Does Better Dental Care Actually Lower Medical Spending?

A 2023 JADA analysis used commercial insurance and Medicaid claims to examine people with diabetes who did and did not receive periodontal treatment. Among commercially insured patients, periodontal treatment was associated with 12% lower overall healthcare costs: $13,915 versus $15,739, with an estimated treatment effect of roughly $2,498. In the Medicaid cohort, treatment was associated with a 14% decrease: $14,796 versus $17,181, with an estimated effect near $2,918. Those are large differences, and they immediately explain why payers are interested. ¹

The same study is also a lesson in how carefully the results should be interpreted. It was a retrospective claims analysis, not a randomized insurance experiment. Researchers used propensity-score methods to reduce measured differences between treated and untreated groups, but unmeasured differences can remain. People who receive periodontal treatment may differ in healthcare engagement, disease severity, access, income, benefit generosity or other ways that also influence medical spending. The association is meaningful; it is not the same as a guaranteed $2,500 savings every time a payer covers periodontal therapy.

An earlier Health Economics analysis using integrated dental, medical and pharmacy claims found that periodontal intervention among people newly diagnosed with type 2 diabetes was associated with lower total healthcare costs by about $1,799, lower non-pharmacy medical costs by about $1,577 and lower diabetes-related healthcare costs by about $408. The study used inverse probability weighting and doubly robust methods, again trying to address differences between people who did and did not receive the intervention. ²

When independent datasets and analytic approaches point in the same direction, confidence in the existence of a signal rises. But replication of an observational association is still not the same as randomized proof. The most defensible interpretation is that periodontal treatment in people with diabetes is consistently associated with lower subsequent medical spending and deserves serious prospective testing.

A systematic review and meta-analysis of economic evaluations published in 2024 included 11 studies assessing periodontal treatment in adults with type 2 diabetes and periodontitis. It concluded that periodontal treatment reduced total healthcare costs, inpatient and outpatient costs, diabetes-related costs and some drug costs with low-to-moderate certainty. Three high-quality model-based cost-utility analyses produced an estimated incremental net benefit of $12,348 in 2022 U.S. dollars, rated high certainty within the review’s economic framework. ³

That is encouraging, but model-based economic evaluations depend on assumptions about treatment effects, costs, time horizons and transferability. A model can be rigorous without proving that a U.S. health plan deploying a periodontal program tomorrow will realize the same savings. The authors themselves called for research to establish transferability across countries and decision contexts. ³

The diabetes-periodontitis relationship has several advantages as an economic use case. The clinical association is mature. Diabetes is expensive and measurable. Periodontal disease is common. Both conditions involve inflammatory pathways, and periodontal treatment has been studied far more extensively in diabetes than oral-systemic interventions have been studied in many other chronic diseases. That creates a plausible chain from oral disease to intervention to measurable medical outcomes, even if every link is not fully causal.

The payer use case is also operationally straightforward. People with diagnosed diabetes can be identified from medical or pharmacy data. Dental utilization can be observed if benefits are integrated. Periodontal treatment can be defined from claims. Outcomes such as total medical cost, emergency visits, admissions, HbA1c testing and drug use can be tracked. Compared with a vague promise that oral health improves overall wellness, this is a testable program.

This is the central confounding problem. A person who completes periodontal treatment may be more likely to refill medications, attend primary-care visits, exercise, follow dietary advice or have stable housing and transportation. Dental treatment can become a proxy for healthcare engagement. Statistical adjustment can reduce this problem but cannot fully eliminate it when important behavioral and social variables are missing.

There is also selection by benefit design. Patients with comprehensive dental coverage are not equivalent to patients without it. Employers offering richer dental benefits may differ from employers with leaner plans. Medicaid programs vary in adult dental coverage, provider access and reimbursement. A claims dataset can make these populations look comparable while leaving meaningful structural differences unresolved.

A strange consequence of the oral-systemic movement is that dentistry sometimes feels pressured to prove it saves the medical system money in order to justify itself. That is an unnecessarily high bar. Treating pain, infection, tooth loss and periodontal disease has direct value. Oral health affects eating, speech, employment, sleep, social function and quality of life. NIDCR’s national work has repeatedly emphasized that oral health is essential to overall health and well-being even before any downstream medical savings are counted. ⁵

The economic question matters because it can reshape benefit design and payer incentives. But failure to prove medical savings would not make dental care unimportant. It would simply mean that the business case should rest more heavily on oral outcomes rather than cross-budget offsets.

A stronger test would identify people with diabetes and periodontitis, randomize or quasi-randomize access to an enhanced periodontal intervention, and prospectively measure both oral and medical outcomes. The intervention could include reduced cost sharing, proactive outreach, appointment navigation and standardized periodontal treatment. Outcomes should include dental treatment completion, HbA1c, medication adherence, outpatient utilization, emergency care, admissions and total medical cost over a sufficiently long horizon.

Crucially, the trial should measure the cost of the intervention itself. A program that reduces medical spending by $600 but costs $900 per member to operate is not a savings program, even if clinical outcomes improve. Conversely, a program can still be worthwhile if it improves outcomes at an acceptable cost without producing net savings. Healthcare often confuses cost effectiveness with cost reduction; they are not the same.

Medical-cost analyses can produce very different conclusions depending on whether they look six months, one year or several years after treatment. Some benefits may appear quickly through reduced acute utilization; others may require sustained periodontal control and better glycemic management. Attrition also matters. People can change insurers or employers before the payer that funded the dental intervention realizes any downstream benefit.

This creates a classic insurance problem: the entity paying for prevention may not be the entity capturing the eventual savings. Medicare, Medicaid managed care, large self-funded employers and vertically integrated insurers may have longer time horizons than fully insured commercial products with high member turnover. The economics may therefore vary by purchaser even if the clinical effect is identical.

The most useful question for payers may not be whether periodontal treatment saves money on average. It may be which patients produce the highest expected value. Someone with well-controlled diabetes, regular medical care and mild periodontal disease may have little opportunity for medical-cost reduction. Someone with poorly controlled diabetes, advanced periodontal disease and fragmented care may have far more.

That suggests a data strategy: combine medical risk, dental history and utilization to target intervention where both disease burden and actionability are high. The model should be validated prospectively and evaluated for bias. But segmentation could make a modest average effect economically meaningful by concentrating resources on people most likely to benefit.

This is where caution becomes even more important. Periodontal disease has been associated with cardiovascular disease and adverse pregnancy outcomes, but the ADA emphasizes that direct causality across many oral-systemic associations remains elusive. Shared risk factors such as smoking and poor diet complicate interpretation. ⁴

Pregnancy is a particularly useful warning. Observational studies can show that periodontitis and adverse outcomes occur together, while randomized trials of periodontal treatment during pregnancy have not consistently reduced preterm birth. An association can be real and still fail to translate into an effective intervention. That is exactly why economic claims should be built downstream from intervention evidence rather than upstream from association alone.

If the strongest evidence sits in diabetes, payers do not need to redesign every dental benefit around an unproven whole-body-health thesis. They can start with targeted programs. A health plan could waive periodontal cost sharing for members with diabetes, proactively connect diagnosed members to dental care, share relevant medical information with dentists and measure results. That is a smaller intervention, easier to evaluate and easier to stop if it does not work.

Employers could do the same. Rather than market enhanced dental coverage as a universal medical-savings engine, they could treat it as a chronic-disease experiment with predefined outcomes. That posture is more credible to benefits leaders and finance teams because it turns a broad narrative into a measurable hypothesis.

For companies building in medical-dental integration, the cost literature matters even before it is conclusive. It provides a reason for payers to engage. Products that connect dental claims to medical risk, navigate high-risk members into periodontal care, integrate benefits or measure cross-domain outcomes can anchor themselves to a concrete buyer problem: prove or disprove whether targeted oral intervention changes medical utilization.

The winning companies will not be those that cite the largest retrospective savings estimate. They will be those that can reproduce value in the buyer’s population, show exactly which members were affected, measure intervention completion and withstand scrutiny from actuarial and clinical teams.

Yes, there is credible evidence that periodontal treatment is associated with lower medical spending in people with diabetes, and a growing body of economic literature suggests the intervention can be cost-effective. The 2023 JADA claims study found roughly 12% to 14% lower overall costs in treated cohorts, while earlier work and a 2024 systematic review point in the same direction. ¹²³

But the intellectually honest version of the story is more valuable than the promotional one. The evidence does not yet justify saying that better dental care broadly reduces medical spending across all populations. It justifies a narrower conclusion: diabetes and periodontitis are mature enough for serious payer experimentation, and the next generation of evidence should come from prospective programs designed to measure causality, implementation and net economic value.