Low-carb diet safety

Low-Carb Diet Safety: A Physician’s Baseline and Re-Measure Plan

Low-carb diet safety is a baseline problem: flag patients on insulin, sulfonylureas, meglitinides, SGLT2 inhibitors or two or more antihypertensives, record autonomic, balance, cognitive and body-composition baselines, and schedule a repeat measurement. Hypoglycemia, euglycemic ketoacidosis and orthostatic falls are the predictable risks.

The patient who quietly cuts carbohydrate arrives at the next visit with a lower glucose, a lower pressure and the same prescriptions. The workup belongs before the diet, not after the first event.

Low-carb diet safety is a primary care problem before it is a nutrition problem. Patients read about carbohydrate restriction for metabolic and brain health, start without telling anyone, and the medication list written for their old diet keeps running. The Starved Brain by Dr. Gurpreet Singh Padda, MD, MBA, MHP, closes with a chapter, and a video, You Can Protect a Struggling Neuron, that put the safety evidence ahead of the protocol for exactly that reason. For the screening physician the lesson is a sequence: identify who is at risk, record the baseline that makes adverse effects visible, and schedule the repeat that shows whether the correction is working. Measura [Cardiometabolic and Autonomic Health Analysis] supplies measurement for that sequence; it does not prescribe diet or manage medication.

What are the risks of a low-carb diet for patients on medication?

• Hypoglycemia on insulin, sulfonylureas and meglitinides. The 2025 ADA Standards call for consistent medical oversight of very-low-carbohydrate eating and recognize that diabetes medications may need adjustment to prevent hypoglycemia. Practical guides published in 2019 and 2021 describe dose reductions at initiation rather than after the first low reading. In a remotely supervised cohort of 262 adults with type 2 diabetes, sulfonylurea use went from 23.7% to 0% in a year under daily glucose review. No study has put a clock on the first hypoglycemic event; Dr. Padda reports seeing it within days in his own patients.
• Euglycemic ketoacidosis on SGLT2 inhibitors. The FDA relabeled the class in December 2015 after 73 reports; median reported glucose was 211 mg/dL, and a low-carbohydrate diet was named among the risk factors. In FAERS, 71% of reports with laboratory data were euglycemic. Among 208,757 new users, ketoacidosis carried HR 2.85 against DPP-4 inhibitors. Published cases include patients sent home because hyperglycemia was not marked.
• Orthostatic symptoms on antihypertensives. Sodium loss dominates the first week. On a metabolic ward, symptomatic orthostatic hypotension occurred in 7 of 7 obese adults on a ketogenic protein diet. In DiRECT, among 69 participants who stopped all antihypertensives at the start of a formula diet, dizziness affected 36% and 27.5% needed a drug restarted, most of them previously on two or more.

Who should be flagged before starting a low-carb diet?

The list is short: patients on insulin, a sulfonylurea or a meglitinide; anyone on an SGLT2 inhibitor; two or more antihypertensives; pregnancy or lactation; kidney, liver, gallbladder or pancreatic disease; a history of disordered eating; a suspected inherited disorder of fat metabolism, where the diet is contraindicated; and moderate or advanced dementia. The 2025 ADA Standards state that very-low-carbohydrate plans are not currently recommended in pregnancy or lactation, in children, in kidney disease, or in people with or at risk for disordered eating. A flag raised at the metabolic or cognitive workup is easier to act on than one raised in the emergency department, and standing orders make that flag reproducible across a panel.

Why does a low-carb diet cause dizziness and falls?

The orthostatic data expose a measurement gap. A seated blood pressure at a routine visit says little about how an older patient on two or three antihypertensives will tolerate a week of sodium and water loss. Cardiac autonomic reflex tests and heart rate variability characterize autonomic regulation of heart rate and blood pressure, and vestibular and balance testing documents balance before a change known to produce dizziness on standing. A patient who already has impaired autonomic regulation or a balance deficit is the one in whom a predictable orthostatic drop is most likely to end on the floor. Recording that baseline ties the metabolic plan to fall prevention, the pairing outlined in cognitive assessment and fall prevention.

Cognitive and body-composition anchors

The USPSTF grades cognitive screening of asymptomatic adults over 65 as insufficient evidence, and the AAN guideline that recommended validated assessment has since been retired. Neither addresses a personal baseline in a patient with a concern who is about to change a major metabolic input. Cognitive assessment in that setting is not population screening; it is the comparison point for the next score.

Stage changes the calculation. In the KDRAFT pilot, 1 of 7 participants with very mild Alzheimer’s withdrew, and all 4 with moderate disease withdrew, each citing caregiver burden. The 2024 ESPEN guideline does not recommend routine ketogenic diets in dementia and treats malnutrition screening as essential. In 37,717 Korean adults newly diagnosed with dementia, being underweight after diagnosis carried HR 1.57 for death. Bioimpedance body composition gives a lean-mass reference, so weight loss in an older patient is read by compartment rather than by the scale.

When should labs be rechecked after starting a low-carb diet?

Laboratory panels drawn before the change should report fasting insulin as a number. In Whitehall II, modeled insulin sensitivity fell steeply over the five years before diabetes while glucose inflected only three years before. A single fasting value has limits: in 4,185 people with normal glucose tolerance it was of limited value for detecting dynamic hyperinsulinemia. The triglyceride-to-HDL ratio performed poorly in 98 overweight African American adults, AUC 0.56 against 0.85 for fasting insulin, so cut-points derived in white cohorts should not be transferred. Every marker then keeps its own clock:

• About ten weeks. In the supervised cohort, 84% of the one-year HbA1c fall, 90% of the fasting-glucose fall and 73% of the fasting-insulin fall were present by day 70.
• Months. Homocysteine; no trial has defined a weeks-scale course.
• About six months. Red-cell EPA reaches steady state after 180 days, so the omega-3 index plateaus late.
• Slowest. hs-CRP did not change significantly across 44 randomized ketogenic trials. A flat CRP at three months is not a failed intervention.

Documentation and workflow

The sequence documents cleanly. A metabolic panel, cognitive score and autonomic or balance baseline recorded before dietary counseling sit naturally within the annual wellness visit and chronic care follow-up, and repeats at ten weeks and six months give the chart a dated trajectory rather than a single value. Findings route to the ordering physician; see getting results into the record. The deeper failure is structural: carbohydrate restriction spreads through patient communities faster than medication plans are rewritten, and a regimen dosed for the old diet keeps working against the new one. The full safety evidence, with case reports and declared industry ties, is in the Chapter 10 companion evidence guide; the cognitive trial evidence is in ketogenic therapy in mild cognitive impairment.

Frequently asked questions

References

• American Diabetes Association Professional Practice Committee. (2025a). 5. Facilitating Positive Health Behaviors and Well-being to Improve Health Outcomes: Standards of Care in Diabetes—2025. Diabetes Care, 48(Suppl. 1), S86–S127. https://doi.org/10.2337/dc25-S005
• Blau, J. E., Tella, S. H., Taylor, S. I., & Rother, K. I. (2017). Ketoacidosis associated with SGLT2 inhibitor treatment: Analysis of FAERS data. Diabetes/Metabolism Research and Reviews, 33(8), Article e2924. https://doi.org/10.1002/dmrr.2924
• Douros, A., Lix, L. M., Fralick, M., Dell’Aniello, S., Shah, B. R., Ronksley, P. E., Tremblay, É., Hu, N., Alessi-Severini, S., Fisher, A., Bugden, S. C., Ernst, P., Filion, K. B., & the Canadian Network for Observational Drug Effect Studies (CNODES) Investigators. (2020). Sodium–glucose cotransporter-2 inhibitors and the risk for diabetic ketoacidosis: A multicenter cohort study. Annals of Internal Medicine, 173(6), 417–425. https://doi.org/10.7326/M20-0289
• DeHaven, J., Sherwin, R., Hendler, R., & Felig, P. (1980). Nitrogen and sodium balance and sympathetic-nervous-system activity in obese subjects treated with a low-calorie protein or mixed diet. The New England Journal of Medicine, 302(9), 477–482. https://doi.org/10.1056/NEJM198002283020901
• Leslie, W. S., Ali, E., Harris, L., Messow, C. M., Brosnahan, N. T., Thom, G., McCombie, E. L., Barnes, A. C., Sattar, N., Taylor, R., & Lean, M. E. J. (2021). Antihypertensive medication needs and blood pressure control with weight loss in the Diabetes Remission Clinical Trial (DiRECT). Diabetologia, 64(9), 1927–1938. https://doi.org/10.1007/s00125-021-05471-x
• Taylor, M. K., Sullivan, D. K., Mahnken, J. D., Burns, J. M., & Swerdlow, R. H. (2018). Feasibility and efficacy data from a ketogenic diet intervention in Alzheimer’s disease. Alzheimer’s & Dementia: Translational Research & Clinical Interventions, 4, 28–36. https://doi.org/10.1016/j.trci.2017.11.002
• Huh, Y., Nam, G. E., Han, K., Jung, J.-H., Kim, B. S., Lee, D., Park, H. J., Yoo, M. Y., Yoon, S. Y., Kang, S. H., Kim, C. K., Kim, S. M., & Park, H. S. (2026). Body mass index levels and changes before and after dementia diagnosis and risk of all-cause mortality: A nationwide cohort study. Alzheimer’s Research & Therapy, 18(1), article 10.1186/s13195–026–02002–x (page not assigned in PubMed record). https://doi.org/10.1186/s13195-026-02002-x
• Hallberg, S. J., McKenzie, A. L., Williams, P. T., Bhanpuri, N. H., Peters, A. L., Campbell, W. W., Hazbun, T. L., Volk, B. M., McCarter, J. P., Phinney, S. D., & Volek, J. S. (2018). Effectiveness and safety of a novel care model for the management of type 2 diabetes at 1 year: An open-label, non-randomized, controlled study. Diabetes Therapy, 9(2), 583–612. https://doi.org/10.1007/s13300-018-0373-9
• Ji, J., Fotros, D., Sohouli, M. H., Velu, P., Fatahi, S., & Liu, Y. (2025). The effect of a ketogenic diet on inflammation-related markers: A systematic review and meta-analysis of randomized controlled trials. Nutrition Reviews, 83(1), 40–58. https://doi.org/10.1093/nutrit/nuad175
• US Preventive Services Task Force (Owens, D. K., Davidson, K. W., Krist, A. H., Barry, M. J., Cabana, M., Caughey, A. B., Doubeni, C. A., Epling, J. W., Kubik, M., Landefeld, C. S., Mangione, C. M., Pbert, L., Silverstein, M., Simon, M. A., Tseng, C.-W., & Wong, J. B.). (2020). Screening for cognitive impairment in older adults: US Preventive Services Task Force recommendation statement. JAMA, 323(8), 757–763. https://doi.org/10.1001/jama.2020.0435

Related reading

• Ketogenic Diet, Mild Cognitive Impairment and the Screening Workup
• Cardiac Autonomic Reflex Tests

Medically reviewed by Dr. Gurpreet Singh Padda, MD, MBA, MHP, medical director of Measura. Last reviewed September 12, 2026.