Longitudinal study of the influence of life events and personality status on diagnostic change in three neurotic disorders.

It has been known for many years that diagnosis within the neurotic spectrum of disorders is temporally unstable and also that life events can be major precipitants of change in symptoms. Reasons for this instability could include inherent inadequacy of current diagnostic practice, the influence of life events as an agent of diagnostic shift, and an innate course of disorder with features dependent on the stage at which disorder presents (e.g., development of panic to agoraphobia). These possibilities were examined in a prospective study that was initially a randomised controlled trial. Two hundred ten patients recruited from primary care psychiatric clinics with DSM-III diagnosed dysthymic, generalised anxiety, and panic disorders were randomly allocated to either drug treatment (mainly antidepressants), cognitive-behaviour therapy, or self-help therapy over a 2 year period, irrespective of original diagnosis. Life events were recorded by using a standard procedure over the period 6 months before starting treatment and at five occasions over 2 years; 181 (86%) of the patients had follow-up data and 76% maintained compliance with the original treatment allocated over the 2 years; and 155 of the 181 patients (86%) had at least one diagnostic change in this period. There was no difference in the number of diagnostic changes between the three original diagnostic groups, but dysthymic disorder changed more frequently to major depressive episode than did GAD or panic disorder (20; 11; 12) (%) and panic disorder changed more frequently to agoraphobia (with or without panic) than did dysthymia or GAD (18; 8; 6) (%). There was no relationship between loss events and depressive diagnoses or between addition events and anxiety diagnoses, but greater numbers of conflict events were associated with diagnostic change. More life events were associated with the flamboyant and dependent personality disorders, reinforcing other evidence that many life events are internally generated by personality characteristics and cannot be regarded as truly independent.