Digital Biomarkers for Depression Screening With Wearable Devices: Cross-sectional Study With Machine Learning Modeling

38 out of 267 (14.2%) participants had PHQ-9 scores of ≥10 at either of the two assessments and were identified as depressed.

“In total, 38 out of 267 (14.2%) participants had PHQ-9 scores of ≥10 at either of the two assessments and were identified as depressed.”

Participants experienced an overall decrease in depressive symptoms over the 2-week period.

“However, participants experienced an overall decrease in depressive symptoms over time; for example, 20 participants with moderate (score 10-14) depressive symptoms at the first assessment shifted to the mild range (score 5-9) at the second assessment, while just 5 participants with mild symptoms at the first assessment appeared to have moderate symptoms at the second assessment.”

A total of 267 participants were included in the analysis.

“A total of 267 participants were included in the analysis.”

• Also stated in: § Characteristics of the Data and Participants · plain

  “The final sample comprised of 267 participants with 14 days of tracking data, on average; the participant flowchart is presented in Multimedia Appendix 4.”

PHQ-9 scores at the two assessments are highly and linearly correlated (r=0.73, P<.001).

“PHQ-9 scores are highly and linearly correlated (r=0.73, P<.001), showing that the change in depressive symptoms over time is quite smooth and gradual rather than quick and sharp.”

Only 8 participants had PHQ-9 scores of ≥10 at both assessments.

“Hence, using the widely adopted cutoff score of ≥10 for both assessments significantly reduced the prevalence rate of depressed participants: only 8 participants had PHQ-9 scores of ≥10 at both assessments.”

22 (8.2%) participants had scores of ≥8 at both assessments.

“First, a meta-analysis by Manea et al [59] showed that a PHQ-9 cutoff score of ≥8 also has acceptable screening properties; in our sample, there were 22 (8.2%) participants who had scores of ≥8 at both assessments.”

For the default cutoff point, the sensitivity range was 3% to 13%.

“For the default cutoff point (ie, either baseline or follow-up PHQ-9 score of ≥10), the accuracy of the models in holdout folds was 86% (equal to the no information rate 86%), the sensitivity range was 3% to 13%, the specificity range was 98% to 100%, and the AUC range was 0.51 to 0.66 (Table 6, models A1-C1).”

For the second cutoff point, the specificity was 100%.

“For the second cutoff point option (ie, both baseline and follow-up PHQ-9 scores of ≥8), the accuracy of the models in holdout folds was 92% (no information ratio 92%), the sensitivity range was 0% to 5%, the specificity was 100%, and the AUC range was 0.54 to 0.67 (Table 6).”

Adding sociodemographic characteristics did not substantially improve performance of symptom-severity prediction models.

“Adding sociodemographic characteristics did not substantially improve performance (Table 5, models A2-C2).”

For the second cutoff point, the sensitivity range was 0% to 5%.

“For the second cutoff point option (ie, both baseline and follow-up PHQ-9 scores of ≥8), the accuracy of the models in holdout folds was 92% (no information ratio 92%), the sensitivity range was 0% to 5%, the specificity was 100%, and the AUC range was 0.54 to 0.67 (Table 6).”

For the default cutoff point (either baseline or follow-up PHQ-9 score ≥10), the accuracy of the models in holdout folds was 86% (equal to the no information rate 86%).

“For the default cutoff point (ie, either baseline or follow-up PHQ-9 score of ≥10), the accuracy of the models in holdout folds was 86% (equal to the no information rate 86%), the sensitivity range was 3% to 13%, the specificity range was 98% to 100%, and the AUC range was 0.51 to 0.66 (Table 6, models A1-C1).”

42 (15.7%) participants had an average PHQ-9 score ≥8.

“Second, given the linearity of PHQ-9 change over time, we used an average PHQ-9 score of ≥8 as another criterion; 42 (15.7%) participants had an average PHQ-9 score ≥8.”

Models with correlation-based feature selection outperformed all models trained without statistical feature selection.

“Among these models, the model with heart rate metrics had the best accuracy, sensitivity, and AUC—70%, 71%, and 72%, respectively—and the model with nonparametric circadian rhythm metrics had the best specificity of 76% (see Table S9 in Multimedia Appendix 1); however, models with correlation-based feature selection outperformed all of these models.”

• AssumptionThe authors state this as an assumption: Excluding participants with midrange (borderline) PHQ-9 scores results in a larger difference between groups and increases the discriminatory power of digital biomarkers.

Correlation analysis revealed that 36 digital biomarkers were significantly associated with the average PHQ-9 score.

“Correlation analysis revealed that 36 digital biomarkers were significantly associated with the average PHQ-9 score—absolute coefficients of Spearman rank correlation were weak and varied from 0.12 to 0.26—and only 17 of them remained significant predictors after the FDR correction (Table 3).”

Using more conservative feature selection criteria consistently resulted in better performance compared to less conservative criteria.

“Similar to the previous step of modeling, using more conservative feature selection criteria consistently resulted in better performance compared to the less conservative criteria.”

In contrasted subsamples of provisionally depressed participants and healthy participants with no risk of depression, the model achieved a sensitivity of 82% in detecting subjects at high risk of depression.

“a sensitivity of 82%”

In contrasted subsamples of provisionally depressed participants and healthy participants with no risk of depression, the model achieved a specificity of 78% in detecting subjects at high risk of depression.

“a specificity of 78%”

Greater severity of depressive symptoms is associated with fewer steps-based daily peaks.

“We found that a greater severity of depressive symptoms was robustly associated with greater variation of nighttime heart rate between 2 AM and 4 AM and between 4 AM and 6 AM; it was also associated with lower regularity of weekday circadian rhythms based on steps and estimated with nonparametric measures of interdaily stability and autocorrelation as well as fewer steps-based daily peaks.”

Across trained symptom-severity prediction models, the range of average MAE in holdout samples was 2.54 to 2.63.

“The range of mean correlations between actual and predicted PHQ-9 scores across trained models was 0.14 to 0.27 (R2=0.03-0.08), the range of mean RMSE was 3.10 to 3.20, and the range of average MAE in holdout samples was 2.54 to 2.63 (Table 5, models A1-C1).”

For the third cutoff point (average PHQ-9 score ≥8), the accuracy range of the models in holdout folds was 85% to 87% (no information ratio 84%).

“Finally, for the third cutoff point option (ie, average PHQ-9 score of ≥8), the accuracy range of the models in holdout folds was 85% to 87% (no information ratio 84%), the sensitivity range was 2% to 24%, the specificity range was 97% to 100%, and the AUC range was 0.62 to 0.74 (Table 6).”

Despite several reliable associations, the evidence showed limited ability of digital biomarkers to detect depression in the whole sample of working adults.

“Despite several reliable associations, our evidence showed limited ability of digital biomarkers to detect depression in the whole sample of working adults.”

For the third cutoff point, the AUC range was 0.62 to 0.74.

“Finally, for the third cutoff point option (ie, average PHQ-9 score of ≥8), the accuracy range of the models in holdout folds was 85% to 87% (no information ratio 84%), the sensitivity range was 2% to 24%, the specificity range was 97% to 100%, and the AUC range was 0.62 to 0.74 (Table 6).”

Adding sociodemographic characteristics did not improve classification accuracy for depression screening.

“Adding sociodemographic characteristics did not improve classification accuracy (Table 6, models A2-C2).”

For the default cutoff point, the AUC range was 0.51 to 0.66.

“For the default cutoff point (ie, either baseline or follow-up PHQ-9 score of ≥10), the accuracy of the models in holdout folds was 86% (equal to the no information rate 86%), the sensitivity range was 3% to 13%, the specificity range was 98% to 100%, and the AUC range was 0.51 to 0.66 (Table 6, models A1-C1).”

In general, the accuracy of models predicting severity of depressive symptoms and depression screening status based on digital biomarkers in the whole sample was poor.

“In general, the accuracy of models predicting severity of depressive symptoms and depression screening status based on digital biomarkers in the whole sample was poor, probably indicating a significant heterogeneity of data within groups of depressed and healthy participants.”

Applying more conservative feature selection criteria, including digital biomarkers most correlated to the outcome, showed relatively better results compared to less conservative criteria and including digital biomarkers less correlated to the outcome.

“However, applying more conservative feature selection criteria, including digital biomarkers most correlated to the outcome, showed relatively better results compared to the less conservative criteria and including digital biomarkers less correlated to the outcome.”

Greater severity of depressive symptoms is robustly associated with greater variation of nighttime heart rate between 4 AM and 6 AM.

“We found that a greater severity of depressive symptoms was robustly associated with greater variation of nighttime heart rate between 2 AM and 4 AM and between 4 AM and 6 AM; it was also associated with lower regularity of weekday circadian rhythms based on steps and estimated with nonparametric measures of interdaily stability and autocorrelation as well as fewer steps-based daily peaks.”

Across trained symptom-severity prediction models, the range of mean RMSE was 3.10 to 3.20.

“The range of mean correlations between actual and predicted PHQ-9 scores across trained models was 0.14 to 0.27 (R2=0.03-0.08), the range of mean RMSE was 3.10 to 3.20, and the range of average MAE in holdout samples was 2.54 to 2.63 (Table 5, models A1-C1).”

The absolute Spearman rank correlation coefficients between the 36 digital biomarkers and average PHQ-9 score were weak, ranging from 0.12 to 0.26.

“Correlation analysis revealed that 36 digital biomarkers were significantly associated with the average PHQ-9 score—absolute coefficients of Spearman rank correlation were weak and varied from 0.12 to 0.26—and only 17 of them remained significant predictors after the FDR correction (Table 3).”

These associations (lower regularity of heart rate circadian rhythm, fewer steps-based daily peaks, greater steepness of heart rate rhythm curve, later sleep midpoint, greater variation of sleep offset time) were less reliable and became nonsignificant in regression models with covariates.

“Additionally, we found that lower regularity of heart rate circadian rhythm, fewer steps-based daily peaks, greater steepness of the heart rate rhythm curve, later sleep midpoint, and greater variation of sleep offset time were associated with a greater severity of depressive symptoms, yet these associations were less reliable and became nonsignificant in regression models with covariates.”

Heart rate variation between 4 AM and 6 AM was positively associated with PHQ-9 scores in the fully adjusted model (β=0.90 per 10% change, 95% CI 0.04-1.78, P=.04).

“Heart rate variation between 4 AM and 6 AM was positively associated with PHQ-9 scores (β=0.90 per 10% change, 95% CI 0.04-1.78, P=.04)”

For the third cutoff point, the sensitivity range was 2% to 24%.

“Finally, for the third cutoff point option (ie, average PHQ-9 score of ≥8), the accuracy range of the models in holdout folds was 85% to 87% (no information ratio 84%), the sensitivity range was 2% to 24%, the specificity range was 97% to 100%, and the AUC range was 0.62 to 0.74 (Table 6).”

Greater severity of depressive symptoms was associated with lower regularity of weekday circadian activity rhythms based on steps and measured with nonparametric measures of IS and autocorrelation.

“We found that greater severity of depressive symptoms was robustly associated with greater variation of nighttime heart rate between 4 AM and 6 AM; it was also associated with lower regularity of weekday circadian activity rhythms based on steps and measured with nonparametric measures of IS and autocorrelation.”

Only 17 of the 36 digital biomarkers remained significant predictors of average PHQ-9 score after FDR correction.

“Correlation analysis revealed that 36 digital biomarkers were significantly associated with the average PHQ-9 score—absolute coefficients of Spearman rank correlation were weak and varied from 0.12 to 0.26—and only 17 of them remained significant predictors after the FDR correction (Table 3).”

Only three digital biomarkers-weekday steps-based IS, autocorrelation based on 15-minute intervals, and CV of heart rate between 4 AM and 6 AM-were consistently associated with symptom severity independent of all confounders (including alcohol consumption, smoking, self-rated health, loneliness, and subjective sleep characteristics).

“Only three digital biomarkers—weekday steps–based IS, autocorrelation (based on 15-minute intervals), and CV of heart rate between 4 AM and 6 AM—were consistently associated with symptom severity independent of all confounders, which additionally included alcohol consumption, smoking, self-rated health, loneliness, and subjective sleep characteristics (Table 4).”

Greater severity of depressive symptoms was robustly associated with greater variation of nighttime heart rate between 4 AM and 6 AM.

“We found that greater severity of depressive symptoms was robustly associated with greater variation of nighttime heart rate between 4 AM and 6 AM; it was also associated with lower regularity of weekday circadian activity rhythms based on steps and measured with nonparametric measures of IS and autocorrelation.”

Similar performance was achieved across all models trained using alternative contrasted subsamples.

“Similar performance has been achieved across all models trained using alternative contrasted subsamples.”

Greater severity of depressive symptoms is associated with lower regularity of weekday circadian rhythms based on steps, estimated with autocorrelation.

“We found that a greater severity of depressive symptoms was robustly associated with greater variation of nighttime heart rate between 2 AM and 4 AM and between 4 AM and 6 AM; it was also associated with lower regularity of weekday circadian rhythms based on steps and estimated with nonparametric measures of interdaily stability and autocorrelation as well as fewer steps-based daily peaks.”

Greater severity of depressive symptoms is associated with lower regularity of weekday circadian rhythms based on steps, estimated with interdaily stability.

“We found that a greater severity of depressive symptoms was robustly associated with greater variation of nighttime heart rate between 2 AM and 4 AM and between 4 AM and 6 AM; it was also associated with lower regularity of weekday circadian rhythms based on steps and estimated with nonparametric measures of interdaily stability and autocorrelation as well as fewer steps-based daily peaks.”

In contrasted subsamples of provisionally depressed participants and healthy participants with no risk of depression, the model achieved an accuracy of 80% in detecting subjects at high risk of depression.

“the model achieved an accuracy of 80%”

For the second cutoff point, the AUC range was 0.54 to 0.67.

“For the second cutoff point option (ie, both baseline and follow-up PHQ-9 scores of ≥8), the accuracy of the models in holdout folds was 92% (no information ratio 92%), the sensitivity range was 0% to 5%, the specificity was 100%, and the AUC range was 0.54 to 0.67 (Table 6).”

Our evidence showed limited ability of digital biomarkers to detect depression in the whole sample of working adults.

“Despite several reliable associations, our evidence showed limited ability of digital biomarkers to detect depression in the whole sample of working adults.”

Autocorrelation (based on 15-minute intervals) was negatively associated with PHQ-9 scores in the fully adjusted model (β=−0.38 per 10% change, 95% CI −0.75 to −0.01, P=.04).

“for autocorrelation: β=−0.38 per 10% change, 95% CI −0.75 to −0.01, P=.04”

Across trained symptom-severity prediction models, the range of R2 was 0.03 to 0.08.

“The range of mean correlations between actual and predicted PHQ-9 scores across trained models was 0.14 to 0.27 (R2=0.03-0.08), the range of mean RMSE was 3.10 to 3.20, and the range of average MAE in holdout samples was 2.54 to 2.63 (Table 5, models A1-C1).”

Weekday steps-based IS was negatively associated with PHQ-9 scores in the fully adjusted model (β=−0.30 per 10% change, 95% CI −0.51 to −0.08, P=.01).

“for IS in the fully adjusted model: β=−0.30 per 10% change, 95% CI −0.51 to −0.08, P=.01”

Greater severity of depressive symptoms is robustly associated with greater variation of nighttime heart rate between 2 AM and 4 AM.

“We found that a greater severity of depressive symptoms was robustly associated with greater variation of nighttime heart rate between 2 AM and 4 AM and between 4 AM and 6 AM; it was also associated with lower regularity of weekday circadian rhythms based on steps and estimated with nonparametric measures of interdaily stability and autocorrelation as well as fewer steps-based daily peaks.”

Physical activity metrics were not associated with PHQ-9 scores.

“Physical activity metrics were not associated with PHQ-9 scores.”

Across trained symptom-severity prediction models, the range of mean correlations between actual and predicted PHQ-9 scores was 0.14 to 0.27.

“The range of mean correlations between actual and predicted PHQ-9 scores across trained models was 0.14 to 0.27 (R2=0.03-0.08), the range of mean RMSE was 3.10 to 3.20, and the range of average MAE in holdout samples was 2.54 to 2.63 (Table 5, models A1-C1).”

For the default cutoff point, the specificity range was 98% to 100%.

“For the default cutoff point (ie, either baseline or follow-up PHQ-9 score of ≥10), the accuracy of the models in holdout folds was 86% (equal to the no information rate 86%), the sensitivity range was 3% to 13%, the specificity range was 98% to 100%, and the AUC range was 0.51 to 0.66 (Table 6, models A1-C1).”

Effects of these digital biomarkers on symptom severity were stable and independent of all confounders including strong predictors of depression, such as sleep quality and loneliness.

“Effects of these digital biomarkers on symptom severity were stable and independent of all confounders including strong predictors of depression, such as sleep quality and loneliness.”

For the second cutoff point (both baseline and follow-up PHQ-9 scores ≥8), the accuracy of the models in holdout folds was 92% (no information ratio 92%).

“For the second cutoff point option (ie, both baseline and follow-up PHQ-9 scores of ≥8), the accuracy of the models in holdout folds was 92% (no information ratio 92%), the sensitivity range was 0% to 5%, the specificity was 100%, and the AUC range was 0.54 to 0.67 (Table 6).”

For the third cutoff point, the specificity range was 97% to 100%.

“Finally, for the third cutoff point option (ie, average PHQ-9 score of ≥8), the accuracy range of the models in holdout folds was 85% to 87% (no information ratio 84%), the sensitivity range was 2% to 24%, the specificity range was 97% to 100%, and the AUC range was 0.62 to 0.74 (Table 6).”

Regularity of circadian rhythm and variation of nighttime heart rate were the most robust digital biomarkers.

“Regularity of circadian rhythm and variation of nighttime heart rate were the most robust digital biomarkers”

Greater severity of depressive symptoms was associated with fewer steps-based daily peaks.

“we first showed that a greater severity of depressive symptoms was associated with a fewer steps-based daily peaks”

Our data did not show reduced levels of locomotor activity in depressed participants in terms of less time spent in moderate to vigorous physical activity, fewer daily steps, or more sedentary time.

“our data did not show reduced levels of locomotor activity in depressed participants in terms of less time spent in moderate to vigorous physical activity, fewer daily steps, or more sedentary time.”

Weekday circadian rhythms were stronger predictors than rhythms based on all days.

“Second, we analyzed weekday circadian rhythms separately that were found to be stronger predictors rather than rhythms based on all days.”

Greater variation of nighttime heart rate between 2 AM and 4 AM and between 4 AM and 6 AM indicates greater severity of depressive symptoms.

“our findings suggest that a greater variation of nighttime heart rate between 2 AM and 4 AM and between 4 AM and 6 AM indicates greater severity of depressive symptoms”

• cites [56] sourceNot checked yet

  Not checked in the free run.

• cites [61] sourceNot checked yet

  Not checked in the free run.

A lower steps-based weekday IS was robustly associated with a greater severity of depressive symptoms independent of confounders.

“we found that a lower steps-based weekday IS—a nonparametric measure of rhythm regularity—was robustly associated with a greater severity of depressive symptoms independent of confounders”

• cites [31] sourceNot checked yet

  Not checked in the free run.

• cites [60] sourceNot checked yet

  Not checked in the free run.

• cites [34] sourceNot checked yet

  Not checked in the free run.

Our best model had 9% higher sensitivity (82% vs 73%) than the model of Tazawa et al.

“Although our best model had 9% higher sensitivity (82% vs 73%)”

• cites [33] sourceNot checked yet

  Not checked in the free run.

Later sleep midpoint and offset time were associated with more severe depressive symptoms.

“the analysis of sleep data showed that later sleep midpoint and offset time were associated with more severe depressive symptoms”

• cites [16] sourceNot checked yet

  Not checked in the free run.

• cites [20] sourceNot checked yet

  Not checked in the free run.

• cites [22] sourceNot checked yet

  Not checked in the free run.

• cites [23] sourceNot checked yet

  Not checked in the free run.

Steps-based autocorrelation during weekdays was robustly correlated to symptom severity independent of all confounders.

“we also found steps-based autocorrelation during weekdays—a complementary and alternative nonparametric measure of rhythm regularity—to be robustly correlated to symptom severity independent of all confounders.”

The set of selected digital biomarkers used in the retrained models remained the same as in models with the full data set.

“the set of selected digital biomarkers used in these models remained the same as in models with the full data set.”

277 of 290 participants (95.5%) had at least 10 days with enough data.

“A total of 77.7% (4025/5180) of days contained at least 20 hours of complete data; however, only 277 participants out of 290 (95.5%) had at least 10 days with enough data (see Table S1 in Multimedia Appendix 1).”

The change in depressive symptoms over the 2-week period is smooth and gradual rather than quick and sharp.

“PHQ-9 scores are highly and linearly correlated (r=0.73, P<.001), showing that the change in depressive symptoms over time is quite smooth and gradual rather than quick and sharp.”

77.7% of days contained at least 20 hours of complete data.

“A total of 77.7% (4025/5180) of days contained at least 20 hours of complete data; however, only 277 participants out of 290 (95.5%) had at least 10 days with enough data (see Table S1 in Multimedia Appendix 1).”

A meta-analysis by Manea et al showed that a PHQ-9 cutoff score of ≥8 has acceptable screening properties.

“First, a meta-analysis by Manea et al [59] showed that a PHQ-9 cutoff score of ≥8 also has acceptable screening properties; in our sample, there were 22 (8.2%) participants who had scores of ≥8 at both assessments.”

• cites [59] sourceNot checked yet

  Not checked in the free run.

Patients with depression are significantly less physically active than healthy controls.

“Systematic reviews and meta-analyses of actigraphy studies demonstrated that patients with depression were significantly less physically active than healthy controls [15,16].”

• cites [15] sourceNot checked yet

  Not checked in the free run.

• cites [16] sourceNot checked yet

  Not checked in the free run.

Mean compliance with wearing the activity tracker was 84.8% (SD 10.4).

“The mean compliance with wearing the activity tracker among participants within the observation period was 84.8% (SD 10.4), whereas the total fraction of complete data was 84.0% (6,265,469 minutes of complete data from 5180 days).”

Close interrelationships between everyday behavior, physiology, and mental well-being make digital phenotyping with wearables attractive for detection of mental disorders.

“Close interrelationships between everyday behavior, physiology, and mental well-being makes digital phenotyping with wearables especially attractive for detection of mental disorders and discovering respective digital biomarkers [8-11].”

• cites [8] sourceNot checked yet

  Not checked in the free run.

• cites [9] sourceNot checked yet

  Not checked in the free run.

• cites [10] sourceNot checked yet

  Not checked in the free run.

• cites [11] sourceNot checked yet

  Not checked in the free run.

Having a PHQ-9 score of ≥10 at any time point means that the score at an adjacent time point is likely to be comparable in magnitude.

“Having a PHQ-9 score of ≥10 at any time point means that the score at an adjacent time point is likely to be comparable in magnitude.”

The poor accuracy probably indicates significant heterogeneity of data within groups of depressed and healthy participants.

“In general, the accuracy of models predicting severity of depressive symptoms and depression screening status based on digital biomarkers in the whole sample was poor, probably indicating a significant heterogeneity of data within groups of depressed and healthy participants.”

The performance of models classifying the outcome of depression screening was relatively better when the third cutoff point (average PHQ-9 score ≥8) was applied.

“Thus, the performance of models classifying the outcome of depression screening was relatively better when the third cutoff point was applied.”

Using feature selection on the entire sample allows mitigation of overgeneralization that is possible in skewed samples.

“This allows us to mitigate overgeneralization that is possible in skewed samples, because models were trained with digital biomarkers inferred from the entire sample.”

Participants with lower stability of circadian activity rhythms had more severe depressive symptoms.

“so participants with lower stability of circadian activity rhythms had more severe symptoms.”

A greater magnitude of heart rate variation between 4 AM and 6 AM indicated higher PHQ-9 scores.

“where a greater magnitude of heart rate variation indicated higher PHQ-9 scores.”

In this study, using statistical analysis and machine learning, some known and novel digital biomarkers based on behavioral and physiological data from consumer wearables could indicate increased risk of depression in a multiethnic working population.

“In this study, using statistical analysis and machine learning, we demonstrated that some known and novel digital biomarkers based on behavioral and physiological data from consumer wearables could indicate increased risk of depression in a multiethnic working population.”

• AssumptionThe authors state this as an assumption: Excluding participants with midrange (borderline) PHQ-9 scores results in a larger difference between groups and increases the discriminatory power of digital biomarkers.

Machine learning models combining these digital biomarkers could discriminate between individuals with a high risk of depression and individuals with no risk.

“Machine learning models combining these digital biomarkers could discriminate between individuals with a high risk of depression and individuals with no risk.”

Novel rhythm stability metrics-autocorrelation and ICV-have value as risk markers of depression.

“Third, we showed the value of novel rhythm stability metrics—autocorrelation and ICV—as risk markers of depression.”

The study included a relatively long period of continuous sensing and activity tracking in free-living settings.

“Strengths of our study include a relatively long period of continuous sensing and activity tracking in free-living settings”

Predictive models based on a combination of these digital biomarkers could quite accurately discriminate individuals with a high risk of depression from individuals with no risk.

“Thus, predictive models based on a combination of these digital biomarkers could quite accurately discriminate individuals with a high risk of depression from individuals with no risk.”

The study used a relatively large workplace-based sample.

“a relatively large workplace-based sample”

An increased arousal of the autonomic nervous system that is possible with depression is likely to affect heart rate dynamics during sleep.

“an increased arousal of autonomic nervous system that is possible with depression is likely to affect heart rate dynamics during sleep [56].”

• cites [56] sourceNot checked yet

  Not checked in the free run.

People with more severe depressive symptoms tended to have less stable circadian activity rhythms.

“In general, results of our study together with previous evidence demonstrated that people with more severe depressive symptoms tended to have less stable circadian activity rhythms.”

Nonparametric indicators appear more robust and sensitive to indicate depressive symptomatology in nonclinical samples than cosinor-based metrics.

“Perhaps this supports the advantage of nonparametric indicators over cosinor-based metrics, where the former appear more robust and sensitive to indicate depressive symptomatology in nonclinical samples.”

The predictive model based on a combination of these digital biomarkers could discriminate individuals with high risk of depression from individuals with no risk.

“The predictive model based on a combination of these digital biomarkers could discriminate individuals with high risk of depression from individuals with no risk.”

Some known and novel digital biomarkers based on data from consumer wearables could indicate increased risk of depression in the working population.

“This study showed that some known and novel digital biomarkers based on data from consumer wearables could indicate increased risk of depression in the working population.”

The absence of participants with clinical depression might cause a lack of contrast in behavioral and physiological data between depressed and healthy participants, which impedes discovery of reliable digital biomarkers.

“The absence of participants with clinical depression might cause a lack of contrast in behavioral and physiological data between depressed and healthy participants, which impedes discovery of reliable digital biomarkers.”

The poor predictability of depressive symptomatology in the whole sample highlights a possible limitation of using this set of digital biomarkers alone for depression screening universally due to prominent interindividual differences and heterogeneity in naturalistic settings.

“the poor predictability of depressive symptomatology in the whole sample highlights a possible limitation, in principle, of using this set of digital biomarkers alone for depression screening universally due to prominent interindividual differences and heterogeneity that appear in naturalistic settings [36].”

• cites [36] sourceNot checked yet

  Not checked in the free run.