Aim: To assess the prognostic accuracy of the FINDRISC for undiagnosed type 2 diabetes mellitus (T2DM) and dysglycaemia among early middle-aged adults from vulnerable groups in a large-scale European cohort.
Methods: Participants were recruited from low-socioeconomic areas in high-income countries (HICs) (Belgium-Finland) and HICs under austerity measures (Greece-Spain) and from the overall population in low/middle-income countries (LMICs) (Bulgaria-Hungary).
Study population comprised of 2,116 parents of primary-school children from families identified at increased risk of T2DM. Sensitivity (Se), specificity (Sp), area under the receiver operating characteristic curves (AUC-ROCs) and the optimal cut-offs of FINDRISC were calculated. Results: The AUC-ROC for undiagnosed T2DM was 0.824 with optimal cut-off≥14 (Se=68%) for the total sample, 0.839 with optimal cut-off≥15 (Se=83.3%) for HICs, 0.794 with optimal cut-off≥12 (Se=83.3%) for HICs under austerity measures and 0.882 with optimal cut-off≥14 (Se=71.4%) for LMICs. The AUC-ROC for dysglycaemia was 0.663 with optimal cut-off≥12 (Se=58.3%) for the total sample, 0.656 with optimal cut-off≥12 (Se=54.5%) for HICs, 0.631 with optimal cut-off≥12 (Se=59.7%) for HICs under austerity measures and 0.735 with optimal cut-off≥11 (Se=72.7%) for LMICs.
Conclusion: FINDRISC can be applied for screening undiagnosed T2DM and dysglycaemia among vulnerable groups across Europe, considering the use of different cut-offs for each subpopulation.