Mitochondrial Health: The Complete Guide
What mitochondria do, what the human trials show for three supplement groups, and why the low-cost fixes still come first.
Published Oct 19, 2026 · Updated Oct 22, 2026 · 18 min read
What You Need to Know
- Mitochondrial dysfunction is one of nine hallmarks of aging, according to the 2013 Cell paper that defined the list (Lopez-Otin et al.).
- In skin samples from 49 people aged newborn to 77, NAD+ fell with age, with negative correlations of 0.71 in men and 0.54 in women (Massudi 2012).
- Blood NAD+ varies widely between people: in an 80-adult NMN trial, the coefficient of variation for the NAD+ change ran from 29% to 113% (Kuerec 2024).
- Walking distance beat placebo at 300, 600 and 900 mg of NMN a day in an 80-adult, 60-day trial (Yi 2022).
- Blood NAD+ rose 22%, 51% and 142% within two weeks of 100, 300 and 1,000 mg of NR a day (Conze 2019).
- In three human trials that measured mitochondrial function directly, that function did not change (Remie 2020, Elhassan 2019, Whitfield 2025).
- Only 12% of people had detectable urolithin A at baseline, and about 40% produced it after drinking pomegranate juice (Singh 2021).
What mitochondria do, and why their output matters
Mitochondria turn the energy in food into ATP, the fuel cells spend on nearly all of their work. NAD+ is one of the carriers that moves electrons into the chain doing that job, so NAD+ levels and mitochondrial output are tied together. Declining mitochondrial function is one of the core features researchers use to describe aging. A 2013 Cell paper that set out nine hallmarks of aging listed mitochondrial dysfunction among them (Lopez-Otin 2013).
When mitochondria lag, a cell has less fuel for maintenance and repair. That gap is what researchers try to measure, and in some trials, what they try to close.
Here is the fact that shapes the rest of this guide. Most of the human data on these ingredients measures blood levels, not how people move, run, or feel. Only a handful of trials measured walking, running, or strength, and those results were mixed.
Go deeper: the NAD+ and cellular energy guide follows the NAD+ pathway from precursor to energy carrier in more detail than this page can.
Who has less NAD+ and mitochondrial activity, and does it change?
Age is the factor these studies track most closely. In skin samples from 49 people aged newborn to 77, NAD+ fell with age. The correlation was negative, and it was stronger in men (0.71) than in women (0.54) (Massudi 2012). A separate study found that NAMPT, an enzyme that helps recycle NAD+, was lower with older age, and that aerobic fitness was its strongest predictor (de Guia 2019).
So who is short? No prevalence figure exists that this guide can cite for low mitochondrial function, because mitochondrial function has no test with a reference range. Check any percentage quoted for it before you rely on it.
Training moves the numbers in the right direction, though the studies are small. In 12 weeks of aerobic or resistance training, NAMPT rose between 12% and 30%, depending on age and training type (de Guia 2019). In 16 middle-aged adults who were untrained at the start, 10 weeks of full-body resistance training twice a week raised muscle NAD+ by 127%. Citrate synthase, a common marker of mitochondrial content, rose 183% (Lamb 2020). That study had no control group, so read it as a before-and-after picture.
Go deeper: the aging puzzle post on NAD+ decline explains how NAD+ levels fall over time, in plainer language than the studies above.
How would you know whether it is working?
There is no clinical test for mitochondrial function that a clinic can run and compare against a target range. The honest answer is that this measurement does not exist yet.
Blood NAD+ is the test people reach for first, and it has a problem. In a post hoc analysis of 80 adults aged 40 to 65 who took NMN for 60 days, the change in blood NAD+ varied so much within each dose group that the coefficient of variation ran from 29% to 113% (Kuerec 2024). The same analysis linked bigger NAD+ rises with bigger gains in six-minute walking distance and in a quality-of-life survey. That is an association inside a post hoc analysis, not proof that one caused the other.
A blood NAD+ number tells you the precursor was absorbed and converted. It does not tell you how well your mitochondria work. Conze 2019 and Christen 2026 show the first part. Nothing in those trials shows the second part from a blood number alone.
What to watch instead are the functional measures the trials used. The six-minute walk test is the measure that picked up a signal in the NMN trial (Yi 2022). Aerobic fitness was the strongest predictor of NAMPT in the training study above, so VO2 is a sensible thing to track alongside training. Grip strength and gait speed were outcomes in NMN trials, and those results were mixed. In the urolithin A trial of adults aged 65 to 90, walking distance and a hand-muscle energy test were the primary outcomes, and neither reached significance (Liu 2022). Pick the measure that matches your goal, and measure it the same way each time.
How often to retest? No schedule has evidence behind it. Retest what you can act on, such as training load, sleep, or a marker your clinician already follows.
Go deeper: the urolithin A clinical data charts lay out the endurance and biomarker results from those trials, including the measures this section describes.
What can food and training do, and where do they stop?
Food matters for urolithin A in one specific way. Gut bacteria make urolithin A from ellagitannins in pomegranate, and not everyone's gut does it. In one study, only 12% of people had detectable urolithin A at the start, and about 40% produced it after drinking pomegranate juice. The people who made it also had more diverse gut microbiota (Singh 2021). Juice is therefore a route with a gut-bacteria filter, not a reliable dose.
Calorie restriction has its own human data. In a six-month trial of 36 overweight adults, a 25% calorie cut raised muscle mitochondrial DNA by 35%, and a calorie cut combined with exercise raised it by 21%. Citrate synthase and other enzyme activities did not change, so the extra DNA did not show up as more enzyme activity. DNA damage fell (Civitarese 2007).
Sleep is harder to put a number on. In a small crossover study of postmenopausal women, four nights of restricted sleep, cutting sleep by 40%, changed the activity of several muscle gene pathways, including energy pathways. The shifts were modest, and gene-level changes do not show what changes in how muscle works (Suryadevara 2026). The sleep and stress guide covers the wider sleep evidence, including why a behavioral program has more support than a pill for most sleep problems.
The limits matter. None of the human studies in this section ran longer than six months, and all of them measured markers rather than long-term health outcomes. That is useful information. It is not proof that the changes last, or that they change how people feel, perform, or age.
Go deeper: the urolithin A complete guide explains the gut-bacteria step in more depth, including what it means for food sources.
When does supplementing make sense, and what do the trials show for each ingredient?
Here is the decision rule the evidence supports. If you have a measurable goal, such as walking distance, a training benchmark, or a marker your clinician already follows, pick the ingredient whose trials measured that goal, and accept that those trials were small. If your goal is general energy, the trials give you nothing to measure against.
NR and NMN: the blood signal is solid, the function signal is not
NR and NMN raise blood NAD+. Conze and colleagues gave 100, 300 and 1,000 mg of NR a day and saw whole-blood NAD+ rise 22%, 51% and 142% within two weeks, with the rise holding through the study (Conze 2019). In a 14-day open-label, placebo-controlled trial of 65 healthy adults, NR and NMN raised circulating NAD+ about equally, and nicotinamide did not produce the same rise. The full text reports 1,000 mg a day of each of NR and NMN (Christen 2026). Its authors are from Nestlé Research, so the trial was company-run.
A 14-day trial of 32 adults aged 55 to 80 found dose-related rises in blood NAD+, with similar side-effect rates across groups (Pencina 2023, 14-day).
Function outcomes are where the trials disagree. Yoshino and colleagues reported better insulin-stimulated glucose disposal after NMN in postmenopausal women with prediabetes (Yoshino 2021). Two NR trials in men with obesity found no change in insulin sensitivity or beta-cell function (Dollerup 2018; Dollerup 2019). A 2024 meta-analysis of eight randomized trials, with 342 adults, found no significant effect on fasting glucose, insulin, HbA1c, or HOMA-IR (Chen 2024).
Walking and muscle results are mixed too. In 80 adults given NMN for 60 days, the six-minute walk beat placebo at all three doses, 300, 600 and 900 mg (Yi 2022). A placebo-controlled trial of 250 mg of NMN a day in men for 6 or 12 weeks raised blood NAD+ and related metabolites, with nominal gains in gait speed and grip strength that the authors said need validation (Igarashi 2022). A 2025 review of NMN and NR trials found no significant effect on muscle index, grip strength, gait speed, or five-times sit-to-stand in adults over 60. Its authors concluded that the evidence does not support either precursor for muscle mass or function in that group (Prokopidis 2025). Taken together, the picture is a few positive signals, several null results, and no consistent gain in older adults.
| Outcome | Human evidence | Grade |
|---|---|---|
| Blood NAD+ rises after NR or NMN | Five randomized trials point the same way: Conze 2019, Yi 2022, Igarashi 2022, Pencina 2023, 14-day, Christen 2026 | Strong |
| Insulin sensitivity or fasting glucose with NMN or NR | Yoshino 2021 positive; Dollerup 2018, Dollerup 2019 and Chen 2024 null | Insufficient (conflicting) |
| Muscle mass or function in adults over 60 | Prokopidis 2025 found no effect; Igarashi 2022 nominal gains only | Insufficient (conflicting) |
| Walking distance with NMN | One trial positive: Yi 2022 | Preliminary |
| Mitochondrial function itself with NR | Remie 2020 and Elhassan 2019 found no change | Insufficient (no gain shown) |
| Urolithin A and muscle endurance or blood inflammation markers | Liu 2022 and Singh 2022 | Preliminary |
| Urolithin A and walking, VO2, or race performance | Liu 2022, Singh 2022 and Whitfield 2025: primary outcomes not significant | Insufficient |
| Blood CoQ10 after supplementing | Cooke 2008: significant rise | Preliminary |
| Endurance after CoQ10 | Cooke 2008: trend only (p = 0.06) | Insufficient |
Urolithin A: endurance and blood markers look better than performance
Urolithin A has only a handful of human trials, and their results are mixed. In a four-month trial of 66 adults aged 65 to 90, 1,000 mg a day improved muscle endurance measures at two months. By four months, it had lowered C-reactive protein (a blood marker of inflammation), acylcarnitines, and ceramides. The two primary outcomes, a six-minute walk and a hand-muscle energy test, did not differ significantly from placebo. Walking distance rose about 61 meters in the urolithin A group and about 43 meters in the placebo group (Liu 2022).
A four-month study in middle-aged adults, at two doses, found that muscle mitophagy proteins rose and CRP fell, while gains in VO2 and six-minute walk distance were not statistically significant (Singh 2022). Among 42 male distance runners on a four-week altitude training camp, 1,000 mg a day did not improve 3,000 m time-trial performance. Perceived effort and a muscle-damage marker were lower, though (Whitfield 2025). A first-in-human study in sedentary older adults found 500 and 1,000 mg for four weeks safe, with changes in blood acylcarnitines and muscle mitochondrial gene activity (Andreux 2019). For the dose question, the dose-response article charts where extra dose stops adding much.
CoQ10: one small trial
In the trials we checked for this guide, CoQ10 has the thinnest human evidence of the three. In one 14-day trial, 41 adults (22 trained and 19 untrained) took 100 mg twice a day or a placebo. Plasma CoQ10 rose significantly. Time to exhaustion trended upward, but not significantly (p = 0.06) (Cooke 2008). The significant result is a blood level, not a performance result.
Go deeper: the NMN monograph, the urolithin A monograph and the CoQ10 monograph each cover their ingredient in more depth than this section can. For a side-by-side look at urolithin A and NMN by goal, see the comparison.
Which groups need extra care?
Older adults. The urolithin A trial in people aged 65 to 90 did not meet its primary outcomes, and the 2025 review found no muscle benefit from NMN or NR in adults over 60. If you are in that group, the current trials do not justify expecting a muscle or walking gain.
Athletes. Cooke 2008 tested CoQ10 in trained and untrained adults, and Whitfield 2025 tested urolithin A in distance runners (Cooke 2008; Whitfield 2025). Neither showed a statistically significant performance gain.
People taking prescription medicines. Ask your clinician before adding any of these, and bring the exact product and dose you plan to use.
Pregnancy and breastfeeding. None of the trials in this guide enrolled pregnant or breastfeeding women, so there is no human evidence here to weigh. Ask your clinician before using any of these.
A diagnosed mitochondrial disorder. Do not start any of these in place of care for a diagnosed mitochondrial disorder. The trials in this guide enrolled generally healthy adults or people with common risk factors, so they say nothing about that group.
Go deeper: the CoQ10 monograph covers CoQ10 dosing and safety in more depth than this section can.
What goes wrong, and what the evidence cannot answer?
Overreading a blood number. A rise in blood NAD+ is one of the more consistent findings here, and it is still a lab result, not a health outcome. Treat it as proof of absorption, not proof of benefit.
Paying for a dose the trials did not reward. The walking signal in the NMN trial appeared at 300 mg, the lowest dose tested. A 28-day trial of 2,000 mg a day of an NMN product found no significant difference from placebo in muscle strength, fatigability, aerobic capacity, or stair-climbing power (Pencina 2023, 28-day). The same trial reported larger decreases in weight, diastolic blood pressure, and LDL cholesterol than placebo, and its authors called for larger trials to test those cardiometabolic outcomes.
Buying a product that does not match its label. A 2024 study tested NMN and urolithin A products against their labels. Measured content ranged from 28.6% above the label claim to 100% below it (Sandalova 2024). A price does not tell you what is in the bottle. A lot-level certificate you can check against the label is a better signal.
Treating animal results as human results. In mice given NMN for 12 months, the authors reported that it mitigated age-associated physiological decline (Mills 2016). That is an animal finding. It is the reason the human trials matter, and those trials are smaller and shorter.
Assuming the effect runs one way. Nicotinamide, a form of vitamin B3 in the same pathway as NAD+, inhibited the SIRT1 enzyme at physiological concentrations in a 2002 test-tube study (Bitterman 2002). That is an in vitro result, so it does not show what happens in a person. It does show that the pathway has more than one switch.
Interactions and conditions. Ask your clinician before you start, especially if you take prescription medicines. Get care promptly if a new symptom appears after you start.
Go deeper: the supplement safety and quality testing guide explains how to read a label, what third-party testing covers, and which red flags to check first.
Which Mt. Angel Vitamins products fit this topic, and where do they stop?
We sell three products in this area. None of them has its own human trial in this guide, so each one is judged by the ingredient trials above. Both buying guides compare products from several brands, including ours.
NMN + NR with resveratrol and TMG. $59.99 for 30 servings, or $2.00 a day. Each serving is two capsules with 500 mg of beta-NMN, 200 mg of nicotinamide riboside, 250 mg of resveratrol, 50 mg of TMG, 50 mg of BioPerine, and 10 mg of zinc. The product page says it is made to support healthy NAD+ levels and cellular energy.† The NMN dose sits inside the 250 to 2,000 mg range in the 2024 meta-analysis, and the NR dose sits inside the 100 to 1,000 mg range that Conze and colleagues tested. The label does not say which form of NR it uses. The resveratrol, TMG and zinc have their own evidence, which this guide does not cover. Shop NMN + NR
Urolithin A 500 mg. $89.99 for 60 servings, or $1.50 a day. One capsule is 500 mg of urolithin A. The product page describes mitochondrial renewal and cellular energy.† The dose matches one of the two doses in the first-in-human study (Andreux 2019), and it is half the 1,000 mg a day used in the Liu 2022 and Whitfield 2025 trials. A capsule skips the gut-bacteria step that food depends on. Its lot certificate is on the lab-testing page. Shop Urolithin A 500 mg
CoQ10 100 mg with alpha lipoic acid. $25.99 for 60 servings, or $0.43 a day. One capsule is 100 mg of CoQ10 and 5 mg of alpha lipoic acid, with 100 mg of soy lecithin. That is half the daily CoQ10 in the 2008 trial, which gave 100 mg twice a day. The capsule contains soy, so check the allergen statement before you buy. The lot certificate for this product is not on the lab-testing page yet. Shop CoQ10 100 mg with alpha lipoic acid
Every batch is made in our own cGMP facility in Mt. Angel, Oregon, and tested in our in-house lab for identity, potency, and purity. Lot certificates for the urolithin A and NMN + NR products are published on the lab-testing page.
None of these products replaces the habits in the sections above, and no trial in this guide tested a product against them.
†These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure or prevent any disease.
Go deeper: the NMN and NR buying guide and the urolithin A buying guide compare these products with other brands on dose, lot testing, form, and cost per day.
Bottom line
Start with regular training and enough sleep. If you add a supplement, choose the ingredient whose trials measured the outcome you care about, check its lot certificate, and expect a blood number to move before anything you can feel. None of the three ingredients has shown a clear, lasting gain in mitochondrial function in people.
Go deeper: the sleep and stress guide covers the sleep side of the first habit.
Related reading
- Epigenetic clocks and the NAD+ precursor evidence
- Resveratrol and the sirtuin claim
- NAD+ precursor trials: what the nulls show
- How creatine got from meat extract to the sports shelf
- The NR monograph: what the trials show for blood NAD+ and function
- NMN and NR side by side, including the 2026 head-to-head trial
- NMN and NR doses compared with the trials
- Why food and supplement urolithin A differ by person
- Copper as a cofactor in energy production, in the copper origin story
- how FMN and FAD, the riboflavin coenzymes, support energy production
- how Rose's amino acid requirements shaped the metabolism story
- how the liver factor became a mitochondrial cofactor
- how vitamin E went from a fertility factor to an antioxidant
- Milk thistle: from Pliny's sillybum to the mixed liver trials
- Selenium from a 1817 mix-up to the glutathione peroxidase story
- the carnitine fat-transport story
- the ginseng research history, from legend to American export
- The history of coenzyme Q10, from beef heart mitochondria to supplements
- The history of cordyceps and its exercise trials
- the research record behind lutein, from marigold petals to eye trials
- the research record behind taurine in bile salts, infant formula and energy drinks
- The thiamin story: how a rice diet fix came before the vitamin
- Where iodine fits in thyroid hormones, from seaweed ash to iodized salt
- Who found biotin, and why raw egg white caused a vitamin mystery
- Who found vitamin B5, and how coenzyme A links it to energy
- Who found vitamin B6, and how rat skin studies led to P5P
- history of maca
Common Myths
"NAD+ goes up, so my cells make more energy."
Blood NAD+ does go up after NR or NMN, and that is well shown (Conze 2019). Whether cells make more energy is a separate question. Two NR trials that measured mitochondrial function found no change (Remie 2020; Elhassan 2019). The rise is real. The energy claim is not supported yet.
"A blood NAD+ test tells me whether the supplement works."
Blood NAD+ varies widely from person to person. In an 80-adult NMN trial, the coefficient of variation for the change ranged from 29% to 113% within dose groups (Kuerec 2024). A rise can show that a precursor was absorbed. It cannot show that you are getting the benefit the trials were measuring.
"Pomegranate juice gives you the same urolithin A as the capsule."
Only 12% of people had detectable urolithin A at the start of one study, and about 40% produced it after drinking pomegranate juice (Singh 2021). In that study, a 500 mg dose gave more than six times the blood exposure of juice. Juice depends on your gut bacteria. A capsule supplies the compound directly.
"CoQ10 is an energy pill."
In the one CoQ10 trial we checked, 100 mg twice a day for 14 days raised plasma CoQ10 significantly. The endurance gain was a trend that did not reach significance (p = 0.06) (Cooke 2008). That is a blood level and a non-significant trend, not a demonstrated energy effect.
"A supplement can stand in for training."
No trial in this guide compared a supplement with training, so that question is still open. What the training studies do show is that resistance training raised muscle NAD+ by 127% in 16 middle-aged adults, and that aerobic and resistance training raised NAMPT by 12% to 30% (Lamb 2020; de Guia 2019). Those are small before-and-after studies, so they are not proof of anything on their own.
"The more I take, or the more I pay, the more it works."
A higher dose was not needed to see the walking signal in the 80-adult NMN trial, which appeared at 300 mg, the lowest dose tested (Yi 2022). At 2,000 mg a day for 28 days, an NMN product did not differ significantly from placebo on strength, fatigability, aerobic capacity, or stair-climbing power (Pencina 2023, 28-day). Price does not tell you the content either. A 2024 study found that measured NMN and urolithin A content ranged from 28.6% above the label claim to 100% below it (Sandalova 2024).
Frequently Asked Questions
What do mitochondria do in a cell, and what changes in how they work as people age?
Mitochondria turn the energy in food into ATP, the fuel cells spend on most of their work, and NAD+ carries electrons into that chain. Mitochondrial dysfunction is one of nine hallmarks of aging in a 2013 Cell paper ([Lopez-Otin 2013](https://pubmed.ncbi.nlm.nih.gov/23746838/)). NAD+ also falls with age in human tissue. In skin samples from 49 people aged newborn to 77, the correlation was stronger in men ([Massudi 2012](https://pubmed.ncbi.nlm.nih.gov/22848760/)).
What are the main ways researchers try to support mitochondrial function, and what is the evidence for each?
Three ingredient groups have human trials: NAD+ precursors such as NR and NMN, urolithin A, and CoQ10. Training and calorie restriction have human data on mitochondrial markers too. NR and NMN reliably raise blood NAD+, but no trial we checked showed a clear gain in mitochondrial function. Urolithin A has preliminary signals on endurance and blood inflammation markers. CoQ10 rests on one small trial. The evidence table in this guide grades each outcome.
What do human trials show for NAD+ precursors, urolithin A, and CoQ10, and how strong is each finding?
NAD+ precursors: blood NAD+ rises in five randomized trials, a strong finding. Walking and muscle results are mixed, and insulin results conflict. Urolithin A: endurance and blood inflammation markers improved in small trials, a preliminary finding, while its primary performance outcomes did not reach significance. CoQ10: one 14-day trial of 41 adults raised plasma CoQ10, a preliminary finding, and its endurance trend was not significant.
Which findings are early, animal-only, or based on a single small trial?
Several are. The NMN result on age-related decline ([Mills 2016](https://pubmed.ncbi.nlm.nih.gov/28068222/)) comes from mice. The CoQ10 endurance result is a non-significant trend from one 14-day trial of 41 adults ([Cooke 2008](https://pubmed.ncbi.nlm.nih.gov/18318910/)). The sleep gene-activity result is one small crossover study ([Suryadevara 2026](https://pubmed.ncbi.nlm.nih.gov/41902670/)). The nicotinamide and SIRT1 finding is a test-tube result ([Bitterman 2002](https://pubmed.ncbi.nlm.nih.gov/12297502/)). The urolithin A endurance signals come from a few small trials.
What should a healthy adult in their 40s, 50s, or 60s realistically expect from these supplements?
Expect blood NAD+ to rise within about two weeks if you take NR or NMN ([Conze 2019](https://pubmed.ncbi.nlm.nih.gov/31278280/)). Do not expect a demonstrated gain in mitochondrial function or in performance from the trials so far. One NMN trial of 80 adults found a walking-distance signal at 60 days ([Yi 2022](https://pubmed.ncbi.nlm.nih.gov/36482258/)). A 2025 review found no muscle benefit in adults over 60 ([Prokopidis 2025](https://pubmed.ncbi.nlm.nih.gov/40275690/)). Talk with your clinician about your goals before you start anything.
Which non-supplement habits have better evidence for mitochondrial function than any product (exercise, sleep, diet)?
No habit has been shown to raise mitochondrial function in a large, long trial, so the honest answer is that exercise fits the evidence best, not that it is proven. Resistance and aerobic training raised muscle NAD+ or NAMPT in small studies ([Lamb 2020](https://pubmed.ncbi.nlm.nih.gov/32369778/); [de Guia 2019](https://pubmed.ncbi.nlm.nih.gov/31207144/)). A 25% calorie cut raised muscle mitochondrial DNA over six months ([Civitarese 2007](https://pubmed.ncbi.nlm.nih.gov/17341128/)). The sleep data is one small crossover study ([Suryadevara 2026](https://pubmed.ncbi.nlm.nih.gov/41902670/)). Start with training and sleep.
How should a buyer choose among these products, and what should they ask the brand?
Start with the ingredient whose trial measured the outcome you care about. Then compare the dose per serving with that trial, the price per day, and whether the label names the form, such as the salt used for NR. Ask for the lot certificate and check that it matches the label, since a 2024 study found large gaps between labels and measured content ([Sandalova 2024](https://pubmed.ncbi.nlm.nih.gov/38935229/)). Our two buying guides compare brands on these points.
Is NMN or NR better for raising NAD+?
On blood NAD+, they performed about the same. In a 14-day trial of 65 healthy adults, NR and NMN raised circulating NAD+ about equally ([Christen 2026](https://pubmed.ncbi.nlm.nih.gov/41540253/)). That trial measured blood NAD+ and related metabolites, so its result does not show which one does more for how you feel or move. Both raise blood NAD+, so that result alone does not pick a winner. The [NMN monograph](/ingredients/nmn-nicotinamide-mononucleotide) covers NMN in more depth.