Human trials and systematic reviews—not petri dishes dressed up as promises.
THE ALL TOLD EVIDENCE YEARBOOK · ISSUE 01
A YEAR OF EVIDENCE.
THE WINS. THE DOUBTS. THE LOT.
Eleven recent human studies and reviews, translated into plain English. Positive, negative and inconvenient findings all included.
READ THE YEAR →A study of ten people does not carry the same weight as a review of thousands.
Dose, population, duration, study design and outcome decide what a result means.
A statistically significant result can still be small, uncertain or irrelevant to you.
THE SHORT VERSION
SUPPLEMENTS AREN'T MAGIC.
CONTEXT IS THE ACTIVE INGREDIENT.
- Most honest win: magnesium may help poor sleepers a little—not transform sleep.
- Biggest disagreement: collagen reviews reached opposite-looking conclusions.
- Clearest warning: a useful ingredient in one setting may do nothing in another.
RECENT RESEARCH · 2025—2026
THE YEAR IN RESEARCH.
Tap any entry for the finding, the catch and our verdict.
01MAGNESIUM · SLEEPA REAL EFFECT. A SMALL ONE.MODEST SUPPORT+
WHAT IT FOUND
In 155 adults who already reported poor sleep, 250 mg of elemental magnesium as bisglycinate improved insomnia scores slightly more than placebo after four weeks. The effect size was small.
THE CATCH
Sleep was self-reported, the trial was short and other psychological outcomes did not differ. This supports “may help a little,” not “will fix your sleep.”
02L-THEANINE · ATTENTIONFOCUS, BUT NOT ACROSS THE BOARD.SUPPORTED WITH LIMITS+
WHAT IT FOUND
A review of 31 randomised trials found that a single 200 mg dose improved choice reaction time, suggesting a short-term attention benefit in healthy adults.
THE CATCH
Not every cognitive test improved, studies varied and evidence for mood or clinical conditions was less certain. “Attention support” is fairer than “brain boost.”
03CREATINE · LEAN MASSTHE SCALE MOVED. MUSCLE GROWTH DIDN'T.NO ADDED BENEFIT+
WHAT IT FOUND
Five grams of creatine daily increased measured lean mass after a seven-day wash-in, but did not increase lean-mass growth beyond resistance training over the following 12 weeks.
THE CATCH
The early change may reflect water, not new muscle. The study had 63 participants and the control group received no placebo, so expectations may have influenced behaviour.
04BEETROOT / NITRATE · EXERCISESOME PERFORMANCE WINS. NOT ALL.MIXED SUPPORT+
WHAT IT FOUND
An umbrella review reported benefits for time to exhaustion, distance, muscular endurance and some power outcomes, especially around at least 6 mmol nitrate daily for more than three days.
THE CATCH
Several other outcomes did not improve, and most included reviews were rated low or critically low quality. A broad “improves performance” claim would overstate it.
05BEETROOT / NITRATE · LIFTINGMORE NITRATE. NO MORE PERFORMANCE.NO CLEAR BENEFIT+
WHAT IT FOUND
In 18 resistance-trained men, low, moderate and high nitrate doses did not improve jumps, squats or bench-press performance versus placebo.
THE CATCH
This was a small acute study in men, so it cannot close the case. It does show why a positive endurance result cannot automatically be borrowed for lifting.
06COLLAGEN · SKINTHE POSITIVE AVERAGE DIDN'T SURVIVE THE QUALITY CHECK.EVIDENCE DISPUTED+
WHAT IT FOUND
Across 23 trials, pooled results favoured collagen for hydration, elasticity and wrinkles. But when the authors restricted analysis to non-industry-funded or higher-quality trials, the benefits were no longer significant.
THE CATCH
This is the year’s clearest reminder that sponsorship and study quality can change the story. The authors concluded current clinical evidence does not support collagen for skin ageing.
07COLLAGEN · SKINANOTHER REVIEW SAW A BENEFIT.CONFLICTING EVIDENCE+
WHAT IT FOUND
A separate review of 10 trials found statistically significant improvements in skin hydration and elasticity, most often at doses around four grams daily.
THE CATCH
The evidence was mostly in women, had moderate inconsistency, unclear risk of bias and too few studies to assess publication bias. Both collagen reviews belong in the conversation.
08PSYLLIUM · CHOLESTEROLA MEASURABLE, SPECIFIC WIN.SUPPORTED WITH VARIATION+
WHAT IT FOUND
Across 41 randomised trials, psyllium lowered LDL cholesterol by about 8.6 mg/dL and total cholesterol by about 9.1 mg/dL on average. Triglycerides and HDL did not clearly change.
THE CATCH
Results varied considerably between trials. This supports a narrow cholesterol finding—not a blanket claim that psyllium “fixes gut health” or prevents disease.
09PROBIOTICS · HEALTHY GUTSNO BROAD DIVERSITY BOOST.NO CLEAR BENEFIT+
WHAT IT FOUND
A systematic review of randomised trials found no significant increase in common measures of gut-microbiome diversity in healthy people taking probiotics.
THE CATCH
Diversity is not the same thing as health, and strain-specific effects or benefits in particular conditions may still exist. The honest conclusion is narrower: a generic diversity claim is not supported.
10OMEGA-3 · HEART HEALTHFORMULATION CHANGES THE ANSWER.CONTEXT DEPENDENT+
WHAT IT FOUND
A 2026 review of high-dose EPA-dominant therapy in high-risk patients found fewer hospital admissions for unstable angina. Other outcomes, including cardiovascular death and stroke, were not clearly reduced.
THE CATCH
Prescription-strength EPA in high-risk patients is not equivalent to an ordinary fish-oil supplement in a healthy customer. We will not borrow a drug-level result for a retail capsule.
11VITAMIN D · BONE MARKERSTHE BIOMARKERS MOVED A LITTLE. OUTCOMES ARE UNPROVEN.EARLY SIGNAL+
WHAT IT FOUND
A meta-analysis found modest reductions of roughly 3–4% in two bone-turnover markers with vitamin D supplementation.
THE CATCH
Biomarkers are not fractures. The authors called for further trials, so this does not prove that supplementing vitamin D prevents fractures in otherwise sufficient adults.