The natural history of depression in general practice: stochastic models

Psychol Med. 1981 Nov;11(4):755-64. doi: 10.1017/s003329170004126x.


Three statistical models are presented to describe different aspects of the natural history of depression (as recognized by a general practitioner) during a 20-year study of a single general practice. After controlling for age, the prevalence of recognized depression in men increased during the 20 years of the study (1957-76), but there was little change in women. In any given year, however, women are both more likely to become depressed than men, and are less likely to recover. The changes in prevalence are shown to be due to changes in observed incidence and recovery rates. Taking into account attendances over the previous 5 years at which a diagnosis of depression was made, the models enable one to predict, for example incidence rates for 'first episodes' of recognized depression, and recovery rates for 'chronic' patients. In all situations there is a significant sex difference (women always being more likely to be recognized as depressed), but this difference is smaller at the end than at the beginning of the 20 years.

MeSH terms

  • Adult
  • Age Factors
  • Depression / diagnosis*
  • Depression / epidemiology
  • Female
  • Humans
  • Longitudinal Studies
  • Male
  • Markov Chains
  • Mathematics
  • Middle Aged
  • Models, Psychological
  • Sex Factors
  • Stochastic Processes