Pregnancy loss occurs in at least 15% of known pregnancies, (1) which is typically before the 20th week ("miscarriage"). Although subsequent emotional reactions are diverse, pregnancy loss may trigger depressive symptoms and disorder (2) and may also be psychologically traumatic. (3,4) In fact, we found recently that pregnancy loss may result in substantial symptoms of posttraumatic stress disorder (PTSD). (5) PTSD involves reexperiencing the traumatic event (eg, intrusive recollections, nightmares, being upset at reminders), avoidance of its reminders and numbing (eg, cognitive avoidance, feeling cut-off from others), and hyperarousal (eg, irritability, sleeping problems). (6) Although usual threatening events associated with psychological trauma are external to the individual, such as war, interpersonal violence, and disasters, there is increasing recognition that threatening medical events may lead to the same constellation of symptoms. (6-8) Consistent with the PTSD literature, (9,10) we found considerable individual differences in PTSD symptom severity, and only part of the variance was explained by the gestational age of the loss.
What factors render women resilient to PTSD symptoms after pregnancy loss? Ehlers and Clark (10) proposed that individuals recover from a traumatic event when (1) memory for the event is integrated into the context of other experiences, (2) avoidance strategies are dropped, and (3) negative appraisals of the event and its aftermath are modified. These propositions are supported by various studies after different events. (10) It is interesting that these factors bear a close resemblance to the "sense of coherence" (SOC) concept. (11) Antonovsky (11 (p. 19)) defined the SOC as the ability to perceive a stressor as (1) comprehensible (stimuli are viewed as structured, predictable, and explicable), (2) manageable (resources are available to meet associated demands), and (3) meaningful (demands are challenges, worthy of investment and engagement). Women with this natural inclination might be at an advantage in coming to terms with pregnancy loss. Retrospective studies report a relationship between a high SOC and less posttraumatic stress after various traumatic events. (12-14) For example, 6 months after a traffic accident, the SOC and PTSD symptoms were strongly correlated, r (51) = -.50, p [less than or equal to] .001. (13) However, in these previous studies, the SOC was assessed after the traumatic event took place, which does not establish causal order: the SOC may affect PTSD symptoms, it may be the other way around, or both. Although the SOC is fairly stable over time, (15) it may be influenced by stressful events, particularly when efforts at mastery fail. (16)
Findings of retrospective studies suggest that the SOC may also protect against depressive symptoms after a stressful life event. (17) However, this may be partly due to content overlap. (16) Depressive symptoms are elevated during pregnancy, (18) are common after pregnancy loss, and often co-occur with PTSD. (6) If the SOC is indeed a resilience factor for depressive symptoms after pregnancy loss, this relationship should still stand after statistically controlling for depressive symptoms before pregnancy loss. It is unclear whether this is the case.
How might the SOC protect against PTSD symptoms? According to Antonovsky, (11) higher levels of the SOC enable an individual to mobilize resources, such as social support, to cope with a stressor. Prior research has shown that more crisis support predicts less posttraumatic reactions after pregnancy loss (3) and other stressful events. (19,20) Important aspects of support are confiding in others, emotional support, and practical support. (20) For example, the use of social interactions to manage upsetting memories (eg, "I ask my friends if they have similar thoughts") is related to less PTSD symptoms. (21) It is unclear whether the SOC predicts the mobilization of crisis support after pregnancy loss. In fact, after pregnancy loss, particularly after a miscarriage, support is often deficient: the woman's partner may deal with the loss differently, (22) friends and family may not have known about the pregnancy, (23) and healthcare workers often underestimate its psychological impact. (24)
It is difficult to examine resilience factors for PTSD because experimental studies cannot replicate real-life threat, and prospective studies are nearly unfeasible because of the randomness of traumatic events. In our previous study of PTSD after pregnancy loss, (5,25) participants enrolled in the study in early pregnancy, which enabled us to measure the SOC and depressive symptoms before pregnancy loss. The present article is part of that study and focuses on (1) the relationship between the SOC in early pregnancy and PTSD symptoms after pregnancy loss, (2) the relationship between the SOC and postloss depressive symptoms when controlling for preloss depressive symptoms, and (3) crisis support as a mediator between the SOC and PTSD symptoms. To our knowledge, this is the first prospective study of the SOC and PTSD symptoms.
METHOD
Participants
A total of 1,372 pregnant women responded to ads in Dutch family magazines recruiting participants ([less than or equal to] 12 weeks pregnant) for study of pregnancy experiences. They completed baseline questionnaires, including the SOC Scale (SOC-13) (11,26) and Beck Depression Inventory (BDI). (27) For every 2 months thereafter until 1 month after the birth due-date, they completed pregnancy-related surveys that also asked them to inform the researchers if they had a pregnancy loss. Of the original sample, 33 women did not complete the baseline questionnaires and dropped out of the study. One hundred and twenty six women (9%) had a pregnancy loss and, about 1 month later, 118 of them (response rate = 94%) completed the PTSD Symptom Scale (PSS-SR), (28) BDI, and Crisis Support Scale (CSS). (29) We sent all of the questionnaires by mail with a postage-paid return envelope. Since 1 woman did not complete the SOC, the results are based on the remaining 117 women. The respondents were on average 31 years old (SD = 4), nearly all were married or cohabiting, about 40% were college-educated, and 30% were childless. The mean gestational age was 8 weeks (SD = 2.0) at the time of enrollment in the study, and it was 12 weeks (SD = 6; range 5-40) at the time of the loss. About 95% of losses occurred before the 20th week. We complied with the American Psychological Association (30) ethical guidelines in the treatment of our sample.
Measures
SOC-13 contains 13 items (eg, "Do you often have very mixed-up feelings and ideas? How often do you have the feeling that there's little meaning in the things you do in your daily life?") that are rated on a 7-point scale. Ratings are summed (higher scores reflect higher coherence). The PSS-SR includes the 17 PTSD symptoms, (6) which are scored on a 4-point severity scale (0 = not at all, 3 = almost always). Five items pertain to the "reexperiencing" symptoms, seven tap "avoidance," and five measure "hyperarousal." These subscales can be combined to yield PTSD symptom severity. The PSS-SR has established reliability and validity. (28) The widely used BDI comprises 21 symptoms of depression that are rated on a 4-point severity scale. In the present study, we excluded 1 item related to weight loss. The CSS is a 6-item self-report scale of received crisis support, which refers to (1) the availability of others, (2) confiding in others, (3) emotional support, (4) practical support, (5) being let down (reverse scored), and (6) contact with others who had a similar experience. Each item is rated on a 7-point Likert-type scale ranging from (1) never to (7) always. Its internal consistency is adequate. (29) The questionnaire instructs respondents to think of the friends and family they turned to for help, advice, and support following a crisis. For the purposes of this study, items (1) through (5) were administered separately for the partner, friends and family, and doctors and nurses after pregnancy loss. We operationalized overall crisis support as the average of these ratings. We calculated contact with peers (6) separately.
RESULTS
