NMN and NAD+ for Fatty Liver (NAFLD/MASH): The Evidence

Fatty liver disease — now formally called metabolic dysfunction–associated steatotic liver disease (MASLD), and its inflammatory form MASH — is one of the most common chronic conditions in the world, and one of the most frustrating, because for years there was no approved drug for it. That gap, plus a genuinely compelling set of mouse studies, has made fatty liver a favorite target for NAD+ marketing: take NMN or nicotinamide riboside (NR), the pitch goes, and you can melt the fat out of your liver.

The animal data behind that pitch is real and, frankly, impressive. But "impressive in mice" and "proven in people" are very different claims, and fatty liver is exactly the place where that distinction matters — because the human trials have now been done, and the results are more sobering than the supplement ads suggest. This page walks through what's genuinely established, what's only preclinical, and why NAD+ for fatty liver remains promising-but-unproven.

Why the liver is a logical NAD+ target

NAD+ (nicotinamide adenine dinucleotide) is a coenzyme in every cell that carries electrons through the reactions turning food into usable energy, and it also fuels DNA-repair enzymes and the sirtuins. Because it sits at the center of energy metabolism, low NAD+ availability can ripple outward into mitochondrial and metabolic dysfunction ¹, and NAD+ tends to fall with age and metabolic stress in human tissue ². The liver is a metabolically ferocious organ, and the leading models of fatty liver involve overwhelmed, dysfunctional mitochondria that can't keep up with a flood of dietary fat. So "restore the liver's NAD+ and you restore its mitochondria" is a clean, mechanistically sensible hypothesis.

The preclinical case is genuinely strong

This is the part the marketing gets right. In animal models, boosting NAD+ does remarkable things to fatty liver.

The landmark study came in 2016: in mice fed a high-fat, high-sucrose diet, nicotinamide riboside both prevented and reversed non-alcoholic fatty liver disease ³. The mechanism was elegant — the high-fat diet lowered hepatic NAD+ and crippled mitochondrial function, and restoring NAD+ triggered a SIRT1/SIRT3-dependent mitochondrial unfolded-protein response that ramped fat-burning back up ³. Subsequent work reinforced the theme, showing that engaging the NAMPT/NAD+/SIRT1 axis ameliorated NASH (the inflammatory, more dangerous form of fatty liver) in disease models ⁴. Across the preclinical literature, the signal is consistent: in rodents, NAD+ repletion reduces liver fat and improves the underlying mitochondrial biology.

If mice were the customers, NAD+ would be a blockbuster. They aren't. (The same "strong in mice, unproven in people" gap runs through the organ next door — see NMN and NAD+ for kidney health, where the protective data is almost all in acute injury in rodents.)

The human trials: a reality check

Here's where the story turns, and where honesty has to be loudest. When NAD+ boosters were actually tested for fatty liver in humans, they did not reduce liver fat.

The most directly relevant trial is a 2023 study published in Hepatology: 111 adults with NAFLD were randomized to placebo or one of two doses of an NR-plus-pterostilbene supplement for six months ⁵. The primary endpoint — hepatic fat fraction — did not change versus placebo at either dose ⁵. That's the headline result, and it's the opposite of what the mouse data predicted. The supplement did move some secondary markers: at the standard dose, liver enzymes (ALT and GGT) and a toxic ceramide lipid fell significantly versus placebo, and the supplement was safe ⁵. Those secondary signals are interesting and worth following up — but a drop in a couple of inflammation markers is not the same as shrinking the fat in the liver, which is the thing the product was supposed to do and the endpoint the trial was built around.

A second human trial points the same direction. A 2023 randomized study gave overweight and obese middle-aged and older adults an oral NMN formulation (MIB-626) for 28 days ⁶. It clearly raised circulating NAD+ and produced some modest cardiometabolic shifts — small drops in body weight, LDL cholesterol, and blood pressure — but it showed no change in liver or abdominal fat and no improvement in insulin sensitivity ⁶. Again: the biomarker moved, the liver fat didn't.

There is one frequently-cited human bright spot worth putting in proper context. A 2021 trial in Science found that NMN improved skeletal-muscle insulin sensitivity in prediabetic women ⁷ — a genuine, controlled metabolic benefit. But note what it was and wasn't: it measured muscle insulin signaling, not liver fat, and even its own authors and commenters debated how to interpret it. It's encouraging for the broader "NAD+ and metabolism" story, but it is not evidence that NMN treats fatty liver.

Why mice and humans diverge here

Why would something that reverses fatty liver in rodents fail to shrink it in people? A few honest reasons. Mouse studies use young animals with diet-induced disease, controlled genetics, and often higher relative doses for shorter, cleaner timelines — conditions that flatter an intervention. Human MASLD is heterogeneous, chronic, and tangled up with obesity, insulin resistance, diet, and genetics that no single coenzyme is likely to untangle. And the broader NAD+ literature shows this exact pattern repeatedly: a systematic review and meta-analysis of NAD+ precursors on metabolic-syndrome parameters found benefits limited and inconsistent despite reliably raised NAD+ ⁸, and even in other organ systems, raising the NAD+ biomarker has repeatedly failed to deliver the matching functional improvement ⁹. We unpack that "biomarker is not benefit" problem across the board in our pillar guide to NAD+ therapy evidence.

So should you take NMN or NR for fatty liver?

Honestly: not as a treatment, because it isn't one — yet.

• •The mechanism and preclinical data are strong ³⁴, which is why this is an active research area and not a dead end.
• •The human trials so far have not reduced liver fat ⁵⁶ — the one endpoint that defines the disease.
• •There are encouraging secondary signals (liver enzymes, lipids, a metabolic-marker improvement here and there) ⁵⁷, but secondary markers are not the disease, and a positive marker doesn't establish a clinical benefit.
• •No NAD+ booster is approved or established for NAFLD/MASH, and none should be relied on in place of the things that genuinely work.

The treatments with real evidence for fatty liver are unglamorous and effective: weight loss (5–10% of body weight meaningfully reduces liver fat), dietary change, exercise, managing diabetes and lipids, and — increasingly — physician-prescribed medications now reaching the clinic for MASH. If you're considering an NAD+ supplement, treat it as an unproven adjunct to discuss with the clinician managing your liver, not a substitute for the proven steps. Before buying anything, understand the whole picture: start with our NAD+ therapy evidence pillar, see how the precursors differ in NMN vs NAD, check the side-effect profile, and — because high-dose niacin-family compounds and liver health intersect with cancer-pathway questions people often ask about — see our honest review of NAD+ and cancer. For product and provider comparisons, see our NAD+ rankings hub.