Evidence is consistent, with known limits.
Across 25 studies and 1,876 people, forest therapy reduced stress with an effect size of 0.71 — larger than most things we check. Nearly every outcome was self-reported by people who knew exactly where they had been.
Melatonin is recognised as first-line therapy for insomnia and buys you about seven minutes. It is also not FDA-approved for anything — and when 31 supplements were measured, content ranged from 83% below the label to 478% above it.
Vitamin D works if you are deficient — strongly, for bone. The enormous trials reported as finding nothing largely enrolled people who already had enough, and two 2026 papers disagree about what that means for the heart.
Beetroot juice lowers blood pressure across 22 randomised trials. But juices with more nitrate produced smaller effects, and isolated nitrate supplements moved nothing at all — which is a problem for what the label says is doing the work.
Seven trials, 608 women. Creatine added 0.8 lb of lean mass over a median of nine months — and 16.5 lb to a leg press. Across 684 trials, placebo groups often reported side effects at least as often.
Two meta-analyses pooling 45 randomised trials found collagen improves skin hydration and elasticity against placebo. Half the trials in the larger one had missing outcome data, and 95% of participants were women.
Meditation and yoga lowered cortisol across 87 randomised trials, every one against an active control. They also lowered blood pressure, heart rate, inflammation and cholesterol — which is exactly why cortisol cannot be shown to be the lever.
Across 139,691 people in 17 countries, every 5 kg of missing grip strength came with 16% higher risk of dying — beating systolic blood pressure as a predictor. The same paper says nobody knows whether improving it changes anything.
Around 117,000 Japanese adults were followed for years, and those who said their life felt worth living were less likely to die. None of those studies used the four-circle diagram, which was first published in 2012 with a different word in the middle: purpose.
A 2025 meta-analysis linked 7,000 steps a day to 47% lower mortality (14 studies) and 38% less dementia (2 studies). For death the curve bends near 5,400 steps; cancer risk did not move. Whether age changes the number is still open.