Ageing is due to a progressive loss of the person's adaptation capability, whereas during this period environmental aggression increases. In the elderly, life events re-present a psychological traumatism that overwhelms the old person and related family, disrupting and fragilising homeostatic balance. A number of authors have suggested a possible link between life traumatisms and the dementia processes. The aim of this study is to reveal the presence of life traumatisms preceding the apparition of the dementia syndrome. Method. This is a retrospective and comparative work based on the PIXEL study on complaints and demands from the principle informal caregivers of Alzheimer patients. It includes 565 patients presenting the criterion of dementia as defined by the DSM IV, and questionnaires filled out by the principle caregivers. One item of the questionnaire referred to life events which could have played a part in the development of the disorder. In a second stage, the reported events were classified into 4 distinct categories: loss, repeated or prolonged stress, psychotraumatism and depression-inducing events. The statistics were produced using SAS and Stat 10 software. Student's test, ANOVA and chi2-test were used. RESULTS: 372 caregivers answered the first item (65%); 76 of them believed there was no event while 296 related the disorder to one or several life events (79% of responders, 52% of the sample). These results confirm Persson and Clement's study which evidenced a higher frequency of stressing life events for subjects afflicted with dementia as compared with older people without any psychic disorder. Reported events and their respective frequency: spouse death (15.39%), parents' death (15%), familial difficulty (10.08%), anaesthesia (8.49%), child's death (4.42%), somatic disturbance (4%), depression (3.89%), retirement (3.89%), financial problems (2.65%), loneliness (2.65%), removal (1.76%), fall (1%), alcohol (0.8%), traumatism (0.53%), spouse care (0.35%), leaving for home care, storm and caregiver change in life (0.17%). Regrouping the data: 82.71% of the answers can be connected to prolonged stress, 62.85% to real or symbolic loss, 39% to a psychotraumatic event and finally, most of these events can potentially induce depression. DISCUSSION: Some of the events refer to difficulties concerning people close to the subject (death of a close relation, hospitalisation of husband or wife) in a period when the loss of autonomy or handicap means greater dependence on surrounding people. General anaesthesia is mentioned by 48 caregivers. In fact, this usually implies surgery, the presence of organic pathology or the need for hospitalisation, which we know has a destabilizing effect in the elderly. It is therefore not surprising that 23 answers mentioned somatic disorders (4% of the sample). According to Leger, it's mostly a loss type event which is implied in the elderly. Such losses would induce a weakening of cognitive stimulation which could decompensate an infraclinical dementia or accelerate an emerging dementia process. Antecedents of depressive illness are considered as an element of risk for the development of Alzheimer's disease. Depression is spontaneously mentioned by 22 caregivers. We must add the many reported life events which are well known to induce depression in older people. Most of the events considered in this study are liable to provoke manic mood swings. Depression resulting from life events can be considered either as an affection occurring along with dementia or as the aggravating factor of an infraclinical process or, finally, as an additional factor of vulnerability. With older people, many events may constitute a trauma because of the proximity of death and because of their sudden onset (fall, hospitalisation, somatic illness). Several studies have pointed out that a particularly traumatic event could enhance the risk of dementia. Life events associated with chronic or repeated stress are characterized by their permanence or their repetition. According to a general psychosomatic biological pattern, psychic distress will engender a series of degradations or an acute or chronic response to an early trauma. According to this hypothesis, prolonged exposure to an excess of glucocorticosteroids at the time of a disadaptative stress would have deleterious effects on the hippocampus. Indeed, the hippocampus plays a part in a number of functions affected by dementia such as memory, learning process and emotional adjustment. This study takes into account stress factors ("stressors") but not factors influencing their impact on the subject such as an individual predisposition (genetic, psychopathologic, coping abilities) and social support. The force of the impact of these events on older people and what is really experienced by them remain unknown. This study strengthens a number of others evincing an unusual frequency of life events in dementia processes. According to an integrative pattern, repeated or prolonged stress could be a deciding factor in the degenerative process or a factor of decompensation with older people presenting a genetic, biological or psychological vulnerability to dementia. The impact of such life events would vary according to the subject's pre-morbid personality, coping abilities and the support he/she can rely on. CONCLUSION: Some life events may be involved in the dementia process as shown by the results of this study, but this relationship does not imply direct causality. It's difficult to appreciate whether these results are not a consequence of the greater attention paid to the patient after the appearance of the first symptoms, leading to a closer observation. Stress could trigger the degenerative process. This argues for the necessity of an early diagnosis taking into account a traumatic event of life either precocious or late.