Insulin degludec

What it is

Insulin degludec is a laboratory-modified version of human insulin, given by injection under the skin once a day. The molecule has a fatty acid chain attached that makes it clump into long chains of multi-hexamers in the fat under the skin, from which single insulin molecules trickle into the blood over more than 24 hours. It comes as a U-100 and a more concentrated U-200 pen, and in fixed combinations with insulin aspart and with liraglutide. It is not insulin the body made; nothing in food contains it and it is destroyed if swallowed.

What the research says

Degludec lowers blood glucose the way the body's own insulin does, by pushing glucose into muscle and fat and switching off glucose production by the liver. Its distinctive claim is flatness, not extra potency. In DEVOTE, 7,637 people with type 2 diabetes and cardiovascular disease, degludec matched insulin glargine U100 on major cardiovascular events (hazard ratio 0.91) and reduced severe hypoglycaemia from 6.6 to 4.9 per cent of patients over about two years. A 2021 Cochrane review, looking across trials in type 1 diabetes, found no clinically relevant difference between degludec and the other long-acting analogues on any of its main outcomes. Hypoglycaemia is by far the commonest harm: in the manufacturer's own type 2 diabetes trials, 28 to 81 per cent of people on degludec had at least one episode.

Evidence grade: Well established.

How it works

Drug class: Ultra-long-acting basal human insulin analogue

Degludec does what insulin does: it makes muscle and fat take glucose out of the blood and stops the liver making more, and it also blocks the breakdown of fat and protein. What is engineered is the timing. Injected under the skin it self-assembles into long multi-hexamer chains that dissolve slowly, and some of it binds to albumin in the blood, so a single injection releases insulin at a nearly flat rate for more than a day. (Source 1)

What it is used for

• Insulin is necessary for survival in type 1 diabetes, and degludec lowers HbA1c as well as the other basal insulins. Its advantage over insulin glargine U100 is a lower rate of severe hypoglycaemia; against insulin glargine U300 no difference has been shown. Evidence: established. (Source 2)
• DEVOTE tested this as a pre-specified secondary endpoint in 7,637 people: severe hypoglycaemia affected 4.9 per cent on degludec against 6.6 per cent on glargine U100, an absolute difference of 1.7 percentage points (about 59 people needing treatment for one to avoid an episode), rate ratio 0.60. Evidence: established. (Source 3)
• The label explicitly says degludec is not recommended for this. Ketoacidosis needs short-acting intravenous insulin, because a once-daily depot cannot be titrated hour by hour. Evidence: not-supported. (Source 4)
• A meta-analysis of three randomised trials (1,171 people with type 2 diabetes) found thrice-weekly degludec worse than once-daily glargine on HbA1c (mean difference 0.27 per cent in favour of glargine), with comparable side effects. Evidence: not-supported. (Source 5)

Interactions

• Alcohol (label): Alcohol can push blood glucose either way with insulin on board, and a low glucose that follows a drink is easy to mistake for being drunk. It is listed as needing dose adjustment and more frequent glucose checks. (Source 6)
• Niacin (nicotinic acid), a widely sold B-vitamin supplement, and high-dose niacin products (label): Niacin raises blood glucose and blunts insulin's effect, so a dose that was right before starting a niacin supplement may no longer be enough. (Source 7)
• Salicylates (including aspirin and willow bark preparations), fish-oil-unrelated fibrates, fluoxetine, GLP-1 receptor agonists, DPP-4 and SGLT-2 inhibitors (label): These increase the risk of hypoglycaemia when combined with insulin, so the insulin dose usually has to come down. (Source 7)
• Beta-blockers, clonidine, guanethidine and reserpine (label): These do not change blood glucose much but they hide the warning signs of a low, so the first thing noticed can be confusion or collapse rather than shakiness and sweating. (Source 6)
• Thiazolidinediones (pioglitazone, rosiglitazone) (label): Combined with insulin these cause dose-related fluid retention that can bring on or worsen heart failure. (Source 8)
• Food, meals and exercise (label): The interaction that matters most every day is with eating and activity. Insulin in excess of what food and energy expenditure need causes hypoglycaemia, and can cause hypokalaemia with it. (Source 9)

Stopping it

• There is no dependence or withdrawal syndrome in the drug-of-abuse sense, but insulin cannot simply be stopped in type 1 diabetes: glucose rises and ketoacidosis follows. Changing insulin at all, including switching brand, type, strength, injection site or method, can tip glucose either way and the label requires closer monitoring while it is done. (Source 10)
• In type 2 diabetes, insulin can sometimes be deliberately cut back. A retrospective study of 61 people in Malaysian primary care ran a structured deintensification clinic, reduced the mean daily dose by 0.41 units per kg and reported less hypoglycaemia and weight loss without losing glucose control. This is uncontrolled observational data during a national insulin shortage, not a randomised trial. (Source 11)

What goes wrong

Hypoglycaemia is the commonest harm of degludec and affected between a quarter and four-fifths of people in the manufacturer's type 2 diabetes trials. (Source 12)

• Randomized trial, Moderate certainty.
• Size: seven trial arms, 226 to 766 patients each, 2,713 adults with type 2 diabetes in the safety pool.
• Who: adults with type 2 diabetes, insulin-naive in most arms, one arm also on insulin aspart.
• How long: 26 to 52 weeks.
• Result: At least one hypoglycaemic episode (below 56 mg/dL or severe) in 46.5, 28.5, 50, 43.8, 50.9, 80.9 and 42.5 per cent of patients across the seven arms; severe hypoglycaemia in 0.3, 0, 0, 0.9, 0.4, 4.5 and 0.4 per cent. The 80.9 per cent and 4.5 per cent figures come from the arm that also received mealtime insulin aspart. The label warns that these open-label rates cannot be compared with other products' trials.
• Funding: industry-funded (trials run by Novo Nordisk; results reported in its label)

Severe Hypoglycemia Percent of patients 0.3% 0 0 0.9% 0.4% 4.5% 0.4% Hypoglycemia § Percent of patients 46.5% 28.5% 50% 43.8% 50.9% 80.9% 42.5%

Allergic reactions including anaphylaxis, and lumps or dents in the skin at injection sites, occur with degludec. (Source 13)

• Randomized trial, Moderate certainty.
• Size: not stated for anaphylaxis; 0.9 per cent figure from the pooled adult trials.
• Who: people with type 1 or type 2 diabetes using insulin.
• How long: 6 to 12 months in the pooled trials, longer for lipodystrophy.
• Result: Hypersensitivity with swelling of the tongue and lips, diarrhoea, nausea, tiredness, itching, and urticaria reported in 0.9 per cent of patients on degludec; severe generalised allergy including anaphylaxis, angioedema, bronchospasm, hypotension and shock reported and may be life-threatening; lipohypertrophy and lipoatrophy at repeatedly injected sites, which then changes how much insulin is absorbed.
• Funding: industry-funded (results reported in the manufacturer's label)

Hypersensitivity (manifested with swelling of tongue and lips, diarrhea, nausea, tiredness, and itching) and urticaria were reported in 0.9% of patients treated with TRESIBA.

All insulins, degludec included, can drive potassium into cells and cause hypokalaemia, which can be fatal if untreated. (Source 8)

• Official position, Certainty not rated.
• Size: not stated.
• Who: anyone treated with insulin, particularly people also on potassium-lowering drugs.
• How long: any time during treatment.
• Result: No rate given; the label states that untreated hypokalaemia may cause respiratory paralysis, ventricular arrhythmia and death, and that an insulin overdose can cause prolonged, life-threatening hypoglycaemia and hypokalaemia.
• Funding: manufacturer's label.

All insulins, including TRESIBA, cause a shift in potassium from the extracellular to intracellular space, possibly leading to hypokalemia.

Sharing an insulin pen, even with a new needle, can transmit blood-borne infections. (Source 14)

• Official position, Certainty not rated.
• Size: not stated.
• Who: anyone using a prefilled insulin pen or a vial.
• How long: any time.
• Result: No rate given; the label states that the disposable prefilled pens should never be shared between patients even if the needle is changed.
• Funding: manufacturer's label.

TRESIBA FlexTouch disposable prefilled pens should never be shared between patients, even if the needle is changed.

What the evidence supports

Degludec did not increase major cardiovascular events compared with insulin glargine U100 in people with type 2 diabetes at high risk. (Source 2)

• Randomized trial, High certainty.
• Size: 7,637 patients (3,818 degludec, 3,819 glargine U100)
• Who: adults with type 2 diabetes, mean age 65, mean diabetes duration 16 years, mean HbA1c 8.4 per cent; 85.2 per cent had established cardiovascular disease, chronic kidney disease or both.
• How long: median about 2 years, event-driven.
• Result: First major cardiovascular event in 325 patients (8.5 per cent) on degludec versus 356 (9.3 per cent) on glargine; hazard ratio 0.91 (95% CI 0.78 to 1.06), P<0.001 for noninferiority against the pre-specified margin of 1.3; absolute difference 0.8 percentage points. On the prespecified secondary outcome, severe hypoglycaemia occurred in 187 patients (4.9 per cent) on degludec versus 252 (6.6 per cent) on glargine, an absolute difference of 1.7 percentage points (rate ratio 0.60, P<0.001 for superiority). Rates of adverse events did not differ between the two groups.
• Funding: industry-funded (Novo Nordisk and others)

The primary outcome occurred in 325 patients (8.5%) in the degludec group and in 356 (9.3%) in the glargine group (hazard ratio, 0.91; 95% confidence interval, 0.78 to 1.06; P<0.001 for noninferiority).

Degludec caused fewer severe hypoglycaemic episodes than insulin glargine U100 at equal glucose control. (Source 3)

• Randomized trial, High certainty.
• Size: 7,637 patients (3,818 degludec, 3,819 glargine U100)
• Who: adults with type 2 diabetes and atherosclerotic cardiovascular disease.
• How long: median about 2 years.
• Result: Severe hypoglycaemia in 4.9 per cent of patients on degludec versus 6.6 per cent on glargine U100 (absolute difference 1.7 percentage points, about 59 people treated to avoid one affected person); odds ratio 0.73 (95% CI 0.60 to 0.89); 3.70 versus 6.25 events per 100 patient-years, rate ratio 0.60 (95% CI 0.48 to 0.76), two-sided P below 0.001.
• Funding: industry-funded (trial run by Novo Nordisk; results reported in its label)

Percent of patients with events 4.9% 6.6% Estimated odds ratio [95%CI] TRESIBA/Insulin glargine U-100 0.73 [0.60; 0.89]*

What the evidence does not support

In type 1 diabetes, Cochrane found no clinically relevant difference between degludec and the other long-acting insulin analogues on any main outcome. (Source 15)

• Systematic review, Moderate certainty.
• Size: 26 RCTs, 8,784 participants randomised overall; 4 RCTs comparing degludec with glargine and 2 comparing degludec with detemir; 1,372 participants received degludec.
• Who: adults and children with type 1 diabetes (8 studies, 21 per cent of participants, were in children)
• How long: 24 to 104 weeks.
• Result: No clinically relevant difference for all-cause mortality, quality of life, severe hypoglycaemia, non-fatal myocardial infarction or stroke, severe nocturnal hypoglycaemia, serious adverse events or HbA1c; no trials at all compared degludec with NPH insulin; data on quality of life and diabetic complications were sparse or missing.
• Funding: independent (Cochrane; the authors obtained clinical study reports from regulators and companies)

Limit of this finding: This "no difference" applies only to the newer analogues compared with each other. No trial in the review compared degludec with NPH (older human) insulin at all, so the review could not say whether degludec differs from NPH. Where the review did find a difference, it was for a different insulin: insulin detemir had less severe hypoglycaemia than NPH insulin on moderate-certainty evidence, and even there the review added that its 95 per cent prediction interval "indicated inconsistency in this finding", meaning the result may not hold in a new population.

insulin degludec versus insulin glargine (4 RCTs): there was no evidence of a clinically relevant difference for all main outcomes comparing (ultra-)long-acting insulin analogues with each other

Degludec showed no advantage over insulin glargine U300 on HbA1c or on any category of hypoglycaemia. (Source 16)

• Meta-analysis, Very low certainty.
• Size: 4 open-label RCTs, 2,727 participants.
• Who: adults with type 1 or type 2 diabetes.
• How long: not stated in the abstract.
• Result: HbA1c mean difference 0.07 per cent (95% CI 0.06 to 0.19, p = 0.29, 3 trials, 2,652 patients); any hypoglycaemia rate ratio 1.02 (95% CI 0.8 to 1.3, p = 0.87, 3 trials, 2,881 patients); nocturnal hypoglycaemia rate ratio 1.13 (95% CI 0.72 to 1.78, p = 0.54) and severe hypoglycaemia rate ratio 1.4 (95% CI 0.41 to 4.73, p = 0.59); degludec lowered fasting plasma glucose more, by 10.27 mg/dL (95% CI 7.25 to 13.29, p < 0.001). Risk of bias yielded some concern or high risk and the certainty of evidence was very low for most outcomes.
• Funding: independent (authors declared no funding from manufacturers)

Limit of this finding: The confidence intervals here are very wide, so "no difference" means "no difference was shown", not "the two insulins are the same". The severe-hypoglycaemia comparison rests on 2 trials and its interval runs from 0.41 to 4.73, which spans a halving and a more than fourfold increase in risk. The review itself rated the evidence very low certainty because of the small number of trials and their risk of bias. The abstract prints this confidence interval as "95% CI 0.06 - 0.19", which drops a minus sign; the paper's own body text gives it as -0.06 to 0.19. Read as printed the interval looks to exclude zero, and it does not. The honest reading is no clear difference in HbA1c between the two insulins.

There is no evidence of a difference between IDeg and IGla-300 in the mean change in HbA1c and the risk of anytime, nocturnal, and severe hypoglycemia.

Given three times a week rather than daily, degludec controlled glucose worse than once-daily glargine. (Source 5)

• Meta-analysis, Moderate certainty.
• Size: 3 RCTs, 1,171 participants.
• Who: insulin-naive adults with type 2 diabetes.
• How long: 16 to 26 weeks.
• Result: HbA1c reduction mean difference 0.27 per cent (95% CI 0.14 to 0.39, p < 0.0001) in favour of once-daily glargine; nine-point self-monitored glucose profile mean difference 0.45 mmol/L (95% CI 0.22 to 0.67, p < 0.0001) in favour of glargine; reaching HbA1c below 7 per cent odds ratio 0.69 (95% CI 0.53 to 0.89, p = 0.005) against thrice-weekly degludec. No difference in fasting plasma glucose, body weight, or quality-of-life scores. Daily insulin dose was slightly lower with thrice-weekly degludec (mean difference -0.07 units, 95% CI -0.13 to -0.01, p = 0.02).
• Funding: not stated.

Limit of this finding: The review describes the five safety comparisons as "identical in the two groups". They are not identical. The risk ratios run from 1.04 to 1.43 and the confidence intervals are wide: serious adverse events 1.43 (95% CI 0.77 to 2.65) and nocturnal hypoglycaemia 1.18 (95% CI 0.49 to 2.84) in an evidence base of only 1,171 patients. The accurate reading is that no statistically significant safety difference was found, in a sample too small to rule out a meaningful increase in harm.

IDeg 3TW was less effective than IGlar OD in HbA1c reduction (MD 0.27%, 95% CI [0.14, 0.39], p < 0.0001), reduction in mean nine-point self-monitored capillary blood glucose profile (MD 0.45 mmol/L, 95% CI [0.22, 0.67], p < 0.0001), and HbA1c reduction <7% (OR 0.69, 95% [0.53, 0.89], p = 0.005).

Where the research disagrees

Whether degludec's lower rate of severe hypoglycaemia is a real, general advantage over other basal insulins

• DEVOTE investigators (2017), reported in the manufacturer's label, rct, industry-funded, in type 2 diabetes with cardiovascular disease: Percent of patients with events 4.9% 6.6% Estimated odds ratio [95%CI] TRESIBA/Insulin glargine U-100 0.73 [0.60; 0.89]* (Source 3)
• Cochrane review of (ultra-)long-acting insulin analogues in type 1 diabetes (2021), systematic-review with GRADE, independent: insulin degludec versus insulin glargine (4 RCTs): there was no evidence of a clinically relevant difference for all main outcomes comparing (ultra-)long-acting insulin analogues with each other (Source 15)
• Systematic review of degludec versus insulin glargine U300 (2023), meta-analysis, very low certainty: There is no evidence of a difference between IDeg and IGla-300 in the mean change in HbA1c and the risk of anytime, nocturnal, and severe hypoglycemia. (Source 16)

How much

• Reference intake: Dosing is set by the prescriber and titrated against blood glucose, not by a reference intake. As a position, the US label's starting dose is 10 units once daily in insulin-naive type 2 diabetes, and about one-third to one-half of the total daily insulin dose (calculated from 0.2 to 0.4 units per kg body weight) in insulin-naive type 1 diabetes. (Source 17)
• Upper limit: The label sets no maximum daily dose; the practical ceiling is what one injection can deliver, 80 units from the U-100 pen and 160 units from the U-200 pen. An excess relative to food intake causes hypoglycaemia and hypokalaemia. (Source 18)
• Studied: DEVOTE gave degludec or insulin glargine U100 once daily between dinner and bedtime, titrated to target, on top of standard care. (Source 19)
• Studied: An overdose relative to food or activity is what causes harm: the label describes severe, prolonged and life-threatening hypoglycaemia and hypokalaemia from excess insulin. (Source 9)

A common belief, and what the research shows

The belief: A newer, longer-acting, more expensive insulin must control diabetes better.

What the research shows: It does not control glucose better. Across trials in type 1 diabetes Cochrane found that comparing the long-acting analogues with each other "there was no evidence of a clinically relevant difference for all main outcomes comparing (ultra-)long-acting insulin analogues with each other", and against insulin glargine U300 the HbA1c difference was 0.07 per cent with a confidence interval crossing zero. What degludec did do, in one large industry-funded trial against glargine U100, is reduce severe hypoglycaemia, from 6.6 to 4.9 per cent of patients. Stretched to three times a week to save injections it becomes worse than daily glargine.

Questions and answers

What is it?

Insulin degludec is a man-made version of human insulin, injected under the skin once a day. A fatty acid chain has been attached so that after injection the molecules link into long chains in the fat, which then release single insulin molecules slowly over more than a day. It is sold as a pen in two strengths and in fixed combinations with a fast insulin and with liraglutide. (Source 1)

What does it do in the body?

It does what insulin does. It makes muscle and fat pull glucose out of the blood and stops the liver producing more, and it also stops the body breaking down fat and protein. The engineering is about timing rather than strength: the slow release from the skin depot and some binding to blood albumin give a nearly flat effect lasting over 24 hours, so it covers the background insulin need rather than meals. (Source 1)

Is it good or bad for you?

In type 1 diabetes insulin is not optional, it is necessary for life. The question is which insulin and how much. Compared with insulin glargine U100 in 7,637 people with type 2 diabetes and heart disease, degludec caused fewer severe low-glucose episodes, 4.9 per cent of patients against 6.6 per cent. The same drug in too high a dose causes the harm it is meant to prevent: dangerous hypoglycaemia, and low potassium with it. (Source 3)

How do you get more of it?

Only by prescription and injection. As a position, the label's starting point in insulin-naive type 2 diabetes is 10 units once daily, and in insulin-naive type 1 diabetes roughly one-third to one-half of the total daily insulin need, with the rest given as short-acting insulin at meals. From there the dose is titrated against blood glucose by a prescriber. There is no behaviour, food or supplement that raises it. (Source 17)

If it is harmful, what reduces it?

Because it is engineered to last, degludec cannot be undone quickly; a dose given works for more than a day. When too much has been given, the treatment is glucose, not removal of the insulin: mild lows are treated with oral glucose and severe ones with glucagon or intravenous glucose, and observation has to continue afterwards because the effect outlasts the first recovery. (Source 9)

Why might someone be low in it or missing it?

Nobody is naturally low in insulin degludec, because the body does not make it. People end up short of it because a dose is missed, a pen runs out or is stored wrongly, a prescription lapses, or a regimen is changed. The label singles out regimen changes as a risk in both directions, including a switch of strength, brand, type, injection site or method. (Source 10)

Which whole foods contain it or feed it?

Does not apply. No whole food contains insulin degludec and nothing you eat can feed it, because insulin is a protein that the gut digests. It only works injected into the fat under the skin, from which it is slowly absorbed. What food does affect is how much you need: meals and activity are what the dose is matched to. (Source 1)

What happens if you do not have it?

Untreated or badly controlled diabetes is what happens. In type 1 diabetes, stopping insulin leads to rising glucose and then diabetic ketoacidosis, which is a medical emergency. The label's pregnancy section spells out the stakes when control is lost: poorly controlled diabetes raises the risk of ketoacidosis, pre-eclampsia, miscarriage, preterm delivery and delivery complications. (Source 20)

How can you test for it?

There is no routine blood test for degludec itself. What is measured is its effect: finger-prick or continuous glucose readings day to day, and HbA1c over about three months. That is reliable for whether the dose is right but tells you nothing about the drug level, and the label requires more frequent glucose monitoring whenever anything about the regimen changes. (Source 10)

References

1. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - CLINICAL PHARMACOLOGY, Mechanism of Action. 2022. Read the source
2. The New England Journal of Medicine. Efficacy and Safety of Degludec versus Glargine in Type 2 Diabetes. 2017. PMID 28605603, DOI 10.1056/NEJMoa1615692. Read the source
3. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - CLINICAL STUDIES, Hypoglycemia Outcomes in DEVOTE. 2022. Read the source
4. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - INDICATIONS AND USAGE. 2022. Read the source
5. British Journal of Hospital Medicine. Thrice-Weekly Insulin Degludec Versus Once-Daily Insulin Glargine in Insulin-Naive Patients With Type 2 Diabetes Mellitus: A Systematic Review and Meta-Analysis. 2025. PMID 40705560, DOI 10.12968/hmed.2024.0716. Read the source
6. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - DRUG INTERACTIONS, Table 5 (alcohol and masking of hypoglycaemia). 2022. Read the source
7. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - DRUG INTERACTIONS, Table 5 (glucose-lowering effect). 2022. Read the source
8. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - WARNINGS AND PRECAUTIONS, Hypokalemia and PPAR-gamma agonists. 2022. Read the source
9. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - OVERDOSAGE. 2022. Read the source
10. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - WARNINGS AND PRECAUTIONS, Changes in Insulin Regimen. 2022. Read the source
11. Malaysian Family Physician. Clinical outcomes of insulin deintensification in a real-world primary care setting: A retrospective observational study (Conclusion section). 2026. PMID 42757279, DOI 10.51866/oa.1088. Read the source
12. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - ADVERSE REACTIONS, Table 4 hypoglycaemia in type 2 diabetes. 2022. Read the source
13. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - ADVERSE REACTIONS, Hypersensitivity and Lipodystrophy. 2022. Read the source
14. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - WARNINGS AND PRECAUTIONS, Never Share a Pen. 2022. Read the source
15. Cochrane Database of Systematic Reviews. (Ultra-)long-acting insulin analogues for people with type 1 diabetes mellitus. 2021. PMID 33662147, DOI 10.1002/14651858.CD013498.pub2. Read the source
16. Frontiers in Endocrinology. Efficacy and safety of insulin glargine 300 units/mL vs insulin degludec in patients with type 1 and type 2 diabetes: a systematic review and meta-analysis (Conclusion section). 2023. PMID 38313835, DOI 10.3389/fendo.2023.1285147. Read the source
17. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - DOSAGE AND ADMINISTRATION, Starting dose. 2022. Read the source
18. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - DOSAGE AND ADMINISTRATION, General Dosing Instructions. 2022. Read the source
19. The New England Journal of Medicine. Efficacy and Safety of Degludec versus Glargine in Type 2 Diabetes. 2017. PMID 28605603, DOI 10.1056/NEJMoa1615692. Read the source
20. DailyMed / Novo Nordisk. TRESIBA (insulin degludec) injection, US prescribing information - USE IN SPECIFIC POPULATIONS, Pregnancy. 2022. Read the source