Setting reference intervals without considering sex and age differences

We compared two sets of sex and age-specific reference intervals obtained from two large reference populations, one set calculated with data from 700,000 reference individuals by the nonlinear optimizing method of Ohgushi and Shibata and the other set calculated with data from 150,000 reference individuals from Shizuoka prefecture by the revised Hoffmann fitting method. Ten laboratory analytes used for health screening were compared. The sex- and age-specific reference intervals for total cholesterol, fasting blood sugar, uric acid, total protein, and albumin from the two large reference sample groups closely resembled each other, but reference intervals for the enzyme analytes aspartate aminotransferase, alanine aminotransferase, lactate dehydrogenase, and gamma-glutamyl transpeptidase only partly corresponded. Surprisingly, new information came from comparison of the sex- and age-specific reference intervals of alkaline phosphatase: low activities were observed in young females and higher activities were observed in older females. If a reference interval is used that does not take this observation into account, misdiagnosis of hyperthyroidism, which is frequently observed in young women, may result. Sex- and age-specific reference intervals should be used to interpret results of laboratory screening tests.