GH-METHODS

Math-Physical Medicine

NO. 032

Using Math-Physical Medicine to Investigate the Relationship Between Weather and Glucose

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

Introduction
This paper describes the relationship between ambient temperatures (weather) and both fasting plasma glucose (FPG) and postprandial glucose (PPG).

Methods
The author collected a total of 6,408 data of 1,068 days (from 10/1/2015 to 9/3/2018), including weather temperature, weight, FPG, PPG.  He further utilized mathematics, physics, engineering, and artificial intelligence with ~400,000 data to develop two prediction models of FPG and PPG.

Results
The following conclusions are observed:

  1. In the author’s other papers, weight is the dominating factor of FPG (contribution 85%) and carbs/sugar and exercise are two dominating factors of PPG (combined contribution 81%). Temperature is the secondary factor (contribution ~10%) of both FPG and PPG.
  2. From the time-series analyses, temperature has negative correlations with weight (-27%) and FPG (-29%), respectively. This medium correlation coefficients are due to temperature’s secondary influential power.
  3. Spatial analysis of FPG vs. Weight shows a “skewed” strong linear relationship existing between them (i.e. higher weight relates to higher FPG). However, the relationship between temperature and both Weight and FPG are “flat” which means regardless of temperature changes, both FPG and Weight remain constant.  Again, this phenomenon is due to temperature’s secondary influential power.
  4. In the time-series analysis, temperature has no correlation with PPG at all (R = -0.3%). However, in the Predicted PPG chart, temperature contributes 8 mg/dL (7%) of total 116 mg/dL of Predicted PPG.  Temperature induced PPG curve has a high 65% correlation with the actual temperature curve.

Summary
Weather temperature indeed has influence on glucose (both FPG and PPG), although its influential power is only secondary.  These quantitative findings are obtained from about 500,000 data collection, various statistical analyses, and glucose formation interpretation.  The results will help healthcare professionals and diabetes patients understand how temperature affects FPG and PPG.

Figure 1: Time-series Analysis of Temperature vs. Weight, FPG
Figure 2: Spatial Analysis of Weight vs. FPG
Figure 3: Spatial Analysis of Temperature vs. Weight
Figure 4: Spatial Analysis of Temperature vs. FPG
Figure 5: Temperature vs. PPG
Temperature vs. temperature induced PPG
Three-years Weather Temperature
Figure 6: Spatial Analysis of Temperature vs. PPG