Identification and reproducibility of dietary patterns in a Danish cohort:: the Inter99 study

We aimed to identify dietary patterns in a Danish adult population and assess the reproducibility of the dietary patterns identified. Baseline data of 3372 women and 3191 men (30-60 years old) from the population-based survey Inter99 was used. Food intake, assessed by a FFQ, was aggregated into thirty-four separate food groups. Dietary patterns were identified by principal component analysis. Confirmatory factor analysis and Bland Altman plots were used to assess the reproducibility of the dietary patterns identified. The Bland Altman plots were used as an alternative and new method. Two factors were retained for both women and men, which accounted for 15.1 -17.4 % of the total variation. The 'Traditional' pattern was characterised by high loadings (>= 0.40) on pate or high-fat meat for sandwiches, mayonnaise salads, red meat, potatoes, butter and lard, low-fat fish, low-fat meat for sandwiches, and sauces. The 'Modern' pattern was characterised by high loadings on vegetables, fruit, mixed vegetable dishes, vegetable oil and vinegar dressing, poultry, and pasta, rice and wheat kernels. Small differences were observed between patterns identified for women and men. The root mean square error approximation from the confirmatory factor analysis was 0.08. The variation observed from the Bland Altman plots of factors from explorative v. confirmative analyses and explorative analyses from two sub-samples was between 18.8 and 47.7%. Pearson's correlation was >0.89 (P<0.0001). The reproducibility was better for women than for men. We conclude that the 'Traditional' and 'Modern' dietary patterns identified were reproducible.