9

Determinants of Health
Thomas McKeown

The question of what determines health is very important. If we, as a society, can agree about
what causes health, then we should be able to agree on where our financial resources should be
placed in order to get the best results from our investment. If there is a single most important
lesson to be obtained from studying the history of health, it must be this one.
It turns out, of course, that defining health, meaning well-being, is not as simple as it might
first appear. The World Health Organization says that health is not merely the absence of
disease but an overall state of physical, mental, and social well-being. You may think such a
definition seems utopian. Most scientists and scholars would agree, however, that levels of
disease in a population (morbidity) and the frequency of death at particular ages in a
population (mortality) represent essential, if gross, measures of health. In other words, improved
health is reflected in the decline of mortality.
We can ask this question in a specifically historical way: What has been the role of clinical
medicine in the improvement of health over time? If by medicine we mean biomedical clinical
care including modern inventions such as X rays, antibiotics, and immunizations, then it is
possible to compare mortality rates before and after significant medical interventions were
introduced. This is exactly the topic of Thomas McKeowns famous book The Role of Medicine:
Dream, Mirage, or Nemesis? (1979), as well as of this selection. McKeown demonstrates that the
greatest advancements in health occurred before specific medical interventions became
available. He suggests (but he doesnt have the data to prove this) that more significant factors
regarding health improvement involved better sanitation, better food, and birth spacing. This
argument has been controversial, largely because there is a widespread presumption in our
society that improved health in developed countries is due to advances in medical knowledge and
the technological and scientific breakthroughs in biomedical science. McKeown disputes this
premise.
Medical anthropologist Stephen Kunitz has done a similar, focused study of historical changes
in health status on the Navajo reservation and the impact of improved medical care because of
Indian Health Service facilities (1983). He argues that the determinants of health are more
complicated than McKeown suggests and that access to medical facilities (in this case, a free
health care system) has a measurable although rather small effect. Other influences on the
health of the Navajo of the dominant white society for example, drinking and automobile
accidentsare also important factors.
In the end, the study of health history does not hinge on an either/or decision. Solutions to
health problems are not going to be found in technology but in the more equitable distribution of
adequate food, sanitation, housing, health information, and medical care services.
As you read this selection, consider these questions:

In the United States today, we spend about 15% of our gross national product on
health care. What do you think McKeown would say about this expenditure?

1
 Do many people actually believe the presumption that McKeown is arguing against
that modern health improvements are caused by modern medicine? What do your
friends think are the determinants of health?
What implications does McKeowns discussion have for combating infectious diseases
in the Third World, where they are still a very significant cause of morbidity and
mortality?

Context: Thomas McKeown (19181988) was a medical historian at the University of

Birmingham. Despite his earlier training in biochemistry and medicine, he became famous for
the thesis that the major mortality declines in industrial societies were due to improved living
standards rather than newly developed biomedical technology. Scholars continue to debate his
demographic methodology and the question of whether he focused too much on economic
change, living conditions, and nutrition while neglecting the development of basic clinical care
(that is, nursing and sanitation within hospitals). This classic and controversial thesis also has
political and policy ramifications. Nevertheless, the McKeown thesis stands as a major
contribution to social medicine and public health.

Modern medicine is not nearly as effective as most people believe. It has not been effective,
because medical science and service are misdirected and societys investment in health is
misused. At the base of this misdirection is a false assumption about human health. Physicians,
biochemists, and the general public assume that the body is a machine that can be protected from
disease primarily by physical and chemical intervention. This approach, rooted in 17th-century
science, has led to widespread indifference to the influence of the primary determinants of
human healthenvironment and personal behaviorand emphasizes the role of medical
treatment, which is actually less important than either of the others. It has also resulted in the
neglect of sick people whose ailments are not within the scope of the sort of therapy that interests
the medical professions.
An appraisal of influences on health in the past suggests that the contribution of modern
medicine to the increase of life expectancy has been much smaller than most people believe.
Health improved, not because of steps taken when we are ill, but because we become ill less
often. We remain well, less because of specific measures such as vaccination and immunization
than because we enjoy a higher standard of nutrition, we live in a healthier environment, and we
have fewer children.
For some 300 years an engineering approach has been dominant in biology and medicine and
has provided the basis for the treatment of the sick. A mechanistic concept of nature developed
in the 17th century led to the idea that a living organism, like a machine, might be taken apart
and reassembled if its structure and function were sufficiently understood. Applied to medicine,
this concept meant that understanding the bodys response to disease would allow physicians to
intervene in the course of disease. The consequences of the engineering approach to medicine are
more conspicuous today than they were in the 17th century largely because the resources of the
physical and chemical sciences are so much greater. Medical education begins with the study of
the structure and function of the body, continues with examination of disease processes, and ends
with clinical instruction on selected sick people. Medical service is dominated by the image of
the hospital for the acutely ill, where technological resources are concentrated. Medical research
also reflects the mechanistic approach, concerning itself with problems such as the chemical
basis of inheritance and the immunological response to transplanted tissues.

2
 No one disputes the predominance of the engineering approach in medicine, but we must now
ask whether it is seriously deficient as a conceptualization of the problems of human health. To
answer this question, we must examine the determinants of human health. We must first discover
why health improved in the past and then go on to ascertain the important influences on health
today, in the light of the change in health problems that has resulted from the decline of
infectious diseases.
It is no exaggeration to say that health, especially the health of infants and young children, has
been transformed since the 18th century. For the first time in history, a mother knows it is likely
that all her children will live to maturity. Before the 19th century, only about three out of every
10 newborn infants lived beyond the age of 25. Of the seven who died, two or three never
reached their first birthday, and five or six died before they were six. Today, in developed
countries fewer than one in 20 children die before they reach adulthood.

FIGURE 9.1 Life expectancy

The increased life expectancy, most evident for young children, is due predominantly to a
reduction of deaths from infectious diseases. Records from England and Wales (the earliest
national statistics available) show that this reduction was the reason for the improvement in
health before 1900, and it remains the main influence to the present day.
But when we try to account for the decline of infections, significant differences of opinion
appear. The conventional view attributes the change to an increased understanding of the nature
of infectious disease and to the application of that knowledge through better hygiene,
immunization, and treatment. This interpretation places particular emphasis on immunization
against diseases like smallpox and polio, and on the use of drugs for the treatment of other
diseases, such as tuberculosis, meningitis, and pneumonia. These measures, in fact, contributed
relatively little to the total reduction of mortality; the main explanation for the dramatic fall in
the number of deaths lies not in medical intervention, but elsewhere.
Deaths from the common infections were declining long before effective medical intervention
was possible. By 1900, the total death rate had dropped substantially, and over 90% of the
reduction was due to a decrease of deaths from infectious diseases. The relative importance of
the major influences can be illustrated by reference to tuberculosis. Although respiratory
tuberculosis was the single largest cause of death in the mid-19th century, mortality from the

3
disease declined continuously after 1838, when it was first registered in England and Wales as a
cause of death.
Robert Koch identified the tubercle bacillus in 1882, but none of the treatments used in the
19th or early 20th centuries significantly influenced the course of the disease. The many drugs
that were tried were worthless; so, too, was the practice of surgically collapsing an infected lung,
a treatment introduced about 1920. Streptomycin, developed in 1947, was the first effective
treatment, but by this time mortality from the disease had fallen to a small fraction of its level
during 1848 to 1854. Streptomycin lowered the death rate from tuberculosis in England and
Wales by about 50%, but its contribution to the decrease in the death rate since the early 19th
century was only about 3%.

FIGURE 9.2 Infant mortality rate

Deaths from bronchitis, pneumonia, and influenza also began to decline before medical
science provided an effective treatment for these illnesses. Although the death rate in England
and Wales increased in the second half of the 19th century, it has fallen continuously since the
beginning of the 20th. There is still no effective immunization against bronchitis or pneumonia,
and influenza vaccines have had no effect on deaths.
The first successful treatment for these respiratory diseases was a sulfa drug introduced in
1938, but mortality attributed to the lung infections was declining from the beginning of the 20th
century. There is no reason to doubt that the decline would have continued without effective
therapeutic measures, if at a slower rate.
In the United States, the story was similar; Thomas Magill noted that the rapid decline of
pneumonia death rates began in New York State before the turn of the century and many years
before the miracle drugs were known. Obviously, drug therapy was not responsible for the
total decrease in deaths that occurred since 1938, and it could have had no influence on the
substantial reduction that occurred before then.
The histories of most other common infections, such as whooping cough, measles, and scarlet
fever, are similar. In each of these diseases, mortality had fallen to a low level before effective
immunization or therapy became available.
In some infections, medical intervention was valuable before sulfa drugs and antibiotics
became available. Immunization protected people against smallpox and tetanus; antitoxin
treatment limited deaths from diphtheria; appendicitis, peritonitis, and ear infections responded

4
to surgery; Salvarsan was a long-sought magic bullet against syphilis; intravenous therapy
saved people with severe diarrheas; and improved obstetric care prevented childbed fever.
But even if such medical measures had been responsible for the whole decline of mortality
from these particular conditions after 1900 (and clearly they were not), they would account for
only a small part of the decrease in deaths attributed to all infectious diseases before 1935. From
that time, powerful drugs came into use and they were supplemented by improved vaccines. But
mortality would have continued to fall even without the presence of these agents; and over the
whole period since cause of death was first recorded, immunization and treatment have
contributed much less than other influences.
The substantial fall in mortality was due in part to reduced contact with microorganisms. In
developed countries an individual no longer encounters the cholera bacillus, he is rarely exposed
to the typhoid organism, and his contact with the tubercle bacillus is infrequent. The death rate
from these infections fell continuously from the second half of the 19th century, when basic
hygienic measures were introduced: purification of water; efficient sewage disposal; and
improved food hygiene, particularly the pasteurization of milk, the item in the diet most likely to
spread disease.
Pasteurization was probably the main reason for the decrease in deaths from gastroenteritis
and for the decline in infant mortality from about 1900. In the 20th century, these essential
hygienic measures were supported by improved conditions in the home, the work place, and the
general environment. Over the entire period for which records exist, better hygiene accounts for
approximately a fifth of the total reduction of mortality.
But the decline of mortality caused by infections began long before the introduction of
sanitary measures. It had already begun in England and Wales by 1838, and statistics from
Scandinavia suggest that the death rate had been decreasing there since the first half of the 18th
century.
A review of English experience makes it unlikely that reduced exposure to microorganisms
contributed significantly to the falling death rate in this earlier period. In England and Wales that
was the time of industrialization, characterized by rapid population growth and shifts of people
from farms into towns, where living and working conditions were uncontrolled. The crowding
and poor hygiene that resulted provided ideal conditions for the multiplication and spread of
microorganisms, and the situation improved little before sanitary measures were introduced in
the last third of the century.

5
FIGURE 9.3 Child death rates reduced by medicine

Another possible explanation for the fall in mortality is that the character of infectious diseases
changed because the virulence of microorganisms decreased. This change has been suggested in
diseases as different as typhus, tuberculosis, and measles. There is no infection of which it can be
said confidently that the relationship between host and parasite has not varied over a specified
period. But for the decline of all infections, this explanation is obviously inadequate, because it
implies that the modern improvement in health was due essentially to a fortuitous change in the
nature of the infections, independent of medical and other identifiable influences.
A further explanation for the falling death rate is that an improvement in nutrition led to an
increase in resistance to infectious diseases. This is, I believe, the most credible reason for the
decline of the infections, at least until the late 19th century, and also explains why deaths from
airborne diseases like scarlet fever and measles have decreased even when exposure to the
organisms that cause them remains almost unchanged. The evidence demonstrating the impact of
improved nutrition is indirect, but it is still impressive.
Lack of food, and the resulting malnutrition were largely responsible for the predominance of
the infectious diseases, from the time when men first aggregated in large population groups
about 10,000 years ago. In these conditions an improvement in nutrition was necessary for a
substantial and prolonged decline in mortality.

6
FIGURE 9.4 Other childhood diseases

Experience in developing countries today leaves no doubt that nutritional state is a critical
factor in a persons response to infectious disease, particularly in young children. Malnourished
people contract infections more often than those who are well fed and they suffer more when
they become infected. According to a recent World Health Organization report on nutrition in
developing countries, the best vaccine against common infectious diseases is an adequate diet.

FIGURE 9.5 Pulmonary diseases

In the 18th and 19th centuries, food production increased greatly throughout the Western
world. The number of people in England and Wales tripled between 1700 and 1850, and they
were fed on homegrown food.
In summary: The death rate from infectious diseases fell because an increase in food supplies
led to better nutrition. From the second half of the 19th century this advance was strongly
supported by improved hygiene and safer food and water, which reduced exposure to infection.

7
With the exception of smallpox vaccination, which played a small part in the total decline of
mortality, medical procedures such as immunization and therapy had little impact on human
health until the 20th century.
One other influence needs to be considered: a change in reproductive behavior, which caused
the birth rate to decline. The significance of this change can hardly be exaggerated, for without it
the other advances would soon have been overtaken by the increasing population. We can
attribute the modern improvement in health to food, hygiene, and medical intervention in that
order of time and importancebut we must recognize that it is to a modification of behavior that
we owe the permanence of this improvement.
But it does not follow that these influences have the same relative importance today as in the
past. In technologically advanced countries, the decline of infectious diseases was followed by a
vast change in health problems, and even in developing countries advances in medical science
and technology may have modified the effects of nutrition, sanitation, and contraception. In order
to predict the factors likely to affect our health in the future, we need to examine the nature of the
problems in health that exist today.
Because todays problems are mainly with noncommunicable diseases, physicians have
shifted their approach. In the case of infections, interest centers on the organisms that cause them
and on the conditions under which they spread. In noninfective conditions, the engineering
approach established in the 17th century remains predominant, and attention is focused on how a
disease develops rather than on why it begins. Perhaps the most important question now
confronting medicine is whether the commonest health problemsheart disease, cancer,
rheumatoid arthritis, cerebrovascular diseaseare essentially different from health problems of
the past or whether, like infections, they can be prevented by modifying the conditions that lead
to them.
To answer this question, we must distinguish between genetic and chromosomal diseases
determined at the moment of fertilization and all other diseases, which are attributable in greater
or lesser degree to the influence of the environment. Most diseases, including the common
noninfectious ones, appear to fall into the second category. Whether these diseases can be
prevented is likely to be determined by the practicability of controlling the environmental
influences that lead to them.
The change in the character of health problems that followed the decline of infections in
developed countries has not invalidated the conclusion that most diseases, both physical and
mental, are associated with influences that might be controlled. Among such influences, those
which the individual determines by his own behavior (smoking, eating, exercise, and the like) are
now more important for his health than those that depend mainly on societys actions (provision
of essential food and protection from hazards). And both behavioral and environmental
influences are more significant than medical care.
The role of individual medical care in preventing sickness and premature death is secondary to
that of other influences; yet societys investment in health care is based on the premise that it is
the major determinant. It is assumed that we are ill and are made well, but it is nearer the truth to
say that we are well and are made ill. Few people think of themselves as having the major
responsibility for their own health, and the enormous resources that advanced countries assign to
the health field are used mainly to treat disease or, to a lesser extent, to prevent it by personal
measures such as immunization.
The revised concept of human health cannot provide immediate solutions for the many
complex problems facing society: limiting population growth and providing adequate food in

8
developing countries, changing personal behavior and striking a new balance between
technology and care in developed nations. Instead, the enlarged understanding of health and
disease should be regarded as a conceptual base with implications for services, education, and
research that will take years to develop.
The most immediate requirement in the health services is to give sufficient attention to
behavioral influences that are now the main determinants of health. The public believes that
health depends primarily on intervention by the doctor and that the essential requirement for
health is the early discovery of disease. This concept should be replaced by recognition that
disease often cannot be treated effectively and that health is determined predominantly by the
way of life individuals choose to follow. Among the important influences on health are the use of
tobacco, the misuse of alcohol and drugs, excessive or unbalanced diets, and lack of exercise.
With research, the list of significant behavioral influences will undoubtedly increase, particularly
in relation to the prevention of mental illness.
Although the influences of personal behavior are the main determinants of health in developed
countries, public action can still accomplish a great deal in the environmental field.
Internationally, malnutrition probably remains the most important cause of ill health, and even in
affluent societies sections of the population are inadequately, as distinct from unwisely, fed. The
malnourished vary in proportion and composition from one country to another, but in the
developed world they are mainly the younger children of large families and elderly people who
live alone. In light of the importance of food for good health, governments might use
supplements and subsidies to put essential foods within the reach of everyone, and provide
inducements for people to select beneficial in place of harmful foods. Of course these aims
cannot exclude other considerations such as international agreements and the solvency of farmers
who have been encouraged to produce meat and dairy products rather than grains. Nevertheless,
in future evaluations of agricultural and related economic policies, health implications deserve a
primary place.
Perhaps the most sensitive area for consideration is the funding of the health services.
Although the contribution of medical intervention to prevention of sickness and premature death
can be expected to remain small in relation to behavioral and environmental influences, surgery
and drugs are widely regarded as the basis of health and the essence of medical care, and society
invests the money it sets aside for health mainly in treatment for acute diseases and particularly
in hospitals for the acutely ill. Does it follow from our appraisal that resources should be
transferred from acute care to chronic care and to preventive measures?
Restricting the discussion to personal medical care, I believe that neglected areas, such as
mental illness, mental retardation, and geriatric care, need greatly increased attention. But to
suggest that this can be achieved merely by direct transfer of resources is an oversimplification.
The designation acute care comprises a wide range of activities that differ profoundly in their
effectiveness and efficiency. Some, like surgery for accidents and the treatment of acute
emergencies, are among the most important services that medicine can offer and any reduction of
their support would be disastrous. Others, however, like coronary care units and iron treatment of
some anemias are not shown to be effective, while still othersmost tonsillectomies and routine
check-upsare quite useless and should be abandoned. A critical appraisal of medical services
for acute illnesses would result in more efficient use of available resources and would free some
of them for preventive measures.
What health services need in general is an adjustment in the distribution of interest and
resources between prevention of disease, care of the sick who require investigation and

9
treatment, and care of the sick who do not need active intervention. Such an adjustment must pay
considerable attention to the major determinants of health: to food and the environment, which
will be mainly in the hands of specialists, and to personal behavior, which should be the concern
of every practicing doctor.

References
Burnet, M. 1971. Genes, Dreams and Realities. New York: Basic Books.
Cochrane, A. L. 1972. Effectiveness and Efficiency. London: Nuffield Provincial Hospitals Trust.
Dubos, R. 1971. Mirage of Health. New York: Harper & Row.
McKeown, T. The Role of Medicine: Dream, Mirage or Nemesis? London: Nuffield Provincial
Hospitals Trust.
. 1977. The Modern Rise of Population. San Diego: Academic Press.
Thomas, L. 1974. The Lives of a Cell: Notes of a Biology Watcher. New York: Viking Press.

10