Instead of studying effects of single nutrients or
foods, the questionnaire also opens for the study of
effects of dietary patters. Using principal component
analysis, Brantsæter et al (14) showed that four dietary
patterns emerged with sufficiently high eigenvalues.
These patterns were given the following names based
on loadings to certain foods: Vegetable, processed
food, potato and fish, and cakes and sweets. Subjects
with high factor scores on vegetables had lowest risk
of pre-eclampsia, while having high factor scores on
processed food was associated with increased risk.
Variability in factor scores for potato and fish and
cakes and sweets had little influence on the occurrence
of pre-eclampsia.