Interpersonal Psychotherapy to Reduce Psychological Distress in Perinatal Women: A Systematic Review

Background: Interpersonal psychotherapy (IPT) is a psychological intervention with established efficacy in the prevention and treatment of depressive disorders. Previous systematic reviews have not evaluated the effectiveness of IPT on symptoms of stress, anxiety, depression, quality of life, relationship satisfaction/quality, social supports, and an improved psychological sense of wellbeing. There is limited information regarding moderating and mediating factors that impact the effectiveness of IPT such as the timing of the intervention or the mode of delivery of IPT intervention. The overall objective of this systematic review was to evaluate the effectiveness of IPT interventions to treat perinatal (from pregnancy up to 12 months postpartum) psychological distress. Methods: MEDLINE(R) and Epub Ahead of Print, In-Process & Other Non-Indexed Citations and Daily (Ovid), EMBASE (Ovid), PsycINFO (Ovid), Cochrane Central Register of Controlled Trials (OVID), CINAHL with Full Text (Ebsco), Social Work Abstracts (Ebsco), SocINDEX with Full Text (Ebsco), Academic Search Complete (Ebsco), Family & Society Studies Worldwide (Ebsco), Family Studies Abstracts (Ebsco), and Scopus databases were searched from inception until 31 January 2019. Two researchers independently screened articles for eligibility. Of the 685 screened articles, 43 met the inclusion criteria. The search was re-run on 11 May 2020. An additional 204 articles were screened and two met the inclusion criteria, resulting in a total of 45 studies included in this review. There were 25 Randomized Controlled Trials, 10 Quasi-experimental studies, eight Open Trials, and two Single Case Studies. All included studies were critically appraised for quality. Results: In most studies (n = 24, 53%), the IPT intervention was delivered individually; in 17 (38%) studies IPT was delivered in a group setting and two (4%) studies delivered the intervention as a combination of group and individual IPT. Most interventions were initiated during pregnancy (n = 27, 60%), with the remaining 18 (40%) studies initiating interventions during the postpartum period. Limitations: This review included only English-language articles and peer-reviewed literature. It excluded government reports, dissertations, conference papers, and reviews. This limited the access to grassroots or community-based recruitment and retention strategies that may have been used to target smaller or marginalized groups of perinatal women. Conclusions: IPT is an effective intervention for the prevention and treatment of psychological distress in women during their pregnancy and postpartum period. As a treatment intervention, IPT is effective in significantly reducing symptoms of depression and anxiety as well as improving social support, relationship quality/satisfaction, and adjustment. Systematic Review Registration: PROSPERO CRD42019114292.


Screening of Studies
Prior to screening, the two reviewers (KSB and EMC) completed a calibration exercise where 10% of studies were reviewed independently and then together assessed for inter-rater agreement. In the calibration exercise, there was 93% agreement. Following the calibration exercise, the two reviewers independently screened the studies for eligibility in two steps. The first step consisted of reviewing all studies' titles/abstracts to identify studies that met the eligibility criteria. The second step consisted of reviewing the provisionally included studies' full text to ensure that they met all the inclusion criteria. Any disagreements were resolved by discussion between the two reviewers. There were 45 studies that met the inclusion criteria ( Figure 1. PRISMA Diagram).

Risk of Bias in Individual Studies
Studies were included regardless of methodological quality. The Effective Public Health Practice Project (EPHPP) Quality Assessment Tool was used for quality assessment. Two reviewers (KSB and

Risk of Bias in Individual Studies
Studies were included regardless of methodological quality. The Effective Public Health Practice Project (EPHPP) Quality Assessment Tool was used for quality assessment. Two reviewers (KSB and EMC) independently assessed all studies for quality and disagreements were resolved by discussion between the two reviewers.
Characteristics of the interventions are presented in Table 2. In most studies (n = 24, 53%), the IPT intervention was delivered individually; in 17 (38%) studies IPT was delivered in a group setting, two (4%) studies delivered the intervention as a combination of group and individual IPT, and two (4%) studies included partners in the delivery of the intervention. Most studies (n = 29, 64.4%) delivered the IPT face-to-face, while two (4.4%) studies delivered IPT over the phone and 14 (31.1%) studies combined face-to-face and telephone calls.
Most interventions were initiated during pregnancy (n = 27, 60%), with the remaining 18 (40%) studies initiated during the postpartum period. IPT was administered individually in 24 (53%) studies and in groups in 17 (38%) studies. Women's partners were included in the intervention in two (4%) studies. Most studies (n = 30, 66.7%) provided IPT in a community setting (e.g., women's recreation facility), 12 (26.7%) studies provided IPT in the clinical setting (e.g., prenatal clinic), and three (6.7%) studies provided IPT in a mixed clinical and community setting. The number of IPT sessions ranged from two to 16 sessions, with an average of eight sessions. Most studies (n = 35, 78%) reported provided IPT according to a study or intervention protocol.
Characteristics of the method of assessment for outcomes are presented in Table 3. In most studies (n = 28, 62.2%), depressive symptoms were assessed using the Edinburgh Postnatal Depression Scale (EPDS), while 16 (35.6%) studies used the Hamilton Depression Rating Scale (HAM-D), 16 (35.6%) used the Beck Depression Inventory (BDI), three (6.7%) studies used the CESD, and three (6.7%) studies used the SCL-20. Symptoms of anxiety were assessed in 18 (40%) studies, most commonly using the State-Trait Anxiety Inventory and Beck Anxiety Inventory. Stress levels were assessed in 10 (22%) of the studies. Maternal-infant attachment was assessed in 16 (36%) of the studies. Eleven (24%) of the studies assessed social support. Relationship satisfaction/quality was assessed in 17 (38%) of the studies.
Characteristics of study methodological quality are presented in Table 4. Methodological quality was assessed using the Effective Public Health Practice Project (EPHPP) Quality Assessment Tool [86]. The study scores ranged from 1 (strong) to 3 (weak), with an average of 2 (moderate). There were 18 studies (40%) categorized as strong overall, 14 (31%) studies were moderate overall, and 13 (29%) studies were weak overall. Study design was assessed as strong in 26 (57.8%) studies, intervention integrity was determined to be strong in 35 (78%) studies, and data analysis was assessed as strong in 20 (44%) studies.                    Among the studies that reported sample demographic characteristics, maternal age ranged from 18 to 38 years old with a mean age of 30 years. The average gestational age for pregnant women ranged from six to 40 weeks, with an average of 23.7 weeks. The weeks postpartum of participants ranged from 0.5 to 96 weeks postpartum, with an average of 24.4 weeks.

Prevention Studies
Among the 13 prevention studies, 12 (92%) were delivered during pregnancy and one (8%) was delivered in the postpartum period.

Treatment Studies
Among the 33 treatment studies, 16 (48.5%) were delivered during the prenatal period and 17 (51.5%) studies were delivered in the postpartum period.

Change in Depressive Symptoms Between Treatment and Comparison Groups
Twelve prevention studies aimed to reduce the risk of depression in participants receiving IPT. Five studies [45,55,56,59,67] reported a significant reduction of depressive symptoms levels over time. These improvements were small to moderate in magnitude. No studies had large effect sizes. Reductions in depressive symptoms were also significantly larger in studies where IPT was delivered in a group format compared to individual IPT.
Thirty-two (71%) treatment studies assessed change in depressive symptoms among participants receiving IPT. Twenty-six studies reported a significant improvement in depressive symptoms over time. The improvements were determined to be in the moderate to large range. Reductions in depressive symptoms were more common in studies where the interventions were initiated in the postpartum period than in studies where interventions were initiated during pregnancy.

Change in Anxiety Symptoms Between Treatment and Comparison Groups
Seven prevention studies aiming to reduce the risk of symptom levels of anxiety addressed the change in symptoms of anxiety among participants receiving IPT. One study (Bowen et al., 2014) reported a significant reduction in the risk level of anxiety symptoms. The effect size of the intervention on symptoms of anxiety was not reported in this study.
Eleven treatment studies assessed the change in symptoms of anxiety among participants receiving IPT. Six studies [47,52,53,60,62,69] reported significant reductions in symptoms of anxiety. There was an overall reduction in symptoms of anxiety among participants receiving IPT, with an effect size in the moderate range. More studies of individual delivery showed a reduction in anxiety than group delivery. Reductions in anxiety were also noted more frequently in studies where IPT was delivered in a medical/clinical setting compared to a community setting.

Change in Stress Symptoms Between Treatment and Comparison Groups
Three prevention studies aimed at reducing the risk of stress levels assessed change in symptoms of stress among participants receiving IPT. Two studies (Bowen et al., 2014;Leung & Lam, 2012) reported a significant reduction in the risk of symptom levels of stress. One study did not report an effect size of the intervention and the other reported a very small effect size (Leung & Lam, 2012).
Seven treatment studies assessed for change in symptoms of stress among women receiving IPT. Two of these studies (Field et al., 2009;Field et al., 2013) reported a significant reduction in symptoms of stress for participants receiving IPT. The effect sizes of the intervention were not reported.

Change in Relationship Quality Between Treatment and Comparison Groups
Three prevention studies aiming to reduce the risk of relationship distress assessed relationship quality/satisfaction among participants receiving IPT. There were no studies that reported a significant improvement in relationship quality/satisfaction.
Twelve treatment studies assessed relationship quality/satisfaction among women receiving IPT. Four studies (Chung, 2015;Field et al., 2013;Hajiheidari et al., 2013;Mulcahy et al., 2010) reported significant improvements in relationship quality, with an effect size in the small range. Studies with married/cohabitating participants were more likely to have greater improvements in their relationship quality than those women without partners. improvements in social support. The effect size was in the medium to large range. Studies with participants who had higher levels of education were more likely to experience significant improvements in social support.

Change in Attachment Levels Between Treatment and Comparison Groups
There were no prevention studies that assessed attachment. There were eight treatment studies that assessed attachment among participants receiving IPT. Three of these studies (Mulcahy et al., 2010;Posmontier et al., 2019;Spinelli, Endicott, Leon, et al., 2013) reported significant improvements in attachment. While these improvements were reported to be statistically significant, the effect size of the IPT intervention was not reported.

Change in the Level of Adjustment Between Treatment and Comparison Groups
There was one prevention study aiming to reduce the risk of poor adjustment that assessed for adjustment among participants receiving IPT. This one study (Crockett et al., 2008) reported that the level of adjustment was statistically significant only between 2-3 weeks and 3 months postpartum. No effect size was reported.
There were 12 treatment studies that assessed for level of adjustment among participants receiving IPT. There were no studies that reported any significant improvements in level of adjustment.

Discussion
This review of the literature provides evidence that IPT is an effective intervention for the prevention and treatment of psychological distress in women during their pregnancy and postpartum period. As a preventive intervention, IPT is superior to comparison conditions, including active interventions, treatment-as-usual, and no intervention, for reducing the risk of depression. As a treatment intervention, IPT is effective in significantly reducing symptoms of depression and anxiety as well as improving social support, relationship quality/satisfaction, and adjustment. IPT is superior to comparison conditions including active interventions, treatment-as-usual, and no intervention for reducing depressive symptoms as well as improving social support and relationship quality.
There is evidence supporting the use of IPT to prevent depression in perinatal women. These findings suggest that IPT is effective as both a prevention intervention and for those women at high risk due to the presence of risk factors including a previous diagnosis of depression   There was one preventive study that reported outcomes for symptoms of anxiety (Bowen et al., 2014). This study found that IPT was effective in reducing anxiety symptoms and worry over time in pregnant women compared to active interventions, treatment-as-usual, and no interventions. Given the far reaching impact of prenatal anxiety on women and their children (Brunton, et al, 2015 [87]; Mughal et al., 2019 [88]; Brunton, Dryer, Field, 2017 [89]; K. Bright & Becker, 2019 [90]), future research exploring preventive interventions in prenatal women would benefit from including assessment of anxiety in addition to depressive symptoms. There is a need for investigating the diagnostic outcomes of anxiety and anxiety-related disorders, including the prevalence of perfectionism and obsessive-compulsive disorder, as preliminary work in this area suggests that there is increased risk for these disorders during the perinatal period ( In this review, group prevention interventions resulted in greater reduction in risk of symptom levels of depressive than individually administered interventions. Groups have a valuable set of therapeutic characteristics where women are provided with a supportive network of peers with shared feelings, thoughts, and problems (Marmarosh, Holtz, & Schottenbauer, 2005) [95]. Women gain insight into the universality of their problems, which helps to normalize their experiences (Reay et al., 2006). Group therapy allows women to increase their coping strategies, knowledge, and skill through vicarious learning. Helping others solve their problems can increase their sense of competence. It may also be that the social skills and competencies gained through group-based IPT prevent the onset of depressive symptoms by specifically moderating relationship challenges.
While RCTs of IPT for mental health disorders show a moderate to large effect on depression compared with control groups, IPT has not been found to be more effective than other psychotherapies such as CBT for depression ( [98]). When pharmacotherapy is combined with psychotherapy, it is not more effective than pharmacotherapy alone, but is more effective than IPT alone (Cuijpers et  There was a trend that more studies of individually administered IPT showed a reduction of anxiety symptoms than group offered IPT. Individual therapy has the advantage of participants receiving greater attention to their individual issues, closer monitoring of symptoms, and more tailored adaptation of the intervention to issues that are particularly  [97,[101][102][103]. Future preventive and treatment research would benefit from including assessment of acceptability of group and individual therapy. Investigation of potential predictors of treatment efficacy should include a history of depressive disorders and anxiety-related disorders as well as their comorbidity to determine if these characteristics are associated with delivery method and differential efficacy. In six RCTs examining the effect of IPT on anxiety, compared to other psychotherapies, this resulted in a small nonsignificant difference in favour of the alternative therapies such as CBT over IPT (Cuijpers et al., 2016;Nillni et al., 2018) [96,99]. There is one study investigating the effect of paroxetine and CBT compared to CBT alone and it was found that there was no significant difference between groups (Misri, Reebye, Corral, & Mills, 2004) [104]. Given the paucity of research in this area, this is concerning given that anxiety symptoms and comorbid symptoms are prevalent in perinatal women, therefore it is important that there is further research on effective treatments.

Strengths
There are numerous strengths of this systematic review, which include explicit methods description and comprehensive database searches to methodologically search for articles exploring the use of IPT/IPT-based interventions in the perinatal population. This transparent and systematic approach to reviewing the literature included the use of a librarian for the search and two reviewers with content expertise for the assessment of inclusion and data extraction attempted to reduce reviewer bias. This rigorous process facilitates a reproducible and objective criteria to select relevant studies and adequately assess their quality.

Limitations
A major limitation of the studies evaluating IPT, whether for prevention or treatment, is that few studies addressed outcomes such as social support, relationships, and adjustment the same way. Improving these interpersonal areas are among the goals of IPT. As such, there needs to be consistency in how these elements are operationalized in a perinatal population. Implications for future IPT intervention studies involve assessing perinatal women's change in interpersonal functioning and involving women's partners in treatment.
Findings from this review of IPT in perinatal women are limited to IPT being delivered face-to-face or via telephone. Literature examining online IPT in non-perinatal populations suggests that despite high dropout rates, internet-delivered self-guided IPT is effective in reducing depressive symptoms (Donker et al., 2013). Future research requires well-designed RCTs that compare internet-delivered IPT to active, treatment-as-usual, and no treatment. Additionally, internet-based IPT trials will need to assess differences in prevention versus treatment, prenatal versus postpartum women, and group versus individual treatment.
This review is limited by the lack of detailed descriptions of recruitment and retention strategies of the individual studies. Further limitations include the inclusion/exclusion criteria of reviewing only English-language articles, which may reduce generalizability to non-English speaking populations. Similarly, this review included only peer-reviewed literature and excluded government reports, dissertations, conference papers, and reviews. This limited the access to grassroots or community-based recruitment and retention strategies that may have been used to target smaller or marginalized groups of perinatal women.

Research Implications
Further studies would benefit from refinement of the perinatal IPT treatment. In future studies, the IPT intervention will need to include a comprehensive IPT manual to promote adherence/competence measures. Perinatal IPT research will also benefit from development of far-reaching training programs for those delivering IPT in research, community, and clinical settings. Improving the structure of IPT and training of clinicians who can deliver evidence-based IPT has the potential to improve outcomes for perinatal women.
Additional research is required to evaluate the efficacy of internet-based treatment compared to telephone and face-to-face delivery. Regardless of the type or mode of delivery, research aimed at exploring the mechanisms of action is necessary for IPT interventions. This will aid in further refining IPT interventions, improving outcomes, and determining whether the intervention is applicable in additional settings.
Studies exploring various techniques for keeping women engaged in treatment for extended periods of time are warranted to ensure that perinatal women can complete the full IPT intervention. This will take into consideration an individual's preference for treatment. Longitudinal studies of different intervention models (varying in length and delivery) and social support are needed. More research into how IPT interventions can be implemented as a part of routine prenatal care is needed.

Clinical Implications
There is a large body of research that demonstrates the effectiveness of treatments for depression and anxiety during the perinatal period (Milgrom, [99,101,102,105]. Given that there is strong evidence for and no difference in the effectiveness for prevention and treatment of various psychotherapies allows for women to determine which psychotherapy they would choose. This choice may also be influenced by the mental health services offered through the trained therapists in their area. Additionally, the decision on whether to use pharmacotherapy in addition to psychotherapy during the perinatal period is complex and requires the consideration of many factors, including the effects of untreated maternal mood and/or medication exposure on both maternal and fetal outcomes. Clinical discussion making around mental health treatment options would benefit from thoughtful conversations between clinicians and the perinatal women as well as their families as no one treatment works for everyone.

Conclusions
This systematic review provides evidence that IPT is an effective intervention for the prevention and treatment of psychological distress in women during their pregnancy and postpartum period. This review also highlights the need for robust, high quality RCTs exploring different intervention models for women during the perinatal period.

Conflicts of Interest:
The authors declare no conflict of interest.