Dietary modification is crucial for managing type 2 diabetes mellitus (T2DM), with carbohydrate reduction being a key strategy. However, access to adequate dietary education and support is often limited due to factors like workforce capacity, location, and cost. Web-based interventions offer a potential solution to overcome these accessibility challenges. This study aimed to evaluate the effectiveness of a web-based healthy low-carbohydrate diet (LCD) education program (T2Diet) in improving glycemic control in adults with T2DM when delivered in conjunction with standard care. The study hypothesized that the web-based LCD program would result in lower HbA1c levels at 16 weeks compared to standard care alone. The increasing prevalence of T2DM and the limitations of existing support systems necessitate innovative approaches to improve patient outcomes and reduce the burden on healthcare systems. Web-based interventions, like the T2Diet program, offer a scalable and potentially cost-effective strategy for delivering widespread dietary education and support.

Existing literature strongly supports carbohydrate reduction as an effective strategy for improving glycemic control and other clinical outcomes in individuals with T2DM. Studies have shown that both ketogenic diets and low-carbohydrate diets (LCDs) lead to improved glycemic control, weight reduction, and decreased medication requirements. While ketogenic diets have shown promise, LCDs have demonstrated greater dietary adherence in several reviews. Despite historical misconceptions, healthy LCDs align with recommendations from organizations like the American Diabetes Association, emphasizing high consumption of non-starchy vegetables and nutrient-dense foods while minimizing added sugar, refined grains, and processed foods. International diabetes guidelines also support LCDs as a suitable dietary option for people with T2DM. However, a significant gap exists in research regarding the effectiveness of web-based LCD interventions delivered through randomized controlled trials (RCTs).

This study employed a 16-week, two-arm parallel randomized controlled trial (RCT) conducted remotely during the COVID-19 pandemic. Participants (n=98), aged 40-89 years, with T2DM not on insulin, and with HbA1c ≥7.0% within six months of enrollment, were recruited through various channels including community organizations, social media, and Facebook advertising. Participants were randomly assigned (1:1) to either the intervention group (standard care plus the web-based T2Diet program) or the control group (standard care only). The T2Diet program was an automated 16-week program recommending carbohydrate intake of 50-100 g per day with ad libitum consumption of nutrient-dense, lower-carb foods. Researchers and participants were blinded to group allocation until after group assignment. The primary outcome was HbA1c, measured at baseline and 16 weeks. Secondary outcomes included weight, BMI, anti-glycaemic medication use (quantified using the Medication Effect Score), dietary intake (assessed via 24-hour recall), and self-efficacy (measured using the Diabetes Management Self-Efficacy Scale). Data analysis was conducted using intention-to-treat principles, with multiple imputation used to handle missing data. Generalized estimating equations (GEE) models were employed to analyze continuous outcomes.

At 16 weeks, the intention-to-treat analysis revealed statistically significant between-group differences favoring the intervention group. The intervention group experienced a significantly greater reduction in HbA1c (-0.65%, 95% CI: -0.99 to -0.30, p < 0.0001) compared to the control group. Large effect sizes (Cohen's d > 0.8) were observed for the primary and secondary clinical outcomes, indicating substantial treatment effects. Specifically, the intervention group demonstrated significant reductions in weight (-3.26 kg, p < 0.0001), BMI (-1.11 kg/m², p < 0.0001), and anti-glycaemic medication requirements (-0.40, p < 0.0001). A substantial proportion of participants in the intervention group achieved clinically meaningful weight loss (≥5% weight loss in 38% of participants, compared to 9% in the control group), and a significant reduction in medication was observed in 25% of intervention participants (≥20% reduction). Analysis of dietary intake showed a significant between-group difference in carbohydrate intake (p < 0.0005), with the intervention group consuming 24.7% of total energy from carbohydrates, aligning with the recommended LCD range. No significant differences were found between groups for energy intake, protein, saturated fat, or dietary fiber, although the intervention group showed a significant increase in monounsaturated and polyunsaturated fat intake. While self-efficacy scores increased in the intervention group, the between-group difference was not statistically significant.

This study provides the first RCT evidence demonstrating the effectiveness of a web-based LCD intervention for improving glycemic control in adults with T2DM. The observed reductions in HbA1c, weight, BMI, and medication requirements are consistent with previous meta-analyses of face-to-face LCD interventions. The significant improvements achieved through a web-based platform highlight the potential for enhancing access to effective dietary management strategies, particularly for individuals in remote areas or with limited access to in-person support. The substantial weight loss observed in the intervention group, even without explicit calorie restriction, is noteworthy and suggests the potential role of hormonal and metabolic changes associated with LCDs. The lack of negative impact on dietary fiber and the increase in beneficial fats indicate that the nutritional quality of the LCD intervention was maintained. The non-significant difference in self-efficacy suggests that other factors, such as increased knowledge or motivation, may have contributed to the observed improvements.

The T2Diet Study demonstrates the efficacy of a web-based, healthy LCD education program in improving glycemic control and related clinical outcomes in adults with T2DM. The web-based delivery method addresses significant barriers to access and availability of dietary support. Future research should focus on larger-scale trials with longer follow-up periods to assess the long-term sustainability of these effects and explore integration within primary care settings to optimize implementation.

The study's limitations include the reliance on self-reported anthropometric and dietary data, the lack of longer-term follow-up, and the absence of additional biomarkers. The sample size, while powered for the primary outcome, may limit the generalizability of findings to certain subgroups. The exclusion criteria may also limit the generalizability to individuals with specific comorbidities or dietary preferences.