Denis Parsons Burkitt (1911-1993)

Denis Parsons Burkitt (1911-1993)
JON A. STORY AND DAVID KRITCHEVSKY*
Department of Foods and Nutrition, Purdue University, West Lafayette, IN 47907-1260
and *The Wistar Institute, Philadelphia, PA 19104
At the beginning of our careers, each of us has the
dream of someday making a discovery that will
change the way our respective disciplines regard
something we now regard as fact. Denis Parsons
Burkitt was able to change the course of our thinking
twice in his career. More remarkably, the two discov
eries were, to the casual observer, in widely different
fields of study, i.e., the elucidation of the origin of a
tumor common in children in parts of Africa and the
observation that dietary fiber deficiency is related to
many of the chronic diseases of the Western world.
Denis Burkitt was born in Enniskillen, Northern
Ireland, on February 28, 1911 and received his
Bachelor of Medicine degree from Trinity College,
studied surgery in Edinburgh, and later received his
medical degree from Dublin University. His father, a
0022-3166/94 $3.00 ©1994 American Institute of Nutrition.
Accepted 20 April 1994. ]. Nutr. 124: 1551-1554, 1994.
surveyor in County Fermanagh, was an ornithologist
and was credited as the first to use banding of birds to
study their territories, an interest in geographical dis
tribution later shared by his son.
After serving in Africa and Sri Lanka as a surgeon
during World War II, Burkitt joined His Majesty’s
Colonial Service in -Africa. It was at Mulago Hospital,
Kampala, Uganda, in the late 1950s that Burkitt was
introduced to a facial tumor usually observed in
young children. The severity of the tumors and the
inadequacy of available treatments led him to further
examine the incidence and distribution of this tumor.
Beginning with a grant of £25,he surveyed hospitals
across Africa by mail, putting together a distribution
map that suggested, for the first time, an ecological
effect on distribution of a cancer. Results of these
early observations of the tumor and of the survey
were published in 1958 (Burkitt) and 1961 (Burkitt
and O’Connor), respectively.
During a trip to England in early 1961 to report his
early results on this tumor, Burkitt met Tony Epstein
of the Bland Sutton Institute at a seminar presented at
Middlesex Hospital. Although Epstein had attended
Burkitt’s seminar almost by chance, he soon realized
that the tumor described by Burkitt might be viral in
origin, supporting his hypothesis of cancers induced
by viruses. At that time Epstein and others had ob
served viral induction of tumors in experimental
animals and had suggested a viral origin for some
human tumors. Epstein and Burkitt met, and Burkitt
agreed to ship biopsies from tumors to Epstein for
further study. This arrangement led to the discovery
of electron microscopic evidence of a virus in the
tumor cells and eventually to th” discovery of the
Epstein-Barr virus, which has since become impli
cated in a wide variety of disease states. Several years
later, when Epstein and Achong published a compi
lation of work dealing with the Epstein-Barr virus,
Epstein gave credit to his attendance at this seminar
as a turning point in its discovery.
Later in 1961, Burkitt and two colleagues departed
on a trip that would take them more than 10,000
miles by car to examine the precise limits of the
occurrence of the tumor in western Africa. This
1551
Downloaded from jn.nutrition.org by guest on July 11, 2013
1552 STORY AND KRITCHEVSKY
monumental task was undertaken with a grant of
£400from several sources, including the British
Medical Research Council (MRC), and was completed
for £678.Initially, altitude and temperature were
thought to be the determining factors, but further
examination suggested rainfall as equally important.
This suggested to Burkitt and his colleagues that the
tumor might be spread by an insect vector. Subse
quent identification of the virus and study of the
lymphoma led to elimination of insects as a direct
vector but demonstrated that the virus was expressed
as a result of the depressed immune system of
children with malaria. It was at a conference spon
sored by the International Union Against Cancer
early in 1963 that the tumor first became commonly
known as Burkitt’s tumor, a name that was later
modified to Burkitt’s lymphoma as a result of a more
thorough understanding of its nature.
Early in 1964, Burkitt resigned his post with
Mulago Hospital to assume a position with the MRC
and make way for African surgeons in the newly
independent Uganda. His new post allowed him to
continue his work with lymphoma patients and
teaching at the Makerere Medical School. He re
mained concerned for treatment of the lymphoma
because surgery only prolonged the patient’s life. The
lack of X-ray treatment facilities in Africa led to
examination of a series of chemotherapeutic agents.
In usual fashion, Burkitt was able to secure a supply
of drugs for his experiments by suggesting to the drug
companies that his patients offered the chance for a
controlled experiment because none had been treated
with X-rays. The success of chemotherapy
(methotrexate, cyclophosphamide and others) was
amazing, with much lower doses being needed than
had been reported for these agents for other tumor
types.
In 1966 Burkitt returned to England and continued
his work with the MRC and, during the following
year, met naval Surgeon Captain T. L. Cleave, who
described his hypothesis concerning refined carbohy
drate as the cause of many of the chronic diseases of
the Western world. Burkitt saw in Cleave’s work a
similarity to his early observations of the lymphoma,
i.e., that several seemingly unrelated diseases might
have a common cause. His own observations of the
absence of these diseases in Africa and his increasing
awareness of their prevalence in Western populations
heightened his interest in this hypothesis.
Burkitt’s foray into the fiber field again entailed
two of his principal traits: a sharp, analytical mind
and unflagging energy. In the preface of Burkitt and
Trowell’s book Refined Carbohydrate Foods and
Disease: Some Implications of Dietary Fibre (1975)
they state, “Our combined experience suggested that
relatively imprecise information can be of epidemiological
value.” The book was dedicated to Surgeon
Captain Cleave.
In the first two chapters of the book Burkitt
delineates the philosophy that led him in his quest for
answers, namely, that one must determine the ge
ographical distribution of a disease in certain commu
nities or in particular groups within that community
that manifest a certain disease and then seek the
environmental factors that are prevalent in areas or
groups exhibiting a high frequency of that disease and
absent where the disease is rare. His observations
(usually illustrated with simplistic cartoons that
made telling points) led him to derive associations
between groups of disease conditions that were as
sociated with Western civilization. From Burkitt’s
work and that of Trowell, there emerged the fiber
hypothesis, which suggests that diets low in fiber may
underlie development of many characteristically
Western diseases.
Two papers published early in Burkitt’s pursuit of
the role of fiber in diseases of the developed world
(Burkitt 1969 and 1971) outlined many of the
epidemiological relationships among gastrointestinal
diseases in various countries. He observed that
similar diseases with similar incidence in similar
populations might have similar causes. The causal
factor that he selected to examine in the populations
he studied was dietary fiber.
Later Burkitt (1979) wrote a short, popular book
that reflected his simple yet forceful writing style and
delineated his nutritional philosophy more
thoroughly. The book was called “Don’t Forget Fibre
in Your Diet” and the subtitle added “to help avoid
many of our commonest diseases.” This book sim
plified the views on fibers and colon cancer that
appeared in his 1975 book with Trowell.
Burkitt hypothesized that the increase in bowel
cancer incidence observed in Western countries was
due to changes in food, meaning reduction in the
amount of cereal fiber and increases in protein, fat,
refined starchy food and sugar. He also suggested that
fecal bacteria might play an important role in the
causation of bowel cancer. The average incidence of
colorectal cancer in the United States, England and
Scotland was, at the time he wrote this book, four
times that in Japan, South India and Uganda. The
level of Bacteroides in the feces was similar for all six
countries, but the three locations with lower cancer
incidence had significantly more Streptococci. Thus
the ratio of Bacteroides to Streptococci in the United
States, England and Scotland was significantly higher
than the ratio in the other three locations. The ratio
of fecal neutral to acidic steroids was much higher in
the three countries with low colon cancer incidence,
and both neutral and acidic steroids were degraded to
a lesser extent in countries with a low incidence.
Burkitt postulated that the amount of carcinogen
formed would depend on variety and number of fecal
bacteria, on the metabolism of bile acids in the gut,
and on the time available for their bacterial degra
dation. Retention of carcinogen in the large bowel
Downloaded from jn.nutrition.org by guest on July 11, 2013
DENIS PARSONS BURKITT 1553
would depend on transit time, frequency of defecation
and inhibition against immediate evacuation when
desire is felt. How many publications in the bowel
cancer field owe their origin to these views?
During this same period of time, Burkitt, with Alec
Walker and Neil Painter, contributed a similar review
to the Journal of the American Medical Association
(1974). In it they described the dietary fiber
hypothesis in a way that undoubtedly launched the
fiber era in the United States. This paper cited several
observations by the authors that indicated a much
lower incidence in Africa of several diseases that were
prevalent in the United States. Included in the list
were heart disease, appendicitis, diverticular disease,
gallstones, varicose veins, hiatus hernia, hemorrhoids,
colon cancer and obesity.
Several hypotheses, based largely on epidemiological
evidence, concerning the potential role of di
etary fiber deficiency in the causation of these dis
orders were presented. A strong case was made for a
link between appendicitis and several other gastroin
testinal diseases (diverticular disease, ulcerative co
litis and colon cancer), with appendicitis being a
predictor of the development of the other problems
within a population. All were related in that a low
fiber diet resulted in increased transit time, reduced
fecal bulk, and drier, less fluid intestinal contents,
which would increase intraluminal pressures, in
crease time of exposure of contents (and potential
toxins) to intestinal mucosa, and cause other similar
changes. Subsequent research has substantiated some
of these hypotheses, but their simplicity made the
message of the relationship of dietary fiber to disease
easily understandable by everyone and created an at
mosphere that resulted in sustainable interest in this
area of research, allowing investigation of these
mechanisms.
Burkitt did not expound at great length on fiber and
heart disease, leaving that to his colleague Hugh
Trowell. He did comment, however, on high fat diets,
high salt intakes, diabetes and obesity as risk factors
for heart disease and suggested that exercise and diets
high in cereal fiber might be protective. The
hypothesis of Trowell and of Burkitt, Walker and
Painter—that lower concentrations of plasma
cholesterol and reduced incidence of heart disease
occurred in populations that consumed high levels of
dietary fiber—was summarized in the Journal review.
They further suggested that these effects resulted
from negative sterol balance as a result of increased
excretion of bile acids, a hypothesis that has been
shown to be true for some sources of dietary fiber. As
with gastrointestinal function, this hypothesis has
been found to be much too simple to explain the
complex relationship between dietary fiber and heart
disease. However, its simplicity made it under
standable to a wide audience and gained support for
the research that was and still is needed to define
these relationships more thoroughly.
The use of epidemiological methods to develop
wide-ranging hypotheses concerning, in this case, di
etary fiber and disease was the essence of Denis
Burkitt. If we examine the literature for evidence of
the impact of Burkitt in the area of dietary fiber we
find a predominance of papers commenting on the
importance of dietary fiber in disease prevention or
suggesting mechanisms by which this effect may be
occurring. In the foreword to Burkitt and Trowell’s
book (1975) Sir Richard Doll wrote, “Once every 10
years or so a new idea emerges about the cause of
disease that captures the imagination and, for a time,
seems to provide a key to the understanding of many
of those diseases whose aetiology was previously un
known… To these we may now add a deficiency of
dietary fiber. But whether it will be as seminal an idea
as that of vitamin deficiency or as sterile as that of
stress, we shall probably not know for another 10
years.”
Almost two decades have passed since those words
were written. The dietary fiber idea has not faded. It
has been refined and redefined, but the basic observa
tions of Denis Burkitt still drive the research. We
know much about fiber analysis, fiber structure and
mechanisms of fiber action under specific physio
logical conditions. We now recognize the presence of
other beneficial substances (carotenoids, phytosterols,
etc.) in a high fiber diet. The discovery of these sub
stances owes something to interest in the fiber
hypothesis.
In addition to his cartoons, Burkitt was well known
for his numerous slides of human feces taken on his
early morning walks in the bush in Africa. He was
often quoted as saying that the health of a country’s
people could be determined by the size of their stools
and whether they floated or sank (complete with a
cartoon), not by their technology. On one occasion he
shocked even the gastroenterologists in the audience
by asking, “How many of you men have any idea of
the size of your wives’ stools?” Although this dis
cussion often led to laughter and to some interesting
responses in the news media, none of us will forget
the message it was intended to deliver, i.e., that in
creased fiber intake would increase stool weight and
reduce stool density, two variables that were epidemiologically
linked with a reduction in the incidence of
several diseases.
Since 1975 the fiber literature has increased ex
ponentially. But this mass of literature originated in
the few lucid papers written on the basis of obser
vation rather than sophisticated research. The stateof-
the-art equipment helps us to learn more about the
fiber phenomenon, but we must never lose sight of
the fact that the origin of this knowledge was a result
of cerebration not instrumentation.
Denis Burkitt received many awards during his
career, being named an honorary fellow of Trinity
College, a Fellow of the Royal Society and a member
of the French Académie de Sciences, and receiving
numerous honorary degrees and, in January 1993, the
Bower Award and Prize in Science. In spite of these
Downloaded from jn.nutrition.org by guest on July 11, 2013
1554 STORY AND KRITCHEVSKY
honors Burkitt remained humble and modest, sug
gesting that in science, as in life, “…attitudes are
more important than abilities, motives… than
methods, character… than cleverness, perseverance…
than power, and the heart always takes precedence
over the head.”
LITERATURE CITED
Burkitt, D. P. (1958) Sarcoma involving jaws in African children. Br.
J. Surg. 46: 218-223.
Burkitt, D. P. (1969) Related disease-related cause. Lancet 2:
1229-1231.
Burkitt, D. P. (1971) Epidemiology of cancer of the colon and
rectum. Cancer 28: 3-13.
Burkitt, D. P. (1979) Don’t Forget Fibre in Your Diet. Martin
Dunitz, London, U.K.
Burkitt, D. P. & O’Connor, G. T. (1961) Malignant lymphoma in
African children. I. A clinical syndrome. Cancer 14: 258-269.
Burkitt, D. P. &. Trowell, H. C., eds. (1975) Refined Carbohydrate
Foods and Disease: Some Implications of Dietary Fibre. Aca
demic Press, London, U.K.
Burkitt, D. P., Walker, A.R.P. & Painter, N. S. (1974) Dietary fiber
and disease. J. Am. Med. Assoc. 229: 1068-1072.
Kellock, B. (1985) The Fiber Man. Lion Publishing, Belleville, MI. by guest