What Is Personalised Medicine?

Every experienced clinician already practises a form of personalised medicine. They adjust a dose for kidney function, choose a medicine a patient can afford, simplify a regimen for someone with memory problems, or frame dietary advice around the foods a family actually eats. What is new is the scale of biological information now available, and the claims made about what it can do. Understanding personalised medicine means separating those 2 things: the long standing clinical skill of tailoring care, and the newer, uneven science of tailoring it to measurable biology.

A confusion of terms

Several overlapping terms are in use.

Personalised medicine is the broadest: care tailored to the individual. It includes biological characteristics but also values, preferences, context and circumstances.

Precision medicine is usually used to mean tailoring based on biological data, particularly genomic and molecular information, often by identifying subgroups of patients who share a mechanism.

Stratified medicine describes dividing patients into groups likely to respond differently to a treatment, for example by biomarker status, and matching treatment to group.

In practice the terms are frequently used interchangeably. The important point is that precision in the molecular sense is one tool of personalisation, not a synonym for it.

Where biological personalisation is well established

Pharmacogenomics

Some of the clearest successes come from understanding how genetic variants affect drug response. Carriers of the HLA B*57:01 allele are at high risk of a serious hypersensitivity reaction to abacavir, an antiretroviral medicine, and screening before prescribing has become standard practice in many countries. Variants in the CYP2C19 gene influence activation of clopidogrel, an antiplatelet drug, and affect how well it works in some patients. Variants in CYP2C9 and VKORC1 influence warfarin dose requirements. International consortia now publish guidelines translating specific genotypes into prescribing recommendations for a growing list of medicines.

These examples share features worth noticing: a specific variant, a specific drug, a well understood mechanism and outcome data showing that acting on the result changes what happens to patients.

Oncology

Cancer care has been transformed by molecular characterisation of tumours. Treatments are increasingly selected according to specific mutations or markers in the tumour rather than by its site of origin alone. This is precision medicine at its most developed, supported by extensive trial evidence.

Where it is promising but less certain

Personalised nutrition

One of the most influential studies in this area, published in 2015, continuously monitored glucose in hundreds of adults and found that blood glucose responses to identical meals varied substantially between individuals. A model incorporating personal characteristics, including the gut microbiome, predicted responses better than carbohydrate content alone, and diets personalised using the model improved glucose responses in a small trial. Later large studies, such as the PREDICT programme, confirmed wide variation in responses to identical meals in both glucose and fat metabolism, and found that genetics explained only a modest part of that variation.

These findings are important. They show that population level dietary advice will not suit everyone equally, and that factors such as meal timing, sleep, activity and the microbiome contribute to individual responses. What they do not yet show is that commercial personalised nutrition programmes produce better long term health outcomes than well delivered standard advice. That evidence is still being gathered.

Nutrigenetic testing

Direct to consumer genetic tests offering diet or supplement recommendations are widely marketed. Most of the variants they report have small effects individually, the evidence linking them to specific dietary recommendations is often weak, and trials generally have not shown that providing genetic information improves dietary behaviour more than standard advice. There are exceptions with clearer clinical significance, such as variants affecting lactase persistence or rare inherited metabolic conditions, but these are the minority. The MTHFR variants often highlighted in such reports are common, and routine testing for them is not recommended by major genetics bodies for most clinical purposes.

Continuous glucose monitoring outside diabetes

Wearable glucose monitors are increasingly used by people without diabetes. They can be educational, showing a patient how different meals, sleep and activity affect their glucose. But glucose varies naturally, reference values for people without diabetes are still being defined, and there is a risk of overinterpreting normal fluctuations or encouraging unnecessarily restrictive eating.

The dimensions of personalisation that do not need a test

Much of the most effective personalisation is not molecular at all.

A plan built around these dimensions is personalised in the ways that most often determine whether it succeeds.

Learning from the individual: N of 1 thinking

When evidence from populations cannot say whether an intervention will help a particular patient, the patient’s own response can. The formal version of this is the N of 1 trial, in which a single patient alternates between an intervention and a comparison in a structured way, with outcomes measured consistently. Formal N of 1 trials are used for some treatment decisions in chronic conditions.

Most practitioners will not run formal trials, but the underlying logic is widely useful. Change one thing at a time where possible. Decide in advance what outcome will be measured and over what period. Record the baseline. Review honestly. If an intervention does not help after a reasonable trial, stop it. This disciplined approach guards against the common trap of accumulating interventions, none of which can be shown to be working.

A composite example of personalisation in practice

2 colleagues in their early fifties, both with prediabetes, receive the same general advice to lose weight and exercise. One works nights as a nurse, sleeps in the day and eats her main meal at 3 am; the other is a lawyer with long evenings at the desk and a large late dinner. A personalised plan for the nurse starts with protecting daytime sleep and restructuring meals around her shift pattern; for the lawyer, with moving the main meal earlier and building activity into the day. Neither needs a genetic test. Both need a plan that fits the physiology created by their lives. Their progress is then tracked with the same simple markers, HbA1c, waist circumference and weight, and adjusted accordingly.

Questions to ask before ordering a personalisation test

1. What specific decision will this result change?
2. Is there good evidence that acting on the result improves outcomes, or only that the marker varies between people?
3. How reliable is the test, and is it validated for this purpose?
4. What will I do with an unexpected or ambiguous result?
5. Could the same information be obtained from the history or from a trial of intervention?
6. What will it cost the patient, and is that proportionate?

The ethical dimension

Personalised medicine raises questions practitioners should keep in view. Genetic information concerns relatives as well as the patient. Data from testing and wearable devices may be stored and shared in ways patients do not expect. And expensive tests can widen health inequalities if they become a prerequisite for good care. Personalisation that relies mainly on time, attention and a careful history is, by contrast, available to almost everyone.

Personalisation within Integrated Functional Medicine

Integrated Functional Medicine is personalised by design: the timeline and matrix are built from one person’s story, and the integrated history includes every form of care they use. Biological tests, including genomic ones, have a place where they will change a decision and where their validity is established. But the core of personalisation remains what it has always been: understanding the individual well enough to offer them the right thing, at the right time, in a form they can use.