Pregnancy and the transition to motherhood are life-changing experiences for most women, and are accompanied by both physical and psychological changes and challenges that make these women vulnerable to mental health problems.Reference Raphael-Leff1,Reference Slade, Cohen, Sadler, Miller and Zeanah2 Correspondingly, depressive symptoms during pregnancy are quite prevalent, with up to 20% of women experiencing antepartum depressive symptoms (ADS).Reference Gavin, Gaynes, Lohr, Meltzer-Brody, Gartlehner and Swinson3–Reference Underwood, Waldie, D'Souza, Peterson and Morton5 Depressive symptoms impair maternal social and physical functioning, and are related to increased maternal distress and poor maternal quality of life.Reference Abbaszadeh, Kafaei Atrian, Masoudi Alavi, Bagheri, Sadat and Karimian6–Reference Da Costa, Dritsa, Verreault, Balaa, Kudzman and Khalifé8 Moreover, if left untreated, ADS can develop into peripartum major depressive disorder, affecting 5–13.5% of women during pregnancy and 7–13.1% in the postpartum period.Reference Woody, Ferrari, Siskind, Whiteford and Harris4,Reference Underwood, Waldie, D'Souza, Peterson and Morton5,Reference Gaynes, Gavin, Meltzer-Brody, Lohr, Swinson and Gartlehner9 Furthermore, ADS have been shown to be associated with harmful maternal health behaviours and pregnancy complications.Reference Zuckerman, Amaro, Bauchner and Cabral10–Reference Lindahl, Pearson and Colpe12 Beyond these adverse effects for the woman herself, ADS also affect the health and development of the unborn infant, as maternal ADS are related to preterm birth and low neonatal birth weight.Reference Jarde, Morais, Kingston, Giallo, MacQueen and Giglia13–Reference Fekadu Dadi, Miller and Mwanri17 Additionally, offspring exposed to maternal depressive symptoms in utero show an increased risk of cognitive, developmental and mental health problems in childhood, adolescence and adulthood.Reference Gentile18,Reference Talge, Neal and Glover19 This illustrates that maternal ADS are a serious public health problem, and that early intervention is crucial.
Interventions for antepartum depressive symptoms
Accordingly, research into the evaluation of the effectiveness of prevention and treatment programmes for antepartum depression has increased. However, evidence regarding the effectiveness of (preventative) interventions for ADS is inconsistent. For major depression,Reference van Ravesteyn, Lambregtse-van den Berg, Hoogendijk and Kamperman20 a meta-analysis found that cognitive–behavioural therapy (CBT) and interpersonal psychotherapy (IPT) are effective treatments. Body-oriented interventions, such as yoga, also seem promising for the treatment of major depression, but because of the limited sample sizes and methodological quality of the included studies, evidence is not conclusive.Reference van Ravesteyn, Lambregtse-van den Berg, Hoogendijk and Kamperman20 Two meta-analysesReference Sockol21,Reference Claridge22 and a systematic reviewReference Nillni, Mehralizade, Mayer and Milanovic23 that included ADS ranging from women at risk for depression to those with major depressive disorder found that, overall, CBT and IPT are effective interventions. However, effect sizes were larger in more clinical study populations,Reference Sockol21–Reference Nillni, Mehralizade, Mayer and Milanovic23 whereas trials directed at non-clinical populations evaluating the effectiveness of IPT and CBT during pregnancy regularly suffered from low adherence and high attrition rates.Reference Spinelli, Endicott, Leon, Goetz, Kalish and Brustman24–Reference Bittner, Peukert, Zimmermann, Junge-Hoffmeister, Parker and Stöbel-Richter27 Therefore, it could be argued that these treatment approaches might not be entirely suitable for women with mild-to-moderate ADS. Possibly, the nature of these interventions might be too intensive, time-consuming or be experienced as stigmatising.Reference Goodman28–Reference Da Costa, Zelkowitz, Nguyen and Deville-Stoetzel31 Moreover, (exposure-based) CBT during pregnancy has previously been criticised for its potential adverse (neuro-endocrine) effects.Reference Arch, Dimidjian and Chessick32 Hence, in a recent randomised controlled trial, an increase of depressive symptoms in pregnant women receiving CBT and a negative effect of CBT on gestational age at birth was found in anxious women, suggesting CBT to be associated with increased hormonal stress reactions resulting from the confrontational elements of CBT.Reference Arch, Dimidjian and Chessick32,Reference Burger, Verbeek, Aris-Meijer, Beijers, Mol and Hollon33 Accordingly, Lever Taylor et alReference Lever Taylor, Cavanagh and Strauss34 proposed that pregnant women with depressive symptoms might benefit more from interventions focusing on the promotion of mental well-being.
Resilience and antepartum depressive symptoms
This recommendation is consistent with the increasing emphasis on the prevention of mental disorders and the promotion of mental well-being both in general and maternal mental healthcare.35,36 One of the key concepts known to be associated with mental well-being is psychological resilience,Reference Davydov, Stewart, Ritchie and Chaudieu37,Reference Jahoda38 which is also a known protective factor against the development of depression.Reference Lee, Nam, Kim, Kim, Lee and Lee39 Although the exact definition varies between disciplines, the American Psychological Association defines resilience as ‘the process of adapting well in the face of adversity, trauma, tragedy, threats or significant sources of stress’.40 Originally, the concept of resilience has been considered as a stable personality trait that enhances the ability to adapt to adverse experiences.Reference Luthar, Cicchetti and Becker41 Yet, recently, resilience is seen as a dynamic and modifiable process that varies across the life course.Reference Davydov, Stewart, Ritchie and Chaudieu37,Reference Southwick, Pietrzak, Tsai, Krystal and Charney42 Consistently, a systematic review of studies in the general population concluded that resilience training based on a combination of mindfulness and cognitive and behavioural skills may be able to improve individual resilience, despite considerable heterogeneity between the trials.Reference Joyce, Shand, Tighe, Laurent, Bryant and Harvey43 Improved resilience, conceptualised as successful stress recovery, positivity and psychological flexibility, is considered to be preventive in the development and recurrence of depression.Reference Waugh and Koster44 Moreover, a low level of resilience during pregnancy is associated with antepartum and postpartum depression.Reference Tobe, Kita, Hayashi, Umeshita and Kamibeppu45–Reference García-León, Caparrós-González, Romero-González, González-Perez and Peralta-Ramírez47 Third-generation behavioural therapies, such as acceptance and commitment therapy (ACT) and mindfulness-based cognitive therapy, aim to enhance mental well-being by utilising resilience factors, including psychological flexibility and mindfulness.Reference Hayes, Luoma, Bond, Masuda and Lillis48,Reference Southwick and Charney49
Because of the modifiable nature of resilience, as well as the changes and challenges associated with pregnancy and the transition to motherhood, we expect that training resilience might be a potential beneficial component of interventions for ADS. In the present systematic review, we therefore provide an overview of trials that evaluated the effectiveness of antepartum resilience-enhancing interventions primarily targeting the improvement of mild-to-moderate ante- and postpartum depressive symptoms. We also investigate whether these (psychological) interventions improve resilience and resilience factors in the ante- and postpartum period.
Method
For this review, we followed the guidelines outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.Reference Moher, Liberati, Tetzlaff and Altman50 The protocol was registered at the international Prospective Register of Systematic Reviews (PROSPERO; registration number CRD42020123592).
Search strategy and study selection
A comprehensive search was developed by an experienced information specialist (R.H.J.O.) for the following electronic databases, up to September 2020: PubMed, EMBASE, CINAHL, APA PsycInfo and Cochrane Library. Search terms expressing ‘depression’ were used in combination with search terms comprising ‘pregnant women’ and search terms comprising ‘resilience’, synonyms of resilience or ‘resilience factors’. The full search strategy is included in Supplementary Appendix 1 available at https://doi.org/10.1192/bjo.2022.60.
Two reviewers (A.L.W. and A.B.W.) independently screened all titles and abstracts for eligibility by using Rayyan, a web-based systematic review application (Rayyan Systems Inc., Cambridge, USA; see www.rayyan.ai).Reference Ouzzani, Hammady, Fedorowicz and Elmagarmid51 The full texts of potential eligible articles were retrieved and independently reviewed by two authors (A.L.W. and A.B.W.) against eligibility criteria. Disagreements were resolved by discussion and consensus or by consulting a third reviewer (J.H.).
Eligibility criteria
To be selected for inclusion, studies had to meet the following criteria: be published in a peer-reviewed journal in English, Dutch or German; utilise a randomised controlled trial or quasi-experimental design with one or more control groups; study pregnant women aged ≥18 years with depressive symptoms; and implement an intervention during pregnancy that (a) aimed to reduce maternal ADS and/or prevent the onset of peripartum major depression, measured by validated self-report or standardised structured clinical interviews, and (b) addressed one or more psychological resilience factors. Based on the literature, the following resilience factors were determined a priori: cognitive/psychological flexibility, mindfulness, acceptance, optimism or positive attributional style, active coping, self-efficacy and self-esteem.Reference Bonanno and Diminich52–Reference Chmitorz, Kunzler, Helmreich, Tüscher, Kalisch and Kubiak56
We only considered internal resilience factors and excluded environmental or non-psychological factors, such as social support or physical activity. Studies examining pregnant women at risk for depression (women with prior depression or anxiety) and/or reporting current ADS were included, provided that depression measurements were used during screening. Also, studies examining pregnant women from the general population were included if they reported a subgroup analysis for participants with current depressive symptoms. Studies that did not use measurements of depression or that exclusively studied women with major depression were excluded.
Quality assessment and data extraction
Methodological quality of the included studies was assessed with the Cochrane Collaboration's tool assessing risk of bias.Reference Higgins, Altman, Sterne, Higgins and Green57 Two authors (A.L.W. and A.B.W.) individually assessed the risk of bias of each study and resolved discrepancies through discussion. Each study was assessed at study level as low risk, high risk or unclear risk for selection bias, performance bias, detection bias, attrition bias, reporting bias and ‘other’ bias, resulting in an overall risk of bias appraisal ranging from low to high risk.Reference Higgins, Altman, Sterne, Higgins and Green57
Data extraction was performed by A.L.W. and verified by A.B.W. Extracted data included details regarding the design of the trial, participants, the description of the intervention and control condition, outcomes and results (change and difference in means, risk ratio and effect size measures). Because of the heterogeneity of the interventions and the outcome measurements, a narrative synthesis was conducted.
Results
The search yielded a total of 6466 articles. Two additional publications were identified through checking references. After removal of duplicates, titles and abstracts of 3975 articles were screened, resulting in 3909 articles not meeting eligibility criteria and leaving 66 articles for full-text review. Of these, 11 articles were identified as eligible, with two articles reporting different outcome measures of the same trial as reported by Muñoz et alReference Muñoz, Le, Ippen, Diaz, Urizar and Soto58 and Urizar and Muñoz.Reference Urizar and Muñoz59 Thus, based on the defined inclusion criteria, ten studies were included in the current review. Fig. 1 illustrates the study selection process, using the PRISMA flow diagram.
Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow chart. *Of the articles included in the qualitative synthesis, two concerned the same study with different outcome and follow-up measures, which were reported as one study in the present review.
Table 1 summarises the characteristics of the included studies. Studies were conducted in Western and non-Western countries. Of the ten studies included, nine studies had a randomised controlled study design, and one studyReference Aslami, Alipour, Najib and Aghayosefi60 used a pre–post test design with a control group. The interventions employed in the included studies are described in Table 2.
Characteristics and findings of the included studies
BDI-II, Beck Depression Inventory-II; BAI, Beck Anxiety Inventory; CBT, cognitive–behavioural therapy; T0, time point 0; T1, time point 1; RCT, randomised controlled trial; BDI, Beck Depression Inventory; EPDS Edinburgh Postnatal Depression Scale; CSEI, Coopersmith Self-Esteem Inventory Adult Form; MBSP, mindfulness-based strengths practice; MAAS, Mindfulness Attention Awareness Scale; STAI I + II, State-Trait Anxiety Inventory I and II; WHO-5, Well-Being Index World Health Organization Five; SCS, Self-Compassion Scale; PSI, Parenting Stress Index; CECPAQ, Comprehensive Parenting Behavior Questionnaire; IBQ-VSF, Infant Behavior Questionnaire-Very Short Form; T2, time point 2; APD, antepartum depression; ADS, antepartum depressive symptoms; PPD, postpartum depression; CES-D, Center for Epidemiologic Studies – Depression Scale; SCID, Structured Clinical Interview; SCL-90, Symptoms Checklist-90; RAC, Risk Assessment Checklist; DASS-sf, Depression Anxiety Stress Scales short form; MMS, Maternal Mood Screener during pregnancy; PANAS, Positive and Negative Affect Schedule; VAS, Visual Analog Scale; GAD-7, Generalised Anxiety Disorder Scale-7; PHQ-9, Patient Health Questionnaire-9; FFMQ, Five Facets of Mindfulness Questionnaire.
a. Effect sizes and P-values are provided when reported. Only primary outcomes for depression and resilience or resilience factors are summarised.
Characteristics of the interventions
CBT, cognitive–behavioural therapy; IPT, interpersonal psychotherapy.
Methodological quality of the studies
The methodological quality of the studies is illustrated in Table 3. One study was evaluated as having an overall low risk of bias,Reference Kozinszky, Dudas, Devosa, Csatordai, Toth and Szabo61 whereas the majority of the studies were evaluated with having some concernsReference Muñoz, Le, Ippen, Diaz, Urizar and Soto58,Reference Urizar and Muñoz59,Reference Yang, Jia, Sun, Ye, Zhang and Yu62–Reference Guo, Zhang, Mu and Ye66 or high risk of bias.Reference Aslami, Alipour, Najib and Aghayosefi60,Reference Fathi-Ashtiani, Ahmadi, Ghobari-Bonab, Azizi and Saheb-Alzamani67,Reference Yazdanimehr, Omidi, Sadat and Akbari68 Except for one study, none of the studies reported blinding procedures. Furthermore, the majority of studies suffered from high and/or unevenly distributed attrition rates. Ratings of ‘unclear’ or ‘high risk of bias’ for the domain ‘Other sources of bias’ were based on baseline imbalance in depression between study groups,Reference Lara, Navarro and Navarrete64,Reference Fathi-Ashtiani, Ahmadi, Ghobari-Bonab, Azizi and Saheb-Alzamani67 no reporting of baseline characteristicsReference Aslami, Alipour, Najib and Aghayosefi60 or poor reporting of analyses and study results.Reference Aslami, Alipour, Najib and Aghayosefi60
Quality assessment of included studies as per domain of risk of bias
+ indicates a low risk of bias regarding this domain; ? indicates that the risk of bias is unclear (identification of a potential risk of bias but its influence on the outcome of the study was appraised as unlikely; or insufficient provision of information on methods and procedures); x indicates a high risk of bias regarding this domain.
a. The study was evaluated as having: +, an overall low risk of bias; ?, some concerns regarding the overall risk of bias; or x, an overall high risk of bias.
Effect of interventions on depressive symptoms
The identified interventions addressing resilience (factors) could be divided into primarily CBT-based and mindfulness-based intervention approaches (see Table 2).
CBT-based interventions
Six studies assessed the effect of a CBT-based intervention on peripartum depressive symptoms, with one study focusing on ADS and five focusing on the reduction and prevention of postpartum depressive symptoms. All six interventions were adapted to pregnancy and included psychoeducation about ante- and postpartum depression. Two interventions were offered individually and four in groups consisting of two to fifteen participants (see also Table 2). Five of the studied interventions were at least partly effective in reducing peripartum depressive symptoms and/or the incidence of depression, whereas one study showed no effect (Table 1). However, except for the study by Kozinszky et al,Reference Kozinszky, Dudas, Devosa, Csatordai, Toth and Szabo61 five of the six studies had a moderate-to-high overall risk of bias.
Kozinszky et alReference Kozinszky, Dudas, Devosa, Csatordai, Toth and Szabo61 studied a 4-week preventive group intervention for pregnant women based on psychoeducation and elements of CBT and IPT. The subgroup analysis of women with ADS (n = 324) revealed an absolute risk reduction of 17.9% of having elevated postpartum depressive symptoms in women in the group intervention compared with women in the control group following group meetings with standard information about pregnancy. Also, Lara et alReference Lara, Navarro and Navarrete64 evaluated an 8-week psychoeducational group intervention with components of CBT, combined with group exercises and supported by a workbook, compared with care as usual plus a self-help book on depression (n = 377). Depression scores in both groups had improved over time, but the cumulative incidence of major postpartum depression in the intervention group was significantly lower than in the control group. However, the study suffered from a high, unbalanced attrition rate and baseline imbalances for depression (see Table 1 for more detail).
Milgrom et alReference Milgrom, Schembri, Ericksen, Ross and Gemmill63 examined the effect of a CBT-based self-help workbook with eight weekly modules and support sessions via telephone, during pregnancy and one time at 6 weeks postpartum, in women at risk for depression and with elevated ADS (n = 143). At 12 weeks postpartum, depressive symptoms in the intervention group had strongly improved (Cohens d = 0.6) and fewer participants scored above the cut-off for depression, compared with the control group that had received usual care. Fathi-Ashtiani et alReference Fathi-Ashtiani, Ahmadi, Ghobari-Bonab, Azizi and Saheb-Alzamani67 (n = 196) also developed an eight-session CBT programme provided individually, using a workbook and video material adapted to the Iranian religious and cultural context. Controlled for baseline scores, post-intervention depressive symptoms in the intervention group were significantly lower than in the control group; however, the results of this study were assessed as highly likely to be biased. Jesse et alReference Jesse, Gaynes, Feldhousen, Newton, Bunch and Hollon65 evaluated a 6-week culturally tailored group CBT intervention for 146 low-income African American, Hispanic and White women in the USA. Participants were stratified into low, medium and high risk, based on their initial depression score. Depressive symptoms (Beck Depression Inventory-II) in the low-risk group had improved post-intervention compared with the control group, whereas no treatment effect was found in the high-risk group. In the subsample of African American women, the high-risk CBT group showed improved depression scores (Edinburgh Postnatal Depression Scale) compared with the high-risk control group. The intervention was thus found to be effective for some of these subgroups. However, the unevenly distributed attrition rate and the small sample size of the subgroups may have influenced these results. Finally, a 12-week CBT group intervention that was culturally tailored to pregnant Latino women with depressive symptoms and/or a past major depressive episodeReference Muñoz, Le, Ippen, Diaz, Urizar and Soto58,Reference Urizar and Muñoz59 showed no differences between groups on depression scores, incidence of major depression, and positive and negative affect, measured at different follow-up time points (n = 45 and n = 57).
Mindfulness-based interventions
Four studies evaluated the effect of mindfulness-based interventions, of which three aimed to decrease ADSReference Aslami, Alipour, Najib and Aghayosefi60,Reference Yang, Jia, Sun, Ye, Zhang and Yu62,Reference Yazdanimehr, Omidi, Sadat and Akbari68 and one aimed to prevent postpartum depressive symptoms up to 12 month after birth.Reference Guo, Zhang, Mu and Ye66 Three interventions targeted women with anxiety and (mild) depressive symptoms.Reference Aslami, Alipour, Najib and Aghayosefi60,Reference Yang, Jia, Sun, Ye, Zhang and Yu62,Reference Yazdanimehr, Omidi, Sadat and Akbari68 Two interventions were provided in groupsReference Aslami, Alipour, Najib and Aghayosefi60,Reference Yazdanimehr, Omidi, Sadat and Akbari68 and two were provided individually online.Reference Yang, Jia, Sun, Ye, Zhang and Yu62,Reference Guo, Zhang, Mu and Ye66 All four mindfulness-based interventions improved depressive symptoms significantly compared with their control conditions. However, there were some concerns regarding the overall risk of bias in two studies,Reference Yang, Jia, Sun, Ye, Zhang and Yu62,Reference Guo, Zhang, Mu and Ye66 and two studies had a high risk of bias,Reference Aslami, Alipour, Najib and Aghayosefi60,Reference Yazdanimehr, Omidi, Sadat and Akbari68 which might have led to an overestimation of the effects.
An individual programme using the approach of mindfulness-based strengths practice significantly improved depressive symptoms (Edinburgh Postnatal Depression Scale score only) at 3 and 12 months postpartum, compared with usual care (n = 314).Reference Guo, Zhang, Mu and Ye66 The 6-week programme included guided exercises focused on self-compassion and was provided online. A second individual online intervention, based on mindfulness-based stress reduction, including an interactive chat function and telephone contact, was effective in reducing post-intervention depression scores (n = 123).Reference Yang, Jia, Sun, Ye, Zhang and Yu62 The 8-week intervention consisted of four sessions plus homework, and was compared with a control group participating in online psychoeducation and a chat group.
Yazdanimehr et alReference Yazdanimehr, Omidi, Sadat and Akbari68 evaluated mindfulness-integrated CBT in group sessions, compared with care as usual. The 8-week intervention significantly improved post-intervention ADS, although the study was rated as having a high risk of bias (Table 3). Moreover, Aslami et alReference Aslami, Alipour, Najib and Aghayosefi60 reported a significantly higher reduction of post-intervention depressive symptoms for women participating in an 8-week group intervention applying mindfulness and Islamic spiritual schemes, compared with both CBT and usual care. However, this study used a quasi-experimental design and risk of overall bias was rated as high (see Table 3).
Effect of interventions on resilience (factors)
None of the studies meeting the inclusion criteria addressed the improvement of resilience directly. Identified resilience components addressed in the selected studies comprised mindfulness, acceptance, coping (positive thinking and self-talk) and self-esteem. Only three of the studies measured change in these resilience factors across the pre- and post-intervention period. Both online mindfulness interventions (mindfulness-based stress reduction and mindfulness-based strengths practice) improved mindfulness scores significantly compared with the control condition post-interventionReference Yang, Jia, Sun, Ye, Zhang and Yu62 and at 3 and 12 months postpartum.Reference Guo, Zhang, Mu and Ye66 In the study examining a CBT programme adapted to the Iranian religious and cultural context, self-esteem had increased in both groups from baseline, at the start of the third trimester, to 2-week postpartum assessments, showing no effect favouring the intervention.Reference Fathi-Ashtiani, Ahmadi, Ghobari-Bonab, Azizi and Saheb-Alzamani67
Discussion
To our knowledge, this is the first systematic review providing an overview of studies evaluating the effectiveness of antepartum resilience-enhancing interventions that aim to reduce ante- and postpartum depressive symptoms and prevent peripartum depression among pregnant women with mild-to-moderate depressive symptoms. According to our criteria, we included ten studies, of which five CBT-based and four mainly mindfulness-based interventions were at least partly effective in reducing peripartum depressive symptoms and/or the incidence of postpartum depression. However, the methodological quality of most of the included studies was low to moderate, which might have led to an overestimation of effects. Only one of the included studies demonstrating the effectiveness of a preventive group intervention based on a combination of psychoeducation and elements of CBT and IPT was rated of high methodological quality.Reference Kozinszky, Dudas, Devosa, Csatordai, Toth and Szabo61 The CBT group intervention adapted to Latina culture evaluated by Muñoz et alReference Muñoz, Le, Ippen, Diaz, Urizar and Soto58 and Urizar and MuñozReference Urizar and Muñoz59 was the only non-effective intervention, yet the sample size was small and the study might have been underpowered to detect a significant difference. Identified resilience factors addressed by the interventions were mindfulness, acceptance, coping (including positive thinking and self-talk) and self-esteem. However, only three studies assessed change in these factors, and the construct of resilience itself was not directly addressed.
Although earlier reviews have revealed a more mixed efficacy of antepartum interventions based on CBT or mindfulness,Reference Sockol21,Reference Lever Taylor, Cavanagh and Strauss34 interventions addressing resilience and resilience factors might be promising in improving peripartum depressive symptoms as revealed by the present systematic review. In contrast to these previous reviews, we included only studies aimed at secondary prevention. This may partly explain the relative consistent pattern of results of the studies included in this review. Moreover, considering the earlier reported negative association between ADS and resilience,Reference Hain, Oddo-Sommerfeld, Bahlmann, Louwen and Schermelleh-Engel46,Reference García-León, Caparrós-González, Romero-González, González-Perez and Peralta-Ramírez47 the findings of the current review indeed support our hypothesis that interventions increasing resilience may be especially beneficial for women with low-to-moderate symptoms, and may secondarily prevent the development of peripartum clinical depression. Our results and earlier literature on antepartum CBT and IPT interventions suggest that psychological approaches, such as classical CBT and IPT, might be more effective for the treatment of clinical peripartum depression,Reference Sockol21–Reference Nillni, Mehralizade, Mayer and Milanovic23 whereas pregnant women with mild-to-moderate depressive symptoms might benefit more from resilience-enhancing interventions.Reference Goodman28,Reference Arch, Dimidjian and Chessick32 Both CBT- and mindfulness-based intervention approaches addressing resilience factors seem promising. However, these findings should be replicated in methodologically rigorous trials.
Attrition rates in four of the CBT-based studiesReference Milgrom, Schembri, Ericksen, Ross and Gemmill63–Reference Jesse, Gaynes, Feldhousen, Newton, Bunch and Hollon65,Reference Fathi-Ashtiani, Ahmadi, Ghobari-Bonab, Azizi and Saheb-Alzamani67 were high or differential, potentially leading to bias. This has previously been described as a methodological problem in trials examining peripartum psychological interventions.Reference Lever Taylor, Cavanagh and Strauss34,Reference O'Connor, Senger, Henninger, Coppola and Gaynes69 In three of the mindfulness-based interventions attrition was low, and occurred slightly more frequently in the control group. Two of these interventions were provided as online programmes, which is in line with low attrition rates reported in previous research on online interventions in the peripartum period.Reference Loughnan, Joubert, Grierson, Andrews and Newby70 All selected interventions, except for the mindfulness-integrated CBT evaluated by Yazdanimehr et al,Reference Yazdanimehr, Omidi, Sadat and Akbari68 were adapted or specifically designed for expectant mothers and included psychoeducational elements on pregnancy, ante- and postpartum depressive symptoms and motherhood. This might enhance engagement and contribute to the effectiveness of these interventions.
The second aim of the present study was to investigate whether the identified psychological interventions improve resilience and resilience factors in the ante- and postpartum period. Interestingly, none of the included studies directly addressed the construct of resilience. However, components of resilience that were addressed by the interventions included mindfulness, acceptance, coping (including positive thinking and self-talk) and self-esteem. Nevertheless, change in these factors across the pre- and post-intervention period was assessed in only three studies. Both online mindfulness-based interventions significantly increased mindfulness post-interventionReference Yang, Jia, Sun, Ye, Zhang and Yu62 and postpartum.Reference Guo, Zhang, Mu and Ye66 Fathi-Ashtiani et al,Reference Fathi-Ashtiani, Ahmadi, Ghobari-Bonab, Azizi and Saheb-Alzamani67 who evaluated the effectiveness of an adapted and culturally-specific mindfulness programme for the Iranian context, observed that the increase in self-esteem did not differ between the intervention group and care as usual. To shed more light on effect mechanisms involved, future research should include process measures alongside primary outcomes.
In addition to mindfulness- and CBT-based interventions, we also expected to identify third-generation behavioural therapies such as ACT for the prevention and treatment of peripartum depression, as ACT is increasingly popular and comprises resilience factors, including psychological flexibility and mindfulness.Reference Hayes, Luoma, Bond, Masuda and Lillis48,Reference Southwick and Charney49 A recent meta-analysis demonstrated the effectiveness of ACT in successfully reducing mild depressive symptoms in the general population.Reference Bai, Luo, Zhang, Wu and Chi71 Bonacquisti et alReference Bonacquisti, Cohen and Schiller72 developed a rationale for an antepartum four-session ACT intervention, suggesting that the emphasis of ACT on the enhancement of psychological flexibility instead of an emphasis on the reduction of depressive symptoms may lead to higher mental well-being. This might be especially beneficial for pregnant women, as it may reduce feelings of (self-)stigmatisation, and positively influence somatic complaints and the adjustments related to the transition to motherhood.Reference Flynn, Henshaw, O'Mahen and Forman30,Reference Bonacquisti, Cohen and Schiller72 However, our search identified only one non-eligible study (because of inadequate study population characteristics) showing that ACT had improved quality of life and anxiety in pregnant Iranian women,Reference Vakilian, Zarei and Majidi73 as well as a pilot study without a control group, which observed that an antepartum ACT group intervention was feasible and had improved both psychological flexibility and depressive symptoms. In addition, we found two ongoing trials evaluating the impact of ACT on depressive symptoms and anxiety.Reference Witteveen, Henrichs, Walker, Bohlmeijer, Burger and Fontein-Kuipers74,Reference Thomas75 In line with a recent review of reviews of psychological interventions for peripartum depression,Reference Branquinho, Rodriguez-Muñoz, Maia, Marques, Matos and Osma76 we conclude that ACT seems promising, but more research assessing its impact on ADS is needed.
Strengths and limitations
This study has several strengths. As far as we are aware, it is the first review of resilience-enhancing psychological interventions during pregnancy. A comprehensive search was conducted, followed by systematic screening, quality assessment and review of the studies performed independently by two researchers. Nevertheless, there are a few limitations. We only included studies that were peer-reviewed and published in English, Dutch or German, which might have biased our results. Furthermore, although we included randomised controlled trials or trials using a quasi-experimental design with control group, the conclusions of the present review are limited because the methodological quality of the individual studies was rated as low to moderate, with the exception of one high-quality study. Moreover, as only two studies reported effect sizes, assumptions about the clinical significance regarding the effectiveness of the successful interventions are difficult to make. Our conclusions also might be influenced by the limited number of studies included. However, we included studies with interventions aimed at pregnant women with elevated depressive symptoms and/or risk factors (women with prior depression or anxiety) only, as implementation of these interventions in obstetric mental healthcare is suggested to be more feasible than primary prevention interventions. The second reason for the limited number of studies was that we only included interventions provided during pregnancy, as early treatment of ADS is essential regarding the adverse effects on the pregnant woman and her unborn child.Reference Walker, de Rooij, Dimitrova, Witteveen, Verhoeven and de Jonge77,Reference Stein, Pearson, Goodman, Rapa, Rahman and McCallum78 Finally, we were only able to conduct a narrative synthesis based on the studies included in the current systematic review, as a meta-analysis was not considered feasible because of the heterogeneity of the interventions.
In conclusion, our results suggest that antepartum psychological interventions addressing the enhancement of resilience factors, such as mindfulness, acceptance, coping and self-esteem, seem effective in improving peripartum depressive symptoms. The ten interventions identified could be divided into primarily CBT-based and mindfulness-based intervention approaches. However, the methodological quality of the included studies was mostly low to moderate, which must be considered when interpreting the results. In contrast to our expectations, no interventions using an ACT approach were included. Considering the adverse effects of peripartum depression on the mental and physical well-being of mothers and their (unborn) children, the promotion of well-being and prevention of exacerbation of depressive symptoms during pregnancy are essential. Therefore, future research should invest in more rigorously designed studies evaluating the effectiveness of antepartum resilience-enhancing interventions, using appropriate process measures, and should report measures of effect to enable future meta-analyses. Moreover, future studies should particularly investigate the role of resilience in ACT interventions for the reduction of ADS, to improve the mental well-being of pregnant women and their children.
Supplementary material
Supplementary material is available online at https://doi.org/10.1192/bjo.2022.60.
Data availability
Data availability is not applicable to this article as no new data were created or analysed in this study.
Author contributions
The study was designed by A.L.W. and J.H.. R.H.J.O. and A.L.W. developed the search strategy. A.L.W. and A.B.W. screened and analysed the data. A.L.W. drafted the manuscript. A.L.W., A.B.W., C.J.V., J.H. and A.d.J. made substantial contributions in revising the manuscript and interpretation of the results. All authors read and approved the final manuscript.
Funding
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Declaration of interest
None.
Open access
eLetters
No eLetters have been published for this article.