Association is not causation.
How to read oral–systemic health claims responsibly—and recognise when an interesting relationship is being presented as a proven cause.
Two conditions can appear together without one causing the other.
Research may find that people with one health condition are more likely to have another. That association can be important. It can generate hypotheses, guide further research and reveal groups that may benefit from closer attention.
But it does not, by itself, tell us that changing one condition will prevent or treat the other. That step—from relationship to cause—requires stronger reasoning and evidence.
THE ESSENTIAL DISTINCTION
Two variables occur together more—or less—often than expected.
Changing one factor produces a change in the other.
WHY A PATTERN MAY APPEAR
One observation.
Several explanations.
A causal relationship
One factor may contribute to the other. This is possible—but the association alone does not establish it.
Reverse causation
The apparent outcome may influence the supposed exposure, rather than the direction suggested by the headline.
A shared factor
Age, smoking, socioeconomic conditions, access to care or another variable may influence both observations.
Bias or chance
How participants were selected, how information was measured or random variation can create or distort a pattern.
READ THE WORDING
“Gum disease causes heart disease.”
“Gum disease and cardiovascular disease are associated; shared risk factors and possible biological pathways are being studied.”
Accurate language does not make a finding less important. It makes the limits of the evidence visible.
BEFORE YOU BELIEVE OR SHARE
Ask six questions.
- 01
Did the exposure happen before the outcome?
- 02
Was the result seen across different studies and populations?
- 03
How large is the effect—and what is the absolute risk?
- 04
Were important shared risk factors measured and addressed?
- 05
Does the study design support the strength of the claim?
- 06
Do the authors say “associated with” or actually demonstrate causation?
THE BOTTOM LINE
An association is a reason to investigate—not permission to overstate.
Good health communication protects both scientific credibility and people’s ability to make informed decisions. Look for the study design, the size and uncertainty of the effect, alternative explanations and language that matches the evidence.