À retenir
Using Fitbit data from All of Us, individualized random forest models separated prepregnancy, pregnancy, postpartum without depression and postpartum depression periods (mAUC = 0.85; κ = 0.80). The cohort with postpartum depression (PPD) had fewer than 20 women, so the results are preliminary.
Study design
Étude transversale
Reading basis
Published abstract
Reading focus
Association & confounding
Cette explication éditoriale existe en anglais, chinois simplifié et chinois traditionnel. Cette interface affiche la lecture en anglais.
Question de recherche
Postpartum depression is underdiagnosed because detection relies on in-person postpartum visits. The authors tested whether heart rate, physical activity and energy expenditure signals from consumer wearables, analyzed with machine learning (ML), can recognize PPD.
Population et méthode
Cross-sectional analysis of the All of Us Research Program Registered Tier v6 data set. Intraindividual ML models used Fitbit digital biomarkers to discern prepregnancy, pregnancy, postpartum without depression and postpartum with depression (PPD diagnosis) periods. Generalized linear models, random forest, support vector machine and k-nearest neighbor were compared by the κ statistic and multiclass area under the receiver operating characteristic curve (mAUC). Specificity was checked in women without PPD, the effect of past depression was assessed, variable importance used Shapley additive explanations, and the approach was compared with a cohort-based ML model.
Résultats
The authors’ reported results. Read each estimate with its population and uncertainty.
- 1The cohorts with valid Fitbit data included fewer than 20 women with PPD and 39 without PPD.
- 2Random forest performed best (mAUC = 0.85; κ = 0.80), ahead of generalized linear models (mAUC = 0.82; κ = 0.74), support vector machine (mAUC = 0.75; κ = 0.72) and k-nearest neighbor (mAUC = 0.74; κ = 0.62).
- 3Performance decreased in women without PPD, which the authors take as evidence of specificity, and previous depression history did not affect performance.
- 4The most predictive biomarker was calories burned during the basal metabolic rate. Individualized models outperformed a conventional cohort-based model.
Limites et incertitude
- 1Fewer than 20 women with PPD is a tiny sample, so performance estimates are unstable and may not generalize.
- 2Models were fitted within each woman's own history and distinguished periods, not predicted depression ahead of time in new women.
- 3All of Us participants who share Fitbit data are a self-selected group, and the abstract does not report cohort-based model results.
- 4This reading is based on the abstract only.
Mise en perspective
This is a proof of concept that wearable patterns shift across pregnancy and the postpartum period in ways a model can pick up. It is not a screening tool and cannot tell any individual whether they have depression. Anyone who feels low, anxious or unable to cope after childbirth should contact a clinician.
Vos données
Fitbit-style heart rate, activity and calorie estimates were the inputs here, but note that the top signal was a modeled basal-metabolism estimate rather than a direct measurement. Major life changes like pregnancy shift these numbers, so compare against your own baseline on one device and treat changes as a prompt to check in, not a diagnosis.
Comprendre ce type de preuve
Measurements are taken in the same period. Direction and causation usually cannot be determined; symptoms can change behaviour as well as the reverse.
Résumé structuré des auteurs English
Le résumé original en anglais est reproduit avec attribution sous la licence ouverte indiquée. Il s’agit d’une source, pas d’une interprétation MoveVitals. L’analyse préparée figure ci-dessus ; vérifiez les conclusions importantes dans la source.
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 using machine learning (ML) and digital biomarkers related to heart rate, physical activity, and energy expenditure derived from consumer-grade wearables for the recognition of PPD.
Methods
Using the All of Us Research Program Registered Tier v6 data set, we performed computational phenotyping of women with and without PPD following childbirth. Intraindividual ML models were developed using digital biomarkers from Fitbit to discern between prepregnancy, pregnancy, postpartum without depression, and postpartum with depression (ie, PPD diagnosis) periods. Models were built using generalized linear models, random forest, support vector machine, and k-nearest neighbor algorithms and evaluated using the κ statistic 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 previous history of depression on model performance. We determined the variable importance for predicting the PPD period using Shapley additive explanations and confirmed the results using a permutation approach. Finally, 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 F_(1)-score.
Results
Patient cohorts of women with valid Fitbit data who gave birth included <20 with PPD and 39 without PPD. Our results demonstrated that intraindividual models using digital biomarkers discerned among prepregnancy, pregnancy, postpartum without depression, and postpartum with depression (ie, PPD diagnosis) periods, with random forest (mAUC=0.85; κ=0.80) models outperforming generalized linear models (mAUC=0.82; κ=0.74), support vector machine (mAUC=0.75; κ=0.72), and k-nearest neighbor (mAUC=0.74; κ=0.62). Model performance decreased in women without PPD, illustrating the method's specificity. Previous 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. 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.
Sources et attribution
Harnessing Consumer Wearable Digital Biomarkers for Individualized Recognition of Postpartum Depression Using the All of Us Research Program Data Set: Cross-Sectional Study.
Hurwitz E, Butzin-Dozier Z, Master H, O'Neil ST, Walden A, Holko M, Patel RC, Haendel MA.
JMIR mHealth and uHealth · 2024-05-02
DOI: 10.2196/54622 ↗PubMed · 38696234 ↗Europe PMC · Licence indiquée: cc by ↗Abstract © the original authors, reproduced under the recorded open licence. Translations support reading; use the English record to check terminology and numerical results. Our methodological interpretation is separate from the authors’ conclusions.
Interprétation éditoriale MoveVitals · contenu éducatif · aucune évaluation clinique individuelle enregistrée
Sources vérifiées 2026-11-01
Guides pratiques
À lire ensemble
Activity–stroke and heart attack curves differ by accelerometer processing method
Among 90,237 UK Biobank adults free of cardiovascular disease at baseline, followed until December 2022 for 1,298 strokes and 2,031 myocardial infarctions, the shape of the link between moderate-vigorous physical activity (MVPA) and each outcome depended on how the accelerometer data were processed. For stroke, hazard fell in a straight line only with the machine-learning metric; for myocardial infarction, two metrics gave a curved decline and activity counts a straight one. The abstract gives no hazard ratios, so it shows that methods disagree, not which one is right, and the findings are observational associations.
Preventive Medicine
Smartwatch data as forensic evidence: heart rate is best validated, the rest uneven
A systematic review and meta-analysis of studies on smartwatches from four major brands, framed around digital forensics, found that heart rate was both the most studied and the most accurate measure. Step count, distance, energy expenditure, activity classification and sleep were less consistently validated, and results varied by brand, protocol and reporting. The abstract gives no pooled numbers, and most studies were run in healthy people in laboratories, so it cannot say how a particular watch performs for a particular person in daily life.
Forensic Science International
Activity trackers: about 1,800 more steps a day, but small effects on other outcomes
An umbrella review of 39 systematic reviews and meta-analyses, covering 163,992 participants, found that activity trackers were associated with more physical activity (SMD 0.3–0.6), equating to about 1,800 extra steps and 40 minutes more walking per day, and with a bodyweight reduction of about 1 kg. Effects on blood pressure, cholesterol, glycosylated haemoglobin, quality of life and pain were typically small and often non-significant. The abstract does not say what else came with the trackers in these programs or how much the reviews overlap, so the gains cannot be credited to the device alone.
The Lancet Digital Health
Seven sleep trackers: good at spotting sleep, weak on wake and sleep stages
In a sleep laboratory, 34 healthy young adults wore a subset of seven consumer sleep trackers for three nights. Against polysomnography every device detected sleep very well (sensitivity 0.93 or higher) but caught wake poorly (specificity 0.18–0.54), and five of the seven matched or beat research actigraphy (specificity 0.39). Sleep-stage estimates were inconsistent, and the participants were healthy adults aged 18 to 35 in a fixed 8-hour bed window, with each device tested in only 11 to 20 of them, so the results say little about people with sleep problems.
Sleep

