Invisible Scars: Psychopathology, Shame and Self-Judgment Following Perinatal Loss—A Cross-Sectional Study

Abstract

Perinatal loss affects 23 million pregnancies worldwide each year, representing a painful experience that disrupts expectations and impacts emotional, physical, social, and spiritual well-being. This cross-sectional observational study assessed symptoms of anxiety, depression, self-judgment (self-criticism, isolation, over-identification), and shame in women who experienced perinatal loss, as well as their predictive value for psychopathology. Participants were 501 women, divided into five groups according to time since loss: 0–6 months, 7–18 months, 19–30 months, 31–42 months, and more than 43 months. Findings showed that women 7–18 months post-loss reported the highest psychopathology levels, with significant differences in anxiety. Isolation and shame were the strongest predictors of depressive and anxiety symptoms. Although symptoms decreased over time, they remained elevated years after the loss. These results underscore the lasting psychological impact of perinatal loss and the importance of sustained recognition, assessment, and intervention to support women’s mental health.

1. Introduction

Perinatal loss refers to the set of situations that occur during pregnancy, during childbirth or within 28 days of childbirth that result in the loss of one or more children. Although there is no consensus in the literature on the classification of perinatal loss, we propose defining miscarriage as the involuntary termination of a pregnancy before the 20th week of gestation [1]. Miscarriage can be early if the loss occurs within the first 12 weeks of gestation or late if the loss occurs between weeks 13 and 20. In the last decade, an incidence of 15 to 25% of early pregnancy losses [2,3,4] is estimated worldwide. The real incidence of miscarriage is difficult to estimate since most of these losses are not accounted for by the authorities, but also because the loss can occur so early that the woman is unaware of her pregnancy [5]. Additionally, limitations in national registration systems contribute to the underestimation of early losses. In Portugal, the National Statistics Institute (INE) registers fetal deaths only according to whether they occur before or after 32 weeks of gestation, which restricts the documentation of losses occurring at earlier gestational ages. In Portugal, no data was found on the prevalence of this loss.

Fetal death refers to losses that occur after the 20th week of pregnancy or during childbirth [6]. Every year, there are around two million of these losses worldwide, although it is estimated that fewer than 5% of stillbirths are officially registered, reflecting substantial global underreporting [7]. According to PORDATA [8], a public database of official demographic and health statistics, in Portugal specifically, there were 214 fetal deaths in 2024.

Neonatal death refers to the loss of a child after less than 28 days of life [9]. This loss can be early, if it occurs before the seventh day of life, or late, if the death occurs between the 7th and 27th day of life [10]. In 2022, 2.3 million children died in the first month of life worldwide [11]. In Portugal, there were 259 neonatal deaths in 2024 [10]. Terminating a pregnancy voluntarily or on medical advice is part of a decision-making process that generates enormous stress and suffering [1]. Worldwide, according to the WHO [12], 29% of all pregnancies end in induced abortion, and it is estimated that, throughout their reproductive life, 40% of women will terminate their pregnancy. In Portugal, voluntary termination of pregnancy or induced abortion can be carried out in the first 10 weeks of pregnancy, and medical termination of pregnancy can be carried out at any time throughout the pregnancy if the situations described in article 142 of the Portuguese Penal Code [13] are verified. The diagnosis of fetal malformation unexpectedly, in the context of a desired and normative pregnancy, can increase the anguish felt by the couple [1]. In 2023, 17,124 induced abortions took place in Portugal, of which 96.7% occurred in the first 10 weeks of pregnancy at the woman’s choice and around 3.2% occurred due to congenital malformation of the fetus or to avoid danger to the woman’s physical or mental health [14]. Multiple systematic reviews and meta-analyses have established that perinatal loss is associated with significant adverse psychological outcomes [15,16]. While this body of evidence consistently identifies psychological risks associated with perinatal loss across different populations, research focusing specifically on the Portuguese population is limited, particularly with regard to constructs such as shame and self-judgment. This study aims to address this gap by examining the psychopathological symptoms experienced by Portuguese women who have had a perinatal loss and the role of self-judgment and shame in these symptoms.

Perinatal loss is a potentially traumatic event with a social, economic, and psychological impact on the couples who experience it. Psychologically, it is associated with high rates of anxiety and depression, with some studies reporting prevalence rates of maternal depression as high as 43% [17,18]. Socially, bereaved parents often experience isolation and a lack of recognition for their grief, which can strain family relationships and reduce community support [19]. Economically, parents may face significant healthcare costs and reduced work capacity due to emotional distress, further affecting their financial stability [20]. Concerning mental health, this loss can trigger sadness, anger, frustration, and emptiness, along with guilt, shame, and helplessness in parents [21]. In addition, after perinatal loss, depressive, anxiety, and post-traumatic stress symptoms can emerge [22]. A study by Mendes et al. [5] revealed that Portuguese participants who had experienced early pregnancy loss showed symptoms of depression, anxiety, and post-traumatic stress, with higher levels of these symptoms one month after the loss, compared to several months after the gestational loss (for example, losses that occurred 7–12 months ago). According to Farren et al. [23], 24% of women reported moderate to severe anxiety and 11% reported symptoms of depression one month after experiencing early pregnancy loss. These indicators gradually decreased over the course of nine months, though they remained higher than those observed in the control group with viable pregnancies [23].

Psychopathological symptoms may arise or increase due to various factors, including self-judgment [24]. Self-judgment refers to negative self-assessment in situations of failure [25]. According to Galhardo et al. [26], when individuals are confronted with suffering, inadequacy, or failure, they judge themselves, invalidate their pain, and criticize themselves. Neff [27] considers that self-judgment, unlike self-compassion, implies that individuals, when faced with situations of failure or pain, are self-critical, isolate themselves, and perceive their experience as separate from the common human condition and over-identify with their painful thoughts and feelings (over-identification), something that can contribute to the development or worsening of depressive and anxiety symptoms [28], and of the grieving process [29]. In addition, shame is an emotion frequently reported after perinatal loss and is related to the feeling of failure or biological defect and loss of control [30]. Shame is a painful emotional experience that concerns the individuals’ perception of being judged negatively and seen as inferior [31], being associated with a tendency to compare themselves with others [32]. After experiencing perinatal loss, women feel ashamed for not having been able to carry their pregnancy to term when they meet pregnant women or women with young children on the street [33]. In addition, shame may lead to the onset or worsening of depression and anxiety [34]. In Portugal, motherhood continues to be highly valued as an integral part of female identity, which can aggravate feelings of inadequacy and shame following a pregnancy loss [35]. This reality may contribute to the emotional isolation of bereaved women, limiting their access to adequate social and psychological support and perpetuating the silence, shame, and stigma surrounding perinatal bereavement.

Due to the lack of studies on the role of shame and self-judgment associated with perinatal loss, this study aimed to evaluate the levels of psychopathology and understand the impact of self-judgment and shame on symptoms of depression and anxiety in Portuguese women who have experienced gestational loss. Most existing research focuses only on a specific type of perinatal loss, excluding losses occurring later in pregnancy or neonatal losses.

2. Materials and Methods

This cross-sectional observational study was previously evaluated and authorized by the Ethics Committee of the University of Beira Interior (CE-UBI-Pj-2022-066). The link to the research protocol, which was created using Microsoft Forms, was disseminated between June 2023 and June 2024 through associations and projects that support parents who have suffered a perinatal loss, social networks, and in person at the (excluded by the anonymity rule). The research protocol consisted of a brief explanation of the research and the objectives of the study, as well as sociodemographic and clinical questions regarding perinatal loss. Assessment scales for the study variables were also used. Participation was voluntary and anonymous, and all participants agreed to the informed consent form.

With regard to the sample, adult women who had experienced perinatal loss were eligible to complete the questionnaire if they met the following inclusion criteria: (1) having suffered a perinatal loss, (2) being 18 years of age or older, (3) being a woman, and (4) being of Portuguese nationality. Participants were not eligible if they (1) had not experienced a perinatal loss; (2) were of a nationality other than Portuguese, (3) were male, and (4) were aged above 49 years (decline in fertility) [36,37]. After the survey was closed, a detailed analysis of the responses was carried out. 545 respondents replied, 44 of whom were excluded (15 because they were of a nationality other than Portuguese, 12 because they were men; nine because they were over 49 years old, three were submissions by women who had not actually experienced a perinatal loss, and five due to missing or incorrect data (e.g., implausible year of perinatal loss or absence of essential sociodemographic or clinical information, such as age and number of losses)). This resulted in a final sample of 501 eligible participants (Figure 1). The final sample size (n = 501) exceeded the recommended minimum for this type of analysis, as sample sizes between 150 and 400 are considered adequate by authors such as Schoemann et al. [38]. Participants who did not meet these criteria (i.e., men, women without a history of perinatal loss, individuals of non-Portuguese nationality, or women over 49 years old) were excluded. While some criteria, such as age and nationality, could be pre-specified in the survey, others—such as confirmation of a previous perinatal loss—could not be automatically verified. Therefore, after the survey was closed, a detailed review of all responses was conducted, and participants who did not meet the inclusion criteria were excluded. After this screening process, 501 eligible participants were included in the analyses.

The research protocol included four self-report questionnaires: the sociodemographic questionnaire, the Brief Symptom Inventory 18 (BSI-18), the Self-Compassion Scale (negative dimension—Self-judgment), and the Personal Attitudes Questionnaire—Shame Scale. Sociodemographic and clinical information was collected via a questionnaire, with questions about age, gender, nationality, socioeconomic status, and marital status, as well as the number and year of the loss(es), the cause(s) of the loss(es), and previous or current psychological or psychiatric support, among other characteristics. Psychological or psychiatric support was defined as having received care from a licensed psychologist and/or psychiatrist; support ‘in the past’ referred to care received at any time prior to the study but not ongoing at the time of data collection, whereas ‘currently’ referred to ongoing care at the time of questionnaire completion.

To assess psychopathological symptoms, the participants answered the Portuguese version of the Brief Symptom Inventory 18 [39]. This instrument consists of 18 items organized in three subscales (Depression, Anxiety, and Somatization). Participants are asked to rate the intensity (0—Not at all to 4—Extremely) with which they have experienced 18 psychopathological symptoms in the last seven days [39]. Higher scores represent greater levels of psychological distress. The subscale scores range from 0 to 24, while the Global Severity Index (GSI) total score ranges from 0 to 72. According to the instrument’s interpretation guidelines, T-scores equal to or above 60 are considered above average across all subscales. In this study, we used the anxiety subscale and the depression subscale, with an internal consistency of α = 0.86 and α = 0.89, respectively. The Self-Compassion Scale (SCS; Portuguese version by Castilho et al., [40]) is a measure of self-compassion that includes 26 items assessed on a 5-point Likert scale. In this study, the self-Judgment subscale was used, which corresponds to the sum of self-criticism (e.g., ‘I am disapproving and judgmental of my own faults and shortcomings’), isolation (e.g., ‘When I think about my shortcomings, it tends to make me feel more separate and isolated from the rest of the world’) and over-identification (e.g., ‘When I fail at something important to me, I am consumed by feelings of inadequacy’). The total score for each subscale is derived from its average score. Higher scores indicate higher levels of self-criticism, isolation, and over-identification. In this study, the internal consistency of the self-criticism subscale was α = 0.81, while the isolation subscale was α = 0.79, and the overidentification subscale was α = 0.79 [40]. The Personal Attitudes Questionnaire—Shame and Guilt Scale was developed by Harder and Greenwald in 1999 and translated and adapted into Portuguese by Geada [41]. It consists of 22 items (e.g., “Feeling ridiculous” and “Feeling “stupid”), which are expressed on a Likert-type scale, ranging from 0 to 4, for a total score of 0 to 40. A higher score means a higher chance of feeling shame, which may indicate a higher level of emotional self-judgment [42]. In this study, we used the shame subscale, which showed an internal consistency of α = 0.87.

Statistical analysis was carried out using the Statistical Package for the Social Science (SPSS) software, version 29.1. Tests of normality revealed p-values < 0.05, suggesting that the variables’ distribution deviated significantly from normality. Homoscedasticity was also rejected for this study’s sample. Given these assumptions were violated, nonparametric procedures were employed where appropriate. Pearson’s chi-square test (χ²) was used to examine associations between sociodemographic and clinical variables across the study groups. This test indicated that the distributions of these variables were independent.

Participants were grouped according to the time elapsed since their last perinatal loss into five intervals (0–6, 7–18, 19–30, 31–42, and >43 months). This temporal categorization was informed by empirical evidence indicating that psychopathological symptoms, particularly anxiety and depression, are typically most intense during the first months following perinatal loss, with subsequent fluctuations and a gradual decrease over time, followed by relative stabilization approximately two years post-loss [5,43] In addition, the selected group boundaries were informed by the distribution of the sample, aiming to ensure adequate group sizes for statistical analyses.

The 501 participants were distributed into five groups based on the time since last perinatal loss: group 0–6 (six months or less); group 7–18 (7–18 months); group 19–30 (19–30 months); group 31–42 (31–42 months) and group 43+ (more than 3.5 years). Kruskal–Wallis tests were used to compare the groups in psychopathological symptoms (anxiety and depression). As no statistically significant global effects were observed, post hoc pairwise comparisons were not conducted. Spearman’s correlation coefficients were used to analyze the relationships between depression, anxiety, shame, and self-judgment dimensions (self-criticism, isolation, and over-identification). In addition, linear regression models were used to test the predictive effect of shame and the self-judgment dimensions on depression and anxiety. No alpha adjustments for multiple comparisons were applied to the correlation and regression analyses, as these analyses were guided by a predefined theoretical framework.

3. Results

3.1. Participants

Regardind participants, the average age was 34 (SD = 4.77), ranging from 21 to 47.

Table 1. Sociodemographic and clinical data according to time (in months) since last perinatal loss.

Variables	Category	0–6 (n = 19)	7–18 (n = 232)	19–30 (n = 131)	31–42 (n = 51)	43+ (n = 68)
Age		M = 32.95 (SD = 4.97)	M = 33.21 (SD = 4.47)	M = 34.23 (SD = 4.84)	M = 33.73 (SD = 4.62)	M = 36.74 (SD = 4.71)
Marital status	Single	__	7.8% (n = 18)	6.1% (n = 8)	7.8% (n = 4)	7.4% (n = 5)
	Nonmarital partnership	36.8% (n = 7)	47.0% (n = 109)	46.6% (n = 61)	37.3% (n = 19)	22.1% (n = 15)
	Married	63.2% (n = 12)	44.8% (n = 104)	45.8% (n = 60)	52.9% (n = 27)	69.1% (n = 47)
	Separated	__	0.4% (n = 1)	0.16% (n = 2)	2.0% (n = 1)	1.5% (n = 1)
Household socioeconomic status	Less than €500	__	__	0.8% (n = 1)	3.9% (n = 2)	1.5% (n = 1)
	Between 500 € and 999 €	5.3% (n = 1)	3.4% (n = 8)	5.3% (n = 7)	2.0% (n = 1)	4.4% (n = 3)
	Between 1000 € and 2499 €.	42.1% (n = 8)	58.2% (n = 135)	58.8% (n = 77)	56.9% (n = 29)	55.9% (n = 38)
	Between 2500 € and 4999 €	47.4% (n = 9)	33.6% (n = 78)	33.6% (n = 44)	27.5% (n = 14)	32.4% (n = 22)
	More than 5000 €	5.3% (n = 1)	4.7% (n = 11)	1.5% (n = 2)	9.8% (n = 5)	5.9% (n = 4)
Academic qualifications	High School (9th grade)	__	1.7% (n = 4)	2.3% (n = 3)	__	5.9% (n = 4)
	High School (12th grade)	10.5% (n = 2)	23.3% (n = 54)	20.6% (n = 27)	23.5% (n = 12)	25.0% (n = 17)
	Bachelor’s degree	36.8% (n = 7)	34.5% (n = 80)	34.4% (n = 45)	41.2% (n = 21)	42.6% (n = 29)
	Master’s degree	52.6% (n = 10)	39.2% (n = 91)	42.0% (n = 55)	35.3% (n = 18)	25.0% (n = 17)
	Doctoral degree	__	1.3% (n = 3)	0.8% (n = 1)	__	1.5% (n = 1)
Professional Status	Unemployed	10.5% (n = 2)	3.9% (n = 9)	2.3% (n = 3)	7.8% (n = 4)	4.4% (n = 3)
	Employed	84.2% (n = 16)	94.4% (n = 219)	91.6% (n = 120)	88.2% (n = 45)	89.7% (n = 61)
	Student	__	__	0.8% (n = 1)	__	__
	Worker/Student	5.3% (n = 1)	1.7% (n = 4)	5.3% (n = 7)	3.9% (n = 2)	5.9% (n = 4)
Household	Living alone	__	1.3% (n = 3)	1.5% (n = 2)	2.0% (n = 1)	1.5% (n = 1)
	Living with child(ren)	__	__	0.8% (n = 2)	3.9% (n = 2)	2.9% (n = 2)
	Living with a partner	73.7% (n = 14)	62.5% (n = 145)	49.6% (n = 65)	17.6% (n = 9)	13.2% (n = 9)
	Living with a partner and child(ren)	26.3% (n = 5)	32.3% (n = 75)	47.3% (n = 62)	74.5% (n = 38)	80.9% (n = 55)
	Other households	__	3.9% (n = 9)	0.8% (n = 1)	2.0% (n = 1)	1.5% (n = 1)
Place of Residence	Village	5.3% (n = 1)	8.6% (n = 20)	9.2% (n = 12)	17.6% (n = 9)	19.1% (n = 13)
	Town	10.5% (n = 2)	18.1% (n = 42)	19.8% (n = 26)	19.6% (n = 10)	11.8% (n = 8)
	Small city	36.8% (n = 7)	26.7% (n = 62)	29.8% (n = 39)	31.4% (n = 16)	38.2% (n = 26)
	Large city	47.4% (n = 9)	46.6% (n = 108)	41.2% (n = 54)	31.4% (n = 16)	30.9% (n = 21)
Number of pregnancies		M = 1.89 (DP = 1.45)	M = 2.20 (DP = 1.15)	M = 2.15 (DP = 1.03)	M = 2.49 (DP = 1.22)	M = 3.01 (DP = 1.37)
Number of perinatal losses		M = 1.47 (DP = 0.77)	M = 1.60 (DP = 0.94)	M = 1.34 (DP = 0.68)	M = 1.37 (DP = 0.82)	M = 1.69 (DP = 1.06)
Type of perinatal loss (most recent loss)	Early miscarriage	78.9% (n = 15)	63.8% (n = 148)	58.8% (n = 77)	72.5% (n = 37)	57.4% (n = 39)
	Late miscarriage	15.8% (n = 3)	18.1% (n = 42)	13.0% (n = 17)	13.7% (n = 7)	16.2% (n = 11)
	Fetal and neonatal death	5.3% (n = 1)	18.1% (n = 42)	28.2% (n = 37)	13.7% (n = 7)	26.5% (n = 18)
Psychological or psychiatric support (in the past)	Yes	63.2% (n = 12)	47.0% (n = 109)	45.8% (n = 60)	47.1% (n = 24)	39.7% (n = 27)
	No	36.8% (n = 7)	53.0% (n = 123)	54.2% (n = 71)	52.9% (n = 27)	60.3% (n = 41)
Psychological or psychiatric support (currently)	Yes	42.1% (n = 8)	43.1% (n = 100)	31.3% (n = 41)	17.6% (n = 9)	22.1% (n = 15)
	No	57.9% (n = 11)	56.9% (n = 132)	68.7% (n = 90)	82.4% (n = 42)	77.9% (n = 53)

M = mean; SD = standard deviation.

describes the sociodemographic information and the clinical variables of the sample in more detail. Most of the participants were married, had a socio-economic level of between 1000 € and 2499 € and lived in a large city. In addition, the sample had a mean of 2.31 pregnancies (SD = 1.203) and 1.52 perinatal losses (SD = 0.884). Specifically, 67.1% of the participants experienced one loss, 20.2% experienced two losses, and 12.7% experienced three or more losses. (see Table 1).

3.2. Anxiety and Depressive Symptoms

We evaluated the progression of psychological symptomatology in women following a perinatal loss, identifying periods of heightened vulnerability and the persistence of distress over time.

Table 2. Descriptive statistics for psychopathology, shame, and self-judgment according to time (months) since last perinatal loss.

Time Since Last Perinatal Loss
	0–6 (n = 19)		7–18 (n = 232)		19–30 (n = 131)		31–42 (n = 51)		43+ (n = 68)
	Median (IQR)	M (SD)	Median (IQR)	M (SD)	Median (IQR)	M (SD)	Median (IQR)	M (SD)	Median (IQR)	M (SD)
BSI-18: Anxiety	6 (4–10)	6.95 (3.84)	8 (5–11)	8.23 (4.65)	6 (3–10)	7.05 (4.87)	6 (4–11)	7.22 (5.12)	6 (3–11)	6.96 (5.16)
BSI-18: Depression	6 (4–11)	7.16 (4.50)	8 (4–12)	8.33 (5.07)	6 (3–10)	7.04 (5.20)	6 (2–11)	6.41 (5.09)	5 (3–10)	6.88 (5.46)
SELFCS: Self-criticism	2.6 (2.2–3)	2.64 (0.70)	2.8 (2.2–3.4)	2.74 (0.79)	2.8 (2.2–3.2)	2.78 (0.85)	2.6 (2.2–3.2)	2.70 (0.84)	2.7 (2.2–3.35)	2.76 (0.88)
SELFCS: Isolation	2.5 (2–3.25)	2.64 (0.86)	2.88 (2.25–3.5)	2.82 (0.88)	2.75 (2–3.25)	2.68 (0.90)	3 (2.25–3.25)	2.81 (0.83)	2.75 (2.25–3.5)	2.86 (0.91)
SELFCS: Over-identification	2.75 (2.25–3.25)	2.83 (0.68)	3 (2.5–3.5)	2.99 (0.85)	3 (2.25–3.5)	2.89 (0.88)	2.75 (2.25–3.5)	2.84 (0.88)	3 (2.25–3.5)	2.88 (0.86)
PAQ: Shame	11 (7–15)	11.68 (5.40)	12 (7.25–16)	12.54 (6.62)	12 (7–16)	12.69 (6.76)	12 (8–17)	12.71 (6.73)	13 (10–19)	14.49 (6.69)

M = Mean; SD = Standard Deviation; IQR = Interquartile Range; BSI-18 Anxiety and Depression: 0–24; Self-Judgment: 1–5; Shame: 0–40.

shows that the highest levels of anxiety and depression occur among women whose loss took place 7–18 months earlier, indicating a critical period of greater emotional vulnerability.

Table 3. Spearman correlation between the BSI-18/respective subscales, the subscales of the negative dimension of the Self-Compassion Scale (SCS), the PAQ—Shame Scale and the sociodemographic and clinical characteristics of the sample.

Variables	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16
1. Age	1
2. Marital status	0.181 **	1
3. Academic qualifications	0.054	0.197 **	1
4. Professional Status	0.091 *	−0.063	0.111 *	1
5. Place of Residence	0.059	0.067	0.242 *	0.029	1
6. Household socioeconomic status	0.153 *	0.181 *	0.339 *	0.083	0.265 **	1
7. Household	0.224 **	0.205 **	0.007	0.014	−0.045	0.094 *	1
8. Number of pregnancies	0.302 **	0.160 **	−0.028	−0.052	0.008	0.037	0.471 **	1
9. Number of perinatal losses	0.225 **	0.046	−0.022	−0.020	0.010	−0.009	0.049	0.627 **	1
10. Type of perinatal loss (most recent loss)	−0.065	−0.011	−0.084	0.009	−0.040	−0.025	−0.068	−0.094 *	−0.127 **	1
11. BSI-18—Anxiety	−0.098 *	−0.095 *	−0.173 **	−0.027	−0.103 *	−0.076	−0.128 **	−0.079	−0.035	0.133 **	1
12. BSI-18—Depression	−0.080	−0.110 *	−0.182 *	−0.053	−0.112 *	−0.109 *	−0.119 **	−0.122 **	−0.015	0.113 *	0.753 **	1
13. SCS—Self-criticism	−0.053	−0.048	−0.019	0.038	−0.017	0.006	0.007	−0.057	−0.039	0.032	0.399	0.433 **	1
14. SCS—Isolation	−0.077	−0.097 *	−0.085	0.036	−0.056	0.009	−0.059	−0.064	−0.010	0.056	0.452 **	0.504 **	0.709 **	1
15. SCS—Over-identification	−0.080	−0.024	0.016	0.022	0.011	0.015	−0.040	−0.043	0.021	0.029	0.408 **	0.402 **	0.724 **	0.707 **	1
16. PAQ-Shame	−0.075	−0.093 *	−0.093 *	0.000	−0.154 **	−0.080	0.038	0.014	0.015	−0.009	0.443 **	0.436 **	0.526 **	0.516 **	0.493 **	1

* p < 0.005; ** p < 0.001.

illustrates the significant positive and negative correlations between anxiety symptoms, depressive symptoms, and the other study variables. Concerning anxiety symptoms, significant differences were found according to time since last perinatal loss in the Kruskal–Wallis test (H(4) = 22.009, p < 0.001, ${\eta }_H^2$ = 0.036), specifically between the 19–30 months and 7–18 months groups (p = 0.002). Other statistically significant differences were found between anxiety symptoms and sociodemographic characteristics of the study sample. In terms of academic qualifications (H(4) = 15.218, p = 0.004), there were significant differences between participants who had a master’s/post-graduate degree and secondary education (12th grade) (p < 0.001), and participants who had a doctorate and secondary education (p = 0.028). Regarding the household (H(4) = 12.327, p = 0.015), there were significant differences between participants who lived with their partner and children and women who lived only with their partner (p = 0.001). Finally, significant differences were found in relation to the type of perinatal loss experienced. There were differences between women who had an early miscarriage and those who had a late miscarriage (p = 0.012), as well as between participants who experienced an early miscarriage and those who experienced fetal or neonatal death (p = 0.036) (see Table 3).

Concerning depressive symptoms, significant differences were also found in the time since the last perinatal loss in the Kruskal–Wallis test (H(4) = 9.509, p = 0.0495), specifically between the 31–42 months and 7–18 months groups (p = 0.012) and between the 19–30 months and 7–18 months groups (p = 0.013). Other statistically significant differences were found between depressive symptoms and sociodemographic characteristics of the study sample. Regarding academic qualifications (H(4) = 17.609, p = 0.001), there were significant differences between women who had a master’s degree and women who had a bachelor’s degree (p = 0.001), as well as between participants who had a master’s degree and secondary education (12th grade) (p < 0.001). With regard to socioeconomic status (H(4) = 12.560, p = 0.014), there were significant differences between women whose income was between €2500–€4999 and women whose income was between €1000–€2499 (p = 0.005), women whose income was between €2500–€4999 and women whose income was over €5000 (p = 0.032), and participants with an income between €2500–€4999 and women whose income was between €500–€999 (p = 0.034). With regard to household (H(4) = 14.347, p = 0.006), there were significant differences between women who lived with their partner and children and women who lived alone (p = 0.008), as well as between participants who lived with their partner and women who lived alone (p = 0.045). Finally, in relation to the type of perinatal loss (H(2) = 7.059, p = 0.029), there were significant differences between women who experienced early miscarriage and women who experienced fetal or neonatal death (p = 0.021) (see Table 3).

The objective of the multiple regression model applied was to assess the effect that levels of shame and self-judgment (self-criticism, isolation, and over-identification) have on the anxiety and depressive symptoms of the sample. The remaining self-judgment components were included in the initial stepwise regression model but excluded automatically during the selection process due to non-significant contribution to the explained variance. As shown in

Table 4. Stepwise Multiple Linear Regression analysis of predictors of anxiety symptoms.

	Predictors of Anxiety Symptoms
	Model I			Model II
	β	R2 (R2adj)	F(df)	β	R2 (R2adj)	F(df)
(Constant)	3.249 *	0.218 (0.217)	139.421 (1.499) *	0.154	0.284 (0.281)	98.582 (2.498) *
Shame	0.339 *			0.221 *
Isolation				1.655 *

* p < 0.001. β = standardized regression coefficient; R² = coefficient of determination; adjusted R² = adjusted coefficient of determination; F(df) = F statistic with corresponding degrees of freedom.

, isolation and shame were the most significant predictors of anxiety symptoms and had a positive relationship with anxiety. As shown in

Table 5. Stepwise Multiple Linear Regression analysis of predictors of depression symptoms.

	Predictors of Depression Symptoms
	Model I			Model II
	β	R2 (R2adj)	F(df)	β	R2 (R2adj)	F(df)
(Constant)	−0.935	0.270 (0.269)	184.661 (1.499) *	−1.275	0.327 (0.324)	121.000 (2.498) *
Isolation	3.049 *			2.156 *
Shame				0.220 *

* p < 0.001. β = standardized regression coefficient; R² = coefficient of determination; adjusted R² = adjusted coefficient of determination; F(df) = F statistic with corresponding degrees of freedom.

, isolation and shame were the most significant predictors of depressive symptoms, explaining that when self-criticism and shame increase, so do symptoms of depression.

3.3. Effects of Psychosocial Predictors Controlling for Clinical and Demographic Variables

To examine the robustness of psychosocial predictors (e.g., shame, isolation, self- criticism, over-identification) of anxiety and depressive symptoms, ANCOVAs were conducted, controlled for relevant clinical (e.g., type and number of losses) and sociodemographic variables (e.g., age, professional situation, socioeconomic status). For anxiety symptoms, only shame (p < 0.001), isolation (p = 0.001), over-identification (p = 0.048), professional status (p = 0.036), academic qualifications (p = 0.013) and household (p = 0.005) remained significant predictors after adjustment. Similarly, for depressive symptoms, only shame (p < 0.001), isolation (p < 0.001), professional status (p = 0.042), academic qualifications (p = 0.003) and household (p = 0.001) remained significant predictors after adjustment.

4. Discussion

The results show that this sample has high psychopathological symptoms, above what would be expected, especially symptoms of anxiety and depression. We conclude that this result is associated with the experience of perinatal loss, influenced by the time that has passed since the event and factors such as shame, isolation, over-identification, and self-criticism. In this study, different types of perinatal loss—from early miscarriage to neonatal death—were analyzed together. This approach follows definitions commonly used in the literature and allows for a general assessment of psychological symptoms following perinatal loss, while acknowledging that specific types of loss may have distinct clinical and psychosocial impacts.

4.1. Psychopathological Symptoms

The mean scores for anxiety and depressive symptomatology in all groups in our study were higher than expected when compared to the nonclinical community sample in the literature [44]. Our findings are supported by those of Bennett et al. [19], who found that women who had experienced perinatal loss reported anxiety and depression levels similar to those of psychiatric patients.

According to the literature, anxiety after perinatal loss is often driven by concerns about future reproductive outcomes [45] and feelings of “biological failure” related to the pregnancy’s non-viability [46]. In our study, participants who had experienced their most recent loss between seven and 18 months prior to participating reported significantly higher anxiety symptoms than those who had experienced their most recent loss between 19 and 30 months prior. These results contrast with studies indicating that anxiety symptoms typically decline within six months and may fully remit after one year [47]. However, our findings align with those of Mendes et al. [5], who observed a resurgence of anxiety symptoms seven to twelve months post-loss.

Similarly, the group of participants who had experienced loss 7 to 18 months prior to participating in this study exhibited high levels of depression, though there were no significant differences between the groups. Our results support those of Mendes et al. [5] and DeMontigny et al. [48], who reported an increase in depressive symptoms one to two years after gestational loss. However, Cumming et al.’s [49] study revealed a decrease in depressive symptoms between six and 13 months after the loss, though symptoms did not completely remit after one year.

These results may be explained by the emotional impact of symbolic dates, such as the anniversary of the loss or the baby’s first birthday [5,50]. The significant levels of anxiety and depression observed in this study highlight the importance of mental health screenings for women who have experienced perinatal loss, regardless of how much time has passed since the event. Although symptoms slightly decrease over time, shame and self-criticism remain elevated, showing that the psychological impact of the loss persists for years after the event. These results highlight the need for long-term psychological follow-up after a perinatal loss.

4.2. Influence of Psychosocial and Sociodemographic Factors on Psychopathological Symptoms

Regarding the psychosocial and sociodemographic factors of the sample, it was possible to verify that variables such academic qualifications, place of residence, socioeconomic level, household, and type of perinatal loss experienced contributed to the presence of anxious and depressive symptoms. Household composition was also associated with different levels of psychological distress. Specifically, women who lived alone exhibited depressive and anxious symptoms compared to those who lived with their partner and children. The scientific literature affirms that living alone is associated with poorer mental health, including elevated symptoms of depression and generalized anxiety [51,52]. Participants with higher levels of education reported fewer symptoms of depression and anxiety compared to those with a high school diploma or less. Finally, late perinatal losses were associated with higher levels of anxiety and depression than early miscarriages. This result is corroborated by Herbert et al. [15], who state that women who suffered losses during later stages of pregnancy had higher levels of anxiety and depression.

The interpretation of these findings should also consider culturally relevant factors within the Portuguese context. Research conducted in Portugal highlights the importance of culturally grounded bereavement rituals in the emotional adjustment to perinatal loss. Oliveira et al. [53] found that the perceived usefulness of bereavement rituals among Portuguese women who experienced pregnancy loss was associated with more adaptive bereavement experiences, underscoring the role of cultural practices in meaning-making and emotional processing. The same study emphasized the relevance of marital satisfaction, showing that lower satisfaction within the couple relationship was associated with more difficult bereavement experiences, reflecting the central role of conjugal and family bonds in Portuguese culture. These findings are particularly relevant to the present results, which indicate lower psychological distress among married women and those living with their families. Moreover, perinatal grief is often socially silenced and disenfranchised, with cultural expectations minimizing the legitimacy of grief, especially in cases of earlier gestational loss [54]. Such cultural dynamics may contribute to feelings of isolation, shame, and self-criticism, which emerged as key predictors of psychopathological symptoms in this study. Taken together, these cultural and relational factors highlight the need for culturally sensitive and family-oriented psychological support following perinatal loss in Portugal, integrating ritual recognition, couple support, and validation of grief experiences within healthcare settings.

It is important to note that sociodemographic variables such as education, income, and household composition were interrelated in our sample. However, the self-judgment and shame measures did not show significant correlations with these demographic characteristics, indicating that these psychosocial factors operate independently of sociodemographic context and represent individual-level predictors of anxiety and depressive symptoms.

4.3. Self-Judgment and Shame

Given the results presented, we can conclude that several factors may have influenced the presence of anxious and depressive symptoms in the participants. This study found that participants with higher levels of self-criticism, isolation, and over-identification after the experience of perinatal loss had elevated levels of anxious and depressive symptoms. This result is consistent with the research by Galhardo et al. [26], which showed that individuals who were more self-critical, isolated or disconnected from others, and over-identified with their negative emotional states, showed increased vulnerability to depressive symptoms. In addition, Neff [25] states that the negative components of self-compassion can lead to psychopathological outcomes such as depression and anxiety. Kamravamanesh et al. [55], in a study of pregnant women, showed that self-criticism, isolation, and over-identification were positively related to depression. The results of these investigations are consistent with those of our study. In addition, our results indicate that a tendency to feel isolated during distressing situations was the strongest predictor of psychopathological symptoms. According to Bellhouse et al. [56], women who experience pregnancy loss report feelings of loneliness and isolation, which can persist for more than four months after the event [57]. Our results can be explained by the fact that the family, friends, and community of these bereaved women may have difficulty understanding, may not know how to approach the subject of perinatal loss or may not recognize the grieving process [58,59]. The stigma felt by these women leads them to feel increasingly alone, and to repress their emotions, leaving them emotionally and psychologically vulnerable and at high risk of anxiety and depressive symptoms.

Furthermore, over the last few decades, several studies have shown that shame is associated with diverse psychopathological symptoms and disorders in clinical and non- clinical samples, specifically depression [60,61] and anxiety [22]. It is known that shame is an emotion that involves a negative perception of the individual, creating psychological vulnerability [62]. Our study showed that shame, like isolation, was a significant predictor of anxiety and depression, explaining that when shame increases, psychopathological symptoms also increase. Regarding the role of shame in the psychopathological symptoms of women who suffer perinatal loss, the results in the literature are not very consistent. Although our study assessed general feelings of shame, previous research suggests that women who experience perinatal loss may feel shame in specific social contexts, such as when confronted with other pregnant women or families with children [28,31,63]. These social comparisons might contribute to symptoms of depression and anxiety, which could partially explain the associations observed in our sample.

Several limitations of the present study should be acknowledged. First, the cross-sectional design prevents conclusions about causality and the evolution of psychological symptoms over time. Longitudinal studies would be needed to better understand changes in anxiety, depression, shame, and self-criticism following perinatal loss. Second, the use of an online survey may have introduced selection bias, as participants with internet access and a higher motivation to share their experiences could be overrepresented. Third, although the sample size of 501 participants is relatively large, it may not fully represent the diversity of all Portuguese women who have experienced perinatal loss, limiting generalizability. Fourth, other unmeasured variables, such as previous psychiatric history, the quality of social support, cultural and familial factors, and coping strategies, could have influenced the results. Fifth, the unification of different types of perinatal loss in the same study, from early miscarriage to neonatal death, may mask important differences in psychological impact. Another limitation concerns the external validity of the study. Since the sample consisted exclusively of Portuguese women, the generalizability of the findings to other cultural or national contexts may be limited. Cultural, social, and healthcare-related factors can influence the psychological impact of perinatal loss, as well as coping strategies and access to support. Nevertheless, the findings provide valuable insights into the persistence of psychological symptoms and key psychosocial predictors, which may inform both local and international research and interventions, particularly in contexts with similar social and healthcare structures.

5. Conclusions

The results of this study show that Portuguese women who have experienced perinatal loss show higher levels of psychopathological symptoms, particularly anxiety and depression, than would be expected in the general population. Not only do these symptoms persist over time, they may also intensify between seven and 18 months after the loss, which contradicts some of the literature that suggests a natural decrease after one year. The study also found that psychosocial and sociodemographic factors, such household composition, educational level, and type of loss, significantly influence the presence of anxiety and depression. Married women, those with higher levels of education, and those with stronger social support networks experienced less psychological distress. Conversely, later-stage losses were associated with greater vulnerability. Additionally, internal variables such as self-judgment, shame, isolation and over-identification were identified as significant predictors of psychological distress. Of these, isolation was the strongest factor, suggesting that difficulty in sharing grief, coupled with stigma and social misunderstanding surrounding perinatal loss, contributes to the persistence of symptoms. Shame also proved to be a critical factor, further increasing vulnerability to depression and anxiety, in line with international research.

Despite some limitations, such as the cross-sectional design and online data collection, the findings highlight the urgent need to recognize the psychological impact of perinatal loss and to develop preventive and intervention strategies. Systematic mental health screenings after perinatal loss are essential for the early detection of anxiety and depression symptoms and for referring women to specialized psychological support, particularly in primary care and obstetric settings. In summary, this research reinforces the idea that perinatal loss is not only an emotionally painful event, but that it may also carry a significant mental health risk. Therefore, it is crucial to invest in training for healthcare professionals and creating support networks to ensure that no woman experiences this alone.