Abstract
Two methods of screening for diabetes mellitus were used in a recent mass survey which checked the efficiency of random urine sugar testing and set standard values using random blood glucose levels. Urine sugar, urine ascorbic acid (A.A.) and blood glucose levels (BG) were measured in 21, 560 examinees, 781 of whom also had 75 g GTT, within one month. The results were as follows:
1) Except for cases under treatment for diabetes mellitus, the prevalence rate of glycosuria was 6.9% in males and 2.2% in females. The rate was almost the same between these with urinary A. A.≥20 mg/dl and these with a value <10 mg/dl. It is thus unnecessary to consider the influence of urinary A. A. levels at the time of mass survey for diabetes mellitus using the urine sugar test.
2) Using the results of the distribution of BG levels in the non-glycosuric group, standard BG levels at screening for diabetes mellitus were 130 mg/dl at 0.5-1 hr., 120 mg/dl at 1.5-2.0 hrs., and 110 mg/dl at 2.5-3.0 hrs. after eating, and 100 mg/dl at fasting or over 3.5 hrs. after eating.
3) The false negative rate was 0.3% in males and 2.0% in females when the screening test was conducted using only random urine sugar, but using random BG levels according to the above mentioned standard BG levels, the rates were 0 % and 1.4% in the two groups.
4) Though the screening test using urine sugar after eating was a good method, the random blood sugar test including fasting was more effective.
