Estimation of Reference Intervals for the Shizuoka Prefecture of Japan and Comparison with Those Obtained from Other Two Large-scale Studies of Reference Populations in Japan

Abstract

Objective We calculated age- and sex-dependent reference intervals for 10 variables on the basis of health examination data of 156, 565 individuals in Shizuoka prefecture of Japan. We compared these reference intervals with age- and sex-dependent reference intervals calculated from two other large scale studies across the country populations.Design and Setting Data were obtained from three laboratories, and intervals were calculated by revised Hoffmanns fitting method. On the other hand, Shibata et al. calculated, by a non-linear optimization method (NLO), age- and sex-dependent reference intervals in 5-year age-based increments from data obtained approximately 750, 000 reference individuals from 45 laboratories all around of Japan Moreover, Suka et al. representing the Japan Association of Health Service (JAHS) established reference intervals and collected data from approximately 2.7 million individuals from metropolitan and branch offices in all prefectures of Japan with the basis of NCCLS criteria. We compared with these three large scale age- and sex-dependent reference intervals each others.Results Reference intervals for all subjects were similar between three studies and especially the serum total cholesterol, uric acid and fasting blood sugar, albumin and lactate dehydrogenase were closely coincided between our study and the other two large scale studies. Levels of serum alkaline phosphatase and postmenopausal increases in this enzyme were similar between our study and Shibata's study.Conclusion Our findings indicate that if compatibility between laboratories can be established by external quality control analysis, reliable reference intervals could be obtained in large-scale cooperative studies, and the use of these reference intervals by multiple laboratories would be possible.