Parenting stress and its correlates in an infant mental health unit: a cross-sectional study

Abstract

Background: An infant’s development is closely linked to the relationship they have with their parents. Parenting stress, affective disorder, and an upbringing with substance-abusing parents can affect parenting quality and increase the risk of children developing behavioral, mental and social problems. The overall aim of the study was to investigate how parents of children attending an outpatient Infant Mental Health (IMH) unit rate their own psychological health and parenting stress, and to explore predictors of parenting stress.

Methods: The sample comprised 197 parents, 129 mothers and 68 fathers, referred with their infant/toddler to an outpatient IMH unit for interplay treatment. On admission, the parents completed self-report questionnaires concerning their own mental health problems and parenting stress.

Results: The mothers reported significantly more psychiatric symptoms and parenting stress than the fathers. Fathers with substance-abusing parents had often experienced divorce in the family of origin, had a low level of education, and had often experienced trauma. Depression was a predictor for parenting stress for both mothers and fathers.

Conclusion: The parents’ situation was strained, presenting a variety of psychiatric symptoms and high levels of parenting stress, making assessment of parental health before starting treatment important. The mothers’ situations were more serious compared with the fathers’, and for both parents depression was a significant predictor for parenting stress. To increase the chances of a positive treatment outcome for the child, both parents should be included in the treatment.

Keywords:

• Infant mental health
• parenting stress
• postnatal affective disorder
• gender

Background

An infant’s development is closely linked to the relationship they have with their parents, as positive emotional bonds are fundamental for an infant’s mental health. There is a general consensus that it is the parent’s quality of sensitivity to the small child, and their ability to attune to and affirm the child’s communication, that helps the child develop appropriate self-regulation, a sense of security, and an organized attachment [1,2].

The stress of any kind impacts parent-child relationships and children’s social-emotional development [3]. High levels of parenting stress, defined as ‘the disparity between the perceived demands of parenting and the resources parents have available to meet those demands’ [4], are particularly problematic because of their direct influence on parenting behavior and consequent child development [5–7]. Studies link parenting stress to harsher and less stimulating parent-child interactions, as well as greater likelihood of child maltreatment [8]. Parenting stress in the child’s earliest years also predicts internalizing and externalizing problems in young children [9].

The degree of parenting stress has been associated with several influential factors. Substance abuse in the family of origin is one such factor that can negatively affect an individual’s ability to develop good parenting skills [10,11]. Other factors are the parent’s affective and cognitive resources and symptoms of anxiety and depression in the child [12–14]. Studies of parenting stress in mothers, who generally experience higher stress levels than fathers [15,16], have shown high levels of perceived parenting stress to be associated with depression [17,18], perceived high workload, and lack of social support [19]. Other research has explored the relationship between a parent’s own experience of maltreatment [20,21], childhood trauma [22–24], and their subsequent parenting styles [25]. All these stress-related factors may negatively affect the ability to show responsiveness and sensitivity in relation to the child, and the self-image as a parent [4,8,26] and, in the long term, the child’s development [27].

The vast majority of studies have focused on parenting stress in mothers, but the subject of parenting stress in fathers is now attracting greater attention. A recently published study found that paternal involvement, directly and by reducing maternal parenting stress, was positive for infant neurodevelopment [28]. Studies have shown that a low sense of coherence and depression in fathers during early pregnancy can lead to increased parenting stress in the fathers [18,29]. Other factors that have been associated with parenting stress in fathers are a low level of education, unemployment, and mental illness [29,30]. Fathers who shared parental leave equally were less affected by perceived parenting stress than those who did not [31].

A growing research focus is on parenting in people with neurodevelopmental disorders, and a recently published meta-analytic review highlights difficulties in parenting with Attention Deficit Hyperactivity Disorder (ADHD) symptoms. ADHD-related deficits, such as poor inhibitory control and working memory, could be associated with increased stress and inconsistent approach to the child’s behavior and harsher and laxer parenting [32].

Sleep disruption, lack of social support in the parental role, and difficulties in the spousal relationship have been shown to be associated with parenting stress for both mothers and fathers [29,33].

Affective disorder in either parent is known to interfere with parenting quality and increase the risk of children developing behavioral, mental and social problems [34–36]. The risk of antenatal and postpartum depression in women is 10 ∼ 15% [37], and indicators of increased risk of postpartum depression may be found already in early pregnancy [38]. Most research focuses on the impact of maternal insufficiency on parenting, but some research on depression in fathers has shown a prevalence of 6–10% in the postnatal period [39–41]. Other studies find a similar association with developmental problems in the child [42–44]. Studies have found depressed fathers to be more withdrawn, unemotional, and less communicative with their children, which was found to have a negative influence on the child’s development [44–48]. A Swedish study found an association between depression in the fathers in the postnatal period and a low level of education, earlier depressive period, stress-filled life events, and problems in spousal relationships [40]. A meta-analysis found a positive correlation between depression in the mother and depression in the father [41].

Another frequent but underemphasized problem is anxiety disorders, with an incidence in mothers of about 8.5% in the postpartum period [49]. In general, mothers report more anxiety symptoms than fathers [50]. Anxiety is common during pregnancy [51] and is often a comorbid state in postpartum depression [52]. Studies have found a connection between a parent’s anxious and excessive controlling behavior towards their child and the increased risk of future anxiety in the child [53,54].

Important support for psychosocial risk families is Infant Mental Health (IMH) units. The aim of these units is to improve the parent-child relationship and to identify, at an early stage, psychiatric symptoms that might influence the parent’s caregiving ability. In most IMH units, treatment models offer interventions designed to support parental development in two areas: firstly, to develop greater sensitivity for their child’s needs and secondly, to increase their ability to reflect over their child’s expressed affect and intentions. The parents all have parenting dysfunctions, but as a group, they are heterogeneous. Individuals with severe affective disorders are offered supplementary treatment outside the IMH unit.

We recently published data on the mothers in an IMH unit [55], comparing those who reported substance abuse in their family of origin to the rest of the mothers. Our conclusions were that female offspring of parents with substance abuse seems to be an especially vulnerable group of patients in IMH units, with more symptoms of depression and ADHD compared with women who denied substance abuse in their family of origin. The parenting stress in the mothers was above the Swedish norm for women, but no differences were found between those with substance abuse in their family of origin and those without.

In the present study, using the same sample, we intended to investigate how both mothers and fathers rate their parenting stress and own psychological health, and to study predictors for parenting stress in the entire sample of parents.

Methods

This study focused on self-reported data from mothers and fathers participating with their small children in an IMH program in a specialized outpatient infant and toddler psychiatric clinic in Malmö, Sweden. The parents were treated in the unit because of deficiencies in parenting behavior and/or dysfunctional patterns of interaction with their child. The IMH unit belongs to the child psychiatric clinic, and parents of treated children are not diagnosed. All participants were referred for treatment either through self-referral (60%) or through referral by a doctor or nurse from primary care or pediatric clinic. Parents diagnosed with intellectual disabilities were not treated at the unit. Data were collected before intervention and stored in a database titled ‘Factors Important for Parenting’ according to the Personal Data Act at Skåne County Council. On admission, the parents were informed about the study by the administrator, who also collected consent forms if they agreed to participate in the study. The administrator was then responsible for distributing and collecting the instruments used.

Subjects

During the study period (1 May 2011 to 1 May 2013), a total of 177 families were admitted to the clinic. Exclusion criteria were psychosis or severe depression with hospitalization (n = 9), or inadequate knowledge of the Swedish language (n = 9), leaving 159 families eligible. Nine families declined participation, and a further 18 failed to complete the questionnaires. This left a total of 132 families, with 129 women and 68 men participating in the study. Forty (38%) of the fathers in the families where the mothers participated in treatment did not complete the questionnaires. The total sample was divided into groups according to gender, and fathers were subdivided into subgroups with (Index) and without (COMP) substance abuse in the family of origin: ‘Index fathers’ (n = 18) and ‘COMP fathers’ (n = 50).

Instruments

A self-reported questionnaire was specifically designed by the research group for this study. This covered sociodemographic and psychosocial risk factors, such as traumatic life events and somatic and psychiatric disorders, including substance abuse of the subjects and in the family of origin [55]. The participants were asked to provide a more detailed answer to any initial positive responses (e.g. ‘Did you or any family member in your family of origin suffer from substance abuse problems (alcohol, prescribed drugs, illegal drugs)? If yes, what kind?’ ‘Have you experienced a severe/intimidating/traumatic life event. If yes, which type of trauma’).

The Swedish Parenthood Stress Questionnaire (SPSQ) is based on the parent domain of the Parenting Stress Index [14]. The SPSQ measures parenting stress and comprises 34 items divided into five subscales: incompetence, role restriction, social isolation, spousal relationship problems, and health problems. The incompetence subscale covers general experiences of caregiving, feelings of incompetence in the parental role, and difficulties of parenthood. The role restriction subscale describes narrowing of interests and activities due to parental responsibilities, e.g. ‘Since I had children, I have almost no time to myself’. The social isolation subscale focuses on social contacts outside the family, e.g. ‘When I get into a problem situation with the children, I have many people I can turn to for help and advice’. Spousal relationship problems concern social experiences with the partner within the family. The final subscale, health problems, concentrates on physical fitness, infections, and fatigue caused by the demands of parenthood [56].

Together, these five subscales give an overall parenting stress score. The response options range from ‘strongly agree’ to ‘strongly disagree’ and are scored on a Likert-type scale from 1 to 5. Higher scores indicate more stress. The instrument has been validated in several studies and shown good psychometric properties [14,57], and stability over time [56,58]. In the present study, the results are compared with the results of a Swedish normal community sample [57]. No cutoff scores were established, but in previous research, items scored as a 4 or a 5 have been considered problematic [59].

The ASRS v. 1.1 Adult ASRS Full Edition (WHO Adult ADHD Self-Report Scale) is the WHO’s self-report rating scale for adult ADHD [60]. The scale consists of 18 items that are consistent with the DSM-IV diagnostic criteria for ADHD symptoms. Items 1 to 9 (Part A) concern symptoms of inattention, and items 10 to 18 (Part B) concern symptoms of hyperactivity or impulsivity. The internal consistency of the ASRS in this dataset on the basis of Cronbach’s alpha was .86 for Part A and .86 for Part B. ADHD symptoms were assessed using the Swedish version of the 18-item ASRS, using the cutoff level of ≥ 24.

The Hospital Anxiety and Depression Scale (HADS) have been shown to be valid and reliable in medical practice and research [61,62]. The HADS is a Likert-style questionnaire consisting of 14 items, seven for anxiety and seven for depression. The score ranges from 0 (no anxiety) to 3 (greatest anxiety) for each question [62]. The range of scores used to define cases are as follows: 0–7, normal; 8–10, mild disorder; 11–14, moderate disorder; and 15–21, severe disorder [61,63]. We used a cutoff score of ≥ 8 in our study [64]. The results were compared with the Swedish norm [65]. Previous studies have supported the two-factor model used in this study [66,67].

The Alcohol Use Disorders Identification Test (AUDIT) is an internationally well-validated questionnaire assessing alcohol use [68]. Alcohol consumption was measured using the Swedish version of the AUDIT. The instrument consists of ten questions, each of which is scored from 0 to 4, so the maximum score is 40. It contains questions relating to the level and frequency of alcohol consumption, heavy drinking, drinking behavior, and alcohol-related problems. Harmful use of alcohol is indicated by a score of 6-13 points for women and 8-15 points for men, and alcohol dependence is indicated by a score of ≥ 18 points. The recommended cutoff scores for alcohol abuse used in this study were ≥ 6 for women and ≥ 8 for men [69].

The Drug Use Disorders Identification Test (DUDIT) is a validated questionnaire evaluating the use of illegal drugs [70]. Patients at risk of drug use were identified using the Swedish version of the DUDIT. The instrument consists of eleven items designed to assess the consumption patterns of illicit drugs and associated problems. The first nine items are scored on a five-point scale ranging from 0 to 4, and the final two on a three-point scale with the values 0, 2, and 4. The total score, therefore, ranges from 0 to 44, with higher scores suggesting a more severe drug problem. The cutoff score for any type of problematic use (i.e. harmful use, substance abuse, or dependence) is generally recommended to be ≥ 6 for men and ≥ 2 for women. The reason for the different cutoffs for men and women is the greater vulnerability, socially and biologically, of women to drug use [70]. The different thresholds have been discussed. The following risk levels are suggested: no drug-related problems (total scores 0-5/1), possible drug-related problems, i.e. risky or harmful drug habits that might be diagnosed as substance abuse/harmful use or dependence (6/2-24), and probable heavy dependence on drugs (scores ≥ 25).

SPSQ, ASRS, HADS, AUDIT, and DUDIT have all been validated for the general Swedish population.

Statistical analysis

A power analysis was carried out before the study start. This showed that at least 50 families must be included to identify group differences on a 5% level with a power of 0.80. However, our study design meant that we had to settle for a lower number of men.

The data were analyzed using IBM SPSS 22.0 software for Windows. First, descriptive statistics were used to describe the findings. Two-sample t-tests and Pearson’s chi-square test for non-parametric data were used to identify significant differences between the whole group of women as compared to all men, and in the analysis of the ‘Index men’ compared with ‘COMP men’.

Multiple regression was used to analyze association between background variables and parenting stress for groups of women and men. The variables were chosen according to whether they showed a high correlation with parenting stress (symptoms of depression, Cohen’s D 0.60; symptoms of anxiety, Cohen’s D 0.44; traumatic life event, Cohen’s D 0.21). One variable, symptoms of ADHD (Cohen’s D 0.28), with fairly high correlation to parenting stress, was not used in the analysis of the fathers, as very few men had high scores in the ASRS test (Table 1). The multicollinearity test for the variables in the multiple regressions showed a low value (variance inflation factor VIF, of 1.1), although there was a high correlation between symptoms of depression and anxiety (Cohens D 0.39).

Table 1. Results of self-assessment instruments on psychiatric symtoms for mothers and fathers (n = 197).

	Mothers n = 129	Fathers n = 68	p-Value
SPSQ parenting stress	3.13 (SD = 0.6)	2.67 (SD = 0.6)	<.001^b
ASRS symptoms of ADHD	24 (19%)	3 (4%)	.006^a
HADS symptoms of anxiety	104 (81%)	23 (34%)	<.001^a
HADS symptoms of depression	52 (40%)	11 (16%)	.001^a
AUDIT harmful alcohol use last 12 months	11 (9%)	10 (15%)	.188^a
DUDIT harmful drug use last 12 months	11 (9%)	2 (3%)	.127^a

^a Pearson’s chi-square test.

^b t-Test.

Finally, a multiple regression was used to analyze association between background variables (symptoms of depression, symptoms of anxiety and traumatic life event) and parenting stress for all participants. Differences were regarded as significant when the p-value was less than 5% (p < .05). We did not perform any analysis of mass significance (e.g. Bonferroni–Holm test).

Ethical approval

Written informed consent was obtained from all participants. The study was approved by the Regional Ethics Committee at Lund University, Sweden (No 2011/193).

Results

Gender differences in terms of demographic and psychosocial data

The female participants were aged between 19 and 46 (median age 32 years; SD = 5.28), and the male participants were between 18 and 47 (median age 34 years; SD = 5.75). Single parents were more common among mothers (16 vs. 6%; p = .048). The mothers had experienced trauma more often than fathers (64 vs. 37%; p = .001). The traumatic life events most frequently reported by the mothers were severe physical abuse, sexual abuse, and death in the immediate family. The fathers reported being the victims of rape, assault and violence in combination with alcohol/drugs and death in the immediate family (Table 2).

Table 2. Demographic and social characteristics for mothers and fathers (n = 197).

	Mothers n = 129	Fathers n = 68	p-Value
Median age of parent, (years)	32.0 (range 19-46)	34 (range 18-47)
Primiparous	86 (67%)	45 (66%)	.945^a
Median age of child (months)	9.0 (range 1-48)	10.0 (range 2-48)
Married/cohabitating	108 (84%)	64 (94%)	.048^a
Educational background			.449^a
Primary, secondary ≤9 years of education	10 (8%)	9 (13%)
Upper secondary 10-12 years of education	50 (39%)	26 (38%)
College/university >12 years of education	69 (53%)	33 (49%)
Unemployment	21 (16%)	11 (16%)	.654^a
Ethnicity			.249^a
Swedish origin	86 (67%)	53 (78%)
European origin, other than Sweden	30 (23%)	11 (16%)
‘Outside Europe’ origin	13 (10%)	4 (6%)
Grew up with substance-abusing parent	35 (27%)	18 (27%)	.896^a
Parents in family of origin divorced	59 (46%)	27 (40%)	.417^a
Experienced traumatic life event	77 (64%)	25 (37%)	.001^a

^a Pearson’s chi-square test.

Fathers with and without substance-abusing parents in their family of origin

Thirty-five (27%) of the mothers and 18 (27%) of the fathers reported that they were raised in families with substance abuse (Table 2). Among the index men, two reported that both parents abused substances during their childhood, 13 reported that only their father abused substances, and three reported that only their mother abused substances. Alcohol abuse was the most common.

A total of 15 index fathers (vs.10 COMP fathers; p < .001) reported trauma. A history of divorce in the family of origin was more common in the index fathers (11 vs. 16; p = .03). The index fathers had a lower level of education compared with the COMP fathers (p = .004). No significant differences were found across groups concerning the other variables.

Perceived parenting stress according to the Parenthood Stress Questionnaire, SPSQ

Both mothers and fathers reported higher perceived parenting stress than the SPSQ norm, p = .0001[57].

The mothers were more affected by symptoms of parenting stress than fathers (3.13 vs. 2.67; p < .001) (Table 1).

Psychiatric symptoms according to ASRS, HADS, AUDIT and DUDIT

The mothers were more affected by symptoms of ADHD, anxiety and depression compared with the fathers. Among the fathers, 15% reported harmful drinking and 3% drug-related problems in the previous 12 months. The corresponding figures for the mothers were 9% for both alcohol and other drugs. Group differences between mothers and fathers in terms of concerning risky alcohol use and drug use were not observed (Table 1).

Variables predicting perceived parenting stress

For the mothers, both symptoms of depression and anxiety predicted high parenting stress. Among fathers, symptoms of depression predicted high parenting stress (Table 3). For the mothers, symptoms of ADHD were common (24 out of 129) but ADHD did not predict parenting stress. There were too few fathers with ADHD to perform an analysis.

Table 3. Summary of regression analysis for variables predicting parental stress for the mothers and for the fathers.

	Mothers (n = 129)			Fathers (n = 68)
Variable	β	SE β	Sign	β	SE β	Sign
HAD Anxiety	0.367	0.119	0.002	0.091	0.136	0.505
HAD Depression	0.574	0.091	0.000	0.623	0.176	0.001
Traumatic life event	0.107	0.089	0.233	0.100	0.123	0.419
	R^2 = 0.381			R^2 = 0.261

Discussion

This study was based on a sample of parents who, due to problematic parenting, had been referred for treatment at an IMH unit. The main findings were that both the mothers and the fathers reported higher perceived parenting stress than the Swedish norm for SPSQ [57] and the mothers were more affected by symptoms of parenting stress than the fathers. The fact that parenting stress is high in this patient group is not surprising, but the factors that contribute to parenting stress are of interest.

Symptoms of depression, anxiety and ADHD

Becoming a parent is one of the most overwhelming events in a person’s life. If the parent does not perceive that the resources available are sufficient to meet the challenges of parenthood, parenting stress can arise. Depression is one predictor of parenting stress [17,18], a correlation confirmed in our study for both mothers and fathers. In line with an earlier study [71], anxiety was also found to be a predictor of parenting stress among the mothers. Symptoms of anxiety and depression were common in both the mothers and the fathers, and the mothers, in particular, reported symptoms of anxiety (81%). Knowledge regarding anxiety among parents to infants is limited, but previous studies have shown that both men and women have an increased incidence of anxiety during the postpartum period [52].

ADHD problems were only overrepresented in the mothers, but ADHD was not a predictor of parenting stress. The mothers were more affected by symptoms of ADHD, anxiety and depression compared with the fathers. Our finding that symptoms of ADHD were more common among the mothers seems unexpected since males are known to suffer more from ADHD. However, recent studies on adult ADHD report a decreasing gender difference (male to female ratio about 2.4) [72].

Our results concerning the prevalence of affective disorders are in accordance with earlier studies confirming a gender difference [73]. Earlier studies indicate that women, during their reproductive years, have twice the risk of major depression compared to men [74–76]. The differences during pregnancy and the postpartum period have been attributed to the effects of burdens of pregnancy, childbirth, female sex steroids and parenting stress [76]. Another aspect is that girls are often socialized into a caregiving and parental role [77,78] and that the primary caregiving relationship usually involves the mother more than the father, which places an extra burden on the woman. Speculation is that this could be one explanation for the low level of participation of the fathers in our study (and in child psychiatric treatment in general). An additional aspect is that guilt and shame seem to be inherent parts of motherhood in our society [79], and an association has been found between feelings of shame and depression [80].

Traumatic life events

Individuals who have been exposed to trauma may experience more stress and socio-psychological problems when they become parents [22]. For example, they express lower satisfaction with being a parent, present a negative picture of the child, and an emotional numbing in their interaction with the child, all shortcomings suggested increasing the risk of child abuse [81–83]. In this study, many mothers (64%) had experienced traumatic life events, e.g. severe physical or sexual abuse and death in the immediate family, a greater figure than for the fathers, but they too had often suffered trauma (37%).

Children of parents affected by PTSD are vulnerable in several ways. A Nordic study showed that parental PTSD had major consequences for their children’s school performance with lower grades than their peers [84]. For the child, the emotional accessibility of the parent is important. A parent’s inability to recover from trauma, as measured by post-traumatic symptoms and inadequate caregiving behavior, affects the impact of trauma on their children [24,85–87].

Substance abuse in parents in the family of origin

A parent’s substance abuse can severely disrupt their caretaking abilities and is often associated with family dysfunction [10,11]. According to a registry study, 17% of all children in Sweden live with parents with high-risk levels of alcohol consumption [88], and 4.6% of all children in Sweden have at least one parent with a DSM-IV SUD [89]. Almost 30% of the mothers and the fathers surveyed in the present study grew up with substance-abusing parent/s, a context associated with an increased rate of mental health problems [90,91]. In our sample, this may have played a role in the family situation that resulted in IMH treatment.

The women's vulnerability in terms of self-rated symptoms of depression and ADHD, and of being a single parent, remained when we studied the subgroup of women who had grown up with substance-abusing parents [55]. The fathers who grew up under the same circumstances did not display the same vulnerability. This is in line with previous studies reporting that girls with substance-abusing parents may be more susceptible to mental disorders and substance abuse than boys [92–94].

However, we suggest that fathers with substance-abusing parents may also be more vulnerable compared to fathers with no such background. The fathers had a lower level of education, had more often experienced their parents’ divorce, and had experienced traumatic life events more often than the rest of the men. Experiencing their parents’ divorce, often with a change of residence and living standards, in addition to the parental conflict, can adversely affect a child and increase the risk of developing mental problems [12,29,95].

In a Swedish study, children of substance-abusing parents were shown to be at high risk. They were less likely to pursue higher-level studies, and later in life, were more often supported by social welfare, and were at greater risk of developing their own addictions and mortality compared with the normal population [96]. However, various social initiatives have been developed to provide support for children with abusive parents. A Finnish study describes the importance of using Beardslee's family intervention as a preventive method in which the child is given the opportunity to reflect on the parent's abuse [97].

In IMH treatment, with a focus on parenting stress, it is important to evaluate depression and anxiety in the parents [71]. A recent study found that the mothers’ attribution of their child is another important factor. A conclusion was that the mother’s ability to reflect on themselves and their children is an important component to successful parenting interventions [98].

Limitations

The present paper has several limitations that challenge generalisability, so we have interpreted our results conservatively, emphasizing results based on validated forms.

The sample was small, especially among the fathers. Only 62% of the fathers in the families where the mothers participated in treatment completed the questionnaires, which challenged the scientific premise to perform an analysis on attrition. The non-participants presumably involved fathers who did not take part in treatment. We chose to conduct some analyses on the fathers but interpreted the results with caution.

The study group, although representative of IMH families, is a selected clinical population where the participants had been referred due to pronounced difficulties with their parental functions and are not necessarily representative of all families with dysfunctional parenting.

One explanation concerning the differences between the mothers and the fathers could be that, in families where the father was insufficient but not the mother, applying for treatment at an IMH was not an option.

Previous studies comparing objective health status with self-reports have shown acceptable correlations [99]. In our study, data are based on five well-known, validated self-report questionnaires (SPSQ, ASRS, HADS, AUDIT, DUDIT) and one self-report questionnaire by the research group covering some sociodemographic and psychosocial risk factors. The IMH unit is a part of the Child and Adolescent Psychiatric Clinic, where parents are not assigned clinical diagnoses. Parents with psychosis and severe depression that had been diagnosed before admittance to the IMH unit by the general psychiatry service were excluded from the sample.

HADS has been criticized for not distinguishing sufficiently between anxiety and depression, and it has been suggested that HADS items should be considered in terms of general distress rather than anxiety and depression respectively [67]. However, some studies [66,67] have lent support to the two-factor model used in our study. It has been suggested that men’s symptoms of depression may differ from those of women [100–102]. This may imply that more fathers than indicated in our study suffered from depression and that the instrument used could not adequately identify the different symptoms. The highest rates of depressive symptoms for mothers occur soon after childbirth, while fathers more often develop symptoms when the child is 3 to 6 months old [41]. Due to the cross-sectional design, we report on parents with children aged from one month to four years.

The reports on substance abuse in the family of origin were based on subjective reports in the self-report questionnaire. Children of the Alcoholics Screening Test, CAST [103] would seem a better choice, but CAST has not been validated for the general Swedish population. The lack of diagnostic measures may have led to a misclassification of substance abuse in the family of origin.

Strengths and future implications

Despite these limitations, the present study of a representative group of IMH families may provide important information about parenting stress and dysfunction in families with small children within child psychiatry. The findings should be useful for health professionals and for the development of interventions to provide parents with adequate support during pregnancy and early parenthood. A key approach in future research would be to perform a thorough psychiatric evaluation, i.e. screening of the parent’s affective symptoms, with an instrument that captures a combination of different kind of symptoms of depression and anxiety symptoms. In view of the correlation between parenting stress and the child’s mental health and development, e.g. executive functions, it is important to also include an assessment of the child’s level of function in future studies [104].

Conclusions

The families referred to the IMH unit were vulnerable in many ways, with traumatized parents presenting a variety of psychiatric symptoms. The situation of the mothers was more serious than that of the fathers and indicated high parenting stress. For both the mothers and the fathers, depression was a significant predictor for parenting stress. For the mothers, symptoms of anxiety were also an important risk factor, making assessment at the start of treatment important. The serious situation of the mothers makes it important to implement targeted multifaceted interventions. It is also essential to include the fathers in the treatment as a possible accessible parent, as the child cannot be left alone with an insufficient mother.

Acknowledgements

The authors gratefully acknowledge the families who participated in the study. The authors thank a lecturer at Department of Statistics, Pierre Carbonnier, for excellent statistical assistance, and Camilla Salomonsson for administering the questionnaires.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The present work was financially supported by Johan Olsson Foundation, Skåne County Council and Lindhaga Foundation.

Notes on contributors

Eva Tedgård

Eva Tedgård, PhD, is clinical psychologist and family psychotherapist working part-time in an Infant Mental Health Unit at the Child & Adolescent Psychiatric Clinic in Malmö, Sweden.

Ulf Tedgård

Ulf Tedgård concluded his PhD in Medical Sciences at Lund University in 1999 on psychosocial and ethical aspects on prenatal diagnosis of hemophilia. He has thereafter participated in research on pediatric hematology. He is a Senior Consultant in pediatric hematology at the Pediatric Department, Skåne University Hospital, Lund, Sweden.

Maria Råstam

Maria Råstam is appointed Full professor in child and adolescent psychiatry at Lund University in 2008, associated with a position as senior consultant psychiatrist. Since 2016 she is senior professor at the same University. In 2015 she was appointed Visiting professor at the University of Gothenburg.

Björn Axel Johansson

Björn Axel Johansson completed his PhD in Psychiatry at Lund University in 2006. Since then, his research has been based on child and adolescent psychiatric in-patients, with a focus primarily on mobile technology, acute psychotic conditions and long-term follow-ups. He is a Senior Consultant at the Child & Adolescent Psychiatric Emergency Unit in Malmö, Sweden.