The author utilizes his research results accumulated since 2015 to summarize some key points regarding type 2 diabetes (T2D) control. This paper is aimed at family clinical practice and public health via lifestyle medicine. The approaches and formulas outlined in this article are based on his collected ~2 million data of his medical and health conditions over a period of 6 years from 7/1/2015 through 6/30/2021. He utilized his developed 4 prediction tools for various basic biomarkers for T2D patients, including body weight, fasting plasma glucose (FPG), postprandial plasma glucose (PPG), and HbA1C. All of the predicted results are then compared against his measured 4 biomarkers during the same time period. His research methodology is based on his developed GH-Method: math-physical medicine approach instead of the traditional biochemical medicine approach. Of course, all of the math-physical medicine derived results have quantitative proof and reliable support from biochemical medicine viewpoints.
In summary, either on a daily basis or a longer time period, all of the predicted biomarker data curves versus the measured biomarker data curves have extremely high correlation coefficients, moving up and down in unison, and high prediction accuracy, where the two datasets have almost identical average results. The following table summarizes the correlation coefficients and prediction accuracies in the format of (Correlation; Accuracy):
Weight : (87%; 99%)
FPG : (99.8%; 100%)
PPG : (88%; 99.8%)
Daily eAG : (91%; 99.8%)
These results have proven that the Prediction models are highly accurate with the ending average results as well as the moving patterns of data curves. For the author himself, who had severe T2D without any diabetic medication interventions from 12/8/2015 to 6/30/2021, this set of prediction tools has demonstrated the usefulness and effectiveness on his T2D control. Therefore, other diabetes patients can also confidently utilize these tools to manage their conditions.