The Oral-Systemic Health Evidence Map: What We Know, What We Don’t, and What’s Overhyped

Oral Signal separates four questions. First: association — do the oral and systemic conditions occur together consistently? Second: causality — is there evidence that one contributes to the other rather than merely sharing risk factors? Third: intervention — does treating the oral condition change a meaningful systemic outcome? Fourth: economics — does that intervention create measurable value through lower cost, better utilization or other outcomes? A condition can score highly on the first question and remain weak on the next three.

The ADA makes the same core caution in its oral-systemic guidance: periodontal disease has been associated with heart disease, diabetes and other conditions, but direct causality remains elusive across many of these relationships, and shared risk factors such as smoking and diet complicate interpretation. ¹

Diabetes and periodontitis sit at the top of the current evidence map because the relationship has been studied extensively in both directions. Diabetes can worsen periodontal health through altered immune response, microvascular changes and impaired healing, while periodontal inflammation has been investigated as a contributor to glycemic control. The ADA describes oral-systemic links involving diabetes as among the best established relationships in the field. ¹

What makes diabetes especially important is that the evidence extends beyond association. Periodontal treatment has been studied as an intervention, and economic analyses repeatedly suggest lower medical spending among people with diabetes who receive treatment. A 2023 JADA claims analysis found 12% lower overall healthcare costs among commercially insured patients receiving periodontal treatment and 14% lower costs in a Medicaid cohort. ⁶

A 2024 systematic review of economic evaluations concluded that periodontal treatment reduced several categories of healthcare cost in adults with type 2 diabetes and periodontitis, while also noting the need for more work on transferability across populations. ⁷

That does not mean periodontal care is proven to prevent diabetes complications or generate a guaranteed savings amount. Claims studies remain vulnerable to confounding, and economic models depend on assumptions. But diabetes is the category where association, intervention and economic evidence overlap enough to justify serious payer experimentation.

Periodontitis has repeatedly been associated with cardiovascular disease, including atherosclerotic disease and related outcomes. The biological hypotheses are plausible: chronic inflammation, transient bacteremia and shared inflammatory pathways could contribute to vascular disease. But plausibility and association do not establish that periodontal disease is a direct cause of cardiovascular events. ¹

This is one of the most important areas for evidence discipline because the public narrative often moves directly from 'gum disease is associated with heart disease' to 'treating gum disease protects your heart.' The second statement requires randomized or otherwise convincing intervention evidence showing that periodontal treatment reduces clinically meaningful cardiovascular outcomes. That evidence is not established at the level required for such a claim.

The practical implication is not that cardiovascular links should be ignored. It is that the current use case is more defensible in risk awareness and integrated care than in promising event prevention. Dentists should understand relevant medical history, physicians should understand that oral inflammatory burden may be part of a patient’s broader health context, and researchers should continue testing mechanisms and interventions.

Pregnancy provides one of the most useful examples in all of oral-systemic science. Maternal periodontitis has been associated in observational literature with outcomes such as preterm birth and low birth weight. Those relationships are biologically plausible and have been repeatedly studied. Yet when researchers have asked the intervention question — does treating periodontal disease during pregnancy reduce those adverse outcomes? — high-quality trials have not consistently shown benefit. ⁴⁵

A BMJ systematic review of randomized trials found that higher-quality studies provided clear evidence that periodontal treatment did not reduce preterm birth or low birth weight, even though lower-quality trials often suggested benefit. ⁵

A later umbrella review reached a similarly cautious conclusion: findings across systematic reviews were differential and unclear. Periodontal treatment remains appropriate for maternal oral health, but it should not be sold as a proven strategy for preventing preterm birth. ⁴

This is the evidence hierarchy in action. The association can be meaningful, the treatment can be appropriate, and the systemic outcome can still fail to improve.

Respiratory health is different from the cardiovascular and pregnancy categories because the strongest oral intervention signal appears in specific high-risk settings, particularly institutionalized older adults and populations vulnerable to aspiration pneumonia. Oral bacterial burden can plausibly contribute to respiratory infection when secretions are aspirated, making the mechanism more direct than some chronic inflammatory associations.

The important qualifier is population. Evidence from nursing homes or hospitalized older adults should not be generalized into a claim that routine dental care prevents pneumonia across the entire population. The intervention, setting and baseline risk matter enormously. Oral-systemic evidence often becomes misleading when a real effect in a narrow group is transformed into a universal wellness message.

Periodontal disease, tooth loss and oral-health burden have been associated with cognitive decline and dementia in observational studies. The hypotheses include chronic inflammation, microbial pathways, nutrition and shared social or behavioral risk factors. But dementia also causes oral-health deterioration through reduced self-care, medication effects, diet changes and dependence on caregivers. That creates a powerful reverse-causality problem.

A patient developing cognitive impairment may stop brushing effectively, miss preventive care or lose the ability to complete treatment. The oral disease can therefore be a consequence or marker of declining health rather than a cause. Until stronger longitudinal and interventional evidence separates those pathways, dementia should remain an emerging-to-moderate relationship rather than a causal oral-health claim. The ADA’s broader warning that direct causality in oral-systemic associations remains elusive applies especially strongly here. ¹

Chronic kidney disease populations often have substantial oral-health burden, and periodontal disease has been associated with kidney outcomes in observational research. But CKD also overlaps with diabetes, cardiovascular disease, age, medication burden, socioeconomic disadvantage and healthcare access. Those shared factors can generate association without a direct periodontal effect.

The clinical argument for good oral care in kidney disease does not require a strong causal claim. Patients may need infection control before transplantation, medication review and coordinated management. The research question is whether periodontal treatment changes kidney-specific outcomes beyond improving oral health itself. That remains far less established than the diabetes relationship.

This distinction is often underappreciated. The mouth can be valuable to healthcare because systemic conditions produce oral manifestations. Medications can cause xerostomia. Immunosuppression can change infection risk. Cancer therapy can affect mucosa and salivary function. Nutritional problems can present orally. Diabetes can worsen periodontal disease. In those cases oral findings can be clinically useful even if the causal arrow runs from body to mouth rather than mouth to body. ³

The first Surgeon General’s report emphasized that many systemic diseases and conditions have oral manifestations and that these can be initial signs of disease. It also noted associations between periodontal diseases and diabetes, cardiovascular disease, stroke and adverse pregnancy outcomes while calling for further research. ³

That framing is still useful decades later: the mouth is a window into health, but a window is not necessarily a driver.

Smoking, diet, obesity, income, education, stress, healthcare access and age can influence both oral and systemic disease. If those variables are incompletely measured, periodontal disease may appear to predict medical outcomes because it is partly capturing the same underlying risk environment. ¹

This does not make the association meaningless. A marker can still be useful even when it is not causal. If poor oral health helps identify people with unmet needs, medication problems or chronic-disease risk, it may have predictive value. But a predictive marker and a therapeutic target are different things. Treating a marker does not necessarily change the outcome it predicts.

The commercial temptation is to jump quickly from association to savings. If periodontal disease is associated with diabetes complications, perhaps treating periodontal disease reduces claims. Sometimes the data support that hypothesis, as in diabetes. But economic claims should generally follow intervention evidence rather than substitute for it.

Claims analyses can be useful because they reflect real-world spending at scale. They are also especially vulnerable to selection bias. People who receive dental treatment may be healthier, wealthier or more engaged with care. The strongest future evidence will connect clinical treatment, oral outcomes, medical outcomes and costs prospectively.

Oral-systemic health does not need every hypothesized relationship to be causal in order to be strategically important. There are at least four durable reasons for integration. First, medical conditions affect oral care. Second, oral infections and medications can affect medical treatment decisions. Third, dentists can identify health risks in people who may not be seeing physicians. Fourth, dental data may add useful information to broader healthcare analytics.

Those use cases are more resilient than any single disease claim. If a future randomized trial shows that periodontal therapy does not reduce cardiovascular events, medication reconciliation between dentists and physicians can still be valuable. If oral biomarkers fail to improve a diabetes prediction model, shared records can still improve oncology coordination. Integration is a broader infrastructure thesis than causality.

The most overhyped statements are those that imply certainty across the entire category: that gum disease causes heart disease, that treating periodontal disease prevents preterm birth, that oral bacteria explain Alzheimer’s disease, or that dental care broadly reduces medical spending. The evidence is not uniform enough for those statements.

The second form of hype is numerical precision without context. A claims study can estimate thousands of dollars in savings, but the population, benefit design, matching method and time horizon matter. A relative-risk estimate from a high-risk nursing-home population should not be applied to healthy adults. Evidence becomes misleading when the denominator disappears.

The underappreciated story is infrastructure. Shared EHRs, interoperable dental data, payer analytics, closed-loop referrals and targeted benefit design can create value without solving every causal question first. Oral health can become more integrated into healthcare because information and workflows are fragmented, not merely because periodontal inflammation might affect distant organs.

The second underappreciated story is that negative evidence is useful. When strong trials show that an intervention does not change a systemic outcome, the category becomes more credible by acknowledging it. Oral-systemic health will mature faster if researchers and companies retire weak claims instead of protecting them.

A useful evidence index is not a ranking carved into stone. Conditions should move as new longitudinal studies, randomized trials, economic evaluations and implementation data emerge. A mature field should expect some hypotheses to strengthen and others to weaken.

Oral Signal’s Evidence Index is therefore designed as a living product. The rating describes the maturity of a relationship, not a treatment recommendation. The individual dossiers separate causality, intervention and economics precisely because a single label cannot capture the whole evidence state.

The strongest oral-systemic story is not that the mouth secretly causes every chronic disease. It is that oral health and general health interact through multiple pathways, with different levels of evidence across different conditions. Diabetes is the most mature relationship and has meaningful intervention and economic evidence. Cardiovascular disease has strong association but weaker proof that periodontal treatment changes events. Pregnancy shows how a real association can fail the intervention test. Dementia and kidney disease remain far more uncertain. ¹⁴⁵⁶⁷

The category becomes more valuable when those differences are made explicit. Association is a reason to investigate. Causality is a higher bar. Intervention is the test of whether changing oral health changes a systemic outcome. Economics asks whether that change creates value. Keeping those questions separate is not skepticism about oral health. It is how oral health becomes credible healthcare.