Rene Schilling ^1* Robyn Cody ¹ Jan-Niklas Kreppke ¹ Oliver Faude ¹ Johannes Beck ² Serge Brand ¹ Lars Donath ³ Martin Hatzinger ⁴ Christian Imboden ⁵ Undine Lang ⁶ Sarah Mans ⁵ Thorsten Mikoteit ⁷ Anja Oswald ⁸ Nina Schweinfurth-Keck ⁹ Markus Gerber ¹

• ¹ Department of Sport, Exercise and Health, University of Basel, Basel, Switzerland
• ² Klinikum Sonnenhalde, Riehen, Switzerland
• ³ German Sport University Cologne, Department of Training Intervention Research, Cologne, Germany, Cologne, Germany
• ⁴ Psychiatrische Dienste Solothurn, Solothurn, Switzerland
• ⁵ Private Clinic Wyss, Muenchenbuchsee, Switzerland
• ⁶ University Psychiatric Clinic Basel, Basel, Switzerland
• ⁷ Psychiatric Services Solothurn and Faculty of Medicine of the University of Basel, Solothurn, Switzerland, Solothurn, Switzerland
• ⁸ Psychiatric Clinic Sonnenhalde, Riehen, Switzerland, Riehen, Switzerland
• ⁹ Center for Affective, Stress and Sleep Disorders, University Psychiatric Clinic Basel, Basel, Switzerland

Major depressive disorders (MDD) are a leading health concern worldwide. While first line medication treatments may fall short of desired therapeutic outcomes, physical activity (PA) interventions appear to be a promising and cost-effective add-on to improve symptoms of depression. This study aimed to address challenges in the assessment of PA in inpatients treated for MDD by examining the correspondence of self-reported and accelerometer-based PA.In 178 inpatients treated for MDD (mean age: M = 41.11 years, SD = 12.84; 45.5 % female) and 97 non-depressed controls (mean age: M = 35.24 years, SD = 13.40; 36.1 % female), we assessed selfreported PA via the Simple Physical Activity Questionnaire (SIMPAQ) for one week, followed by a week where PA was monitored using an accelerometer device (Actigraph wGT3x-BT). Additionally, we examined correlations between PA levels assessed with the SIMPAQ and exercise determinants in both groups.Descriptively, inpatients treated for MDD showed lower levels of light PA on accelerometer-based measures, whereas they self-reported increased levels of certain types of PA on the SIMPAQ. More importantly, there was only a small degree of correspondence between self-reported and actigraphybased PA levels in both in patients (r = .15, p < .05) and controls (r = .03, ns). Only few significant correlations were found for self-reported PA (SIMPAQ subscores) and perceived fitness, whereas self-reported PA and estimated VO2max were unrelated. Furthermore, only weak (and mostly statistically non-significant) correlations were found between exercise determinants and SIMPAQbased exercise behavior in both populations.Our findings emphasize the intricate challenges in the assessment of PA, not only in inpatients treated for MDD, but also in non-depressed controls. Our findings also underline the necessity for a diversified data assessment. Further efforts are needed to refine and improve PA questionnaires for a more accurate data assessment in psychiatric patients and healthy controls.