Ten rules for judging a longevity protocol
Before any of the scoring below, there is a prior question: does an intervention deserve to be in here at all. These ten rules decide that, and they decide what gets recommended to you. They are not aspirational. Every one is already applied in specific entries, and each rule links to where you can check it.
Most of them are ways of resisting the same temptation: treating a promising mechanism as though it were a proven outcome.
Animal evidence is not human evidence
A result in mice is a reason to run a human trial, not a reason to act. Senolytics, acarbose and taurine all have striking animal data and thin human data. Each entry says so in the first paragraph rather than the last.
Doing something is not extending lifespan
A compound can measurably change a biomarker and change nothing you would notice. NAD+ precursors reliably raise NAD+ levels. Whether that improves an outcome is a separate question with a weaker answer, and we keep the two apart.
Association is not causation, and confounding is usually the explanation
Observational findings get labelled as observational. The metformin longevity claim has an obvious alternative explanation in confounding by indication. Coffee looked harmful for decades until studies adjusted properly for smoking.
A surrogate marker is not an endpoint
Some markers predict risk without being the thing causing it. Homocysteine is the clearest case: B-vitamins reliably lower it, and trials doing exactly that have not improved cognitive outcomes. We track it as a signal, not a target.
Formulation and dose can matter as much as the ingredient
REDUCE-IT is cited constantly as evidence for fish oil. It tested purified prescription EPA at a dose consumer products do not reach, in a high-risk population. The omega-3 entry separates the two rather than borrowing the trial’s credibility.
A test is only as good as the base rate it is applied to
A 99%-sensitive test for a rare condition still returns mostly false positives. The full-body MRI entry works the arithmetic through, because the maths is what determines whether screening helps you or starts a cascade.
One study is a hypothesis. Replication is evidence
Findings that survive independent replication get weighted accordingly. FOXO3 appears across many populations, so it appears here. Most candidate-gene results fail to replicate and are left out.
Effect size is judged against the fundamentals, not against zero
The right comparison for any intervention is what training, sleep, protein and body composition already deliver. A supplement that beats placebo can still be irrelevant next to the things it is competing with for your attention.
We say when something is not worth taking
A framework that only ever adds things is a shopping list. Vitamin E and beta-carotene above RDA levels are associated with increased mortality in a 78-trial Cochrane review, so the entry says avoid.
Where the evidence is missing, we say that too
The hydration entry states plainly that it does not map to a validated marker. Bone density is excluded from the risk categories because no tracked marker supports the inference. Gaps are stated rather than filled.
Applied honestly, these rules disqualify a great deal of what the longevity market sells, including several things that would be easy and profitable to recommend. That is the point of writing them down. Everything below is what survives them.
Each marker gets its own 0-100 score
Your blood pressure, VO2 Max, HbA1c, and every other individual marker is scored against published percentile data for your specific age and sex, not a single population-wide "normal" range. A 25-year-old and a 65-year-old aren't compared against the same VO2 Max bar, because that would either be meaninglessly easy for the younger person or unfairly punishing for the older one.
The scoring itself is a smooth ramp, not a hard cliff. Landing one point below a "Top 10%" threshold scores nearly as well as landing exactly on it, and there's no sudden drop-off the moment you cross a boundary in the wrong direction. This matters because real physiology doesn't work in hard cutoffs either, a HbA1c of 5.6% and 5.7% represent almost the same metabolic reality, even though one might sit just inside a labeled category and one just outside it.
Where percentile data doesn't exist at a given resolution, for example if published research groups all adults 60 and older into one bracket rather than breaking out 60-70 versus 70-80, the app uses the bracket that actually exists rather than inventing false precision. Want the full reference data behind a specific marker? Browse every biomarker in the encyclopedia →
Markers roll up into 6 weighted domains
Body Composition, Cardiovascular Health, Compound Strength, Functional Capacity, Metabolic Health, and Systemic Health. Each domain's score is a weighted average of its own member markers, not a simple average of the raw numbers. Cardiovascular Health, for example, blends blood pressure, resting heart rate, and sleep duration, weighted by how strong the published evidence is linking each one to cardiovascular outcomes specifically.
This weighting is deliberate: a marker with decades of large-cohort mortality data behind it (like resting heart rate) carries more weight in its domain than a marker with a thinner, more preliminary evidence base, even if both are directionally useful.
Cardiovascular Health
Largest shareAtherosclerosis and cardiovascular events remain the single largest driver of premature mortality, so this domain carries more of the 100 points than any other.
Metabolic Health
Insulin sensitivity and glucose tolerance sit upstream of vascular damage, cellular repair, and long-term disease risk. Requires basic fasting labs.
Body Composition
Visceral fat, lean tissue mass, and frame load. Waist-to-height is the single easiest marker to track honestly, needing only a tape measure.
Functional Capacity
Real-world movement quality, balance, endurance and cognitive speed. This is what determines physical independence and fall risk in later decades.
Compound Strength
Relative strength across the four major lifts, scored against your own bodyweight and age. The primary defense against sarcopenia and bone density loss.
Systemic Health
Smallest shareSilent inflammation, organ function and nutrient status. Weighted smallest not because it matters least, but because it needs the fullest lab panel to fill in.
Exact weights shift with your tracking mode, since the 100 points are redistributed across however many domains you currently have.
Domains roll up into your Longevity IQ
Every domain has a weight, and all domain weights always sum to 100 points, regardless of how many domains you're actively tracking. That weighted average, on the same 0-100 scale as every marker and domain underneath it, is your score before lifestyle factors are applied. Three of those are scored separately and can reduce it.
Which domains exist for you, and how many of the 100 points each one is worth, depends on your tracking mode.
Body Comp, Cardio Health, Strength, Function. No bloodwork needed to get started.
Everything in Basic, plus Metabolic Health, which requires basic fasting labs.
Everything in Advanced, plus Systemic Health, requiring a fuller lab panel.
Moving up a tier doesn't change the scoring formula for domains you already had, it adds more domains to the same 100-point total, so each existing domain's weight shifts slightly to make room for the new one. Your existing progress and history are never recalculated from scratch when you upgrade tracking modes.
Three lifestyle factors are scored separately
Smoking, drinking and a diabetes diagnosis do not fit the percentile model the rest of the score uses. Most people do not smoke at all, so there is no meaningful distribution to place you in. They are handled differently on purpose: each applies a percentage reduction after the domain average is calculated. It lands on your Lifespan score, not on Healthspan, because the research behind all three is mortality-focused.
Your domain scores are never touched by this. Cardiovascular Health stays a reading of your measured cardiovascular markers, so a measured number always means the same thing.
Keeping it on Lifespan is what makes the split useful. Someone fit who smokes sees a strong Healthspan number beside a materially lower Lifespan one, which is the gap a single blended score would hide.
| Status | Reduction |
|---|---|
| Never smoked | No effect |
| Quit 15+ years ago | About 2.5% |
| Quit 5 to 15 years ago | About 6% |
| Quit under 5 years ago | About 10% |
| Currently smoking | 17% to 30%, rising with cigarettes per day |
| Factor | Reduction |
|---|---|
| Two drinks a day or fewer | Little to none |
| Above two a day | Rising to 17% at six or more |
| Diagnosed diabetes | A flat 12% |
The drinking threshold is not arbitrary. It is where the 2017 Lancet Public Health analysis the app cites shows effects beginning to appear, rather than a round number chosen to look cautious.
Diabetes is set lower than a heavy smoking penalty deliberately. Some of that risk already appears in your HbA1c marker score, and this is meant to cover what a single blood test does not, namely cumulative diagnosis burden and kidney and cardiovascular risk. It is not there to charge you twice for the same thing.
Multiple factors compound multiplicatively rather than adding up, which matches how epidemiological risk actually behaves and means the total reduction can never reach 100% however many factors apply. Whenever a reduction is active it is shown on the score itself, as its own row in Risk Radar, and as a full detail card in the marker breakdown, styled like every other tracked marker. It is never an invisible penalty.
Healthspan and Lifespan are scored separately
Living longer and staying capable are two questions, and one blended number can hide a real gap between them. Alongside your overall score you get both, built from the same marker set and the same scoring model, filtered by what each marker’s own research actually predicts.
| Evidence base | Counts toward | Examples |
|---|---|---|
| A mortality or disease-risk study | Lifespan | Blood pressure, ApoB, HbA1c, waist-to-height ratio, fasting glucose |
| Functional capacity or quality of life | Healthspan | Grip strength, cognition, protein intake, step test recovery |
| Substantial evidence on both sides | Both | VO2 Max, compound strength, gait speed, single-leg balance |
The third row is the honest one. VO2 Max is a real measure of functional fitness and a validated mortality predictor in its own cited research, so forcing it into a single column would misrepresent what the evidence says. Markers like it count toward both scores rather than being made to pick.
Neither number replaces your overall score. They sit alongside it, and the point of showing all three is that you can see which direction is pulling the total. The difference between the two is worth understanding before reading either.
Confidence is about data coverage, not tracking mode
"High" confidence means every domain relevant to your current tracking mode actually has real data in it, not that you've unlocked the most domains available. Someone on Basic mode with all 4 of their domains filled in gets High confidence, full stop. Someone on Complete mode who's only entered data for 2 of 6 domains gets Low confidence, even though they've technically unlocked more categories than the Basic-mode user.
This distinction exists because a score built from thin, patchy data shouldn't look identical to one built from a complete picture, even if the raw number happens to land in the same place. Confidence is the app's way of telling you how much to trust the number in front of you.
"Do Nothing" and "Stay Consistent" are a separate calculation
Your current score reflects where you stand today, full stop, it is never adjusted based on projections. Elsewhere in the app, separate projection lines take that current score and project it forward under two scenarios: what happens if nothing changes, and what happens if you stay consistent with the recommended plan.
Cited research: for Strength, VO2 Max, and Blood Pressure specifically, the decline curve shown is modeled from named published research on age-related physiological change in untrained adults.
Our estimate: for every other tracked marker, the projection is Longevity Coach IQ's own modest, conservative estimate, clearly labeled as an estimate rather than dressed up as a citation it isn't. Both curve types use the same visual style on the chart, which is exactly why the app tells you directly which is which, rather than leaving you to guess.
A workout split that tells you which sessions to keep
Most plans fail for a predictable reason: miss one Wednesday session and the week is "broken," which starts a guilt cycle that ends in abandoning the plan entirely. Longevity Coach IQ runs six sessions and a rest day, and flags the sessions that target whichever domain is lagging. When strength or VO2 max is low, the sessions addressing it are marked high priority. A week where only half of them happen is still a week that went right, provided the right half happened.
- Monday: lift
- Tuesday: cardio
- Life happens, Wednesday and Thursday missed
- Week logged as failed. Streak reset.
- Strength is your weakest domain
- Both lifting days marked high priority
- Life happens, two sessions missed
- Both lifts done. The week did its job.
The plan itself is a sixteen-week periodized program. Block 1 (Tendon Resilience) uses four weeks of controlled-tempo work to build connective tissue durability before heavier loading begins, so the progression is ordered by what your joints can absorb, not by how fast you want to add weight.
Coaching ranked by biological urgency, not a generic checklist
You don't get 20 random wellness habits. Each domain's weakest measured marker is traced to the specific habit or training lever that actually moves it, and only the top few surface on your plan. If nothing in a domain needs attention, nothing from it gets recommended.
The same logic runs in reverse. A marker already sitting in the top decile doesn't generate a habit just to fill the list, it simply reports that nothing there needs attention right now.
Our three rules of scientific integrity
We hold Longevity Coach IQ to research standards rather than wellness-industry ones. In practice that comes down to three rules, and the third is the uncomfortable one.
Open logic, no secret sauce
Every formula, threshold and percentile band used to score you is published on this page and in the encyclopedia, with the source named. There is no proprietary index you have to take on faith.
Conservative projections, labeled
Where a future trajectory is backed by published clinical research, such as age-related strength decline, VO2 Max decline and blood pressure, we model it from that research and cite it. Everywhere else the projection is our own deliberately modest estimate, and it says so on screen.
Aimed at optimal, honest about the data
Your targets are top-decile percentiles, not "within normal range," because normal ranges track a population that is getting less healthy. But we won't pretend the underlying studies are perfect: much of the nutrition research was run on already-metabolically-unwell, largely American cohorts, which changes what a result actually proves.
That last point is worth an example. A trial raises a group's fiber intake and their metabolic markers improve, reported as "fiber improves metabolic health." True as far as it goes, but in a diet dominated by processed food, adding fiber-rich food usually means displacing processed food rather than simply adding a nutrient. Some of the benefit is the fiber. Some is what stopped being eaten to make room for it. A single trial rarely separates the two, which is why populations that already eat little processed food sometimes do fine on lower fiber than the American data would predict.
This isn't a reason to distrust the research. It's a reason to read it carefully, and to be suspicious of any app that quotes a study without mentioning who was in it.
Frequently asked
Do I need a doctor's visit or bloodwork to get started?
No. You can calculate an accurate baseline in about two minutes using basic physical tests at home, a tape measure and a few movement checks. Add bloodwork later, whenever you happen to have routine lab results.
Is my data sold to insurers or data brokers?
No. We don't sell your data and we don't share health data with advertisers, data brokers or any third party for their own purposes. It's stored with industry-standard encryption in transit and at rest, and you can export or delete it at any time. Full detail in the privacy policy.
How often should I update my test results?
Physical tests like push-ups, sit-to-stand and grip hang are worth redoing every 4 to 8 weeks, which is roughly how long a real change takes to show up. Bloodwork panels are typically once or twice a year, in line with normal check-ups.
How is my Longevity IQ score calculated?
Individual markers get 0-100 scores from percentile data, roll up into 6 weighted domains, which roll up into one overall 0-100 score. See the full breakdown above.
What's the difference between tracking modes?
Basic scores 4 domains with no bloodwork required. Advanced adds Metabolic Health. Complete adds Systemic Health. Each tier adds domains to the same 100-point total rather than replacing the scoring method.
What does "confidence" mean on my score?
How complete your data is for your current tracking mode, not how advanced your tracking mode is. Thin data means lower confidence, regardless of tier.
Is the projected decline based on real research?
For Strength, VO2 Max, and Blood Pressure, yes, cited published research. Every other marker uses the app's own clearly-labeled conservative estimate.
How does the workout program adapt to my schedule?
Six sessions and a rest day. When a domain like strength or VO2 max is lagging, the sessions that address it are marked high priority, so on a week when you cannot do all six the labels tell you which to keep.