Can Low-Carb Diets Put Type 2 Diabetes in Remission?

Low-carb diets can put some people with type 2 diabetes into remission. They do not "cure" diabetes, and the result does not automatically last. The five-year evidence points to a more useful conclusion: remission can persist for a subset of people, especially when weight loss holds and pancreatic beta-cell function remains adequate.

That distinction matters. Better glucose readings while taking medication are valuable, but remission sets a higher bar: HbA1c below 6.5% for at least three consecutive months without glucose-lowering drugs. A 2025 systematic review by Das et al. concluded that low-carbohydrate diets, defined as fewer than 130 grams of carbohydrate per day, and ketogenic diets, typically fewer than 50 grams per day, can meet that bar in a meaningful subset of participants.

The review included six randomized or controlled trials with at least 12 months of follow-up. It also discussed five-year follow-up from the DiRECT trial and Virta Health's remote-care model. Those long-term data matter because a diet that works for 12 weeks answers a different question than one a person can live with for years.

Remission can last, but it is not evenly distributed

The central finding is neither "keto cures diabetes" nor "diet never works." Some participants maintained remission beyond one year and into longer follow-up, while others relapsed. Das et al. describe low-carb remission outcomes as qualitatively comparable to bariatric-surgery outcomes, but the review does not provide a head-to-head numerical comparison at matched time points. That language should stay qualitative.

The review also does not provide extractable remission percentages at one, two, three, or five years. It would be tempting to fill that gap with a tidy graph and some impressive-looking decimals. Resist. The evidence supports persistence in a subset, not a precise five-year remission rate that can be quoted from this review.

Weight regain appears to be a practical route back to higher glucose levels. As people gradually loosen carbohydrate restriction or otherwise regain weight, the conditions that initially improved glucose control may reverse. This does not prove that carbohydrate intake alone explains relapse; most low-carb interventions also change total calorie intake, food choices, and the support participants receive.

Why lower carbohydrate intake could help

Type 2 diabetes commonly involves insulin resistance in the liver, excess liver and visceral fat, and declining insulin secretion from pancreatic beta cells. Reducing carbohydrate intake can lower the immediate glucose load and may reduce the need for insulin. Weight loss and reduced visceral fat can also improve the metabolic conditions that sustain high blood glucose.

A 2022 randomized trial by He et al. found that time-restricted eating with or without a low-carbohydrate diet reduced visceral fat and improved metabolic-syndrome markers. This is useful causal evidence that a structured dietary intervention can change visceral adiposity. It does not isolate the effect of carbohydrate restriction, because the trial tested time-restricted eating with and without low-carb eating rather than low-carb eating alone.

The liver-centered explanation remains biologically plausible, but these studies did not directly measure liver fat with imaging, hepatic glucose production, de novo lipogenesis, or hepatic insulin sensitivity with a clamp. The mechanism is background physiology, not a direct measurement from the low-carb trials discussed here.

The beta-cell ceiling is probably the deciding factor

Dietary adherence matters, but biology sets limits. A 2022 replication analysis by Kabisch et al. found that beta-cell failure consistently predicted metabolic non-response to a complex lifestyle intervention. Non-alcoholic fatty liver disease showed an inconsistent relationship with response.

This study predicted non-response rather than long-term low-carb remission, so it cannot tell us exactly who will remain in remission for five years. Still, the inference is sensible: people with more residual insulin-producing capacity may have a better chance of crossing and maintaining the remission threshold. Longer diabetes duration often means more beta-cell loss, although this evidence does not establish a specific duration cutoff or C-peptide value.

For someone with long-standing type 2 diabetes and marked beta-cell failure, a low-carb plan may still improve glucose, weight, or medication needs. It may simply be less likely to produce drug-free remission. That is a physiological constraint, not a failure of effort.

Carbohydrate restriction is not the only plausible explanation

The strongest challenge to a carbohydrate-specific story comes from other diets that create substantial energy deficits. A 2022 systematic review and meta-analysis by Kashyap et al. found clinically relevant improvements in weight and glycemic markers with both very-low-calorie diets and low-fat vegan diets in people with type 2 diabetes.

Those findings do not show that these approaches produce the same remission rates as low-carb or ketogenic diets. They do show that carbohydrate restriction is not the only route to better glycemic markers. Low-carb diets often reduce calorie intake through satiety, so existing studies cannot cleanly separate the effect of fewer carbohydrates from the effect of eating less overall. No matched-calorie trial in this evidence base resolves that question across low-carb, ketogenic, and higher-carbohydrate diets.

Diet comparison also remains thin. A 2024 16-week intervention study by Currenti et al. compared a low-carbohydrate diet with a Mediterranean diet in overweight or obese adults with type 2 diabetes. Its short duration makes it useful for early metabolic comparisons, not for judging remission durability. The Mediterranean pattern remains an evidence-based option, especially for people who find severe carbohydrate restriction hard to sustain.

Signals beyond glucose: visceral fat and BCAAs

A five-year CORDIOPREV analysis by Cardelo et al. followed 183 people with newly diagnosed type 2 diabetes assigned to Mediterranean or low-fat diets. In the Mediterranean-diet group of 80 participants, higher two-hour oral-glucose-tolerance-test levels of isoleucine and valine were associated with remission outcomes: hazard ratios per standard deviation were 0.53 (95% CI, 0.37 to 0.77) for isoleucine and 0.61 (95% CI, 0.45 to 0.82) for valine. Leucine's estimate, 0.75 (95% CI, 0.52 to 1.08), included the null. The corresponding associations were absent in the low-fat group of 103 participants.

Branched-chain amino acids are linked with insulin resistance, but this is an association study within dietary groups, not proof that changing BCAAs causes remission. It does, however, reinforce the idea that remission involves more than carbohydrate arithmetic.

Safety and clinical reality

A 2022 systematic review and meta-analysis by Zaki et al. also found that intermittent fasting combined with a ketogenic diet improved glycemic control and reduced weight in type 2 diabetes. That supports the broader pattern of short- and medium-term metabolic benefit, but it does not settle five-year remission durability.

Long-term safety remains incomplete. Das et al. flagged LDL cholesterol increases in some ketogenic-diet participants, uncertain renal effects for people with diabetic nephropathy, and possible micronutrient deficiencies. The available evidence also lacks long-term data on cardiovascular events, fractures, and bone density. Anyone using insulin, sulfonylureas, or other glucose-lowering medications needs clinician-guided dose adjustment; rapid dietary changes can cause hypoglycemia when drugs stay unchanged.

A low-carb diet is a legitimate remission strategy for some people with type 2 diabetes. It is not a substitute for monitoring HbA1c, medication safety, kidney status, lipids, nutritional adequacy, and the ability to sustain the plan.

The verdict

Low-carb and ketogenic diets can induce drug-free type 2 diabetes remission, and five-year follow-up discussed in systematic-review evidence shows that remission can persist for some participants. The evidence does not support a universal cure claim or a precise five-year success rate from the studies summarized here.

The best candidates may be people with more recently diagnosed disease, preserved beta-cell function, meaningful and sustained weight loss, and clinical support that makes medication changes safe. Others may gain better glucose control without reaching remission. The useful question is not whether keto is magic. It is whether a structured dietary approach can produce durable metabolic improvement for the person sitting across from you.