Heart failure

The failing heart is starving for energy in a specific biochemical sense: it runs low on NAD+, a molecule central to how cells generate ATP and regulate metabolism. Two independent lines of human-tissue evidence establish this depletion as a real phenomenon, and researchers are now testing whether replenishing NAD+ can help.

A depleted salvage pathway

Cardiac muscle from patients with heart failure with reduced ejection fraction (HFrEF) shows roughly a 30% loss of NAD+ compared with healthy myocardium. The mechanism involves a rewiring of the salvage pathway: NMRK2, an enzyme that can synthesize NAD+ from nicotinamide riboside (NR), is upregulated, while NAMPT, the main rate-limiting enzyme of the canonical salvage route, is downregulated [1]. A separate mechanistic study confirmed that failing human hearts carry an improved NADH/NAD+ ratio alongside widespread mitochondrial protein hyperacetylation, a signature of NAD+ insufficiency. In mice, restoring the NAD+ pool reversed that hyperacetylation and improved cardiac function [2]. These two datasets converge on the same conclusion from different angles, which strengthens the case.

What NR does in the lab and in early human trials

In mouse cardiomyopathy models, NR supplementation restores cardiac NAD+ and preserves function [1]. Translating that to humans required first answering whether NR is safe in people with already-compromised hearts and whether it raises NAD+ in that population.

A small double-blind randomized controlled trial in HFrEF patients (n=30) found that NR was well tolerated and roughly doubled whole-blood NAD+ concentrations [3]. That doubling confirms target engagement: the supplement reaches the bloodstream and shifts the biochemistry it is supposed to shift. The same trial noted correlations between NAD+ increases and anti-inflammatory markers, including effects on the NLRP3 pathway, but those correlations were within-group observations, not controlled comparisons, so they cannot be read as proof that NR reduces inflammation in heart failure patients.

An earlier open-label pilot reported lower inflammatory gene expression after NR in a handful of completers [4]. Because that study had no control group, Verhoeven's analysis suggests the inflammatory signal is hypothesis-generating at best, not confirmatory.

What remains unresolved

Neither trial was designed to measure clinical outcomes such as hospitalization rates, ejection fraction improvement, or mortality. The human evidence, while encouraging on safety and biomarker grounds, sits firmly in the early-phase category. Raising NAD+ in the blood is not the same as demonstrating that doing so helps the heart pump better or that patients live longer.

Where the evidence is contested

The NAD+ depletion finding in human failing myocardium is well-supported by primary literature [1][2]. No major contradictory human-tissue studies appear on this page.

The therapeutic question is open. NR doubles blood NAD+ in HFrEF patients [3], and mouse data show functional benefit [1], but the leap from biomarker change to clinical benefit has not been tested in an adequately powered, outcome-driven trial. The anti-inflammatory signals reported in Wang 2022 are correlational [3], and the open-label pilot [4] is too small and uncontrolled to draw conclusions from. Researchers and clinicians reasonably disagree about whether the mechanistic plausibility is strong enough to prioritize larger trials now, or whether the early-phase data are too preliminary to guide any clinical decision.

References

1. Diguet 2018 (Circulation) — NR in heart failure; NMRK2 ~40-fold in DCM; human HFrEF confirmed · PubMed 29217642
2. Lee 2016, Circulation — NAD+ redox normalization as HF therapy (human tissue) · PubMed 27489254
3. Wang et al. 2022 (JACC:BTS) — NR uptitrated to 2g × 12wk in HFrEF (n=30); safety primary = safe; NLRP3 correlation within-NR arm only (no… · PubMed 36644285
4. Zhou et al. 2020 (JCI) — Open-label pilot, n=4 HF; NLRP3/cytokine reduction, no placebo comparator; NOT an RCT for inflammatory endpoints · PubMed 32790648