Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Comparison of longitudinal PSQI and actigraphy-assessed sleep from pregnancy to postpartum

  • Ryan L. Brown,

    Roles Formal analysis, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Department of Human Development and Family Sciences, Texas Tech University, Lubbock, Texas, United States of America

    ⨯
  • Chloe M. Beverly Hery,

    Roles Writing – original draft

    Affiliation Comprehensive Cancer Center, The Ohio State University, Columbus, Ohio, United States of America

    ⨯
  • Aric A. Prather,

    Roles Writing – review & editing

    Affiliation Department of Psychiatry and Behavioral Sciences, University of California, San Francisco, California, United States of America

    ⨯
  • Lisa M. Christian

    Roles Conceptualization, Funding acquisition, Supervision, Writing – review & editing

    Lisa.Christian@osumc.edu

    Affiliations Department of Psychiatry & Behavioral Health, The Ohio State University Wexner Medical Center, Columbus, Ohio, United States of America, Institute of Behavioral Medicine Research, The Ohio State University Wexner Medical Center, Columbus, Ohio, United States of America

    ⨯

Abstract

Despite the importance of sleep for perinatal health, there is limited research examining whether different measurement modalities may yield inconsistent data from pregnancy through postpartum. We aimed to: 1) describe how sleep patterns change across pregnancy and postpartum using self-report PSQI and actigraphy measures and 2) determine the level of correspondence between these two measurement modalities. Pregnant women from the Stress and Health in Pregnancy and Postpartum (SHIPP) study completed visits during the 3rd trimester, 4–6 weeks postpartum, and 4 months, 8 months, and 12 months postpartum. At each study visit, participants completed questionnaires and wore wrist-actigraphy (Actiwatch 2) for one week prior to each visit. Self-reported global sleep quality was measured with the Pittsburgh Sleep Quality Index (PSQI). Actigraphy and self-reported PSQI sleep characteristics were summarized at each of the five study visits. Using generalized linear mixed modeling, we examined if there were differences by sleep measurement (actigraphy vs. PSQI) for the overlapping sleep outcomes: total sleep time, time in bed, sleep latency, and sleep efficiency. Participants (n = 74; 28.9 years ± 4.6) were mostly white (73%), non-Hispanic (96%), married (78.5%) and over 60% had at least one child previously. Average PSQI global score was > 5 (cutoff for poor sleep) at each study visit. Total sleep time, sleep efficiency, and sleep latency measurements were significantly different between self-report and actigraphy throughout pregnancy and postpartum. Actigraphy-assessed sleep may reflect longer total sleep times, shorter sleep latency, and greater sleep efficiency compared to self-reported sleep among pregnant and postpartum women. This may be due to measurement error in actigraphy or recall bias when completing self-reported sleep measures. These factors should be taken into consideration both at the time of study design and when comparing results from different studies to facilitate the highest quality research and clinical decision-making in this population.

Citation: Brown RL, Hery CMB, Prather AA, Christian LM (2025) Comparison of longitudinal PSQI and actigraphy-assessed sleep from pregnancy to postpartum. PLoS One 20(6): e0323489. https://doi.org/10.1371/journal.pone.0323489

Editor: Sascha Köpke, University Hospital Cologne: Uniklinik Koln, GERMANY

Received: October 3, 2024; Accepted: April 8, 2025; Published: June 18, 2025

Copyright: © 2025 Brown et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the manuscript and its Supporting Information files.

Funding: This study was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) [grant numbers HD061644 and HD067670 to L.M.C.]. The project described was also supported by Award Number UM1TR004548 from the National Center for Advancing Translational Sciences. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Advancing Translational Sciences or the National Institutes of Health. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Pregnant and postpartum women commonly experience various sleep issues [1]. Sleep issues that arise during pregnancy tend to be pain-related insomnia, positional discomfort, leg cramps, and frequent urination [2,3], whereas sleep issues during postpartum are more often fragmented or interrupted sleep and shorter sleep durations due to infant sleep schedules [4,5]. Poor maternal sleep can impact the mother’s mental health and functioning as well as increase the risk of adverse pregnancy outcomes, such as intrauterine growth restriction and preterm birth [6–11]. Accordingly, assessing sleep is a focus of many studies on perinatal mental health.

However, the measurement modality used to assess sleep characteristics may be affected differently across pregnancy and postpartum based on stage of pregnancy and the needs of one’s infant. While pregnant, women report more snoring late in their pregnancy and increases in frequency of restless less syndrome and daytime sleepiness [12]. Hormonal changes during pregnancy can also affect sleep, including altering sleep architecture [13]. In contrast, postpartum sleep disturbances are characterized by nighttime awakenings from infant sleep and feeding schedules, resulting in poor sleep quality and shortened sleep duration, most often for mothers [4,5,14].

Despite the substantial literature on sleep during pregnancy and postpartum, most studies in this area are cross-sectional with few studies quantifying changes in sleep over time from pregnancy through postpartum [15,16]. Poor sleep quality generally increases throughout pregnancy, peaks during the first few months postpartum, and tends to remain elevated for the first six months to one year [3,17,18]. Yet, repeated measures using self-report and objective measures of sleep changes, such as actigraphy, across the perinatal period are lacking, with only a few exceptions [19–22]. There are clear benefits and challenges to measuring sleep subjectively in new parents. Self-reported sleep, including questionnaires or sleep diaries, is an inexpensive and low burden option; however, self-report measures are subject to recall bias and depend on the participant to accurately and fully complete all portions of the sleep assessment [23]. While polysomnography is considered the gold standard of sleep research, it can be costly and difficult to schedule in-laboratory visits for new parents. Therefore, more research uses wrist-worn devices to collect actigraphy data to overcome these barriers and provide less obtrusive, objective sleep measures [23,24].

When considering actigraphy-assessed sleep throughout the perinatal period, there are several limitations to the available literature such as small samples and a lack of postpartum assessments. Six studies in the United States have examined changes in parental sleep from pregnancy to postpartum using actigraphy [25–30], of which four had small sample sizes of ≤30 participants [27–30]. Additionally, none of these studies have examined sleep after 16 weeks postpartum, when many sleep complaints do not begin to occur until the third trimester (after 28 weeks) and few report sleep metrics beyond total sleep time and sleep efficiency. In the current study, we report actigraphy-assessed sleep from third trimester to one year postpartum to address this gap in postpartum measurement.

A limited number of studies have examined differences between self-reported sleep measures (e.g., sleep diary or the Pittsburgh Sleep Quality Index (PSQI)) and actigraphy collected sleep measures concurrently among pregnant and postpartum women [21,31–33]. Of these, only two have compared the self-report measure to the actigraphy measures of sleep during pregnancy [21,31]. In these studies, there were discrepancies identified based on measurement modality for sleep duration, wake after sleep onset, and sleep efficiency during pregnancy [31]. Over 60% of participants self-reported sleep duration estimates that differed by over an hour compared to actigraphy [21], actigraphy-assessed wake after sleep onset was 3.5 times greater than self-reported [31], and self-reported sleep efficiency was much higher than actigraphy-assessed [31]. More studies are needed to determine the robustness of these results and extend the question into the postpartum period. Thus, the objectives of this study were to 1) describe how sleep patterns change across pregnancy and postpartum using both PSQI self-reported sleep measures and actigraphy measures and 2) determine the level of correspondence between these two complementary measurement modalities.

Materials and methods

Study design

The Stress and Health in Pregnancy and Postpartum (SHIPP) study included pregnant women from the central Ohio area. Briefly, this study was designed to examine associations between psychosocial factors, mood, health behaviors, neuroendocrine and immune function from the pregnancy through postpartum. Study visits were conducted during the 3rd trimester, 4–6 weeks postpartum, and 4 months, 8 months, and 12 months postpartum. Participants completed questionnaires and psychosocial interviews at each visit and wrist-actigraphy was collected data for one week prior to each in-person study visit. Recruitment occurred from December 8, 2016, to April 9, 2018. The current analyses focus on self-reported sleep as well as actigraphy-based sleep data. The study methods and results follow best practice recommendations and reporting guidelines for Strengthening the Reporting of Observational Studies in Epidemiology (STROBE).

Participants

Women were eligible for this study if they were between the ages of 18–40 years old and were between 29–34 weeks gestation. Women were excluded if they had multifetal gestation, fetal anomaly, chronic health problems with a clear immunological or endocrinological component (e.g., cancers, recent surgeries, strokes, or conditions that require regular use of anti-inflammatories), worked night shifts, or reported alcohol use or illicit drug use (other than marijuana). In addition, women who were underweight (pre-pregnancy body mass index, BMI, < 18.5 kg/m2) or had a BMI > 50 kg/m2 were excluded. Women reporting recent acute illness (e.g., cold) or vaccinations in the past 7 days were rescheduled for a later visit. Written informed consent and Health Insurance Portability and Accountability Act (HIPAA) authorizations were obtained from all participants and each received modest compensation. The study was approved by the OSU Biomedical Institutional Review Board.

A total of 83 women were enrolled in this study with 74 providing actigraphy data at baseline and at least one other time point. For each assessment window, data were only used if ≥ 3 days of valid actigraphy data were collected, as this has been deemed a minimum criterion for reliable and valid data [34]. Across all five study visits, there were only 23 cases (~8% of data points) of individuals with only three to four nights of actigraphy data recorded. Therefore, the analyzed sample included 59 participants at visit 1, 66 at visit 2, 60 at visit 3, 62 at visit 4, and 56 at visit 5.

Demographic data

Age, race/ethnicity, marital status, education, income level, and parity, were self-reported. Pre-pregnancy body mass index (BMI) was calculated using self-reported pre-pregnancy weight and measured height at the first visit. Length of gestation at each visit was calculated from the visit date and the expected due date. Length of gestation at delivery was calculated from the date of delivery.

Self-reported sleep

At every study visit, subjective sleep was measured using the Pittsburgh Sleep Quality Index (PSQI), an 18-item measure of sleep quality over the past four weeks [35]. A PSQI global score is calculated by summing all seven component responses and ranges from 0 to 21. A higher score indicates worse sleep quality and a global score of ≥ 5 is the standard cut-off for poor sleep [35]. The individual PSQI items for hours in bed, sleep duration, sleep latency, and a calculated percent sleep efficiency were used in this analysis.

Actigraphy-measured sleep

Wrist actigraphy data were collected for 7 days prior to each study visit using the Actiwatch 2 (Philips Respironics, Murrysville, PA, USA). Participants were instructed to wear the actiwatch on their non-dominant wrist for 24 hours per day. Actigraphy data collected from the Actiwatch 2 were downloaded using the Actiware Sleep Scoring Program (v6.0.9, Philips-Respironics).

Sleep actigraphy data files for each participant were scored in Actiware using the validated manufacturer’s standard algorithm at medium sensitivity [34]. Two independent scorers determined the validity of each recording based on standard sleep criteria, adapted from other actigraphy studies [36–38]. A recording could be considered invalid for two reasons: 1) an Actiwatch device malfunction indicated by false activity and 2) Actiwatch data was unable to be retrieved. Individual days within a recording could also be considered invalid if an Actiwatch error occurred, such as a failing battery indicated by false activity, or if the participant did not wear the watch for an appropriate amount of time per day (i.e., majority of day off-wrist, or an off-wrist period > 60 minutes within 10 minutes of the determined beginning or end of the sleep period). Two scorers reviewed the automated rest intervals from Actiware and manually adjusted the interval as needed based on evidence of increased or decreased activity levels, rapidly decreased light levels to suggest bedtimes, and rapidly increased light levels to suggest wake times. For example, decreasing activity levels refers to the scorers considering how the activity intensity decreases surrounding periods of suspected sleep within the context of the entire participant’s actigraphy profile. Scorers visually reviewed the entire sleep recording before manually adjusting any sleep periods to consider the activity intensity level of the participant to ensure periods of low movement, but not sleep (e.g., reading in bed, watching TV) were not included. Sleep diary data was used to help suggest, but not confirm, sleep periods as needed.

Each file was reviewed by an unbiased 3rd scorer for interrater reliability. For any interval with bed and wake times that differed by ≥ 15 minutes between scorers, the 3rd scorer reviewed and corrected those intervals accordingly. Any disagreements in scoring differences were adjudicated by the primary study investigator (LMC). Resulting actigraphy variables calculated from the scored sleep intervals included: time in bed, total sleep time, wake after sleep onset, sleep efficiency, and sleep latency.

Statistical analysis

Participant characteristics at baseline were described using means with standard deviations (SD) for continuous variables and frequencies with percentages for categorical variables. Actigraphy and self-reported sleep characteristics were summarized (means with SD) at each of the five study visits. Cross-sectional correlations (i.e., at each time point from late pregnancy to postpartum) reflect the correspondence between actigraphy and PSQI measures at each time point throughout the transition to parenthood.

For the primary analyses, we aimed to determine if there were significant differences in each overlapping sleep outcome measure based on the method of data collection and if these differences emerged at particular time points during pregnancy or postpartum. We used generalized linear mixed modeling (GLMM), a multilevel regression analytic technique, to fit each outcome variable (i.e., time in bed, total sleep time, sleep latency, sleep efficiency) across each model: first as a function of the main effect of method of data collection (self-report vs. actigraphy) and then based on the interaction between time (visit number) and method of data collection (self-report vs. actigraphy). In post-hoc sensitivity analyses, we confirmed that the presented results were unchanged when restricting the sample to only those who had five or more valid nights of actigraphy.

Before testing each primary model, we used a model comparison approach to determine the functional form of the relationship for each outcome with respect to linearity and random effects (i.e., testing whether the slope between time and sleep efficiency varied randomly, with variation based on the individual). The −2 log likelihood was used as an index to determine whether the model with the random intercept for each individual fit the data better than the model without the random intercept. We examined the significance test associated with the likelihood ratios between the models with and without the relevant parameters to assess the relative quality between the models based on the inclusion of random slopes and a linear versus quadratic effect of time. All analyses were conducted in the R statistical computing environment (R Core Team, 2023). Multilevel analyses were performed using the package lme4 [39] with model comparisons tested within the nlme [40] package, and ggplot2 [41] and ggeffects [42] were used for visualization.

Results

Descriptive statistics

There were 74 women who provided actigraphy data for at least two time points. On average, participants were 28.9 years old with a pre-pregnancy BMI of 27.1 kg/m2; a majority were white (75.7%), married (78.5%), had a family income over $75,000 (50.6%), and had at least a college degree (74.7%). Most women had a prior birth (60.7%) and were an average 30.8 weeks gestation at the start of the study (see Table 1). On average these participants had between 6.3 and 6.6 valid days of actigraphy.

thumbnail
Download:
Table 1. Demographic characteristics of participants at baseline.

https://doi.org/10.1371/journal.pone.0323489.t001

Time in bed

Actigraphic time in bed and self-reported time in bed were similar, ranging from 7.97–8.66 hours by actigraphy and 7.72–8.52 hours by self-report (see Table 2; see Fig 1A). Accordingly, actigraphic time in bed and self-reported time in bed were consistently correlated at each time point captured in this study, with the strength of the association varying from low to high: 3rd trimester (r = .58, p < .001), 4–6 weeks postpartum (r = .37, p = .002), 4 months postpartum (r = .62, p < .001), 8 months postpartum (r = .49, p < .001), and 12 months postpartum (r = .37, p = .005).

thumbnail
Download:
Table 2. Mean (SD) of actigraphy and PSQI measures from 3rd trimester to 12 months postpartum (PP).

https://doi.org/10.1371/journal.pone.0323489.t002

thumbnail
Download:
Fig 1. A-F. Mean (SE) values of time in bed (TIB), total sleep time (TST), sleep efficiency (SE), sleep latency (SL), wake after sleep onset (WASO), and PSQI score over visits.

https://doi.org/10.1371/journal.pone.0323489.g001

Overall, there was a significant decrease in time in bed across the study visits (b = −0.16, 95% CI [−0.23, −0.09], p < .001); there was not a statistically significant difference in time in bed based on the method of data collection (b = −0.10, 95% CI [−0.27, 0.07], p = .25) and no significant interaction between time and method (b = 0.00, p = .96; see Fig 2A).

thumbnail
Download:
Fig 2. A-D. Comparison of sleep assessment methods for TIB, TST, SE, and SL. This graph shows model-based predicted marginal means generated from the ggpredict() function in the ggeffects R package (42). Shaded regions represent the 95% confidence intervals for the predicted marginal means.

https://doi.org/10.1371/journal.pone.0323489.g002

Total sleep time

Actigraphic total sleep time was consistently higher than self-reported total sleep time until 12 months postpartum, ranging from 6.78–7.27 hours by actigraphy and 5.86–6.83 by self-report (see Table 2; see Fig 1B). However, actigraphic and self-reported total sleep time were reliably correlated across each time point in this study, with the strength of the association varying from low to moderate: 3rd trimester (r = .34, p = .008), 4–6 weeks postpartum (r = .46, p < .001), 4 months postpartum (r = .40, p = .001), 8 months postpartum (r = .27, p = .040), and 12 months postpartum (r = .48, p < .001).

These data showed a significant linear (b = −0.67, 95% CI [−0.94, −0.39], p < .001) and quadratic (b = 0.11, 95% CI [0.06, 0.15], p < .001) effect of time for total sleep time. There was also a significant difference based on the method of data collection (b = −0.56, 95% CI [−0.71, −0.42], p < .001) such that the actigraphic measure of sleep estimated longer sleep times than self-report measures of sleep across the visits. There was also a significant interaction between time (linear) and method of data collection (b = −0.59, 95% CI [−1.12, −0.06], p = .029), as well as time (quadratic) and method of data collection (b = 0.13, 95% CI [0.04, 0.21], p = .005); an analysis of simple slopes indicated that this difference was significant in 3rd trimester of pregnancy (b = −0.57, p < .001), 4–6 weeks postpartum (b = −1.01, p < .001), 4 months postpartum (b = −0.57, p < .001), and 8 months postpartum (b = −0.63, p < .001), but the two measures converged by 12 months postpartum (b = 0.06, p = .71; see Fig 2B). In sum, self-report and actigraphic estimations of total sleep time may differ by 30–60 minutes from one’s 3rd trimester of pregnancy to 8 months postpartum but estimates are similar by 12 months postpartum.

Sleep efficiency

Sleep efficiency followed a similar pattern in both methods; however, actigraphy-measured sleep efficiency was consistently higher than self-report, ranging from 79.88–86.17% by actigraphy and 69.44–84.18% by self-report (see Table 2; see Fig 1C). Actigraphic and self-reported sleep efficiency were only reliably correlated at 12 months postpartum (r = .35, p = .009), but not at any other study visit: 3rd trimester (r = .06, p = .66), 4–6 weeks postpartum (r = .21, p = .09), 4 months postpartum (r = .20, p = .14), nor 8 months postpartum (r = .18, p = .18).

There was a significant linear (b = −4.22, 95% CI [−7.11, −1.31], p = .004) and quadratic (b = 0.93, 95% CI [0.45, 1.40], p < .001) effect of time, and a significant difference based on method of data collection (b = −5.63, 95% CI [−7.18, −4.08], p < .001) such that the actigraphic measure of sleep consistently estimated a greater sleep efficiency than self-report measures, which is in line with the finding that actigraphy captured a shorter time to fall asleep (latency) and longer sleeping times overall compared with the self-report measures. There was not a significant interaction between time (linear or quadratic) and method of data collection (ps > .16; see Fig 2C).

Sleep latency

Sleep latency was consistently lower by actigraphy compared to self-report, ranging from 8.45–9.17 minutes by actigraphy and 14.17–22.13 minutes by self-report (see Table 2; see Fig 1D). Actigraphic and self-reported sleep latency were not reliably correlated across any study visits: 3rd trimester (r = .17, p = .19), 4−6 weeks postpartum (r = .14, p = .27), 4 months postpartum (r = −.02, p = .87), 8 months postpartum (r = .25, p = .06), nor 12 months postpartum (r = −.06, p = .63).

Overall, there was a significant linear (b = −4.42, 95% CI [−7.88, −0.96], p = .013) and quadratic effect of time (b = 0.79, 95% CI [0.23, 1.35], p = .006) related to sleep latency, and a significant difference based on method of data collection (b = 9.82, 95% CI [8.02, 11.62], p < .001) such that the actigraphic measure of sleep captured less time awake before the sleep period began than was self-reported by the participants. There was also a significant interaction between time (linear) and method (b = −8.25, 95% CI [−14.82, −1.67], p = .015), as well as time (quadratic) and method (b = 1.50, 95% CI [0.42, 2.58], p = .007; see Fig 2D); an analysis of simple slopes indicated that this difference was significant at each visit: 3rd trimester of pregnancy (b = 12.42, p < .001), 4–6 weeks postpartum (b = 5.57, p < .001), 4 months postpartum (b = 7.65, p < .001), 8 months postpartum (b = 10.53, p < .001), and 12 months postpartum (b = 13.68, p < .001).

Wake after sleep onset

Wake after sleep onset (actigraphy only), ranged from 69.56–107.42 minutes (see Table 2; see Fig 1E). There was a significant linear (b = 15.49, 95% CI [3.50, 27.45], p = .011) and quadratic (b = −3.22, 95% CI [−5.18, −1.26], p = .001) effect of time for wake after sleep onset.

PSQI global score

The PSQI Global Score (self-report only), ranged from 5.93–7.73 (see Table 2; see Fig 1F). There was a significant quadratic (b = −0.17, 95% CI [−0.31, −0.03], p = .015), but not linear (p = .10) effect of time related to the PSQI Global Score.

Discussion

Among 74 women sampled five times from their 3rd trimester of pregnancy to 12 months postpartum, we identified differences by sleep measurement method for total sleep time, sleep efficiency, and sleep latency. There were no significant differences for time in bed regardless of whether it was assessed via self-report (i.e., PSQI) or actigraphy. Broadly, actigraphy tended to reflect longer total sleep times, shorter time to fall asleep (sleep latency), and greater sleep efficiency compared to self-reported measures. Understanding how different measurement modalities may affect sleep data collected during pregnancy and postpartum is important for researchers in this area.

Overall, these findings align with other studies of sleep through pregnancy or postpartum where there are decreases in total sleep time and sleep efficiency and increases in wake after sleep onset during the postpartum period (compared to pregnancy), particularly in the first 2–3 months postpartum [43,44]. Our findings indicate that self-reported sleep duration may be underestimated compared to actigraphy in the perinatal population, however the two methods were low to moderately correlated. Others have also shown a moderate correlation between actigraphy-assessed sleep and self-reported sleep in non-pregnant samples [45,46], but these findings are not uniform across the literature and agreement between measurement modalities may also depend on psychological factors such as depressive symptoms, hostility, and perceptions of overall health [47]. Prior research comparing actigraphic and self-reported total sleep time in mid-pregnancy identified that self-reported total sleep time was overestimated compared to actigraphic total sleep time [21]; however, the authors also identified variability in the direction of this effect as 23% underestimated their sleep time by more than an hour while 39% of the sample overestimated their sleep time compared to actigraphy. Variability in self-report measures of sleep is worth considering when implementing sleep diaries; however, it is also possible that there is measurement error from actigraphy that may be contributing to these differences, and that some of this measurement error could be unique to the perinatal period.

While sleep latency and sleep efficiency were consistently estimated differently based on the mode of measurement across the study period, estimates of total sleep time converged by 12 months postpartum. This may reflect less disrupted sleep by 12 months postpartum as infants sleep more throughout the night or may reflect a more general recovery of sleep as one may have developed better strategies to fall back asleep by this point. The differences in sleep latency have similarly been found in other studies, across various populations, and among pregnant women [48–50]. Individuals often overestimate the amount of time it takes them to fall asleep and actigraphy depends on movements for its measurements and will inherently underestimate due to minimal movements when attempting to fall asleep. Assessing and treating sleep difficulties in pregnancy and postpartum is a priority for health of both the mother and developing fetus/infant. Indeed, our results highlight the clinical challenge of using actigraphy in samples that experience substantial time awake during the night.

There are several limitations to our interpretation of these data. First, we did not use polysomnography, which is the gold standard for sleep measurement but not practical for measuring sleep in home; instead, we focused on actigraphy as a well-validated, less burdensome, and more commonly used method of sleep assessment in population studies. The self-reported measure of sleep (PSQI) asks participants to reflect on their sleep over the past four weeks whereas actigraphy was only measured for a week. We do not know how much of differences in the two methods may be from error related to retrospective reporting or error that is inherent when self-reporting sleep. We also were not able to capture sleep disorders that may have been present for some of the women in this sample (e.g., restless leg syndrome, sleep disordered breathing). Lastly, while our sample was larger than some previous studies, it remains a relatively small number of participants and may not be generalizable to other pregnant and postpartum women. For example, in our sample, most women had a prior birth and the variability in the age of that child in their home may affect the sleep experiences reported here.

These limitations are well-balanced by several critical strengths. This study examined the congruence between self-report and actigraphy-assessed sleep in repeated measurements capturing sleep-related changes in the transition to parenthood, which is a sensitive period for women’s health. The only previous examination of perinatal concordance in sleep measurement is limited to pregnancy and did not include the PSQI [30]; thus, our results nicely complement and extend the available literature.

Given the elevated mental health risks of the postpartum period and the well-established links between sleep and mental health risk, many researchers are measuring sleep and mental health symptomology in the postpartum period; our results can inform the ways that choosing to measure sleep via self-report using the PSQI or actigraphy alone may influence various metrics of sleep health. In summary, this study identified that actigraphy-assessed perinatal sleep may indicate longer total sleep times, shorter sleep latency, and greater sleep efficiency compared to self-reported sleep. Sleep latency and efficiency were consistently different across the mode of measurement throughout the entire study period whereas total sleep time was only different by mode of measurement in the first eight months following the transition to parenthood. Using longitudinal assessment of sleep health and knowing the implications of different methods for measurement is critical to advance research on maternal and infant health.

Supporting information

S1 Data. Full data for reproducing reported findings.

https://doi.org/10.1371/journal.pone.0323489.s001

(CSV)

Acknowledgments

We appreciate the contributions of our Clinical Research Assistants and students to data collection. We also thank our study participants and staff at the Ohio State University Wexner Medical Center Prenatal Clinic.

References

  1. 1. Manconi M, van der Gaag LC, Mangili F, Garbazza C, Riccardi S, Cajochen C, et al. Sleep and sleep disorders during pregnancy and postpartum: The Life-ON study. Sleep Med. 2024;113:41–8. pmid:37984016
  2. 2. Lu Q, Zhang X, Wang Y, Li J, Xu Y, Song X, et al. Sleep disturbances during pregnancy and adverse maternal and fetal outcomes: A systematic review and meta-analysis. Sleep Med Rev. 2021;58:101436. pmid:33571887
  3. 3. Sedov ID, Cameron EE, Madigan S, Tomfohr-Madsen LM. Sleep quality during pregnancy: A meta-analysis. Sleep Med Rev. 2018;38:168–76. pmid:28866020
  4. 4. Sobol M, Błachnio A, Meisner M, Szyszkowska J, Jankowski KS. Sleep, circadian activity patterns and postpartum depression: A systematic review and meta-analysis of actigraphy studies. J Sleep Res. 2024;33(4):e14116. pmid:38095248
  5. 5. Carroll JE, Ross KM, Horvath S, Okun M, Hobel C, Rentscher KE, et al. Postpartum sleep loss and accelerated epigenetic aging. Sleep Health. 2021;7(3):362–7. pmid:33903077
  6. 6. Okun ML, Schetter CD, Glynn LM. Poor sleep quality is associated with preterm birth. Sleep. 2011;34(11):1493–8. pmid:22043120
  7. 7. Christian LM, Carroll JE, Teti DM, Hall MH. Maternal Sleep in Pregnancy and Postpartum Part I: Mental, Physical, and Interpersonal Consequences. Curr Psychiatry Rep. 2019;21(3):20. pmid:30826881
  8. 8. Blair LM, Porter K, Leblebicioglu B, Christian LM. Poor Sleep Quality and Associated Inflammation Predict Preterm Birth: Heightened Risk among African Americans. Sleep. 2015;38(8):1259–67. pmid:25845693
  9. 9. Okun ML, Luther JF, Wisniewski SR, Wisner KL. Disturbed sleep and inflammatory cytokines in depressed and nondepressed pregnant women: an exploratory analysis of pregnancy outcomes. Psychosom Med. 2013;75(7):670–81. pmid:23864582
  10. 10. Felder JN, Baer RJ, Rand L, Jelliffe-Pawlowski LL, Prather AA. Sleep Disorder Diagnosis During Pregnancy and Risk of Preterm Birth. Obstet Gynecol. 2017;130(3):573–81. pmid:28796676
  11. 11. Felder JN, Baer RJ, Rand L, Ryckman KK, Jelliffe-Pawlowski L, Prather AA. Adverse infant outcomes among women with sleep apnea or insomnia during pregnancy: A retrospective cohort study. Sleep Health. 2023;9(1):26–32. pmid:36371381
  12. 12. Hedman C, Pohjasvaara T, Tolonen U, Suhonen-Malm AS, Myllylä VV. Effects of pregnancy on mothers’ sleep. Sleep Med. 2002;3(1):37–42. pmid:14592252
  13. 13. Sowers MF, Zheng H, Kravitz HM, Matthews K, Bromberger JT, Gold EB. Sex steroid hormone profiles are related to sleep measures from polysomnography and the Pittsburgh Sleep Quality Index. Sleep. 2008;31(10):1339–49.
  14. 14. Hunter LP, Rychnovsky JD, Yount SM. A selective review of maternal sleep characteristics in the postpartum period. J Obstet Gynecol Neonatal Nurs. 2009;38(1):60–8. pmid:19208049
  15. 15. Quin N, Lee JJ, Pinnington DM, Newman L, Manber R, Bei B. Differentiating perinatal Insomnia Disorder and sleep disruption: a longitudinal study from pregnancy to 2 years postpartum. Sleep. 2022;45(2):zsab293. pmid:34989808
  16. 16. Verma S, Pinnington DM, Manber R, Bei B. Sleep-wake timing and chronotype in perinatal periods: longitudinal changes and associations with insomnia symptoms, sleep-related impairment, and mood from pregnancy to 2 years postpartum. J Sleep Res. 2024;33(3):e14021.
  17. 17. Facco FL, Kramer J, Ho KH, Zee PC, Grobman WA. Sleep disturbances in pregnancy. Obstet Gynecol. 2010;115(1):77–83. pmid:20027038
  18. 18. Wang G, Deng Y, Jiang Y, Lin Q, Dong S, Song Y, et al. Trajectories of sleep quality from late pregnancy to 36 months postpartum and association with maternal mood disturbances: a longitudinal and prospective cohort study. Sleep. 2018;41(12):10.1093/sleep/zsy179. pmid:30535183
  19. 19. Horwitz A, Bar-Shachar Y, Ran-Peled D, Finkelstein O, Ben-Zion H, Bar-Kalifa E. Sleep of mothers, fathers, and infants: a longitudinal study from pregnancy through 12 months. Sleep. 2023;46(9).
  20. 20. Slyepchenko A, Minuzzi L, Reilly JP, Frey BN. Longitudinal Changes in Sleep, Biological Rhythms, and Light Exposure From Late Pregnancy to Postpartum and Their Impact on Peripartum Mood and Anxiety. J Clin Psychiatry. 2022;83(2):21m13991. pmid:35044728
  21. 21. Herring SJ, Foster GD, Pien GW, Massa K, Nelson DB, Gehrman PR, et al. Do pregnant women accurately report sleep time? A comparison between self-reported and objective measures of sleep duration in pregnancy among a sample of urban mothers. Sleep Breath. 2013;17(4):1323–7. pmid:23563909
  22. 22. Tsai S-Y, Lee P-L, Gordon C, Cayanan E, Lee C-N. Objective sleep efficiency but not subjective sleep quality is associated with longitudinal risk of depression in pregnant women: A prospective observational cohort study. Int J Nurs Stud. 2021;120:103966. pmid:34051587
  23. 23. Parsons L, Howes A, Jones CA, Surtees ADR. Changes in parental sleep from pregnancy to postpartum: A meta-analytic review of actigraphy studies. Sleep Med Rev. 2023;68:101719. pmid:36791531
  24. 24. Chinoy ED, Cuellar JA, Huwa KE, Jameson JT, Watson CH, Bessman SC, et al. Performance of seven consumer sleep-tracking devices compared with polysomnography. Sleep. 2021;44(5):zsaa291. pmid:33378539
  25. 25. Gay CL, Lee KA, Lee S-Y. Sleep patterns and fatigue in new mothers and fathers. Biol Res Nurs. 2004;5(4):311–8. pmid:15068660
  26. 26. Gordon LK, Mason KA, Mepham E, Sharkey KM. A mixed methods study of perinatal sleep and breastfeeding outcomes in women at risk for postpartum depression. Sleep Health. 2021;7(3):353–61. pmid:33640360
  27. 27. Park EM, Meltzer-Brody S, Stickgold R. Poor sleep maintenance and subjective sleep quality are associated with postpartum maternal depression symptom severity. Arch Womens Ment Health. 2013;16(6):539–47.
  28. 28. Sharkey KM, Boni GM, Quattrucci JA, Blatch S, Carr SN. Women with Postpartum Weight Retention Have Delayed Wake Times and Decreased Sleep Efficiency During the Perinatal Period: A Brief Report. Sleep Health. 2016;2(3):225–8. pmid:27857964
  29. 29. Sharkey KM, Iko IN, Machan JT, Thompson-Westra J, Pearlstein TB. Infant sleep and feeding patterns are associated with maternal sleep, stress, and depressed mood in women with a history of major depressive disorder (MDD). Arch Womens Ment Health. 2016;19(2):209–18. pmid:26228760
  30. 30. Sharkey KM, Pearlstein TB, Carskadon MA. Circadian phase shifts and mood across the perinatal period in women with a history of major depressive disorder: a preliminary communication. J Affect Disord. 2013;150(3):1103–8.
  31. 31. Okun ML, Kohl V, Feliciano L. Comparison of longitudinal diary and actigraphy-assessed sleep in pregnant women. Sleep Med. 2021;88:149–56. pmid:34753041
  32. 32. Volkovich E, Tikotzky L, Manber R. Objective and subjective sleep during pregnancy: links with depressive and anxiety symptoms. Arch Womens Ment Health. 2016;19(1):173–81. pmid:26250541
  33. 33. Tsai S-Y, Lee P-L, Lin J-W, Lee C-N. Cross-sectional and longitudinal associations between sleep and health-related quality of life in pregnant women: A prospective observational study. Int J Nurs Stud. 2016;56:45–53. pmid:26803171
  34. 34. Marino M, Li Y, Rueschman MN, Winkelman JW, Ellenbogen JM, Solet JM, et al. Measuring sleep: accuracy, sensitivity, and specificity of wrist actigraphy compared to polysomnography. Sleep. 2013;36(11):1747–55. pmid:24179309
  35. 35. Buysse DJ, Reynolds CF 3rd, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res. 1989;28(2):193–213. pmid:2748771
  36. 36. Buxton OM, Lee S, Beverly C, Berkman LF, Moen P, Kelly EL. Work-Family Conflict and Employee Sleep: Evidence from IT Workers in the Work, Family and Health Study. Sleep. 2016;39(10):1871–82.
  37. 37. Buxton OM, Lee S, Marino M, Beverly C, Almeida DM, Berkman L. Sleep health and predicted cardiometabolic risk scores in employed adults from two industries. J Clin Sleep Med. 2018;14(3):371–83.
  38. 38. Olson R, Crain TL, Bodner TE, King R, Hammer LB, Klein LC, et al. A workplace intervention improves sleep: results from the randomized controlled Work, Family, and Health Study. Sleep Health. 2015;1(1):55–65. pmid:29073416
  39. 39. Bates D, Mächler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. Journal of Statistical Software. 2015;67(1):1–48.
  40. 40. Pinheiro J, Bates D, DebRoy S, Sakar D, Team RC. nlme: Linear and nonlinear mixed effects models. 2020.
    • 41. Wickham H. ggplot2: elegant graphics for data analysis. Second ed. Springer; 2016.
      • 42. Lüdecke D. ggeffects: Tidy Data Frames of Marginal Effects from Regression Models. JOSS. 2018;3(26):772.
      • 43. Kang MJ, Matsumoto K, Shinkoda H, Mishima M, Seo YJ. Longitudinal study for sleep-wake behaviours of mothers from pre-partum to post-partum using actigraph and sleep logs. Psychiatry Clin Neurosci. 2002;56(3):251–2. pmid:12047581
      • 44. Bei B, coo calcagni S, Milgrom J, Trinder J. Day-to-day alteration of 24-hour sleep pattern immediately before and after giving birth. Sleep and Biological Rhythms. 2012;10(3):212–21.
      • 45. Lauderdale DS, Knutson KL, Yan LL, Liu K, Rathouz PJ. Self-reported and measured sleep duration: how similar are they?. Epidemiology. 2008;19(6):838–45. pmid:18854708
      • 46. Cespedes EM, Hu FB, Redline S, Rosner B, Alcantara C, Cai J. Comparison of Self-Reported Sleep Duration With Actigraphy: Results From the Hispanic Community Health Study/Study of Latinos Sueno Ancillary Study. American Journal of Epidemiology. 2016;183(6):561–73.
      • 47. Matthews KA, Patel SR, Pantesco EJ, Buysse DJ, Kamarck TW, Lee L, et al. Similarities and differences in estimates of sleep duration by polysomnography, actigraphy, diary, and self-reported habitual sleep in a community sample. Sleep Health. 2018;4(1):96–103. pmid:29332687
      • 48. Smith MT, McCrae CS, Cheung J, Martin JL, Harrod CG, Heald JL, et al. Use of Actigraphy for the Evaluation of Sleep Disorders and Circadian Rhythm Sleep-Wake Disorders: An American Academy of Sleep Medicine Systematic Review, Meta-Analysis, and GRADE Assessment. J Clin Sleep Med. 2018;14(7):1209–30. pmid:29991438
      • 49. Cederberg KLJ, Mathison BG, Schuetz ML, Motl RW. Discrepancies between self-reported and device-measured sleep parameters in adults with multiple sclerosis. J Clin Sleep Med. 2022;18(2):415–21. pmid:34338630
      • 50. Wilson DL, Fung A, Walker SP, Barnes M. Subjective reports versus objective measurement of sleep latency and sleep duration in pregnancy. Behav Sleep Med. 2013;11(3):207–21. pmid:23205562