Gum Disease Is Linked to Everything. Does It Cause Anything?

Periodontitis and systemic disease can appear together because one contributes to the other, because they share causes, or because bias and measurement error distort the picture. Several explanations may be operating at once.

A current umbrella review examined 24 systematic reviews covering 32 noncommunicable diseases. The reported associations ranged from strong to absent, while the certainty of evidence was generally low or very low. There is enough signal to take the subject seriously. There is not enough certainty to use the same causal sentence for every disease (Rodríguez-Medina et al., 2025).

That leaves us with a practical rule: count the rung before you choose the verb.

If I had to choose one oral–systemic relationship that already changes daily periodontal care, it would be diabetes.

The corrected 2026 replacement article, issued through a formal retraction-and-republication, included 28 prospective longitudinal studies from 16 countries and more than 300,000 participants. Periodontitis at baseline was associated with an 18% to 25% higher occurrence of newly incident type 2 diabetes. Diabetes at baseline was also associated with later periodontitis, although those estimates were modest and more imprecise. An earlier independent cohort meta-analysis found similarly modest associations in both directions, with substantial unexplained heterogeneity. Together, these findings support a bidirectional temporal association—not proof of reciprocal causation (Botelho et al., 2026; Stöhr et al., 2021).

The intervention evidence is more useful clinically. A Cochrane review included 35 randomized trials and 3,249 participants. At three to four months, 30 trials with 2,443 analyzed participants found that periodontal treatment reduced HbA1c by a pooled average of 0.43 percentage points compared with no active treatment or usual care (95% CI −0.59 to −0.28). Between-trial heterogeneity was substantial (I² = 71%), so this is not an individual-patient prediction. At six months the estimate was smaller—0.30 percentage points (95% CI −0.52 to −0.08)—and heterogeneity was higher (I² = 80%). Evidence beyond that was sparse (Simpson et al., 2022).

Notice the endpoint: HbA1c. It is a clinically relevant surrogate, but these trials did not show that periodontal treatment replaces glucose-lowering medication or prevents the long-term complications of diabetes.

Cardiovascular disease is where causal language becomes especially tempting.

A 2023 meta-analysis included 39 cohort studies—29 prospective and 10 retrospective—with 4,389,263 participants. Periodontal disease was associated with more major adverse cardiovascular events (RR 1.24, 95% CI 1.15–1.34), myocardial infarction (RR 1.14, 95% CI 1.06–1.22), and stroke (RR 1.26, 95% CI 1.15–1.37) (Guo et al., 2023).

Those pooled estimates were positive, but between-study heterogeneity was substantial: I² was 88.9% for major adverse cardiovascular events, 80.1% for myocardial infarction, and 94.8% for stroke. Periodontal definitions, cardiovascular endpoints, adjustment, and study design varied. The estimates do not mean that an individual patient’s 6 mm pocket caused a stroke.

The same problem appears with hypertension. I unpacked that literature in more detail in this article on gum disease and blood pressure.

Then the blood-pressure trials change the question. A 2026 meta-analysis pooled 12 randomized trials and reported that periodontal therapy reduced systolic blood pressure by 4.64 mmHg (95% CI −5.99 to −3.30), diastolic blood pressure by 1.84 mmHg (95% CI −2.65 to −1.02), and C-reactive protein by 0.58 mg/L (95% CI −0.83 to −0.33). The certainty was moderate to low, and the authors called for trials powered specifically for blood-pressure outcomes (Gandhi et al., 2026).

That final distinction matters. A blood-pressure shift is not a heart attack prevented.

Pregnancy is perhaps the cleanest example of why observational evidence and treatment evidence must be separated.

An overview of 23 systematic reviews found that the reviews at the lowest risk of bias still reported associations between maternal periodontitis and preterm birth (RR 1.6, 95% CI 1.3–2.0), low birth weight (RR 1.7, 95% CI 1.3–2.1), and pre-eclampsia (OR 2.2, 95% CI 1.4–3.4) (Daalderop et al., 2018).

Then we intervene—and the message becomes narrower. The 2017 Cochrane review found no clear reduction in preterm birth before 37 weeks (RR 0.87, 95% CI 0.70–1.10; low-certainty evidence), while a 2024 network meta-analysis produced encouraging estimates for several multicomponent periodontal protocols but judged every preterm-birth comparison low or very low certainty (Iheozor-Ejiofor et al., 2017; Wu et al., 2024).

The most current review included 14 randomized trials and 8,316 participants. It estimated RR 0.85 for preterm birth (95% CI 0.71–1.02), or 20 fewer preterm births per 1,000 people, with a 95% CI from 39 fewer to 3 more. In plain English: Treatment may produce a modest reduction, but prevention is not established because the interval remains compatible with no benefit (Thomas et al., 2026).

This does not make periodontal care during pregnancy unimportant. Scaling and root planing are considered safe when clinically indicated (American Dental Association [ADA], 2025). It makes the promise narrower: we should not tell a pregnant patient that periodontal treatment has been proven to prevent preterm birth.

Poor periodontal health is associated with later cognitive decline. The difficult question is what the association means.

A systematic review of 47 longitudinal studies reported associations with cognitive decline (OR 1.23, 95% CI 1.05–1.44) and dementia (HR 1.21, 95% CI 1.07–1.38). The authors rated the overall evidence as low quality and judged that reverse causality probably explained part of the relationship. The exposure definitions also mixed periodontitis, tooth loss, deep pockets, and bone loss, which are not interchangeable (Asher et al., 2022).

A 2026 umbrella review makes the certainty problem even clearer: 12 of its 14 included reviews were judged at high risk of bias, and GRADE certainty for mild cognitive impairment, Alzheimer disease, and dementia was very low (Rizwan Ali et al., 2026).

Reverse causation is not an abstract statistical problem here. Early cognitive deterioration can reduce brushing quality, dental attendance, nutrition, and the ability to manage complex care long before a formal diagnosis. Brushing also travels with education, income, dexterity, exercise, diet, smoking, and access to healthcare. A cohort can adjust for the variables it measured. It cannot adjust perfectly for everything that makes one person brush more often than another.

The oral–gut axis is biologically fascinating. It is also a place where a mechanism can run ahead of clinical proof.

A directionality-focused review examined whether periodontitis preceded inflammatory bowel disease and whether inflammatory bowel disease preceded periodontitis. For periodontitis to overall IBD, four longitudinal studies covering more than 10 million records produced no clear association (RR 1.04, 95% CI 0.99–1.09). There was a small association with ulcerative colitis (RR 1.12, 95% CI 1.04–1.21), but not Crohn disease (RR 0.98, 95% CI 0.92–1.04).

In the opposite direction, two studies found that IBD was associated with later periodontitis (HR 1.37, 95% CI 1.26–1.49) (Wang et al., 2024).

This is a good reminder that an enormous sample does not, by itself, eliminate exposure misclassification or residual confounding. The oral–gut pathway is plausible. We do not yet have evidence that treating periodontitis treats IBD, or vice versa.

Smoking is a major risk factor for both periodontitis and chronic obstructive pulmonary disease. If smoking exposure is measured crudely, the two diseases can appear more directly connected than they are.

A 2023 review of 22 observational studies and 51,704 participants found a broad association between periodontal disease and COPD (OR 1.20, 95% CI 1.09–1.32). After studies adjusted more rigorously for smoking intensity, the association was no longer clear (OR 1.14, 95% CI 0.86–1.51). The analysis also found no clear association in smokers or never-smokers considered separately (Yang et al., 2023).

That does not make oral care unimportant for a patient with COPD. It means this evidence does not establish periodontal disease as an independent cause of COPD—or periodontal treatment as COPD therapy.

This is confounding in its most clinically recognizable form: a shared cause can imitate a direct relationship.

A patient-friendly explanation is:

“These conditions appear together more often in observational studies. That does not prove that one caused the other. The good news is that your periodontal disease is treatable for important oral-health reasons, and in some conditions—especially diabetes—treatment may also support a measurable systemic outcome.”

Then make the message disease-specific.

• Diabetes: Ask about control, coordinate care, and explain the modest short-term HbA1c evidence.
• Cardiovascular disease and hypertension: Recognize shared risk, encourage appropriate medical follow-up, and do not promise event prevention.
• Pregnancy: Provide indicated periodontal care without claiming it will prevent preterm birth.
• Cognitive impairment: Simplify maintenance and involve caregivers when appropriate; do not sell dementia prevention.
• IBD and COPD: Consider the medical context without inventing a proven treatment effect.

Most importantly, diagnose periodontitis properly. “Poor oral health,” gingivitis, tooth loss, self-reported gum disease, and stage III periodontitis are not the same exposure. If a paper combines them, its headline should not become more precise than its methods.

We do not need weak causal headlines to make periodontology important.

Periodontitis already matters because it destroys attachment and bone, threatens tooth retention, affects function and quality of life, and requires long-term control. Oral–systemic evidence adds another responsibility: it should make us better at risk detection, patient communication, and interdisciplinary care.

The mouth is connected to the body. The evidence is connected too—from association, to temporality, to mechanism, to intervention, to outcomes.

Our job is not to stop at the most exciting rung.

Our job is to tell patients exactly how high the evidence has climbed.