What Is Genetic Variation, Really? Why Population Studies Can Miss Individual Biology

Population-based studies are essential in medicine.

They allow us to test whether a drug, exposure, diet or lifestyle factor is associated with increased disease risk in a group of people. Without them, we would not know whether an effect is general, rare, strong, weak or detectable at all.

But population studies answer population-level questions. They can tell us whether an exposure increases risk across a group.

They usually cannot tell us why one specific individual developed disease while another individual, exposed to the same thing, did not. That is a different biological question.

To answer that question, we need to look below the population average and ask what differs between individuals at the molecular level.

This is where genetic variation becomes important.

Genetic variation does not simply mean having a “disease gene”.

In this context, the focus is not on sporadic mutations that arise in individual cells during life, such as many of the mutations involved in cancer development.

The focus is on gene variants that are part of a person’s biological starting point.

These variants may be inherited from the parents, or they may originate very early during embryonic development. If they are inherited, they are usually present in essentially all nucleated cells of the body, in skin cells, liver cells, immune cells, brain cells and many other tissues. If they arise during early development, how widely they are present depends on how early they occurred.

This matters because a gene variant is not the same thing as a random mutation in one isolated cell.

A somatic mutation is a DNA change that arises in a body cell during life, for example after UV exposure, replication errors, chronic inflammation, chemical exposure or virus-induced DNA damage. It is not inherited from the parents, and it is not present in every cell of the body. It may exist only in one single cell. If that cell divides, the mutation can be passed on to its daughter cells, creating a clone of cells that all carry the same mutation.

This is important in cancer biology, where a mutation may begin in one cell and then become present in the tumour cells that descend from it. But this is not the type of genetic variation discussed here.

A constitutional gene variant is different. Because it can be present across many tissues, it can influence how the body as a whole responds to a drug, hormone, toxin, infection, allergen or environmental exposure.

A somatic mutation may change the behaviour of one cell lineage.

A constitutional gene variant may shape the sensitivity of the organism.

It can influence how strongly a gene is expressed, how well a protein works, how a receptor responds, how efficiently a molecule is metabolized, how the immune system recognizes a signal, how well DNA damage is repaired or how easily a cell crosses a biological threshold.

Some variants may have large effects.

Most have small effects.

But together, they help shape the molecular landscape that determines how a body responds to the world.

Most genetic variants do not cause disease on their own, but they can shift biological sensitivity.

They can make one body slightly more resilient in one context and more vulnerable in another.

This is the important point: genetic variation often does not create a binary outcome where one person is healthy and another is sick. More often, it changes the molecular starting point.

It changes the threshold.

This is also why population-level conclusions may have limited resolution from a molecular point of view.

Imagine a large study where people are exposed to a drug or an environmental factor and then followed for ten years. At the population level, the conclusion may be clear: there is no statistically significant increase in cancer risk.

That conclusion may be completely correct.

But it does not mean that no individual in that population developed cancer. A few people may still have done so.

Statistically, those cases may fall within what we call background incidence. They are not enough to move the result of the whole group and they are not enough to conclude that the exposure increased cancer risk in the population.

But molecularly, each of those cases still happened inside a real body.

And this is where the perspective shifts.

From a statistical point of view, those cases may be treated as part of the expected background incidence.

From a molecular point of view, they are not just background incidence. They are biological events. Biological events for an individual.

Perhaps those individuals carried variants affecting DNA repair, immune surveillance, inflammation, hormone signalling, detoxification enzymes, cell-cycle control or apoptosis. Perhaps their cells were already closer to a threshold where one additional stressor mattered.

In that case, their risk was not meaningless. It was only statistically insignificant when compared with the population as a whole.

If those individuals had instead been analyzed within a subgroup carrying the same relevant genetic variant, the association between the exposure and disease risk might have become much stronger, perhaps even statistically significant.

In the full population, the biological signal may be diluted.

In the susceptible subgroup, it may become detectable.

This is also why animal experiments can still matter.

In human studies, the full molecular background of each individual is usually unknown. We therefore study genetically diverse populations and ask whether an effect is detectable at the group level.

Animal studies are not perfect models of humans. They do not translate directly into human biology. But because the genetic background can be more controlled, they can often reveal mechanisms that are almost impossible to detect in a mixed human population.

A human study may therefore conclude that there is no significant cancer risk in the population. But that is not the same as proving that no genetically susceptible individual could be at increased risk.

A statistically insignificant risk at the population level is not the same thing as a biologically insignificant risk in the individual.