Published on September 20, 2023
This is the second blog post about the causes of insulin resistance. In the first, entitled Causes of Insulin Resistance: The Personal Fat Threshold, I discussed the role of visceral fat, ectopic fat, and low-grade chronic inflammation in the development of insulin resistance, and introduced a concept called the personal fat threshold.
However, it is clear that not all cases of insulin resistance are linked with excess body fat mass. You may have heard on YouTube, social media, or in certain low-carb diet books that insulin resistance is the result of chronically elevated insulin levels (hyperinsulinemia), which in turn is suggested to be the result of excessive consumption of high-carb foods (see figure below).
In this blog post, we are looking at whether there is scientific evidence to support hyperinsulinemia and dietary carbs as a cause of insulin resistance.
We’ll do this by first looking at whether elevated insulin levels (hyperinsulinemia) cause insulin resistance, and then by reviewing the scientific literature on whether consuming high-carb diets indeed causes insulin resistance and – vice versa – whether low-carb diets improve insulin sensitivity.
Does Hyperinsulinemia Cause Insulin Resistance?
I am aware of two lines of evidence that can shed some light on whether elevated insulin concentrations cause insulin resistance. The first is from hyperinsulinemic-euglycemic clamp studies, and the second is based on patients suffering from a condition called an insulinoma.
Insulin Resistance as a Result of Hyperinsulinemia in Clamp Studies
The first line of evidence is data from hyperinsulinemic-euglycemic clamp studies. To explain this procedure, I’ll need to get a little technical. A clamp is a clinical procedure used to measure insulin resistance. It is very involved and requires considerable expertise, so it’s almost only done in the context of research studies.
During the procedure, the investigators continuously administer insulin right into the bloodstream. Insulin is infused to reach specific insulin levels, such as slightly elevated or very elevated. If you infuse insulin continuously into the blood, what do you think happens? Well, the insulin will stop glucose production by the liver, and cause muscle and fat cells to take up glucose from the blood. If this is not clear, I invite you to revisit my blog post about the Regulation of Blood Sugar.
The result of these insulin actions would be that blood glucose levels start to drop. During the clamp, the investigators continually measure the blood sugar levels and infuse glucose right into the bloodstream to make sure that blood sugar levels remain stable within the normal range. In some people, the investigators need to infuse very little glucose to keep blood glucose levels stable, in others, a lot.
So, please think about this for a second: What can we learn about a person’s insulin sensitivity from this test?
Well, the investigators measure how much glucose they need to infuse to keep glucose levels stable within the normal range. The resulting measurement is called the glucose infusion rate, or GIR, and that is a direct measure of insulin sensitivity. Because if a person is very insulin sensitive, then having elevated insulin levels will cause their muscle and fat cells to take up a lot of glucose from the blood. Therefore, they require a lot of extra glucose to be infused to keep blood glucose levels stable. That means that for someone who is very insulin-sensitive, the glucose infusion rate will be high.
However, if someone is insulin resistant, then they will require very little glucose to keep their blood glucose levels stable. Their GIR will be low.
OK, so, this is the gold-standard measure of insulin sensitivity in healthy people. But, why am I mentioning this? Does this tell us something about the causes of insulin resistance?
Yes, potentially. Two published studies show that after many hours of maintaining elevated blood insulin and normal blood glucose levels during such a clamp, the body becomes gradually more and more insulin resistant.
For example, Del Prato and colleagues observed that 4 days of continuous hyperinsulinemic euglycemic clamp reduced insulin sensitivity by 20-40% in healthy young people. Similarly, Rizza and colleagues found that 40 hours of hyperinsulinemic clamp in which insulin levels were kept at about the same level that is seen in obese men and women did cause insulin resistance.
Now, from these data, we cannot be 100% certain that it was really the high insulin levels that caused the insulin resistance. The problem is that one can safely maintain such elevated insulin levels for an extended period of time only by also infusing glucose, and that glucose needs to be taken up into the muscle and fat cells. Thus, it is also possible that the insulin resistance that results from several hours of hyperinsulinemia is the result of excessive uptake of sugar into the cells, i.e., energy toxicity, which we discussed in the blog post about the personal fat threshold hypothesis. I personally don’t believe that energy toxicity fully explains these results, because participants are kept in energy balance during clamps by adjusting the amount of food they can eat for the amount of glucose that is infused. Therefore, my best interpretation would still be that insulin resistance in this setting is the result of hyperinsulinemia per se. We should be clear that these data are not fully conclusive, however.
Insulin Resistance as a Result of an Insulinoma
There is another area of scientific research, however, that does provide additional evidence that elevated insulin levels per se could be a cause of insulin resistance, and that is based on a medical condition called insulinoma.
An insulinoma is a benign tumor of the pancreas that leads to overproduction of insulin. Depending on the size of the tumor, the amount of insulin produced throughout the day may be only slightly elevated or several-fold higher than in healthy people.
Because insulin suppresses the production of glucose by the liver and also enables cells to take up glucose from the blood, the main clinical expression of insulinomas is hypoglycemia, i.e. low blood sugar levels (<70 mg/dL or <3.9 mmol/L). If insulinoma patients remained perfectly insulin sensitive, then they would die very quickly due to hypoglycemia. However, they do not remain perfectly insulin-sensitive. Several clinical studies have shown clearly that these patients become quite insulin-resistant as a result of their condition, in most studies well above the level that would be expected based on their age and body mass index (BMI). Further, once the insulinoma is removed surgically and the chronic hypersecretion of insulin ceases, insulin sensitivity normalizes.
Again, this literature does not perfectly link hyperinsulinemia to insulin resistance. For example, there is at least one publication that shows that the surgical removal of an insulinoma dramatically reduces visceral fat mass. This suggests that the insulinoma may have played a role in the accumulation of visceral fat, and because visceral fat is thought to be a key determinant of insulin resistance, we cannot currently rule out the possibility that insulin resistance in hyperinsulinoma patients is at least partly the result of increased visceral fat mass and not the hyperinsulinemia per se. That is because other published studies in insulinoma patients matched for a general marker of body fat mass such as the body mass index (BMI), but did not measure visceral fat mass and therefore could not assess to which degree the increase in insulin resistance seen in these patients may be explained by increased visceral fat mass.
Conclusion: Does Hyperinsulinemia Cause Insulin Resistance?
Taken together, the evidence from clamp studies and insulinomas strongly suggests that high insulin levels could be a cause of insulin resistance, at least in situations where hyperinsulinemia persists for many hours to many days.
The one question that remains unclear, and this is an important consideration, is whether the normal, physiological up and down in blood insulin levels that occurs when we are eating meals rich in carbohydrates may similarly trigger insulin resistance. In clamp studies and in insulinoma patients, we have persistently high insulin levels for hours to days, and the impact of chronic hyperinsulinemia may very well be quite different from the usual up and down in insulin levels that occurs with eating high-carb meals.
Therefore, while I feel that this literature provides evidence that chronically elevated blood insulin levels likely cause insulin resistance, it remains unclear whether this also applies to dietary carbohydrate-induced temporary increases in insulin.
Do High-Carb Diets Cause Insulin Resistance, and Do Low-Carb Diets Improve Insulin Sensitivity?
What we therefore need to do is evaluate the scientific evidence from clinical feeding studies in which participants ate diets varying widely in their carbohydrate content. The specific questions we are interested in are whether high-carb diets decrease and low-carb diets increase insulin sensitivity.
There are numerous studies that show a substantial increase in insulin sensitivity in people consuming low-carb, very-low-carb, or ketogenic diets. However, some of the most frequently cited studies did not control calorie intake and weight change, and in those studies, study participants improved their insulin sensitivity while also losing weight. This is not a critique of these studies; many were designed for reasons other than investigating the impact of low-carb vs. high-carb diets on insulin sensitivity. And it is certainly practically relevant, as these studies demonstrate a key benefit of low-carb diets, namely weight loss, these studies cannot tell us whether the improvement in insulin sensitivity is the result of the reduced carb intake per se or the reduction in body weight and fat mass.
I, therefore, looked for dietary intervention studies in which
(a) participants were provided with diets differing substantially in their carb content;
(b) in which insulin sensitivity was measured; and
(c) in which the impact of the diet composition on insulin sensitivity could be separated from the impact of body weight loss.
Point (c) means that there were either no weight changes in either group because calorie intake was controlled, or that the dietary study arms were matched for calorie intake and weight loss.
I summarized those studies comparing the body weight-independent impact of high-carb vs. low-carb studies on insulin sensitivity in the table below.
Let’s go through some of these studies to get a sense of what was done, and what was found. Of note, in order to represent these studies accurately, and get a complete picture of these diets on insulin sensitivity, this section will be unavoidably technical when discussing the results of the clamp studies.
There were several very small studies (n=6 in Bischop, n=9 in Lundsgaard, and n=10 in Chokkalingam) that also were quite short in duration (11 days, 3 days, and 6 days). However, these were all cross-over design studies, meaning that each participant completed each dietary intervention phase and can, therefore, be directly compared to him- or herself. And all three studies employed the gold standard measurement of insulin sensitivity, the hyperinsulinemic-euglycemic clamp. And lastly, in all three studies, participants consumed at least five times more carbs on the high-carb diet condition than on the low-carb diet condition, i.e., there were very substantial differences in carb intake between the high-carb and low-carb diet.
If we look at the studies first authored by Bisschop or Lundsgaard, participants ate very little carbs and mostly fat on the low-carb diet, and mostly carbs and very little fat on the high-carb diet (protein and energy intake were matched). The studies differed in one important way: while participants in the Bisschop study were given liquid meals that matched their energy requirements, participants in the Lundsgaard study were provided with regular food, but well in excess of their energy requirements.
Bisschop and colleagues found that hepatic insulin sensitivity was reduced following the low-carb diet, and also saw a trend towards a reduced systemic insulin sensitivity, as GIR was reduced at low/basal insulin concentrations following the low-carb diet, and tended to be reduced under hyperinsulinemic conditions.
Measures of Insulin Sensitivity Based on a Hyperinsulinemic-Euglycemic Clamp
In a hyperinsulinemic euglycemic clamp, we usually obtain two measures of insulin sensitivity, and both are important.
One is the rate at which we need to infuse glucose at high insulin concentrations to keep blood glucose ‘clamped’ within the normal range. That glucose infusion rate, or GIR, is one key measure of the ability of insulin to shuttle glucose from the blood into cells (mostly muscle and adipose cells) and therefore a measure of systemic, whole body insulin sensitivity. Because the rate of glucose infusion is identical to the rate to which cells take up glucose under steady-state conditions (i.e., where insulin and glucose levels do not change much), GIR is often also referred to as the rate of glucose disappearance, or Rd.
During a clamp, the investigators usually also measure the ability of insulin to inhibit glucose release by the liver. This is sometimes called the glucose appearance rate, or Ra. Ra can be measured at low basal insulin concentrations, and during the hyperinsulinemic phase of the clamp.
Taken together, the two main measures of insulin sensitivity are systemic insulin sensitivity by assessing GIR/Rd and hepatic insulin sensitivity by assessing Ra.
Lundsgaard et al. observed lower systemic and muscle insulin sensitivity on the low-carb diet compared to the high-carb diet. However, hepatic insulin sensitivity was greater following the low-carb diet. So that’s why I rated this result as mixed in the table above.
Chokkalingam and colleagues found that after 6 days on the low-carb and high-carb diets, systemic insulin sensitivity was higher after the low-carb diet. Even though no difference in hepatic insulin sensitivity was seen, these data could be taken as evidence of improved insulin sensitivity following a low-carb diet.
A randomized controlled trial using a parallel design by Kirk and colleagues is also notable. Participants with obesity were randomized to a high-carb or low-carb diet, and studied after 48 hours and again at an equivalent weight loss of 7%. At matched weight loss, HOMA-IR was reduced to a greater degree in the low-carb diet, consistent with a greater improvement in hepatic insulin sensitivity on the low-carb diet. However, muscle and systemic insulin sensitivity improved similarly in both diet groups, without being significantly different.
In another parallel-design randomized controlled trial, Bradley and colleagues randomized patients with overweight or obesity to one of two calorie-restricted diets: a low-carb diet and a high-carb diet. Both were followed for 8 weeks, and led to similar reductions in body weight. Systemic and hepatic insulin sensitivity improved in both diet groups with the weight loss, with no differential change between the groups.
In a parallel-design study, Frankenberg and colleagues compared a very high-fat diet fairly low in carbs (27% of total energy intake) to a high-carb diet (62%), with protein and energy intake matched. All food was provided to participants in this trial for 4 weeks. Insulin sensitivity, as measured by clamp, was significantly higher in participants after the high-carb diet.
The largest of the studies considered here was conducted by Gardner and colleagues (the DIETFITS trial), enrolling 609 men and women with elevated BMI. They were randomized to a healthy low-carb or healthy high-carb diet, and followed these diets for an entire year. This study was less well controlled, in that no food was provided to participants, but still, the investigators managed to achieve a remarkable difference in carb intake between the groups for the entire 1-year period in the 481 participants who completed the study. The two diets led to similar reductions in body weight. This study was not primarily focused on insulin sensitivity, and it did not include a gold-standard measurement of insulin sensitivity. However, it reported fasting insulin levels and the insulin concentration at 30 min in a standardized oral glucose tolerance test, and both can be seen as low-quality indicators of insulin sensitivity. Both were reduced following weight loss in both groups, indicating improved insulin sensitivity in both groups, but the degree of change did not differ between the groups for either measure.
The publication by Aronica and colleagues was a secondary data analysis of the DIETFITS trial in a small subgroup of participants who consumed diets that were consistent with a low-carb ketogenic diet or an ultra-low-fat (and therewith high-carb) diet. We need to interpret such a secondary analysis with caution and give it less weight in our assessment of the cumulative evidence. But it’s worth noting that again HOMA-IR improved in both of these groups, with no differential change between the groups.
Another study of note that gives us some clues about whether high-carb diets may trigger insulin resistance is the OMNICARB trial (Sacks et al.). These investigators randomized 163 adults with overweight to consume two or more of four different diets, each for 5 weeks. This was a remarkably large cross-over design study, and all foods and drinks were provided to participants, which is unusual for a study of this size. The diets were either low or high in carbs, and either low or high in the glycemic index of these carbs. Therefore, if we compare the extremes, one diet was relatively lower in carbs (albeit at 40% of total calories not really low-carb) and contained only low-glycemic index foods, resulting in a glycemic load of 64. Another diet was high in carbs, and the high-carb foods were mostly high glycemic index foods, resulting in a glycemic load of 172, which was 2.7-fold higher than on the low-carb low-glycemic index diet. These authors measured glucose tolerance and insulin sensitivity by a standardized oral glucose tolerance test, and observed no significant differences between these two diets (or any other diets) in a measure of insulin sensitivity, the Matsuda-DeFronzo Insulin Sensitivity Index.
There are several more studies we could discuss here, but these either compared diets that were fairly similar in their carb content, such as a trial by Perez-Jimenez and colleagues or a publication from the Women’s Health Initiative led by Shikany; or they had major study design or reporting flaws, such as studies by Cutler et al. and Lovejoy et al.; or they were conducted in patients with manifest type 2 diabetes, such as trials by Garg et al. and Parillo et al. However, it is important to note that considering these studies together with those shown in the table above does not change our overall conclusions.
Summary & Conclusions
Considering the cumulative evidence, it does seem clear that low-carb diets consumed ad libitum usually lead to weight loss in people with overweight or obesity. We also have strong evidence that weight loss is clearly associated with an improvement in insulin sensitivity, particularly in people who lose the most visceral and ectopic fat.
However, such an insulin-sensitizing effect is seen with any weight loss, including weight loss from bariatric surgery, medication, or other diets, and it’s therefore critical to assess the impact of low- vs. high-carb diets on insulin sensitivity under weight-stable or weight-matched conditions. In studies that controlled body weight or matched the degree of weight loss, insulin sensitivity is not clearly higher on low-carb diets, and high-carb diets do not clearly cause insulin resistance.
Many of the available published studies are small, and others employed imperfect measures of insulin sensitivity, so the evidence isn’t perfect at this point. However, a relatively large number of studies used the gold standard hyperinsulinemic-euglycemic clamp technique to measure insulin sensitivity. Also, all food was provided in many studies, and important variables, such as energy or protein intake, were well-controlled in most studies. Therefore, while not fully conclusive, the available evidence suggests that the carbohydrate (and fat) content of the diet is not per se a major determinant of insulin sensitivity.
Now, why may it be that hyperinsulinemia in clamp studies or patients with insulinoma triggers insulin resistance, while eating a high-carb diet does not? Considering the substantial differences in carb content of the tested study diets in these studies, it is pretty certain that insulin concentrations throughout the day must have been higher on the high-carb diet arms.
For one, the nature of hyperinsulinemia in clamp studies and insulinoma patients is chronic, with insulin levels elevated for many hours to days. That’s very different from the regular ups and downs in blood insulin levels that occur with eating a high-carb diet.
It is also possible that other components of high-carb or low-carb diets balance out any effect the carb-induced elevations in blood insulin may have. Because when people eat these low-carb and high-carb diets, in these research studies as well as in real life, they don’t eat isolated carbs or fats. They eat lots of bread, rice, pasta, and fruit on the high-carb diet, or dairy, meat, nuts, oils, butter, and cream on the low-carb diet. We should be clear that even though investigators designed these diets to differ in their carb and fat content, the resulting diets also differ in numerous other ways, many of which could also affect insulin sensitivity. In other words, it is possible that the normal up and down in blood insulin levels that results from eating more carbohydrates does cause some insulin resistance, but that this effect is balanced out by some other component of the diets. That may also explain why there is so much variability in the study results. It basically suggests that some hidden factor, or factors, are at play, and that may well be simply because each study team may have implemented these diets slightly differently.
For this analysis, I have reviewed 21 scientific publications based on 20 clinical trials, and I cannot detect a specific pattern that explains why the results are so heterogeneous. I had a few hypotheses, such as that studies that used more refined carbs in their high-carb diet arms may have had worse outcomes than those that incorporated mostly complex fiber-rich sources of carbs in their high-carb diet. This does not seem to be the case, as most of the studies, as much as can be determined from the papers, seem to have used mostly whole grains, legumes, fruit, and vegetables as high-carb foods in their high-carb diets. I also wondered whether the fat, protein, or fiber contents may have affected the overall results. I cannot confidently rule this out, but I also could not detect a specific pattern.
In other words, from the existing studies, I cannot confidently determine the mechanisms that explain the wide-ranging findings. What we can say clearly, however, is that in weight-stable people, eating more or less carbs will have very little to no consistent impact on insulin sensitivity.
That may be disappointing to some because if lowering carb intake reliably improved insulin sensitivity, we would have a fairly easy tool in our toolbelt to treat insulin resistance. It could also be seen as a positive, however, because these data give people more options in their dietary choices without risking insulin resistance. In other words, if your diet is high in carbs, for whatever reason, there is no good reason to think that this will make you insulin-resistant per se, as is often asserted in low-carb circles.
I also want to be clear that these data should not be seen as suggesting that all high-carb diets are good choices for everyone. Many food sources of carbohydrates, such as refined grains and added sugars, are highly glycemic, low in fiber, and provide little in terms of micronutrients. The data discussed here should not give anyone the impression that eating such foods in large amounts has a neutral effect on their health. Most of the studies discussed above included whole food-based sources of carbohydrates, such as whole grains, legumes, vegetables, and fruit. It shouldn’t come as a surprise, but is worth explicitly stating that investigators in these dietary intervention trials were not feeding their participants mostly candy and soda on their high-carb diets. It is therefore important to be clear that the impact on insulin sensitivity (and other health metrics) of a high-carb diet rich in refined grains and added sugars may well be worse than these data suggest. It is also clear that high-carb diets may be problematic for people with glucose intolerance, such as pre-diabetes and particularly diabetes mellitus.
As importantly, these data should not be seen as suggesting that low-carb diets do not offer any benefits for blood sugar regulation. The fact that low-carb diets consistently help people lose weight is a key benefit, and the low glycemic load of low-carb diets is an intrinsic advantage for people with impaired glucose intolerance. Simply put, if somebody is glucose intolerant, that means they cannot safely keep blood sugar levels within the normal range, and in that case, reducing the amount of glucose that needs to be handled after every meal would seem like a good idea. We’ll talk about this in much detail in a later blog post.
To conclude, the available data strongly suggest that continuously high blood insulin concentrations, as in a hyperinsulinemic clamp or in patients with an insulinoma, causes insulin resistance in these settings. However, the physiologically normal up and down in blood insulin levels that occurs with eating a high-carb diet does not seem to independently trigger insulin resistance. And neither do low-carb diets seem to provide a unique insulin-sensitizing effect when accounting for changes in body weight or fat mass.
Allow me one last remark: I commonly see people argue that carbs cause insulin resistance, or that low-carb diets reverse insulin resistance by citing one or two specific studies. I have also seen claims that high-carb diets improve insulin sensitivity. Many people seem to think that working scientifically means that if you make a statement of “x causes y”, all you need to do is find one or two papers in support of this statement. That is not at all how this works. As you can see in the table above, it is possible to find one or two studies in support of any position. It is critical to be clear that this is not an evidence-based approach. Evidence-based means that we are considering the entirety of the cumulative evidence, not just the few studies that support our opinion. Please protect yourself against misinformation by looking out for that.
- Gastaldelli. Measuring and estimating insulin resistance in clinical and research settings. Obesity 2022; 30: 1549-63.
- Del Prato et al.; Effect of sustained physiologic hyperinsulinemia and hyperglycemia on insulin secretion and insulin sensitivity in man. Diabetologia 1994; 37: 1025-35.
- Rizza et al.; Production of insulin resistance by hyperinsulinemia in man. Diabetologia 1995; 28: 70-5.
- Skrha et al.; Comparison of insulin sensitivity in patients with insulinoma and obese type 2 diabetes mellitus. Hormone and Metabolic Research 1996; 28: 595-8.
- Pontriroli et al.; The glucose clamp technique for the study of patients with hypoglycemia: insulin resistance as a feature of insulinoma. Journal of Endocrinology Investigations 1990; 13: 241-5.
- Furnica et al.; A severe but reversible reduction in insulin sensitivity is observed in patients with insulinoma. Annales d’Endocrinologie 2018; 79: 30-6.
- Saiki et al.; Reduction of visceral adiposity after operation in a subject with insulinoma. Journal of Atherosclerosis and Thrombosis 2004; 11: 209-14.
- Colica et al.; Efficacy and safety of very-low-calorie ketogenic diet: a double-blind randomized crossover study. European Review for Medical and Pharmacological Sciences 2017; 21: 2274-89.
- Paoli et al.; Effects of a ketogenic diet in overweight women with polycystic ovary syndrome. Journal of Translational Medicine 2020; 18: 104.
- Cohen et al.; A ketogenic diet reduces central obesity and serum insulin in women with ovarian or endometrial cancer. Journal of Nutrition 2018; 148: 1253-60.
- Luukkonen et al.; Effect of a ketogenic diet on hepatic steatosis and hepatic mitochondrial metabolism in nonalcoholic fatty liver disease. Proceedings of the National Academy of Sciences U.S.A. 2020; 117: 7347-54.
- Volek et al.; Comparison of very low-carbohydrate and low-fat diets on fasting lipids, LDL subclasses, insulin resistance, and postprandial lipemic responses in overweight women. Journal of the American College of Nutrition 2004; 23: 177-84.
- Ballard et al.; Dietary carbohydrate restriction improves insulin sensitivity, blood pressure, microvascular function, and cellular adhesion markers in individuals taking statins. Nutrition Research 2013; 33: 905-12.
- Bisschop et al.; Dietary fat content alters insulin-mediated glucose metabolism in healthy men. American Journal of Clinical Nutrition 2001; 73: 554-9.
- Aronica et al.; Weight, insulin resistance, blood lipids, and diet quality changes associated with ketogenic and ultra-low-fat dietary patterns: a secondary analysis of the DIETFITS randomized controlled trial. Frontiers in Nutrition 2023; 10: 1220020.
- Lundsgaard et al.; Opposite regulation of insulin sensitivity by dietary lipid versus carbohydrate excess. Diabetes 2017; 66: 2583-95.
- Chokkalingam et al.; High-fat/low-carbohydrate diet reduces insulin-stimulated carbohydrate oxidation but stimulates nonoxidative glucose disposal in humans: an important role for skeletal muscle pyruvate dehydrogenase kinase 4. Journal of Clinical Endocrinology and Metabolism 2007; 92: 284-92.
- Kirk et al.; Dietary fat and carbohydrates differentially alter insulin sensitivity during caloric restriction. Gastroenterology 2009; 136: 1552-60.
- Bradley et al.; Low-fat versus low-carbohydrate weight reduction diets: effects on weight loss, insulin resistance, and cardiovascular risk: a randomized controlled trial. Diabetes 2009; 58: 2741-8.
- Gardner et al.; Effect of low-fat vs low-carbohydrate diet on 12-month weight loss in overweight adults and the association with genotype pattern or insulin secretion: the DIETFITS randomized controlled trial. Journal of the American Medical Association 2018; 319: 667-79.
- Sacks et al.; Effects of high vs low glycemic index of dietary carbohydrate on cardiovascular disease risk factors and insulin sensitivity: the OmniCarb randomized clinical trial. Journal of the American Medical Association 2014; 312: 2531-41.
- Von Frankenberg et al.; A high-fat, high-saturated fat diet decreases insulin sensitivity without changing intra-abdomial fat in weight-stable overweight and obese subjects. European Journal of Nutrition 2017; 56: 431-43.
- Perez-Jimenez et al.; A Mediterranean and high-carbohydrate diet improve glucose metabolism in healthy young persons. Diabetologia 2001; 44: 2038-43.
- Shikany et al.; Effects of a low-fat dietary intervention on glucose, insulin, and insulin resistance in the Women’s Health Initiative (WHI) Dietary Modification trial. American Journal of Clinical Nutrition 2011; 94: 75-85.
- Garg et al.; Comparison of effects of high and low carbohydrate diets on plasma lipoproteins and insulin sensitivity in patients with mild NIDDK. Diabetes 1992; 41: 1278-85.
- Parillo et al.; A high-monounsaturated-fat/low-carbohydrate diet improves peripheral insulin sensitivity in non-insulin-dependent diabetic patients. Metabolism 1992; 41: 1373-8.
- Cutler et al.; Low-carbohydrate diet alters intracellular glucose metabolism but not overall glucose disposal in exercise-trained subjects. Metabolism 1995; 44: 1264-70.
- Lovejoy et al.; Effect of a controlled high-fat versus low-fat diet on insulin sensitivity and leptin levels in African-American and Caucasian women. Metabolism 1998; 47: 1520-4.
- Ge et al.; Comparison of dietary macronutrient patterns of 14 popular named dietary programmes for weight and cardiovascular risk factor reduction in adults: systematic review and network meta-analysis of randomized trials. British Medical Journal 2020; 369: m696.
- Goodpaster et al.; Effects of weight loss on regional fat distribution and insulin sensitivity in obesity. Diabetes 1999; 48: 839-47.
- Magkos et al.; Effects of moderate and subsequent progressive weight loss on metabolic function and adipose tissue biology in humans with obesity. Cell Metabolism 2016; 23: 591-601.
- Johnson et al.; Mechanism by which caloric restriction improves insulin sensitivity in sedentary obese adults. Diabetes 2016; 65: 74-84.