Patients with atrophic glossitis (AG) or burning mouth syndrome (BMS) are frequently encountered in the oral mucosal disease clinic. Our previous studies found a significantly higher frequency (26.7%) of serum gastric parietal cell antibody (GPCA) and a significantly higher frequency (31%) of serum thyroglobulin antibody (TGA) or thyroid microsomal antibody (TMA) in AG patients than in healthy control subjects. Moreover, there is also a significantly higher frequency (13.3%) of serum GPCA or a significantly higher frequency (23.5%) of serum TGA or TMA in BMS patients than in healthy control subjects. Because patients with one organ-specific autoantibody are prone to have another organ-specific autoantibody in sera, we also evaluated whether AG or BMS patients with GPCA are prone to have TGA or TMA in sera and vice versa. We further found that 25.3% of TGA- or TMA-positive AG or BMS patients also have GPCA, 32.3% GPCA-positive AG or BMS patients also have TGA, and 30.6% GPCA-positive AG or BMS patients also have TMA in their sera. Without proper diagnosis and treatment, patients with GPCA are more likely to develop autoimmune atrophic gastritis and subsequently progress to gastric carcinoma, and patients with TGA or TMA may develop autoimmune thyroid disease and finally result in thyroid dysfunction. In addition, previous studies have shown a close association of the HLA-DR or HLA-DQ loci with the presence of autoantibodies (such as GPCA, TGA or TMA) in patients with different types of autoimmune disease. Therefore, in the following 3-year research project, we plan to collect 300 AG and 450 BMS patients from the oral mucosal disease clinic of Department of Dentistry, National Taiwan University Hospital. For each year, 100 AG and 150 BMS patients are collected. A 10-cc blood sample will be drawn from each patient, with 5 cc being used for the determination of the serum levels of GPCA, TGA and TMA and another 5 cc being used for the HLA-DRB1 and HLA-DQB1-genotyping using the polymerase chain reaction with sequence-specific primer (PCR-SSP) typing technique. At the end of this 3-year study, we will realize the frequencies of presence of GPCA, TGA and TMA in sera of our 300 AG or 450 BMS patients. After statistical analyses, we will also know which specific HLA-DRB1 or HLA-DQB1 allele and which specific DRB1-DQB1 haplotype are responsible for the possession of GPCA, TGA or TMA in sera of our AG or BMS patients. In addition, we will understand which specific HLA-DRB1 or HLA-DQB1 allele and which specific DRB1-DQB1 haplotype are responsible for the possession of GPCA in TGA- or TMA-positive AG or BMS patients as well as for the possession of TGA or TMA in GPCA-positive AG or BMS patients. With this important information in mind, we can predict the development of the specific autoimmune diseases such as autoimmune atrophic gastritis and autoimmune thyroid diseases and then adopt proper early diagnosis and treatment to prevent the future occurrence of these diseases and their potential complications (such as gastric carcinoma or thyroid dysfunction).