The authors explore the relationship between reasons for dental extractions and the occurrence of non-fatal myocardial infarction (MI) in a large cohort of male patients followed for over 30 years. They hypothesised that the level of bacteraemiap- during dental extractions, in the presence of infection, may have a relation to the occurrence of MI.

Previous studies revealed that bacteraemia levels during dental extractions might not be different from bacteraemia from brushing, especially in patients with periodontal disease.1,2

The case control design used was adequate for a harmful outcome such as MI. Cases (n=548) consisted of patients with a self-reported history of MI. Controls (625) were adequately matched by age. Other characteristics that could play a role as confounders like BMI, cholesterol levels and systolic blood pressure were not matched as well.

The history of tooth extractions and oral infections was elicited by a questionnaire, introducing the potential for recall bias and open issues with interpretation.

Interestingly data collected in 2000 showed no statistical differences in the levels of C-reactive protein between groups, while the evaluation of the current oral condition reported the presence of periodontal infection being double in the control group (50%) that of the study group (24%). Levels of C-reactive protein have been widely used in research to monitor effectiveness of periodontal therapy,3,4and they are also used in the medical field as a risk indicator for MI.5

The time of the MI occurrence was not reported. Neither the timeline between the extractions and the MI nor antibiotic usage were reported. A temporal relationship could not be established. More cases had teeth extracted than controls did (92.7% versus 88.6%) during the study.

The results were presented as odds ratio (OR) with a calculated value of 1.73, with 95% CI: 1.34-2.23 between having extractions attributed to infections and developing an MI.

The authors do not discard the possibility of what they call a ‘common underlying trait’ for oral infections and MI as part as the effect observed.

With the methodological limitations of this observational study, the researchers concluded that the results indicate a strong association between the two entities and clearly point to the presence of the underlying infection as the factor of importance. The authors' conclusions may be limited because we can only guess that the extractions preceded the MI and that the nature of the bacteraemia is significantly different during this study from when brushing or flossing.

Practice points

  • No possible conclusion about the relationship between extractions and MI can be established.