Correlation vs Causation
Short answer. Correlation vs causation is the difference between two things appearing together and one actually producing the other. In oral microbiome research news the distinction matters constantly, because most studies can show a link but few can show what drives it.
- A link between two things is not proof that one causes the other.
- Third factors, such as diet or smoking, can create false connections.
- The direction of cause can run either way, or both.
- Look for how the study was designed, not just what it found.
What correlation and causation mean
Two things are correlated when they tend to appear together. Causation means that one of them actually brings about the other. Ice cream sales and sunburn rise together in summer, but ice cream does not cause sunburn; hot weather drives both.
The same trap exists in the mouth. A bacterium may be found more often in people with a certain condition without being the reason they have it.
Why mouth studies are full of correlations
Most human studies sample a group of people at one moment, or follow them for a while, and compare the microbes they carry with their health. This is practical and affordable, but it cannot control everything else that differs between people.
Smoking, diet, age, medicines, brushing habits and dental history all shape the mouth's microbial community. Sorting out which factor is doing what is hard, which is why results are often described as associations.
Third factors and reverse cause
A third factor, sometimes called a confounder, can make two unrelated things look connected. If people who smoke have both a particular microbe profile and more gum trouble, smoking may explain the whole pattern.
Reverse cause is the other trap. A condition might change the mouth's microbes, rather than the microbes causing the condition. Inflamed gums, for example, change the environment bacteria live in, so a bacterium appearing there may be a result rather than a cause.
- Is there a third factor that could explain both
- Could the cause run the other way round
- Was the group large and varied enough
- Did anyone test the idea in a different way
How researchers get closer to cause
Stronger evidence usually comes from several kinds of work pointing the same way: studies that follow people over time, trials where participants are assigned an intervention at random, and laboratory work on mechanism. No single type settles it.
Randomised trials are the most direct way to test cause, because chance rather than choice decides who receives what. Even then, a trial tests one product or approach in one group, and its findings do not automatically apply to others.
Questions to ask of any story
When a story says something is "linked to" something else, ask what kind of study it was and whether it could have shown cause. If the article moves from "linked" to "causes" or "protects against" without new evidence, it has overreached.
The most useful step you can take is to read the study's own limitations section, where researchers usually say what they could not conclude. And if your own mouth is a worry, a dentist is better placed than any headline to help.
Questions readers ask
Does the oral microbiome cause disease?
Research is still developing and the evidence is limited or mixed. Some associations are well documented, but showing that a specific microbe causes a specific condition is much harder.
What is a randomised trial in plain terms?
Participants are assigned by chance to receive an intervention or not, so differences between the groups are less likely to be explained by other factors.
Should I ignore every correlation I read about?
No. Associations are often where useful ideas begin. The point is to hold them as leads, not conclusions.
Talk to a professional. Persistent bleeding, pain, swelling, loose teeth or sores that do not heal need a dentist's attention, not a supplement.