NAD+ and Heart Failure: Why SIRT3 May Not Be the Answer

NAD+ supplements can raise circulating NAD+ in people. Whether they improve a failing human heart remains unknown. The distinction matters because much of the marketing for nicotinamide riboside (NR) and nicotinamide mononucleotide (NMN) jumps from a blood biomarker to a cardiac promise without proving that an oral dose reaches heart muscle at a useful concentration.

The cardiac rationale is scientifically credible, especially in animals. Aging and cardiac stress can lower myocardial NAD+, impair mitochondrial energy production, and weaken a cell type that spends most of its working life making ATP. But the best mechanistic mouse evidence also delivers an awkward correction: NR protected overloaded mouse hearts even when the animals lacked SIRT3, the mitochondrial enzyme often presented as the central reason NAD+ supplements should help the heart.

That does not make NR useless. It means the tidy supplement-label story is probably incomplete.

Low NADPH in heart failure is a signal, not a treatment result

The closest human evidence in this area comes from Qian and colleagues, who measured plasma NADPH in 12 patients with New York Heart Association class II–IV heart failure and 12 matched controls. Patients with heart failure had significantly lower circulating NADPH. The same research program reported improved contractility after exogenous NADPH in isolated toad hearts and in mouse models of heart failure caused by isoproterenol or pressure overload.

Those findings make NADPH biologically interesting, but they do not show that NADPH treats heart failure in people. The human component was a small cross-sectional comparison, so low plasma NADPH could be a consequence of heart failure rather than a cause. The contractility experiments occurred in isolated tissue and animals. NADPH is also not an oral, clinically deployable supplement.

The investigators found that a pharmacologic SIRT3 inhibitor, 3-TYP, partly reduced the animal effect. That result cannot establish that SIRT3 caused the benefit: drug inhibitors can affect targets beyond the one on the label, and a genetic knockout test answers a different, stronger question.

Why NAD+ became a cardiac target

NAD+ is required for mitochondrial energy metabolism and for the activity of sirtuins, a family of enzymes that includes SIRT3. In preclinical models of aging, pressure overload, ischemic injury, and diabetic cardiomyopathy, myocardial NAD+ declines under stress; restoring it often improves cardiac function. A 2026 PRISMA-guided systematic review summarizes this broad preclinical rationale while also documenting how thin the cardiac clinical evidence remains.

The traditional explanation centers on SIRT3, the main mitochondrial deacetylase. Lower NAD+ could reduce SIRT3 activity, leaving mitochondrial proteins excessively acetylated and respiratory machinery less efficient. Aging may also disturb SIRT3 localization within cells, potentially leaving less of the enzyme where it needs to be: mitochondria. That model is plausible, but plausible mechanisms are not clinical endpoints.

Animal studies support more than one NAD+-dependent route. In diabetic mice, NR increased cardiac NAD+ and improved mitochondrial fusion through a SIRT1–PGC1α–PPARα–Mfn2 pathway. SIRT1 acts mainly in the cytosol and nucleus, not the mitochondrial compartment dominated by SIRT3. The study reported improved mitochondrial function and lower oxidative stress in diabetic hearts, but it was still a mouse experiment, not a human heart-failure trial. Hu et al. (2022) therefore supports a mechanism worth testing, not a treatment recommendation.

Likewise, NR improved mitochondrial oxidative phosphorylation and ATP-related measures in fetal guinea-pig hearts exposed to maternal hypoxia. Thompson et al. (2024) studied a gestational-hypoxia model, which is useful for mitochondrial biology but differs sharply from age-related human heart failure.

The SIRT3 plot twist: NR worked without SIRT3

A 2023 Circulation study by Walker and colleagues put the dominant mechanism to a direct test. Researchers subjected wild-type mice and Sirt3 knockout mice to surgical pressure overload, then gave NR by intraperitoneal injection at about 500 mg/kg/day. NR produced equivalent cardiac protection in both genotypes. The investigators confirmed increased cardiac NAD+ in isolated perfused hearts with carbon-13 NMR spectroscopy.

If SIRT3 were required for NR's cardioprotective effect in this model, deleting Sirt3 should have erased or materially weakened that protection. It did not. The study instead implicated short-chain dehydrogenase/reductase proteins, particularly MRPP2/HSD17B10, a component of mitochondrial RNase P. Higher NAD+ was associated with restored mitochondrial RNA processing and respiratory-complex assembly.

The researchers also found corresponding SDR-expression changes in failing human left-ventricular tissue collected during transplantation or left-ventricular assist-device implantation, compared with non-failing donor hearts. That is valuable human tissue corroboration, but it is ex vivo evidence. It cannot tell us whether taking oral NR changes SDR proteins, myocardial NAD+, symptoms, ejection fraction, hospitalization, or survival in a patient.

Walker et al. do not prove that SIRT3 never matters in cardiac metabolism. They do show that SIRT3 was dispensable for NR protection in one rigorous pressure-overload mouse experiment. The SDR explanation now needs independent replication and much more work before anyone can use it to select a dose or a patient population.

Blood NAD+ is not heart NAD+

Human pharmacokinetic evidence is real and useful. A randomized comparison by Christen et al. (2026) found that NR, NMN, and niacin all increased circulatory NAD+ in humans, while producing different metabolomic and microbial signatures. The trial measured blood and plasma, not myocardial tissue.

That missing step is the translational problem. The animal studies that most clearly improved cardiac outcomes used intraperitoneal NR at about 500 mg/kg/day. Typical human oral doses of 250 to 1,000 mg/day equal roughly 3 to 12 mg/kg, a per-kilogram gap of about 40- to 150-fold. Injection and oral dosing also differ in absorption, first-pass metabolism, and tissue distribution. A higher blood NAD+ reading does not prove therapeutic delivery to a human cardiomyocyte.

Human trials do show some noncardiac effects. In a six-week randomized, double-blind trial of 48 healthy amateur runners, NMN at 300, 600, or 1,200 mg/day alongside exercise training improved aerobic capacity relative to placebo. Liao et al. (2021) studied fitness in a healthy athletic population, not heart failure. A 2025 meta-analysis of 12 NMN trials involving 513 adults found that NMN increased blood NAD+, while most clinically relevant glucose and lipid outcomes did not differ significantly from control. The review also found risk-of-bias concerns in several included trials.

A separate meta-analysis of NR and NMN for skeletal muscle found at most modest effects on muscle mass and function. Skeletal-muscle results cannot substitute for cardiac outcomes, although they offer a sensible calibration against extravagant claims.

The cardiovascular precedent is sobering

Large cardiovascular-outcome trials of niacin, another NAD+ precursor, provide the strongest reason to resist a simple "raise NAD+, save the heart" theory. AIM-HIGH enrolled 3,414 high-risk participants, and HPS2-THRIVE enrolled 25,673. They found null or unfavorable cardiovascular results despite niacin's ability to increase NAD+ through a pathway that differs from NR or NMN. These trials do not settle the case against NR or NMN, but they show that a plausible NAD+ pathway and a favorable biomarker do not guarantee fewer cardiac events.

The only human NR heart-failure randomized trial described here, HF-AF ENERGY, enrolled a specific group with heart failure with reduced ejection fraction and atrial fibrillation, and used an arrhythmia endpoint. It cannot establish prevention, improved systolic function, or benefit across heart-failure populations.

What a careful buyer can conclude

Oral NR and NMN reliably raise circulating NAD+ in humans. No adequately powered human trial has shown that either supplement improves cardiac function or clinical heart-failure outcomes.

For healthy adults, NR and NMN have a modest human evidence base for circulating NAD+ changes and limited evidence for metabolic or exercise-related outcomes. They do not have an established cardiac indication. For people with diagnosed heart failure or significant cardiovascular disease, the evidence does not support treating NR or NMN as a therapy, and supplement decisions belong in a conversation with the clinician managing the condition.

The studies that would change this conclusion are straightforward to describe and difficult to run: oral-dose studies that measure myocardial NAD+ directly, independent replication of the Sirt3-knockout and SDR findings, and adequately powered cardiac trials that measure ejection fraction, hospitalization, or mortality. Until then, the most accurate description of the NAD+–heart story is promising animal biology with an unproven human destination.