The association between sociodemographic characteristics and dementia in patients with atrial fibrillation.

Association between socio-demographic factors and dementia risk is studied in general but not for atrial fibrillation (AF) patients. We studied AF patients ≥ 45 years in Sweden 1998-2012 (n = 537,513) using the Total Population Register for socio-demographic factors, the Swedish Cause of Death Register, and the National Patient Register (NPR) for incident dementia. Cox regression with hazard ratios (HR) and 95% confidence intervals (CI) was used for the association between exposure and outcome, adjusting for age and comorbidities. Totally 30,332 patients (5.6%) were diagnosed with dementia during the follow-up (mean 5.4 years). Of these, 14,097 were men (4.9%) and 16,235 were women (6.5%). Lower educational levels (reference: highest level) were associated with increased dementia, HRs (95% CI) for basic school for men 1.23 (1.18-1.29) and women 1.36 (1.30-1.42), and middle-level school for men 1.17 (1.11-1.22) and women 1.28 (1.22-1.34). Divorced men and women (reference: married) showed increased risk of dementia, HR 1.07 (1.01-1.13) and 1.12 (1.06-1.18), respectively, while widowed men showed lower risk, HR 0.84 (0.80-0.88). High deprivation neighborhood socio-economic status (NSES; reference: medium level) was associated with increased dementia in men, HR 1.11 (1.05-1.17), and low deprivation neighborhood socio-economic status (NSES) with increased dementia in men and women, HR 1.12 (1.06-1.18) and 1.18 (1.12-1.24), respectively. Some results were expected, i.e. association between lower educational level and dementia. The higher risk of dementia in low deprivation NSES-areas could be due to a higher awareness about dementia, and subsequent earlier diagnosis and treatment of dementia.