NAD+ is a coenzyme every cell uses to turn food into energy. Parkinson's research has taken an interest in it because the disease involves failing mitochondria, the parts of the cell that depend on NAD+.
Three different molecules turn up under the NAD name, and the evidence for each is separate. Nicotinamide riboside (NR) is an oral precursor the body converts to NAD+. NADH is the coenzyme’s reduced form. NAD+ itself is what clinics inject or infuse. The evidence for NAD+ injections in general is covered in NAD injections.
NADPARK: can a pill raise NAD+ in the brain?
The NADPARK trial enrolled 30 people with newly diagnosed Parkinson’s who had not yet started treatment [1]. They took 1,000 mg of NR a day or a placebo for 30 days.
NR was well tolerated. Brain scans showed a significant rise in brain NAD+, but the size of the rise varied between people. Related compounds also rose in the fluid around the spinal cord.
People whose brain NAD+ went up showed changed brain metabolism on PET scans, and this was linked to mild clinical improvement [1]. NR also lowered some inflammatory markers in blood and spinal fluid.
This was a phase I trial of 30 people over one month, so it can show safety and a biological effect, not treatment benefit. The authors call NR a potential neuroprotective therapy and ask for larger trials.
NR-SAFE: a higher dose
The same group then tested 3,000 mg of NR a day, as 1,500 mg twice daily, in 20 people with Parkinson’s [2]. Ten took NR and ten took placebo, for four weeks. Safety was the primary outcome.
No moderate or severe adverse events occurred, and mild events were no more common than on placebo. Whole-blood NAD+ rose about 3.7-fold, with a range of 1.8- to 5.8-fold. Blood homocysteine rose slightly at first.
Parkinson’s rating scores improved in the NR group, but with a catch the authors point out themselves [2]. By day 28, the NR group was being tested on average 49 minutes sooner after its last levodopa dose. The placebo group was tested about 28 minutes later.
Levodopa strongly affects motor scores, so a shorter gap can make symptoms look better. In a subset of 8 per group with closer timing, the NR group’s total score still fell from 46.62 to 35.75 (p = 0.028). The authors still could not rule out levodopa, and call the finding preliminary.
| Study | Molecule and route | People | What it showed |
|---|---|---|---|
| NADPARK (2022) | NR by mouth, 1,000 mg/day, 30 days | 30, newly diagnosed | Brain NAD+ rose; phase I, not built to show benefit |
| NR-SAFE (2023) | NR by mouth, 3,000 mg/day, 4 weeks | 20 | Safe; blood NAD+ up about 3.7-fold; symptom signal confounded |
| NOPARK (completed 2025) | NR by mouth, 1,000 mg/day, 52 weeks | 410 enrolled | No results posted on the registry |
| Dizdar (1994) | NADH by vein, then muscle | 9 (5 vs 4 placebo) | No significant difference from saline |
| Birkmayer (1989), Kuhn (1996) | NADH by vein | 34 and 15, no control group | Improvement reported, with no placebo to compare |
NOPARK: the trial that could answer it
NOPARK is the phase 3 trial meant to test whether NR changes the course of Parkinson’s [3]. It enrolled 410 people diagnosed within two years, who took 1,000 mg of NR a day or placebo for 52 weeks.
Its primary outcome is total Parkinson’s rating score. The registry lists the trial as completed on June 17, 2025. As of September 2026, the registry record shows no posted results, and no results paper appears on PubMed.
Two authors of the NADPARK and NR-SAFE papers have filed patent applications on NR for Parkinson’s, as both papers disclose [1] [2].
The older NADH infusion studies
Claims that NAD infusions treat Parkinson’s go back to the late 1980s. They rest on NADH, not NAD+, and on studies without a comparison group.
In a 1989 open-label study, 34 patients received NADH by vein [4]. The authors reported that every patient improved, 21 of them by more than 30%. No one received a placebo, so improvement from expectation or from other treatment cannot be separated out.
A 1996 open study gave 15 patients NADH infusions for 7 days alongside their usual drugs [5]. Rating scores improved (p = 0.025), and plasma levodopa levels rose. It also had no placebo group.
When a small double-blind trial was run, no clear effect showed up [6]. Five patients got 25 mg of NADH by vein for four days, then by muscle injection; four got saline. Both groups trended better after the infusions, with no significant difference.
A 1998 review weighed this record and concluded that recommending NADH for Parkinson’s was premature [7].
How NAD+ is supplied
No NAD product is FDA-approved for Parkinson’s. NAD+ injections are compounded by pharmacies. The FDA lists NAD in 503A Category 1, the bulk substances still under evaluation for compounding [9].
What this means
The case for NAD in Parkinson’s is early. An oral precursor can raise NAD+ in blood and brain, and it has been well tolerated for a month. Whether that slows or eases the disease is the question NOPARK was built to answer, and its results are not yet public.
Injected NAD+ has no Parkinson’s data at all. What NAD+ infusions do in people is covered in NAD IV therapy. Reported reactions are in NAD side effects. The difference between precursors is in NMN vs NAD.