Article Sidebar
Main Article Content
Abstract
Background. Wrist actigraphy records motor behavior and circadian organization continuously, but it is unclear whether nonlinear complexity summaries can predict within-person change in depressive symptoms.
Objective. To test whether nonlinear actigraphy features improve participant-aware prediction of change in Montgomery–Åsberg Depression Rating Scale (MADRS) scores beyond a training-fold mean benchmark and clinical covariates.
Methods. This secondary analysis used the open OBF-Psychiatric depression subset: 23 adults with unipolar or bipolar depression, 551,716 minute-level activity observations, 359 complete days, and paired MADRS assessments. The outcome was endpoint minus baseline MADRS. Five daily nonlinear complexity features were averaged per participant. Ridge models with clinical, nonlinear, or combined predictors were evaluated by nested leave-one-subject-out validation; mean absolute error (MAE) was primary. Sensitivity analyses allowed an intercept-only option and used metadata-defined recording intervals.
Results. MADRS fell by a mean of 2.74 points (95% confidence interval [CI], −4.35 to −1.22; Wilcoxon p = 0.004). Benchmark MAE was 3.19 (95% CI, 2.26–4.28); the clinical, nonlinear, and combined models had MAEs of 3.31, 3.21, and 3.22 and negative cross-validated R². The nonlinear model's MAE exceeded the benchmark's by 0.021 (95% CI, 0.008–0.035). No nonlinear feature was associated with MADRS change after multiplicity correction. With an intercept-only option, nonlinear and combined models selected it in all 23 folds and matched the benchmark, as did metadata-defined intervals.
Conclusion. In this small cohort, nonlinear actigraphy complexity added no predictive value for short-term change in depression severity. Such features need participant-separated validation against a simple benchmark before being considered clinically informative.
Keywords
Article Details
As our aim is to disseminate original research article, hence the publishing right is a necessary one. The publishing right is needed in order to reach the agreement between the author and publisher. As the journal is fully open access, the authors will sign an exclusive license agreement.
The authors have the right to:
- Share their article in the same ways permitted to third parties under the relevant user license.
- Retain copyright, patent, trademark and other intellectual property rights including research data.
- Proper attribution and credit for the published work.
For the open access article, the publisher is granted to the following right.
- The non-exclusive right to publish the article and grant right to others.
- For the published article, the publisher applied for the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
