Artikel
Stability over time of psychiatric rating scales and questionnaires, but not cognitive tests, in healthy individuals of the PsyCourse Study
Suche in Medline nach
Autoren
Veröffentlicht: | 24. September 2021 |
---|
Gliederung
Text
Introduction: Developing therapies for long-term alleviation of severe mental disorders is a major global task. The affected individuals experience a broad range of symptoms, oftentimes including cognitive impairments. To accurately evaluate changes in psychiatric symptoms and cognition in patients diagnosed from disorders such as schizophrenia, bipolar disorder, or depression, it is critical to understand the course over time of these measurements in healthy control individuals.
Methods: We comprehensively investigated the stability of the longitudinal course of cognitive and symptom measurements (both observer-rated and self-rated) for severe mental disorders in healthy controls of the PsyCourse study (n=326, four assessments over 18 months, male=128, female=198, age range=[18, 77], age mean=37.49, 224 completed all visits) [1]. Cognitive measurements included the Trail Making test, Verbal Digit Span tests (Forward and Backward), the Digit Symbol test, and the German version of the Rey Auditory-Verbal Learning Test (VLMT). Psychiatric rating scales comprised the Global Assessment of Functioning, the Inventory of Depressive Symptomatology, the Positive and Negative Syndrome Scale, and the Young Mania Rating Scale (YMRS). Questionnaires evaluated were the Beck Depression Inventory, and a self-report inventory for mania (Manie-Selbstbeurteilungsskala, MSS). Combined, we investigated the courses of 16 variables. We modelled time effects with Linear Mixed Models for most cognitive tests or Logistic Regression for most questionnaires and rating scales, since their distributions include large spikes in the minimal symptom range. Six variables needed to be transformed to improve the fit of the Linear Mixed Model. For Logistic Regression we dichotomized the data according to symptom categories, when possible, and spikes.The slopes for Linear Mixed Models or odds ratios for Logistic Regression were estimated. We also assessed the robustness of these results with a Longitudinal Non-Parametric Test. In total, 39 statistical tests for time effects were performed, all adjusted for age and sex.
Results: For the cognitive tests small effects were rather consistently detected by both Linear Mixed Models and Longitudinal Non-Paramteric Test, indicating a performance improvement over time (p<0.05). Contrary, in most of the symptom rating scales and questionnaires, no effects were detected. The MSS was the only questionnaire to show a consistently significant change over time (OR=1.36), and deviates from the course of the observer-rated YMRS (OR=0.80).
Discussion: Differences in assessment method and time period may explain varying results of the MSS and YMRS. The positive effects observed in cognitive tests conform with previously described retest effects, and are likely due to increasing familiarity with the series of tests. These retest effects are not solely caused by a single increase in performance between the first two visits, but most performances improve gradually over visits. We also found higher intra-person variability in cognitive tests than in rating scales and questionnaires, which is in line with previous studies of healthy individuals.
Conclusion: Over the course of the PsyCourse study, both questionnaires and rating scales can be assumed to remain stable, while there is an improvement in cognitive tests. Taking these detectable improvements into account is important when interpreting longitudinal data of patients in research studies.
The authors declare that they have no competing interests.
The authors declare that an ethics committee vote is not required.
References
- 1.
- Budde M, Anderson-schmidt H, Gade K, Reich-erkelenz D, Flatau L, Reitt M, et al. A Longitudinal Approach to Biological Psychiatric Research: The PsyCourse Study. 2017 October. p. 1–35.