Examine publishes rigorous, unbiased analysis of the latest and most important nutrition and supplementation studies each month, available to all Examine Members. Click here to learn more or log in.

In this article

Deep Dive: Cutting carbs may help reverse diabetes, but dietary adherence remains a challenge

Low-carb diets can reverse diabetes for some people, but they can be challenging to stick to, and longer-term efficacy remains uncertain.

Study under review: Efficacy and safety of low and very low carbohydrate diets for type 2 diabetes remission: systematic review and meta-analysis of published and unpublished randomized trial data


Type 2 diabetes has been a constant challenge for researchers and clinicians for more than a century, if not longer[1]. Throughout its history, a cornerstone of treatment has been based on nutrition. Before the discovery of insulin 100 years ago in 1921, nutritional fasting was the only treatment available, but as pharmacotherapies (insulin, sulfonylureas, biguanides, etc.) have been developed, treatment has tended to emphasize these new technologies, pushing research on nutritional management to the side.

Perhaps due to an excessive focus on the available medical treatments, diabetes remission was considered impossible for decades. In the 1990s and 2000s, mention of diabetes remission in PubMed was limited to animal models, bariatric surgery, and experimental treatments like islet cell transplantation. Little attention was paid to the few extant human diet studies, and these studies focused on delaying diabetes progression[2], not putting the disease into remission entirely. Only recently have scientific[3], medical[4], and public health communities recognized the possibility of putting the disease in remission with diet and lifestyle changes.

As methodological rigor in nutritional studies has improved, so too has the confidence placed in nutrition therapies by professional organizations increased. For instance, the American College of Lifestyle Medicine[5], American Diabetes Association[6], Diabetes Canada[7], Diabetes UK, European Association for the Study of Diabetes[8], and other organizations now consider the evidence for low carbohydrate diet equally, if not more, compelling as the long standing low-to-moderate fat and consistent-carbohydrate diets recommended by government and other organizations like the American Academy of Nutrition and Dietetics[9].

Diet and lifestyle changes are promoted as a key feature of diabetes management plans because they induce[10] weight[9] loss[2]. Weight loss may decrease insulin resistance and could restore beta-cell function, the purported mechanisms of type 2 diabetes[11] pathophysiology[12]. Dietary macronutrient manipulation can also substantially affect metabolic function and energy substrate, suggesting a possible role for macronutrient modified diets. Moderate fat restriction has been associated with weight loss and lower LDL cholesterol, despite also lowering HDL cholesterol and raising triglycerides[13]. Moderate carbohydrate restriction can induce some positive effects on insulin signaling and blood lipids, but more severe carbohydrate restriction can induce ketogenesis, the metabolic changes of which may positively affect insulin resistance and sensitivity seen in type 2 diabetes metabolism[14], but these effects are likely attributed to the weight loss often associated with stringent carbohydrate restriction. Studies measuring insulin sensitivity in weight stable people with the gold-standard hyperinsulinemic-euglycemic clamp show that moderate carbohydrate restriction has no effect on insulin sensitivity, and very low carbohydrate ketogenic diets may actually reduce insulin sensitivity. Most studies that estimate insulin sensitivity using the homeostatic model assessment of insulin resistance (HOMA-IR) suggest carbohydrate-restriction can increase insulin sensitivity, but this is misleading in the context of low carb diets because they reduce insulin secretion independently of insulin sensitivity[15]. In fact, one study[16] comparing a high-carbohydrate diet to a very-low carbohydrate diet in the context of weight loss showed that carbohydrate-restriction resulted in greater reduction in insulin resistance measured by HOMA-IR, but no differences between diets were observed when insulin sensitivity was measured with the hyperinsulinemic-euglycemic clamp.

A chronic issue in previous cohort studies[17] and meta-analyses[18] on carbohydrate-restricted diets is that the definitions of low carbohydrate diets and very low carbohydrate diets are quite inconsistent. Less stringent definitions of “low carb” applied in research can result in conclusions that may not line up with the real results of people following very low or low carbohydrate diets defined more stringently. The authors use a common definition[19] of very low carbohydrate diets (less than 10% of calories, or 20–50 grams from carbohydrate per day) and low carbohydrate diets (less than 26% or 130 grams per day), but ultimately these thresholds are estimates, not standardized, and may be arbitrary in light of the individualized response to carb cutting.

In addition to the diverse definitions of different tiers of carbohydrate restriction, the physiological response to tiered carbohydrate restriction is not linear. The data[20] suggest that[21] as the proportion of calories from carbohydrates decreases, ketone body concentration increases exponentially, while fasting blood glucose decreases slightly in a linear fashion[21] in healthy individuals. In light of this suggested non-linear relationship between carbohydrate intake and circulating ketone bodies, previous studies that combine the results of very low carbohydrate diets and moderate carbohydrate diets may be missing effects that only appear with very low carbohydrate restriction. For reference, the previously cited trials that suggested increased mortality with low carb diets defined “low carbohydrate” as 39%[18] to 40%[17] of calories, quite far from the meager 10% or less often necessary for ketosis.

Figure 1: The nonlinear increase in ketone production

References: Johnston et al. Am J Clin Nutr. 2006 May.[21]
Puchalska et al. Cell Metab. 2017 Feb.[20]
Harvey et al. Nutr X. 2019 Jun. DOI: 10.1016/j.nutx.2019.100005.

Other systematic reviews and meta-analyses[22][23][24][25][26] (the last of which was covered in Study Deep Dives #72) have tried to answer the question: how does low carb stack up against other diets for diabetes treatment? Lack of consensus on the definition of diabetes remission, the extent and definition of carbohydrate restriction, and the length of time during which people may expect to experience positive effects have all contributed to divergent results. This systematic review and meta-analysis aims to address some of the inconsistencies left by previous meta-analyses: the issues of what constitutes a “low carbohydrate” diet, and what constitutes diabetes reversal.

A number of randomized controlled trials comparing carbohydrate restriction with other diets for diabetes management have provided suggestive evidence that low carb diets may help to not just manage type 2 diabetes, but to actually send it into remission, as opposed to previous strategies that have aimed to slow the progression of the disease. This has been documented by a number of meta-analyses, and reflected by the revision of a number of professional organization position statements. Differences in study methodologies have so far made it difficult to compare the effects of different tiers of carbohydrate intake.

What was studied?

Become an Examine Member to read the full article.

Becoming an Examine Member will keep you on the cutting edge of health research with access to in-depth analyses such as this article.

You also unlock a big picture view of 400+ supplements and 600+ health topics, as well as actionable study summaries delivered to you every month across 25 health categories.

Stop wasting time and energy — we make it easy for you to stay on top of nutrition research.

Try free for two weeks

Already a member? Please login to read this article.

What were the findings?

Become an Examine Member to unlock this article.

Already a member? Please login to read this article.

The bigger picture

Become an Examine Member to unlock this article.

Already a member? Please login to read this article.

Frequently asked questions

Become an Examine Member to unlock this article.

Already a member? Please login to read this article.

What should I know?

Become an Examine Member to unlock this article.

Free 2-week trial »

Already a member? Please login to read this article.

Other Articles in Issue #76 (February 2021)