A Comparison of Ketogenic versus Plant-Based Diet and its Effects on Glucose and Insulin Response in Type 2 Diabetics

Although new cases of diabetes have decreased over the last decade, the overall percentage of people in the United States with diabetes has continued to increase at an alarming rate. The CDC recently published its National Diabetes Statistic Report for 2020, which detailed 1 in 10 Americans having diagnosed diabetes, and 1 in 3 having prediabetes¹. Sadly, the percentage of children and adolescents with diabetes has also been on the rise. The CDC¹ reports that new cases of type 1 and type 2 diabetes have significantly increased, with type 2 seeing the highest growth in non-Hispanic blacks between ages 10-19.  

Diabetes is caused by the body’s inability to create or effectively utilize its insulin, ultimately impairing glucose control and cellular function. Diabetes can cause a myriad of complications, including kidney disease, stroke, heart disease, blindness, impotence, and amputation, among many others. Sadly, it can be a deadly disease, and in 2017 it was listed as the 7th leading cause of death in the United States and was documented as an underlying or contributing cause of death on over 270,702 death certificates². In addition, healthcare costs surrounding diabetes treatment are rising. In 2017 alone, the cost of diabetic care was over 130 billion dollars, often due to billing from ER visits and hospitalizations as a result of uncontrolled diabetes².

Based on the 2020 National Diabetes Statistics Report¹, 89% of adults diagnosed with diabetes were overweight, and 38% were physically inactive. As we move further into the 21^st century, people are relying on quicker, more convenient ways to get their food. Drive-through and fast-food consumption have increased and based on a study published in 2018, around 37% of Americans consume at least one meal of fast food on a given day. Interestingly, there was noted to be a higher percentage of fast-food ingestion among the non-Hispanic black population, the same population that has had the most significant rise in new cases of diabetes3. Also, there was a direct correlation between higher household income and increased frequency of fast-food intake. Unfortunately, fast food meals are high calorie with little nutritional value and directly contribute to increased weight and the development of medical comorbidities. 

While Type 1 Diabetics need exogenous insulin through the form of self-injection or a pump to survive, many people with Type 2 Diabetes can control their glucose and sometimes even reverse their diabetes with dietary modifications and other lifestyle changes. There are many diets that have come and gone over the years that touted their ability to reverse diabetes and provide the best glycemic control. It seems today that two diets are continuously making headlines; the Ketogenic Diet and the Plant-Based diet. This is an interesting comparison, as while sharing some fundamental similarities, they also promote vast differences in what foods can offer the best glycemic control and overall improvements in health. The goal of this study is to provide a thorough look into these diet styles and determine which will ultimately provide the most benefit for the diabetic population. For the purpose of this study, we will look primarily into the glycemic effects of these diets, and if there is one that outperforms the other. 

Methodology

A literature review search was performed using electronic databases such as ProQuest, NCBI, MDPI, and the LMU Reed Health Sciences Library. Additionally, video conferences were viewed via YouTube, including lectures from the “Grand Rounds presentations” hosted by the Institute for Functional Medicine in collaboration with Cleveland Clinic Center for Functional Medicine. Studies were selected using advanced searches, specifically with keywords such as “glycemic impact”, “diabetes”, “Ketogenic diet”, “low carbohydrate”, “plant-based”, and “vegan”. 

Ketogenic Diet 

Initially formulated as a dietary approach to treat pediatric patients with epilepsy, the Ketogenic diet was established in the early 20^th century. Then, prior to the discovery of insulin in 1921, ketogenic or low carbohydrate diets were used for diabetes treatment because of their rapid ability to reduce blood glucose.⁴ In recent years we have seen a steep resurgence of “Keto” that has developed almost a cult-like following in the weight loss and fitness community. The ketogenic diet is based upon high levels of dietary fats, moderate proteins, and low carbohydrates. For reference, in a 2000 kcal per day diet, a person would only consume between 20-50 g of carbohydrates. Another example would be a macronutrient breakdown of 60% fat, 30% protein, and 10% carbohydrates5. Carbohydrates in a ketogenic diet include non-starchy vegetables, nuts, and seeds, a limited amount of berry fruit, and dairy. Ketogenic followers are strongly advised to avoid grains, potatoes, and sugar. Keto advocates promote foods that work with their physiology and ensure that the foods remain filling and enjoyable while staying under 20-50 g of carbs per day. 

In type 2 diabetes, insulin is elevated and associated with insulin resistance, which in turn causes elevations in blood glucose. When looking at the breakdown of macronutrients, carbohydrates cause insulin spikes as they are directly broken down into carbohydrates. Protein causes a delayed insulin response, generally around 2 hours after consumption as they take longer to be broken down to glucose; however, fat intake has minimal post-prandial effects on insulin levels as fat does not readily break down into glucose except in instances of gluconeogenesis. Therefore, in a ketogenic diet where fat is the predominant macronutrient consumed, glucose production and insulin spiking are significantly reduced. Additionally, ketogenic diets have been found to result in weight loss, as the development of ketosis stimulates a physiological state of fasting.⁶

A Low-Carbohydrate, Ketogenic Diet to Treat Type 2 Diabetes⁷ (Yancy et al.)

This study evaluated the outcomes of a low carbohydrate, ketogenic diet (LCKD) in 28 participants with type 2 diabetes over the course of 16 weeks. The primary goals of this study were to determine the glycemic response in these patients, which was measured primarily by HgA1c, in addition to reduction and/or cessation of diabetic medication, and the percentage of body fat loss. Participants were selected via the following parameters: age between 35-75, BMI >25%, and fasting serum glucose >125 along with A1c >6.5%. This study excluded participants with evidence of kidney, liver, or unstable cardiac disease. The participants in this study were educated on how to select or prepare foods that followed a LCKD, and given a carbohydrate goal of less than 20 g per day. In addition to an unlimited amount of protein, the participants were advised to have an unrestricted amount of fats and oils, other than minimizing trans-fat consumption. Diabetes medication was reduced on day 1 as well, and adequate fluid intake and increased aerobic exercise were encouraged. 

Participants were required to keep a food log, which was reviewed by a dietitian at baseline, and then at weeks 2, 8, and 16. Progress throughout the study was tracked bi-weekly through various examinations and laboratory measures. This included measuring urine ketones, the development of hypoglycemic episodes or other symptomatic side effects, weight, blood pressure, and heart rate. Additionally, blood draws were performed at various stages to assess blood and electrolyte counts, kidney function, lipid, and thyroid panels. The primary value that this study focused on was the change in HbA1c before and after the study. 

The outcomes of the study revealed that in addition to overall weight loss, the participant’s A1C measurements were reduced significantly as a result of the ketogenic diet (mean HbA1c of 7.4% at week 0 and 6.3 at week 16), which occurred in the setting of reducing or discontinuing diabetic medications. Fasting glucose was also impacted, with a decrease of 17%. 

Effectiveness and Safety of a Novel Care Model for the Management of Type 2 Diabetes at 1 Year ⁸ (Hallberg et al.)

Three hundred and forty-nine adults with DM2 were enrolled in a 1-year study that evaluated the before and after effects of patients enrolled in a continuous care intervention group, versus a usual care group. Primary outcomes were measured via HbA1c, weight, and medication usage. Secondary outcomes measured fasting glucose and insulin levels, lipid panel, liver and kidney function, and CRP. The purpose of the study was to evaluate the effectiveness of the new continuous care model in addition to a low carbohydrate diet that could safely decrease participants’ HbA1c, weight, and need for medications at the 1-year mark. 

After 1 year, 262 of the participants who enrolled in the CCI model had, on average, a reduction in HbA1c from 7.6 to 6.3, lost 12% of their body weight, and reduced medication use. Ninety-four percent of patients on insulin were able to successfully reduce or discontinue their dosing, and all participants were able to eliminate their use of sulfonylureas. In comparison, the usual care group had no reduction in HbA1c, weight, or medications over the course of the year. 

Effect of the Ketogenic Diet on Glycemic Control, Insulin Resistance, and Lipid Metabolism in Patients with T2DM: a Systematic Review and Meta-Analysis⁶ (Yuan et al.)

In this publication, 13 studies were evaluated to determine the before and after-intervention effects of a ketogenic diet on glucose metabolism, lipid metabolism, and weight control. The biomarker for glucose measurements was fasting glucose and HbA1c. After evaluation, it was noted that fasting blood glucose was decreased by 1.29 mmol/L after a ketogenic diet intervention. Regarding HbA1c, a difference of -1.07% was found after the implementation of the ketogenic diet. It should be noted that this is regarded as the ideal therapeutic effect for diabetes medications. Because there is a strong relationship between obesity to DM2, effects on body weight will be included. The average weight reduction was 8.66kg, waist circumference decreased by an average of 9.17cm, and BMI reduced by 3.13kg/m2. 

Insulin Sensitivity and Glucose Tolerance are Altered by Maintenance on a Ketogenic Diet⁹ (Kinzig et al.)

The focus of this study was to determine the responsivity to peripheral and central insulin, glucose tolerance, and meal-induced effects of consuming a ketogenic diet versus rodent chow in a group of rats. The rats were all-male, 8 weeks old upon the start of the experiment, and weighed between 200-225 g. Upon initial arrival to the laboratory, all rats were contained in individual cages that had the temperature maintained at 25C, along with 12-hour light and 12-hour dark cycles. They were initially given an ad libetum diet of rodent chow but then were split into their respective group. The ketogenic group was placed on a diet in which 5% of calories were from carbohydrates, 80% from fat, and 15% from protein. The control group continued their ad libetum rodent chow diet. The caloric intake and weights of the rats were measured daily. 

After 8 weeks, multiple experiments were performed on groups of rats from each diet cohort. In regard to the effects of glucose tolerance, intraperitoneal glucose tolerance tests were performed. Diets were discontinued 16 hours prior to the testing, and body weight was used to calculate glucose dosing. After glucose injection, the blood samples were obtained at 15, 30, 45, 60, and 120 minutes. The baseline insulin was significantly lower in the ketogenic rats, however, 15 minutes after the glucose injection, insulin levels rose significantly greater than in the chow group. The plasma insulin returned to baseline in the chow group at 120 minutes post-injection, however, in the ketogenic group, the insulin levels remained elevated. Interestingly, although insulin was spiked in the ketogenic group, there were no major differences in blood glucose levels between the two groups. 

Insulin tolerance testing was performed to evaluate the effect of insulin on blood glucose between the two groups. In the chow-fed group, the glucose was significantly decreased 15 minutes after insulin injection and remained suppressed for the remainder of the 120 minutes. In contrast, the glucose in the ketogenic group was not reduced from baseline until 30 minutes after injection but was not reduced to the level of the control. This response demonstrates less insulin-induced hypoglycemia in the ketogenic rats. 

Effects of the glucose and insulin levels were additionally assessed after a test meal. Chow-fed rats were given access to vanilla-flavored Ensure which was 67.6% carbohydrates, while ketogenic rats were given a vanilla-flavored Atkins shake of 7.7% carbohydrates. There was less of an insulin effect after ingestion of the low-carbohydrate Atkins supplement by the ketogenic rats as compared to the insulin effect of the ingestion of the high-carbohydrate test meal by the chow group. When the ketogenic group consumed the high carbohydrate Ensure test meal, the plasma insulin was increased in the first 15 minutes and remained elevated for the remainder of the session. In the chow group, the insulin level peaked at 15 minutes and slowly decreased over 120 minutes. When measuring the blood glucose response, the ketogenic rats consuming the Atkins meal had a minimal increase in blood glucose and stayed fairly steady through the 120 minutes. When consuming the Ensure meal, the ketogenic group had a steady, significant rise in blood glucose during the entirety of the test. The chow rats responded to the Ensure meal with an initial significant rise in glucose with a peak at 60 minutes and then levels decreasing at 120 minutes. 

Plant-Based Diet

Plant-based diets have also become increasingly popular, especially after the boom of documentaries such as “What the Health”, “Forks over Knives” and “Food Inc”, where viewers are often forced to reflect on their food choices based on the shock value of animal processing. Although ethical responsibility is a large factor in the vegan or plant-based lifestyle, there are many other reasons that people consider consuming fewer animal products. Plant-based diets focus on reducing the intake of animal products, and in many cases, eliminating the consumption of all meat, fish, dairy, or eggs. The mainstays of a plant-based diet include vegetables, whole grains, legumes, beans, fruit, nuts, and seeds. Similar to the ketogenic diet, the reduction of highly processed foods and refined sugars is also highly advocated. 

Some of the health claims associated with a plant-based diet include lower body weight, lower cholesterol, and blood pressure, along with a reduced risk of heart disease and diabetes. Many studies have shown that a plant-based diet can prevent the development of type 2 diabetes when compared to a non-vegan diet,10 and improvements in insulin resistance and HgA1c have also been documented with the implementation of a plant-based diet.^{11 12 13} 

The notion that a plant-based diet can prevent or reverse type 2 diabetes stems partially from the reduction in body fat, which is a known risk factor for diabetes, as weight loss improves insulin resistance. Additionally, consuming higher-fiber foods that slow the breakdown of carbohydrates can improve insulin signaling and sensitivity. Saturated fat, which is found in meat and animal products, can cause fat accumulation in hepatic and skeletal muscle cells which in turn impairs insulin signaling and reduced glucose uptake.¹⁴ 

A Low Fat Vegan Diet Improves Glycemic Control and Cardiovascular Risk Factors in a Randomized Clinical Trial in Individuals With Type 2 Diabetes¹¹ (Barnard et al.) 

Dr. Barnard is one of the most well-known names in the plant-based community. He has pioneered multiple studies, performed Ted Talks, and has a podcast series dedicated to improved health through veganism, specifically a whole food plant-based diet. In his 2006 study, individuals with type 2 diabetes were randomly assigned to either a low-fat vegan diet or a diet by ADA (American Diabetes Association) guidelines and evaluated over the course of 22 weeks for changes in their glycemic control and cardiovascular status. Exclusion criteria for this study were established based on an A1c of <6.5 or > 10.5, use of insulin for greater than 5 years, smoking, alcohol or drug abuse, pregnancy, unstable medical condition, or current following a low-fat vegetarian diet. After screening 1049 individuals by phone, 99 met the inclusion criteria and they were randomly assigned to a dietary group. The vegan diet proposed to the participants was comprised of eating around 10% fat, 15% protein, and 75% carbohydrates, with an emphasis on whole grains, vegetables, fruits, and lower glycemic index foods. There were no restrictions set on portion sizes or caloric intake. In comparison, the ADA diet participants were recommended to consume around 15-20% protein, <7% saturated fat, 60-70% carbohydrate and monosaturated fats, and cholesterol <200mg day. This diet was individualized based on the participant’s weight and lipid profile. Any participants in the ADA group with a BMI over 25 were advised to keep a 500-1000kcal/day deficit. 

Both groups were advised to not make any changes to their exercise regimens, and they were required to make meals on their own. The participation did, however, meet with a registered dietitian initially and weekly thereafter to discuss meal options and for cooking demonstrations. Participants were required to make both scheduled and unannounced 24-hour food recall journals. HgA1c and body weight were measured at 0, 11, and 22 weeks, while plasma glucose, lipid panel, and blood pressure were measured at 0 and 22 weeks. At the end of the 22 weeks, of the participants who met complete adherence criteria, the A1c changes reflected -1.20% for the vegan group, and -0.88% in the ADA group. In participants whose diabetes medications remained unchanged throughout the experiment, the vegan group had an A1c reduction of -1.23 percentage points and a reduction of -0.38 in the ADA group. 

A Low-Fat Vegan Diet and a Conventional Diabetes Diet in the Treatment of Type 2 Diabetes: a Randomized, Controlled, 74-wk Clinical Trial¹² (Barnard et al.)

In a follow-up study by Dr. Barnard and his team, the initial group evaluated in the 2006 report11 was followed for an additional 52 weeks to determine the long-term effects of a vegan diet. The participants were evaluated at weeks 0, 11, 22, 35, 48, 61, and 74 for glycated hemoglobin and plasma lipids, and weight was measured at weeks 0, 22, and 74. HgA1c changes from baseline after the 74-week study period were -0.34 for the vegan diet and -0.14 for the ADA diet. The weight changes between the two groups from baseline were -4.4kg and -3.0kg in the vegan and ADA diet groups respectively. The changes in glycated hemoglobin appeared to be more related to weight loss, as there was not a significant difference between the two groups at the conclusion of the study. 

Vegetarian Diets and Incidence of Diabetes in the Adventist Health Study-2¹⁰ (Tonstad et al.)

The Seventh Day Adventists are a religious group of now over 14 million people worldwide¹⁵ who are studied frequently by medical researchers. The community is made up of members who overwhelmingly abstain from smoking, alcohol, drugs, and consuming meat. In this specific study, 15,200 men and 26,187 women throughout the US and Canada were evaluated over a period of 2 years. Prior to the study, all participants were free of diabetes. Participants were given an initial extensive questionnaire, which evaluated medical history, frequency of consumption of various foods, activity level, and smoking and alcohol use among many others. Participants were broken down into a vegetarian status, (which included vegan, lacto-ovo vegetarian, pesco-vegetarian) or non-vegetarian status. After two years, the participants were sent a follow-up questionnaire regarding their development of diabetes. The results of the study showed a substantial reduction in the risk of diabetes development in the vegetarian diet groups. In non-Black participants, vegan, lacto-ovo, and semi-vegetarian diets were associated with a decreased risk. However, in Black participants, only vegan and lacto-ovo vegetarian diets had decreased risk. The two-year incidence of diabetes development was 0.54% in the vegan group, 1.08% of lacto-ovo vegetarians, 1.29% of pesco-vegetarians, 0.92% of semi-vegetarians, and 2.12% in non-vegetarians.  

Effect of a Plant-Based, Low Fat Diet versus an Animal-Based, Ketogenic Diet on an Ad Libitum Energy Intake¹³ (Hall et al.)

In one of the most recent studies comparing a plant-based versus ketogenic diet, the NIH evaluated 20 adults in a highly controlled study to determine the effects of each diet. The participants had an average age of 30, and a BMI of 28, with no history of diabetes. During the 4-week study, participants were admitted as inpatients to the National Institutes of Health Clinical Center and randomized to consume either a plant-based, or ketogenic diet ad libitum (without restriction). It should be noted that both diets consisted of minimally processed, whole foods, and non-starchy vegetables. Each group was on their assigned diet for 2 weeks, evaluated, switched diets, and were again evaluated after 2 weeks. 

All participants wore continuous glucose monitors throughout the study. Postprandial glucose was consistently higher in the plant-based group, with a mean glucose 2hr after meals of 102.5mg/dL, as compared to the ketogenic group of 80.5mg/dL. At the end of each 2-week phase, oral glucose tolerance was measured. The ketogenic group showed a relative impairment of glucose tolerance, with a mean OGTT of 143.3mg/dL, while the plant-based group had a mean OGTT of 115.6mg/dL.

Caloric intake was on average 550-700kcal/day less when participants were in the plant-based group, despite reporting no significant differences in hunger, fullness, or satiety. Bodyweight decreased in both diets, while the ketogenic group had a more rapid weight loss during week 1, and total weight loss after week 2 was 1.77kg. In comparison, the plant-based group had a slower weight loss rate, but after two weeks the total loss was 1.09kg. Interestingly, most of the weight change in the plant-based group appeared to be in a decrease in body fat based on DEXA scan.

Discussion 

The primary aim of this study was to determine the response of insulin levels and glucose control with the application of a ketogenic or plant-based diet, and if one proved superior to the other. Studies were selected to reflect three different categories: biomarker response to diet in patients with diagnosed type 2 diabetes, biomarker response to diet in patients over time to determine the risk of type 2 diabetes development, and biomarker response in patients without diabetes on strictly controlled ketogenic vs plant-based diets. This variety was chosen to better comprehend how dietary adjustments can affect the progression of diabetes and the overall metabolic response of insulin and glucose in the opposing groups. 

Regarding the studies evaluating participants with type two diabetes, there appeared to be a similar initial effect on blood glucose reduction with both diet modifications. In the 16-week study⁷, there was a drop in HgA1c from 7.4-6.3% (-1.1) and a blood glucose level from 163.6-136.4mg/dl (-27.2mg/dl) after the implementation of a ketogenic diet. Whereas in a 22-week plant-based diet, adherent participants saw a drop from 8.07-6.84% (-1.23) and a reduction in fasting plasma glucose from 177.4-128.2mg/dl (-49.2mg/dl).11 When comparing diabetic participants through a greater duration of time, there appeared to be more sustained decreases in biomarkers, and significant medication reduction in the ketogenic group, as the 1-year study by Hallberg⁸ demonstrated.   

The teams of both Dr. Hallberg and Dr. Barnard provided continuous follow-up, education, and management of the participants, so why might there have been a greater sustained reduction in biomarkers after 1-2 years of study? This could potentially stem from easier access to food that fits in the “ketogenic” plan, in both private and social settings. It is idealistic to think that participants are making all meals at home and from scratch, and maintaining a variety of healthier options that falls within a diet plan can be quite difficult when out to dinner with family or friends or at social gatherings. While it may be easier in large cities to find restaurants that cater to plant-based patrons, most areas have a large gap in food choices. Whole food, ketogenic meals can be easily ordered and have high satisfaction in taste and variety. 

In the long-term studies of diabetic risk as represented in the Adventist Health Study, ¹⁰ participants that observed vegan or vegetarian diets observed a substantial reduction in risk of diabetes development compared to their non-vegetarian counterparts. There was also an incremental risk of development as the diet pattern moved from vegan to lacto-ovo vegetarian, to pesco-vegetarian, and so on. This, therefore, demonstrates the importance of preventative dietary adjustments, and a reduction in meat consumption can have some positive benefits regarding the future development of diabetes. 

Regarding the direct comparison of biomarker effects in the 2020 NIH study,¹³ the low carbohydrate diet had a lower glycemic load which was directly reflected in the CGM data showing lower postprandial glucose and insulin levels compared to the low-fat diet. Even though the evaluations were performed over 2-week time periods, there were observed physiologic adaptions. After 2 weeks of implementing a low carbohydrate diet, participants had impaired glucose tolerance compared to the low-fat group. These findings were mirrored in the study by Kinzig,9 as the rats on the ketogenic diet demonstrated decreased insulin sensitivity. The insulin levels in response to a glucose challenge far surpassed the chow-fed group, despite blood glucose levels remaining overall indistinguishable.  Therefore, it appears that long-term maintenance of a ketogenic diet may lead to increased peripheral insulin resistance. 

Health can be found through different forms of extreme dieting or exercise regimens, and each group has its unwavering proponents. This study has shown that both the ketogenic diet and the plant-based diet can have positive effects on blood glucose and insulin levels to better control type two diabetes. There are vast differences in diet preference, access to food, overall health, and potential restrictions that could lead to one diet being preferred over another. It is therefore imperative in the role of a healthcare provider to have thorough discussions with patients to determine how they would best succeed in healthier living. Additionally, education for the younger generations at the greatest risk of diabetes development should be implemented in schools, pediatric offices, and the media. 

The major weaknesses of the majority of these studies were the lack of racial, ethnic, or socioeconomic variation in the participants. Additionally, genetic factors were unable to be addressed. 

The purpose of this study was to solely evaluate the effect on glucose and insulin response; however, those are just two aspects of diabetes. Future studies could include effects on metabolic syndrome as a whole and the long-term effects and sustainability of each diet. Many of these articles evaluated biomarkers such as lipid, cardiac, kidney, and liver enzymes, along with blood pressure and weight measurements- and could therefore be utilized in future research comparisons.