Ketogenic diets: metabolism, health, and performance
Follow the evidence from measured ketosis to appetite, blood glucose, LDL and ApoB, exercise performance, and medication safety.
Fasting Free educational synthesis. Independent clinical review of this site content has not been completed.
THE QUESTION BEHIND THE HEADLINE
What can a ketone result tell us about whether a diet is working?
A ketone result describes part of fuel metabolism. Judging a diet also requires a defined outcome: appetite, body fat, glucose control, cardiovascular markers, performance, or everyday feasibility. These outcomes can move in different directions. The studies below help explain which conclusions follow from the measurements and which questions remain open.
Start a keto paper by writing down the claim it could realistically test. A supervised feeding experiment can measure food intake accurately. A dietary trial in people with diabetes can compare glucose markers. A training study can assess a particular sporting task. Combining these perspectives is useful, provided that a result in one setting does not silently become a promise in another.
This guide follows six questions through original studies and a sports-nutrition position statement. Each answer identifies the measured outcome, explains the comparison, and keeps the important uncertainty beside the finding. The examples are hypothetical exercises in interpretation. They are not patient results, eating plans, or instructions to change medicines.
BEFORE COMPARING RESULTS
Get the terms right.
- Ketogenic diet
A dietary pattern intended to produce nutritional ketosis. Researchers should report the prescribed diet, actual intake, and evidence of ketosis; the label alone does not establish that participants maintained it.
Read the evidence5- Blood ketones
A concentration influenced by ketone production and use. Beta-hydroxybutyrate is commonly measured. A higher reading is not a scale of fat loss, fitness, or overall benefit.
Read the evidence5- Fat oxidation
The use of fat as fuel during the measurement period. Dietary fat can contribute to that fuel, so increased oxidation alone cannot establish a reduction in stored body fat.
- LDL cholesterol and ApoB
LDL cholesterol describes cholesterol carried in LDL particles. ApoB provides information about the number of atherogenic particles. They answer related questions, and a study reporting one has not necessarily measured the other.
Read the evidence3- Exercise economy
The oxygen or energy required to sustain a given pace or workload. An athlete can improve peak aerobic capacity while needing more oxygen at race pace, so the measures must be interpreted separately.
Read the evidence4- Relative change and percentage points
A relative change divides the difference by the starting value. Percentage points describe the subtraction between two percentages. Keto-Med’s figure correction makes this distinction especially relevant when reading its HbA1c results.
Read the evidence2
CLAIM BY CLAIM
What can the research support?
How do researchers establish that a diet is ketogenic?
Look for evidence about both the food pattern and ketosis. A carbohydrate target describes the intervention; a ketone measurement helps establish the metabolic response.
A useful evidence table separates what participants were asked to eat, what they actually ate, and what researchers measured. The ISSN review accepted studies using several criteria, including carbohydrate restriction or ketone evidence. Consequently, a collection of studies called ketogenic does not guarantee identical verification in every experiment.
Ask when blood samples were taken and whether measurements were repeated. A single result cannot describe every day of a long intervention. Also check whether food was supplied or intake was self-reported. These details help distinguish a tightly tested dietary exposure from an instruction that participants followed to varying degrees.
Ketogenic eating and fasting are separate experimental choices. A diet study with regular meals cannot establish that adding long periods without food improves its results. When evaluating a proposed combination, look for a comparison that actually tested the combination rather than borrowing separate findings from unrelated interventions.
Keep this boundary in view
Do not turn a research definition into a personal ketone target. This guide does not evaluate ketone supplements or infer ketosis from hours on a timer.
Saved only in this browser, after you choose Save.
Does burning more fat mean losing more body fat?
Fuel selection, calorie intake, and changes in stored body fat are different outcomes. Each needs its own measurement.
Hall’s inpatient crossover directly measured how much provided food people ate. Participants consumed fewer calories on the plant-based low-fat diet than on the animal-based ketogenic diet, despite established ketosis in the latter condition. This challenges a simple claim that ketosis must produce the lowest spontaneous energy intake under every comparison.
The experimental strength is close control of food and measurement. The limitation is that the diets differed in several ways at once, including food sources, fiber, and energy density. The result therefore belongs to these two whole dietary patterns. It cannot assign the entire intake difference to a single hormone, nutrient, or food property.
Read the body-composition analysis separately. The between-diet difference in total fat lost by the end of each period was not statistically clear, although an analysis of the rate of fat loss favored low fat. Those are distinct analyses. Early scale changes and lean-mass estimates also require attention to water and glycogen, particularly when carbohydrate intake changes.
Keep this boundary in view
Two-week diet periods cannot establish long-term weight maintenance. A statistically clear intake difference does not automatically make every secondary body-composition comparison conclusive.
Saved only in this browser, after you choose Save.
What did Keto-Med show about blood glucose?
Its primary HbA1c comparison did not establish a clear difference between diets. Some exploratory glucose measures and sensitivity analyses favored keto, which adds context without replacing the primary result.
Keto-Med compared two supported dietary patterns that shared several food-quality features. Both emphasized nonstarchy vegetables and restricted added sugars and refined grains. The key comparison was whether further restrictions, including avoiding legumes, most fruit, and whole grains, added benefits or tradeoffs beyond those shared features.
Improvement from baseline on both diets does not by itself identify which change produced it. Participants also changed food choices, received counseling, and lost weight. The randomized comparison between diets addresses a narrower question than a before-and-after photograph or a personal glucose trace.
A nonsignificant difference is not proof that the diets are interchangeable. The sample was small, medication handling complicated interpretation, and COVID-19 disrupted the planned timetable. Read the primary analysis alongside the sensitivity analyses, then ask whether a favorable glucose signal answers the broader health question you actually care about.
Keep this boundary in view
The paper reports biomarkers, not demonstrated reductions in diabetes complications or cardiovascular events. Its results should not be used to reproduce the study’s medication adjustments without clinical supervision.
Saved only in this browser, after you choose Save.
Can improved triglycerides resolve concern about a higher LDL result?
A favorable result for one marker leaves the other marker to be assessed. Check which lipid measurements the paper actually reports and avoid assuming that an unmeasured feature improved.
Keto-Med illustrates the problem: triglyceride changes favored the ketogenic phase, while LDL cholesterol changes favored the Mediterranean-plus phase. Its main report did not provide ApoB results, and the researchers did not measure LDL particle size. Claims that the LDL increase was harmless because particles became larger go beyond those measurements.
Burén’s controlled feeding trial did measure ApoB and LDL subfractions. In the women who completed it, the tested ketogenic pattern raised ApoB alongside LDL cholesterol, including both larger and smaller LDL fractions. This is a useful counterexample to treating a rise in larger particles as automatic reassurance.
Keep the dietary exposure attached to the result. Burén tested a specific pattern rich in saturated fat in a small group of young women. It does not establish identical responses to every ketogenic diet. Equally, uncertainty about other formulations does not erase the measured changes in this one. A clinical interpretation needs the person’s wider risk context.
Keep this boundary in view
Neither study estimates an individual’s future event risk or proves that every ketogenic pattern has the same cardiovascular effects. Biomarker tradeoffs warrant interpretation rather than a single overall health score.
Saved only in this browser, after you choose Save.
What changes when athletic performance is the outcome?
Specify the task. Endurance economy, race performance, maximal strength, repeated efforts, and muscle growth are separate questions.
Burke’s training-camp study makes the distinction visible. Race walkers increased fat oxidation on ketogenic food, yet required more oxygen at race-relevant speeds and did not show the clear race improvement seen in the carbohydrate-supported groups. A metabolic adaptation can therefore coexist with an unfavorable performance comparison.
Its design also deserves close reading. Athletes expressed dietary preferences rather than being randomly allocated, some attended two camps, and race conditions varied. Careful food provision strengthens confidence about what was eaten; it does not remove those other limitations or turn race walkers into representatives of every sport.
The ISSN position stand describes maximal-strength outcomes as commonly similar across diets, with some comparisons favoring the higher-carbohydrate condition. That narrower conclusion does not demonstrate equal muscle growth, recovery, or performance in every demanding session. Ask whether the paper tested the task, training level, and duration relevant to the claim.
Keep this boundary in view
Long-term adaptation and some forms of prolonged, lower-intensity endurance remain uncertain. These findings do not support a general promise that increased fat oxidation improves athletic results.
Saved only in this browser, after you choose Save.
Which safety questions come before interpreting a ketone result?
Medicines, illness, symptoms, and clinical circumstances determine how a ketone result should be interpreted. A reassuring glucose number does not resolve every risk.
The official empagliflozin label identifies ketogenic diets and reduced food intake among circumstances that can precipitate ketoacidosis. It also warns that glucose may be lower than expected for this emergency. This is why a sports discussion of nutritional ketosis cannot substitute for medication-specific safety information.
Research eligibility matters here. Keto-Med excluded people taking certain drugs, including insulin and SGLT2 inhibitors, and used a study physician to manage some other glucose-lowering treatment. A trial conducted within those boundaries does not establish safety for someone outside them.
Before a major dietary change, the useful conversation with a prescribing clinician concerns the proposed restriction, all medicines, relevant conditions, and what symptoms require help. For someone taking an SGLT2 inhibitor, nausea, vomiting, abdominal pain, or shortness of breath warrant urgent medical assessment, even if the glucose reading appears reassuring. This guide cannot determine a safe diet or direct a dose change.
Keep this boundary in view
A drug label identifies hazards; it does not supply a precise risk estimate for every ketogenic diet. Nutrition trials also have limited ability to detect uncommon harms.
Saved only in this browser, after you choose Save.
COMPARE THE METHODS
Different designs answer different questions.
A controlled feeding trial, a training study, and clinical guidance are different kinds of evidence. Start with the design before comparing their conclusions.
| Source and design | People and duration | What was compared |
|---|---|---|
| 1. Ketosis, appetite, and measured calorie intakeRandomized inpatient crossover feeding trial | Adults aged 18–39 without diabetes, living in an NIH metabolic ward21 randomized; 20 completed and were analyzedTwo weeks per diet; four continuous inpatient weeks | An animal-based ketogenic diet versus a plant-based low-fat diet, with unrestricted amounts of provided food and no washout. |
| 2. Keto-Med: glucose and lipids tell different storiesRandomized crossover dietary trial | Adults with prediabetes or type 2 diabetes; primary-analysis participants were aged 41–77Publication cohort: 42 randomized, 40 started, 33 with complete primary-outcome dataPlanned 12 weeks per diet, then 12 weeks of follow-up; some phases extended during COVID-19 | Ketogenic versus Mediterranean-plus diets; both emphasized nonstarchy vegetables and avoiding added sugars and refined grains. No washout. |
| 3. A controlled look at LDL and ApoB on ketoRandomized crossover controlled feeding trial | Healthy young women with normal BMI, recruited from dietetics students in Sweden24 randomized; 17 completed both dietsFour weeks per diet, separated by a 15-week washout | Provided ketogenic food, rich in saturated fat, versus provided food following Swedish dietary recommendations. Ketosis was confirmed; weight fell despite a weight-maintenance goal. |
| 4. More fat oxidation, a higher oxygen costControlled feeding and training intervention; preference-based allocation | International-level male race walkers at supervised training camps21 individual athletes; 29 diet-camp datasets and 26 paired race datasetsThree weeks of intensified training, with testing before and after | Ketogenic food versus consistently high carbohydrate availability or carbohydrate intake varied around training; energy and protein were similar. |
| 5. ISSN: keto, exercise performance, and body compositionProfessional position stand based on a review of controlled exercise studies | Recreationally active adults and competitive athletes in the reviewed studiesEvidence synthesis; no new participant cohortMostly short interventions; strength evidence discussed primarily spans 4–12 weeks | Ketogenic and non-ketogenic diets during exercise or training; diet composition and ketosis verification varied. |
| 6. SGLT2 medicines: why ketogenic-diet safety needs contextOfficial prescription-drug labeling | People prescribed empagliflozin, an SGLT2 inhibitor | Regulatory safety information; not a randomized comparison of diets |
On a narrow screen, scroll the comparison to see every column.
Compare three core papers in the source libraryTHE PAPERS, WITH A READING PLAN
Read the originals.
Each numbered source connects to the claims above. Open its findings, methods, and disclosures, or go directly to the publication. Access notes describe what we could review.
View all references
These references credit the original publications. The explanations and reading notes in this guide are by Fasting Free.
Effect of a plant-based, low-fat diet versus an animal-based, ketogenic diet on ad libitum energy intake
Hall et al. · Nature Medicine
An unusually controlled test of spontaneous food intake.
Read for: Figure 1 participant flow, the primary intake comparison, and the distinction between endpoint fat mass and rate of fat loss.
Published full text inspected; publisher access may require a subscription.
Saved only in this browser, after you choose Save.
Findings, methods, and disclosures
- Design and scope
- Randomized inpatient crossover feeding trial. Adults aged 18–39 without diabetes, living in an NIH metabolic ward; 21 randomized; 20 completed and were analyzed.
- Duration
- Two weeks per diet; four continuous inpatient weeks
- Comparison
- An animal-based ketogenic diet versus a plant-based low-fat diet, with unrestricted amounts of provided food and no washout.
- What was measured
- Primary: average energy intake across each two-week period and its final week. Exploratory: appetite, DXA body composition, and metabolic measures. Daily blood ketones were measured in 15 completers.
- Finding
- Energy intake was 689 ± 73 kcal/day lower on the low-fat diet over two weeks (mean ± SE; P < 0.0001). Ketosis was established on the ketogenic diet. The difference in total body fat lost between diets was not statistically clear (P = 0.085).
- Limits
- Short duration, different food sources, fiber and energy density, and no washout limit interpretation. The significant fat-loss-rate analysis is a different result from the nonsignificant end-period fat-mass comparison. One participant withdrew with hypoglycemia.
- Funding and disclosures
- NIH intramural funding. C. G. Forde disclosed food-industry speaking, advisory, and research relationships; other authors reported no competing interests.
- Design and scope
- Randomized inpatient crossover feeding trial. Adults aged 18–39 without diabetes, living in an NIH metabolic ward; 21 randomized; 20 completed and were analyzed.
- Duration
- Two weeks per diet; four continuous inpatient weeks
- Comparison
- An animal-based ketogenic diet versus a plant-based low-fat diet, with unrestricted amounts of provided food and no washout.
- What was measured
- Primary: average energy intake across each two-week period and its final week. Exploratory: appetite, DXA body composition, and metabolic measures. Daily blood ketones were measured in 15 completers.
- Finding
- Energy intake was 689 ± 73 kcal/day lower on the low-fat diet over two weeks (mean ± SE; P < 0.0001). Ketosis was established on the ketogenic diet. The difference in total body fat lost between diets was not statistically clear (P = 0.085).
- Limits
- Short duration, different food sources, fiber and energy density, and no washout limit interpretation. The significant fat-loss-rate analysis is a different result from the nonsignificant end-period fat-mass comparison. One participant withdrew with hypoglycemia.
- Funding and disclosures
- NIH intramural funding. C. G. Forde disclosed food-industry speaking, advisory, and research relationships; other authors reported no competing interests.
Original source: https://www.nature.com/articles/s41591-020-01209-1
DOI: https://doi.org/10.1038/s41591-020-01209-1
Effect of a ketogenic diet versus Mediterranean diet on glycated hemoglobin in individuals with prediabetes and type 2 diabetes mellitus: The interventional Keto-Med randomized crossover trial
Gardner et al. · The American Journal of Clinical Nutrition
A comparison of two food patterns with shared quality goals.
Read for: Primary versus exploratory outcomes, medication sensitivity analyses, missing data, and the linked percentage-axis correction.
Open full text and correction inspected.
Saved only in this browser, after you choose Save.
Findings, methods, and disclosures
- Design and scope
- Randomized crossover dietary trial. Adults with prediabetes or type 2 diabetes; primary-analysis participants were aged 41–77; Publication cohort: 42 randomized, 40 started, 33 with complete primary-outcome data.
- Duration
- Planned 12 weeks per diet, then 12 weeks of follow-up; some phases extended during COVID-19
- Comparison
- Ketogenic versus Mediterranean-plus diets; both emphasized nonstarchy vegetables and avoiding added sugars and refined grains. No washout.
- What was measured
- Primary: relative HbA1c change. Secondary: blood lipids and other metabolic measures. CGM outcomes were exploratory. ApoB was not reported and LDL particle size was not measured.
- Finding
- The primary HbA1c comparison did not show a clear difference (P = 0.11). Mean triglyceride changes were −16% versus −5%, while LDL cholesterol changed +10% versus −5%, for ketogenic versus Mediterranean-plus diets. These are relative changes from baseline.
- Limits
- Small complete-case analysis, medication adjustments, COVID-19 disruptions, and no washout complicate interpretation. Some sensitivity analyses favored keto. Secondary comparisons lacked multiplicity adjustment. The registry now lists a larger overall enrollment than this publication cohort.
- Funding and disclosures
- Research gifts and NIH/Stanford support; authors reported no conflicts. Funders had no role in design, analysis, or publication.
- Design and scope
- Randomized crossover dietary trial. Adults with prediabetes or type 2 diabetes; primary-analysis participants were aged 41–77; Publication cohort: 42 randomized, 40 started, 33 with complete primary-outcome data.
- Duration
- Planned 12 weeks per diet, then 12 weeks of follow-up; some phases extended during COVID-19
- Comparison
- Ketogenic versus Mediterranean-plus diets; both emphasized nonstarchy vegetables and avoiding added sugars and refined grains. No washout.
- What was measured
- Primary: relative HbA1c change. Secondary: blood lipids and other metabolic measures. CGM outcomes were exploratory. ApoB was not reported and LDL particle size was not measured.
- Finding
- The primary HbA1c comparison did not show a clear difference (P = 0.11). Mean triglyceride changes were −16% versus −5%, while LDL cholesterol changed +10% versus −5%, for ketogenic versus Mediterranean-plus diets. These are relative changes from baseline.
- Limits
- Small complete-case analysis, medication adjustments, COVID-19 disruptions, and no washout complicate interpretation. Some sensitivity analyses favored keto. Secondary comparisons lacked multiplicity adjustment. The registry now lists a larger overall enrollment than this publication cohort.
- Funding and disclosures
- Research gifts and NIH/Stanford support; authors reported no conflicts. Funders had no role in design, analysis, or publication.
Original source: https://pmc.ncbi.nlm.nih.gov/articles/PMC9437985/
DOI: https://doi.org/10.1093/ajcn/nqac154
A Ketogenic Low-Carbohydrate High-Fat Diet Increases LDL Cholesterol in Healthy, Young, Normal-Weight Women: A Randomized Controlled Feeding Trial
Burén et al. · Nutrients
Direct measurement of ApoB and LDL fractions.
Read for: The specific fat composition, participant withdrawals, confirmed ketosis, and the limits of short-term biomarker evidence.
Open full text, methods, results, and disclosures inspected.
Saved only in this browser, after you choose Save.
Findings, methods, and disclosures
- Design and scope
- Randomized crossover controlled feeding trial. Healthy young women with normal BMI, recruited from dietetics students in Sweden; 24 randomized; 17 completed both diets.
- Duration
- Four weeks per diet, separated by a 15-week washout
- Comparison
- Provided ketogenic food, rich in saturated fat, versus provided food following Swedish dietary recommendations. Ketosis was confirmed; weight fell despite a weight-maintenance goal.
- What was measured
- Primary: LDL cholesterol. Secondary: ApoB, LDL subfractions, and other blood markers.
- Finding
- Compared with control, the ketogenic diet increased LDL cholesterol by 1.82 mmol/L (95% CI 1.24–2.39) and ApoB by 0.50 g/L (95% CI 0.35–0.65). Both large and small LDL cholesterol fractions increased.
- Limits
- A small, short study of one high-saturated-fat diet cannot represent every ketogenic pattern or estimate cardiovascular-event risk. Six withdrawals occurred with symptoms during the ketogenic phase; the results describe completers.
- Funding and disclosures
- P. Håkansson’s and Magnus Bergvall’s foundations; no commercial funding or benefits reported. Authors declared no conflicts and no funder role in the research.
- Design and scope
- Randomized crossover controlled feeding trial. Healthy young women with normal BMI, recruited from dietetics students in Sweden; 24 randomized; 17 completed both diets.
- Duration
- Four weeks per diet, separated by a 15-week washout
- Comparison
- Provided ketogenic food, rich in saturated fat, versus provided food following Swedish dietary recommendations. Ketosis was confirmed; weight fell despite a weight-maintenance goal.
- What was measured
- Primary: LDL cholesterol. Secondary: ApoB, LDL subfractions, and other blood markers.
- Finding
- Compared with control, the ketogenic diet increased LDL cholesterol by 1.82 mmol/L (95% CI 1.24–2.39) and ApoB by 0.50 g/L (95% CI 0.35–0.65). Both large and small LDL cholesterol fractions increased.
- Limits
- A small, short study of one high-saturated-fat diet cannot represent every ketogenic pattern or estimate cardiovascular-event risk. Six withdrawals occurred with symptoms during the ketogenic phase; the results describe completers.
- Funding and disclosures
- P. Håkansson’s and Magnus Bergvall’s foundations; no commercial funding or benefits reported. Authors declared no conflicts and no funder role in the research.
Original source: https://pmc.ncbi.nlm.nih.gov/articles/PMC8001988/
DOI: https://doi.org/10.3390/nu13030814
Low carbohydrate, high fat diet impairs exercise economy and negates the performance benefit from intensified training in elite race walkers
Burke et al. · The Journal of Physiology
Fuel use and a real performance test measured together.
Read for: Preference allocation, unique athletes versus datasets, race conditions, and the use of 90% confidence intervals.
Open full text, methods, results, and disclosures inspected.
Saved only in this browser, after you choose Save.
Findings, methods, and disclosures
- Design and scope
- Controlled feeding and training intervention; preference-based allocation. International-level male race walkers at supervised training camps; 21 individual athletes; 29 diet-camp datasets and 26 paired race datasets.
- Duration
- Three weeks of intensified training, with testing before and after
- Comparison
- Ketogenic food versus consistently high carbohydrate availability or carbohydrate intake varied around training; energy and protein were similar.
- What was measured
- Exercise oxygen cost, peak aerobic capacity, fuel oxidation, and performance in an official 10 km race walk.
- Finding
- Fat oxidation rose, but oxygen cost increased at race-relevant speeds. Race performance improved 6.6% and 5.3% in the carbohydrate-supported groups; the ketogenic estimate was −1.6% (90% CI −8.5% to 5.3%), with no clear improvement.
- Limits
- Preference allocation, repeated participation, small groups, and varying race conditions limit causal certainty. Elite male race walking over three weeks does not establish outcomes for recreational training, women, or long-term adaptation.
- Funding and disclosures
- Australian Catholic University and Australian Institute of Sport research funding. Authors reported no competing interests.
- Design and scope
- Controlled feeding and training intervention; preference-based allocation. International-level male race walkers at supervised training camps; 21 individual athletes; 29 diet-camp datasets and 26 paired race datasets.
- Duration
- Three weeks of intensified training, with testing before and after
- Comparison
- Ketogenic food versus consistently high carbohydrate availability or carbohydrate intake varied around training; energy and protein were similar.
- What was measured
- Exercise oxygen cost, peak aerobic capacity, fuel oxidation, and performance in an official 10 km race walk.
- Finding
- Fat oxidation rose, but oxygen cost increased at race-relevant speeds. Race performance improved 6.6% and 5.3% in the carbohydrate-supported groups; the ketogenic estimate was −1.6% (90% CI −8.5% to 5.3%), with no clear improvement.
- Limits
- Preference allocation, repeated participation, small groups, and varying race conditions limit causal certainty. Elite male race walking over three weeks does not establish outcomes for recreational training, women, or long-term adaptation.
- Funding and disclosures
- Australian Catholic University and Australian Institute of Sport research funding. Authors reported no competing interests.
Original source: https://pmc.ncbi.nlm.nih.gov/articles/PMC5407976/
DOI: https://doi.org/10.1113/JP273230
International society of sports nutrition position stand: ketogenic diets
Leaf et al.; International Society of Sports Nutrition · Journal of the International Society of Sports Nutrition
A broader map of exercise outcomes and study limitations.
Read for: Eligibility rules, dietary control, hydration-related measurement issues, research gaps, and author disclosures.
Open full text, review methods, and disclosures inspected.
Saved only in this browser, after you choose Save.
Findings, methods, and disclosures
- Design and scope
- Professional position stand based on a review of controlled exercise studies. Recreationally active adults and competitive athletes in the reviewed studies; Evidence synthesis; no new participant cohort.
- Duration
- Mostly short interventions; strength evidence discussed primarily spans 4–12 weeks
- Comparison
- Ketogenic and non-ketogenic diets during exercise or training; diet composition and ketosis verification varied.
- What was measured
- Endurance performance, muscular strength, and body composition; no pooled treatment effect or formal GRADE rating.
- Finding
- Endurance outcomes were generally unchanged or worse with keto; maximal-strength outcomes were commonly similar. Greater fat oxidation did not establish better performance. Energy, protein, and fluid differences complicate body-composition findings.
- Limits
- A consensus synthesis inherits the weaknesses of its included studies. Long-term, female, and highly trained strength-athlete data remain limited. It does not evaluate exogenous ketone supplements or provide medication-safety guidance.
- Funding and disclosures
- No direct funding for the paper. Disclosures include supplement-industry research and consulting, and one author’s therapeutic-ketosis patents and ketone-company ownership.
- Design and scope
- Professional position stand based on a review of controlled exercise studies. Recreationally active adults and competitive athletes in the reviewed studies; Evidence synthesis; no new participant cohort.
- Duration
- Mostly short interventions; strength evidence discussed primarily spans 4–12 weeks
- Comparison
- Ketogenic and non-ketogenic diets during exercise or training; diet composition and ketosis verification varied.
- What was measured
- Endurance performance, muscular strength, and body composition; no pooled treatment effect or formal GRADE rating.
- Finding
- Endurance outcomes were generally unchanged or worse with keto; maximal-strength outcomes were commonly similar. Greater fat oxidation did not establish better performance. Energy, protein, and fluid differences complicate body-composition findings.
- Limits
- A consensus synthesis inherits the weaknesses of its included studies. Long-term, female, and highly trained strength-athlete data remain limited. It does not evaluate exogenous ketone supplements or provide medication-safety guidance.
- Funding and disclosures
- No direct funding for the paper. Disclosures include supplement-industry research and consulting, and one author’s therapeutic-ketosis patents and ketone-company ownership.
Original source: https://pmc.ncbi.nlm.nih.gov/articles/PMC11212571/
DOI: https://doi.org/10.1080/15502783.2024.2368167
JARDIANCE (empagliflozin) prescribing information and Medication Guide
Boehringer Ingelheim Pharmaceuticals, Inc. · DailyMed, U.S. National Library of Medicine
A medication-specific safety boundary absent from athletic trials.
Read for: The ketoacidosis warning and Medication Guide. This is regulatory information, not evidence that a dietary intervention is effective.
Full prescribing information and Medication Guide available.
Saved only in this browser, after you choose Save.
Findings, methods, and disclosures
- Design and scope
- Official prescription-drug labeling. People prescribed empagliflozin, an SGLT2 inhibitor.
- Comparison
- Regulatory safety information; not a randomized comparison of diets
- What was measured
- Ketoacidosis risk situations and symptoms; volume-depletion and hypoglycemia warnings.
- Finding
- The label identifies ketogenic diets, reduced food intake, illness, and volume depletion among conditions that can precipitate ketoacidosis. Glucose may be below the levels usually expected with ketoacidosis.
- Limits
- This label does not quantify the risk of a particular diet or determine an individual’s suitability. Diet and medicine decisions require the prescribing clinician; the dossier provides no dose-change instructions.
- Funding and disclosures
- Manufacturer prescribing information hosted by the National Library of Medicine; not an independent diet trial.
- Design and scope
- Official prescription-drug labeling. People prescribed empagliflozin, an SGLT2 inhibitor.
- Comparison
- Regulatory safety information; not a randomized comparison of diets
- What was measured
- Ketoacidosis risk situations and symptoms; volume-depletion and hypoglycemia warnings.
- Finding
- The label identifies ketogenic diets, reduced food intake, illness, and volume depletion among conditions that can precipitate ketoacidosis. Glucose may be below the levels usually expected with ketoacidosis.
- Limits
- This label does not quantify the risk of a particular diet or determine an individual’s suitability. Diet and medicine decisions require the prescribing clinician; the dossier provides no dose-change instructions.
- Funding and disclosures
- Manufacturer prescribing information hosted by the National Library of Medicine; not an independent diet trial.
Original source: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=faf3dd6a-9cd0-39c2-0d2e-232cb3f67565
PUT THE EVIDENCE TO USE
Practice asking the better question.
A dashboard improves in one direction
Hypothetical example: after changing food choices, an adult notices lower glucose readings and weight, alongside a higher LDL cholesterol result. A social-media explanation calls every change beneficial because ketosis was present.
- Which observations are measured outcomes, and which are proposed explanations?
- Was ApoB actually measured, or inferred from a different lipid result?
- What comparison would separate the diet’s effects from other changes over time?
The observations justify several separate questions. They do not produce an automatic net-benefit conclusion. Reporting units matters too: a hypothetical HbA1c change from 6.0% to 5.7% is 0.3 percentage points, or a 5% relative reduction. Those illustrative values are not Keto-Med results. In a real clinical conversation, interpretation would also include medicines, the food pattern, and cardiovascular risk context.
The fuel test and the finish line
Hypothetical example: a runner’s laboratory report shows more fat use after a dietary change. The runner assumes a faster race must follow, although the laboratory test used a comfortable pace and no race was measured.
- Did the test measure fuel selection, exercise economy, or actual performance?
- Would the same conclusion apply at the intensity and duration of the race?
- Were training, total food intake, and recovery comparable across conditions?
The laboratory finding answers a fuel-use question. A performance claim requires a performance comparison. Burke’s study provides a concrete reason to keep those steps separate, while the ISSN review shows why one sporting task cannot settle every other one. A stable maximal lift would likewise leave questions about repeated efforts and muscle growth unanswered.
WHERE THE EVIDENCE STOPS
What would make the answer more convincing?
- How do sustained ketogenic patterns compare with credible alternatives for clinical outcomes over years, beyond short-term biomarkers?
- How much do fat sources, fiber, energy intake, and food selection explain differing lipid responses across ketogenic interventions?
- What remains when energy and protein are closely matched, ketosis is repeatedly verified, and adherence is followed long enough?
- How do results differ in women, older adults, and people with conditions or medicines excluded from the featured trials?
- Which athletic conclusions hold beyond short study periods, including prolonged low-intensity events and highly trained strength athletes?
- How sustainable are the tested patterns when free meals, frequent counseling, and research-team support end?
Questions worth taking into a conversation.
Use these to prepare for an individual discussion with a qualified clinician, sports dietitian, or coach as appropriate to the question.
- Which outcome am I trying to understand, and does the cited study actually measure it?
- What was the comparison diet, and were food quality, calories, protein, and support similar?
- Which findings are primary, which are exploratory, and how large is the uncertainty?
- If glucose and lipid results move differently, what additional information would make a clinical discussion more useful?
- What would removing food groups mean for nutritional adequacy and everyday feasibility, rather than the ketone reading alone?
- Do medicines or medical conditions make dietary restriction a question for the prescribing clinician before any change?