The Mediating Role of Sleep Quality in Associations of Smartphone Addiction and Time-Specific Use With Prenatal Depression and Anxiety: Cross-Sectional Observational Study.

The Mediating Role of Sleep Quality in Associations of Smartphone Addiction and Time-Specific Use With Prenatal Depression and Anxiety: Cross-Sectional Observational Study

Abstract

Background

Prenatal depression and anxiety have emerged as pressing global public health challenges. In the digital era, excessive smartphone use may exert complex effects on maternal mental health. Prior research has predominantly focused on adolescents, whereas evidence among pregnant women, particularly regarding total weekly use and use during critical circadian windows, such as after waking and before bedtime, remains limited. Furthermore, the mediating role of sleep quality in these associations has yet to be systematically examined.

Objective

This study aimed to investigate the associations of smartphone addiction and time-specific smartphone use with prenatal depression and anxiety, and to assess the mediating role of sleep quality in these associations.

Methods

This cross-sectional observational study enrolled 537 pregnant women. Smartphone use characteristics, sleep quality, and prenatal depression and anxiety were assessed using the Smartphone Addiction Scale-Short Version (SAS-SV), objective total weekly smartphone use recorded by the device, and self-reported duration of smartphone use after waking and duration of bedtime smartphone use; the Pittsburgh Sleep Quality Index (PSQI), the Patient Health Questionnaire-9 (PHQ-9), and the Generalized Anxiety Disorder-7 (GAD-7), respectively. Binary logistic regression was used to estimate associations, restricted cubic spline (RCS) models were used to explore nonlinear patterns, and exploratory mediation analyses were used to examine whether sleep quality statistically accounted for part of the observed associations.

Results

The prevalence of moderate-to-severe prenatal depression and anxiety was 8.8% (47/537) and 5.8% (31/537), respectively. After adjusting for covariates, smartphone addiction was associated with higher odds of depression (odds ratio [OR] 3.461, 95% CI 1.771‐7.100) and anxiety (OR 7.738, 95% CI 3.064‐23.172). Duration of smartphone use after waking ≥61 minutes and duration of bedtime smartphone use ≥91 minutes were also associated with higher odds of affective disorders. RCS analyses further revealed distinct nonlinear patterns across different exposure dimensions: duration of smartphone use after waking displayed U-shaped and J-shaped associations with depression and anxiety, respectively, while total weekly use displayed a U-shaped association with anxiety. Exploratory mediation analyses showed that the indirect pathways through sleep quality were statistically significant for the associations of bedtime smartphone use, weekly smartphone use, and smartphone addiction with prenatal depression, with mediation proportions of 44.89% (0.0065/0.0145), 31.30% (0.0056/0.0183), and 31.48% (0.0415/0.1315), respectively. The corresponding proportions in the anxiety models were 35.83% (0.0044/0.0123), 27.05% (0.0039/0.0144), and 22.73% (0.0271/0.1187).

Conclusions

Smartphone addiction and excessive smartphone use after waking and before bedtime were associated with higher odds of prenatal depression and anxiety. Sleep quality may statistically account for part of these associations, but longitudinal studies are needed to establish temporal ordering and causal pathways.

Introduction

Pregnancy represents a critically vulnerable period in a woman’s life course, both physiologically and psychologically [1-3]. During this period, women not only endure physiological challenges such as dramatic hormonal fluctuations and early pregnancy symptoms, but also face multiple psychosocial stressors, including shifts in social roles, restructuring of family relationships, and concerns about fetal well-being [4,5]. Although pregnancy is generally regarded as a positive life event, it is frequently accompanied by significant affective disturbances. A systematic review reported that the global average prevalence of prenatal depression is approximately 28.5%, which is substantially higher than that of postpartum depression [6]. In mainland China, a systematic review encompassing 271 studies and 369,477 women found that the average prevalence of prenatal anxiety reached 17.4% [7]. These common prenatal affective disorders, if left unaddressed, can substantially increase the risk of adverse obstetric outcomes, such as preterm birth, low birth weight, and preeclampsia [8,9].

In the digital era, smartphones have become an integral extension of daily life. For pregnant women, smartphones serve as important tools for accessing health information and maintaining social support. Healthy smartphone use can, to some extent, alleviate the sense of isolation and uncertainty brought about by the temporary disruption of social roles [10]. Conversely, the fragmented entertainment offered by smartphones can easily trigger overuse, blurring the self-perceived boundary between functional use and pathological dependence [11]. Against this backdrop, the positive mental health effects of smartphone use may be reversed, turning it into a potential risk factor that jeopardizes maternal prenatal mental health [12].

Smartphone addiction, also referred to as problematic smartphone use, is defined as excessive dependence on smartphones, characterized by core features of behavioral addiction such as withdrawal anxiety, increased tolerance, and impaired social functioning [13,14]. In recent years, the adverse effects of smartphone addiction on mental health have been widely documented across diverse populations. Large-scale studies among adolescents, in particular, have demonstrated significant positive associations of problematic smartphone use with depressive and anxiety symptoms as well as sleep disturbance [15-17]. Moreover, several longitudinal studies have shown that baseline levels of smartphone addiction prospectively predict the worsening of anxiety symptoms and the continued deterioration of sleep quality during follow-up, suggesting that smartphone addiction may not merely be a consequence of psychological distress but a potential cause that precipitates or exacerbates affective disorders [18,19].

Although previous research has established negative associations between smartphone addiction and mental health in adolescents [20,21], studies examining the association between smartphone addiction and prenatal affective disorders among pregnant women, a population characterized by marked physiological and psychological fluctuations, remain scarce. In addition, few studies have distinguished total smartphone use time from use during specific periods. This distinction may be theoretically relevant: circadian rhythm disruption plays a key role in the onset, recurrence, and treatment response of mood disorders, and the central role of circadian rhythms in emotional regulation is widely recognized [22,23]. Screen exposure during different circadian phases may influence emotional states through distinct neurobiological pathways. Specifically, bedtime screen exposure is closely associated with reduced sleep quality [24], whereas screen exposure after waking may have different effects on the daytime stress regulation system [25]. Compared with absolute usage duration, whether screen exposure during key circadian windows, such as bedtime and after waking, is associated with prenatal depressive and anxiety symptoms has yet to be systematically investigated.

In recent years, studies on adolescents have shown that smartphone use is associated with reduced sleep quality and emotional problems [20,26,27]. Experimental studies have further found that nighttime blue light exposure can prolong sleep latency and cause a circadian phase shift by suppressing melatonin secretion [28], while the cognitive and emotional arousal triggered by immersive screen exposure may also interfere with sleep onset and maintenance [29]. It is worth noting that sleep quality in pregnant women undergoes a physiological deterioration as pregnancy progresses, with the prevalence of sleep disturbances rising significantly from the first trimester to the third trimester [30]. This vulnerability may stem from the combined effects of hormonal fluctuations, increased nocturia, and physical discomfort [31]. Furthermore, poor sleep quality is associated with an increased risk of prenatal depression and anxiety [32,33], and its neural mechanisms may involve the effects of sleep deprivation on prefrontal cortex function and amygdala reactivity [34,35]. Based on this prior evidence and biological plausibility, it is reasonable to infer that sleep quality may mediate the relationship between smartphone use and prenatal affective disorders. However, current research on this topic remains scarce, particularly studies that comprehensively assess multiple dimensions, such as smartphone use at specific time points and smartphone addiction.

To address this research gap, this study aims to investigate the associations between smartphone addiction and usage during specific time periods and prenatal depression and anxiety, and to explore the mediating effect of sleep quality on these associations, to provide evidence-based support for the development of digital health guidelines and clinical intervention strategies during pregnancy.

Methods

Study Population

This study used a cross-sectional design and was conducted in the obstetrics outpatient clinic and inpatient ward of Hunan Provincial Maternal and Child Health Hospital. Participants were recruited using convenience sampling. The survey was administered by uniformly trained interviewers, who explained the study’s purpose and procedures to eligible pregnant women while they were waiting to give birth in the admission area or in the ward. After obtaining informed consent, the interviewers provided one-on-one guidance to help participants scan a QR code to complete the electronic questionnaire, which was filled out and submitted on the spot. Investigators provided necessary explanations and assistance but did not interfere with the respondents’ answers. Inclusion criteria were as follows: (1) permanent residents of Changsha, (2) all prenatal checkups completed at the study hospital, (3) age ≥18 years, (4) singleton pregnancy. Exclusion criteria were (1) severe cognitive impairment or mental illness preventing effective communication or understanding of the questionnaire content, and (2) presence of legally reportable infectious diseases such as HIV, syphilis, or hepatitis B. Participants who met all inclusion criteria and none of the exclusion criteria were formally enrolled in the study after signing the informed consent form.

All data in this study were collected by having participants scan a QR code to complete a structured questionnaire. All key research variables in the questionnaire were marked as required fields; participants could not submit the questionnaire unless they had completed all required items. Consequently, none of the variables analyzed in this study had missing values, and all 537 cases ultimately included in the analysis were valid questionnaires.

Ethical Considerations

The study protocol, encompassing participant recruitment procedures, informed consent procedures, data collection, and quality control methods, was reviewed and approved by the Ethics Committee of the Xiangya School of Public Health, Central South University (approval number: XYGW-2024‐19) and the Ethics Committee of the Hunan Provincial Maternal and Child Health Care Hospital (approval number: 2023-S189). After fully understanding the study’s objectives, procedures, potential risks, and benefits, all participants voluntarily signed a paper informed consent form and were informed of their right to withdraw from the study at any time without condition. No remuneration or compensation was provided to any participant in this study.

Exposure Measurement

Smartphone addiction tendency was assessed using the Smartphone Addiction Scale-Short Version (SAS-SV). The scale consists of 10 items rated on a 6-point Likert scale (1=“strongly disagree” to 6=“strongly agree”), yielding a total score ranging from 10 to 60, with higher scores indicating a greater risk of smartphone addiction. According to the developers’ recommendation, a cut-off score of 33 was adopted to define a high risk of addiction for females. The SAS-SV has demonstrated good psychometric properties, with an internal consistency Cronbach α coefficient of 0.91 [36]. The scale has been widely applied and validated in Chinese populations and is suitable for the rapid screening of smartphone dependence among adults [37].

Time-specific smartphone use was assessed using self-developed items, including duration of bedtime smartphone use and duration of smartphone use after waking. Bedtime smartphone use was defined as the cumulative duration of smartphone use from the moment participants lay down in bed with the intention to sleep until they actually fell asleep. Smartphone use after waking was defined as the cumulative duration of use from the moment participants awoke in the morning until they got out of bed. Both definitions were clearly stated in the questionnaire and were verbally explained in a standardized manner by trained investigators before questionnaire completion. Investigators specifically emphasized that the reported duration should only include time spent actively using the phone while lying in bed, rather than the total time spent in bed including nonuse states. Meanwhile, participants were instructed to recall their typical usage patterns over the past week rather than relying on a single day’s experience. Both measures were specific to nighttime sleep periods and did not include daytime naps, and were recorded in minutes. Weekly smartphone use was obtained by asking participants to check the average daily usage duration of all applications over the past week through the screen time management feature built into their smartphone system settings (corresponding to “Screen Time” or “Healthy Phone Usage Time” or “App Management Time” in Android and Apple iOS systems, respectively), and was recorded in hours. According to the current distribution of data and previous studies on smartphone use duration, we classified participants into 4 consecutive groups as follows: (1) duration at bedtime [25]: 0‐30 minutes, 31‐60 minutes, 61‐90 minutes, and ≥91 minutes; (2) duration after waking [38]: 0‐1 minutes, 2‐10 minutes, 11‐60 minutes, and ≥61 minutes; (3) duration of total week [39,40]: 0‐14 hours/week, 15‐28 hours/week, 29‐49 hours/week, and ≥50 hours/week.

Participants’ sleep quality over the past month was assessed using the Pittsburgh Sleep Quality Index (PSQI). Developed by Buysse et al [41], the PSQI comprises 7 dimensions: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep medication, and daytime dysfunction. The total score, ranging from 0 to 21, is calculated as the sum of all component scores. Based on the threshold established by previous studies, a total score of ≥7 was defined as poor sleep quality in this study [42,43].

Outcome Measurement

Depressive symptoms over the past 2 weeks were assessed using the Patient Health Questionnaire-9 (PHQ-9). Developed by Kroenke et al [44] based on the DSM-IV diagnostic criteria for depression, the PHQ-9 comprises nine items. Each item asks respondents to rate the frequency of each symptom over the past 2 weeks on a 4-point Likert scale (0=“not at all” to 3=“nearly every day”), yielding a total score ranging from 0 to 27, with higher scores indicating more severe depressive symptoms. The PHQ-9 has demonstrated good reliability and validity among Chinese pregnant women [45]. As prior research has shown that moderate-to-severe depressive symptoms may lead to more adverse outcomes, a cut-off score of ≥10 was adopted in this study to define moderate-to-severe depressive symptoms, in line with previous studies [46,47]. Hereafter, the term “prenatal depression” refers to moderate-to-severe prenatal depressive symptoms.

Anxiety symptoms were measured using the Generalized Anxiety Disorder-7 (GAD-7). Developed by Spitzer et al [48], the GAD-7 is designed for rapid screening of generalized anxiety symptoms over the past 2 weeks. The scale consists of 7 self-reported items, each requiring respondents to rate the frequency of anxiety symptoms over the past 2 weeks on a 4-point Likert scale (0=“not at all” to 3=“nearly every day”), with total scores ranging from 0 to 21, higher scores reflecting more severe anxiety symptoms. Given that moderate-to-severe anxiety symptoms may be associated with more detrimental effects, a cut-off score of ≥10 was used to define moderate-to-severe anxiety, consistent with previous research [49,50]. Hereafter, the term “prenatal anxiety” refers to moderate-to-severe prenatal anxiety symptoms.

Covariates

Based on previous studies [12,51-55], the following variables that may influence smartphone use and prenatal mental health outcomes were included as covariates in the analyses, covering 3 domains: sociodemographic characteristics, behavioral and lifestyle factors, and health and pregnancy-related characteristics. Sociodemographic characteristics included maternal age (years), maternal educational attainment (junior high school or below, high school or vocational high school, and college or above), and paternal educational attainment (junior high school or below, high school or vocational high school, and college or above). Behavioral and lifestyle factors included prepregnancy smoking status within the 6 months before pregnancy (yes or no), prepregnancy alcohol consumption within the 6 months before pregnancy (yes or no), frequency of physical exercise during pregnancy (<once per week, 1‐3 times per week, 4‐6 times per week, and ≥ once per day), and frequency of social media posting on platforms such as WeChat (Tencent) and Xiaohongshu (Xingyin Information Technology Ltd.) during pregnancy (less than once every 6 months, 2‐5 times per year, 1‐2 times per month, 3‐4 times per month, and several times per week). Health- and pregnancy-related characteristics included gestational age (weeks), prepregnancy BMI (kg/m²), gestational weight gain (kg), gestational diabetes mellitus (yes or no), abortus (times), and gravida (times).

Statistical Analysis

Descriptive statistics were computed for the dataset, with continuous variables presented as means and SD and categorical variables expressed as counts and percentages. Between-group comparisons were performed using independent-samples t tests for continuous variables and chi-square tests for categorical variables. Pearson correlation coefficients were used to assess the bivariate associations among total weekly smartphone use duration, duration of smartphone use after waking, duration of bedtime smartphone use, SAS-SV score, PHQ-9 score, and GAD-7 score.

Subsequently, binary logistic regression models were fitted separately for each smartphone-use dimension and outcome. Two models were constructed to verify the robustness of the associations: Model 1 was a crude model that assessed the initial associations without adjusting for any covariates; Model 2 was a fully adjusted model that included maternal age, maternal educational attainment, paternal educational attainment, prepregnancy BMI, gestational weight gain, gestational age, gestational diabetes mellitus, gravida, abortus, prepregnancy smoking status, prepregnancy alcohol consumption, frequency of physical exercise during pregnancy, and frequency of social media posting during pregnancy as covariates. Results were expressed as odds ratios (ORs) with their 95% CIs.

To further explore potential nonlinear dose-response relationships between smartphone use, sleep quality, and prenatal affective disorders, restricted cubic spline (RCS) functions were fitted on the basis of the above logistic regression models. Following previous studies, 4 knots were placed at the fifth, 35th, 65th, and 95th percentiles, and the median of each independent variable was set as the reference value [56-58]. Finally, mediation models were established to exploratorily examine the indirect pathways of sleep quality in the associations between multidimensional smartphone use and prenatal affective disorders. In our mediation analysis, we clearly distinguished between the total effect, the direct effect, and the indirect effect based on the effect decomposition framework. We used the continuous scores of the PHQ-9 and GAD-7 as outcome variables to preserve the informational integrity of the original scales and enhance the statistical power for detecting indirect effects. Point estimates for direct and indirect effects were obtained using linear regression models, and their 95% CIs were estimated via 1000 Monte Carlo simulations; the mediation proportion was calculated by dividing the indirect effect by the total effect. All models were adjusted for all abovementioned covariates. In addition, to examine the impact of outcome definition on the stability of mediation results, we conducted a sensitivity analysis based on a binary outcome (PHQ-9 and GAD-7 score ≥10). Given the cross-sectional design of this study, the results of the mediation analysis should be interpreted as an exploratory decomposition of association pathways rather than evidence of causal mechanisms.

Given the limited number of events for moderate-to-severe depression (n=47) and anxiety (n=31), two sensitivity analyses were conducted to assess the robustness of the main findings. First, Firth penalized maximum likelihood estimation was applied to refit all models to reduce small-sample bias in rare-events logistic regression [59]. Second, the threshold for the PHQ-9 and GAD-7 outcomes was lowered from ≥10 to ≥5 (ie, including those with mild and above depression and anxiety symptoms) to evaluate whether the observed associations were sensitive to the choice of the cutoff for symptom severity.

Data analyses were performed using SPSS 26.0 (IBM Corp) and R 4.4.1 (R Foundation for Statistical Computing). The primary R packages used included dplyr, rms, mediation, and ggplot2. The significance level was set at α=.05, and P<.05 was considered statistically significant.

Results

Baseline Characteristics of Participants

Table 1 presents the baseline characteristics of the study participants. A total of 537 pregnant women were included, with a mean age of 31.7 (4.0) years. Moderate-to-severe prenatal depression and anxiety were present in 47 (8.8%) and 31 (5.8%) participants, respectively. The mean SAS-SV score (SD) in the moderate-to-severe depression group was 36.4 (9.4), significantly higher than that in the nonmoderate-to-severe depression group (30.8, SD 8.9). The moderate-to-severe depression group also had longer weekly screen time and significantly longer durations of smartphone use after waking and before bedtime compared with the nonmoderate-to-severe depression group (P=.02 and P=.002, respectively). In addition, the PSQI score in the moderate-to-severe depression group was significantly higher than that in the nonmoderate-to-severe depression group (P<.001). The pattern of differences between the moderate-to-severe anxiety group and the nonmoderate-to-severe anxiety group was generally similar. The overall detection rate of sleep disturbance was 46.2% (n=248), and the detection rates in both the moderate-to-severe depression group and the moderate-to-severe anxiety group were significantly higher than those in their respective control groups (P=.003 and P=.008, respectively).

Table 1. Baseline characteristics of the study participants (N=537)a.

Variables Total Prenatal depression(PHQ-9b score≥10) Prenatal anxiety(GAD-7c score≥10)

No Yes P value No Yes P value

Age, mean (SD) years 31.7 (4.0) 31.7 (4.0) 32.0 (3.6) .60 31.7 (4.0) 32.3 (4.5) .40

BMI, mean (SD) kg/m² 21.0 (2.9) 21.1 (2.9) 20.8 (3.0) .48 21.0 (2.9) 21.2(2.8) .78

Gravida, mean (SD) times 2.0 (1.1) 2.0 (1.1) 2.1 (1.4) .29 2.0 (1.1) 2.0 (1.2) .85

Abortus, mean (SD) times 1.6 (0.9) 1.5 (0.8) 1.7 (1.0) .30 1.6 (0.9) 1.7 (0.8) .55

Weight gain during pregnancy, mean (SD) kg 15.2 (5.9) 15.3 (5.9) 14.0 (5.4) .13 15.2 (5.8) 14.6 (7.0) .56

Gestational age, mean (SD) weeks 38.87 (1.43) 38.86 (1.44) 38.98 (1.39) .86 38.87 (1.43) 38.91 (1.46) .94

Gestational diabetes mellitus, n (%) .58 .46

 No 441 (82.1) 401 (81.8) 40 (85.1)   414 (81.8) 27 (87.1)  

 Yes 96 (17.9) 89 (18.2) 7 (14.9)   92 (18.2) 4 (12.9)  

Maternal education attainment, n (%) .02d .61

 Junior high school or below 19 (3.5) 15 (3.1) 4 (8.5)   17 (3.4) 2 (6.5)  

 High school or vocational high school 43 (8.0) 36 (7.3) 7 (14.9)   40 (7.9) 3 (9.7)  

 College or above 475 (88.5) 439 (89.6) 36 (76.6)   449 (88.7) 26 (83.9)  

Paternal education attainment, n (%)  .57 .29

 Junior high school or below 30 (5.6) 26 (5.3) 4 (8.5)   29 (5.7) 1 (3.2)  

 High school or vocational high school 58 (10.8) 52 (10.6) 6 (12.8)   52 (10.3) 6 (19.4)  

 College or above 449 (83.6) 412 (84.1) 37 (78.7)   425 (84.0) 24 (77.4)  

Prepregnancy smoking, n (%) >.99 .09

 No 511 (95.2) 466 (95.1) 45 (95.7)   484 (95.7) 27 (87.1)  

 Yes 26 (4.8) 24 (4.9) 2 (4.3)   22 (4.3) 4 (12.9)  

Prepregnancy alcohol consumption, n (%) .48 .41

 No 373 (69.5) 343 (70.0) 30 (63.8)   354 (70.0) 19 (61.3)  

 Yes 164 (30.5) 147 (30.0) 17 (36.2)   152 (30.0) 12 (38.7)  

Frequency of physical exercise, n (%) .25 .30

 Less than once a week 223 (41.5) 197 (40.2) 26 (55.3)   206 (40.7) 17 (54.8)  

 1‐3 times a week 199 (37.1) 186 (38.0) 13 (27.7)   192 (37.9) 7 (22.6)  

 4‐6 times a week 62 (11.5) 58 (11.8) 4 (8.5)   59 (11.7) 3 (9.7)  

 Once a day or more 53 (9.9) 49 (10.0) 4 (8.5)   49 (9.7) 4 (12.9)  

Frequency of social media posts, n (%) .07 .33

 Less than once every six months 171 (31.8) 159 (32.4) 12 (25.5)   161 (31.8) 10 (32.3)  

 2‐5 times a year 144 (26.8) 125 (25.5) 19 (40.4)   132 (26.1) 12 (38.7)  

 1‐2 times a month 147 (27.4) 138 (28.2) 9 (19.1)   143 (28.3) 4 (12.9)  

 3‐4 times a month 53 (9.9) 50 (10.2) 3 (6.4)   49 (9.7) 4 (12.9)  

 Several times a week 22 (4.1) 18 (3.7) 4 (8.5)   21 (4.2) 1 (3.2)  

PSQIe total score, mean (SD) points 7.6 (3.3) 7.3 (3.2) 9.9 (3.6) <.001 7.4 (3.3) 9.9 (3.4) <.001

Poor sleep quality, n (%) .003 .008

 No 289 (53.8) 274 (55.9) 15 (31.9)   280 (55.3) 9 (29.0)  

 Yes 248 (46.2) 216 (44.1) 32 (68.1)   226 (44.7) 22 (71.0)  

Weekly smartphone use, mean (SD) hours 31.3 (27.6) 30.8 (26.8) 36.4 (34.2) .18 30.8 (26.8) 39.0 (37.4) .11

SAS-SVf total score, mean (SD) points 31.3 (9.1) 30.8 (8.9) 36.4 (9.4) <.001 30.8 (9.0) 38.3 (7.6) <.001

Smartphone addiction, n (%) <.001 <.001

 No 309 (57.5) 295 (60.2) 14 (29.8)   303 (59.9) 6 (19.4)  

 Yes 228 (42.5) 195 (39.8) 33 (70.2)   203 (40.1) 25 (80.6)  

Smartphone use after waking,mean (SD) minutes 29.1 (50.6) 27.4 (50.9) 46.3 (44.2) .02 28.1 (51.0) 45.0 (40.3) .07

Bedtime smartphone use, mean (SD) minutes 73.1 (53.4) 70.9 (52.1) 96.5(61.6) .002 71.3 (52.4) 102.9 (62.4) .001

aContinuous variables (age, BMI, gravida, abortus, weight gain during pregnancy, gestational age, PSQI total score, weekly smartphone use, SAS-SV total score, smartphone use after waking, and bedtime smartphone use) were analyzed using an independent samples t test, while categorical variables (gestational diabetes mellitus, maternal education attainment, paternal education attainment, prepregnancy smoking, prepregnancy alcohol consumption, frequency of physical exercise, frequency of social media posts, poor sleep quality, and smartphone addiction) were analyzed using the chi-square test.

bPHQ-9: Patient Health Questionnaire-9.

cGAD-7: Generalized Anxiety Disorder-7.

dStatistical significance (P<.05).

ePSQI: Pittsburgh Sleep Quality Index.

fSAS-SV: Smartphone Addiction Scale-Short Version.

The Correlation between Different Dimensions of Smartphone Use and Prenatal Affective Disorders

To preliminarily explore the bivariate associations between multidimensional smartphone use characteristics during pregnancy (SAS-SV total score, weekly smartphone use, bedtime smartphone use, and smartphone use after waking) and prenatal affective disorders (PHQ-9 and GAD-7 total scores), Pearson correlation analysis was performed, with the results presented in Figure 1. The correlation matrix showed that both the SAS-SV total score and all dimensions of smartphone use duration were significantly positively correlated with PHQ-9 and GAD-7 total scores (P<.05), providing preliminary evidence that excessive digital media exposure was related to prenatal affective disorders.

The Associations of Different Smartphone Use Dimensions and Sleep Quality With Prenatal Affective Disorders

The results of the independent associations between various dimensions of smartphone use and prenatal affective disorders are detailed in Table 2. Smartphone addiction was associated with higher odds of both outcomes in the crude and adjusted models. In the fully adjusted model, the OR was 3.461 (95% CI 1.771‐7.100) for prenatal depression and 7.738 (95% CI 3.064‐23.172) for prenatal anxiety. For bedtime use, compared with 0‐30 minutes, use of ≥91 minutes was associated with higher odds of depression (OR 3.716, 95% CI 1.467‐10.377) and anxiety (OR 5.141, 95% CI 1.568‐20.808). For use after waking, ≥61 minutes was associated with higher odds of depression (OR 4.758, 95% CI 1.736‐13.773) and anxiety (OR 10.902, 95% CI 2.606‐58.482). Weekly use showed a nonmonotonic categorical pattern: compared with 0‐14 hours/week, the 29-49 hours/week group had lower odds of depression (OR 0.378, 95% CI 0.119‐1.012) and anxiety (OR 0.221, 95% CI 0.033‐0.868), whereas the ≥50 hours/week group demonstrated elevated odds estimates for both outcomes (depression OR 1.687, 95% CI 0.795‐3.597; anxiety OR 1.979, 95% CI 0.830‐4.795). Sensitivity analyses using Firth penalized maximum likelihood estimation and an alternative threshold (PHQ-9 and GAD-7 score ≥5) yielded results consistent with the main findings in terms of direction, supporting the robustness of our results (Tables S1 and S2 in Multimedia Appendix 1).

Table 2.Association between different dimensions of smartphone use and prenatal affective disordersa.

Exposure Prenatal depression (PHQ-9b score≥10) Prenatal anxiety (GAD-7c score≥10)

n/N Model 1d,ORe (95% CI) Model 2f,OR (95% CI) n/N Model 1d,OR (95% CI) Model 2f,OR (95% CI)

SAS-SVg total score (points) —h 1.069 (1.035‐1.105) 1.070 (1.033‐1.110) — 1.091 (1.049‐1.137) 1.104 (1.056‐1.158)

 No 14/309 Refi Ref 6/309 Ref Ref

 Yes 33/228 3.566 (1.897‐7.041) 3.461 (1.771‐7.100) 25/228 6.219 (2.671‐16.984) 7.738 (3.064‐23.172)

Weekly smartphone use (hours) — 1.099 (0.951‐1.256) 1.120 (0.959‐1.295) — 1.139 (0.961‐1.327) 1.184 (0.990‐1.398)

 0‐14 18/189 Ref Ref 13/189 Ref Ref

 15‐28 6/87 0.704 (0.247‐1.749) 0.711 (0.238‐1.880) 2/87 0.319 (0.049‐1.187) 0.362 (0.053‐1.471)

 29‐49 5/133 0.371 (0.120‐0.958) 0.378 (0.119‐1.012) 2/133 0.207 (0.032‐0.765) 0.221 (0.033‐0.868)

 50+ 18/128 1.555 (0.772‐3.133) 1.687 (0.795‐3.597) 14/128 1.663 (0.751‐3.708) 1.979 (0.830‐4.795)

Smartphone use after waking (minutes) — 1.143 (1.004‐1.296) 1.171 (1.000‐1.349) — 1.123 (0.956‐1.279) 1.207 (0.991‐1.433)

 0‐1 8/152 Ref Ref 3/152 Ref Ref

 2‐10 9/133 1.306 (0.485‐3.578) 1.399 (0.498‐4.011) 7/133 2.759 (0.750‐13.006) 3.375 (0.840‐17.270)

 11‐60 16/187 1.694 (0.723‐4.284) 1.637 (0.667‐4.303) 13/187 3.732 (1.176‐16.503) 4.833 (1.375‐23.253)

 61+ 14/65 4.941 (1.998‐13.025) 4.758 (1.736‐13.773) 8/65 6.971 (1.941‐32.674) 10.902 (2.606‐58.482)

Bedtime smartphone use (minutes) — 1.248 (1.079‐1.437) 1.238 (1.046‐1.461) — 1.294 (1.092‐1.520) 1.361 (1.109‐1.666)

 0‐30 7/170 Ref Ref 4/170 Ref Ref

 31‐60 13/148 2.242 (0.892‐6.114) 1.964 (0.746‐5.548) 7/148 2.060 (0.609‐7.999) 2.386 (0.665‐9.938)

 61‐90 8/84 2.451 (0.850‐7.228) 2.073 (0.685‐6.380) 7/84 3.773 (1.106‐14.760) 4.116 (1.115‐17.548)

 91+ 19/135 3.814 (1.618‐10.035) 3.716 (1.467‐10.377) 13/135 4.422 (1.523‐15.999) 5.141 (1.568‐20.808)

aContinuous exposure variables were SAS-SV total score, weekly smartphone use, smartphone use after waking, and bedtime smartphone use. For continuous exposures, ORs represent the change in odds per 1-point increase in SAS-SV total score, per 14-hour increase in weekly smartphone use, and per 30-minute increase in smartphone use after waking and bedtime smartphone use. A total SAS-SV score of 33 or higher is defined as smartphone addiction. Categorical smartphone use variables used the first category as the reference (Ref).

bPHQ-9: Patient Health Questionnaire-9.

cGAD-7: Generalized Anxiety Disorder-7.

dModel 1: crude model.

eOR: odds ratio.

fModel 2: adjustment for maternal age, maternal educational attainment, paternal educational attainment, prepregnancy smoking status, prepregnancy alcohol consumption, frequency of physical exercise during pregnancy, frequency of social media posts on platforms, prepregnancy BMI, weight gain during pregnancy, gravida, abortus, gestational age, and gestational diabetes mellitus.

gSAS-SV: Smartphone Addiction Scale-Short Version.

hNot available.

iRef: reference.

RCS analyses showed that bedtime smartphone use and SAS-SV score had monotonic positive associations with both outcomes, without statistically significant evidence of nonlinearity. Smartphone use after waking showed a U-shaped association with depression (P for nonlinearity=.01) and a J-shaped association with anxiety (P for nonlinearity=.009). Weekly smartphone use showed a U-shaped association with anxiety (P for nonlinearity=.01), whereas its nonlinear association with depression was not statistically significant (Figures 2 and 3).

Furthermore, the independent association between sleep quality and prenatal affective disorders was evaluated, with the results shown in Table 3. In the fully adjusted model, when the PSQI score was analyzed as a continuous variable, each 1-point increase in the score was associated with significantly higher odds of prenatal depression (OR 1.243, 95% CI 1.130‐1.373) and anxiety (OR 1.220, 95% CI 1.087‐1.374). The analysis using a categorical variable yielded consistent results: pregnant women with sleep disturbance (PSQI ≥7) had 2.732 times the odds of depression (OR 2.732, 95% CI 1.389‐5.592) and 2.702 times the odds of anxiety (OR 2.702, 95% CI 1.174‐6.657) compared with those with normal sleep. RCS analyses revealed that the PSQI score was positively associated with both prenatal depression and anxiety, although the nonlinear trends did not reach statistical significance (Figure 2E and Figure 3E).

Table 3.Association between sleep quality and prenatal affective disordersa.

Exposure Prenatal depression (PHQ-9b score≥10) Prenatal anxiety (GAD-7c score≥10)

Model 1d, ORe (95% CI) Model 2f, OR (95% CI) Model 1d, OR (95% CI) Model 2f, OR (95% CI)

PSQIg total score 1.247 (1.142‐1.367) 1.243 (1.130‐1.373) 1.237 (1.115‐1.377) 1.220 (1.087‐1.374)

Poor sleep quality

 No Ref Ref Ref Ref

 Yes 2.706 (1.453‐5.259) 2.732 (1.389‐5.592) 3.029 (1.410‐7.055) 2.702 (1.174‐6.657)

aFor the continuous exposure variable (PSQI total score), the effect size reflects the change in the odds associated with a 1-point increase in the PSQI total score. A total PSQI score of 7 or higher is defined as indicating poor sleep quality.

bPHQ-9: Patient Health Questionnaire-9.

cGAD-7: Generalized Anxiety Disorder-7.

dModel 1: crude model.

eOR: odds ratio.

fModel 2: adjustment for maternal age, maternal educational attainment, paternal educational attainment, prepregnancy smoking status, prepregnancy alcohol consumption, frequency of physical exercise during pregnancy, frequency of social media posts on platforms, prepregnancy BMI, weight gain during pregnancy, gravida, abortus, gestational age, gestational diabetes mellitus, and weekly smartphone use.

gPSQI: Pittsburgh Sleep Quality Index.

Mediating Effect Analysis

In this study, sleep quality was examined as a potential mediator in the associations between multidimensional smartphone use and prenatal affective disorders (Figures 4 and 5). Exploratory mediation analyses showed that the indirect pathways via sleep quality were statistically significant for the associations of bedtime smartphone use, weekly smartphone use, and smartphone addiction with prenatal depression, with mediation proportions of 44.89% (0.0065/0.0145; P<.001; Figure 4C), 31.30% (0.0056/0.0183; P=.04; Figure 4D), and 31.48% (0.0415/0.1315; P<.001; Figure 4A), respectively. Similar indirect patterns were observed in the anxiety models, with corresponding mediation proportions of 35.83% (0.0044/0.0123; P<.001; Figure 5C), 27.05% (0.0039/0.0144; P=.04; Figure 5D), and 22.73% (0.0271/0.1187; P<.001; Figure 5 B1). However, the indirect pathways from smartphone use after waking to depression and anxiety through sleep quality did not reach statistical significance (P=.12; Figure 4B and Figure 5B). Sensitivity mediation analyses using binary outcomes (PHQ-9 or GAD-7≥10) showed that the direction of indirect effects across pathways remained consistent with the main analysis, supporting the robustness of our mediation findings (Figures S1 and S2 in Multimedia Appendix 1).

Discussion

Principal Findings

This study systematically explored the associations of multidimensional smartphone use characteristics with prenatal depression and anxiety, and quantitatively validated the mediating role of sleep quality in these associations. The results showed that smartphone addiction and excessive use after waking and before bedtime were significantly associated with higher odds of prenatal affective disorders, and these associations remained robust after adjustment for multidimensional covariates. RCS analyses further revealed distinct dose–response relationships across different exposure dimensions: smartphone use after waking showed U-shaped and J-shaped associations with prenatal depression and anxiety, respectively; bedtime smartphone use and SAS-SV score showed monotonic positive associations. In addition, sleep quality partially mediated the associations of bedtime use, total weekly use, and smartphone addiction with prenatal depression and anxiety, with the mediation proportions varying across models.

The observed positive associations between smartphone addiction and prenatal depression and anxiety are consistent with findings from adolescents and college students [60-62]. In a meta-analysis encompassing 17 studies and 24,019 college students, Tang et al [63] reported that the pooled OR for the association between problematic smartphone use and depressive symptoms was 2.40 (95% CI 2.19‐2.63), which is lower than the estimate found in this study among pregnant women. This discrepancy suggests that pregnant women may be more susceptible to the adverse mental health effects of smartphone addiction. The potential mechanisms may involve 2 levels. From a neuroendocrine perspective, the dramatic fluctuations in pregnancy-related hormones such as estradiol, progesterone, and cortisol have already rendered the hypothalamic-pituitary-adrenal axis and the emotional regulation circuits of the limbic system in a fragile state of equilibrium; the withdrawal symptoms accompanying smartphone addiction may constitute an additional stress load that further disrupts an emotional homeostasis already operating at a compensatory margin [64,65]. From a sociobehavioral perspective, individuals with addiction tendencies tend to allocate a large amount of time to virtual social environments, thereby encroaching on the time budget that could otherwise be devoted to seeking family emotional support, offline health care visits, and self-care behaviors in the real world, consequently exacerbating the sense of social isolation that is characteristic of the prenatal period [66].

Notably, the associations of time-specific exposure demonstrated marked heterogeneity in this study. In the logistic regression analyses, bedtime smartphone use exceeding 91 minutes and smartphone use after waking exceeding 61 minutes were both significantly associated with higher odds of prenatal depression and anxiety. Previous studies have confirmed that bedtime and nocturnal smartphone use are associated with increased blue light exposure, which in turn has been linked to suppression of melatonin secretion and circadian phase delay [67-69]. However, the underlying mechanisms of digital exposure during the period after waking remain unclear. One possible explanation is that exposure to information on smartphones immediately after waking, such as work notifications and social media updates, may be associated with additional morning psychosocial stressors, which could in turn be related to fluctuations in daytime emotional states. Previous literature has suggested an association between smartphone use and alterations in neuroendocrine function, such as the cortisol awakening response [70]; however, as this study did not directly measure cortisol or other biomarkers, such pathways remain to be verified in future research. Therefore, the mechanisms linking smartphone use after waking to prenatal affective disorders should be interpreted with caution.

In this study, pregnant women with either very low or very high levels of smartphone use exhibited relatively higher odds of depression and anxiety, whereas those with a weekly use duration of 29‐49 hours had relatively lower effect estimates. This nonlinear pattern is similar to the U-shaped trend previously observed in the association between social media use and mental health [20]. From a clinical perspective, the use duration within this range may reflect pregnant women’s active needs for accessing pregnancy-related health information and maintaining social connections, rather than pathological overreliance [10]. Previous studies have suggested that daily smartphone use exceeding 4 hours is significantly associated with prenatal anxiety [12], which appears to contrast with the findings of this study. This discrepancy may indicate that the relationship between smartphone duration and mental health is moderated by multiple factors such as purpose of use, content, and individual susceptibility, rather than being a simple linear accumulation. In addition, the wide CIs for this point estimate and the inability to completely rule out residual confounding warrant a cautious interpretation of the findings.

The RCS analyses further revealed differences in nonlinear patterns across exposure dimensions. The dose-response relationship for duration of smartphone use after waking could not be adequately described by a simple monotonic linear function, with curves showing J-shaped and U-shaped patterns for anxiety and depression, respectively. The association between weekly total use duration and anxiety exhibited a U-shaped curve, whereas its nonlinear trend with depression was not statistically significant, suggesting that the between-group differences observed in Table 2 may partly arise from fluctuations introduced by the categorization strategy. Both duration of bedtime smartphone use and smartphone addiction score showed monotonically increasing associations with depression and anxiety, suggesting that higher exposure levels or greater addiction severity may be associated with higher odds of these outcomes. Thus, the adverse mental health correlates of smartphone use are not solely dependent on cumulative screen time but exhibit notable time-specific differences. Compared with focusing only on total weekly use duration, precise identification of and intervention on screen exposure during critical circadian windows may hold greater clinical value.

Exploratory mediation analyses further indicated that sleep quality partially accounted for the associations between multidimensional smartphone use and prenatal affective disorders, which is consistent with the mechanisms previously observed in nonpregnant populations [20,26,27]. Previous studies have demonstrated that excessive use of electronic devices such as smartphones is associated with increased blue light exposure, which in turn has been shown to suppress the normal secretory rhythm of pineal melatonin, prolong sleep onset latency, and delay circadian phase [67-69]. In addition, research has suggested that cognitive and emotional arousal triggered by immersive screen exposure before bedtime may interfere with sleep initiation and maintenance [29]. For pregnant women who already experience physiological sleep fragmentation, these effects may be further amplified. Previous research has also indicated that sleep deprivation is associated with impaired prefrontal cortex function and increased amygdala reactivity, which may weaken prefrontal-amygdala connectivity and contribute to emotion dysregulation [34,71]; insufficient sleep may also be associated with higher odds of depression and anxiety through activation of the sympathetic nervous system and inflammatory signaling pathways [72,73]. In this study, the mediation proportions of sleep quality in the associations of bedtime use, weekly use, and smartphone addiction with prenatal depression were 44.89% (0.0065/0.0145), 31.30% (0.0056/0.0183), and 31.48% (0.0415/0.1315), respectively; for prenatal anxiety, the corresponding proportions were 35.83% (0.0044/0.0123), 27.05% (0.0039/0.0144), and 22.73% (0.0271/0.1187). These findings suggest that sleep quality may represent a relevant pathway in these associations. Notably, the indirect pathways through sleep quality for smartphone use after waking were not statistically significant for either depression or anxiety. The specificity of these patterns suggests that the associations between digital device use at different times of day and prenatal mental health may operate through distinct biobehavioral pathways.

The association between prolonged and time-specific smartphone use and higher odds of prenatal affective disorders underscores the urgency of digital health management during pregnancy as an important public health issue. In this study, RCS analyses revealed a nonlinear association between smartphone use after waking and prenatal affective disorders, and identified a U-shaped pattern in the relationship between weekly smartphone use and affective disorders. This finding is consistent with the U-shaped association between social media use and mental health observed by Singh et al [74] in adolescent populations, suggesting that moderate smartphone use may be associated with lower emotional distress, whereas excessive use may be related to poorer mental health. These findings provide preliminary evidence for the development of prenatal digital health guidelines. These findings suggest that prenatal mental health interventions should not simply advocate a blanket reduction in smartphone use, but should instead shift toward more precise risk-stratification strategies, such as prioritizing screening for smartphone addiction tendencies and targeting screen exposure behaviors during high-risk periods, specifically after waking and before bedtime. Furthermore, the mediation proportion of sleep quality suggests that improving sleep may be a key intervention target for interrupting the pathway linking digital exposure to affective disorders. Cognitive behavioral interventions targeting bedtime smartphone use behavior may achieve the dual benefits of improving sleep and stabilizing mood simultaneously.

This study has several limitations. First, the study was conducted at the Hunan Provincial Maternal and Child Health Care Hospital, and the study population was limited to permanent residents who completed their entire prenatal care at that hospital; nearly 90 percent (475/537) of them held a college degree or higher, which may have influenced smartphone usage patterns: pregnant women with higher educational attainment may be more inclined to use smartphones to access prenatal health information, but they may also face a higher cognitive load from work-related information processing. Their smartphone usage behavior may reflect a dual characteristic of both health promotion and overreliance, which in turn limits the generalizability of this study’s conclusions to pregnant women with lower educational levels or those in rural areas. Second, the cross-sectional design limits our ability to infer causality; although the pathways in the mediation analysis were established based on evidence from previous longitudinal studies and biological plausibility, we cannot rule out the possibility of reverse pathways. Pregnant women may increase their smartphone use, particularly during periods of nighttime wakefulness, to pass the time or alleviate distress due to preexisting poor sleep quality or symptoms of depression and anxiety. This reverse causality may violate the temporal sequence assumption of the mediation model and could introduce bias into the estimates of indirect effects.

Furthermore, assessments of smartphone usage duration during specific time periods, smartphone addiction, and sleep quality relied on self-report questionnaires, which may introduce recall bias. In this study, total weekly usage duration was objectively recorded using the screen time feature built into the smartphone operating system; however, for usage durations after waking and before bedtime, self-reporting was used as an alternative because the existing operating system could not export data for specific time periods separately. This inconsistency in measurement methods may have some impact on the comparability of the results. Additionally, although we adjusted for several key covariates in our models, some theoretically important confounders (such as caffeine intake, dietary habits, sleep patterns, and occupation type) were not included in the analysis; their omission may lead to residual confounding, which could affect the precision of the effect estimates. Although this study included gestational diabetes mellitus as a covariate to control for the potential impact of common pregnancy complications, other unmeasured complications and lifestyle factors may still constitute residual confounding. However, given that the primary objective of this study was to explore the direction of the association and dose-response trends between smartphone usage patterns and prenatal affective disorders, the impact of these omissions on the directionality and statistical significance of the core conclusions is likely limited; nevertheless, more comprehensive covariate control would help yield more accurate effect estimates. Given the limited sample size in the anxiety group, the CIs for some estimates are relatively wide, indicating limited precision; therefore, the results should be interpreted with caution. Future large-scale, multicenter, prospective longitudinal cohort studies are needed to integrate objective data on screen time tracking and sleep monitoring, comprehensively collect the aforementioned covariates, and explore mediating mechanisms within a longitudinal research framework to further validate the findings of this study.

Conclusion

The results of this study indicate that smartphone addiction and excessive smartphone use after waking and before bedtime were significantly associated with elevated odds of prenatal depression and anxiety, and distinct dose-response patterns exist across different exposure-outcome combinations. Additionally, exploratory mediation analyses indicated that sleep quality may statistically account for part of these associations. Future longitudinal studies are needed to establish temporal precedence and to test the directional assumptions underlying the mediation model.

Supplementary material

Acknowledgments

The authors express their gratitude to all the participants and staff involved in this study. Their efforts and contributions are vital to the success of the research.

Abbreviations

GAD-7Generalized Anxiety Disorder-7

ORodds ratio

PHQ-9Patient Health Questionnaire-9

PSQIPittsburgh Sleep Quality Index

RCSrestricted cubic spline

SAS-SVSmartphone Addiction Scale-Short Version

Footnotes

Funding: The study was funded through grants from Hunan Province Health Research Project (grant number 20255507, 20254300), the Natural Science Foundation of Hunan Province (2025JJ80661), the National Clinical Key Specialty Cultivation Project (Nursing Key Specialty), and the National Natural Science Foundation of China (Grant No. 42407585).

Conflict of Interest Disclosures:
:The authors have no conflicts of interest to declare.

Data Availability: The datasets generated and analyzed during this study are not publicly available due to their containing sensitive information about participants’ social media use and mental health, and to protect participant confidentiality as stipulated by the ethics approval. However, deidentified data underlying the findings presented in this study are available from the corresponding author (MYW) upon reasonable request. Requestors will need to submit a methodologically sound proposal and sign a data access agreement.

Authors’ Contributions: Conceptualization: XXZ, MXOY, XLW, XHY, XYW, YXQC, MMX, YLS, MYW

Data curation: MXOY, XLW, MMX, YLS, MYW

Formal analysis: XXZ, MXOY, MMX, YLS, MYW

Funding acquisition: MXOY, MMX, XLW, YLS, MYW

Investigation: MXOY, XLW, MMX

Methodology: XXZ, YLS, MYW

Project administration: MXOY, XLW, MMX, YLS, MYW

Resources: MXOY, XLW, MMX, YLS, MYW

Software: XXZ, MXOY, MMX, YLS, MYW

Supervision: MXOY, MMX, YLS, MYW

Validation: XXZ, MXOY, XLW, XHY, XYW, YXQC, MMX, YLS, MYW

Visualization: XXZ, MYW

Writing–original draft: XXZ, MYW

Writing–review & editing: XXZ, MXOY, MMX, YLS, MYW

Authors XZ and MO are co-first authors and contributed equally to this manuscript.

MX, YS, and WY are co-corresponding authors of this manuscript and can be reached at:

MX: 1183559609@qq.com; YS: shushu_1012@163.com

Conflicts of Interest: None declared.

References

1.Howard LM, Khalifeh H. Perinatal mental health: a review of progress and challenges. World Psychiatry. 2020 Oct;19(3):313–327. doi: 10.1002/wps.20769. doi. Medline.

2.Dagher RK, Bruckheim HE, Colpe LJ, Edwards E, White DB. Perinatal depression: challenges and opportunities. J Womens Health (Larchmt) 2021 Feb;30(2):154–159. doi: 10.1089/jwh.2020.8862. doi. Medline.

3.Lim G. Perinatal depression. Curr Opin Anaesthesiol. 2021 Jun 1;34(3):233–237. doi: 10.1097/ACO.0000000000000998. doi. Medline.

4.McCarthy M, Houghton C, Matvienko-Sikar K. Women’s experiences and perceptions of anxiety and stress during the perinatal period: a systematic review and qualitative evidence synthesis. BMC Pregnancy Childbirth. 2021 Dec 6;21(1):811. doi: 10.1186/s12884-021-04271-w. doi. Medline.

5.Matvienko-Sikar K, Dijk WV, Dockray S, Leahy-Warren P. Modifiable and vulnerability factors for maternal stress and anxiety in the first 1000 days: an umbrella review and framework. Women Birth. 2025 Jul;38(4):101941. doi: 10.1016/j.wombi.2025.101941. doi. Medline.

6.Al-Abri K, Edge D, Armitage CJ. Prevalence and correlates of perinatal depression. Soc Psychiatry Psychiatr Epidemiol. 2023 Nov;58(11):1581–1590. doi: 10.1007/s00127-022-02386-9. doi. Medline.

7.Yang L, Sun J, Nan Y, et al. Prevalence of perinatal anxiety and its determinants in mainland China: a systematic review and meta-analysis. J Affect Disord. 2023 Feb 15;323:193–203. doi: 10.1016/j.jad.2022.11.075. doi. Medline.

8.Sakowicz A, Allen E, Alvarado-Goldberg M, Grobman WA, Miller ES. Association between antenatal depression symptom trajectories and preterm birth. Obstet Gynecol. 2023 Apr 1;141(4):810–817. doi: 10.1097/AOG.0000000000005125. doi. Medline.

9.Gelaye B, Rondon MB, Araya R, Williams MA. Epidemiology of maternal depression, risk factors, and child outcomes in low-income and middle-income countries. Lancet Psychiatry. 2016 Oct;3(10):973–982. doi: 10.1016/S2215-0366(16)30284-X. doi. Medline.

10.Aftab N. Social media use during pregnancy: risks and benefits for women’s health. Cureus. 2025 Aug;17(8):e90410. doi: 10.7759/cureus.90410. doi. Medline.

11.Costantino C, Mazzucco W, Bonaccorso N, et al. A cross-sectional study on smartphone uses among pregnant women attending childbirth classes in the metropolitan area of Palermo, Italy: the stop-phone study. Ann Ig. 2023;35(3):319–330. doi: 10.7416/ai.2022.2543. doi. Medline.

12.Guo T, Yu Y, Wang X, Sun Y, Yang C. Correlates of antenatal anxiety: smartphone use, depressive symptoms, and hypertensive disorders in a cross-sectional study in Southwest China. Front Med (Lausanne) 2025;12:1682499. doi: 10.3389/fmed.2025.1682499. doi. Medline.

13.Loleska S, Pop-Jordanova N. Is smartphone addiction in the younger population a public health problem? Pril (Makedon Akad Nauk Umet Odd Med Nauki) 2021 Dec 30;42(3):29–36. doi: 10.2478/prilozi-2021-0032. doi. Medline.

14.Lin CY, Ratan ZA, Pakpour AH. Collection of smartphone and internet addiction. BMC Psychiatry. 2023 Jun 14;23(1):427. doi: 10.1186/s12888-023-04915-5. doi. Medline.

15.Ye XL, Zhang W, Zhao FF. Depression and internet addiction among adolescents: a meta-analysis. Psychiatry Res. 2023 Aug;326:115311. doi: 10.1016/j.psychres.2023.115311. doi. Medline.

16.Lopes LS, Valentini JP, Monteiro TH, et al. Problematic social media use and its relationship with depression or anxiety: a systematic review. Cyberpsychol Behav Soc Netw. 2022 Nov;25(11):691–702. doi: 10.1089/cyber.2021.0300. doi. Medline.

17.Nikolic A, Bukurov B, Kocic I, et al. Smartphone addiction, sleep quality, depression, anxiety, and stress among medical students. Front Public Health. 2023;11:1252371. doi: 10.3389/fpubh.2023.1252371. doi. Medline.

18.Kang Y, Liu S, Yang L, et al. Testing the bidirectional associations of mobile phone addiction behaviors with mental distress, sleep disturbances, and sleep patterns: a one-year prospective study among Chinese college students. Front Psychiatry. 2020;11:634. doi: 10.3389/fpsyt.2020.00634. doi. Medline.

19.Wang Y, Luo Y, Yan N, et al. The longitudinal relationships between problematic smartphone use and anxiety symptoms among Chinese college students: a cross-lagged panel network analysis. Addict Behav. 2025 Jan;160:108170. doi: 10.1016/j.addbeh.2024.108170. doi. Medline.

20.Zou Q, Qin J, Zhang L, et al. Time spent on social media applications in relation to depressive symptoms during emerging adulthood and the mediating role of sleep quality: cross-sectional observational study. J Med Internet Res. 2025 Dec 19;27:e75337. doi: 10.2196/75337. doi. Medline.

21.Xiao Y, Meng Y, Brown TT, Keyes KM, Mann JJ. Addictive screen use trajectories and suicidal behaviors, suicidal ideation, and mental health in US youths. JAMA. 2025 Jul 15;334(3):219–228. doi: 10.1001/jama.2025.7829. doi. Medline.

22.Geoffroy PA, Maruani J. Chronobiology of mood disorders: the role of the biological clock in depression and bipolar disorder. Biol Psychiatry. 2025 Dec 1;98(11):830–841. doi: 10.1016/j.biopsych.2025.05.005. doi. Medline.

23.Wang XL, Serchov T. Rhythms of mood: bidirectional regulation of molecular clocks and neurotransmission in affective disorders. Neurosci Biobehav Rev. 2026 Jul;186:106700. doi: 10.1016/j.neubiorev.2026.106700. doi. Medline.

24.Zhong C, Masters M, Donzella SM, Diver WR, Patel AV. Electronic screen use and sleep duration and timing in adults. JAMA Netw Open. 2025 Mar 3;8(3):e252493. doi: 10.1001/jamanetworkopen.2025.2493. doi. Medline.

25.Höhn C, Schmid SR, Plamberger CP, et al. Preliminary results: the impact of smartphone use and short-wavelength light during the evening on circadian rhythm, sleep and alertness. Clocks Sleep. 2021 Jan 22;3(1):66–86. doi: 10.3390/clockssleep3010005. doi. Medline.

26.Lima Santos JP, Soehner AM, Biernesser CL, Ladouceur CD, Versace A. Role of sleep and white matter in the link between screen time and depression in childhood and early adolescence. JAMA Pediatr. 2025 Sep 1;179(9):1000–1008. doi: 10.1001/jamapediatrics.2025.1718. doi. Medline.

27.Feng Z, Diao Y, Ma H, et al. Mobile phone addiction and depression among Chinese medical students: the mediating role of sleep quality and the moderating role of peer relationships. BMC Psychiatry. 2022 Aug 23;22(1):567. doi: 10.1186/s12888-022-04183-9. doi. Medline.

28.Dresp-Langley B, Hutt A. Digital addiction and sleep. Int J Environ Res Public Health. 2022 Jun 5;19(11):6910. doi: 10.3390/ijerph19116910. doi. Medline.

29.León Méndez M, Padrón I, Fumero A, Marrero RJ. Effects of internet and smartphone addiction on cognitive control in adolescents and young adults: a systematic review of fMRI studies. Neurosci Biobehav Rev. 2024 Apr;159:105572. doi: 10.1016/j.neubiorev.2024.105572. doi. Medline.

30.Mislu E, Kumsa H, Tadesse S, et al. Sleep quality disparities in different pregnancy trimesters in low- and middle-income countries: a systematic review and meta-analysis. BMC Pregnancy Childbirth. 2024 Oct 1;24(1):627. doi: 10.1186/s12884-024-06830-3. doi. Medline.

31.Gupta R, Rawat VS. Sleep and sleep disorders in pregnancy. Handb Clin Neurol. 2020;172:169–186. doi: 10.1016/B978-0-444-64240-0.00010-6. doi. Medline.

32.Elahi M, Sanin KI, Rahman MS, et al. Association between mental health and sleep quality among forcibly displaced Rohingya pregnant women living in Cox’s Bazar, Bangladesh: a cross-sectional observation. BMC Psychiatry. 2026 Jan 16;26(1):103. doi: 10.1186/s12888-025-07561-1. doi. Medline.

33.Pauley AM, Moore GA, Mama SK, Molenaar P, Symons Downs D. Associations between prenatal sleep and psychological health: a systematic review. J Clin Sleep Med. 2020 Apr 15;16(4):619–630. doi: 10.5664/jcsm.8248. doi. Medline.

34.Hyndych A, El-Abassi R, Mader ECJ. The role of sleep and the effects of sleep loss on cognitive, affective, and behavioral processes. Cureus. 2025 May;17(5):e84232. doi: 10.7759/cureus.84232. doi. Medline.

35.Sullivan EC, McCall C, Henderson LM, Cairney SA. Disentangling sleep’s role in emotion processing. Neuropsychologia. 2025 Dec 15;219:109269. doi: 10.1016/j.neuropsychologia.2025.109269. doi. Medline.

36.Kwon M, Kim DJ, Cho H, Yang S. The smartphone addiction scale: development and validation of a short version for adolescents. PLoS ONE. 2013;8(12):e83558. doi: 10.1371/journal.pone.0083558. doi. Medline.

37.Luk TT, Wang MP, Shen C, et al. Short version of the Smartphone Addiction Scale in Chinese adults: psychometric properties, sociodemographic, and health behavioral correlates. J Behav Addict. 2018 Dec 1;7(4):1157–1165. doi: 10.1556/2006.7.2018.105. doi. Medline.

38.Lee JJ, Wang MP, Luk TT, Guo N, Chan SSC, Lam TH. Associations of electronic device use before and after sleep with psychological distress among Chinese adults in Hong Kong: cross-sectional study. JMIR Ment Health. 2020 Jun 11;7(6):e15403. doi: 10.2196/15403. doi. Medline.

39.Woo KS, Bong SH, Choi TY, Kim JW. Mental health, smartphone use type, and screen time among adolescents in South Korea. Psychol Res Behav Manag. 2021;14:1419–1428. doi: 10.2147/PRBM.S324235. doi. Medline.

40.Twenge JM, Campbell WK. Associations between screen time and lower psychological well-being among children and adolescents: evidence from a population-based study. Prev Med Rep. 2018 Dec;12:271–283. doi: 10.1016/j.pmedr.2018.10.003. doi. Medline.

41.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 May;28(2):193–213. doi: 10.1016/0165-1781(89)90047-4. doi. Medline.

42.Li C, Cai HB, Zhou Q, Zhang HQ, Wang M, Kang HC. Sleep disorders in the acute phase of coronavirus disease 2019: an overview and risk factor study. Ann Gen Psychiatry. 2023 Jan 31;22(1):3. doi: 10.1186/s12991-023-00431-8. doi. Medline.

43.Huang Y, Jiang ZJ, Deng J, Qi YJ. Sleep quality of patients with postoperative glioma at home. World J Clin Cases. 2020 Oct 26;8(20):4735–4742. doi: 10.12998/wjcc.v8.i20.4735. doi. Medline.

44.Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001 Sep;16(9):606–613. doi: 10.1046/j.1525-1497.2001.016009606.x. doi. Medline.

45.Cheng J, Yu Y, Peng Y, et al. Psychometric validation of the Patient Health Questionnaire-9 for pregnant women: a comprehensive longitudinal study in China. J Affect Disord. 2026 Aug 1;406:121712. doi: 10.1016/j.jad.2026.121712. doi. Medline.

46.Wu D, Chen S, Zhong X, Zhang J, Zhao G, Jiang L. Prevalence and factors associated with antenatal depressive symptoms across trimesters: a study of 110,584 pregnant women covered by a mobile app-based screening programme in Shenzhen, China. BMC Pregnancy Childbirth. 2024 Jul 16;24(1):480. doi: 10.1186/s12884-024-06680-z. doi. Medline.

47.Cheng J, Ren Z, Rehman S, et al. Comparative assessment of Patient Health Questionnaire-9 and Edinburgh Postnatal Depression Scale for screening antenatal depression among Chinese pregnant women. J Affect Disord. 2025 May 1;376:104–112. doi: 10.1016/j.jad.2025.02.004. doi. Medline.

48.Spitzer RL, Kroenke K, Williams JBW, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. 2006 May 22;166(10):1092–1097. doi: 10.1001/archinte.166.10.1092. doi. Medline.

49.Gong Y, Zhou H, Zhang Y, et al. Validation of the 7-item Generalized Anxiety Disorder scale (GAD-7) as a screening tool for anxiety among pregnant Chinese women. J Affect Disord. 2021 Mar 1;282:98–103. doi: 10.1016/j.jad.2020.12.129. doi. Medline.

50.Liu H, Huang F, Gao Y, et al. Network analysis of depression and anxiety symptoms and their associations with cognitive fusion among pregnant women. BMC Psychiatry. 2025 May 26;25(1):537. doi: 10.1186/s12888-025-06978-y. doi. Medline.

51.Mu L, Yu H, Qian G. Associations between physical activity and prenatal depression and anxiety symptoms: a cross-sectional study. Front Public Health. 2025;13:1666312. doi: 10.3389/fpubh.2025.1666312. doi. Medline.

52.Wang R, Cong S, Sha L, et al. Association between social networking site use intensity and depression among Chinese pregnant women: cross-sectional study. J Med Internet Res. 2023 Mar 15;25:e41793. doi: 10.2196/41793. doi. Medline.

53.Yang K, Wu J, Chen X. Risk factors of perinatal depression in women: a systematic review and meta-analysis. BMC Psychiatry. 2022 Jan 27;22(1):63. doi: 10.1186/s12888-021-03684-3. doi. Medline.

54.Herbert D, Young K, Pietrusińska M, MacBeth A. The mental health impact of perinatal loss: a systematic review and meta-analysis. J Affect Disord. 2022 Jan 15;297:118–129. doi: 10.1016/j.jad.2021.10.026. doi. Medline.

55.Toscano M, Royzer R, Castillo D, Li D, Poleshuck E. Prevalence of depression or anxiety during antepartum hospitalizations for obstetric complications: a systematic review and meta-analysis. Obstet Gynecol. 2021 May 1;137(5):881–891. doi: 10.1097/AOG.0000000000004335. doi. Medline.

56.Kang D, Zhao D, Ryu S, et al. Perceived stress and non-alcoholic fatty liver disease in apparently healthy men and women. Sci Rep. 2020 Jan 8;10(1):38. doi: 10.1038/s41598-019-57036-z. doi. Medline.

57.Okosieme OE, Usman D, Taylor PN, et al. Cardiovascular morbidity and mortality in patients in Wales, UK with resistance to thyroid hormone β (RTHβ): a linked-record cohort study. Lancet Diabetes Endocrinol. 2023 Sep;11(9):657–666. doi: 10.1016/S2213-8587(23)00155-9. doi. Medline.

58.Durrleman S, Simon R. Flexible regression models with cubic splines. Stat Med. 1989 May;8(5):551–561. doi: 10.1002/sim.4780080504. doi. Medline.

59.Heinze G, Schemper M. A solution to the problem of separation in logistic regression. Stat Med. 2002 Aug 30;21(16):2409–2419. doi: 10.1002/sim.1047. doi. Medline.

60.Li Y, Li G, Liu L, Wu H. Correlations between mobile phone addiction and anxiety, depression, impulsivity, and poor sleep quality among college students: a systematic review and meta-analysis. J Behav Addict. 2020 Oct 12;9(3):551–571. doi: 10.1556/2006.2020.00057. doi. Medline.

61.Matar Boumosleh J, Jaalouk D. Depression, anxiety, and smartphone addiction in university students- a cross sectional study. PLoS One. 2017;12(8):e0182239. doi: 10.1371/journal.pone.0182239. doi. Medline.

62.Shiferaw BD, Tang J, Wang Y, et al. Impact of digital addiction on youth health: a systematic review and meta-analysis. J Behav Addict. 2025 Sep 30;14(3):1129–1158. doi: 10.1556/2006.2025.00081. doi. Medline.

63.Tang S, Chen R, Ma Z, Li X, Chen J, Zhao J. Associations of problematic smartphone use with depressive symptoms and suicidal ideation in university students before and after the COVID-19 outbreak: a meta-analysis. Addict Behav. 2024 May;152:107969. doi: 10.1016/j.addbeh.2024.107969. doi. Medline.

64.Duthie L, Reynolds RM. Changes in the maternal hypothalamic-pituitary-adrenal axis in pregnancy and postpartum: influences on maternal and fetal outcomes. Neuroendocrinology. 2013;98(2):106–115. doi: 10.1159/000354702. doi. Medline.

65.Zhang C, Xu D, Luo H, et al. Prenatal xenobiotic exposure and intrauterine hypothalamus-pituitary-adrenal axis programming alteration. Toxicology. 2014 Nov 5;325:74–84. doi: 10.1016/j.tox.2014.08.015. doi. Medline.

66.Guo N, Wang MP, Luk TT, et al. The association of problematic smartphone use with family well-being mediated by family communication in Chinese adults: a population-based study. J Behav Addict. 2019 Sep 1;8(3):412–419. doi: 10.1556/2006.8.2019.39. doi. Medline.

67.Dufier JL, Touitou Y. Light pollution. Rev Prat. 2022 Feb;72(2):141–146. Medline.

68.Minich DM, Henning M, Darley C, Fahoum M, Schuler CB, Frame J. Is melatonin the “next vitamin D”?: A review of emerging science, clinical uses, safety, and dietary supplements. Nutrients. 2022 Sep 22;14(19):3934. doi: 10.3390/nu14193934. doi. Medline.

69.Oh JH, Yoo H, Park HK, Do YR. Analysis of circadian properties and healthy levels of blue light from smartphones at night. Sci Rep. 2015 Jun 18;5:11325. doi: 10.1038/srep11325. doi. Medline.

70.Afifi TD, Zamanzadeh N, Harrison K, Acevedo Callejas M. WIRED: The impact of media and technology use on stress (cortisol) and inflammation (interleukin IL-6) in fast paced families. Comput Human Behav. 2018 Apr;81:265–273. doi: 10.1016/j.chb.2017.12.010. doi.

71.Ben Simon E, Vallat R, Barnes CM, Walker MP. Sleep loss and the socio-emotional brain. Trends Cogn Sci. 2020 Jun;24(6):435–450. doi: 10.1016/j.tics.2020.02.003. doi. Medline.

72.Ballesio A. Inflammatory hypotheses of sleep disturbance - depression link: update and research agenda. Brain Behav Immun Health. 2023 Aug;31:100647. doi: 10.1016/j.bbih.2023.100647. doi. Medline.

73.Meerlo P, Sgoifo A, Suchecki D. Restricted and disrupted sleep: effects on autonomic function, neuroendocrine stress systems and stress responsivity. Sleep Med Rev. 2008 Jun;12(3):197–210. doi: 10.1016/j.smrv.2007.07.007. doi. Medline.

74.Singh B, Zhou M, Curtis R, Maher C, Dumuid D. Social media use and well-being across adolescent development. JAMA Pediatr. 2026 Mar 1;180(3):288–297. doi: 10.1001/jamapediatrics.2025.5619. doi. Medline.

Associated Data

Supplementary Materials