Confounders of depression measurement in ALS/MND: metaregression analysis of published literature

Objective: To estimate measurement bias from popular questionnaire measures for depression using meta-regression. Methods: A systematic review of the literature published on the following databases was conducted without time limits: PubMed (MEDLINE), PsycINFO, the Cochrane Central Register of Controlled Trials, CINAHL, Scopus and ISI Web of Science. Mean scores for depression questionnaires were separately meta-regressed on study-level mean times since onset of ALS/MND symptoms in order to test the significance of the slopes and assess the goodness of fi t of the two conditional random models. Results: Our literature search revealed 110 studies that reported depression scores from psychometric questionnaires. Data were included from 103 studies (3,190 patients). The most commonly used questionnaires were The Hospital Anxiety and Depression Scale (HADS) and the Beck Depression Inventory (both the BDI-I and the BDI-II). Average scores on both questionnaires do not indicate a large number of patients with case-level depression and average scores for depression are similar to score from normal populations. Scores for depression increase with duration of illness. The BDI and the BDI-II were highly info uenced (more than 50% of the variance in scores) by level of physical impairment. Scores from the HADS were also infl uenced by physical impairment, but to a lesser degree (19%). Discussion: Meta-evidence from the published literature suggests that commonly used generic depression scales are overly influenced by the presence of impairment. Conclusion: Widely used depression scales over-estimate severity of depression due to confounding with physical symptoms of the underlying physical symptoms in ALS/MND