JMIR Publications

ABSTRACT

Background:

Postpartum depression (PPD) poses a significant maternal health challenge. The current approach to detecting PPD relies on in-person postpartum visits, which contributes to underdiagnosis. Furthermore, recognizing PPD symptoms can be challenging. Therefore, we explored the potential of using digital biomarkers from consumer wearables for PPD recognition.

Objective:

The main goal of this study was to showcase the viability of employing machine learning and digital biomarkers related to heart rate, physical activity, and energy expenditure, derived from consumer-grade wearables for the recognition of postpartum depression.

Methods:

Using the All of Us Research Program (AoURP) Registered Tier v6, we performed computational phenotyping of women with and without PPD following childbirth. Intra-individual machine learning (ML) models were developed using digital biomarkers from Fitbit to discern between pre-pregnancy, pregnancy, postpartum without depression, and postpartum with depression time periods (i.e., PPD diagnosis). Models were built using utilizing generalized linear models (GLM), random forest (RF), support vector machine (SVM), and k-nearest neighbors (KNN) algorithms and evaluated using kappa and multiclass area under the Receiver Operating Characteristic Curve (mAUC) to determine the algorithm with the best performance. The specificity of our individualized ML approach was confirmed in a cohort of women who gave birth and did not experience PPD. Moreover, we assessed the impact of a prior history of depression on model performance. We determined the variable importance for predicting the PPD time period using SHapley Additive exPlanations (SHAP) and confirmed the results with a permutation approach. Lastly, we compared our individualized ML methodology against a traditional cohort-based ML model for PPD recognition and compared model performance using sensitivity, specificity, precision, recall, and F1 score.

Results:

Patient cohorts of women who gave birth, with valid Fitbit data, included <20 with PPD and 39 without PPD. Our results demonstrated that intra-individual models using digital biomarkers discerned between pre-pregnancy, pregnancy, postpartum without depression, and postpartum with depression time periods (i.e., PPD diagnosis), with RF (mAUC = 0.85, kappa = 0.80) models outperforming GLM (mAUC = 0.82, kappa = 0.74), SVM (mAUC = 0.75, kappa = 0.72), and KNN (mAUC = 0.74, kappa = 0.62). Model performance decreased in women without PPD, illustrating the method's specificity. Prior depression history did not impact the efficacy of the model for PPD recognition. Moreover, we found that the most predictive biomarker of PPD was calories burned during the basal metabolic rate (calories BMR). Finally, individualized models surpassed the performance of a conventional cohort-based model for PPD detection.

Conclusions:

This research establishes consumer wearables as a promising tool for PPD identification and highlights personalized ML approaches, which could transform early disease detection strategies.