Depressive symptoms in mothers of preterm infants before and during COVID‐19 restrictions in neonatal intensive care units

Little is known about the specific restriction measures used in intensive care units (NICUs) during the COVID‐19 pandemic and their impact on parental well‐being. Hence, this study aimed to assess the association between restriction measures and mothers' post‐partum depressive symptoms.

presence and participation in the caretaking of preterm infants have been shown to compromise the mothers' psychological wellbeing. 6,7 A meta-analysis showed that the odds of post-partum depression in mothers was 1.79 times higher after preterm birth than full-term birth. 8 Today, many NICUs provide family-centred care, allowing parents to access NICUs 24/7 and supporting parents' participation in the care of their infant. Family-centred interventions have been shown to be associated with decreased post-partum depressive symptoms in mothers of preterm infants. For example, the Close Collaboration with Parents intervention was associated with the reduction of depressive symptoms, as evaluated using the Edinburgh Postnatal Depression Scale (EPDS). The reduction observed was 42% at 4-6 months of corrected age 9 and 37% at 24 months of corrected age. 10 The Family Nurture Intervention reduced the proportion of mothers with possible clinical depression at 4 months of corrected age by 15%. 11 Family-centred care is considered a state-of-the-art approach in neonatal care because of its demonstrated benefits not only for mothers but also for infants. 12 However, COVID-19 caused a worldwide reduction in parents' access to NICUs and participation in the care of their infants. When the pandemic started in March 2020, most NICUs were forced to set up restrictions in accordance with the governmental and hospital policies. 13,14 Many of these measures limited parental presence and involvement in the caretaking of their infants. Few previous studies have investigated whether the COVID-19 pandemic was associated with the increase in post-partum depressive symptoms in mothers in NICUs. Three studies in Belgium, Switzerland and Italy found a slightly higher mean score for depression, assessed using the EPDS, during the pandemic compared to before. However, none of the studies showed a statistically significant difference. [15][16][17] Only the Belgian study revealed a significantly higher proportion of mothers exceeding the cut-off for possible clinical depression during the pandemic than before. 15 In addition, all three were single-centre studies and did not consider potential confounders in their analyses.
Furthermore, none of them focused on the effect of the specific restriction measures implemented in NICUs.
Thus, the effect of the COVID-19 pandemic on the post-partum depressive symptoms of mothers remains unclear. In this study, we evaluated the association between the COVID-19 pandemic and post-partum depressive symptoms in mothers after preterm birth.
In addition, we evaluated the association between the use of NICU restriction measures and depressive symptoms.

| Study design and protocols
This comparative cohort study used two preterm cohorts selected from Estonia before and during the COVID-19 pandemic. The pre-pandemic cohort was recruited between January 2018 and May 2019 for the 2nd International Closeness Survey. 18  These NICUs were located in four hospitals in the two largest cities in Estonia. Out of the seven NICUs, two were not included in both cohorts. One of the Level II NICUs was only included in the pre-pandemic cohort, while another one was only included in the pandemic cohort. These two NICUs were comparable, as they both operated in the same urban region and had similar sizes, patient structures and levels of neonatal care.
Ethical approval was obtained separately for each study.
The study permission of the 2nd International Closeness Survey

| Participants
The inclusion and exclusion criteria for both studies were similar but not identical. In the pre-pandemic cohort recruitment, mothers who gave birth before 35 weeks of gestation were eligible. Families were excluded if the expected duration of hospitalisation was less than 3 days, the infants were triplets, the parents could not speak fluent • The use of hospital visiting policies and their impact on maternal well-being should be studied more widely. regardless of the gestational age. Of those, mothers who gave birth before 35 weeks of gestation were selected for this study. Families were excluded if the expected duration of hospitalisation was less than 3 days or if the mother could not speak fluent Estonian, Russian or English. Of 91 eligible mothers, 59 were enrolled in the study ( Figure 2). The mothers who gave birth before 35 weeks of gestation were selected from the cohorts for this study. Hereafter, the group selected before the pandemic is called the pre-COVID group and during the pandemic-the COVID group.

| Measures
Depressive symptoms were measured using the EPDS when the infant was discharged from the NICU. The EPDS has proven itself a valid method to measure postnatal depressive symptoms. 19 It contains 10 questions, and each item is scored from 0 to 3. The total score ranges from 0 to 30; a higher score indicates more severe depressive symptoms. The mean total score of EPDS was used as the primary outcome of this study. Additionally, as a secondary outcome, the incidence of possible clinical depression among mothers was estimated using the most commonly used total EPDS cut-off score of >12, 19,20 and compared between the groups. Those parents who exceeded the EPDS clinical cut-off score of 12 were contacted and offered support. The Cronbach's alpha was 0.88 in the Pre-COVID group and 0.80 in the COVID group.
These outcomes were also compared between infants exposed to a higher and lower number of restriction measures in NICUs. Each NICU recorded the details of its restriction measures and changes in them throughout the data collection period, which is summarised in  Figure 2). They were classified into the high-restriction group if, for more than 50% of the infant's admission period, they were exposed to NICUs with two or more restriction measures. All others were classified into the mild-restriction group. Those in the Pre-COVID group were classified into the baseline group, although some NICUs had certain restriction measures even before the COVID-19 pandemic.
Background information was collected from medical records or from parents using a questionnaire. Twins were treated differently in the two studies. The infant born first was chosen for the Pre-COVID group. In the COVID group, the twin with the longer length of stay was chosen, or the parents chose one if the twins were discharged home on the same day.

| Statistics
Comparisons were made between the Pre-COVID and the COVID groups. The same comparisons were also made between the high-, mild-restriction and baseline groups.
The difference in the EPDS score was analysed using Wilcoxon's rank-sum test and the Kruskal-Wallis test. The difference in the F I G U R E 1 The number of restriction measures and their changes over time in each NICU (A1-D7). The same alphabet indicates the NICU in the same hospital. The NICU B4 was included only during the COVID-19 pandemic, and the NICU D7 was included before COVID-19, which were comparable as noted in the Methods section. The included restriction measures were targeted to: (1) mothers' overnight stay, (2) mothers' visit to NICU, (3) fathers' overnight stay, (4) fathers' visit to NICU and (5) both parents at the same time in NICU. If an item was permitted with some limitations, it was counted as 0.5 (e.g. parents needed a COVID-19-negative certificate to visit their infant in NICUs). NICU, neonatal intensive care unit.

| RE SULTS
A total of 109 mothers were eligible for the analyses: 54 in the Pre-COVID group and 55 in the COVID group (Figure 2). From the data of all eligible mothers, the median age was 32 years (interquartile range [IQR]: 28-36). Additionally, 3% of them were single parents, and 49% had higher education. These characteristics were comparable to those of the Estonian population, as summarised by the National Institute for Health Development in Estonia. The background characteristics of mothers and their infants in both groups were comparable, except for the earlier gestational age and smaller weight at birth in the COVID group than in the Pre-COVID group (Table 1). Table 2 summarises the results of the analyses comparing the total EPDS scores assessed at infants' discharge between the COVID and the Pre-COVID groups. The median EPDS score was lower in the COVID group than in the Pre-COVID group, but the difference was not statistically significant between the groups. The adjusted mean total EPDS score showed no significant difference.  More mothers were first-time mothers, fewer mothers had higher education and fewer infants were boys in the high-restriction group than in the other two groups. Infants in the mild-restriction group were born earlier with smaller birth weights than those in the other two groups (Table 1). Table 3 summarises the results of the analyses comparing the total EPDS scores assessed at infants' discharge between three re-  post-partum depressive symptoms in NICUs. [15][16][17] All were singlecentre studies, whereas this study used six NICUs for each data collection phase, which covered most preterm births in Estonia.
A study from Belgium found significantly higher adjusted odds of possible clinical depression during the COVID-19 pandemic than before. 15 The Belgian study included mothers of infants born before 32 weeks of gestation, whereas this study included mothers of infants born before 35 weeks of gestation. Furthermore, the Belgian study collected its data during the first lockdown and at 3-6 months after discharge. This may have impacted mothers' social isolation and loneliness, which are known risk factors for post-partum depression. 25 On the other hand, the results of the other two studies were in line with our findings in showing no significant associations. Our results were also consistent with those of a meta-analysis of mothers of full-term infants, which found no significant difference in total EPDS scores between those during the COVID-19 pandemic versus before. 26 This study was the first, to the best of our knowledge, to docu-   Abbreviations: 95% CI, 95% confidence interval; Baseline, Baseline group; EPDS, Edinburgh Postnatal Depression Scale; High, High-restriction group; IQR, interquartile range; MD, mean difference; Mdn, median; Mild, Mild-restriction group; OR, odds ratio.

| CON CLUS ION
This study showed no significant association between COVID-19 restriction measures and post-partum depressive symptoms in mothers of preterm infants in NICUs. The family-centred care culture, the design of single-family rooms and the support provided by healthcare staff may have protected mothers from an increase in depressive symptoms during the COVID-19 pandemic in Estonian NICUs.
Further studies are warranted to understand the negative impact of NICU restriction measures and visiting limitations on maternal wellbeing. Even more importantly, studies are needed to understand how to best support parents' resilience in NICUs during times when certain restrictions are necessary.

ACK N O WLE D G E M ENTS
We thank the doctors and nurses who made significant contributions to patient recruitment and data collection in Estonia, and the participating families.

FU N D I N G I N FO R M ATI O N
The study was funded by the Academy of Finland (no: 332962) and the Pediatric Research Foundation in Finland (no: 210169).

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors have no conflicts of interest to declare.