Open AccessIdentifying depression and anxiety in a 40‐year epidemiological investigation: the Stirling County study
Author(s) -
Murphy Jane M.,
Monson Richard R.,
Laird Nan M.,
Sobol Arthur M.,
Leighton Alexander H.
Publication year - 1998
Publication title -
international journal of methods in psychiatric research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.275
H-Index - 73
eISSN - 1557-0657
pISSN - 1049-8931
DOI - 10.1002/mpr.38
Subject(s) - psychology , anxiety , depression (economics) , sadness , sample (material) , population , longitudinal study , apprehension , psychiatric epidemiology , vocabulary , set (abstract data type) , psychiatry , clinical psychology , medicine , mental health , cognitive psychology , computer science , linguistics , chemistry , philosophy , environmental health , anger , chromatography , pathology , economics , macroeconomics , programming language
As a longitudinal investigation, the ‘Stirling County study’ involves a 40‐year period beginning with the drawing of a sample in 1952 for a survey of psychiatric disorders. In 1970 a new sample was drawn and subjects from 1952 were followed, and in 1992 another new sample was drawn and all 1952 and 1970 subjects were followed. During these years, knowledge about psychiatric disorders has increased in the population at large; changes in everyday language about them have occurred and criteria for identifying them have been improved. This report describes adaptations of methods designed to overcome some of the problems inherent in such changes. A revised computerized algorithm named DPAX‐2 (DP for depression and AX for anxiety) is described. The first version, DPAX‐1, deals with a core set of questions that have been asked in identical fashion throughout the study. Early investigations of validity indicated that DPAX‐1 appeared to be an adequate method for the period from 1952 through 1970. The revised version was constructed after data gathering for the 1990s phase of research documented that vocabulary changes had taken place. DPAX‐2 addresses these changes by having more questions about states of sadness and apprehension. It also takes into account improvements introduced as ‘supplements’ that come after the core questions in the interview schedule. These improvements were available in the data collected for 1970 but have not thus far been utilized in reports about longitudinal findings because of their absence from the 1952 data. Evidence is presented that DPAX‐1 and DPAX‐2 were well correlated in 1970 and became considerably less so by 1992. This result, in conjunction with increased diagnostic sophistication in the responses of subjects to open‐ended questions in the 1990s, suggests that the common parlance, especially about depression, began to change after 1970. It was also mainly after 1970 when antidepressant medications began to be widely used, a fact that may have influenced awareness about depression and the language for describing it. For comparison, depression and anxiety modules of the more recently developed Diagnostic Interview Schedule (DIS) were also part of the interview protocol in the 1990s. To explore validity, the Structured Clinical Interview for DSM‐IIIR (SCID) was administered to a subsample. Agreement with the SCID results was better for DPAX‐2 than for DPAX‐1. Agreement with SCID was somewhat higher for the DIS, but DPAX‐2 and the DIS showed almost equivalent sensitivity. In conclusion, the most satisfactory approach to studying the effects of ‘time's arrow’ in the Stirling County study will be to use DPAX‐1 for the period 1952 to 1970 and DPAX‐2 for the period 1970 to 1992. The level of articulation between the two versions will therefore be discernible at the middle phase of the study. Copyright © 1998 Whurr Publishers Ltd.