Periodontitis, tooth loss and liver cancer: an association in 5.9 million adults

Poor oral health has been linked to several cancers, but the liver has remained an open question. This Swedish study asked whether diagnosed caries, periodontitis and tooth loss are associated with the risk of primary liver cancer (PLC), and whether the two main subtypes behave differently.

The design is a population-based retrospective cohort of 5,893,325 adults registered in the Swedish Dental Health Register between 2009 and 2016, linked to national registers to identify incident liver cancers. Cox regression used attained age as the timescale and adjusted for potential confounders. A sibling comparison analysis was added to control for shared genetic and early-life environmental factors, a useful step in a field where social and behavioural confounding is the usual objection.

Over a mean follow-up of 6.36 years, 2,374 cases of PLC were identified: 1,856 hepatocellular carcinomas (HCC) and 518 non-HCC liver cancers. Periodontitis was associated with a higher risk of PLC (HR 1.25, 95% CI 1.12-1.39), and so was caries (HR 1.15, 95% CI 1.02-1.30). Tooth loss showed a clear dose-response pattern: compared with people who had more than 27 remaining teeth, those with fewer teeth had hazard ratios ranging from 1.15 to 1.84 (p < 0.001). The subtypes diverged: periodontitis was associated with HCC, caries with non-HCC PLC, and tooth loss more strongly with HCC. Results were generally consistent across sibling-controlled, subgroup and sensitivity analyses.

The limits are those of register data. This is an observational association, not proof of cause: residual confounding (alcohol, viral hepatitis, metabolic liver disease, smoking) cannot be fully excluded even with sibling comparisons, and absolute risks remain small, since liver cancer is rare.

Clinical relevance: the size of the cohort, the dose-response with tooth loss and the subtype-specific pattern make this a strong signal in the periodontal-systemic literature. For the periodontist, it does not change cancer screening. It does reinforce that periodontitis and tooth loss are markers worth reading as part of the patient's general health, and a reason to ask about liver risk factors and to collaborate with the physician.