Maternal Depressive Symptoms and Toddler Sleep: Moderating Effects of Sleep Parenting Practices Authors: Hsiao-Yen Wu, Pey-Ling Shieh, Yen-Chun Liu Published: 2026 ABSTRACT AbstractBackground and AimsToddler sleep is more consolidated than infant sleep and may better reflect long-term sleep regulation. Despite growing interest in maternal mental health, evidence linking maternal depression to toddler sleep remains limited. Moreover, sleep parenting practices may shape this relationship. This study aimed to examine the status of maternal depression, toddler sleep, and sleep parenting practices, as well as the moderating role of these practices in the association between maternal depression and toddler sleep.MethodsThis research employed a cross-sectional design conducted between November 2022 and March 2023 among mothers of toddlers in Taiwan. A total of 126 mothers of toddlers aged 1 to 3 years completed an online survey. The mean age of mothers was 33.01 years (SD = 4.33), and the mean age of toddlers was 1.79 years (SD = 0.58). Maternal depression was assessed using the Beck Depression Inventory-II (BDI-II, Beck, 1996). Toddler sleep was measured using the Brief Infant Sleep Questionnaire-Reviesd (Mindell et al, 2019), including four indicators: maternal perceived sleep problem, sleep latency, number of nocturnal awakenings, and nighttime sleep duration. Two sleep parenting practices were assessed: co-sleeping (co-sleep vs not co-sleep) and consistency of sleep routines. Data were analyzed using descriptive statistics, group comparisons, correlation analyses, and multiple regression analyses.ResultsApproximately 33.3% of mothers exhibited depressive tendencies (BDI-II > 13). Most toddlers had nighttime sleep duration shorter than 12 hours (91.3%). Around half of the toddlers had sleep latency longer than 45 minutes (50.8%), 34.9% experienced at least two nocturnal awakenings, and 11.9% were reported to have sleep problem. Co-sleep was reported by 51.6% of mothers, and 45.2% had low consistency in their sleep routines (≤5 nights/week). After controlling for child health status, maternal income, and education level, higher maternal depressive symptoms were associated with greater toddler sleep problem severity and longer sleep latency (Bs = 0.02 and 0.87, ps < 0.05 and 0.01, respectively). Co-sleep moderated the association between maternal depression and toddler sleep problem (B = −0.05, p < 0.05). Simple slope analyses indicated that higher maternal depression was associated with more severe toddler sleep problem in the co-sleep group (B = 0.05, p < 0.01), but not in the not co-sleep group (B = 0.00, p > 0.05). No moderating effect was found for sleep routine consistency.ConclusionMaternal depression is associated with more severe sleep problem and longer sleep latency in toddlers. Co-sleep may amplify the spillover associations of maternal depression on toddler sleep. For mothers with elevated depressive symptoms, encouraging separate sleeping arrangements may help buffer the association between maternal depression and maternal perceived toddler sleep problem. Further studies to validate these findings are warranted.Keywords: bedtime routine consistency, co-sleep, maternal depression, moderation, sleep parenting practices, toddler sleep Introduction Child sleep quality greatly affects family well-being, with links to parental sleep, sense of control, and risk of maltreatment.1 Sleep disturbances in children are associated with adverse physical, neurocognitive, and behavioral outcomes.2 Since toddlers ideally spend about half their day sleeping3 and sleep patterns consolidate from the toddler stage, thus early sleep problem may persist,4 understanding toddler sleep and related factors is essential. Research indicates maternal depression predicts infant sleep problem.5,6 However, both maternal symptoms and child conditions change over time. For example, researchers reported distinct depressive symptom trajectories: 21.7% stable-moderate, 47.5% low-decreasing, and 8.3% increasing depression from pregnancy to two years postpartum.7 Toddler sleep characteristics also differ from those in infancy due to the development of separation anxiety, cognition, and mobility.4 Thus, whether maternal depression still influences toddler sleep remains unclear. There are only limited studies that explore the associations between maternal depression and toddler sleep, and the results are still inconsistent. In this study, the authors reviewed seven relevant studies and summarized their methodologies and key findings in Table 1.8–14 As shown in Table 1, the participants were diverse populations including general families (China and Norway), low-income families (US), indigenous families (New Zealand), and multi-ethnic Asian families (Singapore). Sample sizes ranged from 174 to 14,926 mother-child dyads, with assessments spanning 42 days to 36 months postpartum. Also, as presented in Table 1, five instruments were used to assess maternal depression, and four ways were used to assess children’s sleep. The target sleep indicators include maternal perceived sleep problem, sleep latency, nocturnal awakenings, and sleep duration. Substantial heterogeneity exists in the definition of sleep indicators. Even in a single study, researchers may use different methods to define a single sleep indicator. For example, “sleep problem” was defined by using the scale cut-off,9 the parental perception,9,10,13 or a composite index.13 “Sleep latency” was either assessed separately13 or combined with sleep maintenance.8 Similarly, nocturnal awakenings were either analyzed as categorical14 or as continuous variables.11 Regarding nighttime sleep duration, researchers set different ranges for “night” (7 p.m. to 7 or 8 a.m.). The research findings can be summarized as: maternal depression is associated with nocturnal awakenings,11,14 while it is unrelated to sleep duration.9,11–13 However, the findings regarding sleep problem and sleep latency are divergent. With sleep problem, a cross-sectional study indicated that maternal depression correlated with children’s sleep problem at age three in New Zealand indigenous children,9 while another study showed no correlation at age 12–32 months in US low-income families.10 Also, two longitudinal studies showed no predicted relationships between 3–4 months postpartum depression and sleep problem at ages 2–3.9,13 For latency, one study found that children of highly depressed mothers had greater difficulty in entering and maintaining sleep at age 15–19 months,8 while another study found no correlation between three months postpartum depression and 30-month-old children’s sleep latency.13 In summary, methodological heterogeneity—the study populations, study designs (cross-sectional vs longitudinal), timing of assessments (from 42 days to 3 years postpartum), target sleep indicators, and the assessment/analysis strategies—may contribute to controversial findings, which preclude making conclusions. Table 1The Relationship of Maternal Postpartum Depression and Toddler Sleep: A SummarySource and ParticipantMethodFindingTime PointInstrumentSleep IndicatorCorrelatedNot CorrelatedDepressionSleepYstrom et al (2017)14Norway, Secondary data, N = 14,92618M1. SCL-8a1. Self-designed question1-1. Nocturnal awakening 1: Four categories18M depression can predict 18M awakening 1Covington et al (2018)10American, Low income, N = 28012–32M2. BDI-IIb2. BISQf2-1. Sleep problem 1: Binary as “small/very serious” or not12–32M depression is not associated with 12–32M sleep problem 1Bates et al (2021)8American, Low income, N = 17415–19M3. CES-Dc3. TCSQg3-1. Sleep latency 1: TCSQ total score (sleep latency + maintaining)15–19M depression is related to 15–19M latency 1Gui et al (2022)11China, N = 24342D, 3/6/9/12/18/24/36M4. EPDSd2. BISQf1-2. Nocturnal awakening 2: Numbers4–1. Sleep duration 1: 7 p.m. to 8 a.m.42D depression are associated with 1–3Y awakening 242D depression are not associated with 42D-36M sleep duration 1Ma et al (2022)13China, Secondary data, N = 15833/30M4. EPDSd (3M)2. BISQf2-2. Sleep problem 2: “number of awakening > 3”, or “length of awakening” > 1H, or “sleep duration” < 9 H2–3. Sleep problem 3: Parents’ report as “very serious”, “small”, or “not a problem”3–2. Sleep latency4–2. Sleep duration 2: 7 p.m. to 7 a.m.3M depression is not associated with 30 M sleep duration 2, latency, and sleep problem 2 and 3Carter et al (2025)9New Zealand, Indigenous, Secondary data, N = 8563M, 3Y4. EPDSd (3M)5. K-10e (3Y)1. Self-designed questions4. CSHQh2-4.Sleep problem 4: Total CSHQ score > 412–5. Sleep problem 5: Mother’s perceived problem as moderate-large4–3. Sleep duration 3: Total3Y depression is associated with 3Y sleep problem 4 and 51. 3M and 3Y depression are not associated with 3Y sleep duration 32. 3M depression are not associated with 3Y sleep problem 4 and 5Kotlar et al (2025)12Singapore, Multi-ethnic Asian, Secondary data, N = 11313/12M, 2Y2. BDI-IIb2. BISQf4-2.Sleep duration 2: 7 p.m. to 7 a.m.12M depression cannot predict 2Y sleep duration 3Abbreviations: aSCL-8, Hopkins Symptom Checklist; bBDI-II, Beck Depression Inventory-II; cCES-D, Center for Epidemiologic Studies-Depression Scale; dEPDS, Edinburgh Postnatal Depression Scale; eK-10, Kessler 10 Item Scale; fBISQ, Brief Infant Sleep Questionnaire; gTCSQ, Tayside Children’s Sleep Questionnaire; hCSHQ, Children’s Sleep Habits Questionnaire. The Relationship of Maternal Postpartum Depression and Toddler Sleep: A Summary Abbreviations: aSCL-8, Hopkins Symptom Checklist; bBDI-II, Beck Depression Inventory-II; cCES-D, Center for Epidemiologic Studies-Depression Scale; dEPDS, Edinburgh Postnatal Depression Scale; eK-10, Kessler 10 Item Scale; fBISQ, Brief Infant Sleep Questionnaire; gTCSQ, Tayside Children’s Sleep Questionnaire; hCSHQ, Children’s Sleep Habits Questionnaire. Beyond methodological discrepancies, the inconsistent findings in the literature may be attributed to unmeasured moderating variables that may buffer or exacerbate the link between maternal distress and children’s sleep. Given that sleep parenting practices—specifically sleep place and bedtime routines—are pivotal determinants of early childhood sleep hygiene,15,16 it’s essential to explore whether sleep parenting practices moderate this association. “Sleep place” for children at night can be categorized as co-sleeping (on same surfaces as parents), sleeping on separate surfaces, or in separate rooms. Several studies indicated that toddlers who co-sleep experience more sleep problem, longer sleep latency, more frequent nocturnal awakenings, and shorter sleep duration.4,17 It is possible that bedsharing infants experience increased nighttime mother-infant interactions;18 thus, co-sleep may heighten children’s vulnerability to depressive mothers’ nocturnal restlessness and emotional dysregulation. In another way, the increased mother–child proximity may sensitize mothers’ perception toward children’s sleep problem. Thus, hyper-responsiveness of depressive mothers may lead them to perceive more sleep problem with their children19 under a bedsharing situation. Despite growing evidence linking maternal depression and children’s sleep, limited research has examined how co-sleep may buffer or exacerbate this association, particularly in toddler populations. The role of sleep place in the impact of maternal depression has been overlooked. “Bedtime routines” refers to consistent activities conducted within one hour before lights-out, such as feeding, bathing, reading, singing lullabies, or massage (while feeding or rocking the child to sleep is not included).20,21 The role of routine consistency in promoting sleep continuity is well-recognized.22 Consistent bedtime routines provide a predictable parent-child interaction environment.21 In the context of maternal depression, it’s plausible that consistent bedtime routines may act as a compensatory mechanism to offer the child a sense of security and mitigate the reduced emotional and behavioral consistency often observed in depressed mothers, then lead to better sleep conditions. However, the exploration of the moderating role of bedtime routine consistency is also limited. In Taiwan, about 60% of parents with children under age six suffer from their children’s sleep problem. Specifically, parents reported that over 70% of children exhibit a sleep latency of more than 30 minutes, and nearly 40% wake frequently during the night.23 Compared with Western parents, Taiwanese parents adopt more co-sleeping arrangements and less consistent routines.24 Current research on the association among maternal depression, sleep parenting, and toddler sleep is insufficient, and there is a lack of studies focusing on Taiwanese mothers and infants. Given the suffering Taiwanese parents have about their children’s sleep, and the unique characteristics of sleep parenting within Taiwanese families, this study aims to examine the association between maternal depression and toddler sleep in Taiwan, and the moderating role of sleep parenting in this association. Three hypotheses were addressed: First, maternal depression is associated with toddler nighttime sleep; second, sleep place moderates the association between maternal depression and toddler sleep; third, bedtime routine consistency moderates the association between maternal depression and toddler sleep. Materials and Methods Study Design and Setting and Period This research employed a cross-sectional design. The indicator of maternal depression is depressive symptoms. Indicators of toddler nocturnal sleep are maternal-perceived sleep problem, sleep latency, awakenings, and sleep duration. Sleep parenting practices are indexed by sleep place (co-sleep, sleep on separate surfaces, or sleep in separate rooms) and bedtime routine consistency. We disseminated a research brochure on popular social media among toddler families in Taiwan (for example, mother and baby forums, preschool forums on Facebook and Instagram), pediatric clinics (of a university hospital in Taichung City and a private clinic in New Taipei City), and childcare centers (in Pingtung County). Women can scan the QR code in the brochure to fill out the online questionnaires. Data collection was conducted from November 2022 to March 2023. Data of eligible respondents were analyzed. Sample Size Determination An a priori power analysis was conducted using G*Power version 3.1 for multiple linear regression (fixed Model, R2 deviation from zero). Assuming a medium effect size (f2 = 0.15), a significance level of α = 0.05, statistical power of 0.80, and six predictors, the minimum required sample size was estimated to be 98 participants. A total of 161 respondents were recruited, and 126 eligible participants were obtained, exceeding the minimum sample size required for the planned analyses. Participants Inclusion criteria were: (a) biological mothers of a singleton toddler aged 1–3 years old, (b) aged 20 years or older, (c) proficient in Chinese, and (d) Taiwanese nationals by birth. Of 161 responses, 126 eligible samples remained, and 25 whose children did not meet the age criteria were excluded. Data Collection and Measures Beck Depression Inventory, Second Edition (BDI-II) Maternal depression was assessed by the BDI-II.25 The 21-item inventory uses a 4-point Likert scale (0–3) to indicate the severity of each depressive symptom. The total score (ranging from 0 to 63) reflects the severity of general depressive symptoms, with the higher total score indicating greater depressive severity. The Chinese version has demonstrated robust reliability (split-half r = 0.91; Cronbach’s α = 0.91), with a score above 13 indicating a depressive tendency.26 Criterion-related validity was demonstrated by a significant Spearman correlation with the Chinese Health Questionnaire (r = 0.69).27 In this study, the internal consistency (Cronbach’s α) was 0.90. We applied the total score (continuous variable) to describe the depressive symptom severity and to perform regression analysis. We also used total scores above/not above 13 (categorical variable) to indicate participants with/not with depressive tendency. Brief Infant Sleep Questionnaire-Revised (BISQ-R) Data on toddler sleep and sleep parenting practices were collected using the BISQ-R.28 Participants completed the BISQ-R based on their status over the past week. The literature indicates that data from children aged 5 to 29 months on the BISQ significantly correlated with actigraphy and a sleep diary.29 Furthermore, the test-retest reliability for items of sleep latency, nocturnal awakenings, and sleep duration ranged from 0.82 to 0.94. We obtained a written permission from the author (Dr. Jodi Mindell) to translate the original version of BISQ-R. The translation process included: forward translation by the second author, inviting graduate students to read, the second author discussing and modifying the initial version with the third authors, and finally inviting several infant mothers to complete the modified version. The third author is a Chinese-English bilingual doctoral psychologist and professor. According to feedback from infant mothers, they can easily understand and reply to the modified version of the Chinese BISQ-R. In the present study, the modified version was used. The four sleep indicators used were: mothers’ perceived sleep problem (item 32), sleep latency (item 15), number of nocturnal awakenings (item 19), and nighttime sleep duration (item 26). Except for the sleep problem, which was rated on a 5-point Likert scale (0 to 4 as “not a problem at all”, “a very small problem”, “a small problem”, “a moderate problem”, and “a serious problem”, respectively), the other three indicators were reported as their direct numerical values. We applied the scale score of sleep problem and the numerical values of the other three sleep indicators to perform descriptive analyses, group comparisons, and regression analyses. Based on where the toddler spent most of the night (room for item 16, and surface for item 17), the sleep place was divided into three categories (co-sleep: the same room and same surface; sleep on separate surfaces: the same room and different surface; and sleep in separate rooms). We applied the categorical variables to perform frequency and percentage calculations and to divide groups for regression analyses. Based on the empirical sleep characteristics distribution and typical weekly working schedule in Taiwan, we used “low consistency” (≤ 5 nights/week) or “high consistency” (>5 nights/week) to dichotomize bedtime routine consistency (item 3). On sleep characteristics distribution, we tried different ways of dichotomization. One of the ways was adopting the criterion of “> 5 nights/week”. We found 54.8% of toddlers had the same bedtime routine >5 nights/week and 45.2% had the same bedtime routine ≤ 5nights/week. For the criterion of “> 4 nights/week”, we found 85.7% of toddlers had the same bedtime routine >4 nights/week and 14.3% had the same bedtime routine ≤4 nights/week. Since the former criterion yielded a relatively balanced proportion of participants, we use it as the grouping criterion. On “typical weekly work/non-work schedule”, given that caregivers’ working schedules may affect the establishment and maintenance of family routines, and that a five-day workweek is common among parents of toddlers in Taiwan, we used a cutoff of five nights per week to classify bedtime routine consistency. Background Information Form A self-designed form was used to collect factual demographic and family characteristics. Information collected about mothers was age, income, education level, marital status, proficiency in Chinese, and nationality. Information collected about toddlers was age, gender, health condition, birth order, birth weight, and birth plurality. Statistical Analysis Statistical analyses were performed using IBM SPSS Statistics (Version 29.0). Descriptive statistics were applied to summarize sample characteristics and main variables. Independent samples t-tests and Pearson correlations were initially conducted to examine the associations between demographic factors, maternal depression, and sleep indicators. Finally, a series of hierarchical linear regression analyses were conducted to examine the study’s hypotheses. To mitigate potential multicollinearity, the BDI-II score was mean-centered prior to calculating interaction terms. The regression models evaluated maternal depression as a relevant factor of sleep indicators and investigated the moderating effects of sleep place and bedtime routine consistency. Ethical Considerations The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board (IRB) at Chung Shan Medical University Hospital (CSMUH). The study approval number is CSMUH CS2-22136. Informed consent was obtained from all participants. Participant protection policies included providing a detailed study description, obtaining informed consent prior to participation, participant anonymization, restricted data access, and the right to withdraw at any stage before data analysis. Upon completion, participants could voluntarily provide contact information to enter a lucky draw for NT$ 100 gift vouchers. Results Participants’ characteristics are presented first (as shown in Table 2), followed by descriptive statistics of main variables—maternal depression, toddler sleep, and sleep parenting practices (as presented in Table 3). Comparison of toddler sleep based on participant characteristics was then presented (as shown in Table 2). Finally, hierarchical multiple regression analyses and moderation analyses were shown (as Tables 4–6). Table 2Participant Characteristics and Toddler Sleep ComparisonsParticipant Variable/Sleep IndicatorGroupSleep ProblemaSleep Latency (Minute)Awakening (Number/Night)Sleep Duration (Minute)LabelN (%)M ± SDtM ± SDtM ± SDtM ± SDtMotherAge (Y)>3357 (45.2%)2.26 ± 1.081.0549.95 ± 29.75−0.151.25 ± 1.26−0.22583.77 ± 88.410.92≤3369 (54.8%)2.07 ± 0.9350.74 ± 29.451.29 ± 1.03567.78 ± 104.42Income (thousands/M)>3564 (50.8%)2.02 ± 0.90−1.6449.19 ± 29.00−0.461.11 ± 0.99−1.62555.00 ± 90.97−2.39**≤3562 (49.2%)2.31 ± 1.0851.61 ± 20.131.44 ± 1.25595.68 ± 100.31Education> Undergraduate33 (26.2%)1.82 ± 0.68−2.84**46.54 ± 30.95−0.891.06 ± 0.86−1.24589.55 ± 117.011.00≤ Undergraduate93 (73.8%)2.28 ± 1.0751.77 ± 28.971.34 ± 1.21569.86 ± 89.66In marriageYes123 (97.6%)2.17 ± 1.000.8649.98 ± 29.270.861.26 ± 1.14−0.61576.76 ± 40.421.29No3 (2.4%)1.67 ± 1.1666.67 ± 40.421.67 ± 1.16503.33 ± 126.63ToddlerAge (Y)≥ 249 (38.9%)2.16 ± 1.120.0451.02 ± 28.960.191.35 ± 1.350.61559.22 ± 90.43−1.46< 277 (61.1%)2.16 ± 0.9249.97 ± 29.981.22 ± 0.98585.06 ± 100.96GenderBoy65 (51.6%)2.09 ± 1.00−0.7748.00 ± 31.05−0.941.15 ± 0.92−1.17569.26 ± 78.67−0.68Girl61 (48.4%)2.23 ± 1.0152.92 ± 27.721.39 ± 1.32581.15 ± 114.50HealthGood95 (75.4%)2.03 ± 0.93−2.32*49.51 ± 29.66−0.581.17 ± 0.99−1.45590.82 ± 85.913.31**Fair/Poor31 (24.6%)2.55 ± 1.1253.06 ± 29.201.58 ± 1.48526.58 ± 115.04Birth orderFirst-born95 (75.4%)2.32 ± 0.0871.0559.84 ± 35.322.09*1.65 ± 1.172.15*547.39 ± 116.95−1.84Others31 (24.6%)2.11 ± 1.0447.29 ± 26.791.15 ± 1.10584.03 ± 89.05Birth weight (gram)≥2500109 (86.5%)2.15 ± 1.00−0.3447.96 ± 28.09−2.38*1.21 ± 1.04−1.08579.10 ± 101.681.19<250017 (13.5%)2.24 ± 1.0365.88 ± 34.151.65 ± 1.62548.82 ± 59.78Notes: *p < 0.05, **p < 0.01. aOn the BISQ-R 5-point Likert scale (0–4), sleep problem was rated 0 to 4 as “not a problem at all”, “a very small problem”, “a small problem”, “a moderate problem”, and “a serious problem”, respectively. Participant Characteristics and Toddler Sleep Comparisons Notes: *p < 0.05, **p < 0.01. aOn the BISQ-R 5-point Likert scale (0–4), sleep problem was rated 0 to 4 as “not a problem at all”, “a very small problem”, “a small problem”, “a moderate problem”, and “a serious problem”, respectively. Table 3Description of Maternal Depression, Toddler Sleep, and Sleep Parenting PracticesVariablesTotalGroupM ± SDLabelN (%)M ± SDMaternal depression12.38 ± 9.11Depressivea42 (33.3%)23.17 ± 6.45Non-depressive84 (66.7%)6.99 ± 4.00Toddler sleepSleep problemb1.16 ± 1.00No111 (88.1%)1.89 ± 0.72Yesc15 (11.9%)4.13 ± 0.35Sleep latency (minute)50.38 ± 29.47> 4564 (50.8%)74.50 ± 21.18≤ 4562 (49.2%)25.48 ± 8.72Nocturnal awakening (number/night)1.27 ± 1.13≥ 244 (34.9%)2.52 ± 0.90< 282 (65.1%)0.60 ± 0.49Sleep duration (minute)575.02 ± 97.46< 720115 (91.3%)559.06 ± 83.74≥ 72011 (8.7%)741.82 ± 72.36Sleep parenting practicesSleep placeCo-sleep65 (51.6%)Separate-surface46 (36.5%)Separate-room15 (11.9%)Bedtime routine consistency (night/week)5.70 ± 1.53≤ 557 (45.2%)4.32 ± 1.21> 569 (54.8%)6.84 ± 0.37Notes: aBDI-II > 13 was applied to indicate women with depressive tendencies. bOn the BISQ-R 5-point Likert scale (0–4), sleep problem was rated 0 to 4 as “not a problem at all”, “a very small problem”, “a small problem”, “a moderate problem”, and “a serious problem”, respectively. cOn the BISQ-R 5-point Likert scale (0–4), toddlers rated as having a moderate (3) or serious (4) sleep problem were classified as “Yes”. Description of Maternal Depression, Toddler Sleep, and Sleep Parenting Practices Notes: aBDI-II > 13 was applied to indicate women with depressive tendencies. bOn the BISQ-R 5-point Likert scale (0–4), sleep problem was rated 0 to 4 as “not a problem at all”, “a very small problem”, “a small problem”, “a moderate problem”, and “a serious problem”, respectively. cOn the BISQ-R 5-point Likert scale (0–4), toddlers rated as having a moderate (3) or serious (4) sleep problem were classified as “Yes”. Table 4Regression Analysis: Direct Effect of Maternal Depression and the Moderating Effect of Sleep Parenting Practices on Sleep ProblemPredictor/ModelIIIIIIIVBβBβBβBβToddler healtha−0.77***−0.33***−0.67**−0.29**−0.69***−0.30***−0.65**−0.28**Mother incomea (Thousands/M)−0.12−0.06−0.04−0.02−0.02−0.01−0.05−0.02Mother educationa−0.35−0.15−0.34−0.15−0.35−0.16−0.34−0.15Maternal depression0.02*0.20*0.05**0.47**0.020.20Sleep placea0.000.00Mean-centered maternal depression xSleep placea−0.05*−0.34*Bedtime routine consistencya−0.13−0.06Mean-centered maternal depression xBedtime routine consistencya0.000.00ΔR20.040.040.00R20.160.200.240.20F7.63***7.33***6.20***4.92***Notes: * p < 0.05, ** p < 0.01, *** p < 0.001. aThe reference groups in toddler health, mother income, mother education, sleep place, and bedtime routine consistency are “fair/poor”, “≤60”, “≤ undergraduate”, “co-sleep”, and “≤ 5”, respectively. Regression Analysis: Direct Effect of Maternal Depression and the Moderating Effect of Sleep Parenting Practices on Sleep Problem Notes: * p < 0.05, ** p < 0.01, *** p < 0.001. aThe reference groups in toddler health, mother income, mother education, sleep place, and bedtime routine consistency are “fair/poor”, “≤60”, “≤ undergraduate”, “co-sleep”, and “≤ 5”, respectively. Table 5Regression Analysis by Sleep Place: Direct Effect of Maternal Depression on Toddler Sleep ProblemPredictor/Group/ModelCo-Sleep (N = 65)Not Co-Sleep (N = 57)IIIIIIBβBβBβBβToddler healtha−0.53−0.23−0.40−0.17−0.97**−0.42**−0.95**−0.42*Mother incomea (Thousands/M)−0.33−0.16−0.14−0.070.080.040.090.04Mother educationa−0.41−0.17−0.43−0.18−0.34−0.16−0.34−0.15Maternal depression0.05**0.41**0.000.03ΔR20.150.00R20.160.310.200.20F3.72*6.73***4.69**3.47*Notes: * p < 0.05, ** p < 0.01, *** p < 0.001. aThe reference groups in toddler health, mother income, and mother education were: “fair/poor”, “≤60”, and “≤ undergraduate”, respectively. Regression Analysis by Sleep Place: Direct Effect of Maternal Depression on Toddler Sleep Problem Notes: * p < 0.05, ** p < 0.01, *** p < 0.001. aThe reference groups in toddler health, mother income, and mother education were: “fair/poor”, “≤60”, and “≤ undergraduate”, respectively. Table 6Regression Analysis: Direct Effect of Maternal Depression and the Moderating Effect of Sleep Parenting Practices on Sleep Latency and Nocturnal AwakeningPredictor/Dependent Variable/ModelSleep Latency (Minute)Awakening (Number/Night)IIIIIIIVIIIBβBβBβBβBβBβToddler healtha−8.24−0.12−4.11−0.06−4.26−0.06−2.44−0.04−0.76**−0.29**−0.67**−0.25**Mother incomea (Thousands/M)−0.40−0.012.830.054.230.071.940.03−0.22−0.10−0.15−0.07Mother educationa−4.45−0.07−4.05−0.06−3.64−0.05−4.22−0.06−0.13−0.05−0.13−0.05Maternal depression0.87**0.27**0.720.220.87*0.27*0.020.16Sleep placea−6.36−0.11Mean-centered maternal depression x Sleep placea0.300.07Bedtime routine consistencya−10.84*−0.18*Mean-centered maternal depression x Bedtime routine consistencya−0.14−0.03ΔR20.070.010.020.02R20.020.090.100.120.110.13F0.872.83*2.172.67*4.91**4.50**Notes: * p < 0.05, ** p < 0.01. aThe reference groups in toddler health, mother income, mother education, sleep place, and bedtime routine consistency are “fair/poor”, “≤60”, “≤ undergraduate”, “co-sleep”, and “≤ 5”, respectively. Regression Analysis: Direct Effect of Maternal Depression and the Moderating Effect of Sleep Parenting Practices on Sleep Latency and Nocturnal Awakening Notes: * p < 0.05, ** p < 0.01. aThe reference groups in toddler health, mother income, mother education, sleep place, and bedtime routine consistency are “fair/poor”, “≤60”, “≤ undergraduate”, “co-sleep”, and “≤ 5”, respectively. Participants Characteristics and Descriptive Statistics of Main Variables The mean age of mothers was 33.01 (SD = 4.33), and the mean monthly income was NT$ 39.14 thousand (SD = 41.27). The mean age of toddlers was 1.79 years old (SD = 0.58). As shown in Table 2, 97.6% of women were married, and 73.8% held an undergraduate degree or below (66.7% of the total participants were undergraduates). About half were aged 33 or younger (54.8%) and reported a monthly income of more than NT$ 35 thousand (50.8%). 61.1% of the toddlers were younger than 2 years old, with boys (51.6%) slightly outnumbering girls. 75.4% were first-born and in good health, 86.5% having a birth weight of at least 2,500 grams. As shown in Table 3, the mean BDI-II score was 12.38 (SD = 9.11), and 33.3% of the mothers exhibited depressive tendencies. Regarding sleep characteristics, 11.9% of the toddlers were perceived as having “moderate” to “serious” sleep problem. Furthermore, 50.8% had a sleep latency exceeding 45 minutes, 34.9% woke up at least twice per night, and 91.3% slept less than 720 minutes per night. In terms of sleep parenting practices, more than half of the mothers (51.6%) practiced co-sleep, while 36.5% used separate surfaces, and 12.0% slept in separate rooms. Regarding bedtime routine consistency, 45.2% of the mothers were classified as having “low consistency” (≤5 nights/week). Comparison of Toddler Sleep Based on Participant Characteristics Table 2 presents the independent samples t-tests, which revealed significant differences in sleep indicators across five demographic variables. Specifically, perceived toddler sleep problem varied significantly by maternal education (t = 2.84, p < 0.01) and toddler health (t = −2.32, p < 0.05). Sleep latency differed by birth order (t = 2.09, p < 0.05) and birth weight (t = −2.38, p < 0.05). The number of nocturnal awakenings was significantly associated with birth order (t = 2.15, p < 0.05). Finally, nighttime sleep duration differed by maternal income (t = −2.39, p < 0.01) and toddler health (t = 3.31, p < 0.01). Pearson correlation analysis indicated that maternal depression was positively correlated with sleep problem, sleep latency, and nocturnal awakenings (rs = 0.24 to 0.30, ps < 0.01). These findings suggest that higher levels of maternal depression are associated with increased severity of sleep problem, longer sleep latency, and more nocturnal awakenings. However, maternal depression was not correlated with nighttime sleep duration (r = 0.01, p > 0.05). Therefore, no regression analysis was performed for nighttime sleep duration. Hierarchical Multiple Regression Analyses and Moderation Analyses To examine the associations between maternal depression and three toddler sleep indicators (problem, latency, and nocturnal awakenings), control variables were selected based on the preliminary difference analyses results. Among the five identified variables, toddler health was prioritized over birth weight due to their high correlation (χ2 = 0.05, p > 0.05). Birth order was not included due to its skewed distribution (75.0% first-born). Consequently, toddler health, maternal education, and maternal income were included as covariates. For the moderation analysis, “separate surfaces” and “separate rooms” were collapsed into a single “not co-sleep” group (coded as 1) due to the low frequency of separate-room arrangements. The reference groups (coded as 0) were defined as: “fair/poor” toddler health, “≤ undergraduate” education, “≤ NT$ 35 thousand” income, “co-sleep” place, and “low” bedtime routine consistency. Perceived Sleep Problem As shown in Table 4, hierarchical regression analyses were performed in four steps. In Model I, toddler health was negatively associated with toddler sleep problem (B = −0.77, β = −0.33, p < 0.001), indicating that toddlers in the “fair/poor” health group were perceived as having more severe sleep difficulties. In Model II, maternal depression positively associated with sleep problem (B = 0.02, β = 0.20, p < 0.05). Regarding moderation effects, Model III revealed a significant interaction between mean-centered maternal depression and sleep place (B = −0.05, β = −0.34, p < 0.05). This suggests that children’s sleep place moderated the relationships between maternal depression and sleep problem. In contrast, Model IV showed no significant interaction for bedtime routine consistency (p > 0.05). To further parse the significant interaction, simple slope analyses were conducted. Results as shown in Table 5 indicated that maternal depression was significantly associated with toddler sleep problem only within the co-sleep group (B = 0.05, β = 0.41, p < 0.01). In this group, a one-point increase in maternal depression was associated with a 0.05 increase in sleep problem. Conversely, no significant relationship was observed in the “not co-sleep” group, suggesting that independent sleeping may function as a protective factor that buffers the link between maternal distress and children’s sleep problem. Sleep Latency and Nocturnal Awakenings As shown in Table 6, the control variables in Model I were not associated with sleep latency. While Model II showed that maternal depression was associated with sleep latency (B = 0.87, β = 0.27, p < 0.01), no significant moderating effects were observed for either sleep place or routine consistency in Models III and IV. Regarding nocturnal awakenings, toddler health remained an associated factor across Model I (B = −0.76, β = −0.29, p < 0.01). However, maternal depression did not act as an associated factor. Therefore, the analysis of moderation effects was not performed. Discussion The present study investigated the associations between maternal depression and toddler sleep and the potential moderating effects of sleep parenting practices. Data were collected via online questionnaires completed by 126 mother-toddler dyads. The findings describe toddler mothers’ depression and sleep parenting practices as well as toddler sleep. Specifically, 33.3% of toddler mothers exhibited depressive tendencies. Most toddlers had a nocturnal sleep duration of less than 12 hours (91.3%). Half of the toddlers had a long sleep latency (50.8%), and 35.0% woke up at least twice during the night; 11.9% mothers perceived their children have a sleep problem. In sleep parenting practices, 51.6% of the participants practiced co-sleeping, and 45.2% maintained low bedtime routine consistency. The study found that higher levels of maternal depression are significantly associated with both a greater perceived severity of sleep problem and a prolonged latency. Among the four tested moderation pathways, only one pathway—maternal depression-sleep place-toddler sleep problem—reached statistical significance. This finding suggests that sleep place acts as a buffer against the association between maternal depression and perceived toddler sleep problem. Simple slope analysis further elucidated that the negative association between maternal depression and perceived toddler sleep problem was evident only among co-sleep dyads. Consequently, the first and the second hypotheses were partially supported, and the third was rejected. The finding lends empirical support to the transactional model of children’s sleep.30 In co-sleep arrangements, the physical proximity may facilitate the emotional spillover of maternal restlessness and physiological arousal to children. Conversely, independent sleeping (including separate surfaces or rooms) appears to buffer this association, providing a structural barrier that mitigates the transmission of maternal psychological distress. By establishing a physical boundary during the night, families can effectively buffer the child against the “spillover” of maternal depressive symptoms, thereby safeguarding the child’s sleep integrity even when the mother is experiencing psychological challenges. Interestingly, this moderating effect was exclusive to “perceived sleep problem”—a summative parental perception—and did not extend to “sleep latency.” This divergence suggests that while maternal depression and physical proximity may be related to how a mother perceives and evaluates her child’s sleep quality, the behavioral process of falling asleep (latency) may be governed by more robust biological factors and less susceptible to the maternal-child interaction environmental factors.16 By defining bedtime routine consistency as “consistent activities conducted within one hour before lights-out,” the results indicated that consistency did not moderate the association between maternal depression and toddler sleep. It has been posited that the relevant factors of children’s sleep may be the duration, components, and emotional significance of routines rather than their mere regularity. For instance, routines exceeding 30–40 minutes may paradoxically delay sleep onset and shorten total sleep duration.21 For the components of routines, hygiene and literacy-based routines (eg, brushing teeth, storytelling) are conducive to sleep, while stimulating activities (eg, screen time, high-intensity play) tend to be disruptive.31 Also, bedtime routines embody the characteristics of nurturing care and early child stimulation. They are rich opportunity that is imbued with emotional significance that can generate positive early developmental outcomes (including security and safety, responsive caregiving, etc).21,32 A potential explanation for the non-significant moderation result is that bedtime consistency does not capture the affective synchrony or emotional warmth—elements that are more important in buffering maternal depression. For example, a mother who reads with genuine warmth and one who follows the same routines mechanically would be coded identically in terms of consistency, yet their emotional impact remains different. Regarding nocturnal awakenings, although a bivariate correlation with maternal depression was observed, this relationship vanished when child health status was controlled for, leaving toddler health as the sole robust relevant factor. Similarly, nocturnal sleep duration showed only a marginal association (p = 0.06). These findings suggest that a negative cognitive bias did not uniformly cloud toddler sleep characteristics reported by mothers; rather, mothers provided differentiated and nuanced assessments across various sleep indicators. The findings of this study partially align with several studies while diverging from others. The results regarding toddler sleep duration resonate with findings from diverse cohorts, including indigenous New Zealand, general Chinese, and multi-ethnic Singaporean families.9,11,13 Regarding perceived sleep problem, our results parallel those of one study,9 which indicated that maternal depression was concurrently related to toddler sleep problem in three-year-old. However, this contrasts with other cross-sectional and longitudinal evidence suggesting no concurrent10 or predictive relationships.9,13 In terms of sleep latency, our findings are consistent with research on low-income US families, where maternal depression was significantly associated with a composite index of sleep latency and persistence,8 while the result was different from one study that found no longitudinal association between early postpartum depression and sleep latency at 30 months.13 Finally, regarding nocturnal awakenings, our non-significant results diverge from a Norwegian database analysis14 and a Chinese study,11 both of which found that maternal depression could predict nocturnal awakenings in the toddler stage. In summary, while the lack of association between maternal depression and toddler nocturnal sleep duration aligns with the literature, findings for sleep problem, sleep latency, and nocturnal awakenings diverge from previous research. These discrepancies may stem from the methodological heterogeneity—including variations in assessment timing, instruments, analytical strategies, and the use of cross-sectional versus longitudinal designs. Consequently, it is premature to determine the extent to which the literature findings are moderated by ethnicity or socioeconomic status. In summary, with a focus on Taiwanese toddler families, the findings have several implications. First, the descriptive data can provide references for maternal depression, sleep parenting practices, and toddler sleep. Second, the association between maternal depression and toddler sleep underscores the necessity of integrating maternal mental health screenings into pediatric sleep consultations. Practitioners must pay attention to the association between maternal depression and perceived toddler sleep problem and sleep latency. Third, the moderating role of sleep place in the relationship between maternal depression and perceived toddler sleep problem also has implications for pediatric sleep consultations: promoting a non-co-sleeping arrangement may serve as an effective strategy to safeguard the toddler’s sleep quality, especially in the context of high maternal depressive tendency. Several limitations warrant consideration. First, the cross-sectional nature of this study precludes the exclusion of reverse causality (eg, children’s sleep disturbances exacerbating maternal depressive symptoms). Second, both maternal depression and toddler sleep data relied on maternal self-reports. It remains to be determined whether different results would emerge if other informants assessed maternal depression or if toddler sleep were measured through objective methods, such as actigraphy. Third, no psychometric evaluation was performed for the Chinese version of BISQ-R. Fourth, combining three sleep places (co-sleep, separate surfaces, and separate rooms) into two categories (co-sleep and not co-sleep) may reduce sensitivity of analysis. Also, aggregating data across the 1–3 year age range may overlook the developmental heterogeneity inherent in early childhood sleep. Fifth, several unmeasured confounders—such as the quality of parent-child affective synchrony, child temperament, paternal factors, and marital quality—were not included in the analysis. Finally, the use of online recruitment and a relatively modest sample size may constrain the generalizability of these findings. Future research should employ longitudinal paradigms to clarify causal effects and capture fluctuations over time. Collecting data using a multi-method and multi-informant design, exploring mother-child interaction during bedtime routines and bedsharing, recruiting a larger sample of representative participants, and analyzing data within a narrower age range will be essential to advancing our understanding of these complex developmental dynamics. Conclusions By expanding the focus from infancy to the toddler stage, targeting Taiwanese toddler families, and examining the moderating role of parenting on the associations among maternal depression and toddler sleep, this study establishes descriptive data on maternal depression, toddler sleep, and sleep parenting. Associations between maternal depression, toddler sleep, and sleeping arrangements acting as a critical moderator in perceived sleep problem were noted. The association between maternal depression and toddler sleep problem was found to be significant only among families practicing co-sleep. These results shed light on pediatric sleep consultations. Further studies are warranted to address the limitations and increase the generalizability of the research findings. Acknowledgments We thank all the toddler mothers for their participation. Funding Statement This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Data Sharing Statement The data that support the findings of this study are available from the corresponding author Pey-Ling Shieh upon reasonable request. Author Contributions Hsiao-Yen Wu: Conceptualization, Data curation, Formal analysis, Investigation, Project administration, Writing – original draft. Pey-Ling Shieh: Conceptualization, Formal analysis, Methodology, Resources, Supervision, Validation, Visualization, Writing – original draft and Writing – review and editing. Yen-Chun Liu: Methodology, Resources, and Writing – original draft. All authors gave final approval for the version to be published; have agreed on the journal to which the article has been submitted; and agreed to be accountable for all aspects of the work. Disclosure The authors report no conflicts of interests in the reporting of this work. References Associated Data Data Availability Statement The data that support the findings of this study are available from the corresponding author Pey-Ling Shieh upon reasonable request.