Unlocking the potential of wearable device wear time to enhance postpartum depression screening and detection

Abstract

Postpartum depression (PPD) is a mood disorder affecting one in seven women after childbirth that is often under-screened and under-detected. If not diagnosed and treated, PPD is associated with long-term developmental challenges in the child and maternal morbidity. Wearable technologies, such as smartwatches and fitness trackers (e.g., Fitbit), offer continuous and longitudinal digital phenotyping for mood disorder diagnosis and monitoring, with device wear time being an important yet understudied aspect. Using the All of Us Research Program (AoURP) dataset, we assessed the percentage of days women with PPD wore Fitbit devices across pre-pregnancy, pregnancy, postpartum, and PPD periods, as determined by electronic health records. Wear time was compared in women with and without PPD using linear regression models. Results showed a strong trend that women in the PPD cohort wore their Fitbits more those without PPD during the postpartum (PPD: mean=72.9%, SE=13.8%; non-PPD: mean=58.9%, SE=12.2%, P-value=0.09) and PPD time periods (PPD: mean=70.7%, SE=14.5%; non-PPD: mean=55.6%, SE=12.9%, P-value=0.08). We hypothesize this may be attributed to hypervigilance, given the common co-occurrence of anxiety symptoms among women with PPD. Future studies should assess the link between PPD, hypervigilance, and wear time patterns. We envision that device wear patterns with digital biomarkers like sleep and physical activity could enhance early PPD detection using machine learning by alerting clinicians to potential concerns facilitating timely screenings, which may have implications for other mental health disorders.

Figure 1:. Flow diagram describing the inclusion and exclusion criteria of postpartum depression (PPD) and non PPD cohorts.

*AoURP = All of Us Research Program

Figure 2:. Women in the PPD cohort tended to wear their wearable device more than those in the non-PPD cohort during the postpartum and PPD time periods.

*PPD-eq = PPD-equivalent The percentage of days women in the PPD and non-PPD cohorts wore their wearable device during the pre-pregnancy (top left), pregnancy (top right), postpartum (bottom left), and PPD (or PPD-equivalent; bottom right) time periods. Data in the PPD and non-PPD cohorts were compared using linear regression adjusted for age at PPD diagnosis and race/ethnicity and are expressed as mean ± standard error.

Figure 3:. Women in the PPD and non-PPD cohorts wore their Fitbit for the same number of hours per day across time periods of pregnancy.

*PPD-eq = PPD-equivalent The number of hours per day women in the PPD and non-PPD cohorts wore their wearable device during the pre-pregnancy (top left), pregnancy (top right), postpartum (bottom left), and PPD (or PPD-equivalent; bottom right) time periods. Data in the PPD and non-PPD cohorts were compared using linear mixed-effects models with person ID as the random effect adjusted for age at PPD diagnosis and race/ethnicity and are expressed as mean ± standard error.

Figure 4:. Women in the PPD cohort tended to wear their devices more during sleep than those in the non-PPD cohort during the postpartum and PPD time periods.

*PPD-eq = PPD-equivalent The percentage of days women in the PPD and non-PPD cohorts wore their wearable device to sleep during the pre-pregnancy (top left), pregnancy (top right), postpartum (bottom left), and PPD (or PPD-equivalent; bottom right) time periods. Data in the PPD and non-PPD cohorts were compared using linear regression adjusted for age at PPD diagnosis and race/ethnicity and are expressed as mean ± standard error.

Figure 5:. Women in the PPD and non-PPD cohorts displayed the same pattern of correlation between wear time during earlier and later pregnancy time periods.

*PPD-eq = PPD-equivalent The Pearson correlation coefficient between the percentage of days women wore their Fitbit during pre-pregnancy and PPD (top) in addition to pregnancy and PPD (bottom). The blue line represents the line of best fit and gray shading shows the 95% confidence interval.

Figure 6:. Consistent Fitbit wear time during pre-pregnancy was not correlated with PPD wear time consistency among women in the PPD cohort.

*PPD-eq = PPD-equivalent The Pearson correlation coefficient between the maximum number of days in a row women PPD and non-PPD cohorts wore their Fitbit during pre-pregnancy and PPD (top) in addition to pregnancy and PPD (bottom). The blue line represents the line of best fit and gray shading shows the 95% confidence interval.