ASSOCIATION IS NOT CAUSATION
The Association of Periodontitis With
Cardiovascular Disease Parameters:
A Synthesis of Systematic Reviews
A practical lesson in reading health evidence: when two conditions occur together, what would it take to show that one actually causes the other?
Volume 24 · Issue 2 · 278–306 · 2026
DOI: 10.1111/idh.12885
ORIGINAL PAPER · AT A GLANCE
Volume 24 · Issue 2
The Association of Periodontitis With Cardiovascular Disease Parameters: A Synthesis of Systematic Reviews
Schoenmakers · Weijdijk · Willems · Van der Weijden · Slot
Is periodontitis causally responsible for cardiovascular disease, or do the two conditions occur together because of shared risks and other underlying factors?
This is a PerioMed-designed bibliographic presentation and editorial synthesis, not a reproduction of the publisher’s first page or full abstract.
THE LEARNING LENS
Periodontitis and cardiovascular disease
The relationship examined by the authors provides the real scientific case.
Association does not establish causation
The same reasoning applies whenever research links an exposure, condition, behaviour or treatment to a health outcome.
01 · FIRST, SEPARATE THE QUESTIONS
“Do they occur together?” is not the same as
“Does one cause the other?”
An association is an observed pattern. Causality is an explanation for why that pattern exists.
In this paper, epidemiological studies repeatedly report that periodontitis and cardiovascular disease occur together. That observation matters—but it does not tell us whether periodontitis contributes directly to cardiovascular disease, whether the direction runs the other way, or whether shared factors help produce both.
HOW THE PAPER WAS BUILT
A review
of reviews.
The authors searched MEDLINE–PubMed and Cochrane–CENTRAL for systematic reviews comparing cardiovascular disease or events in people with and without periodontal disease. They then examined both the magnitude of the associations and whether the evidence supported causality.
THE CENTRAL IDEA · APPLICABLE BEYOND THIS PAPER
One observed relationship.
At least three possible explanations.
This is the question to ask whenever a study reports that two health variables are linked.
Periodontitis may contribute to cardiovascular disease.
Inflammatory or microbial pathways could connect periodontal inflammation with cardiovascular biology.
The apparent direction may be reversed.
Cardiovascular disease—or factors surrounding its development and treatment—could influence periodontal health.
A third factor may contribute to both.
Smoking, age, diabetes, obesity, stress and socioeconomic conditions can shape both periodontal and cardiovascular risk.
WHAT THE REVIEW FOUND
An association is present.
Its magnitude is mostly small.
Statistical significance and clinical meaning are not the same thing. The size, consistency and context of an association matter when asking whether it could reflect a causal relationship.
of reported risk ratios and odds ratios were estimated to represent a negligible magnitude of association.
of reported estimates represented a small magnitude of association.
for a predominantly negligible-to-small association—not for a confirmed causal relationship.
Results varied with age, sex, periodontal disease severity, smoking status and geographic region—factors that can influence interpretation of the observed relationship.
BRADFORD HILL · NOT A MATHEMATICAL TEST
Three considerations supported.
Six still insufficient or uncertain.
The Bradford Hill considerations help organise causal reasoning. They are not a checklist in which reaching a particular score automatically proves causation.
Strength
Not satisfiedThe reported associations were predominantly negligible to small.
Consistency
SupportedThe association appeared repeatedly across the reviewed literature.
Specificity
Not satisfiedBoth diseases have multiple causes and overlapping determinants.
Temporality
UncertainThe available evidence did not consistently establish which condition came first.
Biological gradient
SupportedDisease severity provided some evidence of a dose–response pattern.
Plausibility
SupportedInflammatory and microbial pathways offer biologically reasonable mechanisms.
Coherence
InsufficientThe overall clinical evidence did not confirm a coherent causal chain.
Experiment
InsufficientIntervention evidence on major cardiovascular outcomes remains limited.
Analogy
InsufficientComparable relationships do not establish this one as causal.
A REUSABLE READING TOOL
Before interpreting any reported “increased risk,” ask what the study can actually distinguish.
These questions are not limited to oral–cardiovascular research. They help test causal claims across health information.
- 01Was periodontitis measured accurately?
- 02Did it clearly precede the cardiovascular outcome?
- 03Were smoking, age, diabetes and other confounders adequately controlled?
- 04Could reverse causation explain part of the association?
- 05Are results consistent across populations and study designs?
- 06Does treating periodontitis change cardiovascular outcomes?
WHAT THE PAPER ALLOWS US TO SAY
Four statements.
Four different levels of evidence.
“Periodontitis is associated with cardiovascular disease.”
The synthesis identifies a generally negligible-to-small association with moderate certainty.
“Periodontitis causes cardiovascular disease.”
The available evidence did not satisfy enough causal considerations to support this claim.
“Biological mechanisms could connect periodontal inflammation with cardiovascular biology.”
Inflammatory and microbial pathways support a scientific hypothesis, not proof of a clinical causal effect.
“Treating periodontitis prevents cardiovascular events.”
Association evidence cannot demonstrate that periodontal treatment prevents myocardial infarction, stroke or other major events.
THE NEXT RESEARCH QUESTION
Does changing periodontal disease
change cardiovascular outcomes?
Future work must move beyond repeating the same association. Stronger causal inference requires prospective designs, rigorous periodontal measurement, clear temporality, careful control of confounding and adequately powered intervention studies with long-term cardiovascular endpoints—not only surrogate inflammatory biomarkers.
TAKE THIS METHOD WITH YOU
When you encounter a health claim,
read beyond the headline.
Identify the claim
Is the source describing an association, suggesting a mechanism, or claiming a causal effect?
Check direction and timing
Did the proposed cause clearly occur before the outcome—and could the direction be reversed?
Look for shared factors
Could age, behaviour, environment, access to care or another condition influence both sides?
Ask what changes outcomes
Does modifying the proposed cause actually change the outcome in well-designed studies?
PERIOMED BOTTOM LINE
“Finding that X and Y occur together is the beginning of a causal question—not its answer.”
That distinction matters far beyond this paper. Here, the responsible interpretation is neither to dismiss the periodontal–cardiovascular relationship nor to overstate it: current evidence supports an association and plausible biological pathways, but it does not establish that periodontitis causes cardiovascular disease or that periodontal treatment prevents major cardiovascular events.