Abstract
Psychological symptoms tend to change over time, even in the absence of clinical intervention. For example, self-ratings are often higher at screening compared to start of treatment. A plausible hypothesis is that this is due to patients' self-referring when their gradually fluctuating symptoms are worse than usual. That hypothesis predicts that patients that wait longer will have had longer time to return to their average symptom level. On the other hand, other processes related to measurement reactivity, contact with a clinician, or regression towards the mean, do not predict a time-dependent relationship. Our aim was to estimate the extent of this hypothesized symptom reduction in depression, social anxiety disorder, panic disorder, health anxiety and insomnia (both total reduction and the relationship with time). The sample included adults (N = 8744) from an outpatient psychiatric clinic providing ICBT in Swedish routine care. Time-dependent effects were estimated with linear regression for both primary symptoms and secondary depressive symptoms. A simulation of symptom fluctuations was built to estimate power and further contextualize the effects. Patients improved on average from screening to the start of the intervention, but this varied substantially depending on diagnosis and questionnaire used. The waiting time weakly predicted the degree of improvement both for primary depressive symptoms and comorbid depressive symptoms. The estimate for primary depressive symptoms was sensitive to modeling choices, shrinking towards zero when modeled with fat-tailed residuals. The preponderance of "immediate" reductions in symptoms have implications for reporting standards of pre-treatment-measurements, especially in single-group intervention studies.