Perinatal Mental Disorders in a University Clinic Setting: A Retrospective Study

Abstract

(1) Background: Women in the peripartum period are at increased vulnerability to the development or worsening of mental disorders. However, data from specialized psychiatric settings in Southeast Europe on this topic remain limited. In this study, we assessed the distribution of perinatal mental health disorders, diagnostic changes, hospitalizations, obstetric complications, and breastfeeding among women treated at the Perinatal and Reproductive Psychiatry Unit in Belgrade, Serbia. (2) Methods: A retrospective review of medical records from 287 patients (2015–2024), including sociodemographics, ICD-10 diagnoses, hospitalizations, obstetric complications, and breastfeeding data, was performed. (3) Results: The mean age of the women in this cohort was 33.1 ± 5.6 years. Mood disorders (F30–F39) were most common (39.6% before, 42.2% during, and 32.1% after pregnancy), followed by neurotic and stress-related disorders (F40–F48). Postpartum F50–F59 diagnoses (especially mental and behavioral disorders associated with the puerperium, not elsewhere classified—F53) rose significantly (p < 0.001). Overall, 38.7% of patients had at least one psychiatric hospitalization; hospitalization was recorded before pregnancy in 30.0%, during pregnancy in 5.9%, and postpartum in 6.6%, with overlap between periods. Leading obstetric complications were emergency cesarean (1.0%) and fetal loss (1.4%). Only 41.3% of the women whose data were available breastfed (mean 4.7 months). (4) Conclusions: Mood disorders predominate in specialized psychiatric perinatal care. The results of this study indicate the frequency of perinatal mental disorders and the need for both outpatient and inpatient treatment. It is, therefore, essential to continue developing mental health care systems for women in Serbia and to provide ongoing education for health professionals in order to ensure timely diagnosis and treatment of pregnant and postpartum women.

1. Introduction

The perinatal period comprises the period of pregnancy and the first year after giving birth; it represents a period of high vulnerability to the onset or exacerbation of mental disorders [1]. The term perinatal mood and anxiety disorders (PMAD) is often used as an umbrella term to encompass mood disorders (major depressive disorder, bipolar affective disorder, and postpartum psychosis) and anxiety disorders (anxiety and panic disorders, obsessive–compulsive disorder, and post-traumatic stress disorder) [2]. Contrary to the popular belief that pregnancy is a period of emotional well-being and light-heartedness, the perinatal period brings about unique challenges from psychological and physiological perspectives, including hormonal fluctuations, fatigue, insomnia, and possible breastfeeding problems, to changes in partner and family roles [3,4].

The etiology of perinatal mental disorders is multifactorial, involving an interplay of biological, psychological, and social factors [5]. Biological contributors include genetic vulnerability, epigenetic biomarkers, and profound neurobiological changes occurring during pregnancy and the postpartum period. These changes affect the endocrine system, particularly via fluctuations in estrogen, progesterone, thyroid hormones, and oxytocin, as well as the neurosteroid system, including cortisol regulation, and immune system functioning. Psychosocial risk factors are present across the antenatal, perinatal, and postnatal periods. The antenatal risks are anxiety, personal and/or family history of mood disorders, intimate partner violence, low financial income or educational level, single motherhood, and teen or unintended pregnancies. A peripartum factor could be a traumatic delivery experience. Postpartum psychosocial risk factors for PMAD include breastfeeding problems, depression or anxiety during pregnancy, perinatal loss, preterm birth, and traumatic birth experiences, among others. Critically, a low level of social support is an antenatal, perinatal, and postnatal risk factor [1,6,7,8,9].

Epidemiological data on the prevalence and distribution of perinatal mental disorders remain inconsistent throughout the existing literature. Differences are related to applied diagnostic criteria of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5-TR) [10] or International Statistical Classification of Diseases and Related Health Problems (ICD-10 or ICD-11) [11,12]; the definition of the postpartum period, which can range from 4 weeks to a year; and the availability of psychiatric care, among others. The available data show that around 20% of pregnant women develop some form of mental disorder [13]. It has also been shown that the prevalence of postnatal depression is 17% [14], that of antenatal anxiety is 15–20%, and the prevalence of postnatal anxiety is 15% [15]. Additionally, the incidence of the first episode of postpartum psychosis ranges from 0.25 to 0.6 in 1000 births [16].

A few studies have evaluated the prevalence of peripartum depression in Serbia [17,18,19] or peripartum anxiety disorders, including childbirth-related post-traumatic stress disorder [9,20,21,22]. Unfortunately, there are no prospective studies with a large enough sample size to be used to estimate the prevalence or incidence in the global population. The first center for perinatal and reproductive psychiatry in Serbia and the Balkans was opened at the Institute of Mental Health in Belgrade in 2011 [23]. Since then, many national programs and education courses have been organized for the whole region, besides everyday clinical practice. By applying individualized, highly specialized, diagnostic, and therapeutic protocols, the unit provides care to psychiatric patients who are preparing for pregnancy, as well as patients who have developed psychiatric disorders during pregnancy or within one year after childbirth [23]. For some of these women, this is the first time they have received help from the institute’s unit, whilst others have received outpatient or inpatient psychiatric care before.

The aim of this study was to determine the distribution of perinatal mental health disorders among patients treated at the Perinatal and Reproductive Psychiatry Unit. Additionally, changes in the courses of mental disorders during the perinatal period, as well as the frequency of psychiatric hospitalizations, were registered. Further objectives included identifying gynecological and obstetric complications in the perinatal period, as well as the frequency and duration of breastfeeding among postpartum women.

2. Materials and Methods

This was a retrospective study spanning nine years. Data from the medical records of the Unit for Perinatal and Reproductive Psychiatry at the Institute of Mental Health in Belgrade were collected between January 2015 and December 2024. The study included 287 patients treated at the unit, some of whom had been previously treated at the institute due to mental health issues unrelated to the perinatal period. Others were referred from other institutions or received psychiatric treatment for the first time.

This study was approved by the Ethics Committees of the Institute of Mental Health (ethical approval number: 1060/2075/L).

Psychiatric diagnoses were assessed for all patients in the periods before, during, and after pregnancy. It is of particular importance to define the temporal criteria for the periods before and after pregnancy in our work, given that inconsistency in terminology and time frames for peripartum mental disorders within diagnostic criteria is a common problem. In our study, the diagnosis “pre-pregnancy” refers to the last diagnosis established before the patient conceived. Conversely, although ICD-10 specifies the postpartum period as six weeks after delivery, in clinical practice—and therefore in our study—we applied the mental and behavioral disorders associated with the puerperium, not elsewhere classified—F53 diagnosis for the period of up to one year after childbirth. The diagnosis “during pregnancy” refers to the last diagnosis recorded before delivery.

Given that this is a clinical sample, any time frame restrictions would have substantially reduced the number of included patients. Therefore, we did not apply narrow temporal windows and instead sought to include as many clinically examined patients as possible without imposing time restrictions. The pre-pregnancy diagnosis was of particular interest to us because of the personal treatment history and the assessment of prior treatment over the patient’s lifetime. During pregnancy, the women were examined in all three trimesters, and after pregnancy, they were followed throughout the entire first year.

Diagnoses were established clinically according to ICD-10 criteria on the basis of psychiatric examination and medical history. A structured diagnostic interview (e.g., SCID) was not administered, and self-report instruments such as the DASS-21 were not used to assign diagnoses.

Mental disorders were analyzed according to patient age by stratifying participants into three groups: ≤30 years, 31–40 years, and ≥41 years. The participants were grouped into the stated age categories on the assumption that pregnancy, at different ages, has different effects on a person’s mental health status, i.e., that the frequency of mental disorders differs in pregnant women at younger, middle, and older ages [24].

Data analysis was carried out using methods of descriptive and analytical statistics. For measures of central tendency (arithmetic mean, median) and measures of dispersion (standard deviation, percentiles), absolute and relative numbers were used. The normality of distribution was tested using graphical and mathematical methods. Cochran’s Q test and the McNemar test (for nominal data) were used for the comparison of statistically significant differences between data before, during, and after childbirth. Statistical significance was established at the level of 0.05. A tabular and graphical presentation of the results was provided.

Missing sociodemographic data were not imputed. Descriptive analyses were based on available observations for each variable, and the corresponding denominators and numbers of missing values are reported in the table footnotes.

Each ICD-10 diagnostic group was analyzed as a separate binary variable indicating the presence or absence of the corresponding diagnosis. More than one diagnostic group could be recorded for the same patient; therefore, the diagnostic categories were not mutually exclusive.

Cochran’s Q test was performed separately for each ICD-10 diagnostic group to compare the presence of the corresponding diagnosis before pregnancy, during pregnancy, and in the postpartum period. Repeated-measures analyses included patients with available diagnostic information in all three periods (n = 238). When Cochran’s Q test was statistically significant, pairwise comparisons were performed using McNemar’s exact test, with Bonferroni adjustment for three pairwise comparisons.

Differences in the distribution of psychiatric diagnostic categories among age groups were assessed using Pearson’s chi-square test. Analyses were performed using available data, without imputation of missing values. The number of observations included in each analysis is reported in the corresponding table or table footnote. All tests were two-sided, and p-values < 0.05 were considered statistically significant. Age-stratified analyses were considered exploratory, and the reported p-values were not adjusted for multiple comparisons.

The SPSS 21.0 (IBM, Armonk, NY, USA) software package was used for data processing.

3. Results

3.1. Sociodemographic Characteristics

In this study, 287 participants were included. The mean age of the women at examination was 33.1 ± 5.6 years. Most participants lived in the capital city Belgrade (214; 74.6%), and most of them had graduated secondary school (grade 12) (109, 38.0%) or had a university degree (102; 35.5%). Over half of the subjects were married (150; 53.6%), and the majority were permanently employed (147; 57.0%) (

Table 1. Sociodemographic characteristics of the sample.

		n (%)
Place of residence	Belgrade	214 (74.6)
	Other	73 (25.4)
Education	Uncompleted primary school	5 (1.7)
	Primary school	12 (4.2)
	Secondary school	109 (38)
	College	25 (8.7)
	University	102 (35.5)
Marital status	Single	49 (17.5)
	Married	150 (53.6)
	Divorced	15 (5.4)
	Widowed	5 (1.8)
	Cohabitating	61 (21.8)
Employment	Unfit for work	2 (0.8)
	Unemployed	85 (32.9)
	Works, not registered	1 (0.4)
	Employed (fixed-term)	17 (6.6)
	Employed (permanent)	147 (57)
	Student	3 (1.2)

Note: Analyses were performed using available data without imputation. Denominators vary by variable. Education: n = 253 available, 34 missing. Marital status: n = 280 available, 7 missing. Employment: n = 255 available, 32 missing. Percentages are calculated using the available n for each variable.

).

The patients included in the analysis were distributed across three age groups: ≤30 years (n = 96), 31–40 years (n = 168), and ≥41 years (n = 23). We examined the associations between age group and education, marital status (including widowed), and employment status using the χ² test (significance threshold p < 0.05); no statistically significant differences were detected between the groups. University and secondary school education were the most frequent levels of education attained in all age groups. Being married was the most common marital status in all groups (50.5% among patients aged ≤30 years, 56.2% among those aged 31–40 years, and 47.8% among those aged ≥41 years). Cohabitation was reported by 25.3%, 21.0%, and 13.0% of patients, respectively, while unemployment was reported by 43.9%, 27.8%, and 31.8%, respectively (

Table 2. Sociodemographic variables by age.

Variable	Category	≤30	31–40	≥41	p Value (χ2)
		n (%)	n (%)	n (%)
Education	Uncompleted primary school	0 (0.0)	5 (3.4)	0 (0.0)	0.198
	Primary school	8 (9.6)	3 (2.0)	1 (4.5)
	Secondary school	35 (42.2)	64 (43.2)	10 (45.5)
	College	8 (9.6)	16 (10.8)	1 (4.5)
	University	32 (38.6)	60 (40.5)	10 (45.5)
Marital status	Single	19 (20.0)	25 (15.4)	5 (21.7)	0.195
	Married	48 (50.5)	91 (56.2)	11 (47.8)
	Divorced	3 (3.2)	8 (4.9)	4 (17.4)
	Widowed	1 (1.1)	4 (2.5)	0 (0.0)
	Cohabitating	24 (25.3)	34 (21.0)	3 (13.0)
Employment	Unfit for work	1 (1.2)	1 (0.7)	0 (0.0)	0.091
	Unemployed	36 (43.9)	42 (27.8)	7 (31.8)
	Works, not registered	1 (1.2)	0 (0.0)	0(0.0)
	Employed (fixed-term)	8 (9.8)	7 (4.6)	2 (9.1)
	Employed (permanent)	34 (41.5)	100 (66.2)	13 (59.1)
	Student	2 (2.4)	1 (0.7)	0(0.0)

Note: Analyses used available data without imputation: education, n = 253 (34 missing); marital status, n = 280 (7 missing); and employment, n = 255 (32 missing). Percentages were calculated using the available observations for each variable within each age group.

).

3.2. Psychiatric Characteristics

Significant differences in the distribution of psychiatric diagnoses were observed for F20–F29, F30–F39, F40–F48, and F50–F59. Mood disorders (F30–F39) were the most common diagnoses in our sample, followed by neurotic, stress-related, and somatoform disorders (F40–F48). The lowest number of patients were diagnosed with intellectual disabilities (F70–F79) and mental and behavioral disorders due to psychoactive substance use (F10–F19) (

Table 3. Psychiatric diagnoses before, during, and after pregnancy.

	Before Pregnancy n (%)	During Pregnancy n (%)	After Pregnancy n (%)	p Value
F10–F19	12 (4.4)	8 (2.8)	5 (2)	0.078
F20–F29	47 (17.2)	55 (19.5)	27 (10.7)	<0.001
F30–F39	108 (39.6)	119 (42.2)	81 (32.1)	0.036
F40–F48	78 (28.6)	90 (31.9)	47 (18.7)	0.003
F50–F59	9 (3.3)	4 (1.4)	35 (13.9)	<0.001
F60–F69	36 (13.2)	38 (13.5)	28 (11.1)	0.649
F70–F79	4 (1.5)	3 (1.1)	2 (0.8)	0.472

Note: Each ICD-10 diagnostic group was analyzed as a separate binary variable indicating the presence or absence of the corresponding diagnosis. More than one diagnostic group could be recorded for the same patient; therefore, percentages across diagnostic groups do not sum to 100%. Percentages were calculated using the number of patients with available diagnostic information in each period: 273 patients before pregnancy, 282 patients during pregnancy, and 252 patients in the postpartum period. Cochran’s Q analyses were based on 238 patients with available diagnostic information in all three periods. F10–F19 (ICD-10): Mental and behavioral disorders due to psychoactive substance use; F20–F29 (ICD-10): Schizophrenia, schizotypal and delusional disorders; F30–F39 (ICD-10): Mood [affective] disorders; F40–F48 (ICD-10): Neurotic, stress-related and somatoform disorders; F50–F59 (ICD-10): Behavioral syndromes associated with physiological disturbances and physical factors; F60–F69 (ICD-10): Disorders of adult personality and behavior; F70–F79 (ICD-10): Intellectual disability.

). There was a marked increase in the behavioral syndromes associated with physiological disturbances and physical factors (F50–F59) group before pregnancy, during pregnancy, and after childbirth. F50–F59 diagnoses after delivery mainly occurred due to the higher frequency of mental and behavioral disorders associated with the puerperium (F53) in the postpartum period.

Following statistically significant Cochran’s Q tests, pairwise comparisons were performed using the exact McNemar test with Bonferroni adjustment. For F20–F29 disorders, a statistically significant difference was observed between pregnancy and the postpartum period (adjusted p < 0.001), whereas the remaining pairwise comparisons were not statistically significant. For F30–F39 disorders, none of the pairwise comparisons remained statistically significant after Bonferroni adjustment. For F40–F48 disorders, a statistically significant difference was observed between pregnancy and the postpartum period (adjusted p = 0.008). For F50–F59 disorders, statistically significant differences were observed between the pre-pregnancy and postpartum periods and between pregnancy and the postpartum period (both adjusted p < 0.001).

Overall, 111 (38.7%) patients had at least one psychiatric hospitalization, whereas 176 (61.3%) were treated exclusively on an outpatient basis. Psychiatric hospitalization before pregnancy was recorded in 86 (30.0%) patients, hospitalization during pregnancy in 17 (5.9%), and postpartum hospitalization in 19 (6.6%). The hospitalization periods were not mutually exclusive: nine patients were hospitalized in two periods, and one patient was hospitalized in all three periods (

Table 4. Psychiatric hospitalizations in the periods before, during, and after pregnancy.

Treatment Period or Category	n (%)
Before pregnancy	86 (30.0)
During pregnancy	17 (5.9)
Postpartum	19 (6.6)
Any psychiatric hospitalization	111 (38.7)
Outpatient care only	176 (61.3)
Total	287 (100.0)

Note: Hospitalization periods were not mutually exclusive. Values represent unique patients hospitalized in each period. Ten patients were hospitalized in more than one period; therefore, period-specific frequencies should not be summed.

).

In the exploratory age-stratified analyses, a difference in the postpartum frequency of F30–F39 diagnoses was observed among age groups (unadjusted p = 0.028). Postpartum F30–F39 diagnoses were recorded in 18 (22.2%) patients aged ≤30 years, 53 (35.1%) patients aged 31–40 years, and 10 (50.0%) patients aged ≥41 years. However, this isolated finding should be interpreted cautiously because multiple age-stratified comparisons were performed. No other age-stratified comparison had an unadjusted p-value below 0.05 (

Table 5. Psychiatric diagnoses according to age.

Diagnostic Category	Period	≤30 n (%)	31–40 n (%)	≥41 n (%)	p Value (χ2)
F10–F19	Before pregnancy	5 (5.5)	6 (3.8)	1 (4.5)	0.810
	During pregnancy	3 (3.2)	5 (3)	0 (0.0)	0.692
	Postpartum	2 (2.5)	2 (1.3)	1 (5.0)	0.504
F20–F29	Before pregnancy	18 (19.8)	22 (13.8)	7 (31.8)	0.080
	During pregnancy	20 (21.3)	27 (16.4)	8 (34.8)	0.098
	Postpartum	9 (11.1)	13 (8.6)	5 (25.0)	0.083
F30–F39	Before pregnancy	38 (41.8)	60 (37.5)	10 (45.5)	0.674
	During pregnancy	39 (41.5)	71 (43.0)	9 (39.1)	0.925
	Postpartum	18 (22.2)	53 (35.1)	10 (50.0)	0.028
F40–F48	Before pregnancy	28 (30.8)	45 (28.1)	5 (22.7)	0.741
	During pregnancy	30 (31.9)	53 (32.1)	7 (30.4)	0.987
	Postpartum	14 (17.3)	31 (20.5)	2 (10.0)	0.487
F50–F59	Before pregnancy	2 (2.2)	6 (3.8)	1 (4.5)	0.758
	During pregnancy	1 (1.1)	3 (1.8)	0 (0.0)	0.739
	Postpartum	12 (14.8)	20 (13.2)	3 (15.0)	0.937
F60–F69	Before pregnancy	14 (15.4)	22 (13.8)	0 (0.0)	0.152
	During pregnancy	16 (17.0)	20 (12.1)	2 (8.7)	0.422
	Postpartum	9 (11.1)	14 (9.3)	5(25.0)	0.110
F70–F79	Before pregnancy	3 (3.3)	1 (0.6)	0 (0.0)	0.200
	During pregnancy	2 (2.1)	1 (0.6)	0 (0.0)	0.452
	Postpartum	1 (1.2)	1 (0.7)	0 (0.0)	0.822

Note: p-values were calculated using Pearson’s chi-square test. Percentages were calculated using the number of patients with available diagnostic information within each age group and period. Diagnostic information was available for 273 patients before pregnancy (≤30: n = 91; 31–40: n = 160; ≥41: n = 22), 282 patients during pregnancy (≤30: n = 94; 31–40: n = 165; ≥41: n = 23), and 252 patients in the postpartum period (≤30: n = 81; 31–40: n = 151; ≥41: n = 20). More than one diagnostic category could be recorded for the same patient. The age-stratified analyses were exploratory, and the reported p-values were not adjusted for multiple comparisons; therefore, the isolated unadjusted p = 0.028 finding should be interpreted cautiously.

).

3.3. Gynecological and Obstetric Complications

Gynecological and obstetric complications documented in the available psychiatric medical records are presented in

Table 6. Gynecological and obstetric complications documented in the study cohort (N = 287).

Gynecological Complications		Obstetric Complications
	n (%)		n (%)
Unintentional loss of fetus	4 (1.4)	Premature birth	1 (0.3)
Threatened miscarriage and high-risk pregnancy	1 (0.3)	Meconium aspiration syndrome	1 (0.3)
Induced abortion	1 (0.3)	“Traumatic” childbirth	2 (0.7)
Vascular complications	1 (0.3)	Bleeding in the puerperium	1 (0.3)
Infections	1 (0.3)	Emergency cesarean delivery	3 (1.0)
Positive prenatal diagnostics	1 (0.3)	Instrumental revision of the uterine cavus	1 (0.3)
Uterine deformities	1 (0.3)	Phenobarbital treatment of the neonatus	1 (0.3)
Medication-induced complications	1 (0.3)

Note: Percentages were calculated using the total study cohort (N = 287). As these data were obtained retrospectively from psychiatric medical records, the absence of a documented complication does not necessarily confirm that the complication did not occur.

. Unintentional pregnancy loss was documented in four patients (1.4%), while emergency cesarean delivery was documented in three patients (1.0%). Two patients (0.7%) described childbirth as traumatic. All remaining documented complications occurred in one patient each (0.3%). Because these data were collected retrospectively from psychiatric medical records, obstetric outcomes may not have been comprehensively documented for all patients.

3.4. Breastfeeding

Information on breastfeeding was available for 104 of the 287 patients (36.2%). Among patients with available breastfeeding information, 43 (41.3%) breastfed their infants, while 61 (58.7%) did not breastfeed. Breastfeeding information was unavailable for the remaining 183 patients (63.8%). Among women who breastfed and for whom breastfeeding duration was recorded, the mean duration of breastfeeding was 4.7 months, with a range of 1 to 15 months.

4. Discussion

In our sample of patients treated at the Unit for Perinatal and Reproductive Psychiatry, mood disorders were the leading indication for tertiary-level psychiatric care, followed by neurotic, stress-related, and somatoform disorders. These conditions can impair parenting and, beyond affecting postpartum women’s psychophysical health, may also have cognitive, behavioral, and somatic effects on newborns [25]. They can also strain partner relationships and affect broader social functioning. Evidence on diagnostic patterns among pregnant and postpartum women treated in highly specialized psychiatric settings is scarce. However, a Swedish study of pregnant women who screened positive on the Edinburgh Postnatal Depression Scale likewise found mood disorders to be most common, followed by anxiety disorders [26]. This aligns with their high prevalence in the general population [27], and our sample showed a similar overall pattern [28]. The nearly twofold higher frequencies of mood disorders in our cohort likely reflect the fact that our participants were clinical patients (seeking medical help at the institute), whereas this may be different if assessed in the general pregnant and postpartum population.

The distribution of mental disorders was different in the periods before, during, and after pregnancy. A distribution difference is noticeable within almost all disorder groups according to the ICD-10 [11], except for the personality disorders and disorders of adult behavioral (F60–F69) and intellectual disability (F70–F79) groups. In the mental and behavioral disorders due to psychoactive substance use group, pre-pregnancy patients comprised the majority of this sample (p = 0.078). Simultaneously, in the mental and behavioral disorders associated with the puerperium, not elsewhere classified, group, the frequency of disorders increased, which was expected since this group comprises patients with the mental and behavioral disorders associated with the puerperium, not elsewhere classified (F53) diagnosis. The diagnosis of mental and behavioral disorders associated with the puerperium, not elsewhere classified (F53.0–F53.9), refers specifically to mental disorders that occur in the puerperium, i.e., in the first 6 weeks after childbirth. “Postpartum requalification of disorders” is substantiated by a study conducted by Jones et al. [29]: when analyzing bipolar affective disorder as a risk factor for postpartum psychosis, they noticed an increase in the number of diagnoses from group F50–F59 and a simultaneous decrease in group F30–F39. A similar tendency was observed in pregnant women who experienced depressive phases prior to childbirth and developed postpartum depression thereafter [30]. The apparent postpartum shift from mood disorders F30–F39 to behavioral syndromes associated with physiological disturbances and physical factors F50–F59 should not be read as confirmed individual diagnostic conversion. Because this study is retrospective and based on the last recorded diagnosis in each period, we cannot distinguish true changes in psychopathology from changes in coding practice or clinical labeling. The results, therefore, indicate changes in the distribution of diagnostic categories, not patient-level diagnostic transitions. Part of this increase may reflect the structure of the ICD-10 classification and postpartum diagnostic reclassification rather than a true increase in psychiatric morbidity. In clinical practice, conditions previously coded within F30–F39 may be recorded as F53 after delivery, so the observed shift in diagnostic categories should not be interpreted as evidence of individual-level diagnostic conversion or of a genuine rise in the incident of illness.

A large proportion of our patients treated at the unit were initially diagnosed with F53 and later reclassified as F32. No statistically significant differences were found with respect to patient age within the behavioral syndromes associated with physiological disturbances and physical factors group (F50–F59) before pregnancy, during pregnancy, or after childbirth. Future studies would benefit from focusing exclusively on patients with diagnoses of mental and behavioral disorders associated with the puerperium, not elsewhere classified (F53). In clinical practice, this diagnosis is frequently applied beyond the six-week postpartum period, and it is therefore theoretically possible for a pregnant woman to receive an F53 diagnosis as well. The existing literature examining maternal age as a risk factor for postpartum depression is inconsistent: while most authors report a higher risk among younger women, others identify advanced maternal age as a significant risk factor [31]. It has also been suggested by some studies that postpartum women older than 40 years constitute the highest proportion of patients who develop peripartum mood disorders.

The present sample is a specialized clinical cohort of women already receiving psychiatric care, rather than a population-based cohort, which may partly explain the observed rates of diagnosis, treatment, and hospitalization. Population-based and service-level studies report a substantial burden of common mental disorders in pregnancy, as well as adverse obstetric and neonatal outcomes among women with pre-existing mental illness [32,33,34]. Direct comparison with these studies is limited by differences in sampling frames, diagnostic definitions, and healthcare systems; however, they consistently support the need for integrated perinatal mental health care and caution against generalizing clinical-sample estimates to the general pregnant population.

Psychiatric hospitalization was recorded more often before pregnancy than during pregnancy or postpartum. In the present sample, 86 patients (30.0%) had been hospitalized before pregnancy, 17 (5.9%) during pregnancy, and 19 (6.6%) postpartum. These periods were not mutually exclusive, as ten women were hospitalized in more than one period. Langan Martin et al. reported a decrease in hospitalization rates during pregnancy, with an increase during the first two weeks after delivery [28]. The comparatively small numbers of hospitalizations recorded during pregnancy and postpartum in our study may reflect several factors, including reluctance to consent to inpatient treatment, stigma, insufficient family support, concern about separation from the newborn, and the absence of specialized mother–baby units in Serbia. However, these explanations were not directly measured and should be regarded as hypotheses. Moreover, the findings describe a specialized clinical cohort and should not be interpreted as period-specific hospitalization prevalence in the general obstetric population. Because the study recorded whether a patient had been hospitalized in each period, rather than the number of admissions, repeated admissions within the same period could not be evaluated.

Obstetric complications at childbirth and the puerperium are much rarer compared with complications in pregnancy. A major Californian study including pregnant and postpartum women treated for psychiatric disorders reported that women treated for mental disorders are at a significantly higher risk than the general population of having obstetric complications during delivery [13,35]. In our study, the most common complications were emergency cesarean section and unintentional loss of fetus. Emergency cesarean delivery is recognized as one of the most traumatic birth experiences, and so it is associated with an increased risk of developing childbirth-related post-traumatic stress disorder (CB-PTSD) [36] and/or postpartum depression. Miscarriage or pregnancy loss constitutes a separate and significant risk factor for postpartum depression. The trimester in which pregnancy loss occurred in our study patients could not be determined, as this information was not available in the medical records. This limits a more precise interpretation of obstetric complications and should be addressed in future prospective studies.

In our study, fewer than half of the patients included breastfed their infants. A survey conducted in the United States reported that 86% of mothers in the general population breastfeed, with an average duration of three months [37]. The lower breastfeeding rate in our sample may reflect the specific features of a clinical psychiatric population rather than a directly comparable population-level difference. The lower rate in our sample is more consistent with evidence from women with severe mental illness than with general-population estimates: Baker et al. [38] found that women with SMI are less likely to initiate and continue breastfeeding, and that inconsistent professional advice and limited infant-feeding support further reduce breastfeeding rates. In contrast, Haight et al., in a longitudinal community cohort, reported high rates of breastfeeding, as well as a bidirectional association between perinatal depression and feeding [39].

It is known that the poorer a woman’s mental health is, the stronger the negative effects on exclusive breastfeeding and breastfeeding duration are [40,41]. The medical records available in our study showed that psychiatrists or gynecologists advised discontinuation of breastfeeding in some cases, mainly due to concerns about treatment adherence and potential medication side effects. The number and percentage of women for whom discontinuation was explicitly recommended could not be reported, because such advice was not systematically documented in the medical records. As this was a retrospective study based on existing clinical documentation, the finding remains qualitative and cannot be quantified.

Breastfeeding decisions should be individualized, balancing the risks of worsening maternal mental health from early cessation against possible medication effects on the infant. When breastfeeding is continued, careful monitoring of the infant’s psychophysical development is essential.

The high frequencies of mood and mental disorders during the perinatal period highlight the necessity for systematic monitoring of mental health in pregnant and postpartum women at all levels of healthcare. Screening for peripartum mental disorders should therefore become an integral component of routine obstetric care.

In Serbia, an important step in this direction was taken in 2026 with the publication by the Ministry of Health of the National Guide to Good Clinical Practice for Women During Pregnancy for Primary Care Physicians [42]. For the first time, clear recommendations for regular screening for peripartum depression during pregnancy to be carried out by gynecologists are provided. Preventive measures at the primary healthcare level include psychoeducation, the promotion of a healthy lifestyle, and provision of psychotherapy when indicated.

In this study, an overview of the distribution and progression of mental disorders in patients treated at the specialized psychiatric unit, as well as complications and discomforts associated with pregnancy and childbirth, has been provided. The main limitations of this study include its retrospective design and the fact that the sample was restricted to psychiatric patients exclusively. An additional limiting factor is the absence of a national registry for peripartum mental disorders, which could provide more detailed insights. A prospective study design is needed in order to assess the course of mental disorders and corresponding therapeutic approaches more comprehensively across the different phases of the peripartum period. In future studies, it is of particular importance to examine primiparous women separately from multiparous patients, especially when data on the previous treatment of peripartum mental disorders are available. Breastfeeding information was available for only 104 of the 287 patients. Therefore, the reported breastfeeding proportion refers exclusively to those patients with available lactation data and should thus not be generalized to the entire study cohort. In addition, gynecological and obstetric outcomes were obtained retrospectively from psychiatric medical records and may have been incompletely documented; this should be taken into account when analyzing these data. In future research, it would be relevant to use rating scales in addition to clinical diagnosis in order to achieve better objectification of the results.

5. Conclusions

In this study, mood and anxiety-related disorders (F30–F39 and F40–F48) were among the most frequently recorded diagnostic categories before, during, and after pregnancy. The results do not establish the prevalence of perinatal mental disorders in the general population of pregnant and postpartum women; rather, they show the distribution of mental disorders in a specialized Perinatal Disorders Unit.

Shifts in diagnostic categories after delivery, including an increase in the behavioral syndromes associated with physiological disturbances and physical factors (F50–F59), should be interpreted cautiously, as they may reflect ICD-10 coding and postpartum reclassification rather than a true increase in morbidity or individual-level diagnostic conversion. Hospitalization was more often recorded before pregnancy than during pregnancy or postpartum; explanations such as reluctance to consent to admission or the absence of mother–baby units remain hypotheses. Breastfeeding was less frequent than in general-population surveys, consistent with findings in women with severe mental illness, but recommendations to discontinue breastfeeding were not systematically documented.

Taken together, the findings of this study support the need for timely diagnosis, continuity of care, and a multidisciplinary approach to women with perinatal mental disorders. Systematic screening and further development of perinatal mental health services in Serbia and surrounding regions are clinically reasonable implications of this study and of the broader literature. Prospective studies with a standardized diagnostic assessment and individual-level diagnostic trajectories are needed to confirm these observations.