The results of this real-world, retrospective analysis of 1853 patients enrolled in a DT commercial program for T2D over one year demonstrate significant clinical improvements in multiple parameters. A noteworthy finding is that 89.0% of participants achieved an HbA1c level of less than 7%. This reduction in HbA1c, along with substantial decreases in medication usage and improvements in metabolic health markers, aligns well with our study’s hypothesis and objectives. These results are particularly significant as achieving HbA1c levels below 7% is a widely recognized goal in T2D management, indicative of a lower risk of T2D-related complications.5 A decrease of 1% in HbA1c levels was linked with a 37% lower risk of microvascular complications and a 21% reduction in the risk of any diabetes-related endpoint or death.5
In evaluating the effectiveness of the DT intervention for managing T2D, the current study demonstrated an HbA1c reduction of 1.8%, closely aligning with the 1.9% decrease observed after 90 days in a previous study.3 Additionally, the DT intervention in a randomized controlled trial (RCT) reported a more substantial 2.9% reduction in HbA1c levels.4 These results suggest that while real-world applications of the DT intervention consistently yield significant reductions in HbA1c, the more controlled conditions of an RCT can facilitate even greater improvements. When comparing the DT intervention to other diabetes management strategies, it becomes clear that the DT approach offers superior outcomes. For example, the Virta Health program, employing a ketogenic diet approach, reported a 1.3% decrease in HbA1c levels.22 However, the outcomes that Virta achieved were in large part due to the restrictive ketogenic diet versus the liberal and personalized diet used in our program. The Look AHEAD trial, focusing on intensive lifestyle interventions, observed a 0.7% reduction in HbA1c levels and 72.7% of participants achieving an HbA1c below 7%.23 The Diabetes Remission Clinical Trial (DiRECT) trial, utilizing Total Diet Replacement (TDR), achieved a 0.9% reduction in HbA1c levels.24 The digital intervention based on the Low-Carb Program reported HbA1c reductions of 1.2%,25 the Weight Watchers program reported HbA1c reductions of 0.75%,26 and the Why WAIT program, utilizing liquid meal replacements, reported that 21.6% of their participants had an HbA1c less than 6.5%.27
This comparative analysis offers insights into HbA1c reductions across various T2D management interventions, underscoring the varied effectiveness of these strategies. The DT intervention's success can be attributed to its unique integration of advanced technology and personalized health coaching. The CGM data, combined with machine learning algorithms, provided real-time, individualized dietary and lifestyle recommendations. This level of personalization, coupled with regular behavioral nudges and human coaching, facilitated sustained engagement and adherence to healthier habits.3,4 This holistic and dynamic approach to T2D management addresses the multifaceted nature of T2D and offers a more comprehensive solution compared to traditional interventions.
Multivariate regression analysis was performed to overcome selection bias by adjusting for key baseline covariates and to identify factors influencing HbA1c changes in participants of the DT intervention. The analysis reveals key factors affecting HbA1c changes. Older age is linked to slightly higher HbA1c levels at 1 year, indicating that older adults may need more tailored support. Gender differences are minimal, with males showing marginally lower HbA1c levels, suggesting a potential need for gender-specific adjustments. Longer diabetes duration correlates with higher HbA1c levels, highlighting the challenge of managing long-term diabetes and the need for early intervention. Higher baseline HbA1c strongly predicts higher levels at 1 year, emphasizing the challenge of managing patients with high initial HbA1c. Baseline BMI and HOMA2-IR are not significantly related to HbA1c changes, suggesting the DT intervention's broad applicability across different BMI and insulin resistance profiles. Better baseline β-cell function (HOMA2-B) is associated with lower HbA1c, highlighting the importance of β-cell health. Overall, the DT intervention effectively reduces HbA1c levels underscoring the importance of personalization and early intervention. This analysis helps minimize selection bias by adjusting for baseline differences, attributing HbA1c changes more accurately to the DT intervention.
Technological nudges are effective for behavioral change. Research shows that reminder systems in mobile apps can boost medication adherence in T2D patients.28 However, the impact of app interventions on T2D management is mixed. While some report notable HbA1c improvements, challenges include technological advancements outpacing regulations, limited proof of effectiveness, and user barriers such as technical difficulties and lack of awareness.28 The DT app is designed for simplicity and ease, with a non-intrusive nudge system tailored to enhance user experience and engagement. Unlike typical apps, the DT app offers personalized nutrition guidance based on anticipated post-prandial glucose levels for a tailored dietary management approach.
The data on medication elimination and reduction, particularly for GLP-1 receptor agonists, SGLT2 inhibitors, and DPP4 inhibitors, highlights the potential of the DT intervention to alleviate some of the pharmacological burden for individuals with T2D. This is of importance, especially in the context of pharmacoeconomics, as it reduces the reliance on high-cost T2D medications.
Multiple studies have used TIR as an indicator of glucose control to compare different interventions for T2D management.29 DT participants spent more time in the target glucose levels, less time at very high glucose levels and with no severe or symptomatic hypoglycemic episodes reported throughout the program, underscoring the effectiveness and safety of the intervention.
Postprandial hyperglycemia triggers de novo lipogenesis, leading to fat accumulation in the liver and pancreas, which can cause glucotoxicity and inflammatory damage to β-cells.30 Mitigating postprandial hyperglycemia can halt these processes, thus contributing to glycemic control.30 Dietary intake is a central determinant of blood glucose levels, and thus, in order to achieve optimal glucose levels, it is imperative to make food choices that induce normal PPGR. DT platform will suggest the right food to the right participant at the right time. This precise nutrition, activity, and sleep management ensures that the PPGR is consistently maintained within the optimal range. DT algorithms are designed to suggest dietary options that minimize PPGR for individuals, potentially improving β-cell function and reducing insulin resistance, thereby enhancing the T2D management.4,31,32
The strength of this real-world study is that it highlights the program's comprehensive approach, effectiveness and utility, and examines the program's application in real-world settings, focusing on the role of technology across diverse populations with large sample size. The study's limitations include non-randomized participant selection, a one-year observational period, and potential bias due to participant predisposition towards change. Despite these limitations, the preliminary report presents an evaluation of the program's real-world performance, providing an understanding of its practical application.