NAD+, sleep, and exercise

I went to an NAD+ conference in Copenhagen recently to talk about NAD+ pharmacokinetics, IV infusion tolerability, and the comparative physiological effects of two different intravenous formulations. I did NOT expect to come home thinking about sleep (although the jet lag was real). But by the end of three days of the conference, sleep had moved to the center of my thinking about this entire field. And my lab’s own data had a lot to do with that shift.

Here’s why.

The scientists of NAD+ metabolism headed out for a run to naturally boost NAD+

NAD+ is not a static molecule and doesn’t simply decline.

This was one of the foundational insights from Day 1 in Copenhagen. NAD+ concentrations in cells are not fixed. They rise and fall simultaneously. When you give more, you use more. This is a basic principle of homeostasis, but also why tracking NAD+ concentrations are so challenging, especially in blood (which has been a highly controversial headline lately).

NAD+ also oscillates across a 24-hour cycle, rising and falling in coordination with the molecular clock that governs circadian biology. NAD+ fluctuations help regulate energy metabolism, DNA repair timing, and the activity of clock-controlled genes. The enzymes that consume NAD+, particularly the sirtuins, are themselves part of the circadian feedback loop.

In practical terms: when your NAD+ rhythm is disrupted, your circadian clock is disrupted. And when your circadian clock is disrupted, your NAD+ metabolism is further disrupted. It’s a feedback loop.

Sleep is not my field directly, but these data raise a genuine question about the direction of causality in aging. And, as it turns out, is also pretty tightly connected with exercise. We know that NAD+ levels decline with age. We also know that sleep quality degrades with age: less deep sleep, more fragmented sleep, disrupted circadian rhythms, more sedentary behavior. If these phenomena are mechanistically linked through the same feedback loop, that has profound implications for when we should be thinking about NAD+ support, and why.

What my lab’s second NAD+ study actually found.

Although it’s not published yet, I can now tell you about some of the findings from our second manuscript in progress from Restore Labs. This study takes the same cohort from our published IV NAD+ vs. IV NR tolerability paper and adds a new layer: 30 days of continuous Oura ring biometric tracking following four consecutive days of infusions, plus validated patient-reported quality-of-life surveys (PROMIS) at baseline and 30 days post-treatment.

The outcomes we were tracking included sleep architecture (total sleep time, deep sleep, REM, light sleep, sleep onset latency), autonomic function (resting heart rate, heart rate variability), and self-reported measures of cognition, fatigue, pain, and global health.

Here is what stood out:

• The IV NR group showed more total sleep over the 30-day monitoring period, and the increase was particularly pronounced in light sleep (the N1/N2 sleep stages, which are associated with the transition into and between sleep cycles).

• The IV NAD+ group showed a non-significant (but notable) shorter sleep latency, which means they fell asleep faster. But, they didn’t sleep as long as the NR group.

Before I get into the “why,” a note on light sleep, because I know it sometimes gets dismissed as the “unimportant” sleep stage. Light sleep is not nothing. It is the stage you enter many times a night where your brain processes and consolidates newly learned information, where your body regulates heart rate and breathing in preparation for deeper stages, and where a significant portion of total sleep time is spent even in healthy sleepers. More light sleep, particularly when total sleep time is also increasing, can be a sign of improved sleep architecture overall. Your body is getting into the rhythm of cycling through sleep stages more effectively.

The adenosine hypothesis: why IV NAD+ and IV NR may have genuinely different sleep effects.

To understand why the two infusion types might affect sleep differently, you need to know a little about adenosine.

Adenosine is the molecule that accumulates in your brain during wakefulness and drives sleep pressure: the increasing biological urge to sleep the longer you’ve been awake. It builds up throughout the day, and when it binds to adenosine receptors in the brain, it promotes drowsiness and eventually sleep. Caffeine works by blocking adenosine receptors, which is why it temporarily suppresses that sleep drive. (And why it doesn’t actually make you less tired. It just delays when you feel it.)

Here is the mechanistic hypothesis that I’m working on, and that I presented in Copenhagen: when NAD+ is administered intravenously, it exists transiently in the extracellular space (outside the cell, in the bloodstream) where it is rapidly broken down by enzymes including CD38 and CD73. This is the same degradation process that causes the side effects of IV NAD+ infusions I described in Post 2. One of the byproducts of that extracellular breakdown is adenosine.

So the hypothesis is this: IV NAD+ infusion, through its extracellular degradation pathway, may generate a pulse of adenosine that exerts a sleep-promoting effect, one that is independent of, or additive to, whatever intracellular NAD+ eventually gets made. IV NR, by contrast, enters cells directly through nucleoside transporters and bypasses that extracellular breakdown step almost entirely. It generates very little extracellular adenosine. If this hypothesis is correct, it means IV NAD+ and IV NR don’t just differ in their cellular uptake efficiency. They may have meaningfully distinct effects on sleep biology, with IV NAD+ producing a unique, adenosine-mediated sleep signal that IV NR does not share.

The fact that our IV NR group showed more sleep and more light sleep specifically is interesting in this context, but the direction it supports is perhaps counterintuitive. More sleep in the NR group over 30 days may reflect sustained improvements in NAD+ dependent circadian regulation and sleep architecture, while any sleep-promoting (i.e. latency) adenosine effect from IV NAD+ may be more acute and short-lived. Disentangling those two mechanisms is exactly the kind of work that larger, prospective, controlled trials need to do. We’re not there yet. But the hypothesis is mechanistically believable, consistent with our data, and clinically meaningful if it holds up.

Note: I want to be clear that these are pilot-scale findings from a VERY small study, and they need replication. But they are directionally interesting enough that I am actively designing next steps around them.

The early-life sleep connection: a warning signal.

One of the sleep related threads that ran through the Copenhagen discussions involved early-life sleep dysregulation and its downstream consequences for neurodegeneration. The preclinical data is ahead of the human data here, but the signal in humans is being repeated enough to be interesting.

The basic finding: disrupted sleep in earlier life (not just in the elderly, but in midlife and potentially younger) appears to accelerate the kind of NAD+ depletion and circadian dysfunction that is associated with increased neurodegeneration risk later on. Sleep disruption isn’t just making you tired. It may be compromising the biological maintenance systems that protect your brain over decades.

If that framing is correct, it suggests that the window for thinking about NAD+ support may be much earlier than the current trial landscape implies. Right now, most NAD+ human studies enroll people who are already older or already symptomatic. But if the sleep-NAD+-circadian feedback loop is accumulating damage over decades before symptoms appear, waiting until midlife or later to intervene may mean we’re already significantly behind.

The exercise connection.

I promised in Post 1 that I’d come back to exercise, and here is where it fits, because the exercise-sleep-NAD+ interaction is interesting.

It is well established that exercise increases NAD+ levels. Muscle contraction drives up demand for NAD+ in the mitochondria, activates key NAD+ consuming enzymes like SIRT1 and PARP, and stimulates the salvage pathway that recycles NAD+ precursors. One of the more striking data points shared at the conference: exercise in later life can reshape metabolic profiles so substantially that older, consistently active individuals show NAD+ metabolism patterns more similar to much younger people. The biology of exercise is, among many other things, a biology of NAD+ maintenance.

Exercise is also one of the most powerful tools we have for improving sleep quality and circadian rhythm regulation. It promotes deeper sleep, improves sleep efficiency, and helps anchor the circadian clock. When you put those two facts together, it raises the possibility that exercise, NAD+ metabolism, and sleep quality are all running through the same biological infrastructure, reinforcing and amplifying each other.

The question the conference grappled with (and did not fully resolve) is whether supplementation on top of exercise adds anything meaningful, or whether they are redundant. The emerging answer seems to be that they are not redundant, and may be genuinely synergistic, particularly in populations where exercise capacity is limited or declining. The mechanistic logic: exercise increases NAD+ demand and consumption, which is part of how it confers benefit. Supplementation expands the available NAD+ pool before and after that physiological stress, potentially allowing for more robust activation of those same pathways and more complete recovery.

In practice, this may mean that the people most likely to benefit from NAD+ supplementation are not sedentary people looking for a shortcut, but active people (particularly older active adults) who are already generating significant NAD+ demand through exercise and whose capacity to synthesize NAD+ endogenously is declining with age. That framing fundamentally changes the target population, and it has significant implications for how future trials should be designed.

On that front, our team is actively analyzing data examining NAD+ and NR infusions and oral supplementation in athletic populations, looking specifically at performance and recovery outcomes. Stay tuned for more there…

What does this mean for you right now?

Sleep is probably the most underappreciated variable in the entire NAD+ story. If you are not sleeping well (chronically, not occasionally) no supplement is going to compensate for what that’s doing to your NAD+ metabolism, your circadian biology, and your long-term brain health. Fixing your sleep, through behavioral means and medical evaluation if needed, is foundational in a way that supplementation is not.

If you are already a solid sleeper and you’re considering NAD+ infusions with sleep quality as one of your goals, our data hints that oral supplementation, IV NR and IV NAD+ may have different effects on sleep architecture over time, with IV NR associated with more total sleep and particularly more light sleep over 30 days. We can’t prove the mechanism yet, but it’s a hypothesis I’m actively pursuing.

And if you’re younger (in your 30s or 40s) and you have chronic sleep problems, I would encourage you to take them more seriously than perhaps you have been. Not because you need to start NAD+ supplementation, but because the research is increasingly pointing toward the sleep x NAD+ circadian axis as a system where earlier disruption creates longer-term consequences, and where earlier attention may matter more than previously recognized.

Post 4 in this series will cover the biggest unresolved problem in the entire field, the measurement question, along with a surprising new gut microbiome finding and where this science needs to go next.