Adjusting medications
on a low-carb diet

As a reminder, this information is intended for doctors, not for the general public (full disclaimer). Discuss any changes in medication and relevant lifestyle changes with your doctor.

Carbohydrate restriction in the form of a low-carb diet is effective for lowering blood sugar, improving insulin sensitivity, and reducing blood pressure.1 However, without careful monitoring, low-carb diets can be victims of their own success, triggering symptomatic hypoglycemia or hypotension. Therefore, clinicians using carbohydrate restriction need to be proactive about preventing these complications.

There are two major classes of medications that often need to be reduced on a low-carb diet: blood glucose-lowering medications and blood pressure medications.

Blood glucose lowering medications

When patients with type 2 diabetes start a low-carb diet, their blood glucose usually goes down immediately and can continue dropping as weight loss continues and insulin resistance improves.2

As this happens, patients on blood glucose-lowering medications may need to reduce their doses and number of medications to avoid hypoglycemia. It’s important for their doctor to know how to handle this situation. Hypoglycemia due to overdosing of blood glucose-lowering medications, especially insulin, is the biggest risk when starting a low-carb diet.3

Slightly high glucose is safer than too low

Although not all patients reducing carbohydrate intake will exhibit an immediate drop in blood glucose, many will see a profound drop on day one.

Because low blood sugar is far more dangerous in the short term, it’s safer to have your patients’ sugars run a little bit higher than the desired range for the initial few days to weeks.4

Deprescribing

If your patient is on insulin or a sulfonylurea, you will most likely need to decrease the doses or discontinue medications, initially targeting daily blood glucose levels between 145 to 200 mg/dL (between 8.0 and 11.1 mmol/L).5

We recommend assessing the patient’s risk of hypoglycemia in order to determine the magnitude of any dose adjustments. This assessment should include: blood glucose levels, doses of diabetes medications, baseline carbohydrate intake, and age.6

For example, if a patient’s blood glucose levels tend to be more than 250 mg/dL (13.9 mmol/L) throughout the day, that might make you less worried that they will experience significant hypoglycemia upon initiation of a low-carb diet. Is that enough to say that no dose adjustments are needed? No, it is not.

Let’s assume that the above patient is achieving that high level of average glycemia while taking 80 units of insulin glargine, 8mg of glimepiride, and 2000mg of extended release metformin daily. Let’s also assume that the person is eating 300 grams of carbohydrate most days. How would that change your assessment of their risk of hypoglycemia?

The high doses of basal insulin and sulfonylurea, combined with the baseline high carbohydrate intake, put this patient at more significant risk of hypoglycemia once the carbohydrate intake is cut to 20-50 grams/day.

If this patient was elderly, the risk of a significant fall during a hypoglycemic episode would push the overall risk to even higher levels.

Although we recommend using clinical judgment in every situation, if the risk of hypoglycemia is moderate to high, we suggest stopping sulfonylureas and short/rapid-acting insulin. We also recommend reducing intermediate/long-acting insulin by 33-50%, again using baseline glycemic control to guide the magnitude of any dose adjustment. It is also important to take into account the patient’s priorities regarding balancing strict glucose control with fear of hypoglycemia.7

If, on the other hand, you judge your patient’s risk of hypoglycemia to be low (e.g. baseline blood sugars are extremely high on lower doses of insulin/sulfonylurea), it might be reasonable to hold off on any medication adjustments. We would recommend counseling your patient to monitor blood glucose closely and cut their doses by 50% if levels are falling below 145 mg/dL (8.0 mmol/L), or stop medications completely if they have a hypoglycemic episode.

To summarize, here’s the recommended order of de-prescribing diabetes medications for patients with type 2 diabetes.

  1. (due to the risk of ketoacidosis, see below)
  2. (risk of hypoglycemia)
  3. (risk of hypoglycemia)
  4. (risk of hypoglycemia)

You’ll find more details below.

Follow-up adjustments

Within the first few months, blood glucose levels commonly drop back down to an acceptable range. Once the patient’s levels are in the 70-130 mg/dL range (4 to 7.0 mmol/L), it may be time to make medication reductions again, targeting a range of 145 to 200 mg/dL (8.0 to 11.1 mmol/L).8

This cycle can repeat itself until the patient is no longer taking diabetes medications (or is only on metformin). After that, the goal is to bring glucose levels down to the normal range utilizing diet alone, targeting a normal HbA1c.9

If there is no reduction in blood glucose levels between appointments, no adjustment is necessary. Talk to your patient about their diet. Perhaps there are a few improvements to be made that could speed up the process, but sometimes it just takes a bit longer to see the blood sugar levels come down. If patients are slow to respond to a low-carb diet, put things in perspective by reminding them of how long they have had diabetes and how long it takes to develop.

Sometimes a patient’s blood glucose levels temporarily rise above 200 mg/dL (11.1 mmol/L) for various reasons, such as a vacation, relatives visiting from out of town, illness or infection. If their sugars do not quickly normalize, that patient may need a short-term medication increase. Some patients may be resistant to this, but assure them that it’s only for the short term and the goal is to reduce the dose as soon as is safe.

Reversal of type 2 diabetes10

It’s not uncommon for patients to reverse their type 2 diabetes on a strict low-carb diet. What does reversal mean? Diet Doctor defines reversal as having a HbA1c measurement below 6.5% without medications except metformin.

Other sources have coined different terms. For example, one publication describes “partial remission” as “sub-diabetic hyperglycemia (A1C not diagnostic of diabetes [<6.5%], fasting glucose 100–125 mg/dl [5.6–6.9 mmol/l]) of at least 1 year’s duration in the absence of active pharmacologic therapy or ongoing procedures.” The same source defines “remission” as normoglycemia for at least 1 year, as indicated by at least two HbA1c measurements below 5.7% without medications.11

More about type 2 diabetes medications

  

Medication Mechanism Pros and cons
SGLT-2 inhibitors

  • Canagliflozin
  • Dapagliflozin
  • Empagliflozin
  • Ertugliflozin
Inhibit type 2 sodium-glucose cotransporter.

Elimination of glucose in the urine, as reabsorption of glucose by the kidneys is reduced.

No risk of hypoglycemia.
No weight gain, possible weight loss.

Risk of (euglycemic) DKA, especially on low carb.

Intermediate/long-acting insulin

  • NPH
  • Glargine
  • Detemir
  • Degludec
Facilitates movement of glucose from the bloodstream into cells. Effective control of basal/fasting glycemia.

High risk of hypoglycemia.
Weight gain.

Short/rapid-acting insulin

  • Regular
  • Lispro
  • Aspart
  • Glulisine
Facilitates movement of glucose from the bloodstream into cells. Effective control of postprandial glycemia.

High risk of hypoglycemia.
Weight gain.

Sulfonylureas and Meglitinides

  • Gliclazide
  • Glimepiride
  • Glyburide
  • Glipizide
  • Tolazamide
  • Tolbutamide
  • Chlorpropamide
  • Nateglinide
  • Repaglinide
Stimulate insulin secretion by the pancreas. Effective control of overall glycemia.

Risk of hypoglycemia.
Weight gain.

DPP-4 inhibitors

  • Sitagliptin
  • Saxagliptin
  • Linagliptin
  • Alogliptin
  • Vildagliptin
Prolong the action of GLP-1, which stimulates secretion of insulin and inhibits release of glucagon. Low risk of hypoglycemia.
No weight gain.

Weak effect on glucose control.

GLP-1 receptor agonists

  • Liraglutide
  • Exenatide
  • Lixisenatide
  • Dulaglutide
  • Semaglutide
Enhance glucose-dependent insulin secretion.

Slow gastric emptying.

Decrease postprandial glucagon.

Reduce food intake.

Low risk of hypoglycemia.
Weight loss.
Reduced appetite.

Nausea/vomiting.
Risk of pancreatitis.

Alpha-glucosidase inhibitor

  • Acarbose
Inhibit upper GI enzymes that digest carbs.

Slow the absorption of glucose.

No risk of hypoglycemia.
No weight gain.

Intestinal discomfort/diarrhea.

Biguanide

  • Metformin
Inhibit the production of glucose by the liver.

Increase the uptake of glucose in periphery.

No risk of hypoglycemia.
No weight gain.

Intestinal discomfort/diarrhea.


 

A note on SGLT-2 inhibitors

Some doctors choose to have patients stay on SGLT-2 inhibitors and decrease/stop insulin and sulfonylureas first.12 However, given the multiple case reports of euglycemic ketoacidosis with low- carb diets and SGLT-2 inhibitors, the risks of SGLT-2 inhibitors likely outweigh the benefits.13

Should you choose to continue with SGLT-2 therapy, please ask your patients to look out for the following early symptoms of ketoacidosis. They should stop taking the drug and notify you immediately if they develop:

  • Nausea
  • Weakness and fatigue
  • Dehydration

In most cases, however, and in the opinion of most doctors treating patients with low carb, SGLT-2 inhibitors may not be worth the risk of ketoacidosis. Thus, they should be the first medication discontinued when starting a low-carb diet.14

 

DPP-4 vs. GLP-1

If your patient is taking both a DPP-4 inhibitor and a GLP-1 receptor agonist, we recommend asking them which medication they would prefer to reduce or discontinue first.

  • Some people prefer to discontinue the GLP-1 receptor agonist since it is an injectable and is often considered less convenient due to requiring refrigeration.15
  • However, others find that the GLP-1 receptor agonist reduces appetite, so they prefer to reduce or stop their DPP-4 inhibitor first.

Since neither class of medication puts the patient at significant risk for hypoglycemia, the decision can be made according to the patient’s preference. If the patient has no preference, start to reduce and discontinue the DPP-4 inhibitor first, as the GLP-1 receptor agonist may decrease hunger and cause some weight loss.

 


Blood pressure medications

Baseline

Carbohydrate restriction is an effective way to lower blood pressure.16 However, the change happens more slowly than the change in blood sugar, potentially taking weeks to months. Therefore, you do not have to automatically adjust blood pressure medications as you do with glucose-lowering medications.

Instead, start by making sure your patient is aware of potential symptoms of hypotension (such as lightheadedness, dizziness, fatigue, or nausea) and knows to contact you immediately if they occur.

Follow-up

If the patient presents in clinic with low blood pressure or experiences low blood pressures at home, then you will likely need to reduce or discontinue antihypertensive medications.

Determining which blood pressure medication to discontinue first

There isn’t a universal protocol for which drugs to stop first, as patients have different reasons for being on specific medicines – such as ACE inhibitors for proteinuria, beta blockers for coronary disease, or alpha blockers for benign prostatic hypertrophy (BPH). Thus, we recommend individualizing the de-prescribing of anti-hypertensive medications.

However, if there is no additional indication for a medication, then we recommend stopping diuretics first, since low-carb diets frequently have a diuretic effect of their own.17

Instruct the patient to continue careful blood pressure monitoring to ensure there is no rebound increase once they reduce or stop their medications.

  1. For a comprehensive list of studies supporting this statement, see our guide on the science of low-carb and keto.

  2. Not only is this based on clinical experience, but multiple systematic reviews of RCTs support this as well.

    Diabetes Research and Clinical Practice 2018: Effect of dietary carbohydrate restriction on glycemic control in adults with diabetes: A systematic review and meta-analysis [strong evidence]

    BMJ Open Diabetes Research and Care 2017: Systematic review and meta-analysis of dietary carbohydrate restriction in patients with type 2 diabetes [strong evidence]

    The American Journal of Clinical Nutrition 2018: Effects of low-carbohydrate- compared with low-fat-diet interventions on metabolic control in people with type 2 diabetes: a systematic review including GRADE assessments [strong evidence]

    Diabetes, Obesity & Metabolism 2019: An evidence‐based approach to developing low‐carbohydrate diets for type 2 diabetes management: a systematic review of interventions and methods [strong evidence]

  3. This is based on clinical experience of low-carb practitioners and was unanimously agreed upon by our low-carb expert panel. You can learn more about our panel here [weak evidence].

  4. This is based on clinical experience of low-carb practitioners and was unanimously agreed upon by our low-carb expert panel. You can learn more about our panel here [weak evidence].

  5. This is based on clinical experience.

  6. This is based on clinical experience.

  7. This is based on clinical experience of low-carb practitioners and was unanimously agreed upon by our low-carb expert panel. You can learn more about our panel here [weak evidence].

  8. This is based on clinical experience.

  9. Although most medical societies (like the American Diabetes Association) recommend treating people with type 2 diabetes to a HbA1c of 6.5-7.0%, we at Diet Doctor believe that many people can achieve a completely normal HbA1c with a low carbohydrate lifestyle. For some, minimal to zero medication will be necessary.

  10. Some disagree with the use of the word “reverse” when it comes to type 2 diabetes. The concern is that it may imply the disease is completely gone, never to return. At Diet Doctor, we use the term “reverse” to indicate when blood sugars are no longer in the diabetic range. However, we acknowledge that blood sugars will likely return to diabetic levels if a patient goes back to their prior high-carb eating habits. Therefore, “reverse” does not imply a forever cure.

  11. Diabetes Care 2009: How do we define cure of diabetes? [overview article; ungraded]

  12. The rationale may be that the risk of hypoglycemia with SGLT-2 inhibitors is low, compared to the risk with insulin and sulfonylureas.

  13. Diabetes Care: Euglycemic diabetic ketoacidosis: A potential complication of treatment with sodium–glucose cotransporter 2 inhibition [case reports; very weak evidence]

    AACE Clinical Case Reports: Euglycemic diabetic ketoacidosis with SGLT2 inhibitor use in a patient on the Atkins diet: A unique presentation of a known side effect [case reports; very weak evidence]

  14. This is based on clinical experience.

  15. There is one oral GLP-1 receptor agonist available. As for the injectable versions, they require refrigeration prior to first use but generally are stable at room temperature for 2-4 weeks.

  16. Journal of the American Medical Association 2010: A randomized trial of a low-carbohydrate diet vs orlistat plus a low-fat diet for weight loss [moderate evidence]

  17. This is based on clinical experience.