Girls were randomly assigned to studies 1, 2, or 3. Further
randomization was done within each study: either 5 or 6 girls
started with test meal A and 5 or 6 girls with test meal B. Each
test meal was administered as breakfast after the girls had fasted
overnight and again as lunch on 2 consecutive days. The alter-
nate test meals were administered 14 d later according to the
same study protocol. No food or drinks were allowed between
the intake of the labeled breakfast and lunch meals and for 3 h
after the intake of the labeled lunch meals