Screening Methods for Diabetes Mellitus in Aged Group

Abstract

There are many methods of screening for diabetes mellitus, for example, blood glucose test, urine sugar test, hemoglobin A_1C, fructosamine etc. For the purpose of more efficient screening of diabetes mellitus in elderly groups, these data were analyzed to evaluate validity and ability to estimate prognosis. 1) Validity for screening based on diabetic type was analyzed in 27, 074 cases. For screening with the standard level of fasting plasma glucose (FPG)≥120mg/dl, sensitivity was 70.5%, specificity, 94.5%, predictive value, 89.0%, hemoglobin A_1C≥6.5%, 62.7%, 91.2%, 81.5%, respectively, and fructosamine≥290μmol/L, 54.1%, 92.3%, 78.1%, respectivley. FPG had the highest validity. Comparison of validity between the≥65-yr group and the ≤55-yr group was mode. There was significant difference in specificity but sensitivity and predictive value were lower in the ≥65-yr group. 2) Mean blood glucose levels (BG)±S.D. by time after meal were studied. Fasting BG was 84.8±9.8 mg/dl, 0.5-1hr. BG after meal, 100.8±24.5mg/dl, and 4.5hr or more, 84.1±12.8mg/dl. Based on this data, standard levels for screening based on diabetic type using random blood glucose levels for 0.5-1hr BG after meals were ≥130mg/dl, 1.5-2hr BG≥120mg/dl, and 3.5hr or more BG and FBG≥100mg/dl. With screening using this standard, there is no difference in sensitivity according to age, but sensitivity of FBG≥100mg/dl was the highest, the rate being 93.3% in the ≤64-yr group, 83.3% in the ≥65-yr group, and the predictive values of positive tests were 62.2% and 55.6% respectively. 3) Criteria of diabetes in the elderly groups were studied using GTT follow-up and death data. The rate of development of diabetes with FPG≥140mg/dl from diabetic type with FPG<140mg/dl was lower in the ≥65-yr group than in the ≤49-yr group. Comparision of mean age at death among all dead cases, the onset age of diabetes mellitus ≤64-yr group, and the ≥65-yr group was made in cases with age at death ≥65yrs. No difference was found between all dead cases and the group of the onset age ≥65yrs (80.4yrs), but in the group of onset age ≤64yrs, the mean age at death was 73.0yrs, significantly lower than for the overall mean age at death. Mortality rate by FPG was compared between the group of onset age of diabetes mellitus ≤64 yrs and that of ≥65yrs in males. In the former group, mortality rate increased in cases with FPG≥170mg/dl, but in the latter, the rate in creased only with FPG≥200mg/dl. From these data, screening standards in the elderly group were FPG≥140mg/dl and FBG≥120mg/dl. If random blood glucose levels are used as a preliminary screening, the standards levels for 0.5-1hr BG after meal are≥130mg/dl (PG≥150mg/dl), 1.5-2 hr BG≥120mg/dl (PG≥140mg/dl), and 3.5hr or more BG and FBG≥100mg/dl (PG≥120mg/dl).