Objective: To compare actual food intake of elderly people in Europe a
nd further identify food patterns, which mediate favourable or unfavou
rable nutrient intakes from cluster analysis of pooled data. Design: C
ross-sectional analysis of food patterns of participants of the SENECA
follow-up study(1993). Setting: Twelve traditional European towns. Su
bjects: Men (n = 647) and women (n = 710) aged 74-79 y in 1993. Method
s: Food intake data were assessed using the same validated dietary his
tory technique in all SENECA towns. Cluster analysis was used to class
ify subjects into groups based on similarities in dietary variables. R
esults: A northern and southern European eating pattern emerged. The s
outhern food pattern appeared to be the most healthful being rich in g
rain, vegetables, fruit, lean meat and olive oil. The north-south grad
ient did, however, not systematically segregate into the same clusters
. All dietary profiles were represented in all sites. The four dietary
profiles, identified by cluster analysis, were: 'Lean and green eater
s': high carbohydrate and vitamin C intake, 'gourmands': high intake o
f energy and nutrients, 'milk drinkers': high intake of calcium and vi
tamin B-2 and small eaters': marginal energy and nutrient consumption.
Marital status, education, smoking, health status and physical activi
ty level differed between clusters. Conclusions: A north-south gradien
t of food patterns was identified. The southern diet agreed better wit
h guidelines for healthy eating. A sufficient energy intake seemed, ho
wever, to be necessary for an adequate nutrient intake, beside a good
choice of food types.