Cohort studies of the long-term outcome of anxiety, depression and personality status rarely join together.

Two hundred and ten patients recruited with anxiety and depression to a randomised controlled trial between 1983 and 1987 (Nottingham Study of Neurotic Disorder) were followed up over 30 years. At trial entry personality status was assessed, together with the general neurotic syndrome, a combined diagnosis of mixed anxiety–depression (cothymia) linked to neurotic personality traits. Personality assessment used a procedure allowing conversion of data to the ICD-11 severity classification of personality disorder. After the original trial, seven further assessments were made. Observer and self-ratings of psychopathology and global outcome were also made. The primary outcome at 30 years was the proportion of those with no Diagnostic and Statistical Manual of Mental Disorders (DSM) diagnosis.

Data were analysed using multilevel repeated measures models that adjusted for age and gender. Missing data were assumed to be missing at random, and the models allowed all subjects to be included in the analysis with missing data automatically handled in the model estimation.

At 30 years, 69% of those with a baseline diagnosis of panic disorder had no DSM diagnosis compared to 37–47% of those with generalised anxiety disorder, dysthymia or mixed symptoms (cothymia) (p = 0.027). Apart from those with no personality dysfunction at entry all patients had worse outcomes after 30 years with regard to total psychopathology, anxiety and depression, social function and global outcome.

The long-term outcome of disorders formerly called ‘neurotic’ is poor with the exception of panic disorder. Personality dysfunction accentuates poor recovery.