BCG immunization was routinely given to elementary and junior high school students in Japan until it was abruptly discontinued after 2002. Data on the incidence of type 2 diabetes was analyzed to see if there was a change following discontinuation of BCG. The incidence of type 2 diabetes dropped by over 50% in children from Tokyo following discontinuation of BCG immunization, the relative risk reached 2.78 (95% confidence interval 1. 03 -7.480). This drop appears to be more significant given the epidemic of type 2 diabetes in children around the world. The increased secretion of cortisol following immunization is thought to result in metabolic syndrome and type 2 diabetes. Japanese children produce larger amounts of cortisol following immunization than White children. This increased production of cortisol may explain why Japanese children have a high risk of type 2 diabetes but low risk of type 1 diabetes compared to White children. The proposed mechanism may explain the epidemic of obesity which starts in children under 6 months of age.Keywords: Type 1 diabetes, type 2 diabetes, BCG, vaccine, cortisol, inflammation.
RESEARCH DESIGN AND METHODSBCG vaccine has been linked to an increased risk of type 1 diabetes [1]. BCG immunization was routinely given to Japanese elementary school children age 6 to 7 and junior high school children age 12 to 13 since prior to 1982. Immunization was discontinued at the end of 2002 [2]. BCG immunization rates were provided by The Research Institute of Tuberculosis, JATA. The incidence of type 2 diabetes in all elementary and all junior high school students in Tokyo Japan has been recorded since 1974 [3][4][5] and verified data has been published through 2004 [5]. Data on the incidence of type 2 diabetes was analyzed to determine if the incidence declined after discontinuation of BCG vaccine. Only incidence data of type 2 diabetes after 1982 was used because the age of BCG immunization changed in 1982 and this may explain the decreased incidence of diabetes prior to this time. Statistics were performed using Epiinfo 2000, CDC, Atlanta GA and a internet based Poisson calculator. A two stratification analysis was performed with one stratification using data from elementary school children and one stratification using data from junior high school students. Both Poisson and the more conservative Mantel-Haensel Weighted Relative Risk calculation was performed and Greenland/Robins Confidence Limits created. The P values with both methods were less than 0.05, with the later being less significant.
RESULTSThe results (Table 1) show that BCG immunization was associated with an relative risk of 2.78 (1.03