This is a quasi-experimental, quantitative cross-sectional study.

Table 1 presents the participants’ sociodemographic characteristics. Regarding pregnancy data, most of the sample included multiparous women (61.4%, n = 94), with no history of miscarriages (77.1%, n = 118) with expected normal delivery (69.9%, n = 107), unplanned pregnancy (68.6%, n = 105), but almost all reported wanting pregnancy (93.5%, n = 143). Regarding emotional/physical/psychological occurrences, 32% reported at least one, 26.1%, the presence of stressful or traumatic events, 34% emotional or physical complications, 11.1% threatened abortion; 10.5% smoking, 3.9% use of alcohol and 0.7% use of illicit substances.

Table 2 presents the sample characterization results. In the screening assessment for GAD, 22.7% (n = 34) are at risk.

As for mental disorders during pregnancy, 25.9% (n = 42) of the women assessed had PPD symptoms (M = 16.10, SD = 5.00). Women without PPD had a mean of 4.35 points (SD = 2.90) and the difference between the groups with and without depression was significant [F = 319.42 (1.149), p ≤ 0.001, n² = 0.683]. In the current SCID-5-CV manifestations, 18.5% (n = 28) had a major depressive episode, and 0.7% (n = 1) had a manic episode; 9.9% (n = 15) persistent depressive disorder; 10.3% (n = 9) bipolar type one; 1.4% (n = 2) non-active psychotic symptoms; 32.7% (n = 49) had a past episode of depressive mood and 5.3% (n = 8) manic.

Regarding anxiety disorders, 32% (n = 49) met diagnostic criteria for at least one disorder, with 21.5% (n = 32) generalized anxiety, 18.7% (n = 28) panic, 5.4 % (n = 8) current agoraphobia and 4.7% (n = 7) social anxiety; 8.7% (n = 13) for symptoms of adult attention deficit and hyperactivity disorder; 4.8% (n = 7) for current post-traumatic stress disorder and 12.5% (n = 12) for a past disorder and; 2.0% (n = 3) for obsessive-compulsive disorder.

With regard to the use of substances, “other drugs” (anti-allergy, contraceptives, among others) was reported with the highest prevalence, 24% (n = 36), followed by alcohol, 10.6% (n = 16) sedatives, hypnotics and/or or anxiolytics 7.9% (n = 12); opioids and nicotine 7.3% (n = 11) each; cannabis, other hallucinogens and stimulants 2% (n = 3) each. As for substance abuse, 6% (n = 9) reported alcohol, 2.7% (n = 4) nicotine, 2% (n = 3) sedatives, hypnotics and/or anxiolytics, with the same rate for opioids and 1.3% (n = 2) for “other drugs”. For substance use disorder, 6% (n = 9) met diagnostic criteria for alcohol and 2.1% (n = 3) for nicotine (n = 32). Other more frequent possible diagnoses were separation anxiety disorders (21.5%, n = 32) and specific phobias (18.8%, n = 28). Other possible disorders at screening included: intermittent explosiveness (16.9%, n = 25), insomnia (14.2%, n = 21), premenstrual dysphoric disorder (11.6%, n = 17).

As for the risk and protective factors for depression, Tables 3 and 4 show the coefficients of the logistic regression models. Protective factors include absence of emotional neglect in childhood (reduces 27%); and marital adjustment (reduces 21%). Risk factors: previous history of PPD (increases 76%); concerns about lack of confidence (increases 38%); and being multiparous (increases continuous EPDS score).

Only one protective factor stood out in terms of mood disorders: good marital adjustment (reduces 64%); risk factors: concerns about lack of confidence (increases 60%), moderate emotional abuse in childhood (increases 34%) and history of depression (increases 32%). In anxiety disorders, protective factors included: absence of stressful or traumatic events during pregnancy (reduces 21%); and below-average scores for concerns about future perspectives (reduces 8 percent). Risk factors: concerns about lack of confidence (increases 42%); history of emotional abuse (increases 20%) and relationship concerns (increases 13 percent).

Regarding substance use, the protective factors include: not smoking (reduces 50%); and absence of physical abuse in childhood (reduces 15%). On the other hand, risk factors include history of severe physical abuse (increases 55%) and depression in the family (increases 13%). For other mental disorders, the main risk factors include stress-producing financial issues (increases by 13%) and childhood trauma scores (increases by 10%).

The main objective of the study was to identify possible risks and protective factors of mental disorders during pregnancy. As for the sociodemographic results, another national study, carried out in the Midwest region, also found data similar to our sample; on the other hand, most participants had low education (Rezende et al., 2020). A study carried out in the northeast region of Brazil on the quality of life of pregnant women found that lower income would be linked to gestational dissatisfaction, which is consequently linked to gestational risk (Abreu et al, 2019). Such findings show population variability, and it is possible to find a more similar characterization according to regional proximity in the country (Monteiro Neto, 2014).

Among women’s relationships, the majority had good levels of dyadic adjustment; social support contributes to a more positive experience of motherhood (Piccinini et al., 2002), with the family domain being an important factor in assessing the quality of life during pregnancy (Abreu et. al, 2019). In this connection, conflictual relationships (Leite et al., 2020) and low levels of social support (M. L. C. Santos et al., 2022) appear as risk factors for depression.

The results indicated that even without a diagnosis, the prevalence of anxiety disorders was 32%, above depression disorders and in line with another national study (Schiavo et al., 2018). As for substance use, there was an association between cigarette use and substance use disorders. Not smoking cigarettes was noted as a protective factor. Unwanted pregnancy, depression in the family, stressful life events and physical abuse appeared as risk factors. In a Brazilian sample, among some social triggers for alcoholism, physical abuse was one of them (J. C. Lopes et al., 2020). Although studies indicate, in line with our study, that most pregnancies are unplanned (68.6% in this study and 66.7% in a sample from the Northeast), it is observed that most women (93.5% and 76.9 %, respectively) began to desire pregnancy after realizing they were pregnant (Abreu et al., 2019).

Some predictors for each model varied; however, some stood out for their frequency, such as lack of confidence assessed by the Worry Domain Questionnaire is a risk factor for five of the six outcomes. In another study, women reported more concern about lack of confidence than men. Women had a more negative guidance for the problem and greater thoughts suppression (Robichaud et al., 2003). Studies that specifically consider worrying styles as predictive variables of mental disorders still seem to be scarce. Thus, this study adds to the literature the assessment of cognitive factors.

History of depression in the family and some form of violence, as well as stressful life events appear as risk factors in different studies, including in ours. The total score assessed by the EVPE was also identified as a risk factor in two models. A Brazilian survey found an association between the occurrence of EVPE and Common Mental Disorders (CMD), depression and anxiety. With the exception of “death of a close relative”, all other life events were significantly associated with the presence of CMD, with “severe financial problems” having greater strength (C. S. Lopes et al., 2003).

It was found that some predictors appear more frequently in certain outcomes. As is the case of those predictors that are associated with anxiety disorders: presence of stressful and traumatic events during pregnancy and emotional abuse. One study identified their relationship with anxiety and depressive symptoms (Margis et al., 2003).

Stressful events during pregnancy also emerge as a risk factor for anxiety and depression (Kliemann et al., 2017). There are studies that emphasize the clinical overlap between stress, anxiety and depression (Martins et al., 2019); different findings in this sample, indicated a relationship only with anxiety diagnoses. Despite the fact that emotional abuse is recurrently associated with depression (Christ et al., 2019), it can also be associated with anxiety (Ragazzo, 2017).

As for the depression outcome, the risk predictors were a history of previous illness (Ferreira et al., 2018; Ramos et al., 2018), lack of trust and poor marital adjustment. Protective predictors were absence of emotional neglect and receiving material social support. In the CTQ factors, emotional neglect refers to the impossibility of effectively noticing, meeting or responding to the emotional needs of their children (I. M. L. Silva, 2019). Therefore, not having experiences of emotional neglect throughout the child development and perceiving social material support during pregnancy seem to be important variables for mental health.

For other current mental disorders, history of illness in the family, stress-producing life events, concerns about lack of confidence and financial problems, and childhood trauma were identified as risk factors and were also reported in the national (Arrais et al., 2018; Waikamp & Serralta, 2018) and international (Devi et al., 2019) literature. These findings are in line with the results of our study, with emotional and physical abuse associated with anxiety, mood, substances and, mainly, depression. Once again, the events that occurred in childhood appear as factors linked to mental health – emphasizing the importance of investing in programs that aim to encompass psychosocial and family factors from the beginning of the constitution of the first relationships (marital relationship, pregnancy, caregiver-baby interaction, child development, among others).

The SCID-5-CV instrument, referring to the GAD, showed results similar to those of the Worry Domains Questionnaire, with prevalence of 21.5% and 22.7%. The same occurred in relation to EPDS and SCID-5-CV for current depressive episode, 25.9% and 18.5% concomitantly. Therefore, these instruments parallel to the SCID-5-CV may be suitable for measuring symptoms, being self-report instruments and easy to apply.

This study contributed to the survey of risk and protective factors of mental disorders. Among the limitations we include: 1) marital status of the participants: most of them were in a relationship - being single is considered a risk factor; 2) survey in only one municipality makes the description restricted or liable of local influence. The sociodemographic data were not reported in the model with the same frequency as the data associated with the history, emotional and cognitive symptoms, and psychosocial and family variables. Sociodemographic issues seem to change according to the region studied, which may explain the variability. National studies covering different regions of the country are suggested. On the other hand, factors such as unplanned pregnancies persist even with diversity. The importance of validating instruments that assess cognitive issues is enhanced.

Despite the fact that some risk and protective factors have commonly appeared in some outcomes, others were specific to each outcome. Results like these show us the particularity and complexity of the development of mental disorders in pregnant women. It is of particular importance to the area of mental health, the factors that can be subjected to intervention, such as social support, marital relationship, history of traumatic events, beliefs regarding motherhood, among others. Identifying them contributes to the planning of preventive actions and interventions that help reduce the impacts of mental disorders on pregnant women and possible repercussions on their relationship with their baby and child development.