Grief and ruminative thought after perinatal loss among Turkish women: one-year cohort study

Abstract BACKGROUND: Among women who have suffered loss of pregnancy, the level of grief decreases gradually. Age, mental health status and childlessness are the factors known to mostly affect women’s levels of grief. OBJECTIVES: To assess the levels of grief among women who experienced perinatal loss and the changes in their ruminative thought styles over the first year after their loss. DESIGN AND SETTING: One-year follow-up study carried out in a university hospital in Turkey. METHODS: The study population included 70 women who experienced loss of pregnancy in the hospital. The sample size was calculated using G*Power V3.1. Data were collected at 48 hours, at the third month, at the sixth month and at one year after pregnancy loss, between June 2018 and June 2019. A personal information form, the Perinatal Grief Scale and the Ruminative Thought Style Questionnaire were used for data collection. RESULTS: The women’s highest levels of grief and ruminative thought style were in the first 48 hours. Their tendency towards grief and ruminative thought styles decreased over the repeated measurements during the follow-up. Women aged 20-29 years had the highest levels of grief at the third month after perinatal loss. CONCLUSIONS: Nursing assessments regarding grief and ruminative thought style over the first 48 hours after perinatal loss should be integrated into nursing care for these women. Grief follow-up programs for these women can be developed through nursing research.

have experienced perinatal loss. In this regard, there is one previous study assessing the level of grief among women with pregnancy loss in Turkey. 17 To the best of our best knowledge, no studies have evaluated the gradual changes in grief levels and ruminative thought styles among women who suffered perinatal loss in Turkey.

OBJECTIVE
The aim of this study was to evaluate differences in grief levels and ruminative thought styles among women who have experienced perinatal loss, by means of repeated measurements over the first year after the event.

METHODS
This was a one-year follow-up study carried out in a university hospital. The study population consisted of women who had experienced pregnancy loss in the hospital. The number of women who used the hospital's delivery service between April 1 and May 1, 2018, was 49. The sample size was calculated using the G*Power software (version 3.1.9.2; Universität Düsseldorf, Düsseldorf, Germany), in terms of the change in R 2 in multiple linear regression approximation. The minimum sample size required for seven predictors with 80% power and medium effect size (f 2 = 0.15) was calculated as 43 subjects. 18,19 However, through considering abandonment over the course of the repeated measurements of the study, we decided to include a total of 70 women in the study.
The women who were included in this study were voluntary participants who had experienced pregnancy loss in any trimester of pregnancy, and who had the ability to speak and write in Turkish. Women who had previously had a psychological disorder were not included in the study, and women whom the researchers were unable to reach during the repeated follow-ups were excluded.

Data collection forms and tools
Data were collected using a personal information form, the Perinatal Grief Scale (PGS) and the Ruminative Thought Style Questionnaire (RTSQ).

Personal information form
This form was prepared by the researchers. It consisted of 10 questions regarding sociodemographic and obstetric characteristics.

Perinatal grief scale
The PGS was developed by Toedter et al. and assesses the level of grief experienced after perinatal loss. 20 The original scale consists of 33 items on a five-point Likert-type scale and includes three subscales. 21 These three subscales are named Active Grief, Difficulty Coping and Despair. These levels represent progression of the pathological condition on the overall scale.
An Overall Grief score of 91 or higher means that grief is present. An Active Grief score of 34, a Difficulty Coping score of 30 and a Despair score of 27 are used as cutoff points on the three subscales. The Cronbach's α values lie between 0.86 and 0.92. A validity and reliability study was conducted in Turkey by Özgür Köneş et al. in 2017, and its Cronbach's alpha was found to be 0.95. 22 In the present study, the Cronbach's alpha values were between 0.793 and 0.857 for the subscales of the PGS and 0.930 for the total score.

Ruminative thought style questionnaire
The RTSQ was developed by Brinker and Dozois in 2009, and its adaptation to Turkish culture was published by Karatepe et al., in 2013. 14,23 This scale has 20 items. Assessments are made based on the total score. The lowest score is 20 and the highest score is 140. Its Cronbach's alpha in the above studies was 0.907. 14,23 In the present study, Cronbach's alpha was 0.886.

Data collection
Data were collected in four steps from June 2018 to June 2019.
In the first step, the questionnaires were completed for the first time just before the participants were discharged. These women had stayed in the hospital for 48 hours (T 0 ) after their pregnancy loss. The women's phone numbers and addresses were obtained at this time.
The same questionnaires were filled out through phone calls at the third month (T 1 ) and sixth month (T 2 ) and at one year (T 3 ) after discharge. The questionnaires at T 0 were filled out through faceto-face interviews conducted by the researchers. They explained to the participants that later on they would call them by phone, to fill out the same questionnaires again. The first interviews lasted approximately 30 minutes. The repeated interviews lasted approximately 15 minutes each.
At T 0 , 70 participants were involved in the study, but 13 participants did not answer the call at T 1 . The researchers were able to reach all of the T 1 participants again, at both T 2 and T 3 , and thus the study was completed with 57 participants. The researchers called all the women who did not answer the phone, at least three times before deciding to drop them from the study. All the details regarding the study process are explained in the study flowchart (Figure 1).

Data analysis
Statistical analyses were done using the IBM SPSS Statistics software V23 (IBM Corporation, Armonk, New York, United States) and Statistical assessments were made. testing group and time effects, and interactions. The Kruskal-Wallis test and one-way ANOVA test were used for comparisons of more than two independent groups. Relationships between two continuous variables were tested by means of the Spearman rank correlation. The multiple linear regression-backward elimination technique was used for estimation of RTSQ scores via independent variables. The significance level accepted was P < 0.05. University. The procedures used in this study adhered to the tenets of the Declaration of Helsinki. Informed consents were obtained from the participants after we had explained the objectives of the study to them.

RESULTS
The mean age of the participants was 30.34 ± 6.55 years, and the mean week in which pregnancy loss occurred was 15.42 ± 6.61.
Among all the participants, 45.7% had miscarriages and 54.3% had stillbirths. The perinatal deaths all occurred at between 4 and 32 weeks of gestation. Table 1 shows the sociodemographic and obstetric characteristics of the participants.
The median values of the total scores on the PGS (P < 0.001) and the mean scores on the RTSQ (P < 0.001) differed over time ( Table 2).
A statistically significant difference was found between total score median values on the PGS at T 1 in terms of the variables of age and childlessness (P < 0.05). The difference based on the age variable was caused by the 20 to 29-year age group (P < 0.05).
The total median scores on the PGS among women who were unable to have children were higher at T 1 than at other times (P < 0.05) (Table 3; Figure 2). Figure 2 shows the changes in the variables of age and childlessness on the PGS and RTSQ over the one-year follow-up.
According to the correlation results from the scales, positive medium-level correlations were found between the total scores on the RTSQ and the PGS and their subscales at T 0 , T 1 and T 2 (P < 0.05) ( Table 4). Also, it was found that the percentages of the women who had PGS total scores ≥ 91 were 55.7% at T 0 , 21.1% at T 1 , 3.5% at T 2 and 3.5% at T 3 . A score higher than this cutoff point means having grief. The changes in the RTSQ scores based on the PGS cutoff points over time are shown in Figure 3.
The multiple linear regression-backward elimination technique was used for RTSQ score estimation. The regression model included family type, childlessness, working status and PGS T 3 . PGS T 3 contained active grief, difficulty coping and despair. Dummy variables were created for categorical variables in the model. Active grief and family type were found to be statistically significant in the regression model (P < 0.05). They explained 15.6% of the change in RTSQ score ( Table 5).

DISCUSSION
The aim of this study was to evaluate grief and ruminative thought after a perinatal loss, among Turkish women.
This study was specific for perinatal grief in a prospective manner and it makes a valuable contribution to the literature relating to grief, given that this was the first study in Turkey to evaluate grief and ruminative thought style prospectively over a one-year period.
In this study, more than half of the women were experiencing grief at the interview after the first 48 hours. The median values of the grief score periodically decreased over the course of the follow-up measurements.
The grief levels of women who had experienced pregnancy loss have been reported in the literature. 17 26 This result indicated that couples experience active grief within the first three months and that their feelings change positively over time. The results from many studies have indicated that there are high levels of grief among women who recently had the loss of a pregnancy, and that perinatal grief decreases over time. [28][29][30][31] The results from the current study are in line with data in the literature. In Turkish society, individuals believe that they have something in their destiny and that this cannot be changed. The age variable had a significant effect on the median PGS total score at T 1 after the loss. Women aged 20-29 years had higher PGS     total scores than other age groups. Robert et al. stated that maternal age was a significant predictor of the level of grief and that there was a negative relationship between maternal age and perinatal grief level. 30 The results from the current study are consistent with those from that study. At early ages in the cycle of life, individuals may not yet have experienced any loss, and a pregnancy loss might be the first major loss in their lives. Women might become more capable of managing negative emotions as they acquire more experiences of life.

Number of pregnancy losses
In this current study, it was found out that already having children had a significant effect on the median value of the PGS total score at T 1 . Childless women had higher levels of PGS scores at T 1. Childlessness has been determined to be an important factor with regard to the duration of perinatal grief. 25 Moreover, Tseng et al. indicated that being a childless woman is a major risk factor for perinatal grief. 27 The results from our study were similar to those of all these studies. In Turkish society, individuals believe that women who lost their pregnancy may heal through have another living child. Being childless may cause a higher level of grief than that of women who have living children.
We found that the mean RTSQ score of the women was highest at T 0 . The mean scores on the RTSQ showed decreases over the course of the one year of measurements. Thus, the rumination level was highest in the early period of the loss, and it tended to decrease from high to medium rumination gradually.
Rumination is considered a cognitive process, in that it has an important role in various psychiatric disorders such as anxiety and mood state disorders. 23 Studies have reported that among women the risk of experiencing depression increases when their ruminative tendencies increase. 11,14,32,33 There are no studies in the literature that have evaluated ruminative thought styles among women after perinatal loss. However, several studies have assessed the correlations between rumination levels and anxiety, depression and psychological parameters. Previous studies have shown that rumination causes depressive symptoms because of negative thinking, weak problem-solving skills, insufficient coping behaviors and lack of social support. 33,34 Rumination has also been found to have positive correlations with depression, anxiety and negative automatic thoughts, and a negative correlation regarding satisfaction with life. 8,11,[13][14][15][16][33][34][35] Additionally, some studies have found that grief may occur in various pathological forms, and that chronic grief may cause depression, anxiety, phobias, obsessions and psychotic reactions. 36 Therefore, determining the level of rumination after pregnancy loss is quite important in terms of psychopathological conditions such as depression during the early days after the loss.
The results from this study showed that as rumination increases among women in the early period after their loss, grief also increases. This result was supported by the regression analysis in this study. In this current study, the results showed that almost all the women had recovered by the sixth month after the perinatal loss. However, women who are not recovering from grief present more rumination over time than do women who improve. There are no studies evaluating the relationship between perinatal grief and ruminative thought styles in the literature. However, a positive correlation has been reported in the literature between perinatal grief and depression. 24,29,37 It has been observed that the majority of the women who experienced perinatal loss showed depressive symptoms. 24,29 Also, similar studies have shown that the risk of incidence of depression after perinatal loss is high. 25,37 Determining the relationship between perinatal grief and ruminative thought styles might be a starting point for planning qualitative nursing care aimed at protecting and improving women's mental health after perinatal loss.

Limitations
The sample size was limited because of the number of dropouts.
For this reason, the study results can be generalized only for this population.

CONCLUSION
In this study, the levels of perinatal grief and ruminative thought styles among women who experienced pregnancy loss were assessed at the first 48 hours and at three months, six months and one year later. These women had high levels of grief and ruminative thought styles in the first 48 hours. However, their levels of grief and ruminative thought styles tended to decrease over the repeated measurements during the follow-up. Based on the results from this study, the following recommendations can be made: • Nursing assessments regarding grief and ruminative thought style over the first 48 hours after perinatal loss should be integrated into nursing care for these women.
• In clinics, nurses should be trained regarding to how to approach these women, especially during the first 48 hours after the loss.
• Grief follow-up programs for these women can be developed through nursing research, especially during the first six months after the loss.
• Institutions should provide counseling services.  Full model includes the following: family type; childlessness; working status; PGST 3 (Perinatal Grief Scale T 3 ); active grief; difficulty coping; despair.