Etiology and Diagnosis of Major Depression - A Novel Quantitative Approach

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review


Background: Classical psychiatric opinions are relative uncertain and treatment results are not impressive when dealing with major depression. Depression is related to the endocrine system, but despite much effort a good quantitative measure for characterizing depression has not yet emerged. Methods: Based on ACTH and cortisol levels and using clustering analysis and mixture effect modeling we propose a novel and scientifically based quantitative index, denoted the O-index. The O-index combines a weighted and scaled deviation from normal values in both ACTH and cortisol. Results: Using ANOVA we compare the O-index with opinions reach by classical psychiatric diagnostic procedure (sensitivity 83%, specificity 59%, likelihood ratio positive 2.0, and likelihood ratio negative 0.29). The O-index nicely refines the etiology of depression: Combined with clinical data for 29 subjects earlier reported three categories emerge (p = 4.4 × 10-13): hypocortisolemic depressed, non-depressed, and hypercotisolemic depressed. The O-index also reveals why it has been difficult to obtain good markers earlier. It explains that healthy subjects may have an elevated (suppressed) level of cortisol or ACTH, however, the healthy system is able to deal with such elevated (suppressed) levels by compensating through suppressing (stimulating) the other component. In contrast the O-index shows that depressed subjects are incapable of making such compensation to a satisfactory degree. We illustrate how the O-index may be used for diagnostic procedure. Discussion: The methods are discussed and based on the available data material we propose that the O-index may be used to improve the diagnostic procedure and consequently the follow-up treatment.
TidsskriftOpen Journal of Endocrine and Metabolic Diseases
Udgave nummer2
Sider (fra-til)120-127
Antal sider8
StatusUdgivet - maj 2013


  • Depression
  • HPA-Axis
  • Etiology
  • Diagnoses
  • Clustering Analysis
  • Mixture Effect Modeling

Citer dette