Scoping Review of the Associations Between Perinatal Substance Use and Perinatal Depression and Anxiety

Abstract

Objective

To evaluate published literature on the associations between perinatal substance use (PSU), perinatal depression and anxiety (PDA), and known maternal–newborn outcomes.

Data Sources

We conducted a systematic search of health-related databases, including PubMed, CINAHL, MEDLINE, and EMBASE. Search terms included maternal mental health, substance use, postpartum, opioid, alcohol, methamphetamine, addiction, dependence, pregnancy, depression, and anxiety.

Study Selection

We included English-language, peer-reviewed reports of primary research and systematic reviews that were published between 2010 and 2020 and focused on PSU and PDA. We excluded commentaries, non-systematic reviews, and articles on maternal mental health other than PDA. Fourteen of 379 articles met the inclusion criteria.

Data Extraction

We used the Joanna Briggs Institute Review Guidelines to guide extraction of the following data: author(s), year of publication, type of study, country of origin, study sample, targeted substance(s), mental health, key findings, and recommendation(s).

Data Synthesis

In studies of PSU, researchers identified high association with PDA. Likewise, researchers investigating PDA found high association with PSU. Findings from these articles suggested an increasing risk for PSU with increasing severity of PDA, depending on the specific substances of use. Findings also indicated that women with polysubstance use have higher odds for comorbid perinatal mental health conditions. A relationship between PSU and PDA and adverse newborn outcomes such as low birth weight was found.

Conclusion

There is a paucity of published research on co-occurring PSU and PDA. However, polysubstance use appears to be associated with the highest risk for PDA. It is essential to address PSU and PDA together to better understand the effects on maternal and infant outcomes.

Precis

Because of the bi-directional relationship between perinatal substance use and perinatal depression/anxiety, it is imperative to address these conditions together in the perinatal period.

In the United States, 19.5 million women (15.4%) age 18 and older have used illicit drugs in the past year (Center for Behavioral Health Statistics and Quality, 2017). All women, particularly those between the ages of 18 and 29 years, are at risk of developing substance use disorders in their reproductive years, perhaps related to hormonal fluctuations during menstrual cycles, fertility, pregnancy, breastfeeding, and the menopause (National Institute of Drug Abuse [NIDA], 2020; Wright et al., 2016). Women reported using substances for weight control, fighting exhaustion, coping with pain, and attempts to self-medicate to treat mental health conditions (Becker et al., 2017; NIDA, 2020). Approximately 15% of infants are affected by prenatal alcohol or illicit drug exposure (National Center on Substance Abuse and Child Welfare, n.d); data from the National Survey on Drug Use and Health indicated that 5.4% of pregnant women used illicit drugs and 9.4% of pregnant women reported alcohol use within the last 30 days (Center for Behavioral Health Statistics and Quality, 2016). Many known adverse maternal (miscarriage, stillbirth, placental abruption, preterm labor, preterm birth, psychiatric comorbidity, overdose, infections), fetal (congenital disorders, intrauterine growth restriction), and newborn outcomes (low birth weight, neonatal abstinence syndrome, extended hospitalization, sudden infant death) are associated with perinatal substance use (Kotelchuck et al., 2017; Patrick et al., 2012).

CALLOUT 1

Based on the systematic review of Van Niel & Payne (2020), perinatal depression is one of the most common complications of pregnancy and affects approximately one in seven women during pregnancy and one in five women following childbirth. In another systematic review, Fawcett et al (2019) found that the prevalence of experiencing anxiety disorder during pregnancy or the postpartum period is estimated to be 20.7%, HDI_95% (16.7% - 25.4%; Fawcett et al., 2019). Perinatal depression and generalized anxiety disorder are highly comorbid, and many women suffer from both conditions (Wisner et al., 2013). These findings have prompted many health care providers to screen women for perinatal depression and anxiety (PDA) simultaneously (American Academy of Pediatrics, 2017; American College of Obstetricians and Gynecologists, 2018; American College of Obstetricians and Gynecologists Committee on Obstetric Practice, 2017; Siu and U.S. Preventive Services Task Force, 2016; U.S. Preventive Services Task Force, 2019).

In a sample of women with PDA in Nairobi a greater risk (RR = 3.80, 95% CI 1.73–8.37) for pregnancy complications and pre-term births was reported (Mochache et al., 2018). Other examples of adverse perinatal health outcomes associated with PDA included poor adherence to health care, poor nutrition, smoking, and substance abuse (Kendig et al., 2017). Furthermore, PDA affects infants directly and indirectly, increasing the risk for biological effects such as low birth weight and preterm birth (p = 0.012 for PDA vs only depression; p = 0.001 for PDA vs only anxiety; p = 0.000 for PDA vs no psychiatric diagnosis; Uguz et al., 2019) and behavioral deficits in later their life (χ²(l) = 10.7, p < .01; Priel et al., 2019).

It is important to note that the negative effects of PDA are not independent of those of PSU (Metz et al., 2018). Perinatal depression and anxiety have a known linkage with increased use of tobacco, alcohol, and sleep disorder drugs such as sleep aids during pregnancy and the postpartum period (Newport et al., 2012). Engaging in self-medication by smoking tobacco and using substances to relieve symptoms of PDA during pregnancy and the postpartum period has also been reported (Smith et al., 2017).

Perinatal substance use and PDA contribute independently to adverse outcomes for women (preterm labor) and their infants (small for gestational age, prematurity; Tolia et al., 2015), difficult neonatal transition (Tolia et al., 2015), and impaired child growth and development (Substance Abuse and Mental Health Services Administration [SAMHSA], 2012; World Health Organization, 2014). However, the outcomes related to PSU and PDA are typically investigated independently without consideration for associations that may exist between the two or the potentiating effects that may pertain. Although clinicians may assume that PDA and PSU are highly associated, there is limited evidence to document a relationship. The primary objective of our scoping review was to evaluate published literature on the associations between PSU, perinatal depression and anxiety PDA, and known maternal–newborn outcomes.

Methods

We used a scoping review approach (Arksey & O’Malley, 2005; Joanna Briggs Institute, 2015) to accomplish the objective of our review. This approach provided an effective and efficient way to identify and analyze available literature, aggregate knowledge, leverage several study designs, chart the data, and summarize the findings (Arksey & O’Malley, 2005).

We developed a search and review protocol based on the method used by the Joanna Briggs Institute (2015) and the five-step framework developed by Arksey and O’Malley (2005). The approach included identification of the topic of interest, keywords, and data sources, for example, electronic databases (Arksey & O’Malley, 2005). Two authors (RP, GL) conducted the article selection and applied inclusion and exclusion criteria to all the citations identified. The result of this approach led to a narrative format to describe findings regardless of the quality of studies (Arksey & O’Malley, 2005).

Search Strategy

We conducted a search of the literature in four electronic databases for the most recent 10 years (i.e., 2010–2020; PubMed, the Cumulative Index to Nursing and Allied Health Literature [CINAHL], MEDLINE, and EMBASE). The final search terms included combinations of the following: maternal mental health, maternal, substance use, substance abuse, postpartum, perinatal, mental health, opioid, alcohol, methamphetamine, addiction, dependence, pregnancy, depression, anxiety, and bipolar disorder. To meet inclusion criteria, articles were published in English and were peer-reviewed reports of primary research or systematic reviews in which the study samples included women during pregnancy and/or up to one year after childbirth, and there was a focus on PSU and PDA. Articles were included from all countries or geographical areas. Because we focused on PDA, we excluded reports of studies on other perinatal mental health conditions such as postpartum psychosis, obsessive–compulsive disorders, and post-traumatic stress disorder.

Data Abstraction and Synthesis

The following extraction fields were developed based on recommendations from the Joanna Briggs Institute (2015) and Arksey and O’Malley (2005): author(s), year of publication, type of study, objectives, country of origin, study population and sample size (if applicable), targeted substance(s), types of mental health condition(s), key findings, and recommendation(s) that described gaps in the research. We imported studies to Covidence systematic review software (Veritas Health Innovation) and conducted a two-step screening process whereby one of the authors (RP) screened titles and abstracts for exclusion/inclusion criteria and relevancy. When this author determined there was any ambiguity about possibly relevant articles, a second author (GL) screened the article. After screening, two authors (RP, GL) independently conducted a full-text review for the 24 studies using the Covidence software. Subsequently, summaries of articles were entered into an Excel spreadsheet by one author (RP) and reviewed by a second (GL) for consistency. Agreement was reached to exclude 10 studies that did not meet inclusion criteria, resulting in 14 studies for synthesis. One author (RP) abstracted all results (Supplemental Table S1), which was reviewed by the second author (GL). Finally, the abstracted results were synthesized into narrative, and reviewed by both authors (RP and GL).

Results

Study Selection

We identified a total of 379 articles from electronic searches. We removed 189 articles because of duplication and screened 182 articles for potential inclusion (Babineau, 2014). A total of 158 articles were excluded through title and abstract screening, which left 24 articles. The first two authors reviewed the full text review of the 24 articles using the Covidence software. We excluded 10 articles so 14 articles remained for the scoping review (see Figure 1). There were no disagreements between the two authors on inclusion or exclusion of any articles.

Figure 1.

PRISMA diagram for scoping review of the associations between perinatal substance use and depression/anxiety

Table S1 summarizes the included articles. We included one qualitative study with a grounded theory approach, cross-sectional studies (n = 3), secondary data analyses (n = 3), a prospective observational study, one retrospective observational study, retrospective chart reviews (n = 2), one epidemiological study, one retrospective cohort study, one mixed-methods study, and no systematic reviews. Half of the studies (n = 7) were conducted in the United States and four were conducted in Canada. In one study, researchers compared results between the United States and New Zealand, and in another between Australia and South Africa. Seven of the 14 studies were published between 2012 and 2013 and the most recent between 2017 and 2020 (n = 5).

Substances of Use

In most studies (n =10), researchers screened for more than one substance and used general terms such as substance use, lifetime substance use, and drug use in their titles. Polysubstance use included but was not limited to alcohol, cocaine, cannabis, crack cocaine, heroin, crystal methamphetamine, methamphetamine, tobacco/cigarette use, caffeine, phencyclidine, and benzodiazepines (Dennis & Vigod, 2013; Newport et al., 2012; Zhao et al., 2017). However, some researchers used general terms such as illicit substances (Zhao et al., 2017), recreational drugs (Dennis & Vigod, 2013), other substance use (Onah et al., 2016), drug use (Kuo et al., 2013; Prevatt et al., 2017), and illegal drugs (Metz et al., 2018) when discussing substances other than alcohol and/or cigarette use. Newport et al. (2012) categorized substance exposure within the following classes: prescription drugs, such as psychotropic, obstetric, and other medical drugs; nonprescription drugs, such as over-the-counter drugs and herbal medicines; prenatal vitamins; and habit-forming substances such as tobacco, alcohol, caffeine, and illicit substances. Metz et al. (2018) recruited women who reported past use of non-medical opioid or non-opioid illegal drugs other than marijuana, but all participants were asked whether they had polysubstance use. One study title included “opioid-dependent women,” but participants were recruited from a substance abuse treatment program that included other substance use (Holbrook & Kaltenbach, 2012). Young-Wolff et al. (2020) focused on the association between cannabis use and depression/anxiety. Shah et al (2012) and Wouldes et al (2013) focused on prenatal methamphetamine use, while Benningfield et al. (2010) and Faherty et al (2018) evaluated prenatal opioid use and its effects on women and infants.

CALLOUT 2

Categories of Perinatal Mental Health

In nine of the studies, researchers included mental health diagnoses and symptoms in addition to PDA; obsessive–compulsive disorder (Benningfield et al., 2010; Linden et al., 2013; Prevatt et al., 2017; Wouldes et al., 2013), posttraumatic stress disorder (Benningfield et al., 2010; Prevatt et al., 2017; Zhao et al., 2017), trauma (Young-Wolff, 2020), somatization (Wouldes et al., 2013), phobic anxiety (Linden et al., 2013; Wouldes et al., 2013), paranoid ideation (Linden et al., 2013; Wouldes et al., 2013), psychoticism (Linden et al., 2013), hostility (Linden et al., 2013; Wouldes et al., 2013), bipolar disorder (Faherty et al., 2018), schizophrenia (Faherty et al., 2018; Zhao et al., 2017), suicidality (Benningfield et al., 2010; Onah et al., 2016), and bulimia (Benningfield et al., 2010) were described as emotional and mental health illnesses. Dennis and Vigod (2013) focused on the relationship between postpartum depression and substance use, as well as the prevalence of both. Kuo et al. (2013) evaluated the association between depression and anxiety symptoms and cannabis use during pregnancy. In the only qualitative study, Kuo et al. (2013) assessed the treatment needs of women with PSU and depression. Other researchers focused on perinatal depression and/or anxiety and the relationship with PSU or vice versa (Holbrook & Kaltenbach, 2012; Metz et al., 2018; Newport et al., 2012).

Relationship between PSU and PDA

In most of the studies (n = 12), PSU and PDA were significantly associated. Dennis and Vigod (2013) found that 20% of women with perinatal depression had a history of substance use and 22% of women with perinatal depression had used illegal drugs during pregnancy. Study results from Newport et al. (2012) indicated that the severity of both maternal depression and anxiety were positively correlated with the number of weeks that hypnotic drugs were used (r = 0.28 [P < .0001] and r = 0.19 [P = .008] respectively) and tobacco was used (r = 0.21 [P = .003] and r = 0.20 [P= .006] respectively). Perinatal depression was positively correlated with weeks of exposure to opioid analgesics (r = 0.14 [P = .05]) and all prescription drugs (r = 0.17 [P = .02]) (Newport et al., 2012).

Holbrook and Kaltenbach (2012) found that up to 68.8% of women who enrolled in a substance abuse treatment program had a history of depression, and 43.7% of those women screened positive for PPD. Also, nearly 20% of women who had a history of depression had co-morbid anxiety (Holbrook & Kaltenbach, 2012). In the study by Onah et al. (2013), 18% of women who reported substance use during pregnancy were experiencing depressive symptoms, and 19% had a current anxiety diagnosis (Onah et al., 2016), suggesting that mothers who use substances during pregnancy or in early motherhood are at higher risk for depression and anxiety compared to the general population.

Cannabis use during pregnancy was also associated with depression and anxiety disorders (Young-Wolff et al., 2020). In one study, pregnant women with anxiety disorders (aOR, 1.90; 95% CI, 1.76–2.04) and depressive disorders (aOR, 2.25; 95% CI, 2.1–2.41), or PDA (aOR, 2.65; 95% CI, 2.46–2.86) had greater odds for using cannabis compared to pregnant women without PDA (Young-Wolff et al., 2020). Although the risks associated with cannabis use are not well-documented, these study results suggested that early screening for PDA along with appropriate intervention could decrease the risk of cannabis use (Young-Wolff et al., 2020).

Methamphetamine use was associated with a higher risk for having a co-morbid psychiatric diagnosis and/or emotional illness (Shah et al., 2012). Wouldes et al. (2013) reported that 47.6% of U.S. childbearing women who reported methamphetamine use had a psychiatric diagnosis compared to 26.4% of women who did not use methamphetamines. At the same time, having a psychiatric diagnosis significantly increased the odds (OR 2.67, CI 1.63–4.35) of having a substance use disorder (Wouldes et al., 2013).

Opioid use was also a predictor for co-morbid psychiatric symptoms. Benningfield et al (2010) reported that 64.6% of 174 opioid-dependent pregnant women screened positive for one or more psychiatric symptoms: 40% for anxiety, 33% for major depressive disorder, and 48.6% for mood disorder. In another study, women who had an infant with neonatal abstinence syndrome (NAS) were at 2.5 times (95% confidence interval [CI] 1.3–5.0) higher risk for depression (Faherty et al., 2018). Women who had an infant with long-term prenatal opioid exposure without showing signs and symptoms of NAS had a 1.8 times (95% CI 1.2–2.7) higher risk for depression compared to women who had no history of opioid use (Faherty et al., 2018).

Prevatt et al. (2017) found that lifetime drug use was a significant predictor of postpartum anxiety and stress. Bivariate analysis indicated that women had higher levels of postpartum anxiety and depression if they met criteria for lifetime substance abuse and substance dependence. The significance of polysubstance use was discussed in two studies (Metz et al., 2018; Zhao et al., 2017). Metz et al. (2018) used data from the National Survey on Drug Use and Health and found that, among pregnant women, opioid polysubstance use was associated with higher lifetime depression (39.8%), past-year depression (29.1%), lifetime anxiety (27.6%), and past-year anxiety (23.%), compared to opioid only and other illegal drug use. Zhao et al. (2017), found that women who used at least one substance while pregnant were almost five times as likely to report perinatal mental health issues compared to pregnant women with no substance use history (OR = 5.57, 95% CI, 4.57–6.80). Likewise, women who used alcohol and tobacco had significantly higher odds for having perinatal mental health issues (OR 1.71, 95% CI, 1.10–2.66; OR 2.11, 95% CI, 1.89–2.38) (Zhao et al., 2017).

In their qualitative study, Kuo et al. (2013) focused on women receiving care for drug and alcohol use disorders during pregnancy and the postpartum period and who scored a 10 or above on the Edinburgh Postnatal Depression Scale (EPDS). Women identified that they would like to have a specific treatment program tailored for women with dual diagnoses of PSU and PDA (Kuo et al., 2013). Women were reported to have motivational, family, friend, romantic, and agency concerns that either helped or hindered their recovery. However, they identified desired treatment structures such as group treatment, a safe environment, and transportation to treatment (Kuo et al., 2013).

A relationship between PSU and PDA and adverse newborn outcomes such as low birth weight was found in two studies (Shah et al., 2012; Zhao et al., 2017). In the study by Shah et al. (2012), infants who were exposed to methamphetamine had a smaller head circumference and shorter length, were less likely to be breast-fed, and more likely to have a central nervous system diagnosis. Zhao et al (2017) found that mental illness and substance use among pregnant women had a statistically significant association with neonatal birth weight (Zhao et al., 2017).

Discussion

While there is a general assumption that PSU and PDA are highly comorbid, there is little published evidence to support this assumption. Women’s mental health and substance use and dependence during pregnancy and the postpartum period is discussed by numerous researchers, but the evidence is surprisingly limited regarding the relationship between the two. The vast majority of researchers have investigated PSU and PDA separately. Moreover, PDA is often labeled as one of the symptoms of PSU and vice versa, rather than establishing and studying the two as co-occurring diagnoses. Also, PSU and PDA are often treated separately during pregnancy and the postpartum period, despite the assumption that they are highly comorbid. We accomplished the primary objectives of our scoping review to identify and evaluate the associations between PSU and PDA, and any related maternal/newborn outcomes in published studies.

Overall Relationship between PSU and PDA

Overall, our results indicate a relationship between PSU and PDA; pregnant women experiencing PSU are more likely to experience co-existing PDA, and are at greater risk for developing PDA. Likewise, women suffering from PDA are at greater risk for PSU, and more likely to be experiencing concurrent PSU. Between 20% and 60% of women who have PSU suffer from PDA, and nearly 20% of women who suffer from PDA also report PSU. One plausible explanation for the higher prevalence of PDA among women with PSU is that many researchers recruit from a population of women who are already in treatment programs for substance abuse/dependence or have already reported PSU (Benningfield et al., 2010; Holbrook & Kaltenbach, 2012; Linden et al., 2013; Metz et al., 2018; Shah et al., 2012). Recruitment from a general population of childbearing women may reveal more accurate prevalence rates for co-occurring PSU and PDA. There may be an underreporting of substance use applicable to the general population since researchers approached women who are already participating in treatment programs for substance abuse/dependence. As a result, those women who are at risk for PSU but who are not in treatment may not be included in studies creating a barrier to better understanding the true prevalence of PSU (Cook et al., 2017). Furthermore, underreporting of PSU is likely among pregnant women who may be reluctant to disclose PSU due to stigma around alcohol, tobacco, and other substance use during pregnancy. Non-compliance with societal expectations for abstinence from PSU can contribute to feelings of shame and guilt, causing women to avoid reporting any substance use (Cook et al., 2017; Silva et al., 2013; Stone, 2015). Also, possible legal consequences such as criminal charges and convictions, or child custody battles, serve as obstacles to PSU disclosure among childbearing women (Cook et al., 2017; Stone, 2015; SAMHSA, 2018). Our findings support the premise that women with PDA are at high risk for and possibly already suffering from PSU. Therefore, it is important to adequately screen and discuss PSU with pregnant women who suffer from PDA, and vice versa.

Polysubstance and Opioid Use with PDA

Perinatal substance use, especially opioid use, has received a great deal of attention in recent years due to the opioid epidemic. The number of researchers that specifically examined the effects of opioid dependency among women during pregnancy and the postpartum period has, therefore, increased (Benningfield et al., 2010; Faherty et al., 2018; Holbrook & Kaltenbach, 2012; Metz et al., 2018). Furthermore, the importance of assessing for polysubstance use is supported by the findings of our review. Metz et al. (2018) found that, among 818 pregnant women who reported nonmedical opioid use and/or other non-opioid illegal drug use other than marijuana, 36.8% used opioids only, 28.2% were opioid polysubstance users, and 35% used other illegal drugs. Further, the study findings suggested that women who used opioid and other substances had the highest risk of developing depression and anxiety, and opioid substance users had significantly more severe substance use (Metz et al., 2018). The Substance Abuse and Mental Health Services Administration (2018) stated that pregnant women who use opioids frequently misuse other substances that may worsen NAS. In studies focused on substances other than opioids, findings indicated that pregnant women are also using substances such as alcohol and tobacco. Current screening tools for substance use are designed for use in the general population and limited tools are available specifically for screening women during pregnancy, the perinatal period, and postpartum (SAMHSA, 2018). Moreover, there is a limited number of tools available to screen for polysubstance use (SAMHSA, 2018). It is important to assess for polysubstance use to identify higher risk for PDA. Developing and evaluating screening tools for polysubstance use detection is essential.

Gaps and Recommendations

The findings of our review suggest that women should be screened concurrently for both PSU and PDA during pregnancy and the postpartum period. Such an approach has the potential to reduce adverse maternal and child health outcomes. Although pregnant women have been described as highly motivated and open to adapting their behaviors in pursuit of a healthy pregnancy and baby, some women are unable to discontinue substance use due to dependence or addiction to the substance(s), perinatal depression, anxiety, and other mental health conditions (Chang, 2014; Sarkar et al., 2009).

CALLOUT 3

Recommendations from our review include prioritizing the provision of services for both PSU and PDA and identifying the needs, gaps, and barriers to providing care. Comprehensive multi-disciplinary services can address the complex healthcare needs of childbearing women with mental health conditions and substance dependency, in addition to the physical, socioeconomic, and family needs encountered during pregnancy and the postpartum period. Integration of screening and intervention for both PSU and PDA is essential, as is establishing sufficient social support during perinatal care. Further studies are needed to explore evidence-based interventions that address both PSU and PDA integration into perinatal care, as well as cost-effectiveness of care in diverse groups of childbearing women.

Limitations

One limitation of our review is the exclusion of articles published in languages other than English. The World Health Organization (2015) recognized maternal mental health as a global problem, but we have not represented viewpoints from diverse cultures. Another limitation is that not all types of drugs were represented in our review, e.g. cocaine, hallucinogens, and newer drugs (e.g., synthetic cannabinoids). Further, all studies were limited by incomplete information about amount of use and did not use comparable ways to characterize polysubstance use. Finally, we limited our review to articles published from 2010 to 2020. However, inclusion of articles published in the past ten years reflected the increasing rates of opioid use disorder among pregnant women, a disorder that has more than quadrupled from 1999 to 2014 (Haight et al., 2018).

Conclusion

Prevalence rates of PSU and PDA among childbearing women are high. We found that women with PSU have an increased risk for PDA and, likewise, women with PDA have an increased risk for PSU. Among women with PSU, polysubstance use confers the highest risk for PDA. Limited studies are available for infant outcomes, but findings indicate an association between low birth weight and both PSU and PDA. Perinatal depression, anxiety, and other mental health conditions can present a major challenge to a woman’s ability to address substance use. Further studies are needed to explore current treatment and screening gaps/needs in this population. Our review of current literature can help identify gaps in knowledge and guide future research studies to improve health outcomes for women of childbearing age and their families.

Callouts.

1. Although clinicians may assume that perinatal depression and anxiety and perinatal substance use are highly associated, there is limited evidence to document a relationship.

2. A bidirectional relationship exists between perinatal substance use and perinatal depression/anxiety; women experiencing one of these are at higher risk of experiencing the other.

3. Perinatal substance use and perinatal mental health should be addressed together to develop effective interventions and optimize maternal–newborn outcomes.