Li, Zhen
(2015)
Deriving an insulin resistance syndrome score in youth with Type I Diabetes Mellitus based on clinical risk factors.
Master's Thesis, University of Pittsburgh.
(Unpublished)
Abstract
Diabetes is a complicated chronic disease, and it is categorized into type 1 diabetes (T1D) and type 2 diabetes (T2D). Insulin sensitivity (IS) is lower in adults and adolescents with T1D compared to normal people, and a lower IS in T1D has been showed to be associated with longer-term complications. T1D is prevalent in children. The aim of this project was derive an insulin sensitivity (IS) score in children with T1D using noninvasive clinical predictors.
From a sample of 60 children undergoing a euglycaemic-hyperinsulinaemic clamp study at Children hospital of Pittsburgh of UPMC, a linear regression model was derived using clinical and laboratory measurements to predict insulin sensitivity. Because of the limitations of the small dataset, overfitting was an issue. We used a machine learning technique called Cross-Validation to help select predictors and to assess the performance. Data management and analysis were done using SAS 9.4.
One set of models built with only clinical variables were called clinical models. The other models used both clinical and laboratory variables were called research models. Two different outcome variables measures IS, glucose disposal rate (GDR) and glucose disposal rate (GDR) divided by free insulin, were used. After selecting the best models and checking the assumptions, the best model to predict GDR contained diastolic blood pressure percentile, systolic blood pressure percentile, gender, waist circumference, and diabetes duration. When the dependent variable was GDR divided by free insulin, predictors in the best model included DBP percentile, HbA1C at the study time, waist circumference, leptin, and adiponectin/ leptin. These models had much better performance for type 1 diabetes than these models from the literature.
Public Health Significance: Identifying an IS predictive model based on routinely gathered clinical measurements and laboratory value is a valuable alternative to the invasive euglycaemic-hyperinsulinaemic clamp study. The current gold standard of insulin sensitivity, euglycaemic-hyperinsulinaemic clamp, is an invasive intravenous study requiring fasting overnight hospital study. The model makes it practical to use in epidemiological and screening studies.
Share
| Citation/Export: |
|
| Social Networking: |
|
Details
| Item Type: |
University of Pittsburgh ETD
|
| Status: |
Unpublished |
| Creators/Authors: |
|
| ETD Committee: |
|
| Date: |
29 June 2015 |
| Date Type: |
Publication |
| Defense Date: |
20 April 2015 |
| Approval Date: |
29 June 2015 |
| Submission Date: |
9 April 2015 |
| Access Restriction: |
No restriction; Release the ETD for access worldwide immediately. |
| Number of Pages: |
118 |
| Institution: |
University of Pittsburgh |
| Schools and Programs: |
School of Public Health > Biostatistics |
| Degree: |
MS - Master of Science |
| Thesis Type: |
Master's Thesis |
| Refereed: |
Yes |
| Uncontrolled Keywords: |
Insulin resistance; Clinical risk factors; Type I diabetes; Juvenile |
| Date Deposited: |
29 Jun 2015 14:01 |
| Last Modified: |
19 Dec 2016 14:42 |
| URI: |
http://d-scholarship.pitt.edu/id/eprint/24701 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[img]](http://d-scholarship.pitt.edu/style/images/fileicons/application_pdf.png)
![[feed]](/images/feed-icon-32x32.png){kind=icon}