PREDICTIVE VALUE OF AGE-RELATED ACUTE INSULIN-RESPONSE TO GLUCOSE IN SUBJECTS AT RISK FOR TYPE-1 DIABETES - RESULTS OF A 6-YEAR FOLLOW-UP-STUDY FROM WEST-FRANCE

The acute insulin response to i.v. glucose (AIRG) was evaluated in 344 first-degree relatives of patients with Type 1 diabetes. In 318 relat ives aged 3 to 48 years without islet cell antibody and insulin autoan tibody, correlations (p < 0.0006) were found between age and fasting i nsulinaemia, fasting glycaemia, or AIRG, with a peak during puberty. A ssuming that these relatives without islet cells and insulin auto-anti bodies have a low risk of developping Type 1 diabetes, we provided a < < standard age-related chart >> for AIRG with a << low >> AIRG defined as a value below the 1st percentile for each pubertal stage. Using th ese cut-off points, predictive characteristics of a low AIRG for progr ession towards diabetes within 6 years were analysed. Four relatives d eveloped diabetes and one displayed impaired oral glucose tolerance. F our out of these 5 subjects had islet cell and insulin auto-antibodies , but the other one was negative for these markers. Three of these 5 s ubjects had low AIRG at entry (30, 24 and 1 months before diabetes, re spectively). The two others displayed a steady progressive decline (p < 0.02) of age-related during the follow-up before impaired oral gluco se tolerance and diabetes appeared (rate of decline: 15 mu U/ ml/year) . Thus, independently of the presence of Islet cell antibodies, the pr edictive value of a low age-related AIRG during the follow-up is great er than the single low AIRG at entry. Above all, the concomitant prese nce of both islet cell antibodies and low age-related AIRG during the follow-up strongly enhances the prediction of the disease, as compared to the presence of islet cell antibodies alone. Additionally, 4 islet cell antibody positive transient hyperglycaemic subjects without a fa mily history of diabetes were prospectively studied. They rapidely pro gressed to Type 1 diabetes (3-14 months) with a low initial AIRG (3/4 < 1st percentile; 1/4 < 3rd percentile at entry but falling below the 1st percentile 2 months before the onset of diabetes). Finally, in all the subjects who subsequently developed diabetes (5 relatives and 4 t ransiently hyperglycaemic subjects) and in whom the time-relationships between basal glycaemia, oral glucose tolerance and AIRG were analyse d in the short period preceeding the onset of the disease, we observed that some subjects were still glucose tolerant at the time of the fir st blunted AIRG but already displayed fasting blood glucose values at the upper part of the normal range (> 90th percentile of control subje cts). This suggests that the loss of AIRG, and even the slight rise of fasting glycaemia, may precede the impairment of oral glucose toleran ce. We conclude that the limited predictive value of islet cell antibo dies can be overcome by a several-step procedure. As the initial scree ning method, islet cell antibody testing with a low detection threshol d to maximize sensitivity, identifies individuals to which the followu p of age-related AIRG can then be applied to strongly improve the pred ictive value. In subjects with a low age-related AIRG, determinations of fasting glycaemia and oral glucose tolerance could then indicate a worsening of the course and aid to predict the time to development of the disease.