I embarked on a personal journey to understand my glucose levels and the impact of diet and exercise on my health. My initial motivation was to get fitter and healthier after retirement, but I soon realized that my rising glucose levels were a cause for concern. I discovered that my lack of understanding of glycaemic index and load, as well as the interaction between carbohydrates, fibre, fat, and protein, was a significant challenge. This led me to explore the use of artificial intelligence (AI) as a tool to help me interpret the data from my continuous glucose monitor.
The continuous glucose monitor provided valuable insights into my glucose dynamics, showing how levels responded to meals, exercise, sleep, and stress. I learned that my initial approach of making small dietary adjustments was effective, but I needed to understand the underlying mechanisms to sustain change. AI played a crucial role in helping me interpret the data and make informed decisions about my diet.
In the first phase of my experiment, I found that choosing lower-carbohydrate options and substituting them for other foods was a sustainable approach. I never felt deprived, and my glucose levels became steadier with fewer sharp spikes and falls. Walking after meals proved particularly effective in bringing down glucose levels and maintaining a steady baseline.
The continuous glucose monitor provided immediate, concrete feedback, which is essential for building and sustaining new habits. I found that my average glucose level was gradually falling, and this motivated me to continue making changes. The data from the monitor was a powerful tool for understanding the impact of my choices and providing a tangible reward for my efforts.
However, in the second phase, I faced a harder lesson. I assumed that my metabolic health was stable, but I soon realized that my glucose levels were still within the prediabetic range. I had reintroduced more carbohydrates into my diet, and my HbA1c levels had risen again. This relapse highlighted the importance of continuous feedback and the need to interpret the data in the context of physiological and behavioural processes.
I found that exercise had a context-dependent effect on glucose levels, and understanding these variations was crucial. The second phase of my experiment showed that metabolic regulation is not static, and progress is not always linear. The same interventions can produce different effects at different times, and understanding and responding to this variability is essential.
In conclusion, my experience with the continuous glucose monitor and AI has been a valuable learning journey. It has helped me understand the impact of diet and exercise on my health and provided a way to narrow the gap between data and understanding, and between understanding and action. While professional dietary care is important, motivated individuals with access to personalized feedback and reliable explanatory tools may be able to identify and correct early metabolic drift more quickly than traditional referral pathways allow.
