A Doctor's Supplement Cheat Sheet, Fact-Checked Against the Real Trials

A doctor named James DiNicolantonio posted a Supplement Cheat Sheet to Instagram recently, personal 1-10 ratings for twelve supplements, from magnesium at a 10 down to omega-3s and beta-alanine at 7. The list got shared widely, the way these things do, with no citations attached to any of the individual numbers. That's not necessarily a problem, a lot of practitioner shorthand works that way, but it's exactly the kind of claim worth running through the actual trial literature before repeating it to anyone.

So that's what we did, one supplement at a time.

Three of his ratings held up about as well as a claim can. Creatine remains one of the most studied supplements that exists, a loading protocol saturates muscle stores in about six days, and a smaller daily dose reaches the same roughly 20% rise over about four weeks. Lutein and zeaxanthin, for eye health, are backed by a meta-analysis of nine randomized trials and 920 eyes showing real gains in macular pigment density and visual acuity, one of the better-replicated findings on the entire list. Ashwagandha, one of the additions we made to his list, has two separate randomized trials, one of 46 men and a larger follow-up of 176 men, both showing a 14 to 17% rise in testosterone against placebo.

Several more were real but came with a real asterisk. Vitamin D3 and K2 got lumped into one 9/10 rating, but they're not the same claim: D3 deficiency correction is well-established, while K2's artery-calcium benefit is real in two separate trials but unproven for actual heart attacks, a cardiologist involved in one of those trials called recommending it for that purpose a bridge too far. CoQ10's strongest evidence is specifically in people on statins, which deplete it directly, not as a general energy supplement for everyone. Astaxanthin raises HDL and lowers triglycerides in a meta-analysis of eight trials, but never moved LDL or total cholesterol in the same data.

A few were weaker than the rating suggested once you look at the actual dose used in the studies. Inositol's real trials used 12 to 18 grams a day to cut panic-disorder and OCD symptoms, six to nine times what a standard 2-gram supplement serving provides, and the research population was women specifically. Collagen's skin-aging benefit disappeared from a 2026 meta-analysis of 23 trials once you excluded the studies funded by supplement and pharmaceutical companies, the first meta-analysis to report that split directly. A separate trial on collagen for knee pain held up fine, a different claim entirely. Taurine's headline finding, a 2023 paper linking it to multiple markers of aging, is a real and notable result, but it's a cross-species finding in worms, mice, and monkeys, not yet a human supplementation trial.

We also added five supplements with real evidence behind them that didn't make his original twelve: berberine, which matched metformin head-to-head on blood sugar control in a 2008 trial of newly diagnosed diabetics; zinc, which nearly doubled testosterone in a six-month trial, but only in men who started out deficient; rhodiola rosea, with a real but single modest trial in night-shift physicians; ashwagandha, covered above; and tongkat ali, which raised testosterone 37% in one trial, though a later meta-analysis found that effect concentrated in men who started with low testosterone to begin with.

None of this makes the original cheat sheet wrong, exactly. It makes it a starting point, one practitioner's shorthand for a hundred pages of actual research, some of it stronger than the rating suggests, some of it weaker, almost all of it worth reading past the headline number.

If you want to know which of these actually apply to your own numbers instead of a general population average, that's what a real discovery call is for. Book a free call and we'll figure out what's actually worth testing for you.