One Size Has Never Fitted All —
What Cancer Research Is Finally Catching Up To

I read an article recently about the emerging conversation in oncology around metabolic individuality in tumour biology — the idea that what works for one breast cancer patient may actively not work for another, and that treating a category of disease as though it were a single disease because it shares a tissue of origin is producing worse outcomes than treating each patient’s specific metabolic picture.

I am not an oncologist. This is not an article about cancer treatment. I have no business commenting on what people with cancer should or should not do clinically. The oncology researchers working in this space — people like Thomas Seyfried on the metabolic theory of cancer, like Jane McClelland on individualised approaches to treatment support — are doing serious, important work, and it is their expertise, not mine.

But the principle struck me. Because it is the same principle I have been applying in a very different context for thirty-seven years. And if oncology — one of the most rigorous and data-driven fields in medicine — is finding its way toward it, that feels worth writing about.

The Category Error in Everyday Medicine

The category error in question is this: treating a group as though the individuals within it are interchangeable because they share a label.

Breast cancer as a category contains tumours with wildly different metabolic profiles, hormonal sensitivities, immune environments, and genetic drivers. Treating them all with the same protocol because they are all breast cancer produces outcomes that reflect the average of that diversity — which means optimal outcomes for almost nobody.

Now apply that same logic outside oncology. High cholesterol as a category contains people with familial hypercholesterolaemia, people with insulin resistance driving small dense LDL, people with hypothyroidism reducing LDL clearance, people with genetic variants affecting LDL receptor function, and people whose LDL is elevated in the context of otherwise entirely healthy metabolic markers. These are different clinical pictures. They may share a number on a blood test. They do not share a cause, and they should not share a treatment protocol.

IBS as a category contains people with Blastocystis hominis infection, people with SIBO, people with parasympathetic nervous system dysfunction suppressing the migrating motor complex, people with HPA axis dysregulation elevating intestinal permeability, people with bile insufficiency from a sluggish gallbladder, and people with pancreatic elastase deficiency producing incomplete digestion. One label. Multiple distinct physiological pictures. The dietary advice or treatment that helps one will do nothing for another — and will sometimes actively worsen the condition of a third.

One size has never fitted all. This is not a novel insight. It is what the evidence shows in field after field, and what clinical observation confirms in practice every single day. The difficulty has always been in acting on it consistently.

Biochemical Individuality

The scientific foundation for this is not new either. Roger Williams — the biochemist who identified pantothenic acid and coined the term biochemical individuality in the 1950s — demonstrated that the range of normal variation in human enzyme activity, receptor function, and metabolic pathway efficiency is enormous. The “normal” range we use in standard laboratory testing is a statistical construct reflecting the middle 95% of a population sample. It tells us almost nothing about what is normal for a specific individual.

Two people can both fall within the “normal” range for a given marker and differ by a factor of three or four in the actual activity of the pathway that marker represents. The same nutrient, at the same dose, produces dramatically different effects in these two people — because the biological infrastructure processing that nutrient differs significantly between them. The same stress load produces dramatically different cortisol responses. The same dietary pattern produces dramatically different microbiome compositions.

This is not complexity for complexity’s sake. It is the actual biology. And the clinical implication is straightforward: generic recommendations produce generic outcomes. If you want specific outcomes — for a specific person, with a specific clinical picture, with a specific set of underlying drivers — you need specific information about that person.

What Testing Reveals That Generic Advice Cannot

The reason I built the Test, Don’t Guess platform around five specific functional tests rather than around dietary templates or lifestyle protocols is precisely this. I have seen too many people follow excellent, well-researched protocols and feel worse. I have seen too many people eliminate foods that are not actually causing their symptoms while continuing to eat things that are. I have seen too many people take supplements that are genuinely good supplements — for a different person with a different picture.

Testing changes the nature of the conversation. When I look at a GI-MAP, I am not applying an “IBS protocol.” I am looking at which pathogens are present, which keystone species are depleted, whether secretory IgA is adequate, whether zonulin is elevated, and what the beta-glucuronidase level is doing to this person’s oestrogen metabolism. Each finding points toward a specific intervention. The combination of findings — which is different for every person who sits in front of me — points toward a specific sequence.

When I look at a DUTCH Plus, I am not applying a “hormone balancing protocol.” I am looking at where this person’s cortisol pattern is — whether the HPA axis is running high, blunted, or flat — and what that means for their digestion, their thyroid conversion, their oestrogen clearance, and their capacity to mount an immune response. I am looking at specific oestrogen metabolites: whether 2-OH, 4-OH, or 16-OH pathways are dominant, and what that means specifically for this person’s hormonal picture.

The Organic Acids Test tells me things about mitochondrial function, neurotransmitter metabolites, and gut organism metabolic activity that no symptom history alone could reveal. The Randox blood chemistry panel shows me metabolic markers — insulin, homocysteine, ferritin, hs-CRP, free T3 — that the standard GP panel does not include and that often tell the more important story.

None of this produces a population-level recommendation. All of it produces a clinically specific picture of one person’s biology, at this point in time, with these specific findings. That picture changes over time. The interventions that address it are adjusted accordingly. This is what personalised medicine actually means — not a diet tailored to your food preferences, but a protocol tailored to your biology.

Why This Is the Only Approach That Makes Sense

The cancer research context brings something into focus that might otherwise seem like clinical fastidiousness. If the difference between generic and personalised treatment could be the difference between an effective and an ineffective outcome in oncology — and the emerging research suggests it can — then the same principle applied to less acute conditions represents an enormous accumulation of lost opportunity.

Every person who has been given the wrong dietary advice for their metabolic type and suffered for it. Every person whose gut infection was never identified because the right test was never run, and who was told instead to manage their symptoms with a low-FODMAP diet indefinitely. Every person whose fatigue was attributed to “lifestyle factors” when it was driven by mitochondrial dysfunction visible on an OAT. Every person whose hormonal symptoms were treated with HRT without anyone asking why the oestrogen metabolite profile was what it was.

These are not rare edge cases. They are the majority of the people I see in clinic. They have not been failed by medicine. They have been failed by the application of population-level tools to individual-level problems.

Testing before guessing is the principle. Not because testing is commercially convenient. Because without testing, you are applying the average to the individual — and the individual is never the average.

Stephen Duncan FDN-P MSc is a Functional Diagnostic Nutrition Practitioner based in Edinburgh. He is not an oncologist and this article does not address cancer treatment in any clinical capacity. This content is for educational purposes and does not constitute medical advice. For any medical condition including cancer, consult a qualified medical professional.