How We Vet Every Longevity Recommendation: The 5 Checks Behind 160+ Habits

One year a headline tells you coffee causes cancer. The next, coffee is protecting your heart. Eggs are deadly, then eggs are fine. Red wine adds years, then any amount of alcohol takes them away. If you have ever felt like health advice is just weather (changeable, confident, and impossible to plan around) you are not imagining it, and it is not your fault for feeling lost in it.

The good news first: science changing its mind is not the problem. Good science should update as the evidence grows. The harder truth is that most of the advice you run into never went through much of a filter before it reached you. A post gets written on deadline, built around a single eye-catching study, with the caveats trimmed off and the finding rounded up into a promise, and no one ever stops to ask whether following it could hurt someone. Multiply that across a million wellness articles and you get the whiplash. It is a genuinely hard thing to navigate, even for people who care a lot about their health.

That is the exact problem we set out to solve. Longlevity scores nine areas of your life and then hands you a library of 160+ specific actions you can take to move the number, and those actions are the part that changes your day-to-day life. So we hold them to the same standard as the math behind your estimate. This post is that standard, written out in full and with nothing hidden. It also doubles as something we hope is useful far beyond our site: by the end, you will be able to run these same five checks on any health advice you read anywhere.

The short answer

Before any recommendation reaches you, it has to clear five separate filters:

1. Sourced from research, not vibes. It starts from published studies, not from other wellness content.
2. Graded for evidence strength. Every action is graded Core (strong evidence) or Emerging (promising but earlier), and the newer ones wear that label right on the card so you can see which.
3. Citations checked adversarially. We interrogate each source: is it real, does it support the claim, and how strong is it?
4. A two-lens safety screen. Every action is reviewed for both physical and mental harm, and anything that cannot clear both is cut or rewritten with guardrails.
5. No inflation. Nothing is hyped past what its evidence supports, and the whole library is a map for a conversation with your doctor, not a replacement for one.

Want to see which actions apply to you? Take the 2-minute assessment →

Why most health advice fails before it starts

It helps to name the thing we are all up against. In 2005, physician-researcher John Ioannidis published a paper with a title that became famous precisely because it was hard to argue with: "Why Most Published Research Findings Are False." His point was not that scientists are dishonest. It was that small samples, flexible analysis, and a natural pull toward exciting results mean a real share of individual studies do not hold up when someone tries to repeat them.

That is the raw material a lot of health content gets built from, usually with the best of intentions and not enough time. A writer needs a post by Friday, finds a study with a striking result, and writes it up in good faith. What tends to get skipped is the slow part: checking whether that study has been replicated, whether the effect survives once you account for things like wealth and baseline health, and whether the advice could be risky for some of the people reading it. The finished piece reads confident and is often more fragile than it looks.

The answer is not to distrust all research, which would leave you with nothing to act on. It is to have a process that separates the durable findings from the fragile ones, and that holds a recommendation back until it has earned its place. Here is ours, step by step.

Filter 1: Sourced from research, not vibes

Every action in the library starts from the published literature: meta-analyses, randomized trials, and large prospective cohorts. We do not cite other longevity blogs as sources, because a claim copied between ten wellness sites is not any better evidenced than it was on the first one. Repetition can make an idea feel more true without making it more true, and we try to stay honest about the difference.

We also deliberately push past the obvious. "Sleep more" and "eat vegetables" are true, but you already know them, so leaving it there would not help you much. The library reaches for the specific, non-obvious actions that show up in the research and rarely in listicles, so there is something genuinely new even if you have the fundamentals handled.

Filter 2: Graded for evidence strength

Not all evidence is equal, and pretending otherwise is one of the quietest ways health advice misleads people. A finding from a single mechanistic study and a finding from three replicated cohort trials get presented with the same confident voice, and the reader has no way to tell them apart.

So every action carries a plain label, in the spirit of established evidence-grading systems like GRADE:

• Core means strong evidence: randomized trials, meta-analyses, or large, replicated cohorts.
• Emerging means promising but earlier: smaller studies, mechanistic findings, or mixed results.

We show you which is which on purpose. You get to weight an Emerging action differently from a Core one rather than being asked to take everything as equally settled. Transparency about confidence is itself a form of honesty.

You will see this labeling most visibly on the non-obvious, "go deeper" actions, where it matters most and where each one also carries its source study and a safety note you can read for yourself. The everyday fundamentals (morning light, moving more, protecting your sleep) rest on such deep, settled evidence that the grade is not in question. The formal disclosure earns its keep exactly where the science is newer and a reader most needs to know how firm the ground is.

Filter 3: Citations checked adversarially

This is where a lot of otherwise careful content quietly breaks. It is not enough to attach a plausible-sounding reference to a claim. A citation can be real but point to a study that says something narrower than the claim, or that studied a different population, or that found a much smaller effect than the headline implies.

So each action's source is interrogated against three questions. Is the citation real? Does it support the specific claim, rather than an exaggeration of it? And how strong is the underlying study? When the honest answer to any of those is uncertain, we do not paper over it. We grade the action down, or we cut it entirely. A recommendation that cannot survive its own footnote does not belong in front of you.

Filter 4: The two-lens safety screen

This is the filter we care about most, and it is the one most health content skips.

Before an action makes the library, it is reviewed for harm along two separate lenses, physical and mental, and anything that cannot clear both is dropped or rewritten with the right guardrails. Crucially, we screen for category-specific risks, not just generic ones. A few real examples of what that catches:

• Diet: restriction language or framing that could feed disordered eating.
• Sleep: the anxiety that sleep tracking can create, sometimes called "orthosomnia," where worrying about your sleep score makes sleep worse.
• Breathwork: light-headedness or fainting from certain techniques.
• Loneliness, addiction, and low mood: shame and "willpower" framing that blames the person for a hard problem.
• Alcohol: dangerous withdrawal for heavy drinkers who cut back too fast, which can be medically serious.

Where a real risk exists, the action carries a plain-language "before you start" note, and the sensitive areas point to genuine support: a clinician, or a crisis line like 988. When there is any doubt, we err toward caution. Advice that helps the average reader but quietly endangers a vulnerable one is not something we are willing to publish, however good it looks on the page.

Filter 5: No inflation

The last filter is a discipline more than a step: we would rather under-promise. Nothing in the library is hyped beyond what its evidence supports. You will not find an action framed as a miracle, because the honest research almost never supports that framing.

And the whole library is educational, not medical advice. It is a map for a conversation with a qualified professional who knows your situation, not a treatment plan. Several actions are deliberately written as "ask your doctor about…" prompts for exactly that reason. A tool that pretends to replace your physician has failed the honesty test before it starts.

What this standard costs us (and why we pay it)

It would be easy to have a bigger library. Drop the adversarial citation check and we could add hundreds of actions with plausible references overnight. Drop the safety screen and we could confidently recommend things that sound powerful and read well. Grade everything "Core" and the whole library would look more authoritative.

We do the opposite on purpose. The library is smaller than it could be because things get cut. Some actions are labeled Emerging when a louder site would call them settled. The tone is calmer than the genre because we are not trying to sell you a breakthrough. The payoff is the only thing that matters in a health product: when Longlevity tells you to do something, you can trust that it went through all five filters before it reached you.

How to run these 5 checks on any advice you read

The useful part of a standard is that you can borrow it. The next time you read a longevity tip anywhere, run it through the same five questions:

1. Where did this come from? If the "source" is another blog or an influencer's opinion, treat it as a hypothesis, not a finding.
2. How strong is the evidence? One small study is a lead. A meta-analysis or replicated cohort is a conclusion. If the writer will not tell you which they have, that is information too.
3. Does the study say this? Click the citation if there is one. Headlines routinely claim more than their own footnotes support.
4. Could this hurt someone? Ask who this advice is dangerous for, and whether the writer acknowledged them. Silence on risk is a red flag, not a green light.
5. Is it promising too much? Real longevity effects are meaningful but modest. Anything promising to reverse aging or add decades is selling, not informing.

Advice that clears all five is worth acting on. Advice that fails even one deserves a healthy pause. This is the same skepticism we apply to contested areas like alcohol and preventive screening, where we use the most current and methodologically strongest evidence and say plainly when the science is still unsettled.

The bottom line

You should not have to be a research methodologist to take care of your health. But someone in the process should be, and with most health content, no one is. The five filters exist so that the burden of vetting does not fall on you at the moment you are just trying to do something good for your life.

Your estimate tells you where you stand. The library tells you what to do about it. We hold both to the same standard, and now you can hold anyone else's advice to it too. For the full detail on the math and the sourcing, our complete methodology is public, including what Longlevity deliberately cannot tell you.

Frequently asked questions

Citations

• Ioannidis JPA. Why Most Published Research Findings Are False. PLoS Medicine. 2005;2(8):e124.
• Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. 2008;336(7650):924-926.
• For our full sourcing, math, and limits, see the Longlevity methodology.