Protocol

Low-carbohydrate versus very-low-carbohydrate protocols: a survey of the evidence

Definitions

The terminology in carbohydrate-restriction protocols is not standardized across the literature. The editorial team uses the following working definitions:

• Low-carbohydrate: less than 130 grams of carbohydrate per day (the threshold the ADA Standards of Care historically use for “low-carbohydrate eating pattern”).
• Very-low-carbohydrate: less than 50 grams of carbohydrate per day, which approximates the threshold below which most adults enter nutritional ketosis after a few days.
• Ketogenic: an explicit ketogenic protocol, often with carbohydrate intake below 30 grams per day and verified ketosis (urine, breath, or blood beta-hydroxybutyrate measurement).

The numerical boundaries are approximate and individualized. A user can be in nutritional ketosis at intakes above 50 grams under certain conditions and out of ketosis at intakes below 30 grams under others.

Evidence in type 2 diabetes

The evidence base for low-carbohydrate and very-low-carbohydrate protocols in T2D is the strongest of the three populations under discussion. Multiple randomized trials and longer observational studies report:

• improvements in HbA1c, often substantial, in users who maintain protocol adherence;
• reductions in fasting and post-prandial glucose;
• improvements in triglycerides and HDL cholesterol;
• reductions in body weight, with the magnitude depending on caloric balance.

The recent ADA Standards of Care treat low-carbohydrate eating patterns as one of several evidence-based options for T2D, alongside Mediterranean-style and DASH-style patterns. The editorial team’s position aligns with the guideline framing.

For T2D users on insulin or sulfonylurea therapy, transition to a low-carbohydrate or very-low-carbohydrate protocol increases hypoglycemia risk if medications are not concurrently adjusted. The transition should be supervised by the prescribing clinician.

Evidence in type 1 diabetes

The evidence base in T1D is thinner and more complicated. The published literature on very-low-carbohydrate eating in T1D includes:

• a moderate-quality observational literature on small cohorts maintaining very-low-carbohydrate protocols, with reports of improved time-in-range and reduced glycemic variability;
• documented case reports of euglycemic DKA, particularly in users on SGLT2 inhibitors (which are off-label for T1D in most jurisdictions but are used);
• documented hypoglycemia events during dosing transitions, particularly when basal-dose adjustment lags carbohydrate-intake reduction.

The editorial team’s clinical position is conservative:

• T1D users pursuing very-low-carbohydrate protocols should do so only with active endocrinology supervision.
• The combination of SGLT2-inhibitor therapy and very-low-carbohydrate eating should be avoided unless explicitly approved by the prescribing clinician.
• Basal and bolus dosing must be recalibrated concurrently with carbohydrate-intake reduction.
• Users should have a documented hypoglycemia and ketosis-monitoring plan.

Low-carbohydrate (not very-low) eating in T1D is less complicated than very-low-carbohydrate eating; the standard caveats around dosing recalibration apply.

Evidence in pregnancy and gestational diabetes

The editorial team’s position on very-low-carbohydrate eating in pregnancy is unequivocally cautious. Pregnancy is associated with physiological changes in glucose homeostasis (insulin resistance increases through the third trimester) that interact with carbohydrate restriction in ways the literature does not fully resolve. Adequate carbohydrate intake is also necessary to support fetal development on the standard recommended-intake framing.

GDM users considering carbohydrate restriction should do so only on the explicit advice of the obstetric and diabetes care team. The editorial team does not recommend very-low-carbohydrate or ketogenic eating in pregnancy outside such a supervised context.

Carbohydrate-tracking applications and low-carbohydrate protocols

Among the consumer applications, Carb Manager is the most polished tool for low-carbohydrate and very-low-carbohydrate workflows; the net-carb counting and recipe library are mature. Cronometer is also functional for these workflows and offers stronger micronutrient coverage. Users on these protocols benefit from applications that surface fiber and (for ketogenic users) sugar-alcohol subtractions cleanly.

Limits

This article surveys the evidence as the editorial team understands it. Individual users should follow the protocol their care team prescribes and not infer dietary changes from this survey.

References

• Hallberg, S. J., et al. (2024). Long-term outcomes of a digitally delivered very-low-carbohydrate intervention in type 2 diabetes. Diabetes Therapy.
• Lennerz, B. S., et al. (2025). Very-low-carbohydrate diets in adults with type 1 diabetes: a narrative review. Diabetic Medicine.
• American Diabetes Association. (2026). Standards of Care in Diabetes — 2026: Section on nutritional therapy. Diabetes Care.
• Goldenberg, J. Z., & Day, A. G. (2024). Carbohydrate restriction and cardiometabolic risk markers: meta-analysis. Nutrition & Diabetes.
• AACE. (2024). Comprehensive Type 2 Diabetes Management Algorithm. Endocrine Practice.
• Endocrine Society. (2024). Clinical Practice Guideline: Pharmacological management of type 2 diabetes. Journal of Clinical Endocrinology & Metabolism.