Archives of Women's Mental Health

, Volume 11, Issue 3, pp 221–229 | Cite as

New mothers’ thoughts of harm related to the newborn

Original Contribution

Abstract

There are few published studies of new mothers’ experiences of intrusive thoughts of harm related to the newborn. Evidence-based information about the normal phenomenology of intrusive thoughts of harm related to the newborn is needed to facilitate appropriate clinical decision-making. The objective of this project was to assess the phenomenology, prevalence, correlates, and behavioural sequelae of maternal thoughts of harm related to the newborn. One hundred women were recruited during pregnancy. Participants were assessed prenatally and at 4 and 12 weeks postpartum using questionnaires and a semi-structured interview about unwanted thoughts of harm related to the newborn. Postpartum intrusive thoughts of accidental harm to the infant were universal, and close to half of the sample reported unwanted thoughts of intentionally harming their infant. Compared with intentional harm thoughts, accidental harm thoughts were more frequent and more time consuming, but less distressing. High parenting stress and low social support predicted the occurrence of thoughts of intentional harm. Little evidence of an association between these thoughts and aggressive parenting was found. Unwanted intrusive thoughts of harming one’s infant are a relatively normative experience during the early postpartum period, particularly in association with greater parenting stress and low social support.

Keywords

Intrusive thoughts Thoughts of harm Postpartum mental health 

Introduction

Researchers have begun to examine unwanted intrusive thoughts experienced by mothers during the early postpartum period (Abramowitz et al. 2003a, 2006; Larsen et al. 2006). As suggested by the terminology, these thoughts are unwanted, perhaps even ego-dystonic, thoughts about harm coming to the newborn. The thoughts intrude into the individual’s consciousness, often provoking distress. Unwanted intrusive thoughts are not unique to new mothers; they can be observed in everyday life in response to interpersonal loss, as craving-related thoughts, or frightening images related to important upcoming events such as a formal presentation or wedding. Shorter in duration than worry and less frequent than obsessions, unwanted intrusive thoughts nevertheless are capable of provoking intense negative affect.

In the case of new mothers, the thoughts are often distressing to the women who experience them and potentially confusing and alarming to their healthcare providers. Evidence-based information about the phenomenology of normal maternal postpartum thoughts of harm related to the newborn is needed to facilitate appropriate clinical decision-making for women who report these thoughts. The objectives of this project were to document the content and characteristics of postpartum intrusive thoughts of harm related to the infant, to assess predictors of thoughts of intentionally harming the child, and to investigate the relationship between intentional harm thoughts and harsh parenting. A further objective was to provide normative information about how new mothers behaviourally respond to postpartum intrusive thoughts of harm related to their infant.

In the first of the only three extant studies of this topic, researchers at the Mayo Clinic assessed maternal and paternal thoughts of harm related to the newborn via questionnaires (Abramowitz et al. 2003a). Participants were asked to report up to three intrusive thoughts they had experienced in relation to their infant. They also reported on characteristics of the thoughts: duration, degree of interference, distress, and controllability. Participants were mothers (n = 92) and their partners (n = 64) who had experienced a full-term pregnancy and uncomplicated delivery. At the time of the survey, which was an average of 4.3 months after birth, most mothers (69%) and fathers (58%) reported having experienced unwanted intrusive thoughts about their infant. Reported thought content included suffocation or sudden infant death syndrome, accidents, intentionally harming the infant in some way, losing the infant, illness, unacceptable sexual thoughts about the infant, and contamination. In comparison with fathers, mothers were more distressed by the thoughts, and reported longer duration of the thoughts. For mothers, symptoms of depression were positively related to all assessed characteristics of the thoughts (e.g., duration, distress). Fathers’ depression levels, which were lower than mothers’, were not related to characteristics of the unwanted thoughts.

In a separate paper, Larsen et al. (2006) reported on thought control strategies used by participants in the Abramowitz et al. (2003b) study who had reported unwanted intrusive thoughts related to their infant (n = 53 mothers and 22 fathers). The majority of mothers (85%) and fathers (86%) reported intentionally using at least one strategy to try to control their unwanted thoughts related to the infant. Strategies included self-reassurance, distraction, checking, prayer, interacting with the infant, seeking social support, positive thought insertion and avoidance.

In a later study, Abramowitz et al. (2006) assessed unwanted intrusive infant-related thoughts in a sample of 50 first-time mother–father dyads. Researchers administered a diagnostic interview at a prenatal assessment and a semi-structured interview to assess postpartum intrusive thoughts and behavioural responses approximately 3 months after the birth. Participants also completed questionnaires at both time points. The vast majority of mothers (91%) and fathers (88%) reported distressing intrusive thoughts about their infant at some point following the baby’s birth, and most parents (77% of mothers and 60% of fathers) reported engaging in strategies to control or cope with the intrusive thoughts.

The above studies indicate that unwanted intrusive thoughts related to the newborn and behavioural responses to these thoughts are common among new parents. Although they provide important information about the phenomenon of unwanted intrusive thoughts related to the newborn, these studies did not systematically distinguish between thoughts of accidental and intentional harm related to the newborn, a distinction that could bear on caregiving decisions concerning infant welfare. For example, medical professionals may wonder whether thoughts of intentionally harming the infant would presage aggressive parental behaviour. In addition, these studies assessed parents several months after the birth of the baby, but others (Leckman et al. 1999) have reported that parental preoccupation with the infant (defined as thoughts and concerns about the baby that included thoughts of harm and checking behaviours) peaks at 2 weeks postpartum. The research reported here was designed to fill some of these gaps in knowledge.

In each of the above studies, it was hypothesized (implicitly or explicitly) that unwanted and sometimes unacceptable thoughts about infants are common among parents of newborns. This prediction is consistent with data from numerous studies indicating that occasional unwanted intrusive thoughts, images, or impulses are reported by the majority of people and reflect these individuals’ current concerns (Purdon and Clark 1993; Rachman and de Silva 1978; Salkovskis and Harrison 1984). Unwanted intrusive thoughts are distinct from the obsessions that occur as a part of obsessive–compulsive disorder (OCD) in that they are not clinically significant, being shorter in duration, less distressing and easier to dismiss. Although both unwanted intrusive thoughts and obsessions may provoke efforts to control the thought (e.g., suppression, distraction), the former are not generally associated with ritualized behavioural compulsions.

Unwanted intrusive thoughts of harm (either accidental or intentional) related to the infant may be elicited by a variety of factors. Such thoughts may, for example, represent a normal consequence of the stress of new parenthood and the added responsibility that accompanies this unique role. They may also occur as a function of psychological problems known to occur during the early postpartum period (e.g., depression, obsessive–compulsive disorder, psychosis). Although unwanted intrusive thoughts occur as a part of normal psychology, they also occur in these forms of postpartum psychopathology and may at times reflect a psychopathological process preceding child abuse or neglect, thus highlighting the importance of research to understand the phenomenology of these thoughts. Wisner et al. (2003) provide an excellent review of the range of postpartum psychopathology and their relation to thoughts of intentional harm related to the infant as well as actual child harm.

Postpartum OCD is typically characterized by obsessions of harming or of harm coming to the infant (Maina et al. 2000; Abramowitz et al. 2003b). Among our own clients, examples of postpartum obsessions include distressing thoughts or images of smothering one’s infant or touching the infant in a sexual way during diapering. Obsessional thoughts, in spite of their often disturbing or horrific content, have not been associated with an increased risk of violence among individuals with OCD. Obsessions are experienced as unwanted, repugnant, and inconsistent with the person’s belief system and personality. OCD sufferers fear acting upon their thoughts and many engage in high levels of avoidance and rituals in an effort to control their thoughts and ensure that they do not act upon them. Most unwanted intrusions are evoked by an external stimulus (Parkinson and Rachman 1980), and the frequency of unwanted intrusions seems to be higher in the context of stressful situations and negative emotional states like depression (Brewin et al. 1996; Horowitz 1975).

Unwanted intrusive thoughts involving harming the infant are also common in postpartum depression. One study reported that 41% of mothers with major depression disclosed unwanted intrusions related to harming their infant compared with only 7% of non-depressed mothers (Jennings et al. 1999). Compared to non-postpartum major depression, repugnant thoughts occurring during postpartum depression tend to be more violent, with ideation such as putting the baby in the microwave or drowning or stabbing the baby (Wisner et al. 1999).

Thoughts of harming the infant are also common among women suffering from postpartum psychosis, although this problem is in itself fairly rare, occurring in one to two childbearing women out of 1,000 (Nonacs et al. 2005). Postpartum psychotic symptoms typically occur in the context of a psychotic depression or a manic episode (Chaudron and Pies 2003). Unlike postpartum OCD, the aggressive ideation in psychosis is experienced as ego-syntonic, is not subjectively resisted, and is accordingly associated with an increased risk of aggressive behavior.

Child physical abuse, defined as non-accidental use of force by a parent or other caregiver that causes or could cause injury or death, is difficult to study with accuracy because many instances of child abuse go unreported and undetected (Gil 1970). This problem is probably heightened in the case of physical abuse of infants, which is nevertheless thought to occur less frequently than physical abuse of older children (Christensen 1999). Although child abuse researchers have not yet developed a reliable strategy for predicting who will abuse a child, several risk factors have been identified, including those in the domains of demographic factors, parental characteristics, family relationships, and child characteristics (Belsky and Vondra 1989; Lee et al. 2007). For reviews representing different aspects of this literature, see Belsky (1993), Chaffin et al. (1996), or Wolfe (1985). Of significance for this project, there have been no studies of parental ideation at the time of the abuse. Therefore, it is not known if thoughts of harming one’s infant or child are precursors to aggressive parental behaviour.

The primary purpose of this study was to assess the prevalence, nature and predictors of maternal postpartum thoughts of harm related to the newborn and to explore the relationship between these thoughts and harsh parenting behaviours. Women were recruited to the study during the third trimester of their pregnancy. They completed some questionnaires during pregnancy and additional questionnaires on two occasions following the birth of their infant. A semi-structured interview assessed the content, characteristics, and correlates of participants’ thoughts of harm related to the infant at 4 and 12 weeks postpartum.

Materials and methods

Participants

One hundred English-speaking pregnant women who were expecting a healthy first child were recruited to the study. Participants were recruited from two major hospitals, midwifery offices, physician offices and prenatal education classes in a large urban centre. Pregnant women responded to pamphlets describing the “Mother Infant Wellness Project” and inviting them to help researchers learn more about “thoughts of harm coming to their baby” by completing unspecified questionnaires and interviews. Participants’ ages ranged from 23 to 41 years (M = 32.0, SD = 4.3). The average number of years of education was 16.7. Most participants were employed (79%), married or living with a romantic partner (97%) and self-identified as Caucasian (76%) or Asian (13%). Over half (69%) of participants reported annual family incomes of $50,000 or greater.

Procedures

Participants were recruited during pregnancy (at approximately 35 weeks gestation) and completed questionnaires prenatally and twice following the birth of their infants (at approximately 4 and 12 weeks postpartum). In addition, project staff administered a semi-structured interview about postpartum thoughts of harm (see description below) in person at 4 weeks postpartum and over the telephone at 12 weeks postpartum. See Table 1 for listing of measures administered at each time point.
Table 1

Timing of assessments

Prenatal

4 weeks postpartum

12 weeks postpartum

Demographic Questions

Parenting Stress Index

Parenting Stress Index

Eysenck Personality Questionnaire

Social Support

Social Support

 

Postpartum Intrusions Interview

Postpartum Intrusions Interview

  

Conflict Tactics Scale

In all, 98 women completed the prenatal questionnaires. At 4 weeks postpartum, 91 women completed the interview, but four failed to return the questionnaires. At 12 weeks postpartum, 84 women completed the interview, but three failed to return the questionnaires. Two women entered the study after they had given birth, thus providing postpartum data only. Approval for the project was obtained from the appropriate health authorities and university ethics board.

Measures

The Eysenck Personality QuestionnaireRevised Short Scale (Eysenck et al. 1985) is a widely used 48-item, self-report inventory with four subscales. Only the Neuroticism and Psychoticism subscales were used for the present study. The psychometric properties of the scale have been assessed across a range of cultural groups (Francis et al. 1992). This scale was completed during pregnancy along with other questionnaires that are not the subject of this report.

Two questionnaires were administered at both postpartum assessments. The Parenting Stress IndexShort Form (Abidin 1995) is a 36-item self-report questionnaire intended for use with parents of children as young as 1 month of age. The scale yields a Total Stress Score, plus three subscales: Parental Distress, Parent–Child Dysfunctional Interaction, and Difficult Child. The Parenting Stress Index—Short Form shows good evidence of validity and reliability (Reitman et al. 2002). In addition, the Social Support from Partner and Social Support from Others is a 10-item self-report scale designed to assess material, instrumental, emotional, and informational support (as well as satisfaction with support provided) received from the mother’s partner and others in her life. In studies with pregnant and postpartum women, the measure predicts more rapid progress in labour, higher Apgar scores in neonates, and less postpartum depression (Collins et al. 1993; Graham et al. 2002; Yali and Lobel 2002).

The Conflict Tactics Scale Parent–Child Version (Straus et al. 1998) is a 35-item self-report instrument designed to assess abusive and non-abusive parenting practices, as well as child neglect. Straus et al. (1998) report good convergent and discriminant validity, although internal consistency of the subscales is poor (−0.02 to 0.70), perhaps due to the very low base rate of some of the behaviours being assessed. The base rates were similarly low in the present study. Accordingly, the measure was used in a dichotomous fashion for the present study; endorsement of any of the verbal, physical or neglect items was coded as positive for harsh parenting behaviours. Participants completed this measure at the conclusion of their involvement with the project (12 weeks postpartum). In order to maximize the possibility of accurate reporting, numerous steps were taken to render these data anonymous before the responses were examined. Participants were informed about these procedures in advance.

The Postpartum Intrusions Interview was developed specifically for this project in order to assess the content of maternal thoughts of harm related to the newborn, mothers’ affective and behavioural responses to the thoughts, and the degree of functional impairment associated with the thoughts.

The interview is comprised of three sections: (1) thoughts of accidental harm, (2) thoughts of intentional harm (e.g., drowning, stabbing), and (3) questions pertaining to the natural history of the thoughts. Each of the first two sections of the interview includes the following: (a) a checklist of possible harm thoughts including any additional thoughts respondents may have experienced that do not appear on the list, (b) a series of eight questions about respondents’ experience of the thoughts (i.e., frequency, duration, persistence and emotional responses to the thoughts, degree of interference with parenting and more general functioning, perceived likelihood that the content of the thoughts would occur), and (c) overt and covert behavioural responses to the thoughts such as mentally reviewing the child’s safety or avoiding bathing the child. These sections were modeled after the Yale–Brown Obsessive–Compulsive Scale with modifications based on prior research and our clinical experience with new mothers. The natural history section of the interview asks participants about the timing and course of the thoughts. The 4 weeks postpartum interview assessed thoughts occurring “since the baby’s birth” (i.e., the preceding 4 weeks), and the 12 weeks interview assessed thoughts occurring during the preceding week.

To introduce the interview, participants were told that unwanted thoughts of harm are common among new mothers, and examples were given. Interview items were generated from clinical experience with obsessional patients and qualitative interviews with new mothers. Several experts on obsessions provided comments and suggestions on the initial drafts of the interview, as did several researchers and clinicians whose work is relevant to perinatal mental health. Prior to use in this research, the interview was piloted with new mothers and subsequently refined.

Results

Prevalence of thoughts of harm

At 4 weeks postpartum all interview participants (n = 91) reported intrusive thoughts of accidental harm befalling their infant, and 45 of 91 (49.5%) reported thoughts of intentionally harming their newborn (95%CI = 39.18–59.72). Among these 45 women, 39 (86.7%) reported thoughts of intentional physical harm (i.e., 42.9% of the total sample; 95%CI = 32.69–53.03). The remainder (6/45 or 6.6% of the total sample; 95%CI = 1.49–11.69) reported thoughts of intentionally screaming at their infant, but no thoughts of intentional physical harm.

By 12 weeks postpartum, nearly all of the women (80/84 or 95.2% of the sample) reported that they were still experiencing thoughts of accidental harm (95%CI = 90.69–99.79). Of the 45 women who reported intrusive thoughts of intentionally harming their infant at 4 weeks postpartum, 12 (26.7%) reported continued thoughts of intentional harm at 12 weeks. Among the six women who reported thoughts of verbal aggression only (i.e., screaming at their infant) at 4 weeks postpartum, only one reported thoughts of intentional harm at 12 weeks postpartum. An additional four women who had denied thoughts of intentional harm at 4 weeks reported them at 12 weeks postpartum. Altogether, 19.1% (16 of 84) of women reported thoughts of intentionally harming their infant at 12 weeks postpartum (95%CI = 10.65–27.45). As with the 4-week data, most (14/16) reported thoughts of physical harm. Two women among these 16 reported thoughts of verbal aggression (i.e., screaming) only.

Two thirds of the study participants reported that the thoughts of harm coming to their infant began during pregnancy (7.7%), or within 24 h of the birth (58.3%). Most women (63.2%) also reported that the thoughts began abruptly. By 4 weeks postpartum, about half of the sample (54.4%) indicated that the thoughts of harm had declined in frequency or stopped altogether. Most of the remaining women (40.0%) reported that the thoughts remained the same over this time period, but some women (5.6%) described an increase in frequency over the first postnatal month. A similar picture was evident at the 12 weeks assessment, with about half of participants (55.4%) reporting continued declines in the frequency of these thoughts during the second and third months of the infant’s life, 28.9% reporting stable frequency of the thoughts and 15.7% reporting increased frequency of intrusive thoughts about the infant over time.

Content of thoughts of harm

The content of thoughts of accidental harm typically involved suffocation, the baby falling or accidentally being dropped from a high surface, contamination and illness, accidents, neglect (e.g., accidentally forgetting the infant in the car), sexual abuse/assault by another person, drowning, burns, abduction or harm by another person, animal attacks, and accidentally being responsible for harm coming to one’s infant (e.g., stepping on the baby by accident).

The main content areas of reported intentional harm thoughts are listed in Table 2.
Table 2

Percentage of mothers reporting specific thoughts of intentional harm

 

4 weeks

12 weeks

n = 91

n = 84

% (of 91)

% (of 84)

Screaming at your baby

19.78

9.52

Shaking your baby

16.48

5.95

Giving your baby away

13.19

2.38

Intentionally hitting your baby too hard when burping him/her

12.09

0

Dropping or throwing your baby out the window or off the balcony

8.79

1.19

Touching baby’s genitals in an inappropriate way

8.79

0

Intentionally puncturing the soft spot on your baby’s head

7.69

0

Throwing or dropping your baby on purpose

7.69

0

Stabbing your baby

5.50

1.19

Slapping or hitting your baby

5.50

2.38

Intentionally allowing your baby to fall under water in the bath

4.40

1.19

Intentionally smothering your baby

4.40

0

Burning your baby with hot water on purpose

2.20

1.19

Leaving baby somewhere where he/she may not be found right away

2.20

3.57

Strangling your baby

1.10

0

Stepping on your baby on purpose

1.10

0

Other (idiosyncratic)

4.40

4.76

Harm thought characteristics

We assessed characteristics of both accidental and intentional harm thoughts including how frequent, time consuming and distressing participants experienced the thoughts to be, perceptions of the likelihood that the content of the thought could actually occur, how difficult it was for the participant to distract herself from the thoughts, and how much the thoughts interfered with her ability to parent. The means and standard deviations of thought characteristics are presented in Table 3.
Table 3

Harm thought characteristics: percentages per category

 

4-week (past month)

12-week (past week)

Accidental

Intentional

Accidental

Intentional

n = 91

n = 45

n = 80

n = 16

Thought frequency

    

 Less than once a week

1.1

42.2

5.0

31.3

 Once or twice a week

36.3

51.1

43.8

50.0

 Every day

27.5

4.4

20.0

18.8

 Several times a day

29.7

2.2

31.3

0

 Almost constantly

5.5

0

0

0

Time occupied by thoughts

    

No time at all

14.3

62.2

37.5

56.3

<10 min/week

52.7

31.1

41.3

37.5

≤10 min/day

25.3

6.7

15.0

6.3

≤An hour a day

2.2

0

6.3

0

An hour or more per day

5.5

0

0

0

Distress caused by thoughts

    

Not at all upsetting

23.1

8.9

42.5

12.5

Mildly upsetting

40.7

20.0

37.5

31.3

Moderately upsetting

24.2

37.8

13.8

43.8

Severely upsetting

12.1

33.3

6.3

12.5

Extremely upsetting

0

0

0

0

Guilty feelings about thoughts

    

Not at all guilty

78.0

20.0

86.3

37.5

A little bit guilty

15.4

26.7

10.0

31.3

Quite guilty

5.5

24.4

3.8

25.0

Very guilty

1.1

26.7

0

6.3

Extremely guilty

0

2.2

0

0

Interference with parenting

    

No interference

83.5

84.4

90.0

93.8

Slight interference

13.2

15.6

7.5

6.3

Quite a bit of interference

2.2

0

2.5

0

Severe interference

1.1

0

0

0

Extreme interference

0

0

0

0

Likelihood of thought content occurring

    

Not at all likely

53.9

86.7

48.8

62.5

A little bit likely

21.3

6.7

30.0

18.8

Quite likely

19.1

0

17.5

6.3

Very likely

3.4

4.4

2.5

6.3

Extremely likely

2.2

2.2

1.3

6.3

Difficulty getting rid of thoughts

    

Very easy

59.3

65.9

78.8

62.5

A little bit hard

34.1

31.8

18.8

31.3

Quite hard

3.3

0

2.5

6.3

Very hard

3.3

0

0

0

Extremely hard

0

2.3

0

0

Number of thoughts reported

12.77 (4.68)

2.56 (2.26)

9.49 (5.16)

1.81 (0.91)

Some participants provided ratings of their harm thoughts that would indicate clinically significant symptoms (e.g., thoughts take up a lot of time, are extremely distressing, and cause substantial interference with parenting). Participants were classified as having providing ratings indicative of clinically significant symptoms if they scored in either of the top two categories (e.g., several times a day or almost constantly) on three or more of the eight harm thought characteristics. Few participants fell into this category for either accidental (7.7% at 4 weeks and 6.3% at 12 weeks) or intentional (4.4% at 4 weeks and none at 12 weeks) harm thoughts.

Dependent samples t-tests were conducted to compare characteristics of accidental and intentional harm thoughts. Among women who reported both accidental and intentional thoughts of harm at 4 weeks postpartum, accidental harm thoughts were rated as more frequent (t(44) = 9.93, p < 0.001, d = 1.48), more time-consuming (t(44) = 7.53, p < 0.001, d = 1.12), and more likely to actually occur (t(44) = 3.60, p = 0.001, d = 0.54). However, intentional harm thoughts were rated as more upsetting (t(44) = −3.68, p = 0.001, d = 0.55), including more intense feelings of guilt (t(44) = −6.55, p < 0.001, d = 0.98). Accidental and intentional harm thoughts did not differ with respect to how much they interfered with parenting (t(44) = 0.94, d = 0.14) or how much difficulty participants had in distracting themselves from the thoughts (t(43) = 0.63, d = 0.09).

Predictors of occurrence of thoughts of intentionally harming the infant

Logistic regression was used to test whether demographic characteristics, personality, parenting stress (parent distress, difficult child and parent–child relationship), or social support would distinguish women who had thoughts of intentional harm from those who had only thoughts of accidental harm at 4 weeks postpartum. All variables were entered simultaneously into the regression equation. Marital status was not included, as very few (3%) participants were single. The overall regression was significant (χ2(4, n = 75) = 24.54, p < 0.001), correctly classified 76% of participants, and resulted in a Nagelkerke R2 of 0.37. Both parenting stress (specifically maternal perception of the mother–infant relationship) and social support significantly predicted which participants would report thoughts of intentionally harming their infant at the 4 weeks postpartum interview (see Table 4).
Table 4

Final model predicting the occurrence of thoughts of intentional harm at 4 weeks postpartum

Variable

B

SE

Wald

Odds ratio

95%CI for OR

Lower

Upper

Education

−0.11

0.13

0.81

0.89

0.70

1.14

Income

0.24

0.20

1.40

1.27

0.85

1.89

Race (1)

−0.54

0.93

0.35

0.58

0.10

3.56

Race (2)

−0.22

1.33

0.03

0.80

0.06

10.80

Neuroticism

−0.08

0.11

0.54

0.93

0.75

1.14

Psychoticism

−0.17

0.17

0.90

0.85

0.60

1.19

Parent distress

0.09

0.05

3.15

1.10

0.99

1.21

Parent–child relationship

0.21

0.06

11.36*

1.23

1.09

1.39

Difficult child

−0.07

0.05

1.88

0.94

0.85

1.03

Social support

2.02

0.72

7.94*

7.56

1.85

30.85

*p < 0.01

Behavioural responses to unwanted thoughts of harm

Participants were also asked to describe any behaviours they may have engaged in, in response to their reported thoughts. The majority of participants reported at least one behavioural response to both accidental (98.9% at 4 weeks, and 95.0% at 12 weeks) and intentional (71.1% at 4 weeks and 56.3% at 12 weeks) harm thoughts. Behavioural responses to accidental and intentional harm thoughts endorsed by participants included checking on the baby’s health and well-being, reassurance seeking, avoidance behaviours (e.g., avoiding situations in which the thoughts occur, avoiding specific activities with one’s infant), washing, distraction and mental undoing. Participants’ behavioural responses to both accidental and intentional harm thoughts fell, on average, within the “never” to “rarely” (i.e., 0 to 1) range for each set of behaviours (i.e., compulsions, reassurance-seeking and avoidance). This was true for both the 4-week and the 12-week assessment points.

Harsh parenting

At 12 weeks postpartum, 22 women (27.5% of respondents) reported at least one item indicative of harsh parenting on the Conflict Tactics Scale. Items endorsed included verbal aggression (shouting, yelling, or screaming at the infant; n = 13/22), physical aggression (shaking or pinching the infant; n = 4/22), and neglect (leaving the infant home alone, having problems taking care of the infant or showing love because of personal problems or substance abuse, or not making sure the infant got the food s/he needed; n = 11/22).

Chi-square analyses were used to test the degree to which reporting some (versus no) harsh parenting behaviour was associated with reporting the occurrence of thoughts of intentionally harming the baby. These analyses indicated that approximately the same proportion of women with and without intentional harm thoughts at 4 weeks postpartum subsequently reported harsh parenting at some point during the infant’s first 3 months of life (28.2% for women who reported intentional harm thoughts at 4 weeks and 27.5% for women who did not; χ2(1, n = 80) = 0.01, p = 0.92). At 3 months postpartum, 42.9% of the women who reported intentional harm thoughts also reported engaging in harsh parenting, compared with 25.0% of women who did not report intentional harm thoughts at 12 weeks (χ2(1, n = 79) = 1.91, p = 0.17), a difference that is not statistically significant with this sample size, but would be of concern if it is shown to be robust. See Table 5 for details.
Table 5

Thoughts of intentional harm predicting harsh parenting

 

Evidence of physical or verbal aggression or neglect?

Did the participant report thoughts of intentional harm?

Yes

No

n (%)

n (%)

At 4 weeks postpartum

Yes

11 (28.21)

28 (71.80)

No

12 (29.27)

29 (70.73)

At 12 weeks postpartum

Yes

6 (42.86)

8 (57.14)

No

16 (24.62)

49 (75.39)

We also closely examined the degree of correspondence between the type of harsh parenting behaviour reported and the content of the intentional harm thoughts reported among the 11 women who endorsed both. For four of these 11 women, one of the reported thoughts of intentional harm was the same as one of the harsh parenting behaviours indicated. In each of these four cases the thought and the behaviour were “screaming at your baby”.

Discussion

The primary purpose of this research was to document the phenomenology of maternal thoughts of harm related to the newborn. Of particular interest was the comparison of thoughts of accidental harm with those involving intentional harm. The prospective design also allowed for the examination of the degree to which early postpartum intrusive thoughts involving intentionally harming the infant would predict subsequent harsh parenting behaviours. This manuscript presents the first study of new mothers’ thoughts of harm related to the newborn in which accidental and intentional harm thoughts are distinguished and their relation to harsh parenting is examined.

At 4 weeks postpartum every woman in the sample reported unwanted thoughts of accidental harm related to the newborn, and nearly half of the sample reported unwanted thoughts of intentionally harming their infant. Two months later, almost all (95%) participants continued to report thoughts of accidental harm, and nearly one in five reported thoughts of intentional harm. The thoughts of intentional harm ranged greatly in their degree of violence, from verbal or passive aggression (e.g., thoughts of screaming at the baby or intentionally leaving the baby in the sun too long) to graphic images of potentially lethal actions (e.g., poking the baby’s eyes, putting the baby in the freezer, flushing the baby down the toilet, and allowing the baby’s stroller to roll into the ocean). Not surprisingly, although accidental harm thoughts were more frequent and time consuming, women were less distressed by them than by intentional harm thoughts. Of the predictors assessed, only two variables were found to predict the occurrence of thoughts of intentional harm: high parenting stress and low social support.

Approximately 25% of the sample reported some harsh parenting behaviours. The proportion of women who reported some harsh parenting behaviours did not differ between those women who endorsed intentional harm thoughts at 4 weeks postpartum and those who did not. At 3 months postpartum, more women who reported intentional harm thoughts also reported engaging in harsh parenting, compared with the women who did not report intentional harm thoughts (42.86% versus 24.62%), a difference that was not statistically significant. There was no correspondence between the content of intentional harm thoughts and the type of harsh parenting behaviour with the exception of unwanted thoughts of screaming at the infant.

The fact that close to half of this sample reported thoughts of intentionally harming their infant sometime during the first 4 weeks of his/her life strongly suggests that, for first-time parents, these kinds of thoughts are a normative aspect of early parenting. As some women are likely to feel reluctant to disclose that they have experienced thoughts of harming their infant (conversely women are unlikely to falsely report that they have experienced thoughts of intentional harm), it is likely that the true population prevalence of postpartum thoughts of intentional harm among first time parents is even greater. Not surprisingly, the occurrence of intrusive thoughts of accidental harm befalling the newborn appears to be a near universal experience among new mothers.

Both types of thoughts appear to become less frequent over time, but this cannot be unambiguously concluded from the present study because of differences in the sampling time frame of the questions asked. Specifically, at 4 weeks postpartum women were asked about thoughts they had experienced since the baby’s birth, whereas at 12 weeks postpartum, women were asked about thoughts they had experienced in the previous week.

Clinically, intrusive ideation tends to be related to topics that are of current importance to the individual It is therefore not surprising that some of the women in our sample began to experience thoughts of harm related to their infant even prior to birth. It is also not surprising that thoughts of intentional harm were experienced as more distressing and engendered stronger feelings of guilt than thoughts of accidental harm. Thoughts of intentional harm are undoubtedly less socially acceptable than thoughts of accidental harm, and indeed they occur less commonly. Given the paucity of extant empirical research on postpartum intrusive thoughts about the infant, women and perhaps their caregivers may also be more likely to perceive these thoughts as strange or aberrant, leading to stronger feelings of distress and guilt when they occur.

Women who experienced intrusive thoughts of intentionally harming their infant did not differ from those who denied experiencing these thoughts in terms of the demographic characteristics or personality variables measured in this study. Parenting stress, specifically the parent’s perception of her relationship with her infant, and low perceived social support were related to the occurrence of thoughts of intentional harm. This seems to suggest that women who feel frustrated or unhappy with the quality of their interactions with their infant, or who feel unsupported by others in their new role as a mother, are more likely to experience thoughts about harming their infant. Conversely, intrusive thoughts of intentional harm may influence women’s perceptions of the quality of (a) their relationship with their infant and (b) the social support they receive. Future research in this area would benefit from a clarification of the causal direction of the relationship between the quality of the mother–infant attachment and the occurrence of aggressive ideation related to the child.

Clinically, maternal reports of thoughts of intentional harm related to the newborn can raise concerns that the mother may be at risk of harming her infant. This study aimed to gather prospective information to shed light on this issue. There was no evidence to support the notion that the occurrence of thoughts of intentional harm during the first month of the child’s life predicted subsequent harsh parenting behaviours. At 3 months postpartum, however, more women who reported intentional harm thoughts also reported engaging in harsh parenting, compared with the women who did not report intentional harm thoughts at 12 weeks. Because of the potential gravity of harsh parenting, the magnitude of this difference would be meaningful if it is robust; future research should aim to replicate this finding. It would be particularly useful to examine this question in more diverse samples (e.g., broader range of socioeconomic status, more ethnic and racial diversity, solo parent families). What is clear from this study is that the only unwanted thought that corresponded directly to harsh parenting behaviour was unwanted ideation about screaming at the baby, a behaviour some mothers reported engaging in. The present results provide no evidence that unwanted intrusive thoughts of intentionally harming one’s infant represents a risk of abuse or neglect.

In interpreting the results of this study, one should bear in mind that generalizability may be affected by the fact that participants were relatively privileged. As a group, they were well educated and upper middle class, and almost all were parenting with a partner. Replication with a broader and more diverse sample is merited. Furthermore, harsh parenting was assessed dichotomously because of the low base rate of endorsement of specific items of physical and verbal aggression and neglect. A longer follow-up period may allow researchers to better determine the degree to which certain types of ideation or early harsh parenting behaviour are predictive of suboptimal parenting in the long run. Finally, a more complete understanding of unwanted thoughts of intentionally harming the infant would require, in addition to a much larger sample, comprehensive assessment of mental health conditions that are known to occur in the postpartum period and involve upsetting ideation, such as depression, psychotic symptoms, and OCD.

Notes

Acknowledgments

We thank Michael Papsdorf, Elizabeth Horner, Marion MacKay-Dunn, Maria Watson, and the study participants for their invaluable contributions to this project. This research was supported by a grant to Drs. Woody (PI) and Fairbrother (co-investigator) from the North Shore Health Research Foundation. The authors have no conflicts of interest related to this project.

References

  1. Abidin RR (1995) Parenting stress index. Psychological Assessment Resources, Odessa, FLGoogle Scholar
  2. Abramowitz JS, Schwartz SA, Moore KM (2003a) Obsessional thought in postpartum females and their partners: content, severity, and relationship with depression. J Clin Psychol Med Settings 10:157–164CrossRefGoogle Scholar
  3. Abramowitz JS, Schwartz SA, Moore KM, Luenzmann KR (2003b) Obsessive–compulsive symptoms in pregnancy and the puerperium: a review of the literature. J Anxiety Disord 17:461–478CrossRefPubMedGoogle Scholar
  4. Abramowitz JS, Khandker M, Nelson CA, Deacon BJ, Rygwall R (2006) The role of cognitive factors in the pathogenesis of obsessive–compulsive symptoms: a prospective study. Behav Res Ther 44:1361–1374CrossRefPubMedGoogle Scholar
  5. Belsky J (1993) Etiology of child maltreatment: a developmental–ecological analysis. Psychol Bull 114:413–434CrossRefPubMedGoogle Scholar
  6. Belsky J, Vondra J (1989) Lessons from child abuse: the determinants of parenting. In: Cicchetti D, Carlson V (eds) Current research and theoretical advances in child maltreatment. Cambridge University Press, Cambridge, England, pp 153–202Google Scholar
  7. Brewin CR, Hunter E, Carroll F, Tata P (1996) Intrusive memories in depression: an index of schema activation? Psychol Med 26:1271–1276PubMedGoogle Scholar
  8. Chaffin M, Kelleher K, Hollenberg J (1996) Onset of physical abuse and neglect: psychiatric, substance abuse, and social risk factors from prospective community data. Child Abuse Neglect 20:191–203CrossRefPubMedGoogle Scholar
  9. Chaudron LH, Pies RW (2003) The relationship between postpartum psychosis and bipolar disorder: a review. J Clin Psychiatry 64:1284–1292PubMedCrossRefGoogle Scholar
  10. Christensen E (1999) The prevalence and nature of abuse and neglect in children under four: a National survey. Child Abuse Rev 8:109–119CrossRefGoogle Scholar
  11. Collins NL, Dunkel-Schetter D, Lobel M et al (1993) Social support in pregnancy: psychosocial correlates of birth outcomes and postpartum depression. J Pers Soc Psychol 65:1243–1258CrossRefPubMedGoogle Scholar
  12. Eysenck SBG, Eysenck HJ, Barrett P (1985) A revised version of the psychoticism scale. Pers Individ Differ 6:21–29CrossRefGoogle Scholar
  13. Francis LJ, Brown LB, Philipchalk R (1992) The development of an abbreviated form of the Revised Eysenck Personality Questionnaire (EPQR-A): its use among students in England, Canada, the USA and Australia. Pers Individ Differ 34:443–449CrossRefGoogle Scholar
  14. Gil D (1970) Violence against children. Harvard University Press, Cambridge, MAGoogle Scholar
  15. Graham JE, Lobel M, DeLuca RS (2002) Anger after childbirth: an overlooked reaction to postpartum stressors. Psychol Women Q 26:222–223CrossRefGoogle Scholar
  16. Horowitz MJ (1975) Intrusive and repetitive thoughts after experimental stress. Arch Gen Psychiatry 32:1457–1463PubMedGoogle Scholar
  17. Jennings KD, Ross S, Popper S, Elmore M (1999) Thoughts of harming infants in depressed and nondepressed mothers. J Affect Disord 54:21–28CrossRefPubMedGoogle Scholar
  18. Larsen KE, Schwartz SA, Whiteside SP, Khandker M, Moore KM, Abramowitz JS (2006) Thought control strategies used by parents reporting postpartum obsessions. J Cogn Psychother 20:435–445CrossRefGoogle Scholar
  19. Leckman JF, Mayes LC, Feldman R (1999) Early parental preoccupations and behaviors and their possible relationship to the symptoms of obsessive–compulsive disorder. Acta Psychiatr Scand Suppl 00:1–26CrossRefGoogle Scholar
  20. Lee C, Barr RG, Catherine N (2007) Age-related incidence of publicly reported shaken baby syndrome cases: is crying a trigger for shaking? J Dev Behav Pediatr 28:288–293CrossRefPubMedGoogle Scholar
  21. Maina G, Albert U, Bogetto F, Vaschetto P, Ravizza L (2000) Recent life events and obsessive–compulsive disorder (OCD): the role of pregnancy/delivery. Psychiatry Res 89:49–58CrossRefGoogle Scholar
  22. Nonacs RM, Cohen LS, Viguera AC (2005) Diagnosis and treatment of mood and anxiety disorders in pregnancy. In: Cohen LS, Nonacs RM (eds) Mood and anxiety disorders during pregnancy and postpartum. American Psychiatric, Washington, DC, USA, pp 17–51Google Scholar
  23. Parkinson L, Rachman SJ (1980) Are intrusive thoughts subject to habituation? Behav Res Ther 18:409–418CrossRefPubMedGoogle Scholar
  24. Purdon C, Clark DA (1993) Obsessive intrusive thoughts in nonclinical subjects: I. Content and relation with depressive, anxious and obsessional symptoms. Behav Res Ther 3:713–720CrossRefGoogle Scholar
  25. Rachman S, de Silva P (1978) Abnormal and normal obsessions. Behav Res Ther 16:233–248CrossRefPubMedGoogle Scholar
  26. Reitman D, Currier RO, Stickle TR (2002) A critical evaluation of the Parenting Stress Index—Short Form (PSI-SF) in a head start population. J Clin Child Adolesc Psychol 31:384–392PubMedGoogle Scholar
  27. Salkovskis PM, Harrison J (1984) Abnormal and normal obsessions: a replication. Behav Res Ther 22:549–552CrossRefPubMedGoogle Scholar
  28. Straus MA, Hamby SL, Finkelhor D et al (1998) Identification of child maltreatment with the Parent–Child Conflict Tactics Scales: development and psychometric data for a national sample of American parents. Child Abuse Neglect 22:249–270CrossRefPubMedGoogle Scholar
  29. Wisner KL, Peindl KS, Gigliotti T, Hanusa BH (1999) Obsessions and compulsions in women with postpartum depression. J Clin Psychiatry 60:176–180PubMedGoogle Scholar
  30. Wisner KL, Gracious BL, Piontek CM, Peindl K, Perel JM (2003) Postpartum disorders: phenomenology, treatment approaches and relationship to infanticide. In: Spinelli MG (ed) Infanticide: psychosocial and legal perspectives on mothers who kill. American Psychiatric, Washington, DC, pp 35–60Google Scholar
  31. Wolfe DA (1985) Child-abusive parents: an empirical review and analysis. Psychol Bull 97:462–482CrossRefPubMedGoogle Scholar
  32. Yali AM, Lobel M (2002) Stress-resistance resources and coping in pregnancy. Anxiety Stress Coping 15:289–309CrossRefGoogle Scholar

Copyright information

© Springer-Verlag 2008

Authors and Affiliations

  1. 1.Women’s Health Research InstituteVancouverCanada
  2. 2.Department of PsychologyUniversity of British ColumbiaVancouverCanada

Personalised recommendations