Niacin (nicotinic acid)
Summary
Gram doses of nicotinic acid improve cholesterol numbers, but large trials in people already on statins found no fewer heart attacks or strokes and more harms, including new diabetes (no-effect grade, high-quality evidence). Flushing starts at 30–50 mg, sustained-release forms are more liver-toxic, and the supplement upper limit is 35 mg/day.
Key facts
- No reduction in deaths, heart attacks or strokes across 23 RCTs (39,195 people)[5]
- Risk of new-onset diabetes up 34% (RR 1.34) in 11 trials (26,340 people)[3]
- 52% of patients on sustained-release niacin developed hepatotoxicity vs none on immediate-release[4]
- UL for supplemental niacin 35 mg/day for adults, set to avoid flushing[7]
- 30–50 mg or more of nicotinic acid typically causes flushing[7]
- 5 graded human outcomes on this page
- October 13, 2026
How it works
EstablishedAt pharmacological doses, nicotinic acid acts on receptors that change lipoprotein metabolism, raising HDL 10–30% and lowering LDL, triglycerides and lipoprotein(a); nicotinamide does not bind these receptors and has no lipid effect.[7]
What the human research says
Human trials only, one row per outcome. The form column shows what each trial actually tested. Grades follow our evidence-grading policy.
At a glance: strongest evidence per outcome
| Outcome | Grade | Effect | Strongest evidence | Rows |
|---|---|---|---|---|
| Blood sugar | Grade B | Worsened | Meta-analysis of 11 trials · n = 26,340 · 2016 | 1 row |
| Liver | Grade C | Worsened | RCT · n = 46 · 1994 | 1 row |
| Heart | No effect | No effect | Meta-analysis of 23 trials · n = 39,195 · 2017 | 3 rows |
Every graded row
| Outcome | Effect | Grade | Best evidence | PMID |
|---|---|---|---|---|
| New-onset diabetes | Worsened | Grade B | Meta-analysis · n = 26,340 · 2016 Non-diabetic adults in 11 cardiovascular niacin RCTs Form: Niacin therapy · Dose: Pharmacological doses · Duration: Mean 3.6 years New-onset diabetes RR 1.34 (95% CI 1.21–1.49), about one extra case per 43 people treated for 5 years. | 26370223 (opens PubMed)Source [3] |
| Liver toxicity: sustained- vs immediate-release niacin | Worsened | Grade C | RCT · n = 46 · 1994 Adults with high LDL cholesterol Form: Sustained-release vs immediate-release nicotinic acid · Dose: 500 mg rising to 3,000 mg/day · Duration: 6 weeks per dose step Hepatotoxicity in 12 of 23 (52%) on sustained-release vs none on immediate-release; 78% vs 39% withdrew before 3,000 mg/day. | 8309029 (opens PubMed)Source [4] |
| Mortality, heart attack and stroke | No effect | No effect | Meta-analysis · n = 39,195 · 2017 Adults with or without cardiovascular disease in 23 RCTs (median 48% with prior MI) Form: Niacin alone or added to other therapy · Dose: Median 2 g/day · Duration: Median 11.5 months No reduction in overall mortality (RR 1.05), cardiovascular mortality (RR 1.02), heart attacks (RR 0.93) or strokes (RR 0.95). | 28616955 (opens PubMed)Source [5] |
| Major vascular events on top of statins | No effect | No effect | RCT · n = 25,673 · 2014 Adults with vascular disease on statin therapy Form: Extended-release niacin with laropiprant · Dose: 2 g/day niacin + 40 mg laropiprant · Duration: Median 3.9 years Major vascular events 13.2% vs 13.7% (rate ratio 0.96); excess serious diabetes complications (+3.7 points), new diabetes (+1.3), GI, muscle, skin, infection and bleeding events. | 25014686 (opens PubMed)Source [2] |
| Cardiovascular events with intensive statin therapy | No effect | No effect | RCT · n = 3,414 · 2011 Adults with atherosclerotic disease and low HDL on simvastatin Form: Extended-release niacin · Dose: 1,500–2,000 mg/day · Duration: Mean 3 years (stopped early) HDL rose from 35 to 42 mg/dL and triglycerides fell, but the primary endpoint occurred in 16.4% vs 16.2% (HR 1.02). | 22085343 (opens PubMed)Source [1] |
Forms
| Form | Elemental % | Absorption (human data) | GI tolerance | Studied for |
|---|---|---|---|---|
| Immediate-release nicotinic acid | Varies[7] | Almost completely absorbed, even at 3–4 g.[7] | Flushing is common; in the SR-vs-IR trial 39% withdrew before 3,000 mg/day, mostly for flushing, fatigue and skin changes.[4][7] | Lipid trials |
| Sustained- or extended-release nicotinic acid | Varies[7] | Designed to reduce flushing.[7] | More likely to cause liver toxicity; 52% hepatotoxicity in one trial.[4][7] | AIM-HIGH, HPS2-THRIVE (prescription ER forms) |
| Inositol hexanicotinate ("no-flush" niacin) | Varies[7] | Niacin absorption averages about 30% lower than from nicotinic acid and varies widely.[7] | Does not cause flushing.[7] | Not shown to match nicotinic acid in lipid effects |
Studied doses
Doses studied in human trials ranged from 500 mg/day rising to 3,000 mg/day of nicotinic acid over weeks, with outcome trials using 1,500–2,000 mg/day of extended-release niacin for 3–4 years.[1][2][4]
Intake reference: RDA 16 mg NE/day for men and 14 mg NE/day for women (18 mg in pregnancy); the UL of 35 mg/day applies to supplemental niacin only.[7] UL for adults from supplements and fortified foods, based on flushing; it does not apply under medical supervision. NIH ODS fact sheet (updated November 18, 2022).
Interactions
Published interactions only: each row cites a source. No row means none was found in our sources, not that a combination is safe.
Medicines
Medicines that can lower levels
- Isoniazid and pyrazinamideNIH fact sheetThese drugs interfere with niacin production from tryptophan and can lead to deficiency (pellagra).[7]
Who should be careful
- Diabetes or prediabetes
- Niacin raised new-onset diabetes risk 34% and worsened diabetes control in trials.[2][3]
- Liver disease
- High doses can be hepatotoxic, especially sustained-release forms; a case report describes marked liver injury after niacin was used in an attempt to pass a urine drug test.[4][6][7]
- Gout or unexplained stomach symptoms
- ACC/AHA advice cited by ODS is to stop nicotinic acid with acute gout, unexplained abdominal pain or GI symptoms.[7]
Talk to your pharmacist or clinician before starting a supplement, especially if you take prescription medicines.
Frequently asked questions
Does niacin lower heart attack risk?
Not in modern trials: in people on statins, niacin did not reduce heart attacks, strokes or deaths despite better cholesterol numbers.[1][2][5]
Why does niacin cause flushing?
Nicotinic acid widens small skin blood vessels; 30–50 mg or more usually causes redness, warmth and itching.[7]
Is no-flush niacin the same?
No. Inositol hexanicotinate releases less niacin (about 30% lower absorption on average) and does not cause flushing.[7]
What is the upper limit for niacin?
The adult UL for supplemental niacin is 35 mg/day; higher doses should be medically supervised.[7]
Is niacin the same as niacinamide?
Both are vitamin B3, but niacinamide does not cause flushing or change blood lipids.[7]
Sources
- [1]Niacin in patients with low HDL cholesterol levels receiving intensive statin therapy.. N Engl J Med, 2011. RCT
- [2]Effects of extended-release niacin with laropiprant in high-risk patients.. N Engl J Med, 2014. RCT
- [3]Niacin therapy and the risk of new-onset diabetes: a meta-analysis of randomised controlled trials.. Heart, 2016. meta-analysis
- [4]A comparison of the efficacy and toxic effects of sustained- vs immediate-release niacin in hypercholesterolemic patients.. JAMA, 1994. RCT
- [5]Niacin for primary and secondary prevention of cardiovascular events.. Cochrane Database Syst Rev, 2017. meta-analysis
- [6]Hepatotoxicity upon using niacin to pass a drug test: A case report.. J Am Pharm Assoc (2003), 2018. observational
- [7]Niacin: Fact Sheet for Health Professionals. NIH Office of Dietary Supplements, 2022. fact-sheet
What changed
- Editorial evidence re-review: the new-diabetes key fact cited the wrong source (a hepatotoxicity case report); it now cites the 2016 Heart meta-analysis (PMID 26370223). All trial numbers and ODS facts re-checked.
- Second-pass self-review: every PMID re-checked against NCBI esummary (title, year, no retraction or expression-of-concern flag), every dose, n and duration re-read against the PubMed abstract or PMC full text, grades held conservative, and interactions limited to what a cited source supports.
- First draft. Source titles and years taken from NCBI E-utilities esummary; abstracts read for every evidence row.