Abstract
Objective
The current study aims to explore the relationship between Family function and depression, examining the mediation and moderation effects of pregnancy-related anxiety (PRA) and sleep latency, respectively.
Method
A cross-sectional survey was conducted among pregnant women in the Ningxia Hui Autonomous Region between February 2022 and December 2024, obtaining 7622 valid questionnaires. Model 4 and Model 7 in PROCESS were used to analyse the mediating effect of pregnancy-related anxiety and the moderating effect of sleep latency. A bootstrap method was conducted to test the mediating and moderating effects.
Results
After adjusting for covariates, the mediating effect of PRA explained 18.68% (95% CI: -0.094, -0.06) of the association between Family function and prenatal depression (EPD). The results indicate that the interaction term (Family function * sleep latency) has a negative effect on PRA (β = -0.096, 95% CI: -0.160, -0.032). This implies that the protective effect of good Family function in reducing PRA is weakened as sleep latency increases. Further analysis by different pregnancy periods revealed that this regulatory effect was only statistically significant in the third trimester (β = -0.127, 95% CI: -0.226, -0.028).
Conclusion
This paper proved that the Family function negatively predicted depression, and PRA served as a mediator between Family function and depression during pregnancy. Besides, the protective role of sleep quality on individuals’ mental health was identified.
Clinical trial number
Not applicable.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12888-026-07929-x.
Keywords: Family function, Pregnancy-related anxiety, Depression, Sleep latency, Mediation, Moderation
Introduction
Depression and anxiety are the most common mental disorders experienced during the perinatal period. Previous studies have reported a high prevalence of prenatal depression, ranging from 15% to 65% [1]. An increasing number of studies have shown that anxiety and depression during pregnancy can affect the physical and mental health of both the mothers and the fetus. These conditions are associated with an increased risk of adverse pregnancy outcomes, such as complications during pregnancy, preterm birth, poor fetal development, and low birth weight [2, 3]. Therefore, the pernicious and lasting consequences of prenatal depression for pregnant women and their infants present a major social public health challenge.
The existing literature indicates that the factors associated with negative emotions during pregnancy mainly include sociodemographic characteristics, such as maternal age, adverse obstetric history, educational level, and pregnancy complications, as well as social support and Family function [4–6]. As a protective and changeable factor, Family function has been widely studied in relation to negative emotions during pregnancy. Poor Family function is a key risk factor for anxiety and depression during pregnancy, whereas strong Family function can increase emotional support and communication among family members, thus helping pregnant women manage the stress and challenges of pregnancy more effectively [7]. A cross-sectional study in China found that the risk of prenatal depression symptoms with poor Family function was 3.67 times as much as that in the better-functioning group [8].
Pregnancy-related anxiety (PRA) is a unique psychological stress and physical symptom experienced by pregnant women during pregnancy due to concerns about fetal health, their own health and appearance, as well as medical, economic, and social support that may be involved in childbirth and childrearing [9]. It has been established that perceived social support and Family function are important protective factors for PRA [10]. As the basic unit of society, the family provides its members with support, resources, services, and other forms of assistance. Family function is a key variable in the psychological well-being of its members. Additionally, it should be noted that, in general, depression and anxiety are highly comorbid [3]. From the perspective of causal timing, Family function is formed before pregnancy, while PRA is the immediate psychological state under the pressure of pregnancy, and prenatal depression is the result of long-term psychological imbalance. Therefore, the level of Family function in pregnant women may indirectly affect their depressive symptoms through anxiety symptoms.
Sleep quality plays an important role in the overall health of individuals. The data show that the percentage of pregnant women with poor sleep quality is as high as 76% [11]. Increasing evidence shows that sleep quality during pregnancy has a negative effect on the mental health of pregnant women. Loret de Mola et al. [12] found that sleep latency was more likely to be associated with anxiety and depressive symptoms. Shariat [13] and colleagues reported that both state anxiety and trait anxiety were significantly associated with sleep quality during each pregnancy in a sample of 283 women. As a physiological state index, sleep quality is not directly affected by Family function. Sleep-related indicators are more suitable as regulatory variables, which indirectly affect the psychological state of pregnancy by affecting emotional regulation ability. However, few studies have examined whether sleep latency moderates the association between Family function and pregnancy-related affect.
Therefore, based on the above theoretical analysis and the theoretical model (see Fig. 1), this study aims to explore the mediating role of PRA in the relationship between Family function and prenatal depression. Specifically, we propose the following research hypotheses: (1) Family function is negatively associated with prenatal depression; (2) PRA plays a mediating role in the relationship between Family function and prenatal depression; (3) This study intends to examine the moderating effect of sleep latency on the mediating pathway, we hypothesize that sleep latency moderates the first stage of the mediation process (the link between Family function and PRA), such that the protective effect of good Family function on reducing PRA varies with the length of sleep latency. Ultimately, this study aims to suggest potential targets for future intervention studies for alleviating prenatal depression in a timely manner, thereby better safeguarding the psychological well-being of pregnant women and the health of both mothers and infants.
Fig. 1.
Theoretical model diagram
Method
Participants
A cross-sectional survey was conducted at a tertiary hospital in the Ningxia Hui Autonomous Region, between February 2022 and December 2024, using a continuous sampling method. The inclusion criterion was women aged 18 years or older and no intellectual disability. The exclusion criterion was incomplete questionnaires(> 50% missing data). The survey covered general demographic information, family function, prenatal psychological conditions (including pregnancy-related anxiety and prenatal depression), and sleep quality. A total of 7630 questionnaires were collected. After 8 questionnaires were excluded because they were incomplete or all the answers were the same, 7622 valid questionnaires were obtained (Fig. 2). Prior to the commencement of the research, approval was obtained from the Ningxia Medical University (Grant number: No.2022-G007). Informed consent was obtained from all participants before data collection.
Fig. 2.
Flow chart for participant selection
Demographic and socioeconomic characteristics
The demographic data included the following: age, residence, family per capita monthly income, occupation, educational level, duration of pregnancy (early trimester: <14weeks; second trimester: 15-27weeks; third trimester: >28weeks), character trait (introverted, biased, extroverted), pregnancy intention, only child, conceived way, adverse pregnancy, the reaction degree of early pregnancy(lighter, general, serious), pre-pregnancy body mass index (BMI) (kg/m2) was calculated based on participants’ pre-pregnancy self-reported weight.
Family functions
The Family functions of pregnant women were measured by family adaptation, partnership, growth, affection, and resolve (APGAR), compiled by Smilkstein [14]. The APGAR questionnaire consists of five items: adaptation, partnership, growth, affection, and resolve. Each item corresponds to a topic and is scored on a three-level scale: “often” is scored as 2 points, “sometimes” as 1 point, and “almost never” as 0 points. The total score ranges from 0 to 10 points, with higher scores indicating better Family functioning. In this study, the Cronbach’s α of the scale was 0.897.
Pregnancy-related anxiety
Participants filled out the Pregnancy-related Anxiety Questionnaire (PAQ) developed by Xiao Limin [15] from Anhui Medical University. The scale consists of 13 items, related to self-concern, fetal health concern, and childbirth concern. Scores on each item range from 1= “not worried” to 4= “always worried”, and the range of the total sum score is 13 to 52. For the current study, the Cronbach’s α of the scale was 0.732.
Prenatal depression
Depressive symptoms were assessed using a 10-item self-reported Edinburgh Postnatal Depression Scale (EPDS) [16], on a 4-point Likert scale. The total scores range between 0 and 30 points, with a score of more than 12 is a commonly used cutoff value for screening possible depression [17]. In this study, the Cronbach’s α of the scale was 0.763.
Sleep quality
The Pittsburgh Sleep Quality Index (PSQI) was used to assess individuals’ subjective sleep quality during the past month. The scale was developed by Buysse [18] in 1989. The PSQI consists of a total of 24 questions, of which 19 questions were independently assessed by the respondents in combination with their own conditions, and the remaining 5 questions were assessed by their bedmates or roommates. In this study, we mainly focused on 19 self-assessment questions directly related to each dimension of sleep quality, which are divided into seven parts, including seven components: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medications, and daytime dysfunction, with each component is scored from 0 (no difficulty) to 3 (severe difficulty). The PSQI total score ranges from 0 to 21, where higher scores reflect poorer sleep quality. In this study, the Cronbach’s α of the scale was 0.700.
Data analysis
The data in this study were double-entered via Epi Data 3.1 software. SPSS 23.0 and R 4.3.3 were used for statistical comparisons and analyses. The demographic characteristics and the measurement data were calculated via means ± standard deviations(SDs) or numbers(percentages). For comparisons between groups, the study adopted independent samples t-tests and analysis of variance to test the statistical significance of continuous variables. Pearson’s correlation analyses were also conducted to preliminarily assess the correlation among Family function, sleep quality, PRA, and prenatal depression. Model 4 and Model 7 in the SPSS macro program PROCESS v3.4.1 compiled by Hayes [19, 20] were used to test the moderated mediating effect. Specifically, Model 4 was selected to test the mediating role of PRA in the relationship between family function and prenatal depression. Model 7 was adopted to examine the moderated mediating effect of sleep latency, given its suitability for testing models where the moderator acts on the first stage of the mediation pathway. A bootstrap method with 5000 resamples was used to calculate the 95% CIs for the effects, ensuring the robustness of the mediation and moderation results.
The mediating effect model includes unadjusted and adjusted covariate models. With EPD as the final outcome, the adjusted covariates were statistically significant factors for EPD after univariate analysis, including residence, education, occupation, character trait, only-child, pregnancy intention, family per capita monthly income, adverse pregnancy, the reaction degree of early pregnancy, pre-pregnancy physical condition and fetal gender expectation.
Results
Participant characteristics
The socio-demographic profile of the study population is detailed in Table 1. More than half of the participants (67.07%) had a college degree or above. Few pregnant women were advanced-aged pregnant women (age > 34 years). The majority of them live in urban settings(83.84%). In terms of gestational week, 1230 participants (16.14%) were in early pregnancy, 2845 (37.33%) were in late pregnancy, and 3547 (46.54%) were in mid-pregnancy. Among the participants, 5238 (68.72%) had a pre-pregnancy BMI within the normal range, whereas 1473 (19.32%) were classified as overweight or obese.
Table 1.
Demographic profile of the participants (N = 7622)
| Variables | N | % | N | % | |||
|---|---|---|---|---|---|---|---|
| Age | ≤ 34 | 6457 | 84.72 | Conceived way | natural | 7108 | 93.26 |
| > 34 | 1165 | 15.28 | artificial assisted pregnancy | 514 | 6.74 | ||
| Residence | urban | 6390 | 83.84 | Adverse pregnancy | No | 4590 | 60.22 |
| rural | 1232 | 16.16 | Yes | 3032 | 39.78 | ||
| Education | middle school and below | 950 | 12.46 | The reaction degree of early pregnancy | Lighter | 3010 | 39.49 |
| high school or junior college | 1160 | 15.22 | General | 3136 | 41.14 | ||
| college or Bachelor’s | 5112 | 67.07 | Serious | 1476 | 19.36 | ||
| postgraduate and above | 400 | 5.25 | Duration of pregnancy | < 14 weeks | 1230 | 16.14 | |
| Occupation | unemployed | 1886 | 24.74 | 15–27 weeks | 3547 | 46.54 | |
| farmers or workers | 302 | 3.96 | ≥ 28 week | 2845 | 37.33 | ||
| enterprise | 2861 | 37.54 | Pre-pregnancy BMI(kg/m2) | < 18.5 | 911 | 68.72 | |
| others | 2573 | 33.76 | 18.5–24.9 | 5238 | 11.95 | ||
| character trait | introverted | 975 | 12.79 | 25.0-29.9 | 1203 | 15.78 | |
| moderate | 5100 | 66.91 | ≥ 30 | 270 | 3.54 | ||
| extrovert | 1547 | 20.30 | Pre-pregnancy physical condition | Better | 5095 | 66.85 | |
| Only one child | Yes | 1088 | 14.27 | General | 2418 | 31.72 | |
| No | 6534 | 85.73 | Poor | 109 | 1.43 | ||
| pregnancy intention | Planned Pregnancy | 3025 | 39.69 | Fetal gender expectation | No | 6340 | 83.18 |
| Let nature take its course | 3559 | 46.69 | Yes | 1282 | 16.82 | ||
| Accidental pregnancy | 1038 | 13.62 | |||||
| Family per capita monthly income | <¥2000 | 474 | 6.22 | ||||
| ¥2000~¥5000 | 3949 | 51.81 | |||||
| >¥5000 | 3197 | 41.94 |
Univariate analysis of family function, pregnancy-related anxiety, and depression
The distribution of Family function, PRA, and EPD for the whole pregnancy was shown in Fig. 3. Among the 7,622 pregnant women, the depression score was 6.87 ± 3.71. The results of the univariate analysis revealed that the depression scores differed significantly among the groups categorized by inhabitancy, educational level, occupation, character trait, whether being an only child status(only-child), pregnancy intention, family per capita monthly income, adverse pregnancy history, degree of early pregnancy reaction, pre-pregnancy physical condition, and fetal sex expectation (P < 0.05) (Table 2). The score of PRA was 21.07 ± 5.57. The results of the univariate analysis show that age, character trait, pregnancy intention, family per capita monthly income, conceived method, pre-pregnancy physical condition, fetal sex expectation, and duration of pregnancy significantly affected anxiety scores (P < 0.05) (Table 2).
Fig. 3.
The density distribution of family function, PRA and EPD of the study subjects
Table 2.
Associations between sociodemographic factors and Family function, PRA, and EPD
| Variables | Family Functions | PRA | EPD | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Scores (  ) |
F/t | P | ( ) |
F/t | P | Scores (  ) |
F/t | P | |
| Age | 1.906 | 0.057 | 3.141 | 0.002 | 0.693 | 0.488 | |||
| ≤ 34 | 8.32 ± 2.25 | 21.16 ± 5.60 | 6.88 ± 3.71 | ||||||
| > 34 | 8.18 ± 2.35 | 20.60 ± 5.40 | 6.80 ± 3.69 | ||||||
| Residence | 7.992 | < 0.001 | -1.419 | 0.156 | -6.274 | < 0.001 | |||
| urban | 8.39 ± 2.20 | 21.03 ± 5.46 | 6.75 ± 3.69 | ||||||
| rural | 7.82 ± 2.56 | 21.28 ± 6.09 | 7.47 ± 3.77 | ||||||
| Education | 79.655 | < 0.001 | 0.405 | 0.749 | 31.773 | < 0.001 | |||
| middle school and below | 7.49 ± 2.69 | 22.19 ± 6.09 | 7.78 ± 4.06 | ||||||
| high school or junior college | 7.88 ± 2.47* | 22.12 ± 5.84 | 7.21 ± 3.86* | ||||||
| college or Bachelor | 8.48 ± 2.11* | 21.03 ± 5.40 | 6.67 ± 3.56* | ||||||
| postgraduate and above | 9.01 ± 1.73* | 21.24 ± 5.58 | 6.22 ± 3.80* | ||||||
| Occupation | 65.137 | < 0.001 | 39.697 | 0.279 | 17.933 | < 0.001 | |||
| unemployed | 7.87 ± 2.46 | 21.28 ± 5.92 | 7.28 ± 3.84 | ||||||
| farmers or workers | 7.67 ± 2.69 | 21.00 ± 6.06 | 7.53 ± 3.80 | ||||||
| enterprise | 8.71 ± 1.98* | 21.04 ± 5.40 | 6.56 ± 3.66* | ||||||
| others | 8.22 ± 2.28* | 20.96 ± 5.43 | 6.83 ± 3.62* | ||||||
| Character trait | 55.964 | < 0.001 | 26.138 | < 0.001 | 55.612 | < 0.001 | |||
| introverted | 7.61 ± 2.51 | 22.15 ± 6.13 | 7.85 ± 3.97 | ||||||
| moderate | 8.35 ± 2.21* | 21.03 ± 5.38* | 6.87 ± 3.58* | ||||||
| extrovert | 8.54 ± 2.20* | 20.52 ± 5.73* | 6.26 ± 3.82* | ||||||
| Only-child | 7.829 | < 0.001 | -0.581 | 0.561 | -5.584 | < 0.001 | |||
| Yes | 8.79 ± 2.00 | 20.98 ± 5.75 | 6.29 ± 3.74 | ||||||
| No | 8.21 ± 2.30 | 21.09 ± 5.54 | 7.96 ± 3.69 | ||||||
| Pregnancy intention | 37.904 | < 0.001 | 72.094 | < 0.001 | 32.864 | < 0.001 | |||
| Planned Pregnancy | 8.56 ± 2.10 | 20.53 ± 5.29 | 6.51 ± 3.59 | ||||||
| Let nature take its course | 8.16 ± 2.35* | 21.00 ± 5.58* | 6.97 ± 3.71* | ||||||
| Accidental pregnancy | 7.97 ± 2.36* | 22.90 ± 5.95* | 7.55 ± 3.92* | ||||||
| family per capita monthly income | 53.261 | < 0.001 | 6.573 | 0.001 | 25.651 | < 0.001 | |||
| <¥2000 | 7.56 ± 2.72 | 21.97 ± 6.61 | 7.81 ± 4.02 | ||||||
| ¥2000~¥5000 | 8.17 ± 2.32* | 21.01 ± 5.55* | 6.97 ± 3.65* | ||||||
| >¥5000 | 8.56 ± 2.10* | 21.01 ± 5.41* | 6.60 ± 3.71* | ||||||
| Conceived way | -0.408 | 0.683 | 2.078 | 0.038 | -1.500 | 0.880 | |||
| natural | 8.29 ± 2.27 | 21.11 ± 5.59 | 6.87 ± 3.326 | ||||||
| artificial assisted pregnancy | 8.33 ± 2.24 | 21.58 ± 5.30 | 6.89 ± 2.74 | ||||||
| Adverse pregnancy | 4.789 | < 0.001 | -1.454 | 0.146 | -3.938 | < 0.001 | |||
| No | 8.40 ± 2.23 | 21.00 ± 5.57 | 6.73 ± 3.72 | ||||||
| Yes | 8.14 ± 2.31 | 21.18 ± 5.66 | 7.07 ± 3.68 | ||||||
| The reaction degree of early pregnancy | 4.136 | 0.016 | 26.510 | < 0.001 | 24.709 | < 0.001 | |||
| Lighter | 8.33 ± 2.30 | 20.62 ± 5.52 | 6.59 ± 3.70 | ||||||
| General | 8.14 ± 2.22 | 21.11 ± 5.48* | 6.88 ± 3.65* | ||||||
| Serious | 8.09 ± 2.22* | 21.91 ± 5.77* | 7.41 ± 3.79* | ||||||
| duration of pregnancy | 1.219 | 0.295 | 4.573 | 0.010 | 1.428 | 0.240 | |||
| < 14 weeks | 8.38 ± 2.23 | 21.51 ± 5.77 | 6.91 ± 3.85 | ||||||
| 15–27 weeks | 8.27 ± 2.90 | 20.96 ± 5.59* | 6.79 ± 3.72 | ||||||
| ≥ 28 weeks | 8.29 ± 2.25 | 21.02 ± 5.45* | 6.94 ± 3.62 | ||||||
| Pre-pregnancy BMI | 0.680 | 0.564 | 1.132 | 0.335 | 0.345 | 0.793 | |||
| < 18.5 | 8.33 ± 2.24 | 21.27 ± 5.86 | 6.90 ± 3.86 | ||||||
| 18.5–24.9 | 8.31 ± 2.24 | 21.04 ± 5.51 | 6.89 ± 3.67 | ||||||
| 25.0-29.9 | 8.25 ± 2.37 | 20.96 ± 5.56 | 6.78 ± 3.76 | ||||||
| ≥ 30 | 7.14 ± 2.44 | 21.50 ± 5.79 | 6.79 ± 3.66 | ||||||
| Pre-pregnancy physical condition | 76.485 | < 0.001 | 138.200 | < 0.001 | 175.879 | < 0.001 | |||
| Better | 8.51 ± 2.16 | 20.39 ± 5.27 | 6.35 ± 3.56 | ||||||
| General | 7.88 ± 2.40* | 22.30 ± 5.79* | 7.81 ± 3.71* | ||||||
| Poor | 7.28 ± 2.67* | 25.66 ± 7.28* | 10.07 ± 4.83* | ||||||
| fetal gender expectation | 6.660 | < 0.001 | -5.406 | < 0.001 | -5.264 | < 0.001 | |||
| No | 8.37 ± 2.22 | 20.92 ± 5.50 | 6.77 ± 3.68 | ||||||
| Yes | 7.91 ± 2.46 | 21.84 ± 5.83 | 7.36 ± 3.80 | ||||||
The total score for the Family function was 8.29 ± 2.27. Univariate analysis revealed that factors such as habitancy, education, occupation, character traits, only child status, pregnancy intentions, family monthly per capita income, adverse pregnancy, degree of early pregnancy reaction, pre-pregnancy physical condition and fetal sex expectation significantly influenced the family care scores among pregnant women, with statistically significant differences (P < 0.05). The mean sleep quality scores for overall, subjective sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbance, use of sleep medication and the daytime dysfunction were 4.98 ± 3.00, 1.00 ± 0.69, 0.98 ± 0.80, 0.53 ± 0.82, 0.54 ± 0.87, 1.10 ± 0.58, 0.01 ± 0.13, 0.83 ± 0.80, respectively.
Associations between Family function, sleep quality, PRA, and EPD
The correlations between variables were calculated and shown in Fig. 4. Family function levels were negatively associated with PRA (r = -0.179, P < 0.001) and EPD scores(r = -0.304, P < 0.001), suggesting that the higher levels of family dysfunction, the increased likelihood of them suffering from PRA and EPD. Meanwhile, PRA was proven to be significantly positively related to EPD (r = 0.392, P < 0.001), indicating that pregnant women experiencing PRA are more likely to develop depressive symptoms. Additionally, sleep latency was negatively correlated with Family function (r = -0.107, P < 0.001), and positively correlated with PRA (r = 0.178, P < 0.001) and EPD (r = 0.217, P < 0.001). Fig. 4A.
Fig. 4.
Correlation heatmap of Family function, sleep quality, pregnancy-related anxiety, depression and their subscales. A: Total pregnancy: correlations among Family function, sleep quality, pregnancy-related anxiety, depression and their subscales; B: Early pregnancy: correlations among Family function, sleep quality, pregnancy-related anxiety, depression and their subscales; C: Second trimester: correlations among Family function, sleep quality, pregnancy-related anxiety, depression and their subscales; D: Late pregnancy: correlations among Family function, sleep quality, pregnancy-related anxiety, depression and their subscales
Further analysis stratified by different stages of pregnancy, and the results showed that during different pregnancy periods, Family functioning was negatively associated with sleep latency (r = -0.088, P = 0.002; r = -0.092, P < 0.001; r = -0.136 P < 0.001), PRA (r =-0.164 P < 0.001; r = -0.186 P < 0.001; r = -0.177, P < 0.001) and EPD (r =-0.255, P < 0.001; r = -0.306 P < 0.001; r =-0.324, P < 0.001), respectively. Additionally, the correlation analysis results for different stages of pregnancy also showed that the PRA and its various subscales were positively correlated with the EPD (r = 0.185, P < 0.001; r = 0.413, P < 0.001; r = 0.376, P < 0.001) and its respective subscales. Sleep latency was positively correlated with the PRA (r = 0.169, P < 0.001; r = 0.173, P < 0.001; r = 0.192, P < 0.001) and EPD (r = 0.253, P < 0.001; r = 0.193, P < 0.001; r = 0.230, P < 0.001). Fig. 4B-D.
Furthermore, a multiple linear regression model was constructed with EPD as the dependent variable, adjusting for all covariates identified as statistically significant in the univariate analyses. This analysis confirmed that higher levels of PRA (β = 0.204, 95% CI: 0.190, 0.218, P < 0.001) and longer sleep latency (β = 0.593, 95% CI: 0.500, 0.686, P < 0.001) were significantly associated with higher EPD scores, whereas better Family function (β = − 0.331, 95% CI: − 0.364, − 0.297, P < 0.001) was independently associated with lower EPD scores (Fig. S1).
The mediating effect of pregnancy-related anxiety on the relationship between family function and depression during pregnancy
Unadjusted mediation analyses were summarized in Table 3. Family function significantly negatively predicted EPD and PRA, with the total effect(c) was − 0.497 (95% CI: -0.532, -0.462) and the effect (a) was − 0.439 (95%CI: -0.493, -0.384). In addition, the bootstrap 95% CI of the mediating effect (ab) of PRA indicated that PRA explained 20.52%(ab: -0.102, 95% CI: -0.118, -0.086) of the association between Family function and EPD (Fig. 5A; Table 3). The indirect effect of PRA is negative, meaning Family function reduces depression by reducing PRA. Further mediation effect analysis was conducted for different pregnancy periods (including early, middle, and late pregnancy). The results of the unadjusted mediation model analysis showed that the predictive effects of all paths in the mediation model remained significant. The percentages of the mediating effects were 21.16%(ab: -0.095; 95% CI: -0.135, -0.058), 22.47%(ab: -0.111; 95% CI: -0.137, -0.087) and 18.04%(ab: -0.094; 95% CI: -0.119, -0.071), respectively(Fig. 5B, C, D; Table S1, S3, S5).
Table 3.
Testing the mediating role of the PRA in the relationship between Family function and EPD during pregnancy (N = 7622)
| Path | R 2 | F | β | t | LLCI | ULCI | P | Effect (%) |
|
|---|---|---|---|---|---|---|---|---|---|
| Family function(X) - EPD(Y) c | 0.304 | 774.000 | -0.497 | -27.821 | -0.532 | -0.462 | < 0.001 | - | |
| Family function(X) - PRA(M) a | 0.179 | 250.711 | -0.439 | -15.834 | -0.493 | -0.384 | < 0.001 | - | |
| Family function(X) - EPD(Y) c’ | 0.458 | 1012.491 | -0.395 | -23.327 | -0.428 | -0.362 | < 0.001 | 79.48 | |
| PRA(M) -EPD(Y) b | 0.232 | 33.701 | 0.219 | 0.246 | < 0.001 | - | |||
| Family function(X)- PRA(M)-EPD(Y) ab | - | - | -0.102 | - | -0.118 | -0.086 | < 0.001 | 20.52 | |
Fig. 5.
Mediation effect model diagram. A: The overall mediating effect model diagram of unadjusted covariates; B: Mediation effect model diagram of unadjusted covariates in early pregnancy; C: Mediation effect model diagram of unadjusted covariates in the second pregnancy; D: Mediation effect model diagram of unadjusted covariates in the third pregnancy. E: The overall mediating effect model diagram of adjusted covariates. F: Mediation effect model diagram of adjusted covariates in early pregnancy. G: Mediation effect model diagram of adjusted covariates in the second pregnancy. H: Mediation effect model diagram of adjusted covariates in the third pregnancy. The numerical values on the paths represent unstandardized coefficients(β): a = the effect of Family function on PRA; b = the effect of PRA on EPD; ab = indirect effect of Family function on EPD; c = the total effect of Family function on EPD; c’= the direct effect of Family function on EPD after controlling for PRA
After adjusting for the covariates that were statistically significant for EPD after univariate analysis in this study, the predictive effects of all paths in the mediation model remained significant. The total effect(c) of Family function on EPD is -0.423 (95% CI: -0.459, -0.388), the direct effect(c’) is -0.344 (95% CI: -0.378, -0.310), and the direct effect on EPD through PRA(ab) is -0.079 (95% CI: -0.094, -0.065), with the mediating effect accounting for 18.68%(Fig. 5E; Table 4). Further mediation effect analysis was conducted for different stages of pregnancy (including early, middle, and late pregnancy). The results showed that the predictive effects of all paths in the mediation model remained significant, with the PRA explained 19.95%(ab: -0.078; 95% CI: -0.403, -0.224), 21.79%(ab: -0.090; 95% CI: -0.069, -0.041) and 19.95%(ab: -0.070; 95% CI: -0.092, -0.050) of the association between Family function and EPD, respectively (Fig. 5F, G, H; Table S2, S4, S6).
Table 4.
Mediating model test of PRA between family function and EPD during pregnancy after controlling for covariates (N = 7622)
| Path | R 2 | F | β | t | LLCI | ULCI | P | Effect(%) | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Family function(X) - EPD(Y) c | 0.365 | 97.258 | -0.423 | -23.427 | -0.459 | -0.388 | < 0.001 | - | |||||
| Family function(X) - PRA(M) a | 0.274 | 51.343 | -0.366 | -13.050 | -0.421 | -0.311 | < 0.001 | - | |||||
| Family function(X) - EPD(Y) c’ | 0.481 | 176.173 | -0.344 | -19.989 | -0.378 | -0.310 | < 0.001 | 81.32 | |||||
| PRA(M) -EPD(Y) b | 0.217 | 31.208 | 0.204 | 0.231 | < 0.001 | - | |||||||
| Family function(X)- PRA(M)-EPD(Y) ab | - | - | -0.079 | - | -0.094 | -0.065 | < 0.001 | 18.68 | |||||
The moderating effect of sleep quality on the mediating effect
In this study, Model 7 of the PROCESS macro was use to examine whether the interaction term (Family function * sleep latency) has a negative effect on PRA. The results are shown in Table 5. After adjusting for covariates and incorporating sleep latency into the model, it was found that sleep latency had a positive effect on PRA (β = 1.787, 95% CI: 1.245, 2.330). This finding indicates that the longer the sleep latency, the higher the level of anxiety in pregnant women. The interaction effect between Family function and sleep latency had a significant negative effect on PRA (β= -0.096, 95% CI: -0.160, -0.032), which implies that the protective effect of good Family function in reducing PRA is weakened as sleep latency increases. Specifically, longer sleep latency may exacerbate anxiety caused by inadequate Family function, further intensifying depressive symptoms. Furthermore, the differences in the mediating role of PRA between Family function and EPD among pregnant women with varying sleep scores were examined, and the results are presented in Table 6. When pregnant women had a lower sleep score of 0.182 (M − 1SD), the mediating effect of PRA was − 0.055 (95% CI: -0.075, -0.036), which did not include zero and was statistically significant. Conversely, when pregnant women had a higher sleep score of 1.779 (M + 1SD), the mediating effect of PRA was − 0.088 (95% CI: -0.106, -0.070), which was also statistically significant. Therefore, it can be inferred that as pregnant women’s sleep latency scores increase, the mediating effect related to anxiety intensifies. Fig. 6A.
Table 5.
Adjusted moderating effect of sleep latency on the mediating effect (N = 7622)
| Variables | PRA (M) | EPD (Y) | ||||||
|---|---|---|---|---|---|---|---|---|
| β | t | LLCI | ULCI | β | t | LLCI | ULCI | |
| Family function (X) | -0.234 | -5.346 | -0.320 | -0.148 | -0.344 | -19.989 | -0.378 | -0.310 |
| PRA (M) | - | 0.217 | 31.208 | 0.204 | 0.231 | |||
| Sleep latency (W) | 1.787 | 6.463 | 1.245 | 2.330 | - | |||
| XW | -0.096 | -2.940 | -0.160 | -0.032 | - | |||
| R 2 | 0.310 | 0.481 | ||||||
| F | 57.812 | 176.173 | ||||||
Table 6.
Mediating effect of sleep latency at different levels
| Sleep latency (W) | Effect | SE | LLCI | ULCI | |
|---|---|---|---|---|---|
| Overall | 0.182 (M − 1SD) | -0.055 | 0.001 | -0.075 | -0.036 |
| 0.981 (M) | -0.071 | 0.007 | -0.086 | -0.058 | |
| 1.779 (M+1SD) | -0.088 | 0.009 | -0.106 | -0.070 | |
| Early pregnancy | 0.128 (M- 1SD) | -0.061 | 0.025 | -0.111 | -0.014 |
| 0.920 (M) | -0.072 | 0.018 | -0.109 | -0.039 | |
| 1.713 (M+1SD) | -0.084 | 0.022 | -0.130 | -0.045 | |
| Second pregnancy | 0.153 (M- 1SD) | -0.066 | 0.016 | -0.098 | -0.037 |
| 0.912 (M) | -0.082 | 0.012 | -0.106 | -0.060 | |
| 1.670 (M+1SD) | -0.098 | 0.015 | -0.128 | -0.069 | |
| Third pregnancy | 0.257 (M- 1SD) | -0.08 | 0.014 | -0.067 | -0.011 |
| 1.093 (M) | -0.060 | 0.011 | -0.081 | -0.039 | |
| 1.928 (M+1SD) | -0.081 | 0.014 | -0.0111 | -0.054 |
Fig. 6.
Adjusted moderated mediation model diagram. A: The overall moderated mediation model diagram. B: Moderated mediation model diagram for early pregnancy. C: The moderated mediation model diagram for the second pregnancy. D: The moderated mediation model diagram for the third pregnancy. The numerical values on the paths represent unstandardized coefficients (β)
Further exploration of the moderating effect of sleep latency on the mediating effect across different trimesters of pregnancy was conducted. After adjusting for covariates, the results indicated that during early pregnancy (Table 7; Fig. 6B) and second pregnancy (Table 8; Fig. 6C), sleep latency had a positive effect on PRA (β = 1.636, 95% CI: 0.202, 3.071; β = 1.774, 95% CI: 0.947, 2.600). However, the interaction term between sleep latency and Family function did not show a statistically significant moderating effect on PRA (β = -0.068, 95% CI: -0.235, 0.100; β = -0.090, 95% CI: -0.187, 0.008). In the third pregnancy(Table 9; Fig. 6D), sleep latency had a positive effect on PRA (β = 2.030, 95% CI: 1.181, 2.878), and the interaction term between sleep latency and Family function demonstrated a negative moderating effect on PRA (β = -0.127, 95% CI: -0.226, -0.028).
Table 7.
Adjusted moderating effect of sleep latency on the mediating effect (N = 1230)
| Variables | PRA (M) | EPD (Y) | ||||||
|---|---|---|---|---|---|---|---|---|
| β | t | LLCI | ULCI | β | t | LLCI | ULCI | |
| Family function (X) | -0.283 | -2.652 | -0.492 | -0.074 | -0.313 | -6.906 | -0.403 | -0.224 |
| PRA (M) | - | 0.210 | 11.868 | 0.175 | 0.245 | |||
| Sleep latency (W) | 1.636 | 2.238 | 0.202 | 3.071 | - | |||
| XW | -0.068 | -0.789 | -0.235 | 0.100 | - | |||
| R 2 | 0.324 | 0.468 | ||||||
| F | 10.140 | 26.231 | ||||||
Table 8.
Adjusted moderating effect of sleep latency on the mediating effect (N = 3547)
| Variables | PRA (M) | EPD (Y) | ||||||
|---|---|---|---|---|---|---|---|---|
| β | t | LLCI | ULCI | β | t | LLCI | ULCI | |
| Family function (X) | -0.270 | -4.239 | -0.395 | -0.145 | -0.323 | -5.572 | -0.372 | -0.274 |
| PRA (M) | - | 0.233 | 23.099 | 0.213 | 0.253 | |||
| Sleep latency (W) | 1.774 | 4.208 | 0.947 | 2.600 | - | |||
| XW | -0.090 | -1.794 | -0.187 | 0.008 | - | |||
| R 2 | 0.305 | 0.497 | ||||||
| F | 25.785 | 89.350 | ||||||
Table 9.
Moderating effect of sleep latency on the mediating effect (N = 2845)
| Variables | PRA (M) | EPD (Y) | ||||||
|---|---|---|---|---|---|---|---|---|
| β | t | LLCI | ULCI | β | t | LLCI | ULCI | |
| Family function (X) | -0.155 | -2.092 | -0.301 | -0.010 | -0.381 | -13.677 | -0.436 | -0.327 |
| PRA (M) | - | 0.202 | 17.605 | 0.180 | 0.225 | |||
| Sleep latency (W) | 2.030 | 4.691 | 1.181 | 2.878 | - | |||
| XW | -0.127 | -2.506 | -0.226 | -0.028 | - | |||
| R 2 | 0.328 | 0.474 | ||||||
| F | 24.314 | 63.215 | ||||||
Key findings from the stratified analyses indicated that the moderating role of sleep latency in the association between Family function and PRA was not uniform. The protective effect of Family function against PRA was significantly attenuated by longer sleep latency among specific demographic and obstetric characteristics, including those aged ≤ 34 years(β = -0.086, 95% CI: -0.156, -0.015; Table S7), urban residents (β = -0.082, 95% CI: -0.152, -0.011; Table S8), with high school or below(β = -0.124, 95% CI: -0.239, -0.008; Table S9), or those without a history of adverse pregnancy (β = -0.091, 95% CI: -0.176, -0.005; Table S10).
Sensitivity analysis
To further interrogate the robustness of findings, we performed a comprehensive sensitivity analysis in which the moderated-mediation model was re-estimated separately with the global PSQI scores and with each of the six remaining PSQI subscales (Tables S11-16). After adjustment for the full covariate set, only the interaction between daytime dysfunction and Family function emerged as a significant moderator of the indirect pathway from Family function to PRA (β = 0.090, 95% CI: -0.149– -0.030, P < 0.01). See Table 10.
Table 10.
Adjusted moderating effect of overall Daytime dysfunction on the mediating effect (N = 7622)
| Variables | β | t | LLCI | ULCI | R 2 | F |
|---|---|---|---|---|---|---|
| Family function (X) | -0.217 | -5.664 | -0.293 | -0.142 | 0.352 | 76.621 |
| PRA (M) | - | |||||
| Daytime dysfunction(W) | 2.298 | 8.959 | 1.796 | 2.801 | ||
| XW | -0.090 | -2.936 | -0.149 | -0.030 | ||
In addition, 10 pregnant women with self-reported mental illness were excluded for sensitivity analysis. The results of the mediating effect showed that the interaction between family function and sleep latency still had statistical significance on the negative regulation of PRA(β = 0.094, 95% CI: -0.159– -0.032, P < 0.01). See Table S17.
Discussion
In this study, a questionnaire survey was conducted among 7622 pregnant women in a hospital in Ningxia to explore the mediating effect of PRA between Family function and EPD, and the moderating effect of sleep latency. Our cross-sectional data suggest that PRA may partially account for the association between family functioning and EPD; this pattern was observed during the COVID-19 containment and early-recovery periods (February 2022 – December 2024). Furthermore, during late pregnancy, sleep latency moderated the first half of the mediating pathway, specifically the relationship between Family functioning and PRA.
Consistent with the results of Shi et al. [21], the depression score among 7,622 pregnant women was 6.87 ± 3.71. The results of this study show that pregnant women with the following factors have higher EPD scores (P < 0.05), indicating that they are more likely to suffer from prenatal depression. This finding verifies the findings of previous studies, which have shown that unemployment, low education level, and family income are influential factors in prenatal depression among pregnant women [22, 23]. A possible reason for this observation is that pregnant women who are unemployed or have low education levels and low household incomes, typically experience financial hardship. Recent studies have also shown that the prevalence of depression in rural pregnant women is higher than that in urban pregnant women [24]. This may be related to the lack of social support in rural areas, the lack of medical resources, and the pressure of life. Consistent with this study, Bayrampour [25] reported that an introverted personality is a risk factor for depression. Pregnant women with an introverted personality may be more likely to feel lonely and helpless, thus increasing the risk of depression. In addition, this study also found that pregnant women who are not the only child in the family have higher depression scores than those who are the only child. This may be related to the fact that non-only-child pregnant women may face more challenges in terms of family support, interpersonal relationships, economic and social pressure, and pregnancy experience. Previous studies have confirmed that women with unintended pregnancies are more likely to suffer from depression than are those with planned pregnancies [26]. Unintended pregnancy imposes a significant psychological and work-related burden on the entire family, and the emphasis on offspring further exacerbates the stress associated with unintended pregnancy. Pregnant women who have experienced adverse pregnancy outcomes or severe pregnancy reactions are more prone to negative emotions like worry, anxiety, and even fear during pregnancy, which significantly predict prenatal depression [27, 28].
For pregnant women, family members constitute the primary source of social support. The correlation analysis results of this study indicate that Family functioning scores are negatively correlated with both PRA and EPD scores, which is consistent with previous research findings [10, 29]. A high level of Family function directly provides emotional support and psychological comfort, reduces the loneliness and helplessness of pregnant women, enhances their ability to cope with psychological stress, and thus reduces their levels of anxiety and depression. PRA is positively correlated with EPD, suggesting that an increase in PRA levels directly results in an elevation of depression levels. This aligns with the common psychological phenomenon that anxiety and depression frequently coexist, and furthermore, if anxiety remains unalleviated, this coexistence may result in depression [30, 31]. In addition, the results of this study found that PRA, as an intermediary variable, can partially mediate the impact of Family function on depressive symptoms. These findings suggest that insufficient family care may also indirectly aggravate depressive symptoms by increasing anxiety.
In addition to the direct impact of Family function and the mediating role of PRA, this study elucidated the moderating role of sleep during pregnancy. While most existing studies have established a close link between overall sleep quality and prenatal mental state [32, 33], few studies have specifically examined the relationships between sleep latency and Family function and mental state during pregnancy. The results of the correlation analysis in this study further indicated that sleep latency was negatively correlated with Family function and positively correlated with both PRA and EPD. Consequently, sleep latency was employed as a moderating variable to delve deeper into its moderating influence on the mediating effects’ outcomes. The results of the moderated mediation model indicated that the interaction between Family function and sleep latency exerted a significant negative impact on PRA. Further stratification by sociodemographic and obstetric characteristics indicated that this moderating effect was statistically significant, specifically among women who were urban residents, had lower educational attainment (high school or below), were younger (aged ≤ 34 years), or had no history of adverse pregnancy. This suggests that the protective effect of family function against anxiety is most vulnerable to disruption by sleep difficulties in these specific subgroups, which may relate to their distinct psychosocial resource profiles or stress contexts.
Further trimester-stratified analysis revealed that the statistically significant effect of this interaction on PRA was observed only during the third trimester. In the third trimester, the growing size of the fetus leads to increased pressure on the mother’s body, causing back pain and leg cramps, which significantly disrupt sleep [34]. These physical discomforts make it harder for pregnant women to find a comfortable sleeping position, thereby prolonging sleep onset latency. The current study also supported that sleep patterns worsen as pregnancy progresses [35], and poor sleep quality during this period was associated with an increased risk of perinatal mood and anxiety disorders [36]. In addition, women may experience significant stress and worry about the delivery process, pain, and potential complications in the third trimester, contributing to increased anxiety levels. These physiological stresses and poor family function work together to form a double burden, which puts pregnant women in a vicious circle and further aggravates perinatal mental disorders. Sensitivity analyses revealed that, after full covariate adjustment, only the interactions between family function and sleep latency or daytime dysfunction exerted a significant moderating influence on PRA. Among these, daytime dysfunction represents a non-specific, subjective appraisal that is easily confounded by general fatigue or mood state and therefore offers limited insight into the precise sleep pathophysiology characteristic of late pregnancy. In contrast, sleep latency quantifies the interval between settling in bed and sleep onset, thereby capturing the initiation difficulties that arise from third-trimester physiological perturbations. Consequently, interventions that simultaneously optimise family support and target sleep-onset latency are theoretically grounded and clinically expedient. Accordingly, to alleviate PRA and attenuate antenatal depressive symptoms, it is advisable to enhance family function through late-pregnancy family care training for spouses and main family members led by obstetric nurses or psychologists, and provide personalized sleep guidance (e.g., cognitive-behavioral techniques for insomnia, sleep hygiene education, relaxation training) for those with prolonged sleep latency.
While numerous studies have examined the relationships among Family function during pregnancy, the sleep quality of pregnant women, and their mental health status, few have investigated the mediating role of PRA and the interaction between Family function and sleep latency in prenatal depression. Drawing on a large-sample study, this research explores how pregnancy-related anxiety mediates the relationship between Family function and prenatal depression, and examines the moderating effect of sleep latency on this mediation. This study provides new insights for enhancing the mental health of pregnant women. Furthermore, it should be noted that the data collection period of this study covers the latter stages and the post-peak recovery phase of the COVID-19 pandemic. Pandemic-related social restrictions and adjustments in healthcare resources may have influenced the psychological state and family functioning of pregnant women, suggesting that the generalizability of the findings should be considered within this specific context. Therefore, future research needs to further verify the results of this study in different time contexts.
Limitations
However, a few limitations of this study also merit consideration. First, the single-center sampling and the region’s unique cultural context jointly limit the generalizability of findings. Second, the cross-sectional design prevents establishing causal relationships or temporal sequences among family function, pregnancy-related anxiety, sleep latency, and prenatal depression. Third, the reliance on self-report measures for key variables may introduce biases, which could compromise data accuracy.
Conclusion
In summary, the following conclusions can be drawn from this study: Family function has a direct effect on depression; and PRA partially mediates the relationship between family care and depression, and within this mediation model, there exists a moderating effect of the interaction between sleep latency in the third trimester and family care.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We thank all the volunteers who participated in this study, as well as the other students in the group who completed the questionnaire.
Author contributions
Suzhen Guan and Ye Li wrote the manuscript. Ling Ma and Chunyan Wang were involved in data collection, collation and analysis. All the other authors (Can Liu, Zexin Yu, Hao Wang, Shuqin Ma) were involved in the distribution, collection, and data entry of the questionnaires.
Funding
This work was supported by the National Natural Science Foundation of China [no. 82260647], Ningxia Key Research and Development Foundation Project [no. 2023BEG02005], and 2024 College Students’ Innovation and Entrepreneurship Plan -Ningxia Hui Autonomous Region-level scientific research project [no. S202410752028].
Data availability
The datasets generated and/or analyzed during the current study are not publicly available due to our experimental team’s policy, but are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
This study is in line with the ‘Helsinki Declaration’. Before the study, approval of the Ethics Committee of Ningxia Medical University (No.2022-G007) was obtained. All these subjects provided informed consent and filled out a paper questionnaire before enrollment.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Ye Li, Ling Ma and Chunyan Wang contributed equally to this work.
References
- 1.Dadi AF, Miller ER, Bisetegn TA, et al. Global burden of antenatal depression and its association with adverse birth outcomes: an umbrella review [J]. BMC Public Health. 2020;20(1):173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Fekadu Dadi A, Miller ER, Mwanri L. Antenatal depression and its association with adverse birth outcomes in low and middle-income countries: A systematic review and meta-analysis [J]. PLoS ONE. 2020;15(1):e0227323. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Cena L, Gigantesco A, Mirabella F, et al. Prevalence of comorbid anxiety and depressive symptomatology in the third trimester of pregnancy: Analysing its association with sociodemographic, obstetric, and mental health features [J]. J Affect Disord. 2021;295:1398–406. [DOI] [PubMed] [Google Scholar]
- 4.Kasujja M, Omara S, Senkungu N, et al. Factors associated with antenatal depression among women attending antenatal care at Mubende Regional Referral Hospital: a cross-sectional study [J]. BMC Womens Health. 2024;24(1):195. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Martini J, Petzoldt J, Einsle F, et al. Risk factors and course patterns of anxiety and depressive disorders during pregnancy and after delivery: a prospective-longitudinal study [J]. J Affect Disord. 2015;175:385–95. [DOI] [PubMed] [Google Scholar]
- 6.Behera D, Bohora S, Tripathy S, et al. Perinatal depression and its associated risk factors during the COVID-19 pandemic in low- and middle-income countries: a systematic review and meta-analysis [J]. Soc Psychiatry Psychiatr Epidemiol. 2024;59(10):1651–68. [DOI] [PubMed] [Google Scholar]
- 7.Tang X, Lu Z, Hu D, et al. Influencing factors for prenatal Stress, anxiety and depression in early pregnancy among women in Chongqing, China [J]. J Affect Disord. 2019;253:292–302. [DOI] [PubMed] [Google Scholar]
- 8.Zheng B, Yu Y, Zhu X, et al. Association between family functions and antenatal depression symptoms: a cross-sectional study among pregnant women in urban communities of Hengyang city, China [J]. BMJ Open. 2020;10(8):e036557. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Fraga SD, Khan IN, Sharma TA, et al. Predominant approaches to measuring pregnancy-related anxiety in Sub-saharan Africa: a scoping review [J]. BMC Public Health. 2024;24(1):2425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Huang J, Xu L, Xu Z, et al. The relationship among pregnancy-related anxiety, perceived social support, family function and resilience in Chinese pregnant women: a structural equation modeling analysis [J]. BMC Womens Health. 2022;22(1):546. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mindell JA, Cook RA, Nikolovski J. Sleep patterns and sleep disturbances across pregnancy [J]. Sleep Med. 2015;16(4):483–8. [DOI] [PubMed] [Google Scholar]
- 12.Loret de Mola C, Carpena MX, Dias IM, et al. Sleep and its association with depressive and anxiety symptoms during the last weeks of pregnancy: A population-based study [J]. Sleep health. 2023;9(4):482–8. [DOI] [PubMed] [Google Scholar]
- 13.Shariat M, Abedinia N, Noorbala A, et al. The relationship between sleep quality, depression, and anxiety in pregnant women: a cohort study. [J] JSS. 2018;2(1–2):20–7. [Google Scholar]
- 14.Smilkstein G, Ashworth C, Montano D. Validity and reliability of the family APGAR as a test of family function [J]. J Fam Pract. 1982;15(2):303–11. [PubMed] [Google Scholar]
- 15.Limin X. Development of the Pregnancy-specific Anxiety Questionnaire and pregnaney-specifie anxiety status. in early pregnancy [D]; Anhui Medical University; 2010.
- 16.Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression. Development of the 10-item Edinburgh Postnatal Depression Scale [J]. Br J Psychiatry. 1987;150:782–6. [DOI] [PubMed] [Google Scholar]
- 17.Cox J. Thirty years with the Edinburgh Postnatal Depression Scale: voices from the past and recommendations for the future [J]. Br J Psychiatry. 2019;214(3):127–9. [DOI] [PubMed] [Google Scholar]
- 18.Buysse DJ, Reynolds CF 3rd, Monk TH, et al. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research [J]. Psychiatry Res. 1989;28(2):193–213. [DOI] [PubMed]
- 19.Saavedra J, Brzeska J, Matías-García JA, et al. Quality of life and psychiatric distress in people with serious mental illness, the role of personal recovery [J]. Psychol Psychother. 2023;96(2):525–41. [DOI] [PubMed] [Google Scholar]
- 20.Hayes AF. Introduction to mediation, moderation, and conditional process analysis: a regression-based approach (3rd ed.) [J]. The Guilford Press, 2022.
- 21.Shi P, Ren H, Li H, et al. Maternal depression and suicide at immediate prenatal and early postpartum periods and psychosocial risk factors [J]. Psychiatry Res. 2018;261:298–306. [DOI] [PubMed] [Google Scholar]
- 22.Liu Y, Zhang R, Zhang Z, et al. Risk factors and predictive model for prenatal depression: A large retrospective study in China [J]. J Affect Disord. 2024;354:1–10. [DOI] [PubMed] [Google Scholar]
- 23.Ren F, Zhu X, Liu J, et al. Associations of multiple risk factors with prenatal depression and anxiety: Evidence from the Tianjin Birth Cohort (TJBC) study [J]. J Affect Disord. 2024;366:411–22. [DOI] [PubMed] [Google Scholar]
- 24.Chala S, Desalegn M, Oljira R, et al. A comparative study of antenatal depression among urban and rural pregnant women in Gimbi District, Oromia, Ethiopia [J]. Front public health. 2024;12:1393880. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Bayrampour H, McDonald S, Tough S. Risk factors of transient and persistent anxiety during pregnancy [J]. Midwifery. 2015;31(6):582–9. [DOI] [PubMed] [Google Scholar]
- 26.Moreau C, Bonnet C, Beuzelin M, et al. Pregnancy planning and acceptance and maternal psychological distress during pregnancy: results from the National Perinatal Survey, France, 2016 [J]. BMC Pregnancy Childbirth. 2022;22(1):162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Li W, Lin L, Teng S, et al. Path analysis of influencing factors for anxiety and depression among first-trimester pregnant women [J]. Front Psychol. 2024;15:1440560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Charrois EM, Mughal MK, Arshad M, et al. Patterns and predictors of depressive and anxiety symptoms in mothers affected by previous prenatal loss in the ALSPAC birth cohort [J]. J Affect Disord. 2022;307:244–53. [DOI] [PubMed] [Google Scholar]
- 29.Gao Y, Tang X, Deng R, et al. Latent trajectories and risk factors of prenatal stress, anxiety, and depression in southwestern China-a longitudinal study [J]. Int J Environ Res Public Health. 2023;20(5). [DOI] [PMC free article] [PubMed]
- 30.Nielsen-Scott M, Fellmeth G, Opondo C, et al. Prevalence of perinatal anxiety in low- and middle-income countries: A systematic review and meta-analysis [J]. J Affect Disord. 2022;306:71–9. [DOI] [PubMed] [Google Scholar]
- 31.Zhang L, Wang L, Cui S, et al. Prenatal Depression in Women in the Third Trimester: Prevalence, Predictive Factors, and Relationship With Maternal-Fetal Attachment [J]. Front public health. 2020;8:602005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Aukia L, Paavonen EJ, Jänkälä T, et al. Insomnia symptoms increase during pregnancy, but no increase in sleepiness - Associations with symptoms of depression and anxiety [J]. Sleep Med. 2020;72:150–6. [DOI] [PubMed] [Google Scholar]
- 33.Polo-Kantola P, Aukia L, Karlsson H, et al. Sleep quality during pregnancy: associations with depressive and anxiety symptoms [J]. Acta Obstet Gynecol Scand. 2017;96(2):198–206. [DOI] [PubMed] [Google Scholar]
- 34.Kiyoko M, Takayuki A, Kaoru F, et al. Objective and self-reported assessment of sleep quality during each trimester of pregnancy: A prospective observational study [J]. Eur J Obstet Gynecol Reprod Biol. 2024;300:296–301. [DOI] [PubMed] [Google Scholar]
- 35.Al-Musharaf S. Changes in sleep patterns during pregnancy and predictive factors: a longitudinal study in Saudi women [J]. Nutrients. 2022;14(13). [DOI] [PMC free article] [PubMed]
- 36.Osnes RS, Roaldset JO, Follestad T, et al. Insomnia late in pregnancy is associated with perinatal anxiety: A longitudinal cohort study [J]. J Affect Disord. 2019;248:155–65. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated and/or analyzed during the current study are not publicly available due to our experimental team’s policy, but are available from the corresponding author upon reasonable request.