from different study groups evenly over time.