Inter-informant agreement and prevalence estimates for mood syndromes: Direct interview vs. family history method

• Published in: Journal of affective disorders 2015-01, Vol.171, p.120-127
• Main Authors: Vandeleur, C.L, Rothen, S, Lustenberger, Y, Glaus, J, Castelao, E, Preisig, M
• Format: Article
• Language: English
• Subjects: Adult; Adult and adolescent clinical studies; Biological and medical sciences; DSM-5; Europe - epidemiology; Family - psychology; Family Health - statistics & numerical data; Family history method; Family study; Female; Humans; Inter-informant agreement; Interview, Psychological - methods; Male; Medical History Taking - methods; Medical History Taking - statistics & numerical data; Medical sciences; Mood disorders; Mood Disorders - diagnosis; Mood Disorders - epidemiology; Mood syndromes; Prevalence; Psychiatry; Psychology. Psychoanalysis. Psychiatry; Psychopathology. Psychiatry; Reproducibility of Results; Switzerland - epidemiology; Syndrome
• ISSN: 0165-0327; 1573-2517
• DOI: 10.1016/j.jad.2014.08.048

Abstract Background The use of the family history method is recommended in family studies as a type of proxy interview of non-participating relatives. However, using different sources of information can result in bias as direct interviews may provide a higher likelihood of assigning diagnoses than family history reports. The aims of the present study were to: 1) compare diagnoses for threshold and subthreshold mood syndromes from interviews to those relying on information from relatives; 2) test the appropriateness of lowering the diagnostic threshold and combining multiple reports from the family history method to obtain comparable prevalence estimates to the interviews; 3) identify factors that influence the likelihood of agreement and reporting of disorders by informants. Methods Within a family study, 1621 informant–index subject pairs were identified. DSM-5 diagnoses from direct interviews of index subjects were compared to those derived from family history information provided by their first-degree relatives. Results 1) Inter-informant agreement was acceptable for Mania, but low for all other mood syndromes. 2) Except for Mania and subthreshold depression, the family history method provided significantly lower prevalence estimates. The gap improved for all other syndromes after lowering the threshold of the family history method. 3) Individuals who had a history of depression themselves were more likely to report depression in their relatives. Limitations Low proportion of affected individuals for manic syndromes and lack of independence of data. Conclusions The higher likelihood of reporting disorders by affected informants entails the risk of overestimation of the size of familial aggregation of depression.