ABSTRACT
AIMS: To assess if a clinically orientated approach improves screening for latent autoimmune diabetes in adults (LADA) in patients developing diabetes over age 50. METHODS: From a clinic-based cohort of 3327 patients with Type 2 DM diagnosed over age 50 we recruited those with at least one feature suggestive of insulin deficiency: (i) fasting blood glucose > or = 15 mmol/l and/or HbA(1c) > or = 10% in spite of adequate compliance to diet and treatment; (ii) decreasing body weight > or = 10% in the previous 3 months in spite of constant diet; (iii) BMI < 25 mg/kg(2). A control group of 240 patients not presenting any of the previous criteria was randomly selected from the out-patient clinic. RESULTS: We identified 220 (6.6%) patients, of whom 70 were positive for glutamic acid decarboxylase antibodies (GADA) and/or islet cell antibodies (ICA), giving a prevalence of LADA of 31.8% (95% CI 25.7-38.4). In contrast, no patient randomly selected from the remaining cohort had marker positivities. With respect to patients negative for both ICA and GADA, those who were positive had lower C-peptide values (0.53 +/- 0.51 vs. 0.88 +/- 0.42 nmol/l, P < 0.001); the lowest levels were found in patients in whom both antibodies were positive. In linear regression analysis, variables independently associated with fasting C-peptide were GADA (beta = -0.25, P < 0.001), ICA (beta = -0.15, P = 0.04), BMI (beta = 0.03, P < 0.001) and duration of diabetes (beta = -0.02, P < 0.001). CONCLUSION: This study shows that: (i) a clinically orientated approach increases the efficiency of a screening programme for LADA, so that one in three screened patients are classified correctly; (ii) ICA and GADA positivity were negatively associated with residual beta-cell function, independent of BMI and duration of the disease; (iii) positivity for both ICA and GADA identifies patients with the lowest residual beta-cell function.
