5-Year effects of a novel continuous remote care model with carbohydrate-restricted nutrition therapy including nutritional ketosis in type 2 diabetes: An extension study

What this study found

This study followed adults with type 2 diabetes who used a telemedicine program that coaches a very low‑carbohydrate diet aimed at nutritional ketosis. Over five years, many participants saw lasting improvements:

• 20% of those who finished the five years were in diabetes remission by the 2021 international consensus definition (HbA1c <6.5% for at least 3 months without glucose‑lowering drugs). Another 33% had HbA1c <6.5% while taking no medication or only metformin.
• People generally needed fewer diabetes medications, including notable reductions in insulin and sulfonylureas.
• Key health markers shifted in a favorable direction: HbA1c fell slightly, body weight dropped and stayed lower, triglycerides decreased, HDL‑cholesterol rose, and inflammation (hs‑CRP) declined, while LDL‑cholesterol and total cholesterol did not significantly change at five years.

Importantly, many who achieved remission at year 2 were still in remission at year 5, suggesting the approach can be sustained for years.

What changed, in everyday terms

• Diabetes status: - Remission at 5 years: 20% of completers (14% with conservative intent‑to‑treat assumptions). - HbA1c <6.5% with no meds or only metformin: 32.5% of completers (23.1% intent‑to‑treat).
• Weight: - Average weight fell by 7.6% from baseline and stayed lower. Baseline weight was 116.4 kg (256.6 lb) and fell by 8.8 kg (19.4 lb) on average.
• Glycemia: - HbA1c decreased by 0.3 percentage points (absolute). - Fasting insulin dropped 30.6%; insulin resistance (HOMA‑IR) improved.
• Lipids (blood fats): - HDL‑C rose 17.4% from 43.0 mg/dL (1.11 mmol/L) to 50.6 mg/dL (1.31 mmol/L). - Triglycerides fell 18.4% from 202.3 mg/dL (2.28 mmol/L) to 165.1 mg/dL (1.86 mmol/L). - LDL‑C and total cholesterol showed no significant change at five years versus baseline.
• Inflammation: - hs‑CRP decreased 43.6% and white blood cell count fell 9.7%, indicating lower systemic inflammation.
• Medications: - Any diabetes medication use dropped from 85.2% to 71.3%. - Insulin use fell from 26.2% to 13.1%; sulfonylureas from 27.0% to 4.9%. - Overall treatment intensity (Medication Effect Score) decreased 40%.

Retention was strong: 72.1% of those who entered the 3‑year extension completed the 5‑year follow‑up, and among all initially enrolled participants, 47% remained engaged at five years.

Practical relevance

• A very low‑carb approach delivered through continuous remote care can, for a sizable minority, bring type 2 diabetes into remission and for many others reduce medication needs while maintaining or improving blood sugar control.
• Even for people who did not reach remission, the program was associated with lower medication burden, sustained weight loss, and better triglycerides, HDL‑C, and inflammation markers.
• LDL‑C and total cholesterol did not significantly change at five years in this cohort.

Study details and level of evidence

• Design: A 2‑year, open‑label, non‑randomized study comparing the continuous care intervention (CCI) with usual care, followed by a 3‑year extension of the CCI group to reach five years total.
• This is not a randomized controlled trial (RCT) across the full 5 years. After year 2, there was no comparison group, so causality beyond two years is less certain.
• Participants: - Initially enrolled in 2015–2016; 262 people started in the CCI arm. - At two years, 194 remained; 169 consented to the 3‑year extension; 122 (72.1%) completed the 5‑year follow‑up.
• Intervention: - Telemedicine team advised a very low‑carbohydrate diet to achieve nutritional ketosis; initial targets included <30 g carbs/day, protein ~1.5 g/kg reference weight, fat to satiety, then personalized adjustments. Remote monitoring used an app, coaches, and clinical oversight. Metformin was generally not deprescribed unless contraindicated.
• Outcomes assessed at baseline, 1, 2, 3.5, and 5 years: diabetes status (including remission by 2021 consensus), HbA1c, weight, fasting glucose, insulin, lipid profiles (HDL‑C, LDL‑C, triglycerides), inflammatory markers (hs‑CRP, WBC), liver and kidney markers, and medication use.
• Statistics: - Mixed‑effects models for continuous outcomes; McNemar’s tests for status changes; intent‑to‑treat analyses counted missing data conservatively as not meeting targets.
• Funding and interests: - Funding: Virta Health. - Affiliations: Several authors are affiliated with Virta Health; one with The Ketogenic Foundation. The paper declares no competing financial interests. - Readers should be aware that sponsor involvement and author affiliations can introduce potential bias; however, methods and results (including intent‑to‑treat) are transparently reported.
• Limitations noted in the paper: - No control group after year 2; non‑randomized design. - Limited diversity (mostly non‑Hispanic white), affecting generalizability.
  - COVID‑19 overlapped with data collection (to April 2021), possibly influencing behaviors and outcomes. - Metformin retention policy may affect remission categorization; thus the authors also reported the “HbA1c <6.5% with no meds or metformin‑only” outcome.
• Safety: No adverse events were attributed to the intervention in the reported period.

How to interpret these findings

• Strengths: Long duration (5 years); high retention in a lifestyle program; multiple cardiometabolic improvements with less medication; objective lab measures; remission defined by international consensus.
• Caveats: Because the five‑year analysis lacks a randomized control group, we cannot say with certainty that all changes were caused solely by the intervention over years 3–5. Still, the sustained trends (especially medication reductions, weight loss, triglyceride and HDL improvements, stable LDL‑C) are consistent with the program’s aims and earlier controlled phases.
• Bottom line: In a real‑world, long‑term setting, a continuous remote‑care model emphasizing very low carbohydrate intake and ketosis was associated with meaningful, durable benefits for many people with type 2 diabetes, including a notable rate of remission, fewer drugs, and better cardiometabolic markers.