Using Regression Analysis Models to Explore the Possible Hidden Linkage between Finger-Pierced Fasting Plasma Glucose and Three Blood Pressure Components over an Approximate 8-Year Period for a type 2 Diabetes Patient based on GH-Method: Math-Physical Medicine (No. 552)

Abstract

When the author woke up on 11/14/2021 around 07:20AM, he felt lightheaded, nauseated, and had cold-sweat. Being a long-time type 2 diabetes (T2D) patient of over 27 years, he immediately noticed and guessed that it could be a symptom of hypoglycemia or low blood sugar. However, his continuous glucose monitoring (CGM) sensor device showed a normal glucose reading of 102 mg/dL. He then measured his blood pressure (BP) and was shocked to find the low readings for SBP/DBP/HR of 79/47/37. These extremely low readings are considered almost dangerous and in the abnormal range for the combined conditions of hypotension (low blood pressure) and bradycardia (slow heart rate). These kinds of low readings lasted for ~2 hours. This incident peaked his curiosity which caused an immediate interest to identify the relationship existing between glucose, especially fasting plasma glucose (FPG), and three blood pressure components systolic blood pressure (SBP), diastolic blood pressure (DBP), and heart rate (HR). He then started to explore and read several published medical papers regarding this specific subject. In this article, it describes the research efforts and findings by using his own biophysical data from the approximate 8-year period from 1/1/2014 to 11/18/2021. To reduce his data preparation workload for the following task of regression study, he further subdivided the daily data into 16 semi-annual periods with the corresponding biomarker results. In his data table, he labeled them as WY2014, CY2021, etc., where “W” indicates the “Warmer” semi-annual period from April 1st through September 30th of each year. Whereas “C” indicates the “Cooler” semi-annual period from October 1st of one year through March 31st of the next year. In this way, he can then reduce the observation data amount from 2,879 days to only 16 semi-annual periods. As a benefit of organizing the data this way, he could also observe whether ambient (weather) temperature has any influence on his BP and FPG. Of course, at the end of his research and for comparison purposes, he also conducted an additional time-domain analysis of the FPG and BP along with the regression analysis calculations of both correlation (R) and variance (R^2 or R square) based on the daily data of BP and FPG. Incidentally, the FPG values used in the study are finger-piercing FPG, not CGM sensor FPG, because he began utilizing a CGM device on 5/8/2018. In summary, the author’s time-domain analysis and space-domain regression analysis for exploring the possible relationships existing between FPG and three BP components provide the following five observations: (1) It appears that none of the three BP components, SBP, DBP, and HR individually, have a very strong correlation or significant contribution to FPG. FPG has shown two moderate correlations with DBP (R=44%) and HR (R=42%).