Most fitness challenges are built on a hunch and a whiteboard. This one is built on what the research actually found, including the parts that are inconvenient for us. Here is the evidence, the numbers, and where we have deliberately gone against the grain.
This is the finding that shaped the whole product. A meta-analysis of self-weighing trials asked a simple question: does measuring yourself actually make you lose weight?
And across the fifteen trials in that particular comparison, seven instructed daily weighing and eight less often. There was no significant difference between them. Separately, trials that built in accountability produced more weight loss than trials that did not.
Two honest caveats before we lean on that. The accountability comparison was made across trials rather than inside one, which the authors describe as observational, and only two trials sat in the no-accountability arm. And "no significant difference" between daily and weekly is a failure to find one, not proof there is none. The authors' own conclusion was that accountability may improve effectiveness. We think it is the best available read. We are not going to pretend it is settled.
It is still the entire argument for what we do. The scale is not the intervention. Showing up, being watched, being part of something with a scoreboard: that is the intervention. Which is why our challenge scores you on compliance, why coaches watch all 42 days, and why the prize does not go to whoever dehydrated hardest before the final weigh-in.
A cohort study followed 9,768 smart scale users for around three years in the real world, not a lab. People weighed in an average of 2.8 times a week, and more frequent measuring correlated with better weight outcomes, most strongly for the people with the most to lose.
The striking part is what happened when people stopped. Nearly three quarters of participants had at least one gap of 30 days or more without measuring, and during those gaps weight went up in every single group:
Read that as a pattern, not a mechanism. It is an observational study, the correlation between weighing frequency and weight change was weak (it explains about 1% of the variation), and the authors note the relationship may describe weight loss rather than cause it. What it does establish is that the gaps are where the drift lives.
That is why our app offers weekly, fortnightly and three-weekly check-ins, and nothing longer. Three weeks is a deliberate ceiling that keeps every member a full week inside the gap where the data says people drift backwards. It is also why the app pauses your position on the mountain if a check-in goes overdue: not as punishment, but because the evidence says the gap itself is the risk.
Standard practice for body composition scanning is every 8 to 16 weeks, because hydration, glycogen, recent training and last night's dinner move a body fat reading by more than real tissue change does over a fortnight.
So we will say it plainly: a single fortnightly body fat percentage is mostly noise. We could hide that and let you believe every reading is a verdict on your effort. We would rather tell you, because the member who understands this is the member who does not quit in week three over a number that was water.
What is meaningful is the trend across check-ins, and the comparison that our guarantee actually rests on: day 1 against day 42. Six weeks is comfortably long enough for real change to outrun the noise. That is the number we hold ourselves to.
The "21 days to form a habit" line is folklore. It comes from a 1960 book by a plastic surgeon, who was writing about how long it takes a person's self-image to catch up with a changed face, and who said "a minimum of about 21 days". It was never about habits at all.
The actual research puts the median closer to two months, with enormous variation. The best-known study found a median of 66 days across a range of 18 to 254, and that missing a single day did not meaningfully damage the process. A 2024 meta-analysis of 20 studies and 2,601 people landed in the same place on the median, 59 to 66 days, but with individuals ranging from 4 days to 335. So: about two months for a typical person, and anyone quoting you a single number for everybody is guessing.
Forty-two days is our honest compromise: long enough that the behaviour starts to run on rails rather than willpower, short enough that a person with a job and a family can see the end from the start. It is also why our rule is no zero days rather than perfect days, and why missing once is designed to cost you almost nothing. The rule is never perfection. The rule is never zero.
A target of 100% has one failure mode and it is fatal: the first miss ends the streak, the streak was the motivation, and the person is gone by week two. Ninety percent is high enough that it demands real consistency across all 42 days and cannot be salvaged by a heroic final week, but forgiving enough that a bad Tuesday is a bad Tuesday rather than a reason to quit. It is the difference between a standard people chase and a standard people abandon.
The oldest version of this idea we have found is the kaihōgyō, a walking practice of the Tendai monks on Mount Hiei in Japan. The practice is set in 100-day terms, each day a fixed mountain route, and the term is scored on whether the walker completes it, not on how fast or how elegantly. Fewer than 50 monks have finished the full 1,000-day version since 1885. The structure is the same shape as a climb: a fixed term, a daily minimum, a hard finish line, and a rule for the days that go wrong. Ours is the catch-up week. Theirs is less forgiving. We borrowed the shape, not the severity.
Once you have held a climb at 90%, the obvious question is whether to make the next one harder. The answer is yes, and the research points that way. A meta-analysis of goal setting in behaviour-change trials found that difficult goals produced close to double the effect of easy or moderate ones (1.8 times, to be exact about it). Comfortable targets change less than uncomfortable ones. Worth knowing that this is a comparison across studies rather than a trial that raised people's targets and watched what happened, so read it as a strong direction rather than a dose you can prescribe.
That last one is the part that matters more than the headline. The advantage only holds while the goal stays reachable. The same authors flag that improbable goals may work against you, and note that almost every goal they classified as difficult was still something the person could actually do. So the active ingredient is not difficulty. It is difficulty you still meet. Push past that line and you are not adding stimulus, you are manufacturing failure, and a broken streak costs you more than a slightly harder target ever gains.
Here is where we have to be straight with you: nobody has established a number. We went looking specifically for one. The goal-setting literature gives direction, not magnitude, and the step-count and physical-activity work is the same. There is no validated percentage for how far to raise a behavioural target, and anyone quoting you one is extrapolating.
There is a tempting shortcut we chose not to take. A well-known paper found that learning is fastest when you succeed about 85% of the time, which lines up almost too neatly with our 90% ridge. But that 85% is a mathematical result for gradient-descent learners on two-choice tasks, and it moves to 18% or 25% if you change one assumption about the noise. The authors are open about the scope, and they do connect it to older ideas in education, but they tested it on artificial networks and a model of animal learning, not on people changing their habits. Borrowing the number as fitness science would be exactly the sort of thing this page exists to avoid.
So we use the one thing that is genuinely reliable: your own last 42 days. You finished a climb with a record of which targets you actually hit and which you did not. That beats any general rule, because it is you.
Stretch one target. Two at the very most, never more. That might be an extra session a week, or the same frequency at a slightly higher bar. If you came in under 90%, stretch nothing yet: get one climb at 90% on targets you can hold, then raise one.
We deliberately do not give you a percentage. Achievability is the binding constraint and it applies to your whole week, not to each target separately, so raising four at once is not four small stretches, it is one large one. And most people change some of their targets between climbs anyway, which makes any number calculated from the last set a description of a week that no longer exists.
The job is to guarantee 90% again. Hold that climb after climb and the progression takes care of itself: five modest stretches across a year compound into something one ambitious jump never does, because the ambitious one usually breaks the streak that was doing the work.
Fair question, and we went looking for the answer rather than assuming it. A meta-analysis of the psychological outcomes of self-weighing found essentially no association with depression or anxiety, with body-related attitudes, or with disordered eating. There was a small association with poorer psychological functioning: slightly lower self-esteem and slightly higher stress. Small, but real, and we would rather you knew.
Two findings we took seriously. Self-weighing inside a structured, supported programme produced better outcomes than people doing it alone and unprompted, which is an argument for coaching rather than against measuring. And younger and adolescent samples showed a stronger link to disordered eating, which is why our coaches are briefed on it and why any member can switch off their own body fat readings in the app and simply log the numbers without looking at them.
This is the most quietly decisive finding on this page. Researchers put ten overweight adults on an identical calorie deficit for two weeks, twice. The only thing they changed was time in bed: 8.5 hours in one condition, 5.5 hours in the other.
Same food. Same deficit. Same fortnight. On short sleep, three quarters of what they lost was not fat. They also reported significantly more hunger, with measurably elevated ghrelin, the hormone that drives appetite. So the short-sleeper was hungrier, and what they lost for their trouble was largely the tissue they least wanted to give up.
The caveat, because this page does caveats: that was ten people over a fortnight in a sleep laboratory, and the spread around those percentages was enormous. A larger follow-up reproduced the fat-versus-lean direction but not the size of the lean-tissue effect. Treat it as a strong steer rather than a settled number.
Even discounted, it is why sleep sits in our target list next to training and food rather than in a wellness afterthought. If you are running a deficit on five hours a night, you are not dieting. You are just shrinking.
Somewhere between day 16 and day 30, the scale usually stops moving. We call it the dead-zone, and it is the single most common place people quit. It is also almost entirely predictable, which is why our app warns you before it arrives rather than after you have already given up.
Three things happen at once:
The dead-zone is not a signal to change the plan. It is a signal that the plan has reached the part where patience does the work. It usually breaks suddenly rather than gradually. The people who finish as The Unbroken are rarely the ones who avoided it: they are the ones who kept logging through it.
Alcohol carries 29 kilojoules per gram, sitting between protein and carbohydrate (17 kJ) and fat (37 kJ). A standard drink is 10 g of alcohol, so around 290 kJ before a single mixer is added. In practical Australian terms:
Ranges rather than single numbers, because the brand and the strength move it. A 150 mL glass of wine runs roughly 425 to 500 kJ, and a 150 mL pour of a 14% red is 1.6 standard drinks, not one. The mixer matters more than people expect: a spirit with soda is around 257 kJ, while a full-sugar gin and tonic reaches 840 kJ and a rum and cola 930 kJ for identical alcohol. Switch to a zero mixer and both collapse back to about 270. Four schooners on a Friday is roughly 600 to 700 calories, which for most people is a third of a day, and it tends to arrive on top of dinner rather than instead of it. It is also about 6.4 standard drinks, past the NHMRC's limit of four on any one day.
Beyond the calories: alcohol suppresses REM sleep and fragments the second half of the night (deep sleep early actually goes up, which is why a nightcap feels like it worked and then does not), and it reliably loosens food decisions - an Australian meta-analysis put the extra intake at about 343 kJ, with no compensation for the drink's own energy. It also blunts the muscle-building response to training, though the clear evidence for that is at binge doses rather than a couple of beers.
And there is the part the drinks industry would rather stayed quiet. IARC, the World Health Organization's cancer agency, classifies alcohol as a Group 1 carcinogen, alongside tobacco and asbestos. One thing that gets misused: Group 1 describes how certain the evidence is that something causes cancer, not how big the risk is. Alcohol and asbestos share a confidence rating, not a danger rating. What is established is that alcohol causes at least seven cancers, including bowel and female breast, and that no safe threshold has been identified.
You will see it claimed that half of Europe's alcohol-related cancers come from light and moderate drinking. That line is in a WHO press release, but the peer-reviewed studies underneath it put the figure closer to 13 to 14%, so we are not going to repeat the bigger number. The finding that does hold up, and is the one worth knowing: roughly a third of alcohol-caused breast cancers in the WHO European Region came from drinking no more than one or two a day.
We are not going to tell you never to have a beer. We are going to make sure that when you choose one, you are choosing it with the facts rather than without them. That is what the alcohol-free day target is for: not abstinence, just a number you set deliberately.
Nothing in our target list is filler. Here is the reasoning behind each one.
Walking after you eat is the target we would give almost anyone, and it is worth explaining properly because the reason is not the one people assume.
The established finding is about blood sugar. Light walking soon after a meal meaningfully reduces the rise in blood glucose and insulin compared with sitting through it, and it does not take much: a review of the short-bout literature found benefits from as little as two to five minutes, with standing helping a bit and walking helping considerably more. Longer walks do more again. For anyone with blood sugar on their radar, that alone is the case.
Now the honest part, because this page does honest parts. The direct evidence that post-meal walking causes fat loss is much thinner than the blood sugar evidence. Anyone telling you it burns fat in some special way because of when you do it is getting ahead of the data.
Its real value for fat loss is more ordinary and more useful. Meals are the most reliable cue in your day: they happen whether you planned them or not. Attach ten minutes of walking to each one and you have added half an hour of daily movement that never needed scheduling, never competed with anything, and never required motivation. And incidental movement is precisely what your body quietly withdraws when you start eating less, which is one of the main reasons fat loss stalls in weeks three and four. A target anchored to something that already happens three times a day is the most reliable way we know to stop that.
It is also the single easiest target to keep on a bad day, which matters more than it sounds in a system where the only rule is never zero.
10,000 steps was never a research finding. It was a 1960s Japanese pedometer slogan that stuck because it is a round number. The research has since been done, and the number it points to is lower, which is good news for anyone with a job.
The largest review to date pooled 57 studies and more than 160,000 adults and mapped health outcomes against daily step counts. Compared with people walking about 2,000 steps a day, people at around 7,000 steps had a 47% lower risk of dying early, along with 25% less cardiovascular disease, 38% less dementia, 22% less depression, 14% less type 2 diabetes and 28% fewer falls. Beyond 7,000 the heart benefit kept climbing, but for almost every other outcome the curve flattened: more steps were not worse, they just stopped buying much. An earlier pooling of 15 cohorts found the same shape, with the benefit levelling off around 6,000 to 8,000 steps in older adults and 8,000 to 10,000 in younger ones. The honest caveats: these are observational studies, several outcomes rest on only a few of them, and not every study adjusted for age or frailty. The mortality finding is the strongest of the lot and it is consistent across the literature.
So the floor in the app is 7,000. It is roughly an hour of ordinary walking spread through a day, and it is reachable on a desk-bound Tuesday with a walk to the car, a lap of the office and ten minutes after dinner. Treat it as the line you do not go under. Let the good days go higher, and let your Walk after meals target do most of the work of getting you there.
Almost everything on this page fades when the program stops. That is the honest state of behaviour change research and we say so repeatedly. Breaking up sitting is the exception, and it is worth understanding why.
An umbrella review of workplace sedentary-behaviour interventions found sitting time reduced by around 69 minutes a day at three months, 78 at six months, and still 62 at twelve months. Almost nothing else in this literature holds like that. The reason is not that people tried harder: education-only programs were ineffective. It is that the interventions that worked changed the environment, and environment beats willpower over a year every time.
The physiology is separate from your training, which is the part people get wrong. Long unbroken sitting reduces how well your muscles pull glucose and fat out of your blood, and that happens independently of how fit you are. You can train hard at 6am and still spend the day in a metabolic state you would not choose. A short walk every half hour meaningfully changes that, and it does not need to be a workout.
If your team is doing this at work, it is also the target most likely to be noticed by the people around you, which is not nothing.
Fewer than one Australian adult in ten meets the recommended fibre intake. Almost everyone reading this is one of the nine.
The evidence here is unusually consistent for nutrition, which is a field where consistency is rare. Higher fibre intake tracks with lower rates of heart disease, type 2 diabetes and bowel cancer, and the relationship keeps improving up to about 30 g a day before it flattens off. For the climb specifically it does two jobs that make everything else easier: it slows digestion and blunts the blood sugar swing that produces the 3pm hunger that wrecks a good day, and high-fibre food carries more water and bulk per calorie, so less food genuinely feels like enough.
One practical warning, because it is the reason people abandon this target in week one: go up gradually and drink your water. Doubling your fibre overnight is uncomfortable and entirely avoidable.
Five servings of vegetables and two of fruit is the Australian guideline. Under one adult in ten manages both. The pooled evidence points to roughly five servings a day total for the biggest reduction in cardiovascular and all-cause mortality, with little further gain beyond that, so this is a target with a real ceiling rather than a more-is-always-better one.
We include it because it does the same job as the whole-food rule in a form you can actually count while standing in a supermarket, and because it drags fibre, micronutrients and fullness up at the same time. Frozen counts. Tinned counts. The perfect version of this target is not the point.
This is a different target from your daily protein number, and it is the one to add once that number is already solid.
Let's be clear on the order of importance. Your daily total comes first. Hit that and you have done most of the work, and no amount of clever timing rescues a day that came up short. But once the total is solid, how you spread it across the day is where a lot of the gold is. Muscle protein synthesis is switched on feed by feed, and each feed needs enough protein to flip the switch: roughly 0.3 to 0.4 g per kg of body weight, which is about 25 to 40 g for most people. Four feeds that each clear that bar do more for holding onto muscle than the same total loaded into one enormous dinner. If you are in a deficit, that difference is the difference between losing fat and losing the work you did in the gym.
"Every feed" is deliberate wording. Snacks are eating occasions too, and a protein-free snack is a missed opportunity twice over: once for muscle, and once because protein is the macro that actually keeps you full, so the snack you just ate will not hold you.
This is the most commonly broken sleep rule, and the one people are least aware of breaking.
In a controlled trial, 400 mg of caffeine taken six hours before bed reduced total sleep by around an hour compared with placebo. The detail that matters: the people it happened to could not tell. Their own ratings of how they slept showed no difference. That is the entire problem with caffeine and sleep. It does not feel like it is doing anything. It just quietly removes an hour, and you conclude you are someone who does not need much sleep.
Caffeine has a half-life of roughly five hours, so a 3pm coffee still has about a quarter of its dose circulating at midnight. Eight hours before bed is a comfortable cut-off. Six is the minimum the evidence supports. Individual clearance varies a lot, mostly genetically, so the honest version is: if your sleep target keeps missing, test this before you test anything else. It is free, and it is often the whole answer.
We put a social target on the list because leaving it off would have been dishonest about what the evidence says.
Social connection sits alongside the physical risk factors in the mortality literature, and loneliness carries a risk comparable to lifestyle risks we all take seriously. That finding has been replicated enough, and is surprising enough, that it belongs on a page about what actually works.
It also does a specific job inside a 42-day climb. The people who finish are reliably the ones who told someone they were doing it. Accountability is the mechanism behind the whole product, and it does not switch off outside the app.
One proper conversation a day. Face to face or on the phone. Not a text, not a meeting, not a comment on a post. Five minutes counts if it is real.
The usual version of this target is a moral one: processed food is bad, real food is good. We are going to give you the version that is actually supported, because it is more useful and it survives scrutiny.
The flagship study is unusually clean. Twenty adults lived in a research ward at the US National Institutes of Health for a month, eating either an ultra-processed or an unprocessed diet for two weeks each, in random order. The two menus were matched for the calories, sugar, fat, salt and fibre put in front of them. People were told to eat as much or as little as they liked. On the ultra-processed diet they ate about 500 calories a day more, gained around 0.9 kg in a fortnight, and lost about the same on the unprocessed one. Nobody instructed them to change anything.
The mechanism is the interesting part, and it is not "chemicals". Ultra-processed food tends to be softer and denser in calories, so it goes down faster, and you have eaten more before your body notices. A later experiment that changed only the texture of an otherwise identical ultra-processed diet moved intake by 369 calories a day, which points hard at eating rate and energy density rather than processing as a category.
Now the inconvenient bit, because this page does those. Randomised cooking programs have not been shown to improve weight, blood pressure or cholesterol. What they reliably improve is diet quality: what ends up on the plate. So we sell this target on what it does. Cooking is simply the most practical way to end up with food that is bulkier per calorie and slower to eat, and that is the lever the trials actually identified.
This is the best-established dietary cause of weight gain there is, and the easiest thing on this whole page to fix. Pooling 61 cohort studies and 24 randomised trials, each extra daily serving of sugary drink tracks with about 0.42 kg more weight gain a year. In the trials, adding sugary drinks caused weight gain and removing them caused weight loss. Evidence that runs in both directions like that is genuinely rare in nutrition, and it is why this target is worded as plainly as it is: cut the sugary drinks.
What to drink instead is where the reporting goes wrong. Swapping to any non-caloric drink lowered BMI by about 0.31 kg/m² across six randomised trials, roughly a kilo, for as long as people kept it up. Water and diet drinks both work. And in the trials that compared them head to head, the diet drinks came out slightly ahead, not behind: a 52-week trial found 7.5 kg lost on non-nutritive sweetened drinks against 6.1 kg on water, a difference the authors themselves called statistically real but not clinically meaningful.
You will have seen "WHO says diet drinks do not help you lose weight." That is not what happened. WHO issued a conditional recommendation based on low-certainty evidence, and its own review found the randomised trials showed weight going down on sweeteners. Its concern was whether that persists long-term, drawn from observational data that WHO itself flagged as vulnerable to reverse causation. The guideline explicitly states it is not a safety assessment.
Our position, then: cutting sugary drinks is unambiguous and you should do it. Water is excellent. And if a diet drink is the thing that actually gets you off the sugar, that is precisely what the trials support, whatever the headline said.
We bundled these into one target because they take one sitting. They do not carry equal weight, so here they are in order of how well they hold up. The first two are genuinely strong. The third is not, and we will say so.
Goal setting is one of the best-evidenced things in this whole field. A meta-analysis of 141 randomised trials covering 384 comparisons and 16,523 people found that adding a goal to an intervention improved behaviour change at d = 0.34 over and above a matched comparison. For health behaviours specifically it was d = 0.44. In physical activity, goal-setting programmes reached d = 0.55 across 45 studies and 5,912 people. That is not folklore, it is one of the most replicated findings in behavioural science.
The structure matters, and this is the part worth understanding. An outcome goal on its own is the weakest configuration there is. In a meta-analysis of goal setting in sport, a long-term goal set by itself had no measurable effect at all (d = −0.08), while short-term goals worked (d = 0.43) and short-term goals alongside a long-term one worked equally well (d = 0.47). Wanting the thing is not the mechanism. The smaller goals underneath it are.
That is the entire architecture of this program stated in one row of numbers. You set an outcome goal because it tells you which direction to walk. Then you are scored on process goals: the daily targets you actually control. It is why your compliance percentage is the only number we ask you for, and why the leaderboard ranks showing up rather than fitness.
Two caveats we owe you. The process-goal estimate comes from a small set of mostly laboratory sport studies, and the outcome-goal figure rests on only two of them, so treat the size of that gap as indicative rather than precise. And in the broader health-behaviour literature, outcome and behaviour goals did not differ at all across 384 comparisons. The direction is well supported. The exact magnitude is not.
Breaking a target into progressively harder steps has its own name in behaviour change research: graded tasks, defined as setting easy tasks and making them increasingly difficult but achievable until the behaviour sticks. A 2026 meta-analysis of 39 randomised trials found a medium effect on physical activity, g = 0.61, and found something better than a headline: the number of steps in the progression predicted how much activity increased. More rungs on the ladder, more change.
There is a real trial in adults that shows the shape of it. Ninety-six inactive adults were given either a fixed "10,000 steps a day" target or a goal that stepped up from their own recent performance. The fixed goal won the first fortnight. Then it decayed, and by day 110 the progressive group was walking 1,030 more steps a day. A stepped goal loses the sprint and wins the marathon, which is the entire argument for five climbs a year instead of one heroic one.
The honest limits, because they matter: the graded-tasks authors rate their own evidence as low certainty, and the stepping was always delivered alongside goal setting and planning, so what is demonstrated is the package. The same technique has not been shown to help weight loss specifically. And more breakdown is not always better: in one trial, adults given a daily calorie goal lost more than those who also set individual meal and snack sub-goals, and the finer-grained group reported feeling more deprived and dropped out more often.
Two claims you will hear everywhere that the evidence does not support. "Short-term goals beat long-term goals." Two meta-analyses tested exactly that and found no difference. And "breaking your big goal into small ones is proven better than having one big goal." Nobody has shown it. What has been shown is narrower and still useful: a distant goal on its own is the weakest setup, and stepped, achievable targets are among the best-evidenced tools we have for physical activity.
Second in the journal: planning tomorrow tonight. Deciding in advance when, where and how you will do something adds a further small-to-medium effect, around SMD = 0.35 across 35 randomised trials, and it persists after the intervention stops. Planning for what will get in the way matters as much as planning the action: in cardiac rehabilitation, adding coping plans to action plans lifted the share of patients meeting their exercise target from about 43% to 71%.
Third, and weakest: gratitude. Gratitude writing produces a small lift in wellbeing, around g = 0.19. That roughly halves when it is compared against another writing exercise rather than against doing nothing, and against a genuinely active comparison it is indistinguishable from zero. For anxiety it is null in both major meta-analyses. We keep it because a minute costs nothing and some people find it changes the shape of their day. We are not going to call it a performance intervention, because it is not one.
"A Harvard study found the 3% who wrote their goals down earned ten times more." There is no such study. Harvard Library itself describes it as an urban legend. The same tale is told about Yale in 1953. It has been traced repeatedly and it does not exist.
"You are 42% more likely to achieve a goal if you write it down." That figure appears in no peer-reviewed source. It is attached to an unpublished conference presentation that never reported it. We went looking specifically, because it is quoted everywhere.
Goal setting is strong. It never needed the folklore, and quoting the folklore is how you end up being unable to defend the real thing.
We do not sell any of these. Two earn their place, two are here so that what you already take gets counted honestly, and one is on the list mostly so we can tell you the truth about it. Ranked and explained below.
These three carry the thinnest evidence in the whole catalogue, so this section works harder than any other on this page. The genuine benefits are real and we lead with them in the app: measured soreness and recovery effects, a fitness finding for sauna paired with training, a sleep signal for cold water. This page is where we keep the full working: the doses, the trial numbers, the caveats, and the claims the industry makes that do not survive checking. Every claim below is either something we can point at a trial for, or something we are telling you is marketing.
The most useful thing we can give you is the dose, because this is the one area where the gap between what the trials used and what people actually do is enormous. Here it is first, then the reasoning.
Every sauna article you have ever read leans on the same Finnish cohort: 2,315 men followed for 20 years, where the men using a sauna four to seven times a week had markedly lower cardiovascular and all-cause mortality than the men using one once a week. It is a real study and the numbers are striking. Three things about it matter here.
First, it is observational, not a trial. People too unwell to sit in a 90-degree room stop going, so illness reduces sauna use rather than the other way around, and that criticism has been put to the authors formally in print. Second, the same cohort produced a result that nobody quotes: frequent sauna use in men with low fitness did not reach significance. Fitness did the heavy lifting. Third, and decisively for us, those saunas averaged 79°C, and the authors wrote in their own limitations that the results cannot be applied to lower-temperature setups. Infrared cabins run at 45 to 60°C. The people who did the research told us not to make this leap, and the industry made it anyway.
So what does infrared itself have behind it? A much smaller pile. A systematic review of 40 clinical sauna studies found only 13 randomised trials among them, nine of those with fewer than 50 people, and only three at low risk of bias. Its own conclusion on dosing is the sentence we keep coming back to: further study is needed to determine the optimal frequency and duration. As of the most comprehensive review available, nobody knows the optimal infrared dose. Anyone telling you 20 minutes is the magic number, including us, is quoting a convention.
And when passive heating has been pooled properly, it mostly does not do what is claimed. A 2025 meta-analysis of 20 randomised trials found no significant pooled effect on flow-mediated dilation, pulse wave velocity, resting heart rate, heart rate variability, fasting glucose, HbA1c, any cholesterol fraction, triglycerides or C-reactive protein. Systolic blood pressure came out at −2.46 mmHg, not significant. A separate randomised trial in adults with untreated hypertension found neither heat therapy nor exercise training shifted 24-hour blood pressure at all.
What determines the physiological effect of a sauna is total thermal load: how hot the cabin is, multiplied by how long you sit in it. Not the fact that the heat arrives as infrared. Three recent studies nail this down. Researchers put a probe into the quadriceps during 45 minutes in a commercial infrared cabin: muscle temperature rose about 3°C just under the skin and about 1°C at 3.4 cm, but core body temperature did not change at all. Run the same cabin at 65°C for 45 minutes and core temperature rises 1.4°C. And in a head-to-head of three heating methods in the same 20 people, far infrared produced the smallest response of the three, behind both a traditional sauna and a hot bath.
That is why our target says 55 to 65 degrees rather than just "have a sauna". If you sit in a 45-degree cabin for 20 minutes, you have had a pleasant sit down. Nothing wrong with that. Just do not count it as a physiological intervention.
What it does do, honestly stated. You get hot and you sweat. Stiffness and soreness ease while you are in there and for a while after: in the one uncontrolled pilot people cite for arthritis, pain and stiffness dropped 40 to 60% within a session, and then over four weeks the improvement did not reach significance. For training soreness, a network meta-analysis found sauna beat placebo at 48 hours and at no other timepoint. A randomised crossover in 16 basketballers found less soreness, better perceived recovery and a smaller drop in jump height after 20 minutes of infrared at 43°C. A single session also acutely lowers blood pressure and arterial stiffness for around half an hour afterwards, measured in 102 adults at 73°C.
The fitness finding, and it is worth knowing properly. The best sauna results come from pairing it with training rather than using it instead of training. In a randomised trial of 47 low-active adults, adding a 15 minute sauna after each workout for eight weeks improved cardiorespiratory fitness by 2.7 mL/kg/min and lowered systolic blood pressure by 8 mmHg beyond what the identical exercise program achieved alone. In competitive runners, three weeks of ~30 minute post-training saunas expanded blood plasma volume by 7.1% and improved a treadmill endurance test, the same adaptation endurance athletes chase with heat and altitude blocks. The caveats we owe you: the fitness trial used a traditional sauna stepped from 65 to 80 degrees, the runners study had six participants at 90 degrees, and nobody has run these protocols in an infrared cabin. Which is exactly why our dose advice is to run the cabin warm and treat heat times time as the dose: the mechanism these trials rely on is a genuine rise in body temperature, not the badge on the heater.
Heat shock proteins: real biology, honest dose. Heat shock proteins are a genuine cellular response to heat stress and one of the mechanisms behind heat adaptation. In humans they rise when the thermal dose is substantial: protocols that raised core temperature to roughly 38.5 to 39.5 degrees for about an hour, or repeated daily heat acclimation, produced measurable HSP70 elevation, with a pooled effect of g = 0.97 across 12 studies. An hour in 39 degree water was not enough to shift intracellular HSP72. Two honest caveats. First, the outcome link is unfinished science: in the trial where pre-exercise heating did raise HSPs, muscle damage and endurance were unchanged, and in two trials where heat improved blood sugar control, HSPs did not move. Second, the dose problem again: an infrared cabin run cool does not raise core temperature at all, so a meaningful HSP response is another argument for running it hot rather than a benefit of infrared as such.
Sleep: the mechanism is real, the sauna evidence is not there yet. Warm water 1 to 2 hours before bed reliably helps people fall asleep faster across 13 pooled studies, and the mechanism is the rebound: heat opens blood flow to hands and feet, core temperature then falls faster, and that fall is a sleep signal. A sauna that genuinely heats you should be able to ride the same mechanism, but the direct trials have not shown it yet: the basketball study found no difference in night heart rate, HRV, sleep duration or sleep quality after evening infrared sessions, and the survey data linking sauna users to better sleep did not control for the fact that sauna users were younger and healthier to begin with. So we say: plausible, backed by the adjacent warm-bath evidence, unproven for saunas specifically.
Full spectrum and red light, for cabins that have it. Far infrared is absorbed in the first millimetre of skin as heat. Red and near-infrared light, roughly 600 to 1000 nanometres, is a different modality with its own evidence: in the delayed-soreness network meta-analysis, red light therapy was the single best-performing modality at 24 and 48 hours, ahead of everything including cold and heat. A full spectrum cabin emits in that range; a far-only cabin emits none of it. The honest limit is dose: the positive light-therapy trials pressed the device against the skin for measured doses at each site, nobody has published the light dose actually reaching a person sitting in a cabin, and no trial has separated a sauna's light from its heat. Treat the red light end as a plausible bonus riding on top of the heat, not a proven dose.
This is the cleanest split in any of the evidence on this page, and once you see it you cannot unsee it. The key meta-analysis pooled 17 studies of pneumatic compression boots. Look at what moved and what did not, in the same studies, on the same people.
Perceived soreness moves. Jump height, strength and sprint do not, at any timepoint out to 72 hours, and not one of those estimates reached significance. Creatine kinase, the standard muscle damage marker, did not move either. That pattern is what an expectancy effect looks like. It does not prove the effect is placebo, but no study design in this field can rule it out, because you cannot hide from somebody that their legs are being squeezed.
The trial that tried hardest is worth knowing about. Researchers gave elite youth footballers either 30 minutes of boots at 200 mmHg after matches, or an unlabelled cream described to them as a novel recovery product. Nothing differed between the groups on anything: jump height, low-frequency fatigue, creatine kinase, perceived fatigue, soreness or sleep. Everybody recovered. The boots changed nothing, including the subjective measures. A separate randomised trial using boots at 0, 24 and 48 hours after 100 drop jumps found no differences in damage markers or function either.
On dose: the meta-analysis authors' own practical statement is that a single bout of 20 to 30 minutes at about 80 mmHg may be sufficient, which is the dose in our target. Worth knowing that a boot displaying "80" is not delivering 80 mmHg to your muscle, so the number on the machine is a setting rather than a measurement. Worth knowing too that when researchers compared pressures directly, 30 mmHg produced a 90% increase in peak venous flow while 50 mmHg produced only 44%, with signs of the vein being squashed shut in five of twelve people. Turning it up is not a bonus.
The fluid story, which is the best-measured part. This is the mechanism worth talking about, told straight. In 59 healthy adults, a 30 minute boot session at 90 mmHg measurably lowered fluid content inside the leg muscles on MRI, took 6 mm off calf girth, and raised muscle oxygen saturation from about 76% to 81%. So "your legs walk out carrying less fluid" is a claim with imaging behind it, with two caveats we will not hide: the study had no control group and no prior exercise, so lying down for half an hour did some unknown share of the work, and it was funded by a boot manufacturer. Compression garments after damaging exercise also reduce swelling, at a pooled SMD of −0.73. What we will not say is "lymphatic drainage". The one imaging study of lymphatics under pneumatic compression that included healthy legs found no measurable change in lymphatic transport in them, and a separate measurement study found the pressure that shuts superficial lymphatic vessels averages 86 mmHg, which is at or below what most boots run. Nor will we claim the fluid clearance is why legs feel better: in the one trial that separated them, compression alone reduced swelling but did not significantly reduce soreness. Less fluid is measured. Feeling lighter is reported. The line between them is not yet drawn. And on knee joint fluid specifically: synovial fluid sits inside a sealed capsule, and no study has shown boots move it, so that one stays off the page.
So why is it a target at all? Because feeling better is worth something real and we are not going to pretend otherwise. Adherence is what produces a result in 42 days, and legs that feel fresh tomorrow get you back in the gym tomorrow. We just will not dress that up as physiology.
Cold water is the best-studied of the three and it is the one where being specific pays off, because the dose-response curve has a peak and then it turns and goes negative. From a meta-analysis of nine trials, on a ten-point soreness scale:
| Protocol | What happened to soreness |
|---|---|
| 11 to 15°C | +0.27 immediate, +0.32 delayed. Significant. This is the window. |
| 5 to 10°C | +0.14 immediate, +0.06 delayed. Not significant. Colder is not better. |
| 11 to 15 min | +0.23 immediate, +0.32 delayed. Significant. This is the window. |
| 5 to 10 min | +0.65 immediate, but not significant. Too few trials, too much spread. |
| 16 to 20 min | −2.20 delayed. Significantly worse than doing nothing at all. |
Two network meta-analyses since, covering 55 and 87 randomised trials, independently landed on the same 10 to 15 minute window. This is about as replicated as recovery research gets. Note the size of the win though: 0.29 points out of ten. That is below what is normally considered a noticeable change in pain. The direction is well established. The magnitude is small. One honest wrinkle: a separate 52-study analysis found that for blood markers, slightly shorter and cooler worked best after high-intensity exercise, so the edges of the window are softer than the middle. Inside 10 to 15 minutes at 10 to 15 degrees, you are inside every analysis at once.
Where it genuinely earns its keep: after high-intensity work. Split the trials by the kind of session that preceded the plunge and the picture sharpens. After high-intensity exercise, the kind that ends an ESD class, cold water at 24 hours produced less soreness (SMD −0.89), better perceived recovery (SMD 0.66), lower creatine kinase, the standard blood marker of muscle damage (SMD −0.85), and even a small lift in muscular power (SMD 0.22). Those are the strongest recovery numbers on this page for any of the three modalities, and they are the reason the app copy tells you to save the plunge for your hardest sessions. Strength at 24 hours did not move, and after plain eccentric lifting the soreness effect was absent, which is the same message as the interference data below wearing different clothes: this is a conditioning recovery tool, not a lifting one.
The sleep finding. In twelve trained runners, a 10 minute evening immersion at 13 degrees after training led to fewer wake-ups across the whole night than no immersion, measured by sleep lab equipment, and the group immersed to the neck also spent more of the early night in deep slow-wave sleep. Small study, one night at a time, and the deep-sleep comparison involved head-under immersion which we do not allow, but the direction is encouraging and nothing in the trial pointed the other way: we found no study showing an evening plunge harms sleep.
Dose is per session, and nobody has tested weekly frequency. Every trial behind the 11-to-15 numbers used a single immersion after a single training bout, so the dose you are ticking off is one plunge, not a weekly tally. How many per week is best has simply never been compared in a trial: the chronic studies immersed after every session because that is easier to research, and that is precisely the pattern that produced the muscle-growth interference below. Our default of three a week is judgement, not data, and we say so. Put them after your hardest sessions and you are using them the way the evidence was collected.
If you are new to cold water, do not start at 11 minutes. The dose trials used active, mostly trained people. The gasp-and-hyperventilate response to cold entry is the actual risk moment, and it habituates fast: across 17 study groups, most of the cold shock response was gone after about four exposures, with heart rate response down 14 beats and breathing response roughly halved, and short 3 minute immersions were enough to produce that adaptation. So start at 2 to 3 minutes, stay calm on entry, and build toward the full dose across your first couple of weeks. Rushing the first minute is how cold water hurts people; the eleventh minute has never been the problem.
Cold water immersion straight after resistance training blunts muscle growth. This is not a fringe finding. In the landmark trial, 21 men trained twice a week for 12 weeks, taking either 10 minutes at 10°C or an easy cycle after every session. The active recovery group gained 17% in type II fibre cross-sectional area and 26% more myonuclei per fibre. The cold group's gains did not reach significance. In a separate design where each person cooled one leg and not the other, muscle protein synthesis ran roughly 20 to 30% lower in the cooled leg, with the size of the drop tracking how much the microvascular blood volume fell. That is a mechanism, measured within the same person, which is about as clean as human physiology gets.
The nuance, because it cuts both ways. Strength is much less affected than size. One seven-week trial found blunted fibre growth alongside completely intact 1RM gains, and a meta-analysis found the strength decrement was driven entirely by single-limb studies, with whole-body immersion showing no significant effect. And endurance training is not affected at all: six weeks of sprint interval training with cold immersion after every session produced the same improvements in VO2 max, peak power and 2 km time trial as training alone.
One thing we will not do is give you a waiting time. You will read "wait four to six hours" all over the internet. There is no trial behind it. Every study applied the cold within about 15 minutes of finishing, and the researchers say plainly that their conclusions only apply to that window. Nobody has tested whether waiting fixes it. So the honest advice is separation, not a stopwatch: if muscle size is what you are after, keep the plunge off your lifting days.
The irony worth sitting with. The reason people take an ice bath is to reduce inflammation. Reducing inflammation after training is precisely the mechanism by which it blunts adaptation. Those two claims cannot both be selling points. And it may not even be doing the first one: in a study that took actual muscle biopsies, cold water was no better than an easy warm-down at reducing inflammation inside the muscle. Pooled across 11 studies, systemic inflammatory markers went up immediately after immersion, not down.
On mood and mental health, which is where a lot of the current enthusiasm lives: the honest position is that there is not much there yet. The famous noradrenaline figure, up 530%, comes from sitting in 14-degree water for a full hour, and that study measured blood chemistry only. It recorded no mood, alertness or cognitive outcome of any kind. In an earlier study using the same protocol, dopamine did not rise significantly. The fMRI study people cite had no control condition. The depression study had thirteen patients recruited, five who actually attended, no control group and no statistical test. Pooled across trials, cold water improved sleep quality and quality of life but not mood. We would love this to be stronger. It is not yet.
Safety, and we mean this one. The danger with cold water is not hypothermia, it is the cold shock response in the first 30 seconds: the gasp, the drop in blood CO2, the autonomic strain, all of which happen long before your core temperature moves. The Australian position statement from AUSactive, Royal Life Saving and SPASA specifies 10 to 15°C, head and neck out of the water at all times, health screening before you start, supervision, and access to rescue equipment and a defibrillator. Do not go in alone. Skip it entirely if you have a heart or circulatory condition, uncontrolled high blood pressure, or you are pregnant. And if you get hives from cold, this is not a quirk to push through: among people with cold urticaria, 21.5% have had anaphylaxis, and whole-body immersion is the commonest trigger there is.
"Sweating detoxes you." Every study in this area measures the concentration of something in a fluid. Not one measures a health outcome. When researchers compared sweating from exercise against sweating in a sauna, four of the five metals tested came out higher from exercise. The most common industrial chemicals in human blood are not excreted in sweat at all. And sweat is not selective: you lose calcium, magnesium, zinc and iron in it too.
"A sauna burns 600 calories." We could not find a single published measurement of energy expenditure in an infrared sauna by proper calorimetry. The numbers people quote come from heart rate monitors, and heart rate estimates calories by assuming you are exercising. In passive heat your heart rate rises because your skin blood vessels open, without the oxygen use that would justify it. Your watch is guessing, and it is guessing high. The weight you lose in there is water.
"Infrared is like doing cardio." Tested directly in an Australian randomised crossover comparing infrared sauna, cycling and sitting still. No differences in blood pressure, arterial stiffness or heart rate variability between any of them. Separately, eight weeks of infrared sauna three times a week at 60°C produced no gain in muscle size and no gain in leg strength.
"Infrared penetrates 3 to 4 cm." That figure traces to a discussion section of a study that did not measure penetration depth. When someone finally did put a probe in, the thermal effect had fallen 63% by 2.4 cm and was negligible past 3.8 cm, and the warming is best explained by ordinary conduction from hot skin, the same as any hot room.
"Compression flushes lactic acid." Lactate was ruled out as the cause of delayed soreness in 1983. It clears on its own within about an hour. And compression has been measured to raise blood lactate, not lower it.
"Boots promote lymphatic drainage." Borrowed from lymphoedema medicine, where it is real. In healthy legs, the one lymphatic imaging study under pneumatic compression found no measurable change in lymphatic transport, and the pressure that closes superficial lymphatic vessels averages 86 mmHg, at or below what most boots run. The fluid boots move is measurable. Calling it lymphatic is not.
The least flattering finding in rehabilitation research is about the patient, not the treatment. In the studies that have measured it, 50 to 65% of people do not do the home exercises their physio prescribed, and drop-out from physio-led exercise programmes runs as high as 70%. The people who do the work recover better. That is not a surprising sentence, but it is the whole reason this is a target with a number on it rather than a line in your notes.
It matters because the exercises are usually the best-evidenced treatment on offer. For tendons, progressive loading beats rest and beats most of the passive treatments it gets compared with, and the Achilles and patellar loading protocols are among the most replicated interventions in sports medicine. For backs, knees, hips and shoulders the pattern is the same: graded exercise is first-line, and the passive stuff is an add-on at best. The limiting factor is almost never the prescription. It is whether the prescription got done.
Ten minutes is deliberately small. Adherence falls as the programme gets longer, and the programme you finish beats the programme you were meant to do. Set the days per week you can actually own, log it, and let the scoreboard do what your good intentions have not. One rule: your physio's plan outranks your training program for the injured bit. Train as hard as you like around it. Not through it. If you have not seen a physio yet, booking one is the first tick.
We do not sell supplements and we are not going to start, so this section has no reason to be kind. Two of these five have real evidence behind them. Two are on the list because a lot of people already take them and the honest thing is to count them properly and tell you what they do and do not do. One is here mostly so we can say the quiet part out loud. None of them outranks the targets above. If your sleep, protein, fibre, steps and weights are not handled, no capsule is where your money should go.
Creatine monohydrate is the most studied supplement in sport, and one of the very few that does what the tub says. It raises the phosphocreatine your muscles use for short, hard efforts, and across hundreds of trials it adds a little strength, a little lean mass and a few extra reps at the end of hard sets. It is safe in healthy people in every long-term trial that has looked, including the kidney markers people worry about. The dose is 3 to 5 g a day, every day. Timing is irrelevant, it does not need loading, and it takes three to four weeks to fill your stores at that dose, so start on day one rather than day twenty. Expect around a kilo of water weight in the first week. That is your muscles filling up, not fat, and it is why we tell you before you weigh in.
On the brain, the picture is real but smaller than the podcasts. Pooling sixteen randomised trials, creatine improved memory by a small amount (SMD 0.31) and attention and processing speed by a similar margin, at doses from 3 to 20 g a day, with the effect larger in people who eat little meat and in older adults. Against that, a six-week trial that gave healthy young adults 10 or 20 g a day found no cognitive improvement at all, and the largest single trial at 5 g a day found an effect so small it did not reach significance. So: a modest, plausible benefit that some people will notice and many will not.
The 30 g figure you may have heard comes from one study. Fifteen people took a single dose of 0.35 g per kg, about 25 to 30 g for most adults, and then stayed awake for 21 hours. Their memory and processing speed held up better than on placebo, peaking about four hours after the dose. That is a genuinely interesting result for a sleepless night, a night shift or a long drive. It is not a daily dose, nobody has trialled 30 g daily for cognition, and at that amount most people will spend part of the evening in the bathroom. Our recommendation: 5 g daily for everyone, up to 10 g if you are big or you want to test the brain side, and treat the big single dose as an occasional tool rather than a habit.
The "anabolic window" was a good idea that did not survive measurement. When the trials on protein timing were pooled and total daily protein was accounted for, the timing effect disappeared: total protein explained the results, the clock did not. A direct test that gave the same protein either immediately before or immediately after lifting found the same muscle and strength gains either way, and the most recent meta-analysis of that comparison (2025) concluded that timing "does not importantly modify" the change in lean mass. Your daily total comes first, spreading it across the day comes second, and where the session sits in that spread comes a distant third.
So why is it a target? Because the session is the feed people miss. Train at 6am and breakfast becomes a coffee; train after work and dinner lands at 9pm with nothing since lunch. A 25 to 40 g protein feed within a couple of hours either side of training is not magic, it is the simplest way to make a training day a well-distributed protein day. It works alongside your Protein at every feed target. If you have already hit your total and your feeds, you have already done this one.
Carbohydrate timing has a narrower, more honest brief. It matters in three situations: hard sessions longer than about 90 minutes, two sessions closer than about eight hours apart, and any session you need to perform in rather than just complete. In that kind of work, carbs in the hours before improve performance by 7 to 20% in the trials, and a carb feed in the first hour or two after speeds up refilling muscle glycogen for the next one. Outside those situations it does very little: a 45 minute lift or an ESD class followed by a full day's recovery refills on its own within 24 hours if your daily carbs are adequate, and there is no evidence that post-workout carbs build more muscle. Size the carbs to the session. On a fat loss climb, a pre-workout carb feed is still calories, so move them from elsewhere in the day rather than adding them on top.
This is the supplement we would most like to be better than it is. The heart claims did not survive the big trials. The Cochrane review pooled 86 randomised trials and more than 160,000 people and found that EPA and DHA capsules made little or no difference to death or to cardiovascular events, with high certainty, alongside a small possible reduction in coronary events (roughly one event prevented per 167 people treated) and a reliable 15% drop in triglycerides. The muscle claims fared no better: the most recent overview of reviews found no convincing evidence that omega-3 supplements improve lean mass, strength or physical function in healthy adults or clinical populations. A small reduction in soreness after hard sessions shows up in some trials, and that is about the extent of the sports case.
There is also a cost that rarely makes the label. Across seven cardiovascular outcome trials, omega-3 supplementation raised the risk of atrial fibrillation by 25% overall, and by 49% at doses above 1 g a day, with the risk rising roughly 10% per extra gram. That is why the target carries a ceiling, not just a floor.
What survives is simple. Eating oily fish two or three times a week is the version with the best evidence, and it brings protein and a meal with it. If you supplement, 1 to 2 g of EPA plus DHA a day is the range the trials used, and nothing above that unless a doctor put you on it for triglycerides. We include it because plenty of people already take it and it should be counted honestly, not because it will move your numbers.
Greens powders have no independent evidence of any health, performance or longevity outcome. The published trials are funded by the manufacturers, and they show what you would expect a multivitamin dissolved in water to show: blood levels of the vitamins in the tub go up after you drink it, and the gut microbiome shifts a little over a few weeks. No trial has measured whether anyone got fitter, leaner, healthier or lived longer, and a 13 g scoop cannot replace the fibre or the volume of actual vegetables, which is where the evidence for "greens" really lives.
The one thing in this category with decent evidence is the boring one. A plain daily multivitamin produced a small, measurable slowing of memory decline across three trials in older adults, and in the largest trial it did nothing for heart disease or death. The rest of the longevity shelf, NMN, resveratrol, anti-ageing blends, has no human trial showing it extends anything. The supplements with the best longevity evidence are not supplements: they are resistance training, fibre, vegetables, sleep and steps, and they are all targets above this one.
So why is it a target at all? Because plenty of our members already take something, and if you are going to spend the money, the only thing that makes it worth anything is taking it every day. Treat it as insurance for the days your food falls short, not as the plan. If your fruit and veg target and your fibre target are already green, you can skip this one with a clear conscience.
This section used to carry the three figures every wellness pitch in Australia uses. We checked them, and two of the three do not survive checking, so they are gone. We are leaving the explanation here rather than quietly deleting it, because the point of this page is that we show our working.
What we will stand behind is the cost side, from the government body that collects the actual claims:
On return on investment, the honest answer is that nobody has a trustworthy Australian number. The most-quoted one ($2.30 per $1, PwC for beyondblue, 2014) is a model rather than a measurement, it covers mental health programmes rather than fitness, and the large randomised trials since have generally failed to find effects on spending, absenteeism or job performance. We would rather tell you that than sell you a multiple. Judge us on what happens in your own team over a year, which is a thing you can actually measure.
Sports science moves. Things we are confident about today will be refined, and a
few will be overturned outright. When that happens we change the programme and we change this page, rather
than quietly keeping a claim because it sounds good in marketing. If we get something wrong, we would rather
correct it in front of you.
That is the deal: no fluff, no fads, no supplements we happen to sell, no fat-burning zone nonsense. Just the
best evidence available and an honest account of where it is uncertain. We are loving and caring, tough but
fair, and we will always tell you the truth about what actually works, even when the truth is slower and less
exciting than the promise down the road.