Search

Background: Cluster analyses have become popular tools for data-driven classification in biological psychiatric research. However, these analyses are known to be sensitive to the chosen methods and/or modelling options, which may hamper generalizability and replicability of findings. To gain more insight into this problem, we used Specification-Curve Analysis (SCA) to investigate the influence of methodological variation on biomarker-based cluster-analysis results. Methods: Proteomics data (31 biomarkers) were used from patients (n = 688) and healthy controls (n = 426) in the Netherlands Study of Depression and Anxiety. In SCAs, consistency of results was evaluated across 1200 k-means and hierarchical clustering analyses, each with a unique combination of the clustering algorithm, fit-index, and distance metric. Next, SCAs were run in simulated datasets with varying cluster numbers and noise/outlier levels to evaluate the effect of data properties on SCA outcomes. Results: The real data SCA showed no robust patterns of biological clustering in either the MDD or a combined MDD/healthy dataset. The simulation results showed that the correct number of clusters could be identified quite consistently across the 1200 model specifications, but that correct cluster identification became harder when the number of clusters and noise levels increased. Conclusion: SCA can provide useful insights into the presence of clusters in biomarker data. However, SCA is likely to show inconsistent results in real-world biomarker datasets that are complex and contain considerable levels of noise. Here, the number and nature of the observed clusters may depend strongly on the chosen model-specification, precluding conclusions about the existence of biological clusters among psychiatric patients. [ABSTRACT FROM AUTHOR]

BackgroundPredictD is a risk algorithm that was developed to predict risk of onset of major depression over 12 months in general practice attendees in Europe and validated in a similar population in Chile. It was the first risk algorithm to be developed in the field of mental disorders. Our objective was to extend predictD as an algorithm to detect people at risk of major depression over 24 months.MethodParticipants were 4190 adult attendees to general practices in the UK, Spain, Slovenia and Portugal, who were not depressed at baseline and were followed up for 24 months. The original predictD risk algorithm for onset of DSM-IV major depression had already been developed in data arising from the first 12 months of follow-up. In this analysis we fitted predictD to the longer period of follow-up, first by examining only the second year (12–24 months) and then the whole period of follow-up (0–24 months).ResultsThe instrument performed well for prediction of major depression from 12 to 24 months [c-index 0.728, 95% confidence interval (CI) 0.675–0.781], or over the whole 24 months (c-index 0.783, 95% CI 0.757–0.809).ConclusionsThe predictD risk algorithm for major depression is accurate over 24 months, extending it current use of prediction over 12 months. This strengthens its use in prevention efforts in general medical settings. [ABSTRACT FROM AUTHOR]

BackgroundThere are no risk models for the prediction of anxiety that may help in prevention. We aimed to develop a risk algorithm for the onset of generalized anxiety and panic syndromes.MethodFamily practice attendees were recruited between April 2003 and February 2005 and followed over 24 months in the UK, Spain, Portugal and Slovenia (Europe4 countries) and over 6 months in The Netherlands, Estonia and Chile. Our main outcome was generalized anxiety and panic syndromes as measured by the Patient Health Questionnaire. We entered 38 variables into a risk model using stepwise logistic regression in Europe4 data, corrected for over-fitting and tested it in The Netherlands, Estonia and Chile.ResultsThere were 4905 attendees in Europe4, 1094 in Estonia, 1221 in The Netherlands and 2825 in Chile. In the algorithm four variables were fixed characteristics (sex, age, lifetime depression screen, family history of psychological difficulties); three current status (Short Form 12 physical health subscale and mental health subscale scores, and unsupported difficulties in paid and/or unpaid work); one concerned country; and one time of follow-up. The overall C-index in Europe4 was 0.752 [95% confidence interval (CI) 0.724–0.780]. The effect size for difference in predicted log odds between developing and not developing anxiety was 0.972 (95% CI 0.837–1.107). The validation of predictA resulted in C-indices of 0.731 (95% CI 0.654–0.809) in Estonia, 0.811 (95% CI 0.736–0.886) in The Netherlands and 0.707 (95% CI 0.671–0.742) in Chile.ConclusionsPredictA accurately predicts the risk of anxiety syndromes. The algorithm is strikingly similar to the predictD algorithm for major depression, suggesting considerable overlap in the concepts of anxiety and depression. [ABSTRACT FROM PUBLISHER]