Seasonal variations in glycemic control have long been observed in people with type 1 diabetes, typically with higher HbA1c values in winter and improved control in summer months 1. This has been attributed to changes in physical activity, dietary patterns and insulin sensitivity linked to environmental temperature and lifestyle rhythms 1-3. Studies using both HbA1c and continuous glucose monitoring (CGM) data have confirmed this trend, suggesting that fluctuating insulin requirements can represent a barrier to optimal glycemic outcomes 1-3. Automated insulin delivery (AID) systems continuously adapt basal insulin delivery based on glycemia, potentially mitigating such seasonal variability. Here, we analysed real-world data (RWD) from MiniMed 780G system users in Italy to assess whether glycemic control and insulin delivery parameters vary across seasons. Data from users of the MiniMed 780G system was extracted and analysed using a validated methodology as described by Van den Heuvel et al. 4, which has also been applied in other real-world analyses of CareLink Personal. The initial population included all MiniMed 780G users with Type 1 Diabetes in Italy who uploaded data to CareLink before the extraction date (March 2024) and provided consent for their data to be used for research purposes (~9300 users). Users were included if they met both of the following conditions as follows: (1) They activated the AHCL (Auto Mode) feature at least once. (2) They had at least 10 days of sensor glucose data in each of the four seasonal observation periods (Fall, Winter, Spring and Summer 2023), regardless of whether SmartGuard (auto mode) was on or off. After applying the inclusion criteria, the final analysed population consisted of 4801 users. The reduction was primarily driven by the requirement to have sufficient data (≥ 10 days) across all four seasons. Only data actively uploaded to CareLink was analysed. Any metrics or seasonal periods with missing or insufficient data were excluded entirely, not imputed or reconstructed. This approach is consistent with real-world evidence principles. Data were aggregated between December 2022 and November 2023. Seasons were defined as: winter (December–February), spring (March–May), summer (June–August) and autumn (September–November). Metrics of glycemic control and insulin delivery were summarised per user and averaged by season. The following parameters were analysed: mean sensor glucose (SG), standard deviation of SG, glucose management indicator (GMI), time in range (TIR; 70–180 mg/dL), time in tight range (TITR; 70–140 mg/dL), time below range (TBR; 180 mg/dL) and insulin delivery breakdown (autobasal, autocorrection, manual bolus). Users meeting international consensus targets (TIR > 70%, TBR 180 mg/dL) ranged narrowly from 23.0% to 23.6%. The proportion of users meeting consensus glycemic targets showed minimal variation: 69.9%–72.0% achieved TIR > 70%, and 67.7%–69.7% maintained GMI < 7%. Total daily insulin dose (TDD) showed no meaningful differences between seasons (43.3–45.0 U/day). The distribution of insulin delivery modes also remained stable, with autobasal accounting for ~43%, autocorrection boluses for ~16%–17% and manual boluses for ~40% of TDD. Carbohydrate intake was comparable across seasons (176–181 g/day). Historical evidence demonstrates a clear pattern of seasonal variation in glycemic control among people with type 1 diabetes, with HbA1c peaking in winter and improving in summer months 1. Similarly, mean sensor glucose has previously been shown to be higher in the cold seasons when compared to the warm seasons 2 This has been reported across diverse populations and climates. In contrast, the present real-world analysis of over 4800 MiniMed 780G users in Italy showed remarkable stability in CGM-derived glycemic metrics and insulin delivery patterns across all four seasons (Figure 1; Table 1). Mean glucose, GMI, TIR, TITR and TBR remained stable throughout the year, with values consistently meeting international consensus targets 5. Similarly, insulin dosing patterns, including autobasal, autocorrection and manual bolus, remained similar across seasons. By comparison, CGM data collected in Italy over a similar period (2022–2023) from children and young adults treated with MDI 3 revealed measurable seasonal fluctuations in glycemia, including pronounced changes during a summer heatwave (July 2023; which overlaps with our data). These findings highlight the impact of temperature variations on glucose control in people with type 1 diabetes using multiple daily injections. In this context, the stability of glycemic metrics observed in our MiniMed 780G cohort across all four seasons provides real-world evidence that the MiniMed 780G algorithm can maintain consistent control despite expected behavioural and environmental changes. These findings complement prior observations that the MiniMed 780G is capable of swiftly adapting to lifestyle changes, as observed before during and after Ramadan 6 as well as data showing stable glycemic control in or out of school in Italian children using the MiniMed 780G system 7. Interestingly, in our study TDD remained unchanged despite expected seasonal variations in food intake and physical activity. Similar observations have been reported in other real-world data, such as before and during Ramadan 6, where sudden lifestyle changes did not lead to differences in TDD or autocorrection percentages. One explanation is the algorithm's ability to adjust insulin in every 5 min, enabling more efficient insulin delivery. Consequently, good glycemic control is maintained not by increasing TDD, but through optimised distribution throughout the day. A similar pattern is observed in Petrovski et al. 8, where simplified versus fixed meal strategies resulted in identical TDD. Limitations of this study primarily relate to the constraints of the data available in CareLink Personal. Certain sociodemographic variables, such as gender, age and diabetes type, are self-reported due to privacy regulations and age is available only in grouped categories. In addition, detailed clinical and lifestyle information, including dietary intake and physical activity, was not collected, limiting the ability to explore individual-level drivers of glycemic variability. Despite adherence to real-world evidence principles, the potential impact of missing data on outcomes cannot be fully excluded. Furthermore, HbA1c was not available and GMI was used as a surrogate metric. The study also has important strengths, including a large real-world population derived from a well-documented data source, minimising selection bias and enhancing generalizability. Given the large sample size and in the context of existing literature, these findings provide evidence that the MiniMed 780G system can maintain stable glycemic outcomes across seasons. In this large real-world analysis of MiniMed 780G users in Italy, glycemic control and insulin delivery metrics remained stable across all seasons. In contrast to data from individuals on MDI 3, which demonstrate seasonal fluctuations, these findings indicate that the MiniMed 780G system can maintain consistent glycemic outcomes throughout the year and may help minimise the seasonal variability typically observed in type 1 diabetes. The authors have nothing to report. E.B. received research support from Abbott and advisory board and lecture fees from Medtronic, Roche and Sanofi. A.L. received fees for consultancy from Medtronic. J.B.-S., J.C., B.V., V.S., T.H. and O.C. are all currently employed by Medtronic. Author elects to not share data. The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/dom.70834.
Bosi et al. (Sun,) studied this question.