Screening for Depression in Daily Life: Development and External Validation of a Prediction Model Based on Actigraphy and Experience Sampling Method

Abstract

Background In many countries, depressed individuals often first visit primary care settings for consultation, but a considerable number of clinically depressed patients remain unidentified. Introducing additional screening tools may facilitate the diagnostic process. Objective This study aimed to examine whether experience sampling method (ESM)-based measures of depressive affect and behaviors can discriminate depressed from nondepressed individuals. In addition, the added value of actigraphy-based measures was examined. Methods We used data from 2 samples to develop and validate prediction models. The development data set included 14 days of ESM and continuous actigraphy of currently depressed (n=43) and nondepressed individuals (n=82). The validation data set included 30 days of ESM and continuous actigraphy of currently depressed (n=27) and nondepressed individuals (n=27). Backward stepwise logistic regression analysis was applied to build the prediction models. Performance of the models was assessed with goodness-of-fit indices, calibration curves, and discriminative ability (area under the receiver operating characteristic curve [AUC]). Results In the development data set, the discriminative ability was good for the actigraphy model (AUC=0.790) and excellent for both the ESM (AUC=0.991) and the combined-domains model (AUC=0.993). In the validation data set, the discriminative ability was reasonable for the actigraphy model (AUC=0.648) and excellent for both the ESM (AUC=0.891) and the combined-domains model (AUC=0.892). Conclusions ESM is a good diagnostic predictor and is easy to calculate, and it therefore holds promise for implementation in clinical practice. Actigraphy shows no added value to ESM as a diagnostic predictor but might still be useful when ESM use is restricted.

Actigraphically measured psychomotor slowing in depression: systematic review and meta-analysis

“…As psychomotor disturbance predicts better outcome of electroconvulsive therapy and might predict response to different classes of antidepressants, objective assessment of retardation may also inform therapy planning (Brancati et al, 2021; Buyukdura et al, 2011; Heijnen et al, 2019; Hickie, Mason, Parker, & Brodaty, 1996; Schrijvers et al, 2008; Valerio, Szmulewicz, & Martino, 2018; van Diermen et al, 2019). Moreover, actigraphy can enhance detection of depression in primary care settings with limited resources (Minaeva et al, 2020). In their meta-analysis, Burton et al (2013) found an increase in activity after treatment in MDD.…”

Uncovering complexity details in actigraphy patterns to differentiate the depressed from the non-depressed

“…Second, the fact that the complexity measures performed better in discriminating the two groups than simply the mean value of physical activity is worth noting. Reduction in physical activity is known to be a defining characteristic of depression and studies using objective actigraphy have shown that depressed individuals have a lower level of physical activity than individuals without depression 12 , 53 . In this context, the lack of significant differences in physical activity between the depressed and non-depressed groups, which was also observed previously in a sub sample of the same study 54 , is striking.…”

“…Currently, most interventions work under the assumption that it is only the pure level of activity that impacts mood. Other studies, however, argue that, in addition to mean levels of physical activity, diurnal rhythms in activity are relevant in explaining why activity levels are associated with depression 10 – 12 . Healthy people are most active closer to the middle of the day and less active in mornings and evenings.…”

Digital tools for the assessment of pharmacological treatment for depressive disorder: State of the art

“…Data for digital phenotyping are collected through smartphone apps, and digital tools such as smartwatches or other smart applications; the whole of interconnected smart tools or Internet of Things ( Goumagias et al, 2021 ), but also regular text messaging, social media posts ( Sinnenberg et al, 2017 ), web searches, blogs, and chatbots can be informa-tive ( Colder Carras et al, 2018 ;Torous et al, 2021 ). The nature of digital data can be diverse and can include the level of physical activity, actigraphy and circadian rhythm ( Minaeva et al, 2020 ). Social contacts and engagement in social media can be informative too ( Dissing et al, 2018 ), often in addition to Bluetooth and GPS information which can provide an insight in spatial proximities.…”

“…Different features of depression can be captured by different digital tools in a context of digital phenotyping. This table highlights some examples, the traits they measure and their phenotypic correlates, throughout the different stages of pharmacological treatment ( Ayers et al, 2017 ;Birk and Mandryk, 2019 ;Bernert et al, 2017 ;Büscher et al, 2020 ;Brietzke et al, 2019 ;Cohen et al, 2020a ;Colder Carras et al, 2018 ;Cho et al, 2019 ;Cheng et al, ;Dissing et al, 2018 ;Doryab et al, 2019 ;Faurholt-Jepsen et al, 2013Henry et al, 2020;Wallace et al, 2017 ;Insel, 2018 ;Izmailova et al, 2018 ;Jacobson et al, ;Kalantarian et al, 2016 ;Kivelä et al, 2019 ;Kramer et al, 2014 ;Minaeva et al, 2020 ;Pennebaker et al, 2003 ;Pratap et al, 2019 ;Perna et al, 2020 ;Pham et al, 2022 ;Philip et al, 2020;Vaidyam et al, 2019 ;Rohatagi et al, 2016 ;Stange et al, 2018 ;Sinnenberg et al, 2017 ;Sano et al, 2018 ;Sachan, 2018 ;Torous et al, 2021 ;Washington et al, ;Widdershoven et al, 2019 ;Yoshikawa et al, 2021 ;Zhou et al, 2022 ;Zulueta et al, 2018 ;Zijp et al, 2019 ).…”