Comparison of cluster and principal component analysis techniques to derive dietary patterns in Irish adults

摘要

The aims of the present study were to examine and compare dietary patterns in adults using cluster and factor analyses and to examine the formatnof the dietary variables on the pattern solutions (i.e. expressed as grams/day (g/d) of each food group or as the percentage contribution to totalnenergy intake). Food intake data were derived from the North/South Ireland Food Consumption Survey 1997–9, which was a randomisedncross-sectional study of 7 d recorded food and nutrient intakes of a representative sample of 1379 Irish adults aged 18–64 years. Cluster analysisnwas performed using the k-means algorithm and principal component analysis (PCA) was used to extract dietary factors. Food data were reduced tonthirty-three food groups. For cluster analysis, the most suitable format of the food-group variable was found to be the percentage contribution tonenergy intake, which produced six clusters: ‘Traditional Irish’; ‘Continental’; ‘Unhealthy foods’; ‘Light-meal foods & low-fat milk’; ‘Healthynfoods’; ‘Wholemeal bread & desserts’. For PCA, food groups in the format of g/d were found to be the most suitable format, and this revealednfour dietary patterns: ‘Unhealthy foods & high alcohol’; ‘Traditional Irish’; ‘Healthy foods’; ‘Sweet convenience foods & low alcohol’. In sum-nmary, cluster and PCA identified similar dietary patterns when presented with the same dataset. However, the two dietary pattern methods requiredna different format of the food-group variable, and the most appropriate format of the input variable should be considered in future studies.