Background: Dyslipidemia, an important risk factor for cardiovascular disease, may be associated with atrial fibrillation (AF). Cross-sectional studies that have examined this association, however, have produced controversial results, and few longitudinal studies have been conducted.
Methods and Results:Using annual health examinations in Japan, the association between lipid profile and the risk of new-onset AF was investigated in the general population. A total of 28,449 individuals who did not have AF at baseline were included in the study. During a follow-up of 4.5±2.7 years, 265 individuals (0.9%) developed AF. In multivariate models, low high-density lipoprotein (HDL) cholesterol was associated with the development of AF in women (hazard ratio [HR], 2.86; 95% confidence interval [CI]: 1.49-5.50) but not in men (HR, 1.35; 95%CI: 0.77-2.38). Women had a 28% higher risk of AF with each 10% decrease in HDL cholesterol. Neither triglycerides nor lipid ratios were associated with AF. After excluding individuals with risk factors for AF, including those who were taking anti-hypertensive drugs, had diabetes, and structural heart disease, the association between low HDL cholesterol and AF remained significant in women.
Conclusions
Methods
SubjectsThis community-based observational cohort study was based on a program of voluntary annual health examinations in the Niigata prefecture of Japan. 5 In this prefecture, annual health examinations supported by the administration are available to residents aged ≥20 years. The population of the prefecture is approximately 2,400,000, and approximately 250,000 residents (approximately 10%) receive the examinations at Niigata Health Foundation every year. The annual examination consists of a detailed medical history obtained through interviews; physical examinations; blood examinations including blood cell count and biochemical markers; urine tests; chest X-rays; and a 12-lead electrocardiogram (ECG). Low-density lipoprotein (LDL) cholesterol was calculated using the Friedewald formula (LDL cholesterol = total cholesterol -HDL cholesterol -0.2 × triglycerides). 10 The HDL cholesterol was subtracted from the total cholesterol to obtain the non-HDL cholesterol. 11 The present study cohort included individuals who did not meet the exclusion criteria, who had had at least one fasting blood test between 1996 and 1998 (which constituted the baseline examination for the present study), and who subsequently received at least one annual examination from the baseline examination to 2005. AF was diagnosed from a 12-lead ECG recorded at an annual follow-up visit. The ECG diagnoses were made by physicians, and any abnormalities were confirmed by cardiologists. Exclusion criteria included a past and/or current history of AF (or atrial flutter), the presence of AF, and having a permanent pacemaker at the initial examination. To study the effects of lipid profile on the development of AF, we excluded individuals who received anti-dyslipidemia drugs that might have affected baseline serum lipid ...