Search strategy

We conducted a literature search in PsychINFO, PubMed, and CINAHL, according to guidelines in the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement [13]. The terms ‘agency’, ‘autonomy’, ‘choice’, ‘control’, ‘domestic decision-making power’, ‘economy’, ‘empower’, ‘empowerment’, ‘fertility intervention’, ‘leadership’, ‘power’, ‘pregnancy’, ‘resources’, ‘transformation’, ‘voice’, and ‘women’s health’ were combined with the search terms ‘birth’, ‘birthweight’, ‘preterm’, ‘depression in pregnancy’, ‘perinatal depression’, ‘postpartum depression’, and ‘premature birth’. The search terms ‘CenteringPregnancy’, ‘child marriage’, and ‘sexual activity’ also emerged as relevant due to their presence in the initial search results, and the three terms were thus added to the final search.

Inclusion and exclusion criteria

To be included in this review, studies had to be peer-reviewed and include (1) a sample of women who were either pregnant or within the first year postpartum; (2) a measure of empowerment or lack thereof or an intervention aimed at supporting women’s empowerment; and (3) a measure of perinatal depressive symptoms or of prematurity (PTB, LBW). Of note, whereas PTB/LBW is determined at the time of birth, maternal depressive symptoms can occur in the postpartum period as well. Thus, studies of empowerment in the first year postpartum were only relevant to the literature on depressive symptoms. We identified 1 study assessing depressive symptoms during pregnancy, 11 in the postpartum period, and 4 in both pregnancy and postpartum. Empowerment was measured or an intervention supporting empowerment was administered postpartum for 9 studies, in pregnancy for 3 studies, and in both for 4 studies. No limits were set for date of publication, maternal age, study design, or language of publication; nevertheless, only English language publications were identified.

Selected studies

The literature search identified 150 records in PubMed, 74 records in PsycINFO and 40 records in CINAHL (see Fig. 1 for PRISMA flow chart). After removing 30 duplicates, 234 articles were screened for eligibility. Five records were excluded based on title and abstract reviews because they were animal studies. Of the remaining 229 full-text articles, 202 records were excluded because they did not include a measure of empowerment or an intervention supporting empowerment (n = 74); did not measure perinatal depressive symptoms, PTB, or LBW (n = 72); or lacked a measure of empowerment as well as perinatal depressive symptoms and PTB or LBW (n = 56). The remaining 27 articles were included in this manuscript. These articles report on 25 independent studies with a total of 17,795 participants. Across studies, sample sizes ranged from 30 [14] to 6155 [15], with a mean of 693 participants (SD = 1196). Most studies were conducted in the US (70.83%), and the remaining studies in Iran, Pakistan (both 8.33%), Taiwan, Canada, and the Netherlands (all 4.17%). Of the studies conducted in the US, one did not provide ethnicity information [16]. In the remaining studies, the majority of participants were African-American (36.27%), followed by White (28.13%), Hispanic (24.42%), Asian (7.01%), Native American (<1%), and Other (4%). Average participant ages ranged from 15 [17] to 33 years old [18]. Across studies, participants were an average of 25.01 years old (SD = 3.79).

Quality ratings

Study quality ratings were conducted according to National Institute of Health standards for controlled intervention studies as well as for observational, cohort, and cross-sectional studies [19]. We considered studies with a quality score of eight to 14 of good quality, and those with scores of five to seven as fair. While the National Institute of Health standards recommend to not base quality ratings on sum scores alone, we contend that this categorization also matched our individual readings of each study. No studies were considered to be of poor quality.

Conceptualizations of empowerment

We identified six studies that assessed empowerment in cross-sectional/observational study designs and 19 studies that reported on empowerment interventions. Observational studies used mostly self-report or interview measures, such as the Decision-Making Dominance Subscale of the Sexual Relationship Power Scale [20] and critical ethnographic interviews [14]. Studies used different types of empowerment interventions (see Table 1 for details). The majority report on CenteringPregnancy (CP), a group prenatal care model. “CenteringPregnancy unifies the components of prenatal care—risk assessment, education, and support within the group—and encourages women to take responsibility for their own health” ([21], p. 46). CP was included in this review because it aims to empower women by increasing their health self-care efficacy during pregnancy. “The woman’s involvement in self-care activities, the discussion and education format, the worksheets and handouts, and the sharing among the women all lead to her enhanced sense of empowerment. This, in turn, results in a sharing of power between the provider and the consumer” ([21], p. 53). Other studies report on the Creating Opportunities for Parent Empowerment (COPE) program for parents with premature infants in Neonatal Intensive Care Units. COPE aims to inform parental decision-making and increase parental knowledge about infant developmental outcomes, consequently easing parent–infant interactions. Thus, it aims to support women’s empowerment by increas[es]ing parenting confidence, participation in care, self-confidence, and control, affecting health outcomes including reducing parenting anxiety and depressive symptoms as well as reducing their stay in the Neonatal Intensive Care Units.

The remaining studies used various other interventions. The first is the Guided Participation program. It is included in the framework of empowerment-related models because it acts on autonomy and self-determination by informing mothers of premature infants of effective premature infant care and increases parenting self-efficacy, feeding competencies, and positive infant behaviors to reduce anxiety and depressive symptoms associated with parenting a premature infant [22]. This has empowerment implications in terms of promoting self-confidence among mothers of premature infants who may feel helpless, scared, and might lack confidence in their ability to care for their premature newborn. The second is an internal health locus of control intervention that aims to promote empowerment of women through education about self-care such as pregnancy stage monitoring, nutrition, and health [12]. One study reports on a parent-to-parent and parent-to-provider dialogue intervention related to participation, collaboration, awareness, sense of control, self-efficacy, and self-help as well as access to resources and personal action to shift the personal health decision-making power from healthcare provider to client [18]. Finally, Mom Power is a parenting and self-care skills group program for high-risk mothers and their young children that aims to empower by increasing parenting and self-care knowledge [23].

Maternal perinatal depressive symptoms

Intervention studies

Among the intervention studies were seven randomized controlled trials [12, 22, 24–28], two quasi-experimental studies [18, 29], and one prospective cohort study [23].

Three of the randomized controlled trials, all by the same team of authors, used the COPE program as an experimental intervention. The first of this set of studies evaluated COPE in a small sample of 42 mothers of premature infants [26]. When depressive symptoms were assessed 4 to 8 days after admission and 1 to 4 days prior to discharge, women receiving the intervention had fewer depressive symptoms than those receiving traditional care. However, no differences were observed when scores were aggregated across all phases of the intervention. A second study of 260 families confirms that mothers in COPE report fewer post-hospital depressive symptoms than those in the control group, and further expands by suggesting that COPE may also reduce parental stress [27]. The last article reporting on a subset of 246 mothers found participation in COPE was related not only to mothers’ decreased post-hospital depressive symptoms, but also to reduced anxiety [28]. Thus, while COPE was conceptualized to improve premature infant health and developmental outcomes, the findings of these three studies suggest that COPE may also be an effective tool for reducing maternal postpartum symptoms of depression, stress, and anxiety.

Two studies reported on CP randomized controlled trials [24, 25]. Among women who initially reported high stress, those in a ‘CP+’ group, which added information about HIV prevention as well as components of psychosocial functioning such as behavioral risk assessment, goal setting, communication, and negotiation skills, had significantly fewer depressive symptoms at 1 year postpartum than those in the standard CP and control groups [24]. These findings may suggest that CP could be useful in terms of improving maternal depressive symptoms if it also addresses the specific needs of the patient population. Another CP study of 322 pregnant women in the military found no differences in prenatal or postnatal depressive symptoms between those in a CP condition and those receiving standard care [25]. Similarly, a smaller quasi-experimental study of 49 Hispanic women suggests no significant differences in perinatal depressive symptoms between those in CP and those receiving traditional care [29]. However, in addition to the small sample, the study groups differed in that those in the CP group had more primigravidas.

Of the remaining studies, three found significant associations between empowerment and perinatal depressive symptoms. The strongest support comes from a prospective cohort study of 80 mother–child pairs who participated in the Mom Power intervention. Results suggest that participation in Mom Power is not only associated with significant reduction in depressive symptoms, but also with lower symptoms indicative of post-traumatic stress disorder risk and reduced caregiving helplessness [23]. Another study empowered women by providing an internal health locus of control intervention and found significant increases in internal health locus of control as well as lower postpartum depressive symptoms among those in the intervention group [12]. Similarly, a study of 70 Taiwanese parents of preterm infants, using a parent-to-parent and parent-to-provider dialogue, found that post-intervention depression scores were significantly lower for those in the intervention than for those in the control group. Those in the intervention group also had higher childrearing self-efficacy [18]. In contrast, findings from a small study of 42 mothers with very LBW infants (≤1200 g) indicate no differences in maternal perinatal depressive symptoms between those in the Guided Participation and control groups [22]. However, those in the intervention group were better able to regulate infant negative affect and feeding behaviors than those in the control group.

Observational studies

Among the 6 observational studies, 2 were longitudinal [30, 31], 3 cross-sectional [16, 20, 32], and 1 used a qualitative study design [14]. Despite variations in definitions of empowerment, all quantitative studies yielded significance, and studies suggest a positive association between perinatal depressive symptoms and domestic violence [20, 30], emotional and economic relationship dominance by an intimate partner [14], and external locus of control [16] as well as negative associations with domestic decision-making power [32] and the power to manage household finances [31].

Summary

In sum, all observational and all but three of the intervention studies [22, 25, 29] suggest reduced perinatal depressive symptoms with empowerment. More specifically, studies using the COPE, Mom Power, internal health locus of control, parent-to-parent and parent-to-provider dialogue interventions were effective in reducing maternal perinatal depressive symptoms. In contrast, the CP intervention did not significantly improve maternal affect, unless adaptations specific to the study population were made. Similarly, the Guided Participation intervention did not improve maternal health outcomes, likely because it was not primarily developed to address maternal health outcomes, but to improve infant feeding competencies and affect regulation.

PTB and LBW

Among the eight retrospective cohort studies, five provided evidence of at least some associations between empowerment and prematurity. One of these studies found CP to be associated with lower PTB and LBW rates in a sample of 124 adolescents [17]. Three studies, including samples of 216 Latinas [39] as well as ethnically diverse samples of 4083 [38] and 379 [36] women, had mixed findings, reporting reductions in the rate of PTB but not of LBW. A final study reports higher birth weight and higher gestational age at birth as well as significantly reduced risk of very LBW in the CP compared to the control group [15].

The remaining three retrospective cohort studies, all with homogenous samples including 268 African-American women [37], 487 Latinas [40], and 404 US military women [41], suggest no impact of CP on either PTB or LBW. These null findings may, at least in part, be explained by confounding variables. For example, one study reports that those in CP were significantly younger than those in standard care [37], and another reports that those in CP were more likely to be in active duty, younger, and primiparous [41], but these variables were not statistically controlled for.

Summary

Six of the CP studies found significant benefits for PTB and/or LBW, with an additional two studies suggesting benefits for gestational age and birthweight. Specifically, a retrospective cohort study and a randomized controlled trial utilizing the CP+ intervention found significant benefits of CP for both PTB and LBW. An additional four studies found an association between CP and reduced PTB, but not LBW. Similarly, two studies found significantly higher gestational age at birth and birthweight among CP participants compared to those in the control group; however, effects for PTB and LBW did not reach significance. Three other CP studies found no significant association with gestational age, birthweight, PTB, or LBW. Confounds related to sample characteristics and study design may play an important role in the effectiveness of CP.

Recommendations for future work

It is encouraging that the overall pattern observed in this small literature review is suggestive of significant associations between maternal empowerment-related concepts, perinatal depressive symptoms, and prematurity. To move this literature forward, both theoretically and in terms of the development of successful interventions, we propose several future directions that seem particularly promising, in our view.

First, while most studies provide evidence for an association between empowerment-related concepts and perinatal depressive symptoms or prematurity, we also identified some studies that report null findings. These inconsistencies could be the result of different definitions of empowerment and of differences in how empowerment was measured or supported through interventions. It is also possible that the divergent findings are the result of differences in sample composition, including, for example, age, ethnicity, and socioeconomic status. Moreover, a portion of the studies reviewed here included relatively small sample sizes and some were underpowered because they were originally designed to investigate outcomes other than PTB/LBW and perinatal depressive symptoms [22, 29]. We recommend that future, adequately powered studies carefully test the role of possible moderators in the link between women’s empowerment, perinatal depressive symptoms and prematurity, and that more detailed attention will be given to methods of measuring or supporting empowerment.

Moreover, we note that existing studies have focused on empowerment as it specifically relates to women’s parental role. While the parental role is perhaps most salient in the context of empowerment of pregnant women and new mothers, other significant life changes occur with the birth of a child, in particular the birth of a first child. Across many cultures, new mothers often scale back on or halt their involvement in non-parental societal roles such as work outside the home, leading to decreases in financial independence as well as changes in social relationships and social status. We could not identify any studies that tested whether the empowerment of pregnant or postpartum women in domains other than the parental one is associated with perinatal depressive symptoms or prematurity. It would be an important contribution to the literature to study the relative impact of different facets of empowerment during the perinatal period on maternal and infant birth outcomes, because this knowledge would provide the groundwork for targeted interventions.

Related to the above, we further observe that studies sometimes use a vague conceptualization of empowerment that may be confounded with other interpersonal processes in group intervention studies. For example, the use of group prenatal care models gives rise to companionship and social support as confounding variables because these group processes are distinct from empowerment, but also may be associated with risk of perinatal depressive symptoms [52, 53]. Moreover, many interventions aimed to empower through provision of information, knowledge, and health literacy, but they did not include a measure of empowerment. Including a measure of empowerment in the intervention studies would allow for the identification of particular components of empowerment, such as increases in self-competence that may be particularly beneficial for reducing perinatal depressive symptoms and PTB or LBW. In observational studies, the operationalization of empowerment is somewhat clearer and includes increases in domestic decision-making power, financial autonomy, and internal locus of control.

Finally, we note that the literature on women’s empowerment and PTB/LBW is distinct from that on empowerment and perinatal depressive symptoms. This may be because interventions either focused on improving infant outcomes or on reducing postpartum depressive symptoms in mothers of infants that were already born premature. However, there is evidence that maternal perinatal depression and prematurity are correlated and may, in fact, share pathophysiological pathways [11, 54, 55]. The role of perinatal empowerment as a factor associated with physiological pathways related to both perinatal depressive symptoms and prematurity risk therefore merits further investigation.