- Top of page
Objective To compare maternal and paternal psychological responses following birth of a healthy baby; and to explore predictors of parental psychological distress.
Design A prospective, longitudinal, population-based cohort study.
Setting A Norwegian district general hospital.
Population One hundred and twenty-seven mothers and 122 fathers were included.
Methods Eligible consenting parents were enrolled. The assessments, which were performed zero to four days after birth, at six weeks and at six months, included General Health Questionnaire-28 (GHQ-28), State Anxiety Inventory and Impact of Event Scale. The response rates at the three occasions were 97%, 85% and 71%.
Main outcome measures Symptoms of intrusion, avoidance, arousal and psychological distress including anxiety, depression, social dysfunction and somatisation.
Results Clinically important psychological distress was reported by 37% of the mothers and 13% of the fathers a few days after childbirth (P < 0.001). Severe intrusive stress symptoms were reported by 9% and 2% of mothers and fathers, respectively (P= 0.002). Level of intrusive stress was the outcome that differed most clearly between mothers and fathers at all three points of time. Being a single parent, multiparity and a previous traumatic birth were significant independent predictors of acute maternal psychological distress. After six weeks and six months, the level of psychological distress including symptoms of depression fell to levels found in the general population.
Conclusions Childbirth does not seem to trigger long term psychological distress in most parents. Clinically important psychological distress occurred more frequently in mothers than in fathers. Acute maternal psychological distress was predicted by being a single parent, being multiparous, and having a previous traumatic birth.
- Top of page
Postnatal maternal depression may adversely influence infant and child development1,2. The prevalence of postnatal maternal depression ranges from 5.5% to 49%3–7 with consistently higher prevalence among mothers than fathers6,8–12. Paternal depression13 and stress responses may also influence the family atmosphere and thus child development. Nevertheless, most studies have had a limited focus, mostly on maternal postpartum depression weeks to months after birth. Recently, research focus has been expanded to include other psychological distress responses such as excessive worrying and post-traumatic stress symptoms in mothers14–16. Fathers' responses have not been addressed within this perspective, however. Post-traumatic stress disorder (PTSD)-like responses have been reported in 1–6% of women following childbirth14–18.
Provision of postnatal support in the community, in addition to the usual care provided by community midwives, has had no additional health benefit compared with traditional community midwifery visiting19. However, as long as we do not know the phenomenology and prevalence of various clinically important psychological outcome measures, we cannot know the clinical relevance of such studies. For example, the intervention may be inappropriate simply due to a mismatch between the treatment offered and the problems experienced. Recent reviews considering the effect of suggested interventions to prevent postnatal depression have failed to show significant improvement in the intervention group20,21.
Knowledge of the phenomenology of psychological distress in both parents to the birth of a healthy child, and the prevalence of clinically abnormal responses, are important. Delineation of abnormal responses requires knowledge of normal responses. Population-based studies of the full range of maternal and paternal psychological responses to childbirth are required. Such studies should address the cognitive, emotional and physiologic symptoms.
In order to cover the whole spectrum of psychological and behavioural responses in a clinically meaningful way, social role functioning, somatisation and psychological distress, such as anxiety and depressive symptoms, should be measured. The assessments should also include specific measurement of stress-related cognition, behaviour and symptoms of physiologic arousal. This information is needed for prevalence estimates of PTSD-like responses. To differentiate strong but brief normal responses from those heralding more long term psychological distress, a long term perspective is needed.
The aim of the present study was to meet those requirements. Based on clinical experience, data from previous studies of psychiatric disorder after birth3,5,7,22–27 and studies of mothers' responses to pregnancy termination and perinatal loss28, we expected that the prevalence of clinical psychological distress after six months should not exceed that reported for the general population. We did expect an increase in psychological distress acutely, however, but this response should be time limited. Our main hypothesis was that mothers were expected to demonstrate higher scores on all measures compared with fathers. Secondarily, we hypothesised that current birth complication, being a single parent, primiparity and previous traumatic births are risk factors for increased level of psychological distress after six months.
- Top of page
During November and December 1998, all parents of healthy babies born at Aust-Agder County Hospital were invited to participate in the study. The hospital is the only hospital in the county serving a population of 100,000 people. A healthy baby was defined as a baby without congenital malformation who was not admitted to the neonatal intensive care unit (NICU).
Psychological distress was assessed by the 28-item version of the General Health Questionnaire (GHQ-28)29. The GHQ-28 has four subscales: social dysfunction; health perceptions (somatisation); anxiety and insomnia; and depression. Each item has four answering categories. Likert's score (item score 0–1–2–3) and Case score (item score 0–0–1–1) were calculated. Total GHQ-28 score can be used to estimate the prevalence of clinically important psychological distress30. Psychological distress was measured by total GHQ-28 score, and clinically important psychological distress was defined as a Case score ≥6. Clinically important depression was defined as GHQ-28 depression subscale Case score ≥2. A Likert's score of 2–3 on either GHQ item 27 or 28 indicates suicidal ideation, and according to the study protocol, these subjects were contacted for provision of psychiatric help.
The State Anxiety Inventory (STAI-X1) consists of 20 items, which are rated on a four-level scale (1–2–3–4) with a possible score range 20–80. STAI-X1 measures current anxiety and reflects current subjective feelings concerning tension, apprehension, nervousness and worry31. Clinically important state anxiety was defined as a STAI-X1 score ≥40.
Stress-related cognition and behaviour was assessed by the Impact of Event Scale (IES)32. IES is a 15-item questionnaire with two subscales: seven items measure intrusion and eight items measure avoidance. The scoring range for each item is 0 (not at all) to 5 (very much). Intrusion deals with unbidden thoughts and images, troubled dreams, strong pangs or waves of feelings and repetitive behaviour related to the birth experience. Avoidance includes ideational constriction related to the birth experience, denial of the meanings and consequences of the birth, blunted sensation, behavioural inhibition and awareness of emotional numbness. A subscale score of 0–8 usually denotes minor responses, 9–19 moderate responses, and ≥20 severe responses33,34.
Subjective physiologic arousal was assessed by four items on GHQ-28 measuring irritability, startle response, hypervigilance and concentration difficulties35. Each item was dichotomised using Case score. Hyperarousal was defined as an arousal sum score ≥2 (range 0–4). Disturbance of sleep is common for both parents during the postnatal period15, thus sleep disturbance was not defined as part of arousal in the present study.
PTSD-like responses were defined as the triad of intrusion score ≥20, avoidance score ≥20 and hyperarousal at six weeks or six months, which is in accordance with the DSM-IV criteria28.
The study was approved by the Regional Ethics Committee and by the Regional Health Authorities. Information about the study was given orally and by written material. Informed consent to participate was obtained. Each parent completed a set of questionnaires. They were explicitly asked not to co-operate when completing the questionnaires. The majority of the participants completed the questionnaire the first day following birth or the next day and all within day four. Follow up questionnaires were mailed to the parents six weeks and six months after birth. If no response was obtained, the parents were encouraged to complete the questionnaire by a phone call.
Prenatal and postnatal factors were extracted from the medical charts. The independent variables included parental gender, age, parity, late miscarriage, previous traumatic birth, mode of delivery, current birth complications, previous maternal psychiatric illness and being a single parent.
Based on the paired nature of our data set, we applied paired Student's t test and McNemar's test for comparisons of mothers and fathers. Statistical analyses were conducted as two-tailed tests with a 0.05 level of significance. 95% confidence intervals were calculated. Multiple linear regression was performed separately in mothers and fathers to identify predictors of psychological distress. Calculations were performed by SPSS version 11.0 (Statistical Package for the Social Sciences, Chicago, Illinois).
- Top of page
During the study period, 274 parents of 138 healthy babies were candidates for the study. A further 38 parents were ineligible for the study because their baby was either admitted to the NICU (n= 17) or was stillborn (n= 2). Of the 268 parents available at the hospital, 127 mothers (93%) and 122 fathers (93%) gave their consent. Sixteen declined, two were excluded due to language problems and one due to other reasons. The response rates on the three occasions were 97%, 85% and 71%, respectively. Ninety mothers (71%) and 81 fathers (66%) responded on all three occasions. The number of pair responders (both parents in a couple) was 113, 102, and 84 on the three occasions, respectively.
The mean age of participating mothers and fathers was 28.2 years (range 18–39) and 31.2 years (range 17–49), respectively. The median number of previous pregnancies and births in the mothers were 2 (range 1–7) and 1 (range 0–3), respectively. One hundred and two had a normal vaginal delivery, seven had an instrumental vaginal delivery, 11 had an emergency caesarean section and seven had an elective caesarean section. Eight participating women and five participating fathers had no actual partnership. Three single fathers did not participate in the study. Four fathers were not attending in the delivery room. Most mothers had maternity leave throughout the study period.
Statistical comparisons between the 249 parents included in the study and the 25 parents who were excluded from the study showed no significant differences in age, number of previous children, mode of delivery, previous psychiatric illness according to medical records and previous traumatic birth. However, the mean number of previous pregnancies was significantly higher in mothers excluded from the study than in participating mothers (3.2 vs 2.4, P= 0.006).
Mothers reported significantly higher levels of psychological distress (22.0 vs 16.4, P < 0.001) than fathers acutely (Table 1). A similar pattern was found with anxiety (GHQ-28), somatisation, social dysfunction and intrusive stress with significantly higher scores in mothers than in fathers.
Table 1. Psychological responses in pairs of mothers and fathers after childbirth.
|Psychological response variable||Gender||0–4 days after birth||6 weeks||6 months|
|Mean||(95% CI)||P||Mean||(95% CI)||P||Mean||(95% CI)||P|
|Anxiety and insomniaa||Mothers' responses||0.864|| || ||0.605|| || ||0.581|| || |
|Fathers' responses||0.656|| || ||0.656|| || ||0.565|| || |
|Difference||0.208||(0.08, 0.34)||0.002||0.051||(−0.172, 0.070)||0.405||0.016||(−0.099, 0.132)||0.779|
|Depressiona||Mothers' responses||0.137|| || ||0.055|| || ||0.108|| || |
|Fathers' responses||0.101|| || ||0.137|| || ||0.079|| || |
|Difference||0.036||(−0.023, 0.095)||0.226||−0.083||(−0.138, −0.028)||0.004||0.029||(−0.040, 0.097)||0.413|
|Somatisationa||Mothers' responses||0.881|| || ||0.743|| || ||0.704|| || |
|Fathers' responses||0.622|| || ||0.703|| || ||0.628|| || |
|Difference||0.259||(0.132, 0.387)||<0.001||0.040||(−0.088, 0.168)||0.533||0.076||(−0.059, 0.210)||0.266|
|Social dysfunctiona||Mothers' responses||1.285|| || ||1.044|| || ||1.030|| || |
|Fathers' responses||0.971|| || ||1.064|| || ||1.016|| || |
|Difference||0.314||(0.223, 0.405)||<0.001||−0.020||(−0.100, 0.061)||0.633||0.014||(−0.064, 0.091)||0.725|
|Psychological distressa||Mothers' responses||22.0|| || ||17.2|| || ||16.7|| || |
|Fathers' responses||16.4|| || ||17.9|| || ||15.9|| || |
|Difference||5.6||(3.3, 7.9)||<0.001||−0.7||(−2.9, 1.3)||0.473||0.8||(−1.4, 3.0)||0.451|
|State anxietyb||Mothers' responses||30.5|| || ||29.2|| || ||29.8|| || |
|Fathers' responses||29.5|| || ||30.2|| || ||30.1|| || |
|Difference||1.0||(−0.9, 2.9)||0.314||−1.0||(−3.1, 1.1)||0.335||0.3||(−2.4, 1.7)||0.737|
|Intrusionc||Mothers' responses||7.8|| || ||4.4|| || ||4.2|| || |
|Fathers' responses||4.6|| || ||2.4|| || ||1.8|| || |
|Difference||3.2||(1.6, 4.8)||<0.001||2.0||(0.6, 3.3)||0.005||2.4||(1.3, 3.5)||<0.001|
|Avoidancec||Mothers' responses||3.4|| || ||1.6|| || ||2.7|| || |
|Fathers' responses||3.0|| || ||1.2|| || ||1.4|| || |
|Difference||0.4||(−0.7, 1.5)||0.495||0.4||(−0.2, 0.9)||0.215||1.3||(−0.1, 2.6)||0.072|
A significantly higher proportion of mothers than fathers (37 vs 13%, P < 0.001) reported clinically important psychological distress in the acute phase (Table 2). Looking at the 111 pair respondents regarding clinically important psychological distress in the acute phase, we identified seven pairs with both parents being cases, 31 pairs with the mother only being a case, eight pairs with the father only being a case and 65 pairs with both parents being without clinically important psychological distress. McNemar's test for paired data showed that the frequency of clinically important psychological distress was significantly increased in mothers compared with fathers (38/111 vs 15/111, P < 0.001).
Table 2. Prevalence of parental psychological casedness on selected psychological outcomes after childbirth. The numbers presented include all valid respondents. Values are given as % (n/N).
|Outcome||Gender||0–4 days after birth||6 weeks||6 months|
|Psychological distressa||Mother||37 (46/124)#||21 (23/108)##||19 (17/91)|
|Father||13 (15/115)||11 (11/102)||11 (9/84)|
|State anxietyb||Mother||12 (15/125)||8 (10/108)||12 (11/92)|
|Father||11 (12/114)||13 (13/101)||12 (10/85)|
|Depressionc||Mother||6 (7/126)||1 (1/109)||2 (2/91)|
|Father||2 (2/115)||2 (2/103)||1 (1/84)|
|Suicidal ideationd||Mother||0 (0/126)||0 (0/109)||0 (0/92)|
|Father||0 (0/115)||3 (3/102)||1 (1/82)|
|PTSD-like responsese||Mother||0 (0/126)||0 (0/108)||1 (1/92)|
|Father||0 (0/115)||0 (0/103)||0 (0/85)|
Clinically important state anxiety (STAI-X1) related to childbirth was reported by 12% of the mothers and 11% of the fathers. Clinically important depression (GHQ-28) was reported by 6% of mothers and 2% of fathers. None of the parents had scores indicating suicidal ideation (Case score ≥ 1 on GHQ item 27 or 28) during the first days after childbirth.
Acutely, 9% of the mothers compared with 2% of the fathers (P= 0.001) reported severe intrusive stress. Moderate or severe intrusive stress was observed in 34% and 18% of mothers and fathers, respectively. Severe avoidance symptoms were reported by one mother and two fathers only (Table 3). The specific pattern of stress responses is presented in Tables 4 and 5. Mothers scored significantly higher than fathers on six of seven items assessing intrusion. However, regarding avoidance, mothers and fathers had similar scores.
Table 3. Postnatal stress scores in mothers and fathers after childbirth expressed as the percentage with minor (0–8), moderate (9–19) and severe (≥20) stress responses.
| ||0–4 days after birth||6 weeks||6 months|
|Intrusion in mothers||109||66.1||24.8||9.2||98||86.7||12.2||1.0||82||89.0||8.5||2.4|
|Intrusion in fathers||109||81.7||16.5||1.8||98||93.9||4.1||2.0||82||97.6||0||2.4|
|P|| ||0.002|| || || ||0.118|| || || ||*|| || |
|Avoidance in mothers||100||90.0||9.0||1.0||97||97.9||2.1||0||82||95.1||2.4||2.4|
|Avoidance in fathers||100||91.0||7.0||2.0||97||96.9||3.1||0||82||98.8||0||1.2|
|P|| ||1.000|| || || ||1.000|| || || ||*|| || |
Table 4. Frequency of endorsement (item score ≥1) and mean score of postnatal intrusive symptoms after childbirth according to the Impact of Event Scale.
|Item|| ||0–4 days after birth||6 weeks||6 months|
|Frequency (%)||Mean score||Frequency (%)||Mean score||Frequency (%)||Mean score|
|1. I've had waves of strong difficult feelings about the birth of my baby||Mother||62||1.5||32||0.6||31||0.7|
|P|| ||0.002|| ||0.202|| ||0.008|
|2. Things I saw or heard suddenly reminded me of the birth||Mother||75||1.5||61||1.4||62||1.4|
|P|| ||0.015|| ||<0.001|| ||<0.001|
|3. I have thoughts of the birth of my baby when I didn't mean to||Mother||50||1.1||20||0.3||16||0.2|
|P|| ||0.036|| ||0.921|| ||0.133|
|4. Pictures about the birth have popped into my mind||Mother||77||1.7||74||1.5||74||1.3|
|P|| ||<0.001|| ||<0.001|| ||<0.001|
|5. Any reminder brought back difficult feelings about the birth of my baby||Mother||31||0.6||11||0.2||17||0.3|
|P|| ||0.059|| ||0.892|| ||0.083|
|6. I've had trouble falling asleep or staying asleep because of pictures or thoughts about the birth of my baby come into my mind||Mother||37||0.7||2||0.0||7||0.2|
|P|| ||<0.001|| ||0.368|| ||0.333|
|7. I've had bad dreams about the birth||Mother||18||0.4||6||0.1||10||0.2|
|P|| ||0.005|| ||0.843|| ||0.181|
Table 5. Frequency of endorsement (item score ≥1) and mean score of postnatal avoidance symptoms after childbirth according to the Impact of Event Scale.
|Item|| ||0–4 days after birth||6 weeks||6 months|
|Frequency (%)||Mean score||Frequency (%)||Mean score||Frequency (%)||Mean score|
|8. I know that a lot of unresolved feelings are still there, but I've kept them under wrap||Mother||28||0.5||12||0.2||21||0.3|
|P|| ||0.049|| ||0.095|| ||0.090|
|9. I've avoided letting myself get emotional when I thought about the birth of my baby or was reminded of it||Mother||43||1.0||31||0.6||32||0.7|
|P|| ||0.857|| ||0.963|| ||0.314|
|10. I've wished to banish memories about the birth of my baby||Mother||17||0.3||10||0.2||12||0.2|
|P|| ||0.068|| ||0.055|| ||0.035|
|11. I have tried not to talk about the birth of my baby||Mother||8||0.2||8||0.1||11||0.2|
|P|| ||0.428|| ||0.657|| ||0.116|
|12. I've felt unrealistic about it, as if the birth of my baby hadn't happened or as if it was not real||Mother||26||0.5||29||0.4||21||0.4|
|P|| ||0.883|| ||0.004|| ||0.051|
|13. I have stayed away from things or situations that might remind me of the birth of my baby||Mother||6||0.1||3||0.0||6||0.2|
|P|| ||0.426|| ||0.320|| ||0.139|
|14. My feelings about the birth of my baby were kind of numb||Mother||10||0.2||1||0.0||6||0.2|
|P|| ||0.288|| ||0.320|| ||0.326|
|15. I have not let myself have thoughts about the birth of my baby||Mother||16||0.5||5||0.1||15||0.4|
|P|| ||0.348|| ||0.254|| ||0.419|
Irritability (30%vs 9% in mothers and fathers, respectively, P < 0.001) and hypervigilance (11%vs 6%, P= 0.15) were the two most common arousal symptoms in the acute phase. A significantly higher proportion of mothers compared with fathers reported hyperarousal immediately after the delivery (14%vs 3%, P < 0.001).
Although each dimension of postpartum stress was endorsed relatively frequently, the triad of severe intrusion, severe avoidance and hyperarousal was not observed acutely.
In general, the differences between mothers and fathers on several psychological outcomes were diminished at follow up (Table 2). The proportion of mothers with clinically important psychological distress remained increased compared with their partners. Three fathers reported suicidal ideation at six weeks, and one persisted at six months. None of the mothers reported suicidal ideation at follow up.
During follow up, less than 3% reported severe intrusive or avoidance distress (Table 3). The mean intrusion scores were significantly higher in mothers than fathers at follow up (Table 1). At six months, one mother suffered from the triad of severe intrusion, severe avoidance and hyperarousal suggesting a PTSD-like response.
Multiparous women had significantly higher GHQ-28 scores in the acute phase than primiparous women on all subscales except for depression. Mode of delivery was not related to stress response patterns in mothers or fathers.
Mothers with previous mental health problems, as recorded in the hospital charts, scored significantly higher concerning somatisation and total GHQ-28 score at birth and concerning state anxiety (STAI-X1) at six weeks and six months.
Eight pairs of mothers and fathers were living isolated from each other, and increased total GHQ-28 score was found in both single mothers and single fathers.
Separate multivariate equations were constructed for mothers and fathers, and the results are presented in Table 6. In mothers, the model explained 47% of the variance in the acute phase. Six independent predictors of acute maternal psychological distress (total GHQ-28 score) were identified. Out of four dichotomous clinical predictors, the effect was strongest for previous traumatic birth (β= 10.7) followed by current birth complication, being a single mother and multiparity. When controlling for the other five predictors, women with a previous traumatic birth scored 10.7 (95% CI 3.4–18.0) higher on total GHQ-28 in the acute phase than those women who had not. Previous traumatic birth was also an important predictor of total GHQ-28 at follow up with β= 6.8 and β= 11.0, although significance was not reached at six weeks. Regarding the psychometric predictors (continuous variables), a one unit increase in maternal acute IES score predicted an increase in acute total GHQ-28 score by 0.44. Thus, if acute maternal IES score increased by 10 units, an increase by 4.4 in total GHQ score would be predicted from the regression equation.
Table 6. Predictors* of parental psychological distress (total GHQ-28 score) in the first six months after childbirth.
|Parent||Predictor||0–4 days after birth||6 weeks||6 months|
|β||95% CI||P||β||95% CI||P||β||95% CI||P|
|Mothers||Current birth complication||7.1||(1.6, 12.5)||0.012||4.3||(−1.2, 9.8)||0.125||2.7||(−4.0, 9.5)||0.424|
|Single mother||6.9||(0.3, 13.6)||0.041||3.3||(−3.1, 9.8)||0.306||0.38||(−7.8, 8.6)||0.928|
|Multiparity||6.9||(3.8, 9.9)||<0.001||0.14||(−2.9, 3.2)||0.930||1.3||(−2.6, 5.2)||0.515|
|Previous traumatic birth||10.7||(3.4, 18.0)||0.004||6.8||(−0.9, 14.5)||0.085||11.0||(2.0, 20.0)||0.017|
|Intrusion score acutely (IES)||0.44||(0.23, 0.65)||<0.001||0.04||(−0.17, 0.26)||0.683||0.24||(−0.03, 0.51)||0.077|
|Anxiety score acutely (STAI)||0.46||(0.21, 0.70)||<0.001||0.32||(0.08, 0.57)||0.011||0.22||(−0.08, 0.53)||0.148|
|R2||0.47|| || ||0.14|| || ||0.17|| || |
|Fathers||Single father||15.4||(9.4, 21.3)||<0.001||2.5||(−5.7, 10.8)||0.546||−3.5||(−15.8, 8.9)||0.578|
|Previous children||1.2||(−0.8, 3.2)||0.234||4.0||(1.0, 7.0)||0.010||−0.06||(−3.0, 2.8)||0.968|
|Previous traumatic birth||4.3||(−1.6, 10.2)||0.154||−7.1||(−16.9, 2.8)||0.156||−0.22||(−8.9, 8.4)||0.960|
|Intrusion score acutely (IES)||0.30||(0.09, 0.51)||0.005||0.28||(−0.02, 0.59)||0.067||0.30||(0.01, 0.59)||0.045|
|Anxiety score acutely (STAI)||0.38||(0.26, 0.50)||<0.001||0.16||(−0.06, 0.37)||0.147||0.17||(−0.03, 0.37)||0.091|
|R2||0.53|| || ||0.17|| || ||0.13|| || |
Immediately after birth, the paternal regression equation, which included five potential predictors, explained 53% of the variance. Three independent predictors were identified acutely. Being a single father predicted an increase by 15.4 in total GHQ-28 score, and high intrusion score (IES) and high state anxiety (STAI-X1) were associated with increased total GHQ-28 score, all reaching significance. In fathers, the total GHQ-28 score at all three time points could be predicted by acute IES intrusion score, an increase by one unit in acute IES score predicted an increase in total GHQ-28 score by approximately 0.3 at all three occasions.
- Top of page
The present study clearly demonstrates that within a six month perspective, the birth of a healthy child only provokes long term psychological distress in 19% of mothers and 11% of fathers as measured by GHQ-28. Analyses of stress-specific self-reported cognitive, emotional and behavioural symptoms (IES) and the low state anxiety (STAI-X1) confirm this finding.
Several studies have addressed psychological response to adverse events, using similar methods33–36. As expected, mothers who experienced pregnancy termination due to fetal anomalies or perinatal loss28 had markedly higher stress scores (IES) than the mothers of healthy babies in the present study.
Women who had a previous history of a subjective a traumatic birth were at significantly increased risk of developing clinically important psychological distress. This finding agrees with a recent study by Hughes et al.37, who found that a previous third semester stillbirth was a risk factor for depression and anxiety in a subsequent pregnancy. The consequences of negative birth experiences have been further investigated in a recent Swedish study, which showed that a negative birth experience is related to a reduced probability of having a subsequent child38.
Increased acute stress responses among single parents are expected because poor social circumstances and social roles are associated with self-reported postpartum depressive symptoms39,40. Single mothers may have less social support7, but may also suffer from more psychological and social problems, each increasing the likelihood of an increased acute stress response.
In a meta-analysis based on 59 studies, no association was found between parity and the mothers' postpartum mental health24. A Brazilian study demonstrated an association between lower parity and maternal postpartum maternal depression41. In contrast to these findings, we found that multiparous women reported significantly higher GHQ-28 scores in the acute phase than primiparous women on all subscales except for depression, even when the effect of a previous traumatic birth was controlled for. Recently, a large Danish study reported a significant association between high parity (>2) and postpartum depression at four months, but no significant difference between nullipara and para1–2 combined4. The association between increased psychological distress and having previous children may be a reflection of the increased demand on mothers who have to care for both the newborn child and the family at home. We cannot rule out, however, that modern women may experience additional births as more distress provoking than their first birth.
Acutely, about 10% of these mothers report increased level of intrusive symptoms suggesting clinically important stress. Mothers consistently scored higher than fathers on most measures, a finding which is understandable in terms of the different biologic roles. This finding is supported by an increase in the percentage of high total GHQ-28 Case score among mothers acutely. Nevertheless, persisting high avoidance and arousal scores, which are typical of clinically important stress responses33, were almost non-existent. These findings confirm clinical experience that following the birth of a healthy child, acute stress responses may occur, but such responses are infrequent and time limited. They seldom indicate the need for specific psychological interventions.
Wijma et al.14 conducted a cross sectional study on 1640 mothers applying the Wijma Delivery Expectancy/Experience Questionnaire 1–13 months after childbirth, and a PTSD profile was found in 1.7% of the women. Czarnocka and Slade15 recently published a comprehensive study of 264 mothers assessing post-traumatic stress symptoms, anxiety and depression. In that study, follow up was restricted to the first six weeks after birth, and 3% had high scores on all three dimensions of PTSD and a further 24% had elevated scores on at least one of the three dimensions of PTSD. Our finding of one mother with a PTSD-like response at six months is in accordance with the studies by Wijma and Czarnocka, but our sample size is insufficient for the purpose of prevalence estimation of PTSD-like responses.
In a prospective study, Astbury42 found that maternal state anxiety increased during late pregnancy with a marked fall after birth. However, the routine use of prenatal ultrasonography may have altered the natural course of pregnancy and postpartum anxiety. We have no data from our study that can answer whether childbirth reduces or increases psychological distress compared with levels occurring throughout pregnancy.
To our best knowledge, there is no study that has addressed the psychological response to the birth of a healthy child applying modern concepts of stress response patterns in both mothers and fathers for six months after the birth. Fathers' responses to childbirth have largely been neglected in previous studies43, despite the fact that the vast majority of fathers today attend the birth.
Our study indicates that for most fathers, the birth of a healthy child does not provoke a clinically important stress response either in the acute or in the later phase. The paternal psychological distress scores showed little variation over time expect for intrusion and avoidance. Clinically important psychological distress and state anxiety was reported by about 10% of the fathers on all three occasions, and clinically important depression was rare (1–2%). However, six weeks after childbirth, three fathers demonstrated suicidal ideation and one persisted at six months.
Not unexpectedly, being a single father predicted increased acute psychological distress. Interestingly, in fathers, increased acute intrusion score (IES) and acute anxiety score (STAI-X1) were found to be independent predictors for psychological distress at six weeks and six months. Previous studies have found an association between acute stress responses and personal vulnerability and psychopathology. Used longitudinally, GHQ has been found to be a valid measurement of clinically important psychopathology. Based on the above findings, we speculate that the association between increased acute intrusion in fathers and subsequent increased psychological distress measured by GHQ is most likely explained by birth-independent vulnerability or psychopathology in a subgroup of fathers.
The observed prevalence of maternal postnatal depression was low compared with previous reports6,9,12,44. GHQ-28 has been demonstrated to have a sensitivity and specificity comparable to other self-reported assessments of depression and mental health. Nevertheless, in particular, the GHQ-28 depression subscale delineates the more severe forms of depression (major depression) compared with other self-report questionnaires. By using Likert's score and the 10-item Edinburgh Postnatal Depression Scale (EPDS) with depression defined as a score above 9, Bergant et al.45 found that 20% of the women were depressed five days after delivery, and in accordance with other investigators probably included both cases with minor and major depression5. Lee et al.46 found that 11.3% of the women were depressed according to EPDS (scores above 12) six weeks after delivery, but only 5.5% met the DSM-III-R criteria for major depression assessed by Structured Clinical Interview (SCID).
An alternative explanation is that previous studies may have confounded measures of psychological distress with clinical depression. This interpretation is supported by the fact that the prevalence of psychological distress immediately after birth in the present study is in accordance with recent reports on maternal depression39,45.
We included an assessment zero to four days after birth which has been reported rarely in the literature8,11,39,45. The finding of peak prevalence of psychological distress immediately after birth in the present study is in accordance with recent reports on maternal depression39,45.
The strength of this study is the population-based, longitudinal design and a high response rate. By including consecutive parents in a well-defined geographic area, selection bias is minimised. Moreover, by including different ways of assessing psychological responses, we have obtained clinically meaningful and sensitive measures, strengthening our conclusions. Furthermore, the present study is a comprehensive study on father's response to childbirth.
Some restrictions to the generalisability of the findings do exist, however. The present study was conducted in a country with high standards of hospital care. Furthermore, hospital care is free of charge for the mothers and their partners. The social benefits for mothers of newborn children are good. Thus, our findings of low levels of postnatal distress in most mothers and fathers are most likely valid only for countries were the social security system for mothers of newborns is optimal. The application of paired tests excluded about 10% of the women, but only a few fathers. The respondents who were excluded because paired statistical tests were used appeared to have higher scores than the included pair respondents, thus a bias with under-estimation of psychological distress probably has been introduced. However, this bias does not invalidate our findings, but gives a more conservative estimate.
The sample of the current study is of moderate size. For the analysis of continuous outcome variables, which was the main objective of the study, we consider the sample size appropriate based on a priori power calculations. However, for the analyses of caseness of rare outcomes, the sample size is insufficient. For instance, this holds true for PTSD-like responses, which were not among the main outcomes of the study.
According to Medline, 954 studies have been published using GHQ and 563 studies using STAI studying different populations' responses to all kinds of stress exposures ranging from daily hassles via upsetting life events to distress related to physical illness and disasters. Similarly, 484 studies have applied IES to different stressful life events. Based on those findings and clinical experience, we did not expect to find high levels of distress following the birth of a normal baby. Thus, we decided not to replicate previous studies3 by including a control group. One may argue that we should have included an age-matched control group of parents living with children from the general population. However, considering the studies which have applied the same methods as used in the present study, which also includes studies looking at normal populations' level of psychological distress, we did not think that such a control group would add to our understanding or interpretation of our findings. Accordingly, we decided to focus our study on the comparison between mothers and fathers and discuss those responses in relation to previous studies.
The study has shown that in most mothers and fathers long term psychological distress triggered by the birth of a healthy child does not occur. Although the study is not an intervention study, these findings suggest that psychological interventions beyond routine psychosocial care are not warranted. In fact, it is even possible that compulsory psychological interventions during the first days may interfere negatively with the natural course of falling levels of distress. This is the lesson learned from debriefing studies aimed at subjects exposed to stressful events like injuries47, cardiac events48 and accidents49. However, we found that single mothers, those who have experienced a previous traumatic births and multipara are at increased risk of postnatal psychological distress, and may benefit from being offered additional emotional and social support during and after delivery.