iglu: an R package for CGM Data

Header row

Interpreting Blood Glucose Data in R with iglu
Steve Broll, Irina Gaynanova stevebroll@tamu.edu github.com/stevebroll/iglu

Body row

Introduction

Type 2 Diabetes

85-90% of Diabetes cases
Insulin resistance, high blood glucose levels
Increased risks for many diseases/conditions

Treatment

Diet and exercise, possibly insulin therapy
Monitoring blood glucose levels
Can use Continuous Glucose Monitors (CGMs)

CGM Profiles

Blood glucose levels every 5-15 minutes
Normal blood glucose range is 70-120 mg/dl
> 170 mg/dl - Hyperglycemia

Metrics

Many metrics used to assess blood glucose levels
No consensus for best metric, most are correlated

Goals of iglu

Provide functions for all relevant metrics
User friendly documentation and function names
Default outputs produced by 1 line of code

Workflow – Single Subject

# load example data (single)
library(iglu); data("example_data_1_subject")
above_percent(example_data_1_subject)

	above_140	above_180	above_200	above_250
Subject 1	26.14065	8.198971	3.396226	0.3773585

# accepts user input of targets
above_percent(example_data_1_subject, targets = c(100,150,200))

	above_100	above_150	above_200
Subject 1	72.72727	20.06861	3.396226

# Continuous Overall Net Glycemic Action
conga(example_data_1_subject)

	conga
Subject 1	36.86259

# Mean Amplitude of Glucose Excursion
mage(example_data_1_subject)

	mage
Subject 1	53.36908

# Upper plot in "Visualizations"
plot_glu(example_data_1_subject)

Workflow – Multiple Subjects

# load example data (multiple)
data("example_data_5_subject")
in_range_percent(example_data_5_subject)

	in_range_80_200	in_range_70_180	in_range_70_140
Subject 1	96.02058	91.66381	73.722127
Subject 2	42.55921	26.44044	3.358077
Subject 3	85.19243	81.34377	49.836921
Subject 4	96.45197	95.11463	67.740175
Subject 5	69.57265	62.11966	30.119658

# Mean difference between glucose values
# obtained at the same time of day 
modd(example_data_5_subject, lag = 2)

	modd
Subject 1	27.88716
Subject 2	44.05254
Subject 3	48.47412
Subject 4	24.87520
Subject 5	61.36327

# Lower two plots in "Visualizations"
plot_glu(example_data_5_subject)

Visualizations

Footer row

Continuous Glucose Monitor

Future Updates

Fleshed out documentation (above_percent as example)
Input catches and more advanced automatic testing
Unit Conversion function
Rate of Change calculation function
Missing metrics- SD subtypes
More versions of lasagna plots in plot_glu

Repository QR Code

Acknowledgements

I am grateful to Dr. Irina Gaynanova for her mentoring during the project and for providing some of the code used in development of the package.
I am supported by the Texas A&M University Department of Statistics.