Cardiovascular diseases are largely underreported in Danish centenarians.

Background: the substantial decline in oldest old mortality has led to more people surviving to very old age. As morbidity and disability generally increases with age epidemiological research in ageing has focused on the health of oldest olds. However, most studies are based on self-reported or physician-reported information, not objective health information.

Objective: to estimate and compare the prevalence of cardiovascular diseases (CVDs) in Danish centenarians using three different sources of information: self-reported, physician-reported and objective data.

Design: the population-based clinical-epidemiological study of 100-year-old Danes.

Methods: all eligible participants were interviewed (self-report) in their domicile and offered a clinical examination, including an electrocardiogram (ECG) and blood pressure measurement. Further health information was retrieved from general practitioners' medical files and the Danish National Discharge Register (physician report).

Results: out of 276 eligible, 207 (75%) participated. Blood pressure and ECG were measured in 76 and 69%, respectively. There was poor agreement between self-reported and physician-reported CVDs, and between physician-reported CVDs and clinical objective CVD diagnoses. Only angina pectoris reached a Kappa value of 0.5. ECG revealed twice as many cases of myocardial infarction and ischaemia compared with physician-reported. Using both physician-reported and ECG 95 (46%) of the centenarians suffered from at least one of the diseases myocardial infarction, angina pectoris or atrial fibrillation. Adding physician-reported heart failure and hypertension increased the prevalence of CVD to 80%.

Conclusion: self-reported information largely underestimates the CVD life-time prevalence in Danish centenarians. Objective clinical examinations are necessary to evaluate true disease prevalence in oldest old.