PROCEEDINGS 
OF THE 
NATIONAL ACADEMY OF SCIENCES 
Volume 4 JUNE 15, 1918 Number 6 
EFFECTS OF A PROLONGED REDUCTION IN DIET ON 25 MEN 
I. INFLUENCE ON BASAL METABOLISM AND NITROGEN 
EXCRETION 
By Francis G. Benedict and Paul Roth 
Nutrition Laboratory, Carnegie Institution op Washington, Boston 
Read before the Academy, April 22, 1918 
A year ago, realizing that this nation faced a food shortage, several members 
of the staff of the Nutrition Laboratory decided that positive evidence re- 
garding the effect of a prolonged restriction in diet would give knowledge of 
possible use in an imminent emergency. Such data seemed especially impor- 
tant as exact experimentation on a large number of men and women, including 
many with peculiar dietetic habits and a supposedly low metabolism, had failed 
to indicate that the basal or maintenance metabolism of any particular class 
of persons or, indeed, of any single individual (making due allowance for dif- 
ferences in weight), is materially lower in energy requirement than the basal 
metabolism of the average individual. Since the law of the conservation of 
energy obtains in the human organism, it is clear that with uniform mainte- 
nance metabolism, the food requirement must also be fairly uniform. On the 
other hand, no evidence is available as to the actual effect of a reduction in 
diet, continued over a considerable period of time. Accordingly, a research 
was carried out by the Nutrition Laboratory during the past winter in which 
the effect of a low diet upon a group of normal adults was studied for a period 
of several months. 
If the food intake is reduced below the maintenance level and the basal 
requirement remains constant, it is plain that there must be drafts upon pre- 
viously-stored body reserves. In any study of the effect of a reduced diet, 
since we are dealing primarily with the question of energy rather than with a 
specific ingredient of the food, there must be the strictest control of the diet, 
so that the exact intake of energy over relatively long periods may be known. 
This involves, with human subjects, a degree of personal integrity and veracity 
that cannot be assumed but must be demonstrated. 
149 
