Endocrine Abstracts

Introduction: Lifestyle, exercise and diet vary hugely between summer and winter months population-wide. We compared diabetes control in children with type 1 diabetes (T1D), during summer and winter months.

Methods: All children under 18 years with T1D attending our diabetes clinics were eligible. Those diagnosed within one year, changed insulin regimen between seasons or with incomplete data were excluded. Summer or winter months were defined by British Summer Time [25 March 2018–28 October 2018] or Greenwich Mean Time [29 October 2018–31 March 2019] use respectively. Clinical data for the previous year (glycated haemoglobin (HbA1c), weight, insulin dose and regimen) and HbA1c matched Diasend downloads (average glucose level, number of tests/day, proportion of tests in euglycaemia, hypoglycaemia and hyperglycaemia) was collected retrospectively. HbA1c and Diasend downloads within first 6 weeks of the season were not included, to avoid overlap between seasons.

Results: All 197 patients were audited; 97 excluded as diagnosis within 1 year (60), non-T1D (8), change in insulin regimen (2), over 18 years (1) and incomplete clinic data (26). Patients (n=100) ranged from 3 to 17 years (mean=12.3, S.D.=3.4). Paired sample t-tests were used to compare data in summer and winter months. Mean HbA1c (mmol/mol) in summer (mean=63.6, S.D.=12.2) was slightly lower than in winter (mean=65.6, S.D.=13.3), but not significant (P=0.093). Diasend downloads (n=82) revealed a significantly high percentage of tests in euglycaemia range (summer mean=29.6, S.D.=13.3; winter mean=25.4, S.D.=13.6) (P=0.03). There was significantly lower insulin dose used (units/day) in summer, both for bolus insulin (summer mean=25.8, S.D.=12.5, winter mean=28.0, S.D.=13.2) (P=0.01) and basal insulin (summer mean=24.7, S.D.=14.3, winter mean=26.3, S.D.=14.6) (P=0.03). But no difference noted when total insulin used was adjusted for weight (units/kg/day) between summer (mean=0.9, S.D.=0.3) and winter (mean=0.9, S.D.=0.2) (P=0.94).

Conclusions: Overall we found lower mean HbA1c in summer (by 2 mmol/mol), but it was not significant. There was greater proportion of tests in euglycaemia range in summer, which may explain the slightly lower mean HbA1c. Insulin requirements appeared lower in summer, but no difference noted when adjusted for weight. However, multiple factors affecting glycaemic control were not individually studied.