Aims/hypothesis <p>Screening for islet autoantibodies is an effective method for identifying individuals with pre-symptomatic (stage 1 and 2) type 1 diabetes. This approach offers a valuable opportunity for education and monitoring, which can help to reduce the severity of clinical manifestations at clinical onset (stage 3), including diabetic ketoacidosis. The aim of the study was to evaluate the progression to stage 3 and the incidence of diabetic ketoacidosis in relatives of individuals with type 1 diabetes screened and followed up at a single institution in Italy.</p> Methods <p>This was a single-centre observational study conducted at San Raffaele Hospital, Milan, Italy, within the international multisite TrialNet Natural History Study-Pathway to Prevention. First-degree (aged 1–45 years) and second-degree (aged 1–20 years) relatives were screened primarily for GADA, IAA and IA-2A. In the event of a positive result, subsequent testing was conducted for ICA and ZnT8A. Periodic autoantibody testing, metabolic monitoring and educational support were offered to all autoantibody-positive participants. Participants were screened between July 2005 and February 2020, with the latest update obtained between January 2023 and June 2024.</p> Results <p>In total, 4046 relatives were screened at a median (IQR) age of 17.6 (7.9–38.0) years. At first screening, 4.9% were found to be positive, with 3.1% having a single autoantibody and 1.8% multiple autoantibodies. Follow-up data were available for 78.5% of the participants, with a median (IQR) follow-up time of 9.9 (6.5–13.5) years. Progression to stage 3 was observed in 51 (1.6%) participants. Disease onset occurred in 0.4% of autoantibody-negative, 6.5% of single-positive and 43.1% of multiple-positive participants after a median (IQR) time of 7.8 (5.4–10.4), 7.9 (2.1–11.8) and 2.9 (0.9–6.5) years, respectively (<i>p</i>=0.012). The Kaplan–Meier survival free of clinical diabetes at 15 years was 99.5% (95% CI 99.1, 99.7), 87.3% (95% CI 74.4, 94.0) and 45.9% (95% CI 31.1, 59.6), respectively (<i>p</i>&lt;0.001). At the time of disease onset, no occurrences of diabetic ketoacidosis were documented. Median (IQR) HbA_1c was 64 (52–86) mmol/mol (8.0 [6.9–10.0]%) and median (IQR) venous pH at onset was 7.37 (7.35–7.39). Hospitalisation occurred in 22 paediatric participants, as part of standard practice for newly diagnosed patients at our institution aiming to provide disease education and insulin therapy optimisation.</p> Conclusions/interpretation <p>The early identification of individuals at risk for type 1 diabetes through a single-centre approach, combining autoantibody screening and regular monitoring, completely prevented diabetes-associated ketoacidosis at disease onset in relatives of individuals with type 1 diabetes.</p> Trial registration <p>ClinicalTrials.gov NCT00097292</p> Graphical Abstract <p></p>