Niacin and heart outcomes: why better lipids were not enough

The outcome matters more than the attractive marker

Niacin provides a clear example of why changing a familiar biomarker does not guarantee the clinical outcome people want. Large trials in patients already receiving statin-based treatment found no added cardiovascular benefit from the investigated niacin regimens, despite favorable changes in lipid measurements. ^{[1] [2]} The result matters to NAD-related shopping because niacin is also discussed as a precursor. It does not mean every NAD-related molecule has the same effects. The useful lesson is about evidence: a biological change needs to be connected to the actual outcome, and benefits need to be assessed alongside harms. A desirable laboratory direction is a reason to investigate, not a substitute for that investigation.

This distinction changes how a comparison should be presented. A product may improve a marker that is associated with lower risk in observational data without improving outcomes when used as an intervention. Association describes who tends to experience events; a randomized treatment comparison asks what happens when the marker is deliberately changed. Those questions can produce different answers without either measurement being fraudulent or useless.

AIM-HIGH tested an add-on to statins

AIM-HIGH randomized 3,414 people with established atherosclerotic disease and residual lipid abnormalities despite intensive statin-based treatment. The active regimen used oral extended-release niacin 1,500–2,000 mg/day; background simvastatin was 40–80 mg/day, with ezetimibe when needed. These describe the trial, not a personal regimen. After roughly three years, primary events occurred in 16.4 percent assigned niacin and 16.2 percent assigned placebo (hazard ratio 1.02, 95% CI 0.87–1.21). The study stopped early for lack of expected benefit despite improved lipid values. ^[1] Background treatment is part of the finding, so it cannot answer every possible historical use of niacin or be extrapolated to an unrelated NAD product.

HPS2-THRIVE tested a combination regimen

HPS2-THRIVE randomized 25,673 high-risk participants receiving statin-based treatment to oral extended-release niacin 2 g plus laropiprant 40 mg daily or placebo. Median follow-up was 3.9 years. These pharmacologic trial amounts are not supplement advice. Major vascular events occurred in 13.2 percent assigned the combination and 13.7 percent assigned placebo (rate ratio 0.96, 95% CI 0.90–1.03; P=0.29). ^[2] Laropiprant was included to affect flushing, so the experiment did not isolate every effect of niacin alone. It nevertheless tested a real clinical strategy at substantial scale. Favorable lipid changes did not translate into a statistically significant reduction in major events under those conditions.

For the broader interpretation, read why a biomarker does not guarantee benefit.

The harms belong beside the null result

The null vascular result was accompanied by serious harms. Relative to placebo, HPS2-THRIVE reported absolute excesses of 3.7 percentage points for serious disturbances in diabetes control, 1.3 points for new diabetes, 1.4 points for infection and 0.7 points for bleeding. ^[2] A later analysis examined these events in greater detail. ^[3] These are estimates for niacin–laropiprant in this high-risk, statin-treated population; they are not risk estimates for NR, NMN or every supplement amount. They explain why familiar vitamin terminology cannot establish safety. Net value requires both meaningful benefits and harms, rather than favorable laboratory changes shown separately from the participants’ experience.

Run-in selection affects what safety means

The HPS2-THRIVE program used a prerandomization run-in that screened for the ability to tolerate and adhere to the regimen, and many potential participants did not proceed to randomization. ^[4] That detail affects how a reader interprets tolerability among those ultimately enrolled. A selected group that has already passed an initial test is different from everyone who might start the treatment in ordinary care. Run-in designs can improve the ability to answer an efficacy question, but they also make it important to examine what happened before the randomized phase. The safety story should include selection and discontinuation, not only event counts among the people who remained.

This is nicotinic acid, not every NAD product

The niacin discussed in these trials was nicotinic acid in a pharmacologic treatment strategy. It should not be casually renamed nicotinamide, NR, NMN or direct NAD. Those are distinct interventions with different evidence bases. The cardiovascular trials do not prove that an NR capsule shares the same benefit-harm profile, and they do not support copying their amounts into a supplement routine. The converse is also true: a short NR biomarker trial cannot erase the niacin outcome findings. Keep each preparation and clinical question intact. A broad pathway connection is insufficient to transfer either a positive promise or a negative risk estimate across every product in the category.

The lesson for a biomarker-based purchase

For a consumer, the practical question is whether the advertised result is a measured benefit or an appealing intermediate marker. A higher HDL value and a higher NAD value are different measurements, yet both can be misused if the seller assumes that moving them necessarily improves health. The niacin trials illustrate why the outcome needs direct testing. Evaluate products using their actual ingredients, applicable evidence and manageable purchase terms. Do not replace prescribed cardiovascular care or adopt a research regimen from a review. The strongest conclusion is measured: better-looking laboratory values can coexist with no added clinical benefit and meaningful harms. A credible product claim should show how it crosses that gap.

For a practical product comparison, see our Life Extension NAD+ Cell Regenerator review; its purchase terms do not establish a clinical result.