DYNAMIC RISK MEASURES FOR SEVERE HYPOGLYCEMIA

点击翻译按钮获取中文摘要

DESCRIPTION: Among individuals with insulin dependent diabetes mellitus severe hypoglycemia (SH--stupor/unconsciousness) is triggered by a number of dynamically interacting biological and behavioral factors, that include, but are not limited to: failure to hormonally counterregulate, intensive therapy to normalize blood glucose (BG), hypoglycemia unawareness. The occurrence of SH is a stationary over time process--a history of SH and/or frequent low BG readings are followed by future SH. Understanding extreme BG fluctuations as a process in a specific time frame leads us to the dynamic risk models proposed in this application. Recent discussions with the National Science Foundation Center for Biological Timing suggest that, with some modifications, the investigators' current data collection provides an ideal opportunity to develop dynamic risk measures of SH involving the relationships between timing of the self-care behaviors of insulin injection, food consumption and exercise performance with low BG. Similarly, the investigator would be in the unique position to develop a dynamic model of the deterioration in behavior as BG falls and EEG changes. Phase I would involve the development of these models, subsequent computational methods, algorithms and software. The investigators have recently been funded to perform cross-sectional comparisons, both in the field and laboratory, of adults with Type I diabetes who either do or do not have a recent history of multiple SH. The intent is to determine what biological, psychological and behavioral factors differentiate these two groups. During Phase 2A they would modify their field data collection in order to test and verify two dynamic models of SH developed during Phase I. During Phase 2B, time series analyses of driving behavior and EEG data, collected in the investigators' laboratory study under conditions of euglycemia-induced hypoglycemia, will be performed. This proposed study would take advantage of these unique opportunities by supporting the personnel necessary to 1) develop such dynamic models and subsequent data analysis methods, algorithms and software, and 2) to test and validate the models with real data.