GH-METHODS

Math-Physical Medicine

NO. 337

A detailed illustration on the prediction accuracy of postprandial plasma glucose value using a sample lunch meal via GH-Method: math-physical medicine

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

Abstract
This article is an extension of the author’s previous two research papers on a developed linear equation to predict postprandial plasma glucose (PPG) and using a simple, practical card as a highly accurate PPG prediction tool for type 2 diabetes (T2D) patients.

As his research methodology, the GH-Method: math-physical medicine (MPM) has been utilized repeatedly for the last ten years (Reference 1).

This article uses his lunch meal on 9/29/2020 as an example to illustrate step by step how he predicts his PPG with 100% accuracy in comparison with his finger-piercing measured PPG value.

The predicted PPG formula is based on the average fasting plasma glucose (FPG) for the past 4 months, carbs/sugar intake amount of the meal, post-meal walking steps after eating, and weather temperature on the same day, which is described as follows:

  • Predicted PPG =  0.97 * FPG + (carbs/sugar grams * 1.8) – (post-meal walking steps in thousand * 5) + ((weather temperature – 77)/5)*0.3

The predicted PPG using the formula above, prior to his first bite of his lunch on 9/29/2020, was 169.99 mg/dL.  His finger-piercing measured PPG at two hours after the first bite of lunch was 170 mg/dL.  The prediction accuracy was 100%.

The author highly recommends T2D patients to measure their FPG at least several times a quarter, in order to get a quarterly average FPG value, which can serve as their PPG baseline values.  By applying various methods mentioned in his reference section, the other three PPG input values such as carbs/sugar, exercise, weather temperature (this input element can be safely disregarded) can be estimated easily and quickly.  Therefore, they can utilize the formula-based predicted PPG or “linear equation” to control their overall diabetes conditions.

The described method for the predicted PPG formula along with the post-meal walking exercise and carbs/sugar intake amount can assist patients in controlling their diabetes without the painful and troublesome finger-piercing glucose measurements.  The author uses one random example of his lunch meal on 9/29/2020 to illustrate and explain step by step how he calculates his predicted PPG and how he has developed his glucose prediction APP tool on his iPhone.

The author has been measuring his glucoses for over 8 years with the finger-piercing technique, while spending 30,000 hours to study and research diabetes during the past decade.  This simple PPG prediction method is an example of his diligent and persistent research work.  By using one lunch meal as an example in his developed linear equation, he hopes that it can provide useful guidelines and motivate other T2D patients to take back their lives from this dreadful chronic disease, diabetes.

Introduction
This article is an extension of the author’s previous two research papers on a developed linear equation to predict postprandial plasma glucose (PPG) and using a simple, practical card as a highly accurate PPG prediction tool for type 2 diabetes (T2D) patients.

As his research methodology, the GH-Method: math-physical medicine (MPM) has been utilized repeatedly for the last ten years (Reference 1).

This article uses his lunch meal on 9/29/2020 as an example to illustrate step by step how he predicts his PPG with 100% accuracy in comparison with his finger-piercing measured PPG value.

Method
1. Background
To learn more about the GH-method: Math-physical medicine method (MPM) developed by the author, readers can review his article, Biomedical research methodology based on GH-Method: math-physical medicine (No. 310).  In addition, the outlined history of his chronic diseases, diabetes research, and application tools development can be found in another article, Glucose trend pattern analysis and progressive behavior modification of a T2D patient using GH-Method: math-physical medicine (No. 305).

2. History of Diseases
The author is a 73-year-old male who has a long history of three severe chronic diseases for 25 years.  In addition, he experienced five cardiac episodes during 1994 – 2008 and was diagnosed with an acute renal problem in early 2010.  He also suffered from foot ulcer, bladder infection, diabetic retinopathy, and hypothyroidism.  He weighed 220 lbs. in 2000 and his HbA1C level was 10.0% in 2010.

This article describes his predicted and measured PPG from his lunch meal on 9/29/2020.  The PPG prediction utilized his developed simple linear equation and a simple glucose control card for PPG prediction based on above-mentioned GH-Method: MPM approach.

3. Diabetes Research
The author has spent the past 10 years to self-study and research metabolism, endocrinology, and diabetes.

Starting from 2015, he spent three consecutive years to discover the characteristics and behaviors of this complex “wild beast” of glucose.  His major objective is to truly understand the “inner characteristics” of glucose, not just using medications to control its “external symptoms”.  As a result, during this period of 3 years from 2015 to 2017, he developed four diabetes prediction models, which include Weight, PPG, FPG, and HbA1C to reach to his personal goal of understanding glucoses.

The author estimated and proved that PPG contributes approximately 75% to 80% towards HbA1C.  Therefore, he tried to unravel the mystery of PPG first.  Through his research, he has identified at least 19 influential factors associated with PPG formation.  Among those 19 influential factors, diet (carbs/sugar intake amount) would contribute ~39%, while exercise (he chose walking) would contribute ~41%.  In summary, these two primary influential factors add ~80% of the total PPG formation.  Among the rest of 17 secondary factors, the weather temperature contributes ~5%, stress and illness only make noticeable contributions when they occur.  But, when stress or illness occurs, its impact on glucose would supersede both carbs/sugar and exercise which are two daily required elements.

Two useful scientific discoveries were identified by using his collected big data of ~ 2 million data and trial-and-error analytics method.  First, one gram of carbs/sugar intake amount would increase 1.8 to 2.2 mg/dL of PPG level, depending on the variety of food factors.  Second, every one-thousand post-meal walking steps would decrease 5 mg/dL of PPG level.

For the past 4 years since 2016, his average carbs/sugar intake amount has been maintained around 15 grams or less, and his post-meal walking steps have been maintained at a rate of around 4,000 steps or more per meal.

From 2016 to 2017, he discovered a solid connection between FPG and weight with >90% of correlation.  In addition, similar to his PPG research, he also recognized that there are about five influential factors of FPG formation with weight alone contributing ~85% and cold weather temperature influencing ~5%.

Since July 2019, he launched his investigation based on the degree of damage to his pancreatic beta cells.  For over the past year of researching this subject, he noticed that his FPG has been decreasing in the past 6 to 8 years at an annual rate of 2.3% to 3.2% (average at 2.7% per year).  In other words, his pancreatic beta cells have been self-regenerating approximately 16% for the past 6 years and 27% for the past 10 years. Since there are no food intake and exercise while sleeping, he then thought about FPG as being a good indicator for the healthy state of his insulin functionality in the pancreatic beta cells.  He then formed a hypothesis that FPG carries an important message about the baseline status of his combined health state for both the liver and pancreas which also can be served as the “baseline for his PPG”.

In early 2015, he further utilized a total of eight identified influential factors with artificial intelligence (AI) technology and optical physics principle to develop an AI-based Glucometer APP to predict PPG.  This APP can automatically guesstimate the carbs/sugar amount from the meal photos he took, where each photo contained 20 million pixels with 160 million digits of information.  His AI predicted PPG values have reached to a 98.76% prediction accuracy based on his measured PPG values from his 5,838 accumulated meals data.

Furthermore, he utilized the observed food contents visually by his eyes, along with his acquired knowledge of food nutrition and his accumulated glucose research findings stored in his brain, to predict the PPG level.  He named this approach as the “natural intelligence” or NI approach.  His NI approach has achieved a slightly better prediction accuracy than his AI approach, with a prediction accuracy of 99.24% (0.48% more accurate than AI).  This NI results is also based on the same 5,838 accumulated meals data (Figure 1).  These findings have further proved his long-time hypothesis that human brains are smarter than computers, especially in cognition and judgement.

Results
In Figure 1, it shows the comparison results between NI versus measured PPG along with AI versus measured PPG during a period of 1,946 days with 5,838 meals.

Figure 1: PPG prediction using NI vs. measured PPG & AI vs. measured PPG

Figure 2 illustrates the nutritional contents of his lunch: beef soup-based meal that contained 4 tablespoons of pickled mustards and 7 fish balls imported from Korea on 9/29/2020. The food nutritional contents are included and it is best to focus on the carbohydrates and sugar amounts only for studying the glucose. 

Figure 2: Ingredients of lunch on 9/29/2020

These AI predicted PPG of 106.6 mg/dL, NI predicted PPG of 106.99 mg/dL, and his finger-piercing measured PPG of 107 mg/dL of this specific lunch meal are also shown on Figure 3.  

Figure 3: Meal Photo and Predicted PPG (both NI and AI) of the lunch on 9/29/2020

Next, he describes his developed “linear equation” of PPG prediction formula which uses the average FPG during the past 4 months, carbs/sugar intake amount of the meal, post-meal walking steps after eating, and weather temperature on that day, as follows:

  • Predicted PPG =  0.97 * FPG + (carbs/sugar grams * 1.8) – (post-meal walking steps in thousand * 5) + ((weather temperature – 77)/5)*0.3

Once he plugs his input data (see below) of his lunch meal on 9/29/2020, he can get the result of the predicted PPG very quickly.  The following demonstrates his step by step calculation using this linear equation.

  • Input data:
    4-month average FPG = 96.4 mg/dL
    Carbs/sugar intake amount = 20 grams
    Post-meal walking = 4,550 steps
    Weather on 9/29/2020 = 81-degree F
  • Predicted PPG Calculation steps:
    = (0.97 * 96.4) + (20 * 1.8) – (4550/1000 * 5) +((81-77)/5) * 0.3
    = (((93.5 + 36 – 22.25) + 0.24)
    = (129.50 – 22.75) + 0.24
    = (106.75 + 0.24)
    = 106.99

The numerical operations from above illustrate the following phenomena:

  • The PPG baseline for his body starts at 93.5 mg/dL (97% of his averaged FPG during the past 4 months).
  • After eating 20 grams of carbs/sugar from his lunch, this incremental amount of 36 pushes his PPG up to 129.5 mg/dL.
  • After lunch, his 4,450 walking steps, this reduction amount of 22.75 pushes his PPG down to 106.75 mg/dL.
  • The warmer weather temperature of 81-degree Fahrenheit increase slightly of 0.24 pushes up his PPG to 106.99 mg/dL.
  • His post-lunch PPG via finger-piercing and testing strip yields a reading of 107 mg/dL.

As a result, his predicted PPG value based on his developed linear equation is 169.99 mg/dL, whereas his finger-piercing measured PPG value is 170 mg/dL (without decimals) yields a 100% of prediction accuracy rate.

Conclusions
The author highly recommends T2D patients to measure their FPG at least several times a quarter, in order to get a quarterly average FPG value, which can serve as their PPG baseline values.  By applying various methods mentioned in his reference section, the other three PPG input values such as carbs/sugar, exercise, weather temperature (this input element can be safely disregarded) can be estimated easily and quickly.  Therefore, they can utilize the formula-based predicted PPG or “linear equation” to control their overall diabetes conditions.

The described method for the predicted PPG formula along with the post-meal walking exercise and carbs/sugar intake amount can assist patients in controlling their diabetes without the painful and troublesome finger-piercing glucose measurements.  The author uses one random example of his lunch meal on 9/29/2020 to illustrate and explain step by step how he calculates his predicted PPG and how he has developed his glucose prediction APP tool on his iPhone.

The author has been measuring his glucoses for over 8 years with the finger-piercing technique, while spending 30,000 hours to study and research diabetes during the past decade.  This simple PPG prediction method is an example of his diligent and persistent research work.  By using one lunch meal as an example in his developed linear equation, he hopes that it can provide useful guidelines and motivate other T2D patients to take back their lives from this dreadful chronic disease, diabetes.

References

  1. Hsu, Gerald C. eclaireMD Foundation, USA. No. 310: “Biomedical research methodology based on GH-Method: math-physical medicine”
  2. Hsu, Gerald C., eclaireMD Foundation, USA, No. 305: “Glucose trend pattern analysis and progressive behavior modification of a T2D patient using GH-Method: math-physical medicine”
  3. Hsu, Gerald C., eclaireMD Foundation, USA, No. 226: Community Medicine, Doctors without distance:   Using AI tools and Internet to assist T2D patients in remote rural areas  (GH-Method: math-physical medicine approach)”
  4. Hsu, Gerald C., eclaireMD Foundation, USA, No. 301: “A simple formula based on postprandial plasma glucose prediction using 5,640 meals data via GH-Method: math-physical medicine”
  5. Hsu, Gerald C., eclaireMD Foundation, USA, No. 264: “Community and Family Medicine via Doctors without distance:  Using a simple glucose control card to assist T2D patients in remote rural areas  (GH-Method: math-physical medicine)”
  6. Hsu, Gerald C. eclaireMD Foundation, USA. No. 97: “A simplified yet accurate linear equation of PPG prediction model for T2D patients using GH-Method: math-physical medicine”
  7. Hsu, Gerald C. eclaireMD Foundation, USA. No. 99: “Application of linear equation-based PPG prediction model for four T2D clinic cases using GH-Method: math-physical medicine”
  8. Hsu, Gerald C. eclaireMD Foundation, USA. No. 89: “Using GH-Method: Math-Physical Medicine to Conduct the Accuracy Comparison of Two different PPG Prediction Methods”
  9. Hsu, Gerald C. eclaireMD Foundation, USA. No. 106: “Accuracy of Predicted PPG by using AI Glucometer and GH-Method: math-physical medicine”
  10. Hsu, Gerald C. eclaireMD Foundation, USA. No. 234 Article “Travel, Weather, Meals”: “Using Math-Physical Medicine to Study Traveling vs both Metabolism and Glucose, Weather Temperature vs both FPG and PPG, Food & Meals vs PPG”