[Cause of mortality in schizophrenic patients: prospective study of years of a cohort of 150 chronic schizophrenic patients].

UNLABELLED: Overmortality in schizophrenic patients in comparison to the reference population has been found. At the present time this over mortality is mainly due to suicide or certain natural causes such as respiratory, cardio-vascular and cerebro-vascular diseases. In France there are not psychiatric cas registers that could allow us to study the mortality of psychiatric patients. The aim of the study was first to determine the standardized mortality ratio (SMR) in a group of 150 chronic schizophrenics followed during 8 years and secondly to detect the variables that could predict this mortality.

METHOD: The subjects filled out the RDC criteria for definite chronic schizophrenia and were included from 1991 to 1995. The subjects were inpatients or outpatients and their evaluation was made by psychiatrist. The subjects were selected from the different departments of two psychiatric hospitals corresponding to two French geographic areas (the Somme and Oise, two French "département"). At the initial assessment socio-demographic, clinical and psychometrical variables were collected: sex, age, educative level, number of hospitalizations, mean duration of the illness, scores on the Physical Anhedonia Scale, Brief Psychiatric Rating Scale (BPRS), Positive and Negative Syndrome Scale (PANSS). For the BPRS and PANSS, negative, positive and general subscales were extracted. In May 1999 all the subjects were contacted in order to know if they are alive or not and if they are death to know the date and the causes of their death. For the subjects that were still alive we used either direct assessment by interview of their psychiatrist or general practioner or indirect assessment by interview of their family. For the deceased subjects, we obtained informations about the date and the causes of the death by their psychiatrist or general practioner. If the patients were lost sight of we send a letter to the city of their place of birth in order to know if they are alive or not and if they are dead to know the date of their death. Moreover demographic data concerning the French and the Somme populations as well as the corresponding data concerning the mortality according to age and gender were obtained. A comparison of global mortality between patients and the French general or the Somme populations was made by the SMR. Moreover the deceased subjects and the survivors were compared by unidimensional statistical tests (chi 2 analyses for qualitative variables or Student's t test for quantitative variables) for the sociodemographic, clinical or psychometric variables. For each significant difference at p level < or = 0.05, the corresponding variable was retained for a multivariate step by step discriminant analysis.

RESULTS: We found 13 deaths (10 males, 3 females): 3 suicides, 3 cardiac diseases, 2 cancers, 1 respiratory disease, 1 car crash, 1 homicide, 1 infectious disease, 1 respiratory arrest. The mortality rate (without correction for age) were 1.08% for both sexes, 1.44% and 0.6% for males and females respectively. The mortality rates (corrected for age) were 2.47% in our cohort and 0.988% in the Somme population. The corresponding SMR was 2.5. (chi 2 = 3.15, df = 1, p < 0.01). The overmortality was found only for males (chi 2 = 2.57, df = 1, p < 0.01) and not for females (chi 2 = 0.034, df = 1, p > 0.05). Concerning the comparisons between the deceased subjects and the survivors, there were five significant differences: gender, age, duration of the illness, neuroleptic dosage, negative symptoms (BPRS negative subscale). The deceased subjects were older, there was more men, the duration of the illness and the neuroleptic dosage were higher and the BPRS negative subscale was lower. These five variables were introduced in the discriminant analysis to explore notably their respecting weight. The corresponding power of the five variables were in decreasing order: neuroleptic dosage, negative symptoms, age, gender, duration of the illness.

DISCUSSION: Our study confirm the overmortality in schizophrenic patients, this overmortality was especially explained by natural and non natural causes of death. The overmortality concerned only schizophrenic males patients whereas schizophrenic females did not have an overmortality. This negative result could be explain by a bias selection, the males being overrepresented in our cohort. Among the variables that were linked to the overmortality, the low level of negative symptomatology confirmed previous studies that have shown a low suicide rate in deficit schizophrenic. Moreover a high level of positive symptomatology could lead to high risk behaviors (suicide attempts, sexual disinhibition...). The neuroleptic dosage was the variable whom discriminate power was the highest. At least two explanations can be proposed. (ABSTRACT TRUNCATED)