Skip to main content

Table 4 Associations between fat depots patterns, identified through principal component analysis, with incident atrial fibrillation

From: Imaging-based body fat depots and new-onset atrial fibrillation in general population: a prospective cohort study

  Per SD of factor scores Per quartile of factor scores P for trend
Q1 Q2 Q3 Q4
Subcutaneous fat dominant pattern (PC1)
 Model 1 1.27 (1.06–1.53) 1.00 (Ref.) 0.83 (0.47–1.45) 1.35 (0.81–2.25) 1.72 (1.00–2.94) 0.02
 Model 2 1.24 (1.02–1.51) 1.00 (Ref.) 0.75 (0.43–1.35) 1.27 (0.75–2.13) 1.58 (0.91–2.74) 0.04
 Model 3 1.21 (0.98–1.49) 1.00 (Ref.) 0.76 (0.43–1.34) 1.25 (0.74–2.11) 1.50 (0.85–2.65) 0.07
Visceral fat dominant pattern (PC2)
 Model 1 1.20 (0.99–1.50) 1.00 (Ref.) 1.77 (0.95–3.30) 2.07 (1.11–3.85) 1.86 (0.97–3.55) 0.09
 Model 2 1.16 (0.93–1.47) 1.00 (Ref.) 1.83 (0.97–3.45) 1.90 (0.99–3.63) 1.69 (0.85–3.38) 0.20
 Model 3 1.13 (0.90–1.43) 1.00 (Ref.) 1.82 (0.96–3.42) 1.84 (0.96–3.53) 1.58 (0.78–3.18) 0.31
  1. Values are shown as hazard ratios and 95% confidence interval. N = 1297
  2. Abbreviations: SD standard deviation, PC principal component
  3. Model 1 was adjusted for age and sex. Model 2 was additionally adjusted for high-density lipoprotein cholesterol, total cholesterol, smoking, total alcohol intake, lipid-lowering medication and cardiac medication and history of hypertension, left ventricular hypertrophy, diabetes mellitus, heart failure, and coronary heart disease. Model 3 was additionally adjusted for total lean mass