Longitudinal path analysis for the directional association of depression, anxiety and posttraumatic stress disorder with their comorbidities and associated factors among postpartum women in Northwest Ethiopia: A cross-lagged autoregressive modelling study

Introduction Vulnerability for depression, anxiety and posttraumatic stress disorder symptoms due to perceived traumatic birth increase during the postpartum period. Traumatic birth has been defined as an event occurring during labour and birth that may be a serious threat to the life and safety of the mother and/or child. However, the comorbidity and multimorbidity of depression, anxiety and PTSD with their direct and indirect predictors is not well investigated in the postpartum period. In addition, the longitudinal directional association of depression, anxiety and PTSD with their comorbidities is not studied in Ethiopia. Objective The aim of this study was to assess prevalence of postnatal comorbid and multimorbid anxiety, depression and PTSD. It also aimed to determine the directional association of postnatal anxiety, depression and PTSD with the comorbidity and multimorbidity of these mental health problems over time and to explore the factors that are directly or indirectly associated with comorbidity and multimorbidity of anxiety, depression and PTSD. Methods A total of 775 women were included at the first, second and third follow-up of the study (6th, 12th and 18th week of postpartum period) during October, 2020 -March, 2021. A crosslagged autoregressive path analysis was carried out using Stata 16.0 software in order to determine the autoregressive and cross-lagged effects of depression, anxiety and PTSD with their comorbidities. In addition, a linear structural equation modelling was also carried out to determine the direct and indirect effects of independent variables on the comorbidities of depression, anxiety and PTSD. Results Comorbidity of anxiety with depression was the most common (14.5%, 12.1% and 8.1%) at the 6th, 12th and 18th week of postnatal period respectively. With regard to the direction of association, comorbidity of PTSD (due to perceived traumatic birth) with depression, PTSD with anxiety, depression with anxiety and triple comorbidity predicted depression and anxiety in subsequent waves of measurement. Direct and indirect maternal morbidity, fear of childbirth and perceived traumatic childbirth were found to have a direct and indirect positive association with comorbidities of depression, anxiety and PTSD. In contrast, higher parity, higher family size and higher social support had a direct and indirect negative association with these mental health disorders. Conclusion Postnatal mental health screening, early diagnosis and treatment of maternal morbidities, developing encouraging strategies for social support and providing adequate information about birth procedures and response to mothers’ needs during childbirth are essential to avert comorbidity of anxiety, depression and PTSD in the postpartum period.

As part of the health facility linked community based prospective follow-up study conducted in Northwest Ethiopia to determine the effect of maternal morbidities on maternal health related quality of life, functional status and mental health problems [56,57], postpartum women were recruited in four hospitals of south Gondar zone, Northwest Ethiopia. The data collection took place between October 1, 2020 and March 30, 2021. South Gondar is located at 650 km Northwest from Addis Ababa the capital city of Ethiopia. A total of 775 women consented to participate in the study and participated at the first, second and third follow-up of the study (6th, 12th and 18th week of postpartum period). Recruitment of the study participants was done after child birth and before the time of discharge within the hospitals where women gave birth. All women who were asked for consent agreed to participate in the study. The selected women were among those who were attending their antenatal care (ANC) in the study hospitals and among those who came for delivery in the study hospitals from different health centers and/or hospitals through referral. Therefore, the total number of delivering women in the study hospitals were not known before the initiation of the study. Women differ in their socio-demographic characteristics, reproductive, obstetrics and medical variables. The effect of these variations on the outcome variables were controlled with the use of multivariable analysis in this study (multivariable linear structural equation modelling). Sample size was calculated using a two-population proportion formula with Epi-Info version 7. Accordingly, a minimum sample size of 746 (249 exposed and 497 non-exposed) was calculated by taking 0.05 alpha (α), power of 90%, odds ratio of 4.23, 2.3% of mothers without depression during pregnancy who develop PTSD in postpartum period, 1:3 ratio of exposed to non-exposed (since the controls were 3 times the cases in a previous study [58]) and by adding 10% non-response rate. These parameters used for the sample size calculation were taken from a previous study [58]. Women aged 15 years and above, with preterm birth (between 28–36 weeks), term or post term delivery and with live birth, still birth or fetal death were included in the study. Women who are unable to communicate (having hearing problem and cannot communicate with sign language) were excluded from the study. The direct and indirect maternal morbidities were identified based on the WHO maternal morbidity working group criteria [59]. According to the WHO maternal morbidity working group criteria, the direct maternal morbidities included in this study were: gestational hypertension, pre-eclampsia, eclampsia, placenta previa, placental abruption, postpartum hemorrhage, mastitis, puerperal sepsis, urinary tract infection, perineal tear, episiotomy wound infection, vaginal wall/perineal laceration and caesarean section wound infection. The indirect maternal morbidities included in this study based on the WHO maternal morbidity working group criteria were: asthma, tuberculosis, influenza, pneumonia, malaria, HIV/AIDS, candidiasis, hepatitis, hypertension, anemia and diabetes mellitus. Women who were diagnosed with any direct and/or indirect maternal morbidities were treated accordingly within the hospitals where they were diagnosed. Women diagnosed with any of the direct and indirect maternal morbidities were taken as exposed mothers and included in the study. Women without the direct and indirect maternal morbidities were taken as non-exposed mothers. All women with direct maternal morbidities were included in the study and women without direct maternal morbidities were selected by simple random sampling method using their chart number on daily bases. The chart numbers of women without direct maternal morbidities were entered into computer to generate random numbers using Microsoft Excel for random selection of women. The recruitment of women continued prospectively until the required sample size was fulfilled. Women were asked to give written consent to participate in the study and after getting their consent and full address, appointments were made at their home to collect the data for the follow up study. Comorbidity of depression, anxiety and posttraumatic stress disorder was taken as the outcome variable. Direct maternal morbidities(obstetric hemorrhage, hypertensive disorders, obstructed labour, puerperal sepsis, gestational diabetes mellitus, perineal tear), indirect maternal morbidities (anemia, malaria, hypertension, asthma, tuberculosis, HIV), socio-demographic variables (age, educational status, marital status, religion, ethnicity, occupation, monthly expenditure), residence, obstetric variables (parity, mode of delivery, gestational age at birth, birth weight, birth interval, fetal death, unwanted pregnancies, antenatal care visit, history of abortion) and psychosocial factors (social support and fear of child birth) were taken as the main independent variables. The short version of depression, anxiety and stress scale 21 (DASS-21) questionnaire was used to measure depression, anxiety and stress. DASS-21 is a psychological screening instrument which is capable of differentiating symptoms of depression, anxiety and stress. It is a validated and reliable instrument with 21 items in three domains. Each domain comprises seven items assessing symptoms of depression, anxiety and stress. Participants were asked to indicate the presence of symptoms in each domain over the past week scoring from 0 (did not apply at all). to 3 (applied most of the time). Scores from each dimension were summed. Then, the final score was multiplied by 2 and then categorized according to the DASS manual as normal, mild, moderate, severe and extremely severe. Accordingly, for participants with depression, a depression score of 0–9 is considered normal, 10–13 as mild, 14–20 as moderate, 21–27 as severe and 28 and above as extremely severe. In this study a score ≥ 10 was considered for a mother to have a symptom of depression. For participants with anxiety, an anxiety score of 0–7 is considered normal, 8–9 as mild, 10–14 as moderate, 15–19 as severe and 20 and above as extremely severe. A cut-off score of ≥ 8 was considered to have symptoms of anxiety for this study. For participants with stress, a stress score of 0–14 is considered normal, 15–18 as mild, 19–25 as moderate, 26–33 as severe and 34 and above as extremely severe. A score of ≥ 15 was considered to have symptoms of stress for this study. This instrument was validated and used previously in Ethiopia [60,61]. The childbirth stressor was operationalized by using the Traumatic Event Scale (TES) [62,63]. In this scale, the items concerning criterion A (stressor) were formulated as follows: There were four alternative answers for each statement stated as: “not at all = 1,” “somehow = 2,” “much = 3,” and “very much = 4.” If the mother’s answer is “much” or “very much” for item 1, for item 2 and item 3 or either of the two, then Criterion A is fulfilled [62,63]. The A2 criterion requiring the experience of “fear, helplessness, and horror” to qualify as a traumatic event was included in this study to examine the differences in rates of traumatic childbirth according to the DSM-IV versus DSM-V trauma criteria due to the removal of criterion A2 in DSM-V [64,65]. Following the criterion A questions, we have used the Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5) comprising the 20 PTSD symptoms (criterion B, C, D and E) to measure PTSD over the past month. The instrument contains 20 items, including three new PTSD symptoms (compared with the PTSD Checklist for DSM-V): blame, negative emotions and reckless or self-destructive behavior [66]. A total-symptom score of zero to 80 can be obtained by summing the items. A score of 31–33 is optimal to determine PTSD symptoms and a score of ≥33 is recommended when further psychometric testing is not available [67,68]. Therefore, a score of ≥ 33 was considered to have symptoms of PTSD for this study. The cut-offs for the instrument were validated by a previous study in Ethiopia [68]. The Wijma Delivery Expectation/Experience Questionnaire (W-DEQ) was used to measure fear of child birth. The W-DEQ has been designed specially to measure fear of child birth operationalized by the cognitive appraisal of the delivery. This 33-item rating scale has a 6-point Likert scale as a response format, ranging from ’ not at all’ (= 0) to ’ extremely’ (= 5), yielding a score-range between 0 and 165. The Internal consistency and split-half reliability of the W-DEQ was checked in previous studies in Ethiopia with the Cronbach’s alpha score of 0.932 [69,70]. A score of ≥ 85 was considered to have fear of child birth for this study [69,70]. The Oslo 3-items social support scale with scores ranging from 3 to 14 was used to measure social support. The social support scores were categorized into poor or no social support for scores less than nine. Scores between 9 and 14 were considered moderate to strong support and merged together as “yes” for social support. The Oslo 3-items social support scale was validated and previously used in Ethiopia [71–73]. Administering baseline questionnaire and diagnosis of direct and indirect maternal morbidities based on the WHO criteria, were done by health professionals (BSC Nurses and Midwives) working in the Gynecology and Obstetrics wards of the study Hospitals. The questionnaire included a patient interview and record review. The interview was on socioeconomic status, medical and obstetric history and clinical symptoms. The record review was intended to extract information on selected laboratory tests and results for hemoglobin, HIV, malaria (rapid diagnostic test or smear) and glucometer (random blood sugar). The follow up data on DASS-21 and PCL-5 were administered by health extension workers (community health workers) at the first, second and third home visit (6th, 12th and 18th week of postpartum period). Two different groups of data collectors were used for data collection to avoid bias. The first group were health professionals with BSC degree in Nursing and Midwifery who collected the baseline data and the direct and indirect maternal morbidities based on the WHO maternal morbidity working group criteria. The second group were health extension workers (community health workers) who collected the follow up data by house-to-house visit (home visit). Supervision was done by the principal investigator. Training was given for data collectors in scale administration. During the training process, data collectors carefully reviewed each question and conduct pretest before the study commenced. The investigator and data collectors checked the questionnaire during the pretest and amendment was done as required. We used a three-wave, cross-lagged autoregressive structural equation modeling. The analysis was carried out by using Stata version 16 software. The Autoregressive Cross-lagged (ARCL) modeling strategy was used to examine the longitudinal association of depression, anxiety and PTSD with their comorbidities including all other scales used for this study. This modeling strategy incorporates three main components. First, the stability/autoregressive effects (e.g., effects of T1 comorbid depressive, anxiety and PTSD symptoms on their respective T2 variables). That means, later measures of a construct are predicted by earlier measures of the same construct. Second, the cross-lagged effects (e.g., effect of T1 depressive symptoms on T2 comorbid PTSD and anxiety symptoms and of T1 comorbid PTSD and anxiety symptoms on T2 depressive symptoms). That means, earlier measures of depression predict later measures of comorbid PTSD and anxiety symptoms. This model can be extended to examine bi-directional relations such that earlier measures of PTSD predict later measures of comorbid depression and anxiety as well. Third, the synchronous associations between the unexplained variances of these variables at T1, T2 and T3 [74,75]. We estimated the model fitness by using the comparative fit index (CFI), Tucker-Lewis Index (TLI) and the root-mean-square error of approximation (RMSEA) based on the Satorra-Bentler correction. Both the TLI and CFI should be greater than 0.90, but the RMSEA value should be less than 0.08 to judge the model as reasonably fitting the data [21,74]. The aim of the analysis was to examine the cross-lagged and autoregressive association of depressive, anxiety and PTSD with their comorbidities, controlling for the confounder variables. As this study involved repeated measurements, we have carried out test of temporal invariance (homogeneity) for depression, anxiety, PTSD and their comorbidities across the three data points in time [76]. Homogeneity is to mean that statistical properties of any one part of an overall dataset are the same as any other part. Therefore, homogeneity of paths was conducted to test the regression coefficients for each observed variable is stable over time. It can be tested by comparing if the impact of the variable (regression coefficient) at the time of [t-1] on the variable at the time of [t] is the same with the impact of the variable (regression coefficient) at the time of [t] on the variable at the time of [t+1]. The regression coefficient can be labeled as stability coefficient [77,78]. Accordingly, Model 1 served as the base model with no invariance constraints (freely estimated factor loadings, autoregressive paths, cross lagged paths and residual covariance at a different time point) to test configural invariance. In Model 2, we constrained the auto-regressive and cross-lagged paths for depression, anxiety, PTSD and their comorbidities across time to be equal. We then compared the model fit indices of the two models to select the best model. Test of difference in CFI, TLI and RMSEA (ΔCFI, ΔTLI, ΔRMSEA) were used to compare the fit of the nested models. If a difference of less than 0.01 is found in the ΔCFI, ΔTLI and ΔRMSEA index between nested models, it is indicated that the more constrained model should be supported [79–81]. In addition, the direct and indirect relationships between the independent and dependent variables was also explored using a linear structural equation modeling. This allowed us to assess the strength of the hypothesized direct and indirect causal pathways. Estimated effects for which p <0.05 were considered as being statistically significant. The direct effect is the pathway from the exogenous (independent) variable to the outcome (endogenous) variable while controlling for the mediator. Therefore, in the indicated path diagram of Fig 1 bellow, y and z are the direct effects of anxiety and depression on PTSD respectively [82]. The indirect effect describes the pathway from the exogenous (independent) variable to the outcome (endogenous) variable through the mediator. This path is represented through the product of x and y in the path diagram of Fig 1 bellow [82]. This indicates the indirect effect of depression on PTSD through anxiety which is the mediator between depression and PTSD. Finally, the total effect is the sum of the direct and indirect effects of the exogenous (independent) variables on the outcome (endogenous) variable, Z + XY [82]. Ethical approval was obtained from Institutional Review Board of Bahir Dar University (Reference number: 00225/2020). All institutions of the study area were also approached for permission. Each study participant has given written informed consent to participate in the study before taking part after giving birth and before discharge. Written informed consent for women of less than 18 years was obtained from their guardians or care givers. In addition to written informed consent from their guardians or care givers, assent was also obtained from teenage mothers whose age is less than 18 years, to participate in the study. The written informed consent and assent were taken after clear explanation about the purpose of the study. The study participants and guardians or care givers of women whose age is less than 18 years were informed that they have full right to withdraw from the study without losing any of their right as a client in the health institutions. Data collection (scale administration) at hospitals took place in a private room of delivery wards and the follow up data collection at household were also carried out in a private place within the mothers’ living compound. Using codes, passwords and limiting access to the data only for the investigators were the measures taken to ensure the confidentiality of the data. Data collectors read out and assisted participants to fill out the consent form if participants were unable to read and write. This was witnessed by a neutral observer to prevent coercion.

• malaju, marelign tilahun
• alene, getu degu
• azale, telake