Psychotic disorders in the Lundby population 1947-1997: Incidence, life-time prevalence and predictors related to personality and behaviour

University dissertation from Lund University Sweden, Department of Clinical Sciences, Faculty of Medicine

Abstract: The aims were to analyze the first incidence, age-at-onset, prevalence and personality related predictors of psychotic disorders in the Lundby population (n=3563) – a community population followed through four waves of field study 1947-1997. Case ascertainment included interviews, registers, case notes and key informants. Consensus diagnoses of mental illness were assessed according to a classification adapted to the Lundby Study, but in 1972-1997 also according to the DSM-IV. Mental illness 1947-1972 was rediagnosed according to the DSM-IV. Follow-ups were feasible in the majority of subjects; 99% in 1947-1972 and 94% in 1972-1997. The incidence rate was higher in males than in females for ‘substance-induced psychoses’. The incidence rates of the other psychoses did not differ between the sexes. The male mean age-at-onset was lower than that for females for ‘psychotic disorder due to a general medical condition’, ‘nonaffective psychoses’ and ‘schizophrenia’. Males and females had different incidences by age patterns for ‘nonaffective psychoses’, ‘schizophrenia’ and ‘other nonaffective psychoses’. The period prevalence 1947-1997 was higher in males than in females for ‘substance-induced psychoses’. For the other psychoses the period prevalences did not differ between the sexes. The lifetime prevalence in 1997 was higher in males than in females for the general group including any psychotic or bipolar disorder. For the specific psychoses the lifetime prevalences in 1997 did not differ significantly between the sexes. The influence of constructed dichotomous predictor variables related to premorbid personality traits in the subjects on time to occurrence of ‘functional psychosis’ and ‘schizophrenia’, respectively, was assessed by Cox regression. In multivariate and/or univariate models the predictors nervous-tense, down-semidepressed, blunt-deteriorated, paranoid-schizotypal, sensitive-frail, easily hurt and tired-distracted were associated with ‘functional psychosis’. When ‘schizophrenia’ was analyzed separately, nervous-tense remained significant and abnormal-antisocial reached significance. The findings suggest similar overall risks in males and females to contract psychosis during the life course, but with putative disorder delaying mechanisms in females and/or different aetiologies in males and females. Certain personality traits – related to anxiety-proneness, affective/cognitive blunting, poor concentration, personality cluster-A like traits and interpersonal sensitivity – may be associated with general psychosis vulnerability. Limitations: changing sample representativeness, small sample size, attrition, recall bias, changing data sources, inter-rater reliability and changing diagnostic systems during the study.