Glucose Variability in Type 1 Diabetes (T1D) and Glycaemic Responses to Food Composition.

Overview

The study proposed to recruit approximately 435 children and young people who have T1D and who regularly use Dexcom continuous glucose monitoring (CGM). Recruitment was be via their local dietitian. The dietitian was asked to provide baseline information about the participants which will include demographic data and information on clinical data, treatment and anthropometrics. Participants will be asked to provide access to Dexcom CGM data throughout the period of recording. Participants were issued with, for seven consecutive days, two survey questionnaires, one in the morning at breakfast time and the other in the evening. The morning survey will include questions on the breakfast meal (including a photograph of the meal) and insulin dosage, similarly the evening survey will also include questions on diabetes management and food and fluid intake in addition to questions on activities all of which took place during the four-hour postprandial period. These data will be statistically described using univariate, bivariate and multivariate analysis.

This is a quantitative observational study. Dexcom CGM data was collected for the period recording. Retrospective CGM data was collected to assess glucose variability during the day and night time periods. The Participants were then asked to provide information for seven days about their breakfast meal and the four hour post meal period. This included information on the type of meal consumed, a photograph of the meal and their diabetes management for the meal for example the insulin dose amount and timing. For the four hour post meal period the participants were asked about any adjustments they had to make to their diabetes management for example the treatment of high or low glucose readings, any snacks they ate and any activities they took part in. They were also asked to describe their mood during the post meal period. Recruitment - Paediatric diabetes dietitians, working across the United Kingdom (UK), were enrolled to help recruit participants and become principal investigators (PI) for their site. As this was an observational study there was no power calculation. The sample size was calculated from the number of dietitians that were likely to be available to help with recruitment; this was estimated to be around 32 and was based on the number that had already been approached. It was felt that they would have capacity to recruit 13-14 participants and this would result in 435 participants. Methodology Baseline data In order to make comparisons between relevant variables and glucose levels, the following baseline data will be collected from the dietitians and sent to the chief investigator, along with the artificial identifier on the Excel spreadsheet as discussed earlier at the stage of recruitment of participants: Parent's email address Sex, date of birth and recent weight and height (for calculation of BMI and BMI centile) and date of when this was taken Date of diagnosis of T1D Last four HbA1c Total daily insulin dose (TDD) Insulin: carbohydrate ratios (ICR) and Insulin Sensitivity Factor (ISF) Current insulin regimen - including the type of insulin prescribed and if applicable type of insulin pump i.e. open or closed loop system. Glucose measurement data on interstitial glucose was be collected via Dexcom CGM. The Dexcom CGM data will be accessed by a research 'Clarity Clinic' with Dexcom CLARITY® Clinic Portal (Dexcom In, San Diego, California (CA), USA). The researcher was the administrator of this clinic. Once the local dietitian had recruited, the participant's parent's email address will be sent to the researcher along with the baseline information/data as described above. The researcher, as administrator of the Clarity Clinic account, then invited the participant, via email, to be added to the clinic. Once the invite has been accepted, it stood for the period of the recording i.e. until all the seven day breakfast recording period was completed. Once the participant had submitted their last questionnaire, they were removed from the Dexcom CLARITY® Clinic Portal. The participants were not asked to make changes to their diabetes management and their breakfast meal. The participants were asked to submit two daily questionnaires (a morning and post-meal questionnaire) for seven days about their breakfast meal and the postprandial period and email a photograph of the breakfast meal. The morning questionnaire was about the breakfast meal and included the following questions: Participant's artificial identifier Date (of when meal was taken) The total average daily insulin dose (both bolus and basal insulin) if known (participant will only be asked to provide this once) Food eaten at breakfast including portion size, type and brand Drinks taken at breakfast including volume, type and brand The total carbohydrate count of breakfast in grams The number of units of insulin taken with breakfast How many units were given for a correction dose The type of bolus (for those on an insulin pump) The time of when the insulin dose was administered The time that breakfast started and ended How the bolus was calculated The amount of breakfast that was consumed (all, ¾, ½, ¼, none) The post meal questionnaire included the following questions: Participant's artificial identifier Date (of when meal was taken) Details of any hypo's in the four-hour postprandial period (time, treatment) Details of correction doses given in the four-hour postprandial period (whether in response to Dexcom CGM, times and number of units if insulin) Details of snacks taken in the four-hour postprandial period (type and amount of carbohydrate, responses to Dexcom CGM, number of units of insulin) Details of drinks taken in the four-hour postprandial period (volume, type/brand, carbohydrate amount if applicable) Details of any other adjustments made to insulin and or carbohydrate Type and duration of all physical activities undertaken during the 4-hour post prandial period Details of insulin adjustments made for physical activities Details of any addition information e.g. mood, menstruating, illness, stressors Data analysis This will be a mix of univariate, bivariate and multivariate analysis as this is best suited to describing, summarising and visualising these data. Outputs will include the distribution of glucose levels post-breakfast to determine the spread and dispersion of the data.