Sitagliptin

What it is

Sitagliptin is a synthetic small-molecule drug taken as a tablet, sold as Januvia and in fixed-dose combinations with metformin. It is a dipeptidyl peptidase-4 (DPP-4) inhibitor, the first of the drug class often called "gliptins", first approved in the United States in 2006. It is absorbed quickly, is about 87% bioavailable, is not meaningfully affected by food, and is cleared largely unchanged by the kidneys, so the dose is cut when kidney function is low. It is prescription-only and does not occur in food.

What the research says

The evidence is that sitagliptin lowers blood sugar modestly and does not change the risk of heart attack, stroke or death. A network meta-analysis of 87 randomised trials put the HbA1c reduction against placebo at about 0.64 percentage points. In the 14,671-person TECOS trial it was no better and no worse than placebo for major cardiovascular events, and in the head-to-head GRADE trial it was the weakest of four drugs added to metformin for holding HbA1c under 7.0%. Its selling point is tolerability rather than potency: it causes much less hypoglycaemia than a sulfonylurea, but the literature carries an unresolved pancreatitis signal and class-level reports of bullous pemphigoid, severe joint pain and acute kidney injury.

Evidence grade: Well established.

How it works

Drug class: Dipeptidyl peptidase-4 (DPP-4) inhibitor (an oral "gliptin"; an incretin enhancer)

After a meal the gut releases incretin hormones (GLP-1 and GIP) that tell the pancreas to make insulin and tell it to make less glucagon. An enzyme called DPP-4 breaks those hormones down within minutes. Sitagliptin blocks DPP-4, so the hormones last longer. Because the hormones only act when glucose is normal or high, the effect on insulin is glucose-dependent, which is why sitagliptin on its own rarely causes low blood sugar. (Source 1)

What it is used for

• Sitagliptin reliably lowers HbA1c by roughly half a percentage point against placebo, with high-confidence evidence from 87 randomised trials. In the one trial that compared it head-to-head against three other add-ons to metformin it was the least effective at keeping HbA1c under 7.0%. Evidence: established. (Source 2)
• TECOS was designed to show sitagliptin was not worse than placebo, and that is what it showed: 11.4% versus 11.6% over a median 3 years, hazard ratio 0.98. It did not reduce cardiovascular events, and it did not reduce hospitalisation for heart failure either. The approved indication is glucose lowering, not cardiovascular protection. Evidence: not-supported. (Source 3)
• A 2026 meta-analysis of four small randomised trials in 285 patients found DPP-4 inhibitors as a class increased complete ulcer healing (RR 1.63) and shortened healing time by about 11 days. The authors themselves limit the conclusion to the small number and size of the trials, and the analysis is of the class, not of sitagliptin alone. Evidence: limited. (Source 4)
• A meta-analysis of eight randomised trials in 295 people found sitagliptin plus insulin lowered HbA1c by 0.36 percentage points more than insulin alone, but heterogeneity was very high (I-squared 91.6%) and the authors call the methodological quality of the included trials limited. Evidence: limited. (Source 5)
• The label states sitagliptin should not be used in type 1 diabetes. No trial evidence supporting that use was found in this search. Evidence: not-supported. (Source 6)

Interactions

• Sulfonylureas such as glimepiride, glipizide and glyburide (case reports): Adding sitagliptin to a sulfonylurea raises the risk of low blood sugar beyond what either causes alone, and the label says a lower sulfonylurea dose may be needed. The sentence quoted here is the opening premise of a 196,138-person cohort study rather than a result of it: that study set out to test whether some sulfonylureas or some DPP-4 inhibitors make the interaction worse than others, and found no difference between them. Limit: The quoted sentence is this paper's opening statement of an interaction already accepted in the literature, not a finding this study produced. The study's own comparisons - long-acting against short-acting sulfonylureas, and peptidomimetic against non-peptidomimetic DPP-4 inhibitors - both came out null (adjusted hazard ratios 0.87, 95% CI 0.65 to 1.16, and 0.96, 95% CI 0.76 to 1.22). It therefore gives no basis for preferring one drug over another within either class. (Source 7)
• Insulin and other insulin secretagogues (clinical trial): When sitagliptin was given with insulin or a drug that makes the pancreas release insulin, more people had an episode of low blood sugar than on placebo plus the same background drug. (Source 8)
• Food, including high-fat meals (pharmacokinetic study): Food does not change how much sitagliptin gets into the blood, so it can be taken with or without a meal. This was measured directly in a pharmacokinetic study of a high-fat meal. (Source 9)

Stopping it

• No withdrawal syndrome or rebound effect is described for sitagliptin in the sources reached. What the label does say about stopping is that it should be stopped promptly if pancreatitis is suspected. (Source 10)
• The label also tells prescribers to stop it if bullous pemphigoid is suspected, and reports that people whose severe joint pain was caused by a DPP-4 inhibitor got better after stopping. (Source 11)
• The deprescribing literature for diabetes drugs in older adults singles out insulin and sulfonylureas as the high-risk medicines to de-intensify; DPP-4 inhibitors such as sitagliptin are not named among them in this 2026 evidence synthesis. (Source 12)

What goes wrong

Severe hypoglycaemia happened on sitagliptin but less often than on the three comparator drugs. (Source 2)

• Randomized trial, High certainty.
• Size: 5047 participants.
• Who: Adults with type 2 diabetes on metformin.
• How long: Mean 5.0 years.
• Result: Severe hypoglycaemia in 0.7% on sitagliptin vs 2.2% on glimepiride, 1.3% on glargine and 1.0% on liraglutide. Against glimepiride that is 1.5 percentage points fewer, about 1 severe episode avoided for every 67 people treated with sitagliptin instead over 5 years.
• Funding: independent (NIDDK and others)

Severe hypoglycemia was rare but significantly more frequent with glimepiride (in 2.2% of the participants) than with glargine (1.3%), liraglutide (1.0%), or sitagliptin (0.7%)

A 2020 meta-analysis of cardiovascular outcome trials found DPP-4 inhibitors raised the odds of acute pancreatitis against placebo. (Source 13)

• Meta-analysis, Moderate certainty.
• Size: 5 cardiovascular outcome trials vs placebo, 47,714 participants; 6 trials vs placebo or active comparator, 53,747 participants.
• Who: Adults with type 2 diabetes in cardiovascular outcome trials.
• How long: Trials of at least 12 months.
• Result: Odds ratio 1.81 (95% CI 1.21 to 2.70) against placebo and 1.54 (95% CI 1.08 to 2.18) against placebo or active comparator, without significant heterogeneity. By contrast GLP-1 receptor agonists showed no increase (OR 1.05, 95% CI 0.77 to 1.42). Pancreatitis is rare, so the absolute excess is small, and the papers do not give it.
• Funding: not stated.

Limit of this finding: The published abstract prints 'OR, 1.81; 95% CI, 1.21-2.70; p = 0.04'. Those figures cannot all be right: an interval of 1.21 to 2.70 around 1.81 corresponds to a p-value of about 0.004, ten times smaller than the 0.04 printed. The slip is in the publication itself, and the other two estimates in the same sentence are internally consistent. Read this result from the confidence interval, which excludes 1 and so does indicate a real increase, and ignore the printed p-value. The increase is attributed to DPP-4 inhibitors only - the GLP-1 receptor agonists analysed in the same paper showed no increase, and the two classes must not be merged.

In contrast, meta-analysis of five CVOTs comparing DPP-4 inhibitors with placebo (N = 47,714) and six CVOTs comparing DPP-4 inhibitors with placebo or active comparator (N = 53,747), found a significant increase (OR, 1.81; 95% CI, 1.21-2.70; p = 0.04 and OR, 1.54; 95% CI, 1.08-2.18; p = 0.02, respectively) in AP without any significant heterogeneity.

DPP-4 inhibitors as a class are associated with bullous pemphigoid, a blistering skin disease, but sitagliptin on its own was not. (Source 14)

• Meta-analysis, Low certainty.
• Size: 5 studies with cases and controls.
• Who: People with diabetes taking DPP-4 inhibitors, compared with controls.
• How long: Not stated; the included studies were retrospective.
• Result: Class adjusted OR 2.13 (95% CI 1.59 to 2.86, I-squared 46%). By drug: vildagliptin OR 5.08 (95% CI 1.70 to 15.19), linagliptin OR 2.87 (95% CI 1.06 to 7.79), sitagliptin OR 1.29 (95% CI 0.79 to 2.08, P=0.31, not significant).
• Funding: not stated.

Limit of this finding: This is a class-level estimate. In the same analysis sitagliptin on its own was not associated with bullous pemphigoid (odds ratio 1.29, 95% CI 0.79 to 2.08, an interval that crosses 1), and the drugs driving the class signal were the others. Bullous pemphigoid is also rare, so even a real class effect means a small absolute risk.

Adjusted meta-analysis revealed significant association between DPPI use and BP (OR 2.13, 95% CI 1.59-2.86, I2 = 46%, P < 0.00001). This association was stronger between vildagliptin and BP (OR 5.08, 95% CI 1.70-15.19, P = 0.004) compared to linagliptin (OR 2.87, 95%CI 1.06-7.79, P = 0.04), and no association was found between sitagliptin and BP (OR 1.29, 95%CI 0.79-2.08, P = 0.31)

The authors of the bullous pemphigoid meta-analysis say their own evidence base is retrospective and bias-prone. (Source 15)

• Meta-analysis, Low certainty.
• Size: 5 studies with cases and controls.
• Who: People with diabetes taking DPP-4 inhibitors.
• How long: Retrospective.
• Result: No new effect size; this is the review's own statement about its quality.
• Funding: not stated.

Limitations were that studies reviewed were retrospective by design which are susceptible to bias and lack of randomisation.

The label records postmarketing reports of serious harms for which no rate is known: acute pancreatitis, acute renal failure, anaphylaxis and Stevens-Johnson syndrome, severe disabling joint pain, and bullous pemphigoid. (Source 16)

• Official position, Certainty not rated.
• Size: Spontaneous postmarketing reports, denominator unknown.
• Who: People prescribed sitagliptin or another DPP-4 inhibitor.
• How long: Time to onset of joint pain varied from one day to years.
• Result: No rates are given. Spontaneous reporting cannot produce a rate, which is why this is recorded as a regulatory position rather than evidence of frequency.
• Funding: not stated.

There have been postmarketing reports of serious allergic and hypersensitivity reactions in patients treated with JANUVIA such as anaphylaxis, angioedema, and exfoliative skin conditions including Stevens-Johnson syndrome.

The label carries a heart-failure caution for sitagliptin borrowed from two other drugs in its class, not from sitagliptin trials. (Source 17)

• Official position, Certainty not rated.
• Size: Not applicable.
• Who: Regulatory position, label dated 2024-11-14.
• How long: Not applicable.
• Result: No effect size for sitagliptin. TECOS found a heart-failure hazard ratio of 1.00 for sitagliptin itself.
• Funding: not stated.

An association between dipeptidyl peptidase-4 (DPP-4) inhibitor treatment and heart failure has been observed in cardiovascular outcomes trials for two other members of the DPP-4 inhibitor class.

What the evidence supports

Sitagliptin 100 mg/day lowers HbA1c by about 0.64 percentage points more than placebo. (Source 18)

• Meta-analysis, High certainty.
• Size: 87 randomised controlled trials, 64,517 participants; 61 trials and 28,715 participants in the primary analysis.
• Who: Adults with type 2 diabetes.
• How long: Trials of at least 12 weeks; primary analysis restricted to outcomes at 24 to 54 weeks.
• Result: HbA1c mean difference -0.64% (95% credible interval -0.75 to -0.53), rated high confidence; fasting plasma glucose -15.16 mg/dL (95% CrI -18.82 to -11.55)
• Funding: not stated.

Compared with placebo, both sitagliptin 100 mg/day and linagliptin 5 mg/day demonstrated significantly superior efficacy in reducing hemoglobin A1c. The mean difference (MD) was -0.64% [95% CrI, -0.75, -0.53] (high confidence) for sitagliptin 100 mg/day

DPP-4 inhibitors as a class improved healing of diabetic foot ulcers in four small trials. (Source 4)

• Meta-analysis, Low certainty.
• Size: 4 randomised controlled trials, 285 patients.
• Who: People with diabetic foot ulcers.
• How long: Not stated in the abstract.
• Result: Complete ulcer healing RR 1.63 (95% CI 1.30 to 2.06); healing time mean difference -10.72 days (95% CI -14.61 to -6.84); ulcer size -2.36 cm² (95% CI -2.95 to -1.77). Side effects were rare and similar between groups.
• Funding: not stated.

Limit of this finding: This pooled result rests on four small randomised trials with 285 patients between them, and the authors limit their own conclusion to 'the number of included studies and the small sample size'. It is also a class-level result for DPP-4 inhibitors as a group, not a trial of sitagliptin in foot ulcers.

DPP-4 inhibitors significantly increased complete ulcer healing (RR = 1.63, 95% CI 1.30-2.06, P < .0001) and shortened healing time (MD = -10.72 days, 95% CI -14.61 to -6.84, P < .00001)

What the evidence does not support

Adding sitagliptin to usual care did not change the rate of major cardiovascular events compared with placebo. (Source 19)

• Randomized trial, High certainty.
• Size: 14,671 participants.
• Who: Adults with type 2 diabetes and established cardiovascular disease.
• How long: Median follow-up 3.0 years.
• Result: Primary composite outcome 839/7332 (11.4%; 4.06 per 100 person-years) on sitagliptin vs 851 (11.6%; 4.17 per 100 person-years) on placebo; hazard ratio 0.98 (95% CI 0.88 to 1.09). Absolute difference 0.2 percentage points, which does not exclude zero. HbA1c difference only -0.29 percentage points.
• Funding: industry-funded (Merck Sharp & Dohme, the manufacturer)

Overall, the primary outcome occurred in 839 patients in the sitagliptin group (11.4%; 4.06 per 100 person-years) and 851 patients in the placebo group (11.6%; 4.17 per 100 person-years). Sitagliptin was noninferior to placebo for the primary composite cardiovascular outcome (hazard ratio, 0.98; 95% CI, 0.88 to 1.09; P<0.001)

Sitagliptin did not increase or decrease hospitalisation for heart failure in TECOS, and pancreatitis and pancreatic cancer rates did not differ significantly. (Source 19)

• Randomized trial, High certainty.
• Size: 14,671 participants.
• Who: Adults with type 2 diabetes and established cardiovascular disease.
• How long: Median 3.0 years.
• Result: Heart-failure hospitalisation hazard ratio 1.00 (95% CI 0.83 to 1.20), P=0.98; acute pancreatitis P=0.07; pancreatic cancer P=0.32.
• Funding: industry-funded (Merck Sharp & Dohme)

Rates of hospitalization for heart failure did not differ between the two groups (hazard ratio, 1.00; 95% CI, 0.83 to 1.20; P=0.98). There were no significant between-group differences in rates of acute pancreatitis (P=0.07) or pancreatic cancer (P=0.32).

Of four drugs added to metformin, sitagliptin was the least effective at keeping HbA1c below 7.0%. (Source 2)

• Randomized trial, High certainty.
• Size: 5047 participants.
• Who: Adults with type 2 diabetes of under 10 years' duration, HbA1c 6.8 to 8.5%, already on metformin; 19.8% Black and 18.6% Hispanic or Latinx.
• How long: Mean 5.0 years.
• Result: Rate of reaching confirmed HbA1c 7.0% or higher per 100 participant-years: glargine 26.5, liraglutide 26.1, glimepiride 30.4, sitagliptin 38.1 (P<0.001 for the global test across groups)
• Funding: independent (National Institute of Diabetes and Digestive and Kidney Diseases and others)

the rates with glargine (26.5 per 100 participant-years) and liraglutide (26.1) were similar and lower than those with glimepiride (30.4) and sitagliptin (38.1)

Over 5 years none of the four add-on drugs, sitagliptin included, differed in microvascular complications, major adverse cardiovascular events or death. (Source 3)

• Randomized trial, Moderate certainty.
• Size: 5047 participants.
• Who: Adults with type 2 diabetes on metformin.
• How long: Mean 5.0 years.
• Result: Overall rates per 100 participant-years: MACE 1.0, heart-failure hospitalisation 0.4, cardiovascular death 0.3, all deaths 0.6, with no difference between groups. For 'any cardiovascular disease' the sitagliptin group had the highest rate (2.0) and a hazard ratio of 1.2 (95% CI 1.0 to 1.5) against the other three combined, on confidence limits the authors say are not adjusted for multiple comparisons.
• Funding: independent (NIDDK and others)

Limit of this finding: Each hazard ratio in this passage compares one drug with the other three pooled together, not with any single drug head to head, and the authors say the confidence limits are not adjusted for making several comparisons at once. A figure from this analysis is not a head-to-head result and should not be read as one.

The treatment groups did not differ with respect to MACE (overall rate, 1.0), hospitalization for heart failure (0.4), death from cardiovascular causes (0.3), or all deaths (0.6). There were small differences with respect to rates of any cardiovascular disease, with 1.9, 1.9, 1.4, and 2.0 in the glargine, glimepiride, liraglutide, and sitagliptin groups, respectively. When one treatment was compared with the combined results of the other three treatments, the hazard ratios for any cardiovascular disease were 1.1 (95% confidence interval [CI], 0.9 to 1.3) in the glargine group, 1.1 (95% CI, 0.9 to 1.4) in the glimepiride group, 0.7 (95% CI, 0.6 to 0.9) in the liraglutide group, and 1.2 (95% CI, 1.0 to 1.5) in the sitagliptin group.

The FDA label states sitagliptin should not be used in type 1 diabetes and has not been studied in people with a history of pancreatitis. (Source 6)

• Official position, Certainty not rated.
• Size: Not applicable.
• Who: Regulatory position held by the FDA, label dated 2024-11-14.
• How long: Not applicable.
• Result: No effect size; this is a regulatory limitation of use, not a trial result.
• Funding: not stated.

JANUVIA should not be used in patients with type 1 diabetes (1) JANUVIA has not been studied in patients with a history of pancreatitis.

A larger 2024 meta-analysis of randomised trials found no increase in pancreatitis or pancreatic cancer with DPP-4 inhibitors. (Source 20)

• Meta-analysis, Moderate certainty.
• Size: 81,737 participants with type 2 diabetes.
• Who: Adults with type 2 diabetes in randomised trials lasting over 24 weeks.
• How long: Mean follow-up 24 to 520 weeks.
• Result: Pancreatitis Peto odds ratio 0.97 (95% CI 0.74 to 1.27); pancreatic cancer Peto OR 0.88 (95% CI 0.59 to 1.30). A subgroup analysis found sitagliptin specifically associated with a reduction in pancreatitis risk.
• Funding: not stated.

The use of DPP-4 inhibitors was not associated with an increased risk of pancreatitis (Peto-OR 0.97; 95% CI: 0.74, 1.27) or pancreatic cancer (Peto-OR = 0.88; 95% CI: 0.59, 1.30).

Pooling 29 randomised trials, DPP-4 inhibitors as a class were not associated with heart failure, though saxagliptin was. (Source 21)

• Meta-analysis, Low certainty.
• Size: 29 studies, 67,873 pooled patients.
• Who: Adults with type 2 diabetes, with or without a history of cardiovascular disorders.
• How long: Varied by trial.
• Result: Risk ratio 1.06, reported with a 95% CI of 0.94-1.12 which cannot be correct for that estimate (the same paper reports 0.94 to 1.20 for its longer-follow-up subgroup). Either way the interval crosses 1 and the association is not statistically significant. A subgroup analysis found a significant association for saxagliptin specifically; the paper also reports that excluding one trial moved the overall estimate to 1.00 (0.90-1.12).
• Funding: not stated.

Limit of this finding: One of this paper's own numbers is impossible as printed: a risk ratio of 1.06 cannot sit inside an interval of 0.94 to 1.12, because the interval is lopsided around it. The same paper prints 0.94 to 1.20 for a subgroup, and 1.20 is almost certainly what the main upper bound should read. Either version crosses 1, so the class result is still no association with heart failure. The saxagliptin subgroup signal needs the same caution: the paper's entire conclusion is one sentence saying the results should be interpreted with caution because of several limitations, and dropping a single trial moved the overall estimate to exactly no effect.

The analysis for total of pooled 67,873 patients showed statistically insignificant association between the use of DPP-4 inhibitors and heart failure with Risk Ratio (RR) = 1.06, 95% Confidence Interval (CI) 0.94-1.12. Sub-group analysis found significant association of saxagliptin use with increased heart failure risk.

Across 43 randomised trials, no DPP-4 inhibitor including sitagliptin changed the risk of diabetic retinopathy. (Source 22)

• Meta-analysis, Low certainty.
• Size: 43 trials, 74,546 participants, 779 retinopathy events.
• Who: Adults with type 2 diabetes.
• How long: Varied by trial.
• Result: Odds ratios versus placebo ranged from 0.31 (omarigliptin) to 2.31 (saxagliptin), with every credible interval crossing 1.0; no dose-response signal.
• Funding: not stated.

Limit of this finding: The individual odds ratios here look dramatic - 0.31 for omarigliptin up to 2.31 for saxagliptin - but every one of their credible intervals crosses 1, which means not one of them can be told apart from no effect. The rankings the paper reports come out of a network that could not distinguish the drugs, so they are not evidence that any gliptin is better or worse than another for retinopathy.

No DPP-4 inhibitor significantly changed the DR risk versus placebo (ORs ranged from 0.31 [omarigliptin] to 2.31 [saxagliptin]; all CrIs crossed 1.0)

Where the evidence is mixed

Sitagliptin added to insulin in latent autoimmune diabetes in adults lowered HbA1c slightly across 8 small trials, but the trials disagreed with each other so strongly (I-squared 91.6%) that the pooled figure is not reliable, and the C-peptide result showed no effect. (Source 5)

• Meta-analysis, Very low certainty.
• Size: 8 randomised controlled trials, 295 participants.
• Who: Adults with latent autoimmune diabetes in adults (LADA)
• How long: Not stated in the abstract.
• Result: HbA1c weighted mean difference -0.36 (95% CI -0.61 to -0.10) with I-squared 91.6%; fasting C-peptide 0.08 (95% CI -0.02 to 0.17) with I-squared 88.8%, an interval that includes zero. The paper then states there was no significant association with either HbA1c or C-peptide regardless of duration or sample size. 8 trials, 295 participants in total.
• Funding: not stated.

Limit of this finding: The source contradicts itself on C-peptide: it says the combination 'increased fasting C-peptide' while giving an interval of -0.02 to 0.17, which includes zero, and then says in the next sentence that there was no significant association. Treat the C-peptide result as no demonstrated effect. Both pooled results also come from only 8 trials in 295 people with I-squared of 91.6% and 88.8% - a measure of how far the trials disagreed with one another, where anything above about 75% means the trials are telling different stories and a single pooled number should not be trusted.

Sitagliptin and insulin treatment lowered hemoglobin A1c (weighted mean difference -0.36, 95% confidence interval -0.61 to -0.10, I2 = 91.6%)

Where the research disagrees

Whether DPP-4 inhibitors such as sitagliptin cause acute pancreatitis

• Singh and colleagues, 2020 meta-analysis of cardiovascular outcome trials, meta-analysis of 5 to 6 cardiovascular outcome trials, 47,714 to 53,747 participants: In contrast, meta-analysis of five CVOTs comparing DPP-4 inhibitors with placebo (N = 47,714) and six CVOTs comparing DPP-4 inhibitors with placebo or active comparator (N = 53,747), found a significant increase (OR, 1.81; 95% CI, 1.21-2.70; p = 0.04 and OR, 1.54; 95% CI, 1.08-2.18; p = 0.02, respectively) in AP without any significant heterogeneity. (Source 13)
• 2024 updated meta-analysis of randomised controlled trials, Clinical Therapeutics, meta-analysis of randomised trials, 81,737 participants: The use of DPP-4 inhibitors was not associated with an increased risk of pancreatitis (Peto-OR 0.97; 95% CI: 0.74, 1.27) (Source 20)
• TECOS investigators, 2015, single randomised controlled trial, 14,671 participants: There were no significant between-group differences in rates of acute pancreatitis (P=0.07) or pancreatic cancer (P=0.32). (Source 19)

How much

• Reference intake: There is no reference intake for a prescription medicine. The dose is set by the prescriber. As a position, the FDA label in force on 2024-11-14 gives 100 mg once daily, with or without food. (Source 23)
• Upper limit: As a position, the label's maximum is 100 mg once daily, reduced to 50 mg once daily when eGFR is 30 to under 45 mL/min/1.73 m2 and to 25 mg once daily below that or on dialysis (label dated 2024-11-14). No independent upper limit is set by any scientific body. (Source 24)
• Studied: The dose pooled across 87 randomised trials in the 2026 network meta-analysis was sitagliptin 100 mg/day. (Source 18)
• Studied: TECOS added sitagliptin or placebo to the existing therapy of 14,671 people with type 2 diabetes and cardiovascular disease and followed them a median of 3.0 years. (Source 25)
• Studied: GRADE randomised 5047 people already on metformin to add sitagliptin, glimepiride, liraglutide or insulin glargine and followed them a mean of 5.0 years. (Source 26)

A common belief, and what the research shows

The belief: Because sitagliptin works on the gut hormones that also protect the heart, it must protect the heart too.

What the research shows: It does not, on the trial evidence. TECOS was a non-inferiority trial and reported that "Sitagliptin was noninferior to placebo for the primary composite cardiovascular outcome (hazard ratio, 0.98; 95% CI, 0.88 to 1.09; P<0.001)" - meaning no worse, not better. The same trial found "Rates of hospitalization for heart failure did not differ between the two groups (hazard ratio, 1.00; 95% CI, 0.83 to 1.20; P=0.98)." Five years of GRADE found no cardiovascular advantage either.

Questions and answers

What is it?

Sitagliptin is a prescription tablet for type 2 diabetes, sold as Januvia and in combinations with metformin. It belongs to a class called DPP-4 inhibitors, or gliptins, and was first approved in the United States in 2006. It is not a nutrient and is not found in food. (Source 6)

What does it do in the body?

It blocks an enzyme called DPP-4 that destroys the gut hormones GLP-1 and GIP. With the enzyme blocked those hormones last longer, so the pancreas releases more insulin and less glucagon after a meal. Crucially the effect is glucose-dependent: it only pushes insulin up when blood glucose is normal or high, which is why it rarely causes low blood sugar by itself. (Source 1)

Is it good or bad for you?

Good for lowering blood sugar with little risk of hypoglycaemia; not good as heart protection, because it does not provide any. In the GRADE trial, where four drugs were added to metformin in 5047 people over 5 years, sitagliptin was the weakest at holding HbA1c under 7.0% - but it also had the lowest rate of severe hypoglycaemia, 0.7% against 2.2% on glimepiride. Whether it causes pancreatitis is genuinely unsettled in the literature. (Source 2)

How do you get more of it?

You cannot get more of it outside a prescription; it is not in any food or supplement. The amount a person takes is decided by their prescriber. As a position, the label in force on 2024-11-14 gives 100 mg once daily, and says food makes no difference to timing. (Source 23)

If it is harmful, what reduces it?

Sitagliptin leaves the body on its own within about a day - its half-life is 12.4 hours - and more than three quarters of a dose is cleared unchanged by the kidneys, so levels build up when kidney function is poor and the dose is cut for that reason. The label's instruction when a serious reaction is suspected is to stop the drug, and people whose joint pain was caused by a gliptin recovered after stopping. Stopping any diabetes drug changes blood sugar control, so it is a prescriber's decision. (Source 27)

Why might someone be low in it or missing it?

This question does not apply in the way it does to a nutrient: nobody is naturally low in sitagliptin. The nearest equivalent is being given a lower dose than the standard one, and the commonest reason for that is reduced kidney function, where the label drops the dose to 50 mg or 25 mg once daily. People with type 1 diabetes are not given it at all. (Source 23)

Which whole foods contain it or feed it?

No whole food contains sitagliptin or supplies it. Food does matter in one narrow sense: it was tested directly and makes no difference to absorption, so the tablet does not have to be timed around meals. The gut hormones sitagliptin preserves are released by eating, so the drug's effect follows meals whatever those meals are. (Source 1)

What happens if you do not have it?

Without sitagliptin (and without something in its place) blood sugar sits higher: across 87 randomised trials the drug lowered HbA1c by 0.64 percentage points more than placebo. What does not happen if you stop is any change in heart attack, stroke or death risk - TECOS found the same rate of cardiovascular events on sitagliptin and on placebo, 11.4% against 11.6% over three years. (Source 19)

How can you test for it?

There is no routine blood test for sitagliptin itself; what gets measured is the thing it is meant to change. HbA1c is the standard measure and in GRADE it was checked every three months, with a result confirmed on a repeat before it counted. Kidney function is the other test that matters, because the label requires it before starting and periodically after, and it sets the dose. Neither test tells you the drug is present, only whether it is working and whether the dose still fits. (Source 26)

References

1. DailyMed / Merck Sharp & Dohme LLC (FDA label). JANUVIA (sitagliptin) tablets, for oral use - full prescribing information. 2024. Read the source
2. The New England journal of medicine. Glycemia Reduction in Type 2 Diabetes - Glycemic Outcomes.. 2022. PMID 36129996, DOI 10.1056/nejmoa2200433. Read the source
3. The New England journal of medicine. Glycemia Reduction in Type 2 Diabetes - Microvascular and Cardiovascular Outcomes.. 2022. PMID 36129997, DOI 10.1056/nejmoa2200436. Read the source
4. The international journal of lower extremity wounds. Efficacy of Dipeptidyl Peptidase-4 Inhibitors in Diabetic Foot Ulcer Healing: A Meta-Analysis of Randomized Clinical Trials. [structured-abstract section 'Results']. 2026. PMID 41528934, DOI 10.1177/15347346251413943. Read the source
5. Journal of diabetes investigation. Efficacy and safety of sitagliptin and insulin for latent autoimmune diabetes in adults: A systematic review and meta-analysis.. 2022. PMID 35445591, DOI 10.1111/jdi.13814. Read the source
6. DailyMed / Merck Sharp & Dohme LLC (FDA label). JANUVIA (sitagliptin) tablets, for oral use - Highlights of Prescribing Information, INDICATIONS AND USAGE. 2024. Read the source
7. Clinical pharmacology and therapeutics. Pharmacologic Heterogeneity and Risk of Severe Hypoglycemia with Concomitant Use of Sulfonylureas and DPP-4 Inhibitors: Population-Based Cohort Study. [abstract: background, aim and methods]. 2023. PMID 37326010, DOI 10.1002/cpt.2975. Read the source
8. DailyMed / Merck Sharp & Dohme LLC (FDA label). JANUVIA (sitagliptin) tablets, for oral use - Full Prescribing Information, section 6.1 Clinical Trials Experience, Hypoglycemia. 2024. Read the source
9. DailyMed / Merck Sharp & Dohme LLC (FDA label). JANUVIA (sitagliptin) tablets, for oral use - Full Prescribing Information, section 12.3 Pharmacokinetics, Effect of Food. 2024. Read the source
10. DailyMed / Merck Sharp & Dohme LLC (FDA label). JANUVIA (sitagliptin) tablets, for oral use - full prescribing information. 2024. Read the source
11. DailyMed / Merck Sharp & Dohme LLC (FDA label). JANUVIA (sitagliptin) tablets, for oral use - full prescribing information. 2024. Read the source
12. Clinical interventions in aging. Frailty-Informed Deprescribing of Glucose-Lowering Therapy in Older Adults with Type 2 Diabetes: A Pragmatic Clinical Framework. [a targeted evidence synthesis with an author-constructed decision framework, not a systematic review or meta-analysis; structured-abstract section 'Results']. 2026. PMID 42436877, DOI 10.2147/cia.s622440. Read the source
13. Expert review of clinical pharmacology. Risk of acute pancreatitis with incretin-based therapy: a systematic review and updated meta-analysis of cardiovascular outcomes trials.. 2020. PMID 32129106, DOI 10.1080/17512433.2020.1736041. Read the source
14. The Australasian journal of dermatology. Dipeptidyl peptidase-4 inhibitors and bullous pemphigoid: A systematic review and adjusted meta-analysis.. 2020. PMID 31215644, DOI 10.1111/ajd.13100. Read the source
15. The Australasian journal of dermatology. Dipeptidyl peptidase-4 inhibitors and bullous pemphigoid: A systematic review and adjusted meta-analysis.. 2020. PMID 31215644, DOI 10.1111/ajd.13100. Read the source
16. DailyMed / Merck Sharp & Dohme LLC (FDA label). JANUVIA (sitagliptin) tablets, for oral use - full prescribing information. 2024. Read the source
17. DailyMed / Merck Sharp & Dohme LLC (FDA label). JANUVIA (sitagliptin) tablets, for oral use - full prescribing information. 2024. Read the source
18. Journal of evidence-based medicine. Comparison and Evaluation of Efficacy and Safety of Sitagliptin and Linagliptin in Adults With Type 2 Diabetes Mellitus: A Bayesian Network Meta-Analysis.. 2026. PMID 41918146, DOI 10.1111/jebm.70128. Read the source
19. The New England journal of medicine. Effect of Sitagliptin on Cardiovascular Outcomes in Type 2 Diabetes.. 2015. PMID 26052984, DOI 10.1056/nejmoa1501352. Read the source
20. Clinical therapeutics. Pancreatitis and Pancreatic Cancer Risk Among Patients With Type 2 Diabetes Receiving Dipeptidyl Peptidase 4 Inhibitors: An Updated Meta-Analysis of Randomized Controlled Trials.. 2024. PMID 39084911, DOI 10.1016/j.clinthera.2024.06.015. Read the source
21. Endocrinology, diabetes & metabolism. Dipeptidyl Peptidase-4 Inhibitors Associated Heart Failure Events in Adult Patients With Type-2 Diabetes Mellitus Treated With Dipeptidyl Peptidase-4 Inhibitors: A Systematic Review and Meta-Analysis.. 2026. PMID 42614024, DOI 10.1002/edm2.70316. Read the source
22. Journal of diabetes and its complications. Dipeptidyl peptidase-4 inhibitors and diabetic retinopathy in type 2 diabetes: A network meta-analysis of randomized clinical trials.. 2026. PMID 41314124, DOI 10.1016/j.jdiacomp.2025.109234. Read the source
23. DailyMed / Merck Sharp & Dohme LLC (FDA label). JANUVIA (sitagliptin) tablets, for oral use - Highlights of Prescribing Information, DOSAGE AND ADMINISTRATION (the flattened cells of the renal dose table are omitted; the dose-to-eGFR pairing is recorded in full in januvia-label-dosing-renal). 2024. Read the source
24. DailyMed / Merck Sharp & Dohme LLC (FDA label). JANUVIA (sitagliptin) tablets, for oral use - Full Prescribing Information, section 2.2 Recommendations for Use in Renal Impairment. 2024. Read the source
25. The New England journal of medicine. Effect of Sitagliptin on Cardiovascular Outcomes in Type 2 Diabetes.. 2015. PMID 26052984, DOI 10.1056/nejmoa1501352. Read the source
26. The New England journal of medicine. Glycemia Reduction in Type 2 Diabetes - Glycemic Outcomes.. 2022. PMID 36129996, DOI 10.1056/nejmoa2200433. Read the source
27. DailyMed / Merck Sharp & Dohme LLC (FDA label). JANUVIA (sitagliptin) tablets, for oral use - Full Prescribing Information, section 12.3 Pharmacokinetics, Absorption. 2024. Read the source