The CGM, food intake data, Internet of Things technology, and machine learning algorithms in the Twin Precision Treatment Program optimized combinations of nutrients and provided nutritional guidance to type 2 diabetes patients that helped them consume foods that do not produce glucose spikes and avoid foods that cause blood glucose spikes.
The program participants in the current study reduced mean systolic blood pressure by 7.5% (p < 0.0001) and reduced diastolic blood pressure by 4.3% (p = 0.002) from baseline over the course of the 90-day study period. The TPT Program recommended patient-specific meal plans that depended on the likes and dislikes of the patient and that were balanced across macro, micro, and biota nutrients to reduce glucotoxicity and lipotoxicity. This helped to heal inflammation and may be a reason for the improved blood pressure. Patients were also provided with supplements to ensure sufficient micronutrients were consumed. Nutritional, activity, and sleep counselling were provided by trained health coaches through the app and via telephone. Additionally, the digital twin technology enabled precise management of nutrition, activity, and sleep and helped the coach focus on the most important lifestyle variables for that patient for the improvement of blood pressure. Daily home blood pressure monitoring was done by the patient, and measurements were transmitted through Bluetooth-enabled equipment to the patient’s mobile app and to the platform and were used by the coaches to measure the impact of these interventions. Hence a feedback loop was established which enabled the efficient reduction of blood pressure.
Improvements over the 90 days were also seen in BMI (6.2% reduction, p < 0.0001). BMI reductions may have stemmed from the nutritional interventions, increases in physical activity (mean (SD) steps taken per day increased from 4677.4 (2804.9) at baseline to 7004.1 (3999.0) at 90 days)26, the 63% reduction in fasting insulin level, and the reduction in average homeostatic model assessment of insulin resistance (HOMA-IR) from 7.4 (3.5) at baseline to 3.1 (2.5) at 30 days and 3.2 (2.8) at 90 days26. Additionally, the percent of patients taking antihypertensive medications decreased from 35.9 to 4.7% (p < 0.0001). For most patients, glycemic variability was stabilized during the first week of the program, and %CV, LBGI, and HBGI were maintained within normal or low thresholds over the 90 days.
Hypertension is found in more than half of patients with diabetes. Hypertension increases diabetes patients’ risk of micro- and macrovascular disease and chronic kidney disease and increases costs. Diabetes and hypertension both have insulin resistance in common, and both improve with lifestyle intervention. Controlling blood pressure in patients with diabetes prevents and delays micro- and macrovascular complications30,31. Specifically, among patients with type 2 diabetes, each 10-mm Hg reduction in blood pressure has been associated with improved mortality, reductions in cardiovascular events, coronary heart disease, stroke, albuminuria, and retinopathy32.
The literature suggests that blood pressure goals of systolic/diastolic blood pressure < 130/80 mm Hg are rarely attained in patients with diabetes and that treatments employing at least two medications are needed for most patients31. In the present study, patients reduced their mean systolic blood pressure from 134.7 to 124.6 mm Hg and reduced mean diastolic blood pressure from 83.9 to 80.3 mm Hg at 90 days, with less than 5% of patients taking antihypertensive medication by the end of the study. After 10 weeks of intensive nutrition, behavioural counselling, digital coaching, and medication management, McKenzie et al. found similar blood pressure reductions in a much more obese population (a reduction from 132 to 125 mm Hg in systolic blood pressure and a reduction from 82 to 78 mm Hg in diastolic blood pressure)1.
Monitoring BMI in patients with type 2 diabetes is important, as increased BMI is associated with decreased life expectancy33. Prior studies have reported some improvements in BMI after certain types of intervention. In the intensive nutritional study by McKenzie et al., BMI decreased 7.2% in a population of morbidly obese patients with type 2 diabetes1. After 12 weeks of low-carbohydrate or low-glycemic diets, type 2 diabetes patients in another study22 reduced BMI from 37.8 to 34.4 (9.0%) and from 37.9 to 36.5 (3.7%), respectively. Over a similar time period, the present study found a 6.2% reduction in the BMI of patients whose initial mean BMI was 29.
Limited interventional studies were found that reported %CV, LBGI, and HBGI results. In a study by Ohara et al. of normal treatment of type 2 diabetes patients with mean (SD) duration of diabetes of 11.6 (9.2) years, mean %CV began at 24.3% and decreased to 21.7% at 24 weeks34. In patients with mean duration of diabetes of 9.0 (6.0) years, Inzucchi et al. reported an increase in %CV from 21.0 to 25.7% and a decrease in HBGI from 15.4 to 5.3 in a 24-week study of type 2 diabetes patients who received an intensification of their insulin treatment35. In patients with duration of diabetes of 8.43 (6.52) years who participated in the TPT Program, the current study found that %CV over the 90-day program was substantially lower at 17.3%.
In a study of 549 patients with type 2 diabetes taking insulin or oral hypoglycemic medications, McCall et al. found that mean daily LBGI and HBGI values over weeks 8 to 26 of the study were approximately 0.5 (minimal range) and 5.5 (moderate range), respectively15. In the present study, mean LBGI and HBGI over days 1 to 90 were 1.37 (low range) and 2.13 (low range), respectively, and most patients were able to eliminate hypoglycemic medication use over that time26. As reported previously, all 12 patients who had been taking insulin were able to discontinue its use, two-thirds of the patients on metformin discontinued taking metformin, and almost every patient on other oral hypoglycemic medications was able to stop taking them26. Other metrics for this population, such as HbA1c, fasting glucose, time in range, and weight, were reported previously26.
No cases of diabetic ketoacidosis and no episodes of symptomatic hypoglycemia were observed in the current study. Additionally, no new cases of gout were reported. The overall average patient-reported program satisfaction score was 4.4 out of 5.
The strengths of this study include significant reductions in BMI and blood pressure and the maintenance of %CV, LBGI, and HBGI below low thresholds, while eliminating antihypertensive and hypoglycemic medication use in nearly all patients. The out-patient nutritional intervention employed home-cooked foods, making the intervention sustainable. Accurate assessment of nutrition was aided by the patients’ app-based logging of food.
Limitations of this study include its smaller study population, retrospective nature, and the lack of a control group. Additionally, increasing the length of the program will be a goal of subsequent studies.