What Actually Causes Type 2 Diabetes Remission?
Introduction
For decades, type 2 diabetes was commonly described as a chronic, inevitably progressive disease. Once blood glucose crossed the diagnostic threshold, treatment was largely expected to intensify over time. We now know that this trajectory is not inevitable. In at least a subset of people—particularly earlier in the disease—type 2 diabetes can enter remission.
The RESET for Remission trial provides a striking recent example. Researchers randomized 96 adults aged 18–45 with obesity and early-onset type 2 diabetes to usual care or an intensive 24-week lifestyle intervention combining an 800–900 kcal/day diet with structured aerobic and resistance exercise. Diabetes medications and most antihypertensive medications were withdrawn during the intervention. At 24 weeks, 27 of 50 participants in the intervention group—54%—achieved diabetes remission, compared with only 2 of 46 participants receiving usual care.
One feature of the intervention makes the study particularly useful for understanding remission: the diet was not ketogenic or even particularly low in carbohydrate. During the low-energy phase, approximately 50% of calories came from carbohydrate, 30% from protein, and 20% from fat. Yet more than half of the intervention group achieved remission. This provides strong evidence against the idea that carbohydrate elimination is required for type 2 diabetes remission.
But it does not prove that macronutrient composition is irrelevant. Consuming 50% carbohydrate while eating 800–900 calories per day produces a very different metabolic environment from consuming 50% carbohydrate during chronic energy surplus. Likewise, carbohydrate-restricted diets can improve glycemia through mechanisms that begin before substantial weight loss occurs, particularly by reducing dietary glucose exposure and insulin demand. Different nutritional strategies may therefore approach the same disease through different physiological routes.
The more useful question is not whether low-carbohydrate or low-fat diets win. It is what must actually change inside the body for diabetes to enter remission. Evidence increasingly points toward several interconnected processes: reduced liver fat, improved hepatic insulin sensitivity, lower ectopic fat exposure, decreased metabolic demand on pancreatic beta cells, recovery of beta-cell function in people with sufficient remaining capacity, and improvements in skeletal-muscle insulin sensitivity and body composition.
Understanding those mechanisms also explains why remission is possible for some people but considerably harder for others—and why remission should not be confused with a permanent cure. Before comparing different strategies for achieving it, we first need to define what type 2 diabetes remission actually means.
🎧 Listen to the Episode: Does Diabetes Remission Require Keto?
Ketogenic diets can dramatically reduce glucose and insulin demand, but does that mean nutritional ketosis is required for type 2 diabetes remission? A lifestyle trial in which 54% of participants achieved remission despite consuming roughly half their calories from carbohydrates provides an important reminder: diet labels don't tell the entire metabolic story.
In this episode of The Health Pulse, we explore the twin cycle hypothesis, ectopic liver and pancreatic fat, beta-cell function, resistance training, and the personal fat threshold to understand what may actually determine whether type 2 diabetes moves toward remission. We also explain why a normal HbA1c on medication and medication-free remission are not interchangeable outcomes.
▶️ Click play below to listen, or keep reading to discover why the path to diabetes remission may be less about choosing the “correct” diet camp—and more about changing the physiology that allowed diabetes to develop.
What Does Type 2 Diabetes Remission Actually Mean?
Diabetes remission does not simply mean that glucose improved or that someone reduced their medication dose. An international consensus convened by the American Diabetes Association, European Association for the Study of Diabetes, Endocrine Society, and Diabetes UK defined remission as an HbA1c below 6.5% that persists for at least three months without glucose-lowering medication. When HbA1c is unreliable, alternative measures such as fasting plasma glucose may sometimes be used.
That definition makes an important distinction between controlled diabetes and remission. Someone taking metformin, insulin, a GLP-1 receptor agonist, or another glucose-lowering medication may achieve an HbA1c of 5.8%, which represents excellent glycemic control, but it cannot establish medication-independent remission under the consensus definition. This does not make medication-assisted control less valuable; it simply answers a different question. Remission asks whether glucose can remain below the diabetes threshold after the glucose-lowering treatment has been withdrawn.
Remission is also deliberately different from a cure. A person can maintain normal glucose for months or years while retaining the underlying susceptibility that originally allowed type 2 diabetes to develop. If substantial weight is regained, ectopic fat accumulates again, physical activity falls, beta-cell function continues to deteriorate, or other metabolic pressures return, hyperglycemia can recur. Long-term follow-up from remission studies such as DiRECT has demonstrated this relationship particularly clearly: maintaining substantial weight loss is strongly associated with maintaining remission, while weight regain increases the likelihood of relapse.
There is another reason the distinction matters. HbA1c below 6.5% tells us that glycemia has improved enough to fall below the diagnostic diabetes threshold, but it does not necessarily mean that every component of metabolic health has normalized. Insulin resistance, compensatory hyperinsulinemia, fatty liver, visceral adiposity, dyslipidemia, hypertension, or cardiovascular risk may persist to varying degrees. A person in glycemic remission therefore still benefits from ongoing metabolic monitoring rather than assuming the underlying biology has permanently disappeared.
The RESET trial is compelling precisely because glucose-lowering medications were withdrawn and more than half of participants in the intensive intervention reached remission. The next question is therefore not whether their blood sugar improved—it clearly did. The more important question is what changed physiologically to allow the body to regulate glucose without medication?
Why Type 2 Diabetes Develops: Insulin Resistance Meets Beta-Cell Failure
Type 2 diabetes develops through the interaction of two major abnormalities: insulin resistance and inadequate pancreatic beta-cell compensation. Insulin resistance means tissues require a stronger insulin signal to achieve the same metabolic effect. In skeletal muscle, insulin-stimulated glucose uptake becomes less effective; in the liver, insulin becomes less able to suppress glucose production; and in adipose tissue, impaired insulin action can increase the release of fatty acids into the circulation. None of these changes necessarily produces diabetes immediately because the pancreas can compensate by secreting more insulin.
This compensated state can persist for years. A person may therefore have substantial insulin resistance while fasting glucose and HbA1c remain relatively normal because higher insulin concentrations are holding glucose down. The problem becomes clinically visible when beta cells can no longer produce an adequate insulin response relative to the degree of insulin resistance. Post-meal glucose begins to rise, fasting hepatic glucose production becomes increasingly difficult to suppress, and eventually HbA1c crosses the diagnostic threshold. Type 2 diabetes is therefore not simply “too much sugar in the blood”; hyperglycemia is the measurable result of a system in which insulin demand has exceeded the body's ability to compensate adequately.
This framework helps explain why remission is biologically possible. If insulin resistance can be reduced substantially, the amount of insulin required to regulate glucose also falls. At the same time, reducing chronic metabolic stress may allow beta cells that are dysfunctional but still viable to recover part of their glucose-responsive insulin secretion. Human remission studies have shown that this recovery can occur, particularly in people with shorter diabetes duration and greater remaining beta-cell capacity. The pancreas does not necessarily need to return to some theoretical state of perfect function; it needs to regain enough function relative to the reduced metabolic demand to maintain glucose below the diabetic range.
The liver appears especially important in the early response. Hepatic insulin resistance allows inappropriate glucose production during fasting and contributes directly to elevated fasting glucose. Large reductions in energy intake can improve hepatic insulin sensitivity remarkably quickly—sometimes before major changes in body weight have occurred. Over a longer period, substantial weight loss can reduce liver fat, visceral fat, and potentially excess lipid exposure affecting the pancreas, further changing the metabolic environment in which insulin resistance and beta-cell dysfunction developed.
This gives us a more useful model of remission: reduce the metabolic demand enough, improve insulin sensitivity enough, and preserve or recover sufficient beta-cell function to restore glucose regulation without medication. Different interventions can potentially accomplish those goals through different routes. In RESET, the most obvious starting point was the intervention's profound negative energy balance—and its rapid effects on the liver.
What the RESET Trial Actually Tested
RESET was not simply a comparison between a low-calorie diet and usual care. The trial enrolled 96 adults aged 18–45 with obesity and early-onset type 2 diabetes and randomized them to usual care or a 24-week intensive lifestyle program designed to produce substantial weight loss while maintaining physical fitness and lean tissue. That distinction matters because the remission result reflects the entire intervention, not one isolated dietary variable.
During the intensive phase, participants consumed approximately 800–900 kcal per day, with roughly 50% of energy from carbohydrate, 30% from protein, and 20% from fat. The program also included structured aerobic and resistance exercise. Diabetes medications and most antihypertensive medications were withdrawn under study supervision, allowing investigators to determine whether glycemic control could be maintained without glucose-lowering therapy rather than simply documenting medication-assisted improvement.
After 24 weeks, 27 of 50 participants assigned to the intensive intervention achieved diabetes remission, compared with 2 of 46 receiving usual care—54% versus 4%. That is a large difference for a randomized intervention and particularly notable because these were relatively young adults with early-onset T2D, a phenotype often associated with aggressive disease progression. The trial therefore provides strong evidence that substantial metabolic improvement—and medication-free remission—is possible even in this higher-risk population.
However, the design also tells us what the study cannot establish. Because caloric restriction, weight loss, aerobic exercise, resistance training, and medication withdrawal were incorporated into the same intervention, RESET cannot tell us how much of the remission rate was caused independently by each component. Nor can the study be interpreted as a controlled comparison of high-carbohydrate versus ketogenic diets. The fact that remission occurred while approximately half of dietary energy came from carbohydrate demonstrates that carbohydrate restriction was not required under these conditions; it does not establish that carbohydrate quantity has no effect on glycemia, insulin demand, hunger, adherence, or long-term remission.
The intervention also involved an unusually large energy restriction. At 800–900 kcal per day, a diet containing 50% of calories from carbohydrate provides only about 100–113 grams of carbohydrate daily. That is very different metabolically from obtaining 50% of calories from carbohydrate on a 2,500-kcal diet, which would provide more than 300 grams. Percentages alone therefore obscure an important part of the intervention: total energy and absolute substrate exposure changed dramatically.
RESET's most useful contribution is consequently not evidence that one macronutrient strategy defeated another. It demonstrates that when a sufficiently powerful intervention changes the underlying metabolic environment, type 2 diabetes can enter remission even without carbohydrate elimination. The next question is what such profound energy restriction does first—and one of the earliest changes appears to occur in the liver.
What Happens to Liver Fat During Major Energy Restriction?
One of the fastest metabolic responses to a large reduction in energy intake occurs in the liver. In type 2 diabetes, excess hepatic fat is strongly associated with hepatic insulin resistance. When insulin can no longer adequately suppress hepatic glucose production, the liver continues releasing glucose into the circulation even when the body does not need additional fuel. This inappropriate endogenous glucose production is a major contributor to elevated fasting glucose.
Severe energy restriction changes that environment rapidly. When incoming energy falls substantially below expenditure, the body begins mobilizing stored energy, hepatic triglyceride content can decline, and hepatic insulin sensitivity can improve. Classic metabolic studies using very-low-calorie diets have demonstrated that fasting glucose and hepatic insulin sensitivity can improve within days, well before a person has reached their eventual weight-loss target. This is one reason fasting glucose can fall remarkably quickly during an intensive dietary intervention: the early improvement cannot be explained solely by becoming dramatically thinner. The metabolic flux through the liver has already changed.
This observation is central to the twin-cycle hypothesis developed from work by Roy Taylor and colleagues. In this model, chronic energy surplus promotes accumulation of liver fat, which contributes to hepatic insulin resistance and higher insulin concentrations. A fatty, insulin-resistant liver also exports more triglyceride through VLDL particles, increasing lipid delivery to other tissues, including the pancreas. In susceptible individuals, excess pancreatic fat exposure is proposed to impair beta-cell function. Substantial weight loss can potentially reverse these processes in the opposite direction: liver fat falls first, hepatic insulin sensitivity improves, VLDL-triglyceride export decreases, and the metabolic environment surrounding the beta cell becomes more favorable.
The mechanism is more nuanced than saying that “fatty liver causes diabetes.” Many people with hepatic steatosis never develop T2D, while genetic susceptibility, adipose-tissue storage capacity, visceral fat distribution, skeletal-muscle insulin sensitivity, and beta-cell resilience all influence who progresses. The amount of fat an individual can tolerate before metabolic dysfunction develops also appears to vary substantially. Nevertheless, intervention studies consistently show that reducing excess liver fat accompanies major improvements in glucose metabolism and is strongly associated with remission.
This also explains why the RESET results should not be interpreted simply as evidence for a high-carbohydrate approach. An 800–900 kcal/day diet produces a powerful negative energy balance regardless of its carbohydrate percentage. The liver suddenly receives far less incoming substrate while stored energy is mobilized. Under those conditions, hepatic metabolism can improve even though carbohydrate has not been eliminated.
But normalizing fasting glucose is only part of remission. The liver can reduce its inappropriate glucose output relatively quickly, while restoring normal responses to a meal requires something else: the pancreas must still be capable of producing an appropriately timed insulin response. That brings us to the second major determinant of whether remission succeeds—how much beta-cell function remains and how much of it can recover.
The Pancreas and Recovery of Beta-Cell Function
Improving insulin resistance alone does not fully explain type 2 diabetes remission. Many people live with substantial insulin resistance without developing diabetes because their pancreatic beta cells can increase insulin secretion enough to compensate. Clinical diabetes emerges when that compensation becomes inadequate. For remission to occur, therefore, reducing insulin resistance must leave the remaining beta-cell population capable of producing enough insulin, at the right time, to control glucose.
Importantly, beta-cell dysfunction in type 2 diabetes is not necessarily equivalent to complete beta-cell destruction. Chronic exposure to elevated glucose, excess fatty acids, altered intracellular lipid metabolism, oxidative and endoplasmic-reticulum stress, and persistently high secretory demand can impair how beta cells sense glucose and release insulin. In some people—particularly earlier in the disease—part of this dysfunction appears to be reversible when the metabolic environment improves. This helps explain why diabetes duration is such an important predictor of remission: the earlier the metabolic pressure is reduced, the greater the probability that sufficient recoverable beta-cell capacity remains.
Studies of substantial dietary weight loss have demonstrated this recovery directly. In the Counterpoint and subsequent DiRECT-related work, reductions in liver fat and fasting glucose occurred relatively rapidly, whereas improvements in beta-cell glucose responsiveness and first-phase insulin secretion developed over a longer period. Among people who achieved sustained remission, beta-cell function continued to improve after the initial weight-loss phase. This suggests that remission is not simply the consequence of forcing glucose downward through caloric restriction; the physiology controlling glucose can actually recover to a meaningful degree.
The concept also helps explain why identical weight loss does not guarantee identical outcomes. Two people may lose the same percentage of body weight and achieve similar reductions in liver fat, yet only one enters remission. If one retains enough functional beta-cell capacity while the other has more advanced beta-cell failure, reducing insulin resistance may be sufficient for the first person but not the second. Remission depends on the relationship between insulin demand and remaining insulin-secretory capacity, not on body weight alone.
This is particularly relevant to early-onset type 2 diabetes. Younger age should not automatically be interpreted as milder disease; early-onset T2D can follow an aggressive course and expose patients to decades of cumulative metabolic risk. RESET nevertheless demonstrates that substantial remission is possible when intensive intervention occurs relatively early. One plausible reason is that early intervention may reduce metabolic demand while a meaningful amount of beta-cell function remains recoverable.
This gives us a useful way to think about remission: the goal is not simply to make glucose lower—it is to reduce the workload imposed on the beta cell enough that its remaining capacity can once again meet demand. Major energy restriction can accomplish this partly by reducing liver and ectopic fat and improving insulin sensitivity. Carbohydrate restriction can reduce that workload through another immediate mechanism: dramatically decreasing the amount of dietary glucose that requires disposal after each meal. Before comparing those approaches, however, RESET adds another important variable to the equation—skeletal muscle.
Why Muscle Matters in Type 2 Diabetes Remission
The liver and pancreas receive much of the attention in diabetes remission, but skeletal muscle is one of the largest sites of insulin-stimulated glucose disposal in the body. After a carbohydrate-containing meal, insulin helps move glucose into muscle, where it can be oxidized for energy or stored as glycogen. When skeletal muscle becomes insulin resistant, considerably more insulin may be required to handle the same glucose load, increasing the secretory burden placed on pancreatic beta cells.
This becomes particularly important during aggressive weight loss. A large caloric deficit does not exclusively remove body fat; without an adequate stimulus, some lean mass can also be lost. Losing metabolically active muscle while trying to improve glucose metabolism is not ideal. Resistance training provides a mechanical signal that helps preserve muscle during energy restriction, while exercise itself can improve glucose uptake and insulin sensitivity. Muscle contractions can stimulate glucose transport through pathways that are at least partly independent of insulin, and repeated training produces adaptations that improve mitochondrial capacity, glycogen handling, strength, and insulin responsiveness.
RESET therefore combined its low-energy diet with both aerobic and resistance exercise rather than relying on caloric restriction alone. That makes physiological sense. The dietary intervention created the negative energy balance needed for substantial fat loss, while structured exercise could help preserve functional lean tissue and improve peripheral glucose disposal. The combination potentially attacks metabolic dysfunction from several directions simultaneously: less ectopic and visceral fat, improved hepatic insulin sensitivity, lower beta-cell demand, and a larger or more metabolically capable sink for circulating glucose.
Muscle also helps explain why the number on the scale cannot be the entire story. Two people who each lose 15 kilograms may experience different metabolic outcomes if one primarily loses adipose tissue while preserving muscle and the other loses substantially more lean mass. Body composition, fitness, and muscle quality matter alongside total weight loss. This is particularly relevant when very-low-calorie diets or powerful weight-loss medications produce rapid reductions in body mass.
However, exercise should not be portrayed as the sole explanation for RESET's 54% remission rate. Because caloric restriction and structured exercise were delivered together, the trial cannot isolate the independent contribution of resistance training, aerobic exercise, or diet. What it does demonstrate is that a remission strategy can be designed around more than lowering body weight: reducing excess energy stores while protecting the tissue responsible for much of glucose disposal is metabolically coherent.
At this point, RESET gives us one successful route to remission: create a profound energy deficit, substantially reduce excess fat stores, improve insulin sensitivity, preserve muscle, and reduce the demand placed on the beta cell. But it raises an obvious question for anyone following the low-carbohydrate literature: can carbohydrate restriction reach some of these same physiological endpoints through a different route?
How Carbohydrate Restriction Can Reach Some of the Same Endpoints Differently
RESET demonstrates that carbohydrate restriction is not required for type 2 diabetes remission. But that does not mean carbohydrate intake is physiologically irrelevant. A very-low-calorie diet and a carbohydrate-restricted diet can reduce glycemia through partially different mechanisms, while ultimately converging on several of the same metabolic endpoints: lower insulin demand, improved insulin sensitivity, reduced ectopic fat, and less stress on pancreatic beta cells.
Carbohydrate restriction has an immediate effect that does not require substantial weight loss: it reduces the amount of dietary glucose entering the circulation after meals. For someone with impaired glucose disposal, consuming less carbohydrate generally means a smaller postprandial glucose load and consequently less insulin required to manage it. This can rapidly reduce glucose excursions and exogenous insulin requirements in people using insulin. In that sense, carbohydrate restriction directly addresses a central problem in T2D: the body is struggling to process a glucose load that exceeds its current metabolic capacity.
Over time, carbohydrate-restricted diets can also produce many of the changes observed with conventional weight-loss interventions. When they reduce total energy intake and body fat, liver fat and visceral adiposity can decline, hepatic insulin sensitivity can improve, fasting insulin may fall, and glycemic control can improve substantially. Some people spontaneously eat less when carbohydrate is restricted because protein intake increases, food choices narrow, or hunger and satiety change. In those individuals, carbohydrate restriction simultaneously reduces glucose exposure and creates the negative energy balance needed to mobilize excess stored fat.
This is where RESET provides an important counterpoint. Its participants did not need to minimize carbohydrate to produce a large remission response because the intervention attacked energy surplus very aggressively. At 800–900 kcal per day, even a diet providing approximately 50% of energy from carbohydrate contained only about 100–113 grams of carbohydrate daily. Total substrate delivery was dramatically lower than before the intervention. Liver fat could fall, hepatic glucose production could improve, body fat could be mobilized, and insulin demand could decline even though carbohydrate remained the largest macronutrient by percentage.
The two approaches therefore should not be treated as metabolic opposites. A low-energy diet primarily creates a large energy deficit, while carbohydrate restriction directly reduces glucose and insulin demand and may also create an energy deficit. Both can ultimately reduce the metabolic pressure responsible for maintaining hyperglycemia. The relative importance of each mechanism will vary according to the individual, the severity of insulin resistance, remaining beta-cell capacity, medication use, body composition, dietary adherence, and how much weight or ectopic fat is actually lost.
There is also an important distinction between inducing remission and maintaining it. An 800–900 kcal diet can produce a powerful short-term metabolic response, but it is not intended to remain an 800–900 kcal diet indefinitely. Long-term success requires transitioning to a sustainable eating pattern capable of preventing substantial weight regain and renewed metabolic deterioration. Conversely, some people find carbohydrate restriction easier to maintain because glycemic excursions and hunger improve, while others find it unnecessarily restrictive. The best long-term strategy is therefore not necessarily the diet that produces the fastest initial change, but one that preserves the physiological conditions that allowed remission to occur.
RESET consequently moves the discussion beyond the simplistic question of whether carbohydrates or calories cause diabetes. Both substrate exposure and total energy balance matter, but they describe different parts of the system. The deeper question is whether remission occurs because a person loses weight itself—or because weight loss is a visible marker of much more important changes occurring inside the liver, pancreas, adipose tissue, and muscle.
Is Weight Loss the Mechanism—or a Marker of Metabolic Change?
Weight loss is one of the strongest predictors of type 2 diabetes remission, but saying that diabetes improves simply because someone “lost weight” does not fully explain the biology. A kilogram measured on a scale tells us that body mass changed; it does not tell us which tissues changed, where stored energy was removed, how insulin sensitivity responded, or whether beta-cell function recovered. The more important question is what weight loss represents metabolically.
DiRECT illustrates this distinction particularly well. Greater weight loss was associated with a progressively higher probability of remission, yet people can respond differently despite losing similar amounts of weight. Mechanistic analyses found that remission was associated with reductions in liver and pancreatic fat, normalization of hepatic insulin sensitivity and changes in hepatic VLDL-triglyceride metabolism, provided sufficient beta-cell capacity remained to recover. In other words, weight loss appears to be a powerful means of removing the metabolic conditions sustaining diabetes rather than the scale itself being the biological mechanism.PubMed Central (PMC)
This leads to the concept of a personal fat threshold. Subcutaneous adipose tissue is normally a relatively safe place to store excess energy, but its storage capacity differs considerably between individuals. Once that capacity is exceeded—or adipose tissue becomes dysfunctional—more lipid can accumulate in visceral and ectopic locations such as the liver. This helps explain why BMI alone is an imperfect measure of metabolic health. One person may remain metabolically healthy at a relatively high BMI, while another develops fatty liver, insulin resistance, and T2D at a much lower body weight. PubMed Central (PMC)
Remission may therefore depend less on reaching a universally “normal” body weight than on losing enough excess energy to move below the individual's threshold for metabolic dysfunction. Someone does not necessarily need to become lean. They may need to reduce liver and ectopic fat sufficiently for hepatic insulin sensitivity to improve and for the remaining beta-cell capacity to once again match insulin demand. This framework also helps explain why relatively modest weight regain can sometimes precede recurrence of diabetes: the person may cross that threshold again.
At the same time, we should be careful not to reduce all of T2D to ectopic fat. Genetics, adipose-tissue biology, skeletal-muscle insulin resistance, physical activity, sleep, medications, hormonal disorders, age and beta-cell susceptibility can all modify the phenotype. Even pancreatic fat itself is more complicated than a simple “more fat equals more diabetes” relationship; a 2026 international consensus emphasized that intrapancreatic fat is biologically heterogeneous and that its measurement and interpretation require greater precision. T.co
This changes how we should interpret the RESET trial. The participants were not cured because the scale moved downward. The intensive intervention likely changed energy flux, liver fat, insulin sensitivity, glucose exposure, body composition and beta-cell workload simultaneously. Weight loss is an easily observable marker that many of these processes are moving in the right direction, but the metabolic changes underneath it are what ultimately determine whether glucose regulation can become independent of medication.
And that explains one of the most important observations in remission research: not everyone who loses substantial weight achieves remission. The next question is why.
Why Some People Achieve Remission and Others Don't
Even with substantial weight loss, type 2 diabetes remission is not guaranteed. RESET makes that clear: 54% achieved remission, meaning nearly half of the intensive-intervention group did not, despite receiving the same structured program. Similar variation has appeared across other remission studies. This tells us that the response depends not simply on whether someone follows a particular diet, but on the underlying metabolic phenotype and how much reversible dysfunction remains.
One of the strongest factors appears to be remaining beta-cell capacity. If insulin resistance falls substantially, the pancreas no longer needs to produce as much insulin to maintain glucose control. For someone whose beta cells are dysfunctional but retain meaningful capacity, that reduction in demand may be enough for glucose-responsive insulin secretion to recover and remission to occur. But when beta-cell dysfunction is more advanced, even a major improvement in insulin sensitivity may leave insulin secretion insufficient. This helps explain why shorter diabetes duration is generally associated with a greater probability of remission: intervention is occurring while more recoverable beta-cell function may remain.
The severity and distribution of insulin resistance also differ considerably between individuals. One person may have prominent hepatic insulin resistance and fatty liver, another may have severe skeletal-muscle insulin resistance, and another may have relatively modest insulin resistance but disproportionately impaired insulin secretion. Visceral adiposity, ectopic fat, genetics, physical activity and adipose-tissue storage capacity can further modify these patterns. Two people with the same HbA1c—and even the same BMI—can therefore arrive at diabetes through somewhat different combinations of metabolic abnormalities.
The amount and durability of fat loss also matter. Remission studies consistently show that larger sustained reductions in body weight are associated with higher remission rates, but the amount required varies between individuals. Someone who crosses back below their personal threshold for ectopic fat accumulation may experience a dramatic improvement in hepatic insulin sensitivity, while another person may require substantially greater fat loss before reaching the same physiological point. If weight is subsequently regained and liver or visceral fat begins accumulating again, diabetes can recur.
There is another important clinical issue: not everyone diagnosed with “type 2 diabetes” necessarily has identical disease biology. Adults with hyperglycemia can occasionally have autoimmune diabetes, monogenic diabetes, pancreatic disease, medication-induced hyperglycemia, endocrinopathies, or other causes that resemble conventional T2D. Even within genuine type 2 diabetes, substantial heterogeneity exists. A remission strategy built primarily around reducing insulin resistance will predictably work better when insulin resistance and reversible ectopic-fat accumulation are major drivers than when severe insulin-secretory failure dominates the phenotype.
This is why remission should ultimately become more individualized than simply prescribing weight loss to everyone. HbA1c tells us how high glucose has been; it does not tell us why it became elevated. Fasting insulin or C-peptide, glucose patterns, triglycerides, liver markers, body composition and other metabolic measurements can provide additional context, although none individually determines whether someone will achieve remission.
The practical goal is therefore not merely to get HbA1c below 6.5% once. It is to identify and continually control the metabolic pressures that allowed diabetes to develop in the first place. That distinction becomes especially important when we ask whether someone who reaches remission should consider their diabetes permanently gone—or whether remission and cure are fundamentally different concepts.
Remission Is Not the Same as a Cure
When someone achieves normal or near-normal glucose without diabetes medication, it is tempting to say the diabetes has been “reversed” or cured. Clinically, however, remission is the more accurate term. The international consensus definition intentionally uses remission because glucose can return to the diabetic range in the future, particularly if the metabolic conditions that originally drove the disease reappear.
This distinction becomes clearer when we separate glycemia from susceptibility. An HbA1c below 6.5% tells us that average glucose is currently below the diagnostic diabetes threshold. It does not demonstrate that the person's tendency toward insulin resistance, ectopic fat accumulation, or beta-cell dysfunction has permanently disappeared. Someone may have dramatically reduced liver fat and insulin demand while retaining less beta-cell reserve than a person who never developed diabetes. If insulin resistance increases again, that reduced reserve may once again become insufficient.
Long-term remission studies reinforce this point. In DiRECT, remission was strongly related to maintaining substantial weight loss. As weight was regained over subsequent years, remission rates declined. The five-year extension continued to show that sustained remission was possible for some participants, but maintaining the underlying weight loss remained a major challenge. These findings support a model in which T2D can remain metabolically quiet for prolonged periods without assuming that the underlying susceptibility has been permanently eliminated.
This also means that remission should not be treated as permission to stop monitoring metabolic health. Periodic HbA1c and fasting glucose remain important, while additional markers can provide context depending on the individual. Blood pressure, triglycerides, HDL cholesterol, liver enzymes, kidney function and cardiovascular risk factors may remain relevant even when glucose has normalized. In selected cases, fasting insulin or C-peptide can help characterize insulin secretion and metabolic compensation, although these tests are not part of the formal definition of remission.
The language matters because remission should be viewed as a metabolic state that needs to be maintained, not simply a laboratory milestone that permanently closes the diabetes diagnosis. The encouraging message from trials such as RESET is not that everyone with T2D can be cured in 24 weeks. It is that the physiology underlying diabetes can sometimes be altered enough for glucose regulation to function without medication—and understanding whether those changes persist requires looking beyond HbA1c alone.
That is where laboratory testing becomes particularly useful: not simply determining whether glucose has crossed a diagnostic threshold, but tracking which parts of the metabolic system are actually improving.
How Lab Testing Can Track the Physiology of Remission
HbA1c remains one of the most useful markers for determining whether glycemic control has improved enough to meet the formal definition of remission, but HbA1c alone cannot explain why remission occurred. Two people can reach the same HbA1c while having very different insulin requirements, beta-cell reserve, liver health, lipid profiles, and cardiovascular risk. A more complete laboratory picture can help distinguish improvement in glucose from broader improvement in the metabolic abnormalities that contributed to T2D.
Fasting glucose and HbA1c provide the foundation. HbA1c estimates average glycemic exposure over the preceding several months, while fasting glucose is particularly sensitive to the balance between hepatic glucose production and insulin's ability to suppress it. When fasting glucose improves rapidly during a major energy deficit, that can be consistent with improving hepatic insulin sensitivity, although fasting glucose alone cannot directly measure liver insulin resistance. In selected situations, an oral glucose tolerance test or continuous glucose monitoring can reveal post-meal abnormalities that fasting measurements may miss.
Fasting insulin and C-peptide can add another layer of information. A person whose fasting glucose remains normal only in the presence of substantial hyperinsulinemia is metabolically different from someone maintaining the same glucose with much lower insulin concentrations. C-peptide can be particularly useful when endogenous insulin production needs to be assessed because it is released alongside insulin from pancreatic beta cells and is not contained in injected insulin. Neither fasting insulin nor C-peptide, however, provides a complete measurement of beta-cell function by itself; interpretation depends on the simultaneous glucose concentration, medication use, kidney function, and clinical context.
Liver and lipid markers can help characterize another part of the remission process. ALT, AST, GGT, triglycerides and HDL cholesterol may change as liver fat, insulin resistance and metabolic health improve, although normal liver enzymes do not exclude metabolic dysfunction-associated steatotic liver disease. When cardiovascular risk assessment is appropriate, ApoB, LDL-C, non-HDL cholesterol, Lp(a), blood pressure and kidney function remain relevant because achieving glycemic remission does not automatically erase the cardiovascular risk accumulated during years of diabetes or metabolic syndrome.
Laboratory monitoring is also valuable during intensive treatment because the intervention itself can change medication requirements rapidly. Someone taking insulin, sulfonylureas, antihypertensive medications or other therapies may require medical supervision as glucose, blood pressure and body weight fall. RESET withdrew diabetes medications within a controlled research protocol; that should not be interpreted as a recommendation for patients to discontinue medication independently while beginning a low-calorie or carbohydrate-restricted diet.
For patients in Miami working with their healthcare professionals, QuickLab Mobile can make repeated blood collection more accessible by bringing laboratory collection to the home. Depending on the tests ordered and available, longitudinal measurements such as HbA1c, fasting glucose, insulin or C-peptide, metabolic and liver markers, kidney function, and advanced cardiovascular markers can help build a much richer picture than a single glucose measurement.
The goal of testing is therefore not to collect the largest possible panel. It is to answer specific questions: Has glycemia normalized? Has insulin demand fallen? Is endogenous insulin secretion adequate? Are liver and metabolic markers improving? And are cardiovascular and renal risks being addressed even after remission? When those measurements are followed over time, remission becomes more than an HbA1c result—it becomes a measurable change in the physiology underlying type 2 diabetes.
Conclusion
The RESET for Remission trial reinforces one of the most important changes in how we think about type 2 diabetes: for some people, particularly earlier in the disease, progression is not inevitable. After only 24 weeks, 54% of participants receiving intensive dietary and exercise intervention achieved remission compared with 4% receiving usual care. Just as importantly, they did so without a ketogenic diet.
That finding should not be reduced to the claim that carbohydrates do not matter. RESET used an extremely low-energy intervention of approximately 800–900 kcal per day combined with aerobic and resistance exercise. The intervention dramatically changed the participants' metabolic environment. Energy intake fell, stored fat was mobilized, insulin sensitivity could improve, beta-cell demand decreased, and exercise provided an additional stimulus for glucose disposal while helping preserve muscle.
Carbohydrate restriction approaches the problem somewhat differently. By reducing dietary glucose exposure, it can immediately decrease post-meal glucose excursions and insulin requirements, sometimes before substantial weight loss occurs. If it also produces sustained fat loss, it can eventually converge on many of the same physiological changes produced by intensive energy restriction. Different nutritional strategies can therefore reach overlapping metabolic endpoints through different pathways.
This is why asking whether low-carbohydrate or low-fat diets are superior can miss the more important question. Type 2 diabetes remission appears to depend on whether an intervention sufficiently changes the conditions maintaining the disease: hepatic and peripheral insulin resistance, ectopic fat accumulation, excessive beta-cell demand, impaired insulin secretion, physical inactivity, and chronic energy imbalance. The relative contribution of each factor differs between individuals, which helps explain why the same intervention can produce remission in one person and not another.
Remission also should not be confused with cure. Glucose can remain below the diabetic threshold while an underlying susceptibility to insulin resistance and beta-cell failure persists. Maintaining the metabolic changes that produced remission—and continuing to monitor glucose, cardiovascular risk, liver health, kidney function, and other relevant markers—remains important even after medications are no longer required.
Perhaps the most useful lesson from RESET is therefore not that one particular diet can reverse type 2 diabetes. It is that the metabolic state responsible for type 2 diabetes can sometimes be changed profoundly enough for normal glucose regulation to return. Understanding which physiological abnormalities are driving the disease in each person may ultimately be more useful than arguing over the name of the diet used to get there.
For people in Miami who want to follow their metabolic markers over time, QuickLab Mobile provides convenient at-home blood collection for laboratory testing, making it easier to track changes without repeated trips to a collection center.
👉 Need a specimen collection?. Book Now
Disclaimer:
The information provided in this blog, podcast, and associated content is for educational and informational purposes only and is not intended as a substitute for professional medical advice, diagnosis, or treatment. The content shared is based on reputable sources, medical literature, and expert insights, but it should not be used as a replacement for direct consultation with a licensed healthcare provider.
No Doctor-Patient Relationship: Engaging with this content does not create a doctor-patient relationship between you and QuickLabMobile or any contributors. Always consult with a qualified physician, specialist, or healthcare professional before making any medical decisions, changing your treatment plan, or starting/stopping any medications.
Not a Substitute for Medical Advice: While we strive to provide accurate and up-to-date information, medicine is constantly evolving. New research, treatments, and medical recommendations may emerge, and individual health conditions can vary. Do not rely solely on this content for health decisions. If you are experiencing symptoms, have concerns about your health, or require medical assistance, seek immediate care from a licensed medical professional.
Emergency Situations: If you are experiencing a medical emergency, such as difficulty breathing, chest pain, signs of a stroke, or any other life-threatening condition, call 911 (or your local emergency services) immediately. Do not delay seeking emergency care based on information provided here.
Liability Disclaimer: QuickLabMobile, its contributors, and any associated entities do not assume liability for any damages, harm, or adverse outcomes resulting from the use, interpretation, or misuse of the information provided in this content. You are responsible for your own healthcare decisions and should always verify information with a trusted medical professional.
External Links & References: This content may include links to external sources, medical studies, or third-party websites for further reading. These links are provided for convenience and informational purposes only. QuickLabMobile does not endorse, control, or take responsibility for the accuracy of external content. Always verify information with authoritative sources such as the CDC, NIH, WHO.
Final Note: Your health is unique, and what works for one person may not be suitable for another. Stay informed, ask questions, and always prioritize professional medical guidance
