Mitochondria and longevity

Mitochondria are the cell's energy factories. They make roughly 90 percent of your ATP, the molecule that powers basically everything you do. They do it with an electron assembly line that turns food and oxygen into usable energy. They also carry their own DNA, build new copies of themselves on demand, and recycle their own broken parts.

They started out as free-living bacteria 1.5 billion years ago, moved into our cells, and never left. They even kept their own little loop of DNA.

How they actually make energy. Picture a tiny assembly line called the electron transport chain. Food gets broken down, and the electrons it releases flow down this line through four stations. CoQ10 is the courier that shuttles those electrons from the early stations to the next one, and oxygen catches them at the end, which is the whole reason you breathe. As the electrons move, the chain pumps charged particles to one side of a membrane, building up a kind of battery (about 150 to 180 millivolts of stored charge). A final machine, ATP synthase, lets that charge flow back and spins like a turbine to crank out ATP. Burn one molecule of glucose all the way down and you net about 30 to 32 ATP. Burn a fat molecule and you get around 106. Fat is the denser fuel.

A little exhaust is unavoidable. Some of those electrons slip off the line early and form reactive oxygen species, or ROS (think of them as sparks, mostly from two leaky spots on the chain) ^[1]. One California lab (at the Buck Institute for Research on Aging) has mapped at least 11 separate spots inside mammalian mitochondria that can produce these sparks ^[2]. Here is a neat trick: the reduced form of CoQ10 also works as the membrane's main built-in antioxidant, mopping up some of that exhaust on the spot.

The mitochondrial genome is tiny and lives in a bad neighborhood. It's a small ring of 16,569 base pairs (DNA letters) carrying just 37 genes. Thirteen of them build parts of the energy-making machinery: 7 for the first station, 1 for the third, 3 for the fourth, and 2 for the ATP synthase turbine. The other 24 (22 tRNAs and 2 rRNAs) run the local protein-making machinery. No protective packaging, no spare parts. And it sits right up against the membrane where the sparks fly, so its mutation rate runs 10 to 20 times higher than the DNA in your cell's nucleus. You inherit it only from your mother. You carry hundreds to thousands of copies per cell. And there is a buffer built in: trouble usually only shows up once 60 to 90 percent of the copies are damaged.

Quality control runs nonstop. Mitochondria constantly merge together and split apart, which lets them share parts and isolate damage. When one goes bad and loses its charge, a tag-and-remove system flags it and the cell hauls it off for recycling. This targeted cleanup is called mitophagy, basically a trash-and-recycle program built just for worn-out mitochondria. A 2008 study was the first to catch the key cleanup protein, Parkin, homing in on damaged mitochondria ^[10]. Break the proteins that run all this and you get real, named diseases of the nerves, the optic nerve, and an inherited early-onset form of Parkinson's.

You can also build new ones on demand. The master switch for making fresh mitochondria is a protein called PGC-1α, first identified in 1998 from cold-exposed brown fat ^[9]. Flip it on and your cells ramp up production of new energy machinery. Two upstream sensors flip that switch: AMPK (your cell's low-fuel alarm, which fires when energy runs short) and SIRT1 (an enzyme that depends on a molecule called NAD⁺). What flips them in real life? Exercise, cold, fasting, and eating less. The best-proven triggers for new mitochondria. None of them come in a capsule.

Mitochondria slow down with age in just about every tissue anyone has looked at. The pattern is rock solid. The reason behind it is still a fight.

For decades, the leading idea was the free radical theory of aging. Denham Harman floated it in 1956 and sharpened it in 1972 to blame mitochondrial sparks (those reactive oxygen species again) for chewing up DNA, fats, and proteins until you age. It ruled the field for thirty years. In its strong form, it is now dead.

Here is what killed it. Researchers built a mouse, the so-called mutator mouse (two landmark papers, one in Nature in 2004 and one in Science in 2005), whose mitochondria can't proofread their own DNA ^{[3, 4]}. These mice rack up 3 to 5 times more mitochondrial DNA mutations than normal and age fast: gray fur, hunched spine, muscle wasting, enlarged heart, anemia. Their median lifespan is around 12 months versus about 30 for controls. Now the twist. The 2005 Science paper found no overall jump in oxidative-stress markers in these mice ^[4]. So you can wreck the mitochondrial DNA, speed up aging, and still never see the spark damage the theory predicted. A 2009 Biochimica et Biophysica Acta study went further and tweaked 18 different antioxidant genes in mice; only one (deleting a key antioxidant enzyme, Sod1) shortened lifespan ^[54]. And big Cochrane reviews of antioxidant pills in people show no survival benefit and possible harm at high doses ^[56].

What survives the wreckage. A model called mitohormesis, from Michael Ristow, flips the script. Small, brief, well-contained sparks (the kind you get from exercise, fasting, and eating less) act as a signal that tells your cells to toughen up. They switch on your own antioxidant defenses, build new mitochondria, and run more cellular cleanup. In worms, this kind of signal extended lifespan ^[5]. In people, high-dose vitamin C and E blunted the insulin-sensitivity gains from exercise ^[55]. The bad kind of ROS, the steady high flood, is a symptom of broken mitochondria and does feed disease. But it is not the thing that starts the fire. A 2013 theory paper from Douglas Wallace (J Clin Invest) reframes it: the real drivers are a buildup of mixed healthy-and-mutant DNA, falling energy output, and broken signaling, not raw spark damage ^[59].

Some numbers that show the decline. CoQ10 in the human heart drops by roughly 50 percent between ages 20 and 80 ^[22]. NAD⁺ in human skin falls steadily with age: a 2012 study of samples from newborns to age 77 found a strong negative correlation in both men and women ^[53]. You see the same shape of curve in fat, muscle, and the brain. And increasingly this looks like a using-it-up problem, not a making-it problem. An enzyme called CD38 that burns through NAD⁺ rises with age, driven by the low-grade inflammation that worn-out cells throw off. A 2016 Cell Metabolism study linked the rise to CD38, and a 2020 Nature Metabolism paper showed that senescent (aged, no-longer-dividing) cells switch on CD38 in nearby immune cells. A 2023 study reported that blood taurine drops with age in mice, monkeys, and people ^[36]. A 2025 NIH-led study found it rose or held steady instead, in all three ^[37]. So the taurine-decline story is now disputed (more on that later).

So what do you do with this? Mitochondria failing with age is real. Calling that the cause of aging goes past what the evidence supports. And that distinction matters for treatment: blanket antioxidant pills are a dead end. The moves that actually improve mitochondrial quality control (exercise, fasting, cold, NAD⁺ precursors in some cases, sleep) have far stronger evidence behind them.

Mitochondrial dysfunction is one of the twelve official hallmarks of aging ^[7], the core breakdowns that drive getting older. It touches most of the other eleven. But a hallmark is a consistent companion of aging, not a proven single cause.

In 2013, a landmark Cell paper organized modern aging research around nine "hallmarks," the core breakdowns that show up as we get older ^[6]. A 2023 update in the same journal added three more (stalled cellular cleanup, chronic inflammation, and a disrupted gut microbiome) for a total of twelve ^[7].

Mitochondrial trouble is one of those twelve. And it brushes up against most of the others. When mitochondria fail, proteins misfold, your cells throw off inflammatory signals (some of it from leaked mitochondrial DNA), the fuel-sensing systems get thrown off, and your stem cells wear out faster. It sits in the middle of the web.

But a hallmark is not a cause. The framework asks three questions: does it show up with age, does making it worse speed aging up, and does fixing it slow aging down? "Mitochondria failing with age" passes all three. "Mitochondria are the cause of aging" does not. It is one of twelve linked breakdowns, and they all lean on each other.

One repair system worth knowing about. Mitochondria have their own emergency response for when too many proteins inside them go wrong. Researchers study it hard because tweaking it can stretch lifespan in lab animals. In a tiny worm called C. elegans (the workhorse model of aging research), a sensor protein keeps tabs on how smoothly proteins are getting imported into the mitochondria. When import stalls under stress, the sensor reroutes to the cell's command center and switches on a cleanup-and-repair crew. A 2013 Nature paper showed that gently dialing down the mitochondria's own protein-making machinery fires up that repair response and stretches the worm's lifespan by 50 to 60 percent ^[8]. Boosting NAD⁺ copies a slice of this in worms and mice. Whether it works the same way in people is plausible but unproven.

Some tissues lean on mitochondria more than others. Your heart, brain, slow-twitch muscle, and liver are packed with them. So when people inherit mitochondrial diseases (with hard-to-pronounce names like MELAS, MERRF, and LHON, each tied to a specific glitch in the mitochondrial DNA), those are exactly the tissues that fail first. Here is the link to ordinary aging: the same organs that crash in mitochondrial disease are the ones that fade with age. Neurons and heart-muscle cells barely replace themselves, so they quietly stockpile damaged mitochondria and a mix of healthy and mutant DNA (called heteroplasmy) over decades.

Clearly symptomatic mitochondrial DNA disease hits roughly 1 in 5,000 people. And about 1 in 200 healthy people carries a disease-linked mitochondrial variant at a low enough level that it never causes symptoms ^[48]. So this is not a rare-disease footnote. It sits underneath a big chunk of neurology and cardiology.

No supplement comes close to what training, sleep, and how you eat do for your mitochondria. The system that builds new mitochondria (PGC-1α and its two upstream sensors, AMPK and SIRT1) answers to muscle contraction and energy stress. Capsules barely nudge it. Movement shouts at it.

Zone 2 cardio. The founding study is from 1967 ^[11]. Rats ran five days a week, building up to 2 hours a day, for 12 weeks, and roughly doubled both their muscle mitochondrial protein and the activity of their energy enzymes. That is exercise growing new mitochondria, captured for the first time. A 2014 review boiled down four decades of follow-up ^[12], and it splits cleanly: how much you train drives mitochondrial quantity (more of them, denser, more enzymes); how hard you train drives mitochondrial quality (each one breathes better). You want both. A 2018 Sports Medicine study put elite cyclists next to people with metabolic syndrome and found the athletes kept burning fat at far higher power before they tipped into sugar-burning. "Metabolic flexibility" is really a story about how many good mitochondria you have.

How to actually do zone 2: 3 to 4 sessions a week, 30 to 60 minutes each, at an effort of about 5 to 6 out of 10. The talk test nails it: you can hold a conversation but not sing. (The full how-to lives in the zone 2 and VO₂ max training guide.) You are aiming for a heart rate just below your first lactate threshold (roughly 60 to 70 percent of max heart rate in trained adults, where blood lactate sits around 2 mmol/L). The reason to favor zone 2 is simple: you can do a lot of it without wrecking yourself. Harder efforts give you similar signals per minute, but you can't repeat them nearly as often.

HIIT (high-intensity intervals). The go-to review is from 2017 ^[13]. Sprints and hard intervals raise your mitochondrial enzymes and total mitochondrial count about as much as steady endurance work, in less time. A classic protocol is 10 rounds of 1 minute hard (around 90 percent of max heart rate) with 1 minute easy, and it hits your fuel-stress sensors harder per minute than easy steady cardio. HIIT adds to zone 2. It does not replace it. Two HIIT sessions a week is a sensible ceiling for most adults.

Strength training. A 2017 Cell Metabolism study compared HIIT, weights, and both together in young and older adults ^[14]. HIIT lit up the broadest set of mitochondrial genes and, in the older group, pushed their mitochondrial protein profile back toward a younger one. Weights boosted muscle building but didn't match HIIT for the mitochondrial signals. Translation: lift weights for muscle and strength as you age; do HIIT and zone 2 for your mitochondria. Doing both is fine, just split them across different days.

Fasting and eating windows. A 2018 Cell Stem Cell study showed that a 24-hour fast flips intestinal stem cells into fat-burning mode. A 2018 Cell Metabolism trial tested a 6-hour eating window (last meal before 3 PM) for 5 weeks in 8 men with prediabetes. It improved insulin sensitivity, blood pressure, and oxidative stress, without any weight loss, even though blood sugar itself did not budge ^[15]. The TREAT trial is the cautionary tale on the other side ^[16]: 16:8 with a noon-to-8-PM window produced only about 1 percent weight loss over 12 weeks, and a worrying chunk of that was muscle. The takeaway: the early window is the version with the signal. A 2022 head-to-head trial found an early window beat a midday one for insulin sensitivity ^[58]. Late-window 16:8 on its own is underwhelming. The intermittent fasting guide walks through the windows in detail.

Eating less overall. The CALERIE 2 trial had healthy, non-obese adults cut calories by about 12 percent for 2 years ^[17]. Every heart-and-metabolism risk factor improved more than weight loss alone could explain. A 2023 Nature Aging re-analysis ran a "pace of aging" epigenetic clock (a blood test that estimates how fast you are aging) on the same people and found the calorie-restriction group aged 2 to 3 percent slower ^[18]. In projected reduction of death risk, that is in the same ballpark as quitting smoking. The catch: two other aging clocks (PhenoAge and GrimAge) did not budge, and the authors call the effect small. An earlier 6-month calorie-restriction trial in healthy overweight adults found more mitochondrial biogenesis in their muscle biopsies ^[46].

Cold. A 2012 J Clin Invest study confirmed that adults still have working brown fat, the heat-generating fat that switches on in the cold by deliberately burning energy as warmth instead of storing it as ATP. A 2013 study showed 10 days of mild cold gets that brown fat more active. But the ice-bath crowd oversells the mitochondrial angle: a 2021 Cell Reports Medicine study found that seasoned winter swimmers mostly got better at regulating their temperature, not at building more mitochondria. Cold helps at the edges. It is not a stand-in for zone 2.

Sleep. A 2018 Science Advances study showed that a single bad night changes how your body-clock genes are switched and dials down energy-production genes in human muscle ^[47]. A 2022 Frontiers in Endocrinology study in 20 young men found that 5 nights of short sleep shifted muscle gene activity toward inflammation and immune pathways ^[60]. That is gene-expression data, not a direct measure of mitochondrial function. Sleep is the cheapest mitochondrial upgrade you can give yourself.

Stop the obvious damage. Smoking jams one of the chain's final stations directly, through cyanide and carbon monoxide. Alcohol shuts down mitochondrial protein building, leaks more electrons, and throws off the NAD⁺ balance in your liver and muscle. Heavy ultra-processed food intake tracks with a disrupted gut, more inflammation, and downstream mitochondrial trouble at the population level. That last link is more inferred than nailed down at the cellular level, so hold it a bit loosely.

What that adds up to for a healthy adult. Zone 2 plus HIIT plus enough protein plus 7 to 9 hours of sleep plus the occasional fast will move every mitochondrial marker that matters (your VO₂ max, lactate threshold, mitochondrial enzyme activity, and energy-machinery content) more than any supplement stack on the shelf. Anyone selling you a mitochondrial stack as a replacement for that is selling you the weaker option at the higher price.

The four supplements with the strongest mitochondrial mechanism are CoQ10, riboflavin (B2), an NAD⁺ precursor (NR or NMN), and taurine. Each sits at a load-bearing spot in the energy chain, so the biology is unusually solid. How much they help in the clinic, though, ranges from one strong heart-failure trial (CoQ10) to mechanism-only longevity claims (NAD⁺). Here is each one.

CoQ10 (ubiquinone / ubiquinol)

Remember the courier that ferries electrons down the chain? That's CoQ10, and nothing else can do its job. It's also the membrane's main fat-soluble antioxidant. Your body makes it, but statins (cholesterol drugs) cut into that supply, lowering circulating CoQ10 by 16 to 54 percent across trials. And tissue levels in the human heart drop by roughly 50 percent from age 20 to 80 ^[22].

The headline trial is Q-SYMBIO (2014, 420 patients with moderate-to-severe chronic heart failure, 300 mg/day for 2 years) ^[19]. Serious heart problems were cut roughly in half (hazard ratio 0.50, 95% confidence interval 0.32 to 0.80). Over two years, 10 percent of the CoQ10 group died from any cause versus 18 percent on placebo. Deaths from heart causes: 9 versus 16 percent. Symptoms and hospital stays improved too. A separate trial, KiSel-10 (2013, with 12-year follow-up published in 2018), found lasting reductions in heart-related deaths among elderly Swedes taking selenium plus CoQ10 together ^[20]. The selenium muddies it, but the direction matches.

Where the evidence is mixed: for statin-related muscle pain, a 2015 meta-analysis came back inconclusive on both muscle-damage markers and pain. For male fertility, a 2013 meta-analysis found better sperm movement but no boost in actual pregnancies. For preventing migraines, a 2005 trial looked decent: 48 percent of people on CoQ10 cut their migraine days in half versus 14 percent on placebo (you'd need to treat 3 people for one to benefit). Where the evidence is clearly negative: Parkinson's disease. The big NIH QE3 trial (2014, 600 patients on 1200 or 2400 mg/day) was stopped early because it wasn't working, and the treated group actually trended worse ^[21]. So do not claim CoQ10 helps Parkinson's.

Ubiquinone versus ubiquinol (the two forms you'll see on labels): in healthy adults, your blood already runs about 95 percent ubiquinol anyway. The popular line that ubiquinol is 2 to 3 times better absorbed leans mostly on studies funded by the companies selling it. An independent 2019 comparison found that how the capsule is formulated matters more than which form it is. A well-made ubiquinone softgel can beat a poorly-made ubiquinol capsule. The dedicated CoQ10: Ubiquinol vs Ubiquinone guide has the full funding map and the DACH product detail.

Dose: 100 to 300 mg/day with a meal that has some fat, split into two if you go above 200 mg (you stop absorbing more past that). Q-SYMBIO used 300 mg/day. Trials have gone up to 1200 to 2400 mg/day safely. One interaction to watch: if you take the blood thinner warfarin, check your clotting numbers more closely above 100 mg/day.

Where the EU stands: food regulators reviewed CoQ10 in 2010 and turned down all six proposed claims (energy, blood pressure, antioxidant protection, brain function, cholesterol, and endurance) ^[23]. So there is no approved health claim for CoQ10 in the EU. The fair way to write about it is to report what the studies measured (Q-SYMBIO, the statin trials, even the rejection itself) and skip label-speak like "supports cardiovascular function."

Riboflavin (vitamin B2): The unsexy heavyweight

Riboflavin is vitamin B2, and your cells turn it into two helper molecules that the energy chain literally cannot run without. They sit at the entry point of the first station, lock into the second station, and feed the whole fat-burning pathway into the chain. So riboflavin props up three load-bearing spots in mitochondrial metabolism, and acts as a built-in part for roughly 90 different human enzymes.

Riboflavin actually treats some mitochondrial diseases. This is one of the cleanest "a vitamin fixes a genetic disease" stories in medicine. In a rare inherited disorder that blocks fat-burning enzymes (RR-MADD), oral riboflavin at 100 to 400 mg/day, under specialist neurology care, can turn things around. A 2007 study identified the genetic cause ^[25], and later groups report response rates near 98 percent in one subtype. In another rare disorder caused by a broken riboflavin transporter (Brown-Vialetto-Van Laere syndrome), high-dose oral riboflavin under specialist neurology care can be life-saving ^[26]. A 2017 case report describes an adult who went from paralyzed and on a ventilator to living independently on 1200 mg/day.

Migraine prevention is the best evidence outside the rare diseases. A 1998 trial: 55 patients, 400 mg/day for 3 months, and 59 percent cut their migraine days in half versus 15 percent on placebo (p=0.002, treat about 2.3 people for one to benefit) ^[24]. Trials in children at 200 mg/day came back negative (2008 and 2010). The main neurology bodies rate riboflavin a solid "probably effective" for preventing migraines in adults.

A cardiovascular niche. A 2013 trial in people already on blood-pressure drugs who carry a specific gene variant (MTHFR 677TT) found that just 1.6 mg/day of riboflavin for 16 weeks dropped top-number blood pressure by a further 5.6 mmHg ^[27]. Why that variant: it produces a wobbly enzyme that loses its riboflavin-based helper, and extra riboflavin shores it back up.

Dose, safety, EU rules: you only absorb about 27 mg from a single dose, so 400 mg/day works best split across the day. There is no EU upper limit because no toxic dose has ever been found. Take too much and your pee turns bright fluorescent yellow, which is harmless and just proof you absorbed it.

And riboflavin has one of the broadest sets of approved EU claims on this list: nine authorised Article 13(1) claims (Regulation 432/2012), including "contributes to normal energy-yielding metabolism" and "contributes to the reduction of tiredness and fatigue."

NAD⁺ precursors: Nicotinamide riboside (NR) and NMN

NAD⁺ is a helper molecule that hundreds of enzymes need, and it's the required fuel for the sirtuins (the same anti-aging enzymes from earlier), for DNA repair crews, and for that NAD⁺-burning enzyme CD38. NAD⁺ falls with age in several tissues. In human skin, a 2012 study found NAD⁺ correlated strongly and negatively with age across samples from newborns to age 77 ^[53]. You see the same shape in fat, muscle, and the brain. And again, this increasingly looks like a using-it-up problem driven by CD38 ramping up in age-related inflammation, not a making-it problem.

The mechanism. The longevity pitch goes like this: more NAD⁺ feeds the sirtuins, which switch on PGC-1α (the build-new-mitochondria master switch) and tune up the energy chain and your antioxidant defenses from inside the mitochondria. That's solid molecular biology. The trouble is the leap to real human results, which is much more modest.

Human trials for NR:

• 2016, single 1000 mg dose: blood NAD⁺ rose up to 2.7-fold. That 2.7x ceiling came from one participant in a pilot, not a group average. Across the 12-person cohort the dose-response was shallower ^[28].
• 2018, 1000 mg/day for 6 weeks: NAD⁺ in immune cells rose about 60 percent. The only blood-pressure and arterial-stiffness signal was a non-significant trend in an exploratory subgroup with higher baseline blood pressure, not a confirmed result ^[29].
• 2018, 2000 mg/day for 12 weeks in obese, insulin-resistant men: clearly negative on the metabolic endpoints, even though NAD⁺ went up ^[30].
• NADPARK (2022, Parkinson's, 1000 mg/day for 30 days): brain NAD⁺ rose, but the clinical improvement was small and exploratory.

Human trials for NMN:

• 2021 Science trial, 250 mg/day for 10 weeks in postmenopausal women with prediabetes: muscle insulin sensitivity improved on a gold-standard test ^[31].
• 2023 trial, 300 to 900 mg/day for 60 days: 6-minute walking distance improved dose by dose, leveling off at 600 mg ^[32].
• 2022 trial, 250 mg/day for 12 weeks: higher blood NAD⁺ and a slight bump in walking speed.

The 2PY/4PY caution. A 2024 Nature Medicine study looked at 4,000+ heart patients across three groups ^[33]. It tracked two breakdown products of niacin, shorthand 2PY and 4PY. When their levels sat in the top quartile (the highest 25 percent), they carried 1.6 to 2 times the 3-year risk of a major heart event (heart attack, stroke, or death). That held even after accounting for the usual risk factors. One of them, 4PY, irritates the lining of blood vessels. The catch: every NAD⁺ precursor (NR, NMN, and the older niacin forms) drains out through the same pathway that ends in 2PY and 4PY. Whether supplement doses of NR or NMN push those products into the danger zone has not been measured in trial groups yet.

Where the EU stands. NR (the branded Niagen) is approved as a Novel Food up to 300 mg/day for adults (under Regulation EU 2017/2470 and a 2022/1160 extension). NMN is not yet on the EU list of novel foods. EFSA gave it a positive safety opinion in May 2026 for food supplements up to 300 mg/day in adults (not during pregnancy or breastfeeding) ^[57], but the Commission hasn't formally authorised it, so selling it in the DACH region is currently borderline. Neither has an approved health claim.

Pairing it with a methyl donor (TMG). Taking NR makes your body spend more of its methyl groups to process the leftover nicotinamide. Trimethylglycine, also called betaine, refills that methyl pool. It's real biochemistry, but the case for routinely pairing TMG is mechanistic, not proven in trials. TMG does carry its own approved EU claim, "contributes to normal homocysteine metabolism" at 500 mg or more per portion and 1.5 g/day intake. That's the methyl-buffer angle, not a mitochondrial claim.

Where that leaves NR and NMN: both reliably raise NAD⁺ in people. The health payoffs in already-healthy adults are modest, with the cleanest signals in older or sick populations. The build-new-mitochondria story is mechanism, not outcome. And the 2PY/4PY caution is real and still unresolved.

Taurine: The underrated translation story

Taurine has a precise job inside your mitochondria: it gets attached to the little adapter molecules that read the mitochondrial genetic code and turn it into protein. Without taurine's tweak, those adapters start misreading certain letters, and the mitochondria can't build their own machinery properly ^[34]. So taurine quietly keeps your energy-chain parts coming off the line.

MELAS is the strongest anchor. A single mutation in one of those adapter molecules causes about 80 percent of MELAS, a brutal disease of systemic energy failure with stroke-like episodes, lactic-acid buildup, hearing loss, diabetes, and heart muscle disease. A 2019 phase III open-label trial gave 9 to 12 g/day of oral taurine for 52 weeks to 10 MELAS patients ^[35]. The yearly rate of stroke-like episodes fell from 2.22 to 0.72 (p=0.001), and the taurine tweak on those adapter molecules recovered in their blood cells. Japan approved taurine for MELAS in 2019. Outside Japan it's off-label but standard in specialist mitochondrial-disease centers.

Heart evidence: mice bred without a working taurine transporter develop a shrunken, weakened heart and tire out faster (a 2008 study). The mechanism is mitochondrial: the missing taurine tweak breaks the local protein building, and the energy chain fails.

The 2023 Science paper and its 2025 rebuttal. A 2023 Science paper reported that blood taurine drops with age across several species and that 1 g/kg/day of oral taurine extended the median lifespan of middle-aged mice by about 10 to 12 percent ^[36]. It got huge coverage as a likely longevity move. Then a 2025 Science paper, using the Baltimore Longitudinal Study of Aging and other groups, challenged the core claim and found that taurine actually rose or held steady with age in all three human cohorts they sampled, and in monkeys and mice too ^[37]. The NIH summary on the rebuttal concluded taurine is "unlikely to be a good aging biomarker." The MELAS biology and the mouse-heart story stand untouched. But the "you're running low, just refill it" narrative that fueled the 2023-to-2024 supplement wave has not held up.

Dose, safety, EU status: trials use 1 to 6 g/day; MELAS uses 9 to 12 g/day under specialist care. The observed safe level, the highest daily intake with strong evidence of no harm, is 3 g/day. Above that, long-term safety data are thin ^[52]. Taurine is an ordinary, long-used food ingredient in the EU (think energy drinks and infant formula) but has no approved health claim. EFSA rejected every proposed claim in 2009 and 2011. So in the EU you can't market it with heart, mitochondrial, energy, or anti-fatigue claims.

The core four in one paragraph

CoQ10 is biochemically central and has the single best positive trial here (Q-SYMBIO). It is not a longevity drug. Riboflavin is the cheapest defensible mitochondrial helper, with a broad set of EU claims and a genuine role in treating mitochondrial disease. NAD⁺ precursors reliably raise NAD⁺, but the human longevity case is still preclinical and the 2PY/4PY question is open. Taurine has a beautiful mechanism and a phase III trial in MELAS; the broader longevity case took a real hit in 2025.

These have real mitochondrial biology behind them and weaker human evidence. Useful, but keep your expectations calibrated.

Magnesium (bisglycinate or similar)

Magnesium is the partner mineral that makes ATP usable, and it props up two of the energy chain's stations. The EU upper limit for supplements is 250 mg/day; the US figure is 350 mg. The form you pick matters more for how your gut tolerates it than for absorption:

• Oxide is laxative and only about 14 to 23 percent absorbed.
• Citrate sits in the middle.
• Bisglycinate is gentle on the stomach. Its glycine carrier brings along a mild sleep-friendly molecule for free (see the deep sleep guide).

L-threonate (Magtein) is priced at a premium on the strength of one rat brain study (2010) and an industry-funded sleep-ring trial in humans (2024). The price doesn't match the independent evidence.

Magnesium carries ten approved EU claims, including "contributes to normal energy-yielding metabolism," "normal muscle function," and "reduction of tiredness and fatigue."

Creatine monohydrate

Creatine doesn't build new mitochondria. It works more like a rechargeable ATP battery: it stores energy and snaps it back to ATP exactly where demand spikes, in muscle fibers mid-contraction and in neurons mid-thought. The strongest evidence is in muscle performance. A 2018 meta-analysis also found brain benefits in older adults, sleep-deprived people, and vegetarians (who start out lower on creatine). A 2024 study found that a single 0.35 g/kg dose during 21 hours of sleep deprivation improved thinking and measurably shifted brain energy stores on a specialized scan. The old kidney-safety scares in healthy adults have been knocked down over and over.

Dose: 3 to 5 g/day for muscle; 5 to 10 g/day if your goal is brain. Monohydrate is the only form with solid evidence. The EU has two approved claims: "increases physical performance in successive bursts of short-term, high-intensity exercise" (3 g/day or more) and "can enhance the effect of resistance training on muscle strength in adults over the age of 55" (3 g/day or more plus resistance training, Regulation EU 2017/672) ^[50]. A brain-function claim was rejected by EFSA in 2024.

Methylcobalamin (vitamin B12)

B12 powers an enzyme that lives right inside the mitochondria and feeds certain carbon fragments into your central energy cycle. When you're low on B12, a marker called methylmalonic acid (MMA) rises, which is a direct, measurable sign that this mitochondrial step is struggling. Quietly low B12 is common: 10 to 20+ percent of people over 60, people on the diabetes drug metformin (a 2010 BMJ trial), people on acid-blocking PPI drugs (a 2013 JAMA study), and vegans (50+ percent without supplements).

Forms: methylcobalamin, adenosylcobalamin (the actual mitochondrial form), hydroxocobalamin (the injection standard), and cyanocobalamin (cheapest, fine for most people). In practice, the dose matters far more than the form. The EU has eight approved claims, including energy metabolism, nervous system, homocysteine metabolism, and reduction of tiredness and fatigue.

Curcumin

Curcumin nudges some of the same anti-inflammatory and mitochondria-building pathways from earlier, including PGC-1α indirectly. But the human results are mostly about inflammation, not direct mitochondrial endpoints. The big problem is absorption: plain curcumin is only about 1 percent absorbed. Boosted forms (Theracurmin, Meriva, Longvida, CAVACURMIN, C3 Reduct) get 27 to 65 times more into your blood. The best clinical data are for osteoarthritis pain (a 2016 meta-analysis), as an add-on for depression (a 2017 meta-analysis), and for metabolic syndrome and blood lipids (a 2017 meta-analysis). Direct mitochondrial endpoints in humans are thin.

Safety: rare but real reports of liver injury with the high-absorption forms (registry entries from Italy, Australia, and the US since around 2018). Be careful if you're on statins or other liver-stressing drugs. No approved EU health claim. The "contributes to normal joint function" application was rejected by EFSA in 2017.

PQQ (pyrroloquinoline quinone)

In cell cultures and rodents, PQQ switches on PGC-1α and grows new mitochondria (a 2010 study). But the human evidence is thin: one small crossover study (2013, just 10 people) showing some biomarker shifts. There's no human trial measuring real clinical endpoints at normal supplement doses. In the EU it's an approved Novel Food up to 20 mg/day (Regulation 2018/1122) ^[49], with no approved health claim. The "mitochondrial biogenesis nutrient" marketing isn't backed by human trials. Treat it as preclinical to weak.

These got swept up into mitochondrial-supplement marketing, but in the human evidence they aren't really mitochondrial agents. Each one has a legitimate job. None of those jobs is "directly improves mitochondrial function in healthy adults."

Citicoline (CDP-choline) helps build neuron membranes and supplies a brain messenger chemical. The big ICTUS stroke trial (2012) was negative and stopped early for not working. It only brushes mitochondria through a membrane fat called cardiolipin, and only after injury. It's an approved Novel Food in the EU up to 500 mg/day, with no approved health claim. Peripheral, not a mitochondrial ingredient for healthy people.

Rhodiola rosea shows faint mitochondria-building signals in rodent muscle in a dish. No human study has shown Rhodiola changing mitochondrial breathing, mitochondrial growth, or PGC-1α in actual tissue. A 2011 review concluded the effects were "not convincingly demonstrated." EFSA rejected the "reduction of mental fatigue" claim, and the EU medicines agency lists Rhodiola as a traditional herbal medicine based on long use, not modern trials. It's an adaptogen with a mitochondrial story stretched from preclinical pathway data.

P5P (the active form of vitamin B6) is a helper for around 150 enzymes, some of them in the mitochondria (for building the oxygen-carrying part of blood, for one-carbon chemistry, and for shuffling amino groups). But it's permissive, not a driver: topping it up in people who aren't deficient doesn't boost mitochondrial output. EFSA's 2023 review lowered the upper limit to 12 mg/day (from 25 mg), because long-term high B6 causes nerve damage. Vitamin B6 has ten approved EU claims, and the form you pick (P5P versus pyridoxine HCl) matters more for nerve-damage risk at chronic high doses than for any clinical benefit at normal supplement levels.

5-MTHF (the ready-to-use form of folate) sidesteps a common gene variant (MTHFR 677TT) that slows folate processing. There is a real folate cycle inside mitochondria (a 2017 review laid out how it exports a key carbon building block to the rest of the cell), but 5-MTHF supplements act mainly on methylation outside the mitochondria, with only indirect knock-on effects inside. Best human result: a 2012 study found that 15 mg of L-MTHF added to an antidepressant produced a 32 percent response in treatment-resistant depression versus 15 percent on placebo. Folate carries several approved EU claims, including "normal homocysteine metabolism."

Betaine (TMG, trimethylglycine) is a methyl donor. It works outside the mitochondria, mostly in the liver, so it's not a mitochondrial ingredient. Its sensible pairing is with NAD⁺ precursors: taking NR or NMN drains your methyl pool, and TMG refills it. The approved EU claim is "contributes to normal homocysteine metabolism" (500 mg or more per portion, 1.5 g/day intake). That's methyl-pool buffering, not mitochondrial. Best used alongside NR/NMN, never sold solo as a mitochondrial product.

Myo-inositol is a cellular signaling molecule that helps with insulin signaling. Strongest evidence: polycystic ovary syndrome (a 2017 meta-analysis) and possibly gestational diabetes (a 2023 Cochrane review). Not a mitochondrial agent. An insulin sensitizer at best.

Piperine (BioPerine, black pepper extract) blocks the enzymes and pumps that clear drugs from your body. Its only real job is to raise the absorption of whatever it's taken with, like curcumin (a 1998 study found roughly a 2000 percent jump in curcumin blood levels when paired with 20 mg of piperine). It has no direct mitochondrial action. And it carries a non-trivial interaction risk: it raises blood levels of tacrolimus, cyclosporine, simvastatin, midazolam, calcium channel blockers, theophylline, phenytoin, and more. Cap it at 5 mg/day or less in supplements, and avoid it entirely if you're on any of those drugs.

Bottom line on this group: these play supporting roles. Useful in specific situations (the MTHFR gene variant, refilling methyl groups for NR/NMN, boosting curcumin absorption, treating a PCOS overlap), but not headline mitochondrial ingredients. If the marketing calls any of them a "mitochondrial booster," push back.

For almost everything mitochondrial, food beats single-compound pills. Three categories are worth knowing about.

Urolithin A (Mitopure)

Urolithin A is a postbiotic, a compound your gut bacteria make for you. They produce it from plant compounds in pomegranate, walnuts, and berries. The catch: roughly half of Western adults can't make much of it no matter how much pomegranate juice they drink, because they lack the right gut bugs. Taking it directly skips that microbiome lottery.

The mechanism is unusually specific for a supplement: it directly triggers mitophagy, that targeted recycling of worn-out mitochondria, through the same cleanup system from earlier. Researchers at EPFL found it by screening more than 1,500 compounds for exactly this effect. A 2016 Nature Medicine paper then showed it induces mitophagy, extends lifespan in C. elegans, and improves muscle function in rodents ^[51]. Three trials, all funded by the maker (Amazentis), form the human evidence:

• A 2019 Nature Metabolism paper: a first-in-human safety and absorption study in healthy older adults, showing dose-dependent blood levels and shifts in muscle mitochondrial genes and fat-burning byproducts ^[38].
• A 2022 JAMA Network Open trial: 66 adults aged 65 to 90, 1000 mg/day for 4 months. It missed both co-primary endpoints (the 6-minute walk test and muscle ATP production), but improved muscle endurance (more contractions before fatigue in hand and leg tests) and lowered inflammation and fat-burning byproduct markers ^[39].
• A 2022 Cell Reports Medicine trial: 88 middle-aged adults, 500 or 1000 mg/day for 4 months. It missed its main peak-power goal, but produced about a 12 percent muscle strength gain, improved peak aerobic fitness and walking distance, and lowered those same byproducts plus CRP (an inflammation marker) ^[40].

Where that leaves urolithin A: this is the most distinctive recent mitochondrial-supplement story, with human trials in top-tier journals. But all three were funded by the maker and all three missed their primary goals. The case rests on secondary results and biomarkers, the effects are modest, and the mechanism is unusually clean for a supplement.

Dose: 500 to 1000 mg/day. EU: approved as a Novel Food (Mitopure, a synthetic version); no health claim.

Omega-3 EPA and DHA

This is one of the few entries on the list with approved EU health claims at a specific intake (creatine and betaine also have them, see the claims section). The mitochondrial angle runs through cardiolipin, a membrane fat that holds the energy chain's big assemblies together. DHA in particular gets built into cardiolipin and reshapes those assemblies (a 2018 study).

What the human data show:

• A 2021 study: the long-running Framingham Offspring group, about 2,200 adults followed for 11 years. People with more omega-3 in their red blood cells were less likely to die from any cause, with an effect on the same scale as smoking. This one is observational, so it shows a link, not proof ^[43].
• REDUCE-IT (2019): 4 g/day of pure EPA in statin-treated patients with high triglycerides cut major heart events by 25 percent ^[41].
• VITAL (2019): 1 g/day of EPA+DHA for general prevention was null overall, with hints of benefit in people who ate little fish ^[42].
• STRENGTH (2020): 4 g/day of a different EPA+DHA formulation was null. Some read REDUCE-IT's win as partly an artifact of its mineral-oil placebo.

Approved EU claims (Regulation 432/2012):

• "EPA and DHA contribute to the normal function of the heart" at 250 mg combined per day.
• "DHA contributes to maintenance of normal brain function" and "normal vision" at 250 mg DHA per day.
• At higher intakes (2 to 3 g/day): blood pressure and triglyceride claims.

Safety: up to 5 g/day combined is safe (EFSA 2012). There's a modest signal of irregular heartbeat at gram-level doses across REDUCE-IT, STRENGTH, and OMEMI.

Polyphenols: Drink the tea, eat the berries

Resveratrol has loads of hype and modest human evidence. The original anti-aging-enzyme story (a 2006 mouse paper) took a hit when a 2010 paper showed much of that lab-dish activation was just a measurement artifact. A 2011 trial (150 mg/day in obese men) found small gains in mitochondrial breathing and liver fat; later trials are mixed to null. It's barely absorbed (under 1 percent stays free in the blood) and it blocks a key drug-clearing enzyme, so it can interact with statins, blood thinners, and immune-suppressing drugs. EU Novel Food up to 150 mg/day; no approved claim.

EGCG (green tea) switches on protective pathways in preclinical models. In people, the heart-and-metabolism effects are small. EFSA's 2018 risk assessment flagged a liver-injury risk for green tea extracts above 800 mg EGCG/day, especially on an empty stomach. Drink the brewed tea (no liver injury reported from a cup of tea); just be careful with concentrated extracts.

Fisetin clears out senescent (worn-out, no-longer-dividing) cells (a 2018 study extended both median and maximum lifespan in mice) ^[45]. Mayo Clinic Phase 2 trials at 20 mg/kg/day for 2 days a month are running. As of September 2026, ClinicalTrials.gov still lists the main ones (the AFFIRM frailty trials) as ongoing, and none has published primary results. Don't pay for "senolytic" branding before the trial results land.

Sulforaphane (from broccoli sprouts) is the strongest food-based switch for your antioxidant defenses known. Human trials in autism (a 2014 study) and recurrent prostate cancer (a 2015 trial) are encouraging but not mitochondria-specific. A daily handful of broccoli sprouts (with a little mustard powder to activate it) delivers about 30 to 60 mg of sulforaphane equivalent.

The Mediterranean pattern: Top-grade whole diet

PREDIMED (the corrected 2018 reanalysis) randomized about 7,400 adults at high heart risk to a Mediterranean diet with extra-virgin olive oil or mixed nuts, versus a low-fat control ^[44]. Major cardiovascular events fell about 30 percent over 5 years. The polyphenol-rich Mediterranean pattern beats any single-compound supplement in long-term outcome data. And you can't pin the benefit on one ingredient: it's the whole mix (fiber, healthy fats, marine omega-3, polyphenols, lower blood-sugar load, even the social side of eating well). Which is exactly the point.

This is an editorial guide, not a supplement label. And that line matters legally. Regulation (EC) No 1924/2006 governs health claims on commercial communications (product packaging, advertising, sponsored content). Editorial journalism is carved out by Article 1(2) and Recital 4, as long as the writing stays scientific rather than promotional.

The safe way to write it: describe what the studies measured. "A trial reported that 300 mg/day CoQ10 cut major adverse cardiac events by half in moderate-to-severe heart failure ^[19]." That's reporting on a study. It is not a claim that CoQ10 lowers your own heart-failure risk.

The unsafe way: "CoQ10 supports cardiovascular function," or "take 300 mg of CoQ10 to support your heart." That reads like a label.

Authorised Article 13(1) claims (Regulation 432/2012) for nutrients discussed in this guide:

• Magnesium: "contributes to normal energy-yielding metabolism," "to a reduction of tiredness and fatigue," "to electrolyte balance," "to normal functioning of the nervous system," "to normal muscle function," "to normal protein synthesis," "to normal psychological function," "to the maintenance of normal bones," "to the maintenance of normal teeth," and "has a role in the process of cell division."
• Zinc: claims on metabolism, cognitive function, immune function, cell division, protection from oxidative stress, and more.
• Riboflavin (B2): nine claims including normal energy-yielding metabolism, normal functioning of the nervous system, reduction of tiredness and fatigue, protection of cells from oxidative stress.
• Niacin (B3), Pantothenic acid (B5), Vitamin B6, Vitamin B12, Folate: each carries energy-metabolism and fatigue-reduction claims; B6 was downgraded to UL 12 mg/day in 2023.
• Vitamin C: energy metabolism, immune function, collagen, iron absorption, oxidative-stress protection.
• Vitamin D: bone, muscle function, immune system, calcium absorption.
• EPA + DHA: "contribute to the normal function of the heart" at 250 mg/day combined. DHA alone has claims for brain function and vision at 250 mg/day.
• Creatine: "increases physical performance in successive bursts of short-term, high-intensity exercise" at ≥3 g/day; "can enhance the effect of resistance training on muscle strength in adults over 55" at ≥3 g/day plus resistance training.
• Betaine: "contributes to normal homocysteine metabolism" at ≥500 mg per portion, 1.5 g/day intake.

No authorised claim, in the EU, for:

• CoQ10 (EFSA rejected all six categories in 2010).
• NR (Niagen has Novel Food authorisation at ≤300 mg/day; no health claim).
• NMN (EFSA positive safety opinion May 2026 for adults up to 300 mg/day ^[57]; Commission Implementing Regulation has not authorised; sale in DACH is borderline).
• Taurine (EFSA rejected all proposed claims).
• Resveratrol (Novel Food ≤150 mg/day; no claim).
• Curcumin (food additive E100; "joint function" claim rejected by EFSA 2017).
• L-theanine, glycine: no authorised claims.
• Urolithin A (Mitopure): Novel Food authorised; no health claim.
• PQQ (Novel Food ≤20 mg/day; no claim).
• Citicoline (Novel Food ≤500 mg/day; no claim).
• Fisetin: uncertain Novel Food status; no claim.

EU Tolerable Upper Intake Levels (the EFSA figures are scientific advice, not binding law):

• Magnesium 250 mg/day supplemental (NIH: 350)
• Zinc 25 mg/day total (NIH: 40)
• Vitamin B6 12 mg/day (revised down 2023; NIH: 100)
• Vitamin D 4000 IU/day
• Selenium 255 µg/day (revised down 2023)
• EPA+DHA 5 g/day combined is safe
• Riboflavin: no UL (no toxicity threshold)
• B12: no UL

What this means specifically in DACH. Germany applies the Federal Court of Justice's "doppelte Zweckbestimmung" doctrine (dual-purpose classification). If a food supplement is labeled or advertised with disease-related effects, it can be reclassified as a functional medicinal product under §2(1) AMG, the German Medicines Act. Then the drug regulator BfArM steps in. Melatonin shows how this plays out (relevant to the companion sleep guide). A May 2023 ruling (OLG Koblenz 9 U 1947/22) eased earlier strictness on products at 1 mg or below. But a 17 September 2024 statement from the BfR, Germany's federal risk-assessment institute, still warns that melatonin is "not a gentle sleep aid." The practical line: 1 mg or less is acceptable as a food supplement in Germany. Above 1 mg, you drift toward medicinal-product territory.

The pattern we follow on this site: talk about the science openly, cite the trials, cite the regulators (including their rejection notices), state the safety thresholds whenever a dose comes up, and never write "take X to do Y" in a product-selling voice.

The evidence-first mitochondrial stack, in order of evidence-to-effort: start with lifestyle (zone 2, HIIT, lifting, sleep, protein, an early eating window), then a Mediterranean-pattern diet plus omega-3, then defensible cofactors (a B-complex multivitamin, magnesium bisglycinate, vitamin D, B12 if at risk). Everything past that is fine-tuning. Foundation always beats pills. Here is the full tier list.

Before the tier list, one ground rule. The Tier 2 to 5 doses below are the doses commonly studied or sold over the counter. They are not a personal recommendation. Talk to your family doctor (in Germany, your Hausarzt) before starting a stack, especially if you take medication or have a chronic condition.

Tier 0: Foundation (not supplements). Zone 2 endurance 3 to 4 sessions per week. HIIT 1 to 2 sessions. Resistance training 2 to 3 sessions. Sleep 7 to 9 hours. Protein at least 1.6 g/kg body weight. Time-restricted eating with an earlier window if it fits your life. Don't smoke. Moderate alcohol or none. Limit ultra-processed food. This one tier produces measurable gains that no supplement matches: higher VO₂ max, a higher lactate threshold, more mitochondrial protein, and more activity in the genes that run the energy chain (OXPHOS genes). Skip this tier and the supplement tiers below will not rescue it.

Tier 1: Foods with the best evidence.

• A Mediterranean-pattern diet: vegetables, legumes, whole grains, fish, extra-virgin olive oil, nuts, modest red wine if you drink at all. PREDIMED-grade evidence.
• Fatty fish 2 to 3x per week or omega-3 EPA+DHA ≥250 mg/day (EU-authorised heart claim).
• Berries, walnuts, pomegranate: they carry ellagitannins, the plant compounds your gut bacteria turn into urolithin A. Half of you won't make much urolithin from them; the other half will.
• Broccoli sprouts with a pinch of mustard powder (its enzyme, myrosinase, switches the sulforaphane on).
• Green tea (brewed, not concentrated extracts above 800 mg EGCG).

Tier 2: Defensible cofactor coverage. Especially relevant for older adults, vegans, metformin or PPI users, statin users, or anyone whose diet does not consistently cover the basics.

• A multivitamin with riboflavin, B12, folate, and the rest of the B-complex at roughly the recommended daily intake. Riboflavin has a broad EU claim portfolio and the cleanest mitochondrial cofactor story.
• Magnesium bisglycinate: typical consumer dose 200 to 400 mg elemental in the evening. EU authorised claims for muscle, nervous system, and fatigue.
• Vitamin D: dose set individually to keep your blood vitamin D level (25(OH)D) at 30 to 50 ng/mL. Test, don't guess.
• Methylcobalamin or cyanocobalamin: studied dose 500 to 1000 µg sublingual or oral in deficiency-risk groups (vegan, metformin, PPI, age 60+). MMA and homocysteine are the functional markers for a quantitative read.

Tier 3: Discretionary mitochondria-leaning additions.

• Creatine monohydrate: studied dose 5 g/day for muscle endpoints, with 5 to 10 g/day used in cognitive or sleep-deprivation studies. Two authorised EU claims. ATP buffer, not biogenesis.
• CoQ10: studied dose 100 to 200 mg/day taken with fat. The case is strongest for statin users, adults over 60, and anyone with cardiovascular risk. No authorised EU claim. This is evidence-based personal choice territory, not a prescribed mitochondrial booster.
• Omega-3 EPA+DHA: typical consumer dose 1 to 2 g/day in people who do not eat fatty fish twice a week.

Tier 4: Higher-cost, higher-uncertainty experiments.

• NR: typical consumer dose 300 mg/day for testing the NAD⁺ story (EU Novel Food authorised; only Niagen / ChromaDex). Usually paired with TMG 500 to 1000 mg to cover the methyl-pool tax.
• Urolithin A (Mitopure): studied dose 500 mg/day, most relevant for adults over 60 worried about age-related muscle loss (sarcopenia) who have the budget. Three RCTs missed their primary endpoints but produced reasonable secondary signals.
• Curcumin in a bioenhanced form (Meriva, Theracurmin, Longvida, CAVACURMIN): typical consumer dose 500 to 1000 mg/day for joint or anti-inflammatory use cases. Not a primary mitochondrial agent. Watch for liver injury.
• Taurine: studied dose 1 to 3 g/day, cheap and safe at that range. The MELAS evidence is real; the longevity case took a hit in 2025. Plausible adjunct, not a headline ingredient.

Tier 5: Where the evidence is preclinical or contested.

• NMN. EFSA opinion suggests up to 300 mg/day is safe; not yet on the EU Novel Food list; sale in DACH is regulatorily borderline. Comparable mechanistic case to NR but weaker EU legal footing.
• PQQ: typical consumer dose 10 to 20 mg/day. Marketed as the biogenesis nutrient; human evidence is one n=10 biomarker study.
• Fisetin. Skip until Mayo Clinic Phase 2 results publish.
• Resveratrol. The SIRT1 story has largely unwound. Skip unless you're keen on polyphenol mechanism research.

If you only do one thing from this guide: do Tier 0. If you only buy one thing: a quality multivitamin covering riboflavin, B12, folate, and magnesium bisglycinate. Everything else is fine-tuning. And fine-tuning never beats foundations.

The main supplements at a glance

The EU column matters for labels more than for you. A missing claim means no regulator accepted the evidence for a label statement. It does not automatically mean the compound is useless.

A note on marketing versus mechanism

The mitochondria-supplement space is among the most aggressively marketed corners of consumer health. The mechanism stories are real. CoQ10 really is the electron courier the chain can't run without. Riboflavin really is the raw material for the helper molecules (FMN and FAD) at the first and second stations. Taurine really does tag the mitochondria's code-reading adapters (mt-tRNAs). NAD⁺ really does decline with age. But a mechanism is not the same as longevity outcome data. Every time a label or an influencer turns "X is required for the energy chain" into "X extends your healthspan," they skip the step that needs a randomized controlled trial in humans.

The best you can buy with money is high-quality food, time to exercise, time to sleep, and competent medical care. The cofactor supplements above are second-order optimizations. PQQ is not going to add a decade to your life. Lifting weights twice a week and getting eight hours of sleep might.