Lunch intake was followed in 31 matched pairs of hospitalized diabetic
patients over four consecutive days. Pairs of patients were matched f
or type and duration of diabetes, gender, age and body mass index. Lun
ches were composed of appetizer, meat, vegetables, starch, cheese, bre
ad and dessert; water, coffee, tea and lemon were available. One patie
nt per pair was randomly ascribed to the experimental group and was se
rved vegetable and starch dishes added with 0.6% monosodium glutamate
(MSG). Lunch intake was measured by weighing amounts served and left-o
vers. Patients in the experimental group ingested more starch food tha
n their matched controls, and less lemon juice and yogurt. However, th
e total energy load at lunch was not different between groups. This ef
fect on meal time food selection replicates earlier observations made
on elderly persons. It is suggested that manipulating palatability of
various foods within a meal, and especially by using MSG, is an effici
ent way to affect food selection in the meal, without inducing hyperph
agia. (C) 1996 Academic Press Limited