Background Heart failure prevalence increases sharply at older ages. The section of the population aged 85 and over represents the most rapidly increasing demographic worldwide. Previous epidemiological studies of ventricular dysfunction and heart failure have included only small numbers of the “oldest old”, and have generally been conducted in hospital-based settings, potentially introducing ascertainment biases. We conducted a community-based study of the oldest old using domiciliary echocardiography to estimate the prevalence of cardiac dysfunction and heart failure. Since in elderly people with multiple comorbidities, heart failure may more frequently be incorrectly diagnosed, we cross-referenced our findings to preceding diagnoses present in general practice records.

Methods Four hundred and twenty-seven individuals aged 86–89 years (mean age 87.9 years; 39.1% (n=167) men, 60.9% (n=260) women) were visited in their usual place of residence. A full cardiovascular and medical history, including current medication, was taken; symptoms were graded using the NYHA classification. Previous diagnoses of heart failure (HF) were abstracted from the GP record. Participants underwent 2-D and Doppler echocardiography, including tissue Doppler measurements of LV long axis velocities, using a portable instrument (Vivid-I, GE Healthcare). LV systolic and diastolic dysfunction were graded according to American Society of Echocardiography guidelines.

Results LV systolic function could be quantified in 93.2% (n=398) participants and diastolic function (classified as normal, mild, moderate or severe dysfunction) in 88.1% (n=376). 37.2% of participants (n=140/376) had normal cardiac function or isolated mild diastolic dysfunction; 19.6% (n=78/398) had moderate or severe LV systolic dysfunction, which was commoner in men (27.4%) than women (14.5%); and 14.4% (n=54/376) had isolated moderate or severe diastolic dysfunction. 65.1% (278/427) of participants had valid data on previous diagnosis of HF, NYHA class and echocardiographic assessment of cardiac dysfunction. Of these, 37.4% (104/278) had clinical evidence of HF, which was defined as NYHA class II, III, or IV symptoms with underlying systolic dysfunction (29.5% (82/278)) or isolated moderate or severe diastolic dysfunction (7.9% (22/278)) on echo. Only 7.6% (21/278) had a previous diagnosis of HF. 33.1% (n=92/278) had no previous diagnosis of HF but had clinical evidence of HF and an additional 21.6% (n=60/278) had no previous diagnosis but evidence of pre-clinical HF (NYHA class I with systolic or moderate/severe diastolic dysfunction). Of those with a previous diagnosis of HF, 23.5% (n=5/21) had no echocardiographic evidence of cardiac dysfunction.

Conclusions Systolic and diastolic dysfunction and HF were commoner in our population than previous studies in the “younger old” have suggested. There are significant levels of both undiagnosed and misdiagnosed HF in this age group.