My interest in stress and cancer began around 55 years ago, when I had a Fellowship at Hans Selye’s Institute of Experimental Medicine and Surgery at the University of Montreal. His magnum opus Stress had just been published in 1950, expanding on his theory of the “General Adaptation Syndrome” and its resultant “Diseases of Adaptation”. One of the hallmarks of his “Alarm Reaction”, the first phase of this syndrome, was marked dissolution of lymphoid tissue and atrophy of the thymus gland. However, the significance of this was not clear with respect to its implications for altered immune system function that might predispose to malignancy.
During dinner at his home one evening, Selye indicated that he had been intrigued with the possible relationship between stress and cancer. He cited various anecdotal reports of the onset of malignancy following emotional stress. He was fascinated by the phenomenon of spontaneous remission, which he thought was due to a strong faith that buffered the harmful effects of stress, and we discussed various aspects of this with respect to its relevance to the “General Adaptation Syndrome”. I suggested to him that cancer might possibly represent another “Disease of Adaptation”, or a response to stress that had gone awry, but our conversation soon turned to more pressing priorities.
The subject did not come up again until over a quarter century later in 1977, when Selye’s International Institute of Stress and Sloan-Kettering Institute co-sponsored a symposium entitled Cancer, Stress, and Death. Selye called to indicate that he would be coming to New York and wanted to have dinner with me to discuss my participation in this event. When we met, he suggested that I contribute a paper on possible relationships between stress and cancer, particularly with respect to the hypothesis that cancer might represent another “Disease of Adaptation”. I pointed out that I had been involved in clinical practice for over two decades, and no longer had the time, resources, or training required for such an undertaking. In addition my immediate reaction was that attempting to prove any relationship between stress and cancer in humans would be an insurmountable task for numerous reasons. There was still no satisfactory scientific definition of stress. Cancer comprised a variety of distinct malignancies that most likely had different etiologies. Cancer also caused stress and it might be difficult to determine which came first. More importantly it was obviously impossible to establish exactly when malignant growth began as opposed to when it was first detected, making it equally difficult to prove any temporal relationship.
However, as usual, Selye was persistent and persuasive, and reminded me of our dinner discussion at his home. He had also brought a reprint of “The Growth and Development of the Stress Concept”, a chapter he had asked me to write for Modern Trends in Endocrinology in 1958, where I had again referred to the possibility that cancer might represent a “Disease of Adaptation”. During dessert and coffee, he confided that he had a very personal interest in this subject. About five years previously, he had been diagnosed as having a histiocytic reticulosarcoma, a tumor that is usually fatal. He attributed his apparent complete recovery to his own faith and determination to remain well in order to continue his research, rather than any surgery or cobalt therapy. Since then, he had been collecting a great deal of material dealing with the subject of stress and cancer, and asked if I would at least look it over before turning down his request. I acquiesced but again emphasized why I was ill equipped to do justice to this complex topic. We reminisced about so many other things as I drove him back to the airport that I dismissed the matter completely.
However, two weeks later Selye sent me a warm note offering to put his research staff at my disposal, together with a package of reprints containing comments penciled in at various locations that were cleverly designed to pique my curiosity. I would normally not have had the time to go through it thoroughly but was able to do so since it serendipitously coincided with a two-week vacation. I found it fascinating reading and after checking some of the references cited that had not been included, which his staff quickly supplied I was even more intrigued. I had taken the bait hook, line and sinker and was now anxious to pursue the assignment for several reasons.
The belief that cancer might in some way be related to stress or distressful emotions is as old as the history of recorded medicine. Over 2,000 years ago, in his dissertation on tumors, De Tumoribus, Galen noted that women who were melancholy were much more susceptible to cancer than other females, presumably because they had too much black bile (mélas chole). It was difficult to find much written about cancer in the English medical literature until 1701, at which time a British physician, Gendron, emphasized the effect of “disasters of life as occasion much trouble and grief” in the causation of cancer. Eighty years later, Burrows attributed the disease to “the uneasy passions of the mind with which the patient is strongly affected for a long time.”
Early nineteenth century physicians such as Nunn emphasized that emotional factors influenced the growth of tumors of the breast, and Stern noted that cancer of the cervix in women was more common in sensitive and frustrated individuals. Walshe’s major treatise The Nature and Treatment of Cancer called attention to the “influence of mental misery, sudden reverses of fortune and habitual gloomings of the temper on the disposition of carcinomatous matter. If systematic writers can be credited, these constitute the most powerful cause of the disease.” One hundred years ago, Snow’s review of over 250 patients at the London Cancer Hospital concluded that “the loss of a near relative was an important factor in the development of cancer of the breast and uterus.”
I attach particular importance to these observations, particularly the last, because the practice of medicine one or two hundred years ago was much more personalized. Physicians had to rely more upon their own understanding of the significance of the history, emotional background, and life‑style of the patient, in contrast to today’s emphasis on detached diagnostic high tech laboratory and imaging procedures. In addition, their education was more apt to include a heavy background in literature, the humanities, and philosophy, rather than the current accent on basic science. They were much more likely to be familiar with the patient’s family and social relationships, and the influences of other psychosocial environmental factors. They also spent much more time observing and talking to patients, and asking pertinent questions about such details, than is possible in the frenetic pace of today’s super specialized and relatively perfunctory practice setting. Thus, by virtue of education, orientation, and a more personalized approach, they might well be expected to have had a greater sensitivity and appreciation of certain subtle nuances that could suggest any possible relationship between emotional stress and cancer.
During the 20th century, emphasis shifted to external agencies as the cause of cancer. Currently, a host of carcinogens in the air we breathe, the foods we ingest, or various viruses have been incriminated. All of these approaches imply some physical assault on us from without, consistent with the germ theory of disease, which is quite understandable. Pasteur’s discovery of microbes and clinical achievements, and the proof afforded by Koch’s Postulates confirmed the direct causal relationships between microorganisms and infectious diseases. The subsequent success of various vaccines and dramatic life‑saving effects of antibiotics seemed to settle any doubts. People became sick because something attacked them from without. Little attention was directed to what determined resistance or susceptibility to disease. Few questioned why certain individuals, similarly exposed to the same tubercle bacillus, hepatitis virus, or carcinogen, remained healthy.
Nevertheless, over the past several decades, numerous clinical and animal research studies have continued to confirm the important influences stressful emotions can exert with respect to the development and progression of different diseases, and particularly malignant growth. Some of the major characteristics of cancer prone individuals appear to be frequent feelings of hopelessness and helplessness, an inability to express anger or resentment, an unusual amount of self dislike and distress, and having suffered the loss of a meaningful emotional relationship. Everson et al. evaluated hopelessness in some 2500 men and found that six years later those who scored high were almost 3.5 time higher to have died from cancer or heart disease. For purposes of this discussion, I should like to concentrate on Snow’s observation about the significance of loss of an important emotional relationship as a precursor to cancer.
Implicit in Cannon’s “fight or flight” theory, is the teleological premise that our automatic and involuntary responses to stress have been progressively developed over the lengthy course of man’s evolution. It is postulated that they represent adaptive changes which were essential for the survival of our ancestors when faced with a life threatening physical threat. The outpouring of adrenalin and stimulation of the sympathetic nervous system caused the pupils to dilate and promote better vision, blood clotting was quickened to reduce loss from lacerations or internal hemorrhage, blood pressure and heart rate rose increasing flow of blood to the brain to facilitate decision making, and carbohydrate and fat stores in the body were broken down raising blood sugar and lipid levels to provide fuel for more energy. The circulation of blood was shunted away from the gut where it was not immediately needed for purposes of digestion, to the large muscles of the extremities. This produced greater tension and strength in the arms and legs to assist in physical combat, or speed of locomotion away from a scene of potential peril.
However, the nature of stress for modern man is not a potentially lethal, physical encounter with a sabre toothed tiger or a warring tribe every few months, but rather a host of emotional stresses which often occur several times a day. The tragedy is that these still often result in the same “fight or flight” responses which are now no longer appropriate or purposeful. Repeatedly invoked, it is not difficult to understand how they could contribute to “Diseases of Civilization” such as hypertension, diabetes, heart attack, strokes, peptic ulcers, muscle spasms, etc. Many of our responses to stress don’t seem to make any sense in terms of having ever provided any benefits. When severely frightened, some people experience “goose flesh”, or the hairs on the back of the neck may stand up, and what good does that do?. However, the stimulation of those same arrector pili muscles is responsible for the flying fur on the arched back of an aroused cat, which makes it look more ferocious to an assailant. They also produce the bristling of the quills of the porcupine, providing a very effective defense mechanism. Thus, all of our responses to stress undoubtedly served some useful purpose at some time during the lengthy course of human evolution.
It is equally apparent that we often overreact to a stimulus with responses that are damaging. We see this in the occasional development of disfiguring keloids during excessive scar formation in wound healing. Similarly, lip cancer may develop in clay pipe smokers at the site of heat injured tissue which is attempting to repair itself. There are other instances where adaptational evolutionary changes may eventually prove detrimental. In my 1958 chapter discussing Selye’s concept of “Diseases of Adaptation” I referred to the theory of “opportunism” in the evolutionary process. This refers to the organism’s response to fill a need with whatever means are available, even if that response may ultimately prove harmful. The illustration cited at that time was the tremendous variation in the development of different horns by some twenty-three species of African Antelopes. Some horns are obviously too small to be effective, such as those of the duiker, while others are prohibitively unwieldy, as in the kudu. As one examines this tremendous variation, the marked alterations in anatomical configuration and functional effect do not appear to serve any useful or rational adaptive purpose, and are more of a detriment. If I were to rewrite that article today, I would select the development of malignancy in man as perhaps a more dramatic example of “opportunism” in the evolutionary process, for the following reasons.
As one descends the phylogenetic scale, the incidence of cancer progressively decreases, and it is absent in primitive forms of life. Conversely, the ability of the organism to regenerate injured or lost tissues increases proportionately. Simpler organisms, including some invertebrates, are able to sever parts of their anatomy when they are injured. Obviously, this capability would have survival value only if the animal possessed an equally remarkable ability to regenerate the cast off portion from available cell remnants. Thus, a starfish can grow a new appendage, and the salamander or newt can grow a new tail or leg if it is severed. Humans, however, do not have such reparative or regenerative powers, except perhaps for the liver and spleen which are similar in nature to organs found in lower forms of life.
I believe that some cancers may represent a vestigial remnant of this primitive, purposeful, regenerative potential. When we suffer a loss or injury, an attempt to respond with similar purposeful replacement activities is triggered. Unfortunately, this new growth, or neoplasia, may prove to be harmful rather than functional. Experiments with chemicals known to cause cancer when applied to the skin or injected into laboratory animals and humans support this hypothesis. When these same carcinogens are injected into the leg of a salamander, it does not result in cancer, but surprisingly causes the growth of a new accessory limb at that site. If injected into the lens of the eye, the salamander will regenerate a new lens. Thus, the identical carcinogenic stimulus can produce either purposeful regeneration, or a fatal malignancy, depending upon the evolutionary development of the organism.
The emotional distress associated with an anticipated traumatic incident is often greater than that encountered as a result of the physical event itself. Some examples are a child awaiting a well deserved spanking, or sitting in the dentist’s waiting room prior to some procedure that proves practically painless. Therefore, the leap from physical to emotional loss should not be troublesome. The ability to regenerate lost or injured tissue in lower forms of life obviously involves something more than a simple local response. The message that tissue has been damaged or lost must be relayed to higher centers in the central nervous system which then initiate appropriate and coordinated reparative responses. With man’s highly developed cerebral cortex, emotional loss may well be perceived as being as significant or even greater stress than a physical separation. The same signals may be sent to activate endocrine, immune, and central nervous system mechanisms to continue to respond in some manner to repair the damage. However, our attempts to stimulate replacement or purposeful new growth are futile. What may result instead, is new growth in the form of neoplasia which is malignant and beyond control.
In the Holmes-Rahe Scale, the four most stressful life change events all involve loss of important emotional relationships, with death of a spouse and divorce heading the list. If stress can cause cancer, one would therefore expect that affected individuals would demonstrate significantly higher rates of malignancy. It has long been recognized that widowed and divorced individuals die at much higher rates for all the leading causes of death including cancer. It is also quite clear that depression of immune system function predisposes to cancer, as is vividly illustrated by a host of AIDS related malignancies, including the rare Kaposi’s sarcoma. Over the past two decades, a variety of studies have demonstrated that following loss of a spouse there is a prompt and impressive decline in immune system defenses, and possibly, this is aberrant adaptive response is a mechanism that may explain some stress related malignancies.
There is also evidence that increased stresses associated with progressive civilization, contribute to cancer. I do not refer here to such things as smoking, air pollution, asbestos, radiation hazards, or other carcinogenic concerns, but rather to psychosocial stresses that were evident long before these modern problems. This concept is far from new, and was proposed in Tanchou’s “Memoir on the Frequency of Cancer” delivered to the French Academy of Sciences over one hundred and sixty years ago. Tanchou noted that “cancer like insanity increases in a direct ratio to the civilization of the country”. He noted that in Paris, the annual cancer mortality rate over an eleven year period was .80 per thousand. While it was only .2 per thousand in London. Thus he proudly concluded that the data “proved that Paris is four times more civilized than London”. Powell’s The Pathology of Cancer (1908) stated: “There can be little doubt that the various influences grouped under the title of civilization play a part in producing a tendency to Cancer.” Similarly, Roberts wrote in Malignancy and Evolution (1926), “I take the view commonly held that, whatever its origin, cancer is very largely a disease of civilization”.
The renowned medical missionary, Dr. Albert Schweizer, wrote “on my arrival in Gabon in 1913, I was astonished to find no cases of cancer”, over the years, cases began to appear in growing numbers, and he concluded “my observations incline me to attribute this to the fact that the natives are living more and more after the manner of the whites”.
The celebrated anthropologist and Arctic explorer, Vilhjalmur Stefansson, in his book which was actually entitled, Cancer: Disease of Civilization?, noted the absence of cancer in the Eskimos upon his arrival in the Arctic, but a subsequent increase in the incidence of the disease as closer contact with white civilization was established. He quoted Sir Robert McCarrison, a physician who had studied 11,000 Hunza natives in Kashmir from 1904-1911. Cancer was unknown, and these individuals seemed to preserve their youthful physique and appearance well into their sixties and seventies, and to enjoy unusual longevity. McCarrison attributed this to the fact that they were “far removed from the refinement of civilization…..and endowed with a nervous system of notable stability”. Both Stefansson and Schweitzer believed this had nothing to do with diet, but resulted entirely from the stresses associated with progressive civilization.
In an July, 1927 article in Cancer, Dr. William Howard Hay noted: “A study of the distribution of cancer, among the races of the entire earth, shows a cancer ratio in about proportion to which civilization living predominates; so evidently something inherent in the habits of civilization is responsible for the difference of cancer incidence compared with the uncivilized races and tribes. Climate has nothing to do with this difference, as witness the fact that tribes living naturally will show a compete absence until mixture with more civilization, even so does cancer begin to show its head”. One of the most persuasive arguments is to be found in Dr. Alexander Berglas’ work, Cancer; Its Nature, Cause and Cure (1957). Throughout this book runs the theme that cancer is a disease from which primitive peoples are relatively or wholly free, and that we are “threatened with death from cancer because of our inability to adapt to present day living conditions. Over the years, cancer research has become the domain of specialists in various fields. Despite the outstanding contributions of scientists, we have been getting farther away from our goal, the curing of cancer. This specialized work, and the knowledge gained through the study of individual processes, has had the peculiar result of becoming an obstacle to the whole. More than thirty years in the field of cancer research have convinced me that it is not to our advantage to continue along this road of detailed analysis. I have come to the conclusion that cancer may perhaps be just another intelligible natural process whose cause is to be found in our environment and mode of life”.
Our latest government figures report a puzzling increase in the incidence of breast cancer in middle-aged females. The experts have no explanation, but I believe this may also be related to the stress of “civilization”. It has been well established that the younger a woman is when she has her first child or even becomes pregnant, the less likely she is to develop breast cancer. Pregnancy lowers prolactin, a pituitary hormone that stimulates breast tissue growth and promotes breast cancer in experimental animals. As more and more women enter the work force, they tend to remain single, marry and decide not to have children, or do so only when they are much older. The per cent of women having their first child after the age of 35 has more than quintupled since 1970. Similarly, career oriented women, especially those with no children, have a much higher incidence of deadly ovarian cancer. Single working women have fourteen times the average risk of ovarian cancer than a matched group of homemakers. Job stress itself may be a factor, sometimes because of overt and covert sexual harassment. Many married women have to juggle work responsibilities with being a wife, supermom, single parent, or providing custodial duties for an aging parent or relative. In addition, they find that despite equal or superior training, experience and ability, they are paid less than their male counterparts, and usually reach a dead end when they try to reach the upper rungs of the corporate ladder. Other demographic groups ranging from children, adolescents and the elderly also have unique stresses not experienced generations ago as a consequence of changes imposed by the pressures of contemporary civilization. One can only speculate as to whether this may have also have implications for an increase in certain malignancies.
Is it all bad news? I don’t think so. All the great integrative systems of the body operate on a system of checks and balances. The autonomic nervous system has balancing antagonistic but complementary sympathetic and parasympathetic components. The endocrine system is regulated by feedback mechanisms between pituitary and target gland hormones that operate much like a thermostat to maintain homeostasis. We know much less about how such homeostasis is achieved in the immune or central nervous system, but it appears plausible that if distress can cause adverse effects, there is quite likely good stress, or what Selye termed “eustress” that promotes health. Sir William Osler noted that the course of tuberculosis depended more on “what the patient has in his head than what he has in his chest”. Ishigami in Japan came to a similar conclusion in his paper “The Influence of Psychic acts On The Progress of Pulmonary Tuberculosis”, which appeared in the American Review of Tuberculosis in 1919. Some stable patients often deteriorated and died after learning of the loss of a loved one. In other, more severe cases, a surprisingly complete recovery came about, despite the fact that no specific therapy was available. “These patients are found to be optimistic and not easily worried”, he wrote.
A firm faith, feeling of social support from family and friends, all appear to be powerful stress buffers. It is not surprising, therefore, that such attributes have also been reported to be associated with a lower risk of cancer. A lack of emotional support as well as certain other traits were convincingly demonstrated by both Eysenck and Grossarth-Maticek to be highly predictive of cancer. More importantly, they have shown in extensive, long term prospective studies that stress reduction strategies were effective in reducing malignancy by 50% in individuals assessed as being cancer prone. Spiegel’s study similarly demonstrated that metastatic breast cancer patients who participated in group social support activities had an 18-month increase in survival compared to controls who received only routine treatment. Fawzy and coworkers found that if a 6-week stress management intervention was added to the treatment for early stage melanoma it enhanced immune system function when compared to controls. After 6 years, the stress management group had less than half the rate of recurrence and deaths. How can one explain the numerous well documented cases of spontaneous remission of cancer? Ikemi’s meticulous studies of such patients suggested that a firm faith and a strong positive belief system was the common denominator. Anecdotal but irrefutable reports of cancer cures from shrines, faith healers, comfrey, krebiozen, laetrile, coffee enemas, acupuncture, macrobiotic diets, and other alternative treatments abound. Yet, like spontaneous remission they are extremely rare, and can never be predicted. Here again, a strong faith in whatever the individual believes in, may provide the best explanation. But how is this mediated? How does the placebo effect work? How are the benefits of faith healing or “therapeutic touch” achieved? Is there such a thing as psychic healing? No consistent immune, neuroendocrine, or central nervous system changes have ever been demonstrated in connection with such responses.
Good health essentially depends on good communication – good communication within the internal environment, as well as with the external environment, in order to preserve homeostasis. That holds true for all living systems, ranging upward from the cell to an organ, person, family, corporation, nation, or a society. What we often fail to appreciate, is that these systems are in constant communication, and problems at one level, can reverberate up and down the line. Essentially, the basic problem with the cancer cell is that it does not communicate properly, as evidenced by these quotes from Yamasaki’s article on non-genotoxic mechanisms of carcinogenesis: “Cancer can be regarded as a rebellion in an orderly society of cells when they neglect their neighbors and grow autonomously over surrounding normal cells”.
“Since intercellular communication plays an important role in maintaining an orderly society, it must be disturbed in the process of carcinogenesis”.
“Evidence suggests that blockage of intercellular communication is important in the promotion process of carcinogenesis”.
While we cannot define stress, all of our research confirms that the sense of being out of control is always distressful. That also happens to be the best definition of the cancer cell – it is essentially a cell out of control, because it does not communicate. Could it be that the beneficial effects of a firm faith, or visual imagery, are somehow related to the development of a sense of control? Can that message somehow filter down through the body’s complex informational network to cancer cells? We know that the brain has hard wired as well as humoral connections with the immune system that may transmit such messages. However, it is clear that there are receptor sites on cell membranes for very subtle electrical energies similar in nature to those that are generated in the body. EEG wave patterns may be much more than simply the noise of the machinery of the brain. They may well represent messages being sent to other parts of the body. This is consistent with Nordenstrom’s theory of an internal electrical circulatory system and his dramatically successful treatment of metastatic lung tumors using weak electrical energies. Understanding how such mind/body interactions are mediated, may help us to learn how to stimulate, simulate, or emulate such mechanisms, to tap into the wisdom of the body and its awesome potential for self healing. Considerable evidence suggests that such forces play an even more important role in stress-cancer relationships because of their ability to control cell growth at its very basic level.
We are all exposed daily to a host of potential physical carcinogens in our environment but are there psychosocial carcinogens as well? What determines resistance or susceptibility to cancer? Behavioral factors and inappropriate responses to stress must also be considered along with genetic factors in attempting to understand why some individuals develop cancer, or what the clinical course will be. Impaired host resistance due to disturbances in immune system function seem to be an important factor as evidenced by the increase in malignancies in patients with AIDS and the melanoma studies by Fawzy and the benefit of stress reduction has been demonstrated in both of these fatal disorders. Similarly, Cohen’s detailed experiments on the effect of stress on the development of colds found that rates for both laboratory evidence of infection and clinical colds correlated precisely with the magnitude of psychological stress scores for each of the five rhinoviruses used in healthy volunteers. In addition, he also demonstrated the protective effects of strong social support.
Our current preoccupation is with cancer epidemiology, the roots of which epi (on), demos (people), logos (reason), connote something that has been thrust upon us from outside. What we must now begin to appreciate is what I have referred to as the endemiology of cancer, and those influences emanating from within the individual which may be equally significant and potentially under our control. Louis Pasteur, the great proponent of the germ theory of disease, engaged in many debates with his famous contemporary Claude Bernard. On his deathbed, he allegedly stated: “Bernard avait raison, Le germe n’est rien, c’est le terrain quiest tout”. (Bernard was right. The microbe is nothing, the soil is everything). In the final analysis, we are left with what every “compleat” physician eventually learns, namely, that, “Many times it is much more important to know what kind of patient has the disease, than what kind of disease the patient has”.
Paul J. Rosch, M.D., F.A.C.P.
President, The American Institute Of Stress
Clinical Professor of Medicine and Psychiatry
New York Medical College
(The above is based on prior chapters and articles that are listed in the following Suggested Readings)