Figure - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Source publication
Glycemia assessment in people with type 1 diabetes (T1D) has focused on the time spent in different glucose ranges. As this time reflects the relative contributions to the finite duration of a day, it should be treated as compositional data (CoDa) that can be applied to T1D data. Previous works presented a tool for the individual categorization of...
Contexts in source publication
Context 1
... use data of 30 weeks, approximately, recollected in different periods between 2020 and 2022. The demographic characteristics of patients are presented in Table 1. All patients provided written informed consent to participate in this study. ...Context 2
... 2 shows the total number of days analyzed, sensors, and valid periods of 24 h and consecutive 6 h periods for each patient. 1 90 Dexcom G6 85 83 83 84 2 90 MiniMed 640G 54 51 54 53 3 226 MiniMed 640G 57 44 62 56 4 90 Dexcom G6 74 81 81 75 5 134 Dexcom G6 83 82 85 81 6 232 Dexcom G6 76 68 72 73 7 115 MiniMed 640G 80 79 71 81 8 556 FreeStyle Libre 229 231 232 227 2.2. CoDa A composition is a vector X = (x 1 , x 2 , . . . ...Context 3
... example of specific results of these measures for patient 1 (P1) can be observed in Tables A1 and A2 of the Appendix A. ...Context 4
... compositional geometric mean is calculated for each of these groups (Equation (5)). This vector provides a quantitative interpretation of each of the groups (Table A1). The 24 h and 6 h periods were qualitatively characterized in terms of the relative time spent in the different glucose ranges, according to the log-ratio approach. ...Context 5
... the 24 h and 6 h periods were classified taking into account the standardized metrics [8] where the following glucose targets are pursued: <54 mg/dL (<1%), 54-70 mg/dL (<4%), 70-180 mg/dL (>70%), 180-250 mg/dL (<20%) and >250 mg/dL (<5%). If we analyze the example shown in (Table A1) we can see that group 1 has an average of 2.61% in hypo level 1, 87.38% in TIR and 10% in hyper level 1. In view of these data, we assign classification A, which we will qualitatively define as periods with a moderate percentage of hypo and hyper level 1. ...Context 6
... us consider the patient at 18 h when they have been categorized with four clusters. First, we analyze the glucose composition of the previous 24 h period using Table A1 (Compositional center of each group of 24 h periods) and verify that this period is categorized as type D (86.47% in normoglycemia and 13.52% in hyperglycemia, no hypos observed). Then the probability that the category of the next 6 h period (from 18 h to 0 h) is of type D 75% can be known. ...Context 7
... the probability that the category of the next 6 h period (from 18 h to 0 h) is of type D 75% can be known. Table A1 (Compositional center of each group of 6 h periods) demonstrates how group D was characterized by having 83.28% time in normoglycemia and 16.72% in hyperglycemia. In other words, P1 is expected to continue in normoglycemia with a tendency to hyperglycemic excursions for the next 6 h. the evaluation resulted in higher accuracy compared to the evaluation with 5 clusters. ...Citations
... Compositional data (CoDa) are data that transmit information about the parts of a whole expressed in proportions or percentages, as is the case of the vector of daily times in each of the glucose ranges: time below range (TBR) (<70 mg/dL), TIR (70-180 mg/dL), and time above range (TAR) (>180 mg/dL) [14], where all the components are positive and of constant sum. Previous studies have treated the percentage of time in the glucose range as a composition, yielding favorable outcomes, and this variable is of paramount importance in this field [15][16][17]. Furthermore, regression models have demonstrated favorable results overall, both in scalar variables and CoDa, due to their simplicity of implementation and robustness in prediction outcomes. ...
... The compositional input could contain zeros if some of the parts of the CoDa vector were zero; therefore, a pre-treatment was done because CoDa is based on log-ratios of parts. The detection matrix (dL) used in the imputation of the zeros was interpreted as in [17], taking into account the consecutive zeros. In this case, where we are only analyzing three parts, there could only be two consecutive zeros; the dL value will then be calculated by dividing 5 min (sensor measurement interval) by 120 min, which is the time analyzed from the previous 2 h, dL = 0.04166. ...
... DSSs have proven to be useful tools for patients and physicians [2,46,47]. Although glucose profiles have been treated as CoDa vectors in previous studies [15][16][17], there is no application in this branch of mathematics that is focused on predicting the mean and the CV as an information system or DSS tool for patients with T1D at specific hours of the day oriented to wide PH (2 h and 4 h). In this work, CoDa variables and transformed scalars have been used to predict the mean and CV of glucose in patients with T1D. ...
This paper presents an individualized multiple linear regression model based on compositional data where we predict the mean and coefficient of variation of blood glucose in individuals with type 1 diabetes for the long-term (2 and 4 h). From these predictions, we estimate the minimum and maximum glucose values to provide future glycemic status. The proposed methodology has been validated using a dataset of 226 real adult patients with type 1 diabetes (Replace BG (NCT02258373)). The obtained results show a median balanced accuracy and sensitivity of over 90% and 80%, respectively. A information system has been implemented and validated to update patients on their glycemic status and associated risks for the next few hours.
