Supplements Are Punctuation, Not the Sentence

I used to treat supplements like the sentence itself.

Back then I thought a long enough list and enough bottles meant the nutrition was written — done, handed in. Each morning a row of capsules into my palm, like turning in homework so I could stop thinking about it. It took years of blood panels and papers to change my grammar. Food is the sentence. Supplements are only punctuation. Punctuation can make a sentence clearer — a comma to let you breathe, a period to settle the tone — but it cannot write the sentence for you. If the sentence isn't there, no amount of punctuation saves it. You just get a scatter of marks that mean nothing.

So I write the sentence first. The base has always been food: salmon, sardines, boquerones, goat kefir straight cold from the fridge, eggs, legumes, colored vegetables and berries, olive oil, the whole fermented shelf, herbs and spices. Only once that sentence is full do I look for where a mark belongs. And belongs isn't a feeling — it's a number. I run baseline panels, read the real values on a schedule, ask my AI, go to the papers. And I read the papers specifically for their edges: who this helps, how much, and where the evidence quietly stops. What I'm reading for is the boundary.

I do take four things. I'm not going to hand you a dose, and I'm not naming a brand — that isn't what this is. What I want to show you is how I decided to let each one onto the table at all.

First, creatine. I do resistance training, sauna, cold plunge, and creatine pairs with the training line. The part of the evidence I trust most is steady: creatine, combined with strength training, tends to improve strength — and that holds in women, in the peri-menopausal, in older women too (I anchor to the kind of work in PMC11547435), not just the young men in the gym. That's the stronger tier. As for the possibility that it also helps cognition and mood — the proposed mechanism runs something like restoring the brain's energy stores, and the recent reviews (Nutrients 2025, the Oxford work in 2026) look at it carefully — I file that under preliminary: may, possibly, the studies in women still thin. I take it for the harder evidence about training. The brain part I treat as a bonus, not a promise.

Second, collagen peptides. Let me keep this one restrained, because it slides so easily into a conversation about looks, and that isn't the conversation I'm having. Multiple RCTs and meta-analyses observe that collagen peptides tend to improve skin hydration, elasticity, and wrinkles (low molecular weight, taken over some weeks, formulations and doses varying widely, with no notable adverse events — I anchor to the Am J Med 2025 meta and work like PMC6073484), with some joint and whole-body evidence besides. Note may support, tends to — not reverses. I care about skin as tissue: it's the largest organ I have, a barrier, something that repairs and ages. That's a different thing from looking younger. Whatever glow I'm after has never been a size or a face — it's the feeling of being clear, awake, a body running the way it should. That mark I set down lightly.

Third, vitamin D — which is almost a live demonstration of this whole metaphor. There's one rule: test, then supplement. Correcting a real deficiency has clear value — for bone, for immunity and the upper respiratory tract, especially in people whose baseline was low to begin with. But the other direction: if you were already replete and you top up anyway, the large trials (VITAL, ViDA, D2d — I anchor to the Nat Rev Endocrinol 2021 review) found no extra benefit on most hard endpoints. And there it is, the entire meaning of supplement as punctuation: it fills a real gap. It doesn't press more periods onto a sentence that was already complete. Where the gap is, the labs tell me. Where there's no gap, I don't add the mark.

Fourth, magnesium. This one goes at the end of the day. Magnesium tends to offer a mild benefit for sleep, for anxiety and stress — especially in people running low at baseline — and it is linked to lower cortisol (the PMC5452159, PMC11136869 line of work). I said mild. Tends to. Not a sleeping pill, not a switch that turns anxiety off. More like a soft period at the end of a sentence, letting the day's tone come down. Some nights it's there. Some nights the sentence is written steadily enough that it doesn't need one.

Let me put the boundary where it's brightest: none of the four is writing the sentence for me. Every one of them stands after the food is written. The moment I let punctuation stand in for the sentence — eat carelessly and expect the bottles to fill nutrition in — all I'm holding is a string of floating marks. Decorative. Meaningless. The great failure of a supplement isn't that it doesn't work. It's being put in the wrong place — mistaken for the subject.

In this one, I want Foodna to land somewhere new. In the last few pages it read what's in a meal, whether the order was right, whether a mouthful of iron would be used. Here I want it to step back and read the daily food matrix: across these weeks, these dozens of meals, is the sentence itself full — enough protein, and the boundary lines like omega-3, B12, iodine, caught by the food itself? It reads the sentence, not the marks. And only when it sees a genuine gap — one that food most likely can't close — does it gently flag: here, maybe, a mark belongs. Whether to actually place it, and how heavy, that sentence goes back to the labs, never to a photograph. What it sorts for me is only this: this gap, food can fill versus this gap needs a mark. Nothing more.

So the sentence, I believe it now, all the way down: write the sentence first. Punctuation always comes after — and it stays punctuation.