## GH-METHODS

Math-Physical Medicine

### NO. 007

Using GH-Method: Math-Physics Medicine to Predict Fasting Plasma Glucose

Corresponding Author: Gerald C. Hsu, eclaireMD Foundation, USA.

Introduction
The author spent 8.5 years and 23,000 hours to research his diabetes conditions. Using the GH-Method: math-physical medicine approach, he developed a fasting plasma glucose (FPG) prediction model to improve his type 2 diabetes.

Methods
The author examined the correlation coefficients between FPG versus postprandial plasma glucose (PPG), carbs and sugar intake, and exercise amount but found all were very low. Finally, after struggling for nine months, due to his “out-of-box thinking” personality, he accidentally discovered that Weight is the major influential factor of FPG formation.

He applied energy theory of physics, engineering modeling, and statistical tools, such as time-series, spatial analysis, frequency domain, to construct this FPG prediction model.

Results
Based on ~31,000 data of 1,825 days, (1/1/2014 – 12/32/2018), he found 83% correlation between FPG and Weight. In the time series diagram, the FPG curve followed the Weight curve like its “twin”. In spatial analysis diagram of BMI vs. FPG (without time factor), there is a “quasi-linear” equation existing between two coordinates of weight and FPG: from point A (165, 98) to point B (185, 148). The stochastic (random) distribution of data has two clear “concentration bands” stretched from lower left corner toward upper right corner. The +/- 10% band covers 65% of total data and the +/- 20% band covers 94% of total data. Only the remaining 6% of total data are influenced by other five secondary factors, see Figure 1.

###### Figure 1: Weight and FPG

After capturing basic characteristics, he then developed a practical tool to predict each day’s FPG value based on weight. The final prediction accuracy is 99.99% with 99.6% correlation between predicted and actual FPG values, see Figure 2.

###### Figure 2: Predicted FPG vs. Measured FPG

Conclusion
Based on his research, the author developed two glucose prediction tools. He was able to reduce his FPG from 142 mg/dL in 2014 to 107 mg/dL in 2018 (with 12 lbs. weight reduction), daily glucose from 279 mg/dL in 2010 to 117 mg/dL in 2018, and A1C from 10% in 2010 to 6.6% in 2018.