[Cancer, a defect of the psyche?]

Patients and families often put forward psychosocial determinants to explain occurring or progression of cancer disease. The idea of a psychogenesis in cancer is widely spread by the surrounding culture. Nevertheless, the scientific research in this field brought very contradictory results, and often rests on debatable methods. Some authors hypothesize that a type C personality predisposes to cancer. But so far, most of prospective studies analysing the association between C traits and this disease have been negative. Results of semi-prospective studies, in which patients are enlisted before a diagnostic investigation, are difficult to interpret, because the starting features can partly result from the stress of this investigation. The same problem interferes with most studies about the association between alexithymia and cancer. The reliable works analysing the impact of depression most often conclude in favour of a non-existent or a weak risk, and do not allow defining it as a well established risk factor. Most of prospective studies about the link between stress and cancer are negative. Moreover, they index stressful life events occurred for a few years before the diagnosis, that is to say during a too short delay to be allowed to incriminate them in the supervening of the disease. The question of the possible impact of coping styles with cancer on its evolution also remains debated. Some studies for instance plead in favour of a beneficial effect of fighting spirit or denial, but they are contradicted by other studies. All this forces us to be very careful when discussing the possible links between psyche and cancer. It is important to underline the hypothetical nature of these relationships. Psychosomatic explanations risk to be used to fill in gaps of knowledge, and to give us the illusion that we can avoid or control a disease which escapes us. They especially risk to make the patients guilty of their cancer or its evolution, through an inadequate point of view which does not take the complexity in stake into account.