The association between labor epidural analgesia and postpartum depression: a systematic review and meta-analysis.

Background Previous studies have demonstrated that appropriate treatment for postoperative pain can lead to improvement in depressive symptoms, however the association between adequate intrapartum pain control and the development of postpartum depression is not clear. The purpose of the study was to examine the effects of labor epidural analgesia and postpartum depression. Methods We performed a quantitative systematic review in compliance with the PRISMA statement. We conducted a search of PubMed, Embase, the Cochrane Database of Systematic Reviews and Google Scholar databases. The primary outcome was a positive screen of postpartum depression among women who received labor epidural analgesia up to 3 months into the postpartum period. Meta-analysis was performed using the random effect model. Results Of the 148 studies available, 9 studies with 4442 patients were included in the analysis. The use of labor analgesia on positive depression screen compared to control revealed no significant effect, OR (95% CI) of 1.02 (0.62 to 1.66, P = 0.94). Conclusion Based on current literature, the use of epidural analgesia for pain relief during labor doesn’t appear to affect the likelihood of postpartum depression. Future studies are warranted to further investigate these findings and identity other possible preventative interventions that reduce postpartum depression.

Given the strong association between pain and depression, one can argue that optimal labor analgesia (e.g. use of epidural infusions of local anesthetics and opioids) may reduce the incidence of postpartum depression.
The main objective of the current study is to evaluate a possible association between optimal labor analgesia using epidurals and the development of postpartum depression. We hypothesized that patients who received epidural analgesia would develop a lower incidence of postpartum depression.

Methods
This systematic review was conducted following the guidelines of the preferred Reporting Items for Systematic Reviews (PRISMA) statement [8]. Institutional review board approval and patient consent were not required.

Systematic search
Published articles evaluating the effects of peripartum pain control via labor analgesia on postpartum depression were searched using the electronic PubMed database, Embase, the Cochrane Database of Systematic Reviews and Google Scholar up to July 2019 (Additional file 1). Key words and MeSH descriptor terms "pain", "labor", "analgesia" and "depression"' were used individually and in various combinations. An attempt to identify additional studies was made by reviewing the bibliographies from identified studies. No language restrictions were used. Unpublished studies were not investigated. There was no limitation on sample size.

Inclusion and exclusion criteria
The inclusion and exclusion criteria were defined prior to the initiation of the systematic review. We included peer-reviewed observational-prospective trials that investigated peripartum pain control via labor analgesia in women who were experiencing postpartum depression up to 12 months postpartum as defined by the Diagnostic Statistical Manual of Psychiatric disorders (DSM) [9]. We excluded studies in which woman experienced early pregnancy loss or stillbirth. Studies that did not implement the use of formal screening measure for depression or did not report specific postpartum depression assessments were omitted.

Selection of the included clinical studies and data extraction
The investigators individually assessed the abstracts and outcomes extracted from the initial query. Articles that did not meet the inclusion criteria or fulfilled the exclusion criteria were omitted. Discrepancies among the reviewers were finalized by discussion and if resolution was not met, the final decision was determined by an additional investigator who was not present or aware of the discussions held by the previous investigators.
The authors independently evaluated the full manuscripts of all potentially eligible trials and performed data extraction using a data collection form specifically developed for this review. Data explored in the literature included the study design, number of participants, inclusion and exclusion criteria for the study, sample size, number of subjects in treatment groups, type of labor analgesia administered, dose of analgesia administered, psychiatric screening tools (Edinburgh Postnatal Depression Scale, DSM diagnosis) time period of postpartum depression assessments, and the conclusion to each study. Disagreements on extracted variables were resolved by consulting with an additional investigator.

Methodological quality assessment of clinical studies
The investigators assessed the quality of the included studies using the Newcastle-Ottawa Quality Assessment Scale [10]. This tool assesses aspects regarding population/sampling methods, exposure/outcome collections, as well as statistical matching/adjustments of the data. Quality scores were assigned for each domain, and an overall score was assigned for each study, with a maximum possible score of 9. Studies with a score of ≥7 were categorized as high-quality studies, and those with a score of < 7 were categorized as low-quality studies.

Primary outcome
A positive screen of postpartum depression among women who received or did not receive labor epidural analgesia up to 3 months into the postpartum period.

Meta-analyses
For positive postpartum depression screen (dichotomous data), the Peto odds ratio (to account for the potential of zero counts in the cells for low frequency outcomes) and 99% CI are reported. A significant effect compared to control required that the 99% CI for continuous data did not include zero and for dichotomous data, the confidence interval did not include 1.0. Due to the inclusion of different study populations, we choose to use the random effects model to generalize our findings to studies not included in our meta-analysis. The random effect model has an advantage to the fixed effect model since it does not rely on the assumption that a true effect size is the same in all combined studies [11]. Publication bias was evaluated by examining for asymmetric funnel plot using the Egger's regression test [12,13]. A one-sided P < 0.05 was considered as an indication of an asymmetric funnel plot.
Of the included studies, heterogeneity was considered to be high if the I 2 statistic was greater than 50%. If heterogeneity was high, we performed a sensitivity analysis by removing individual studies according to its methodological quality and examining its effect on the overall heterogeneity. A P value < 0.05 was required to reject the null hypothesis. Analyses was performed using Stata version 15 (College Station, Texas) and Comprehensive Metaanalysis software version 3 (Biostat, Englewood, NJ).

Results
The initial search yielded 184 studies and 145 articles were omitted that did not fulfill the inclusion criteria upon further evaluation of the study abstracts. The full text of 39 articles were evaluated, and 30 articles were excluded because they did not meet the inclusion criteria. The flowchart of selected studies with specific reasons for exclusion is shown in Fig. 1.
A total of 9 studies that were published between 2014 and 2019 with 4442 women were eligible and included in the analysis (Table 1) [14 -22]. The overall incidence of a positive screen for postpartum depression in the total sample was 9.51%. All nine studies reported on postpartum depression within 3 months following delivery. All studies used the Edinburgh Postnatal Depression Scale (EPDS) as the screening tool for presence of postpartum depression. The methodological quality of included trials is presented in Table 2.

Methodological quality of studies
The quality of studies ranged from 4 to 9 points, with a median of 8 (7 to 8). There were7 studies that scored ≥7 representing high-quality with 2 studies scoring below 7 representing low quality studies. Five studies did not clearly demonstrate the presence of depression prior study enrollment. Two studies scored a zero for comparability on basis of analysis for not adjusting for known confounding variables such as past or family history of depression or an additional factor including difficulty during childbirth. The assessment of outcome was selfreported in four studies. In two studies, the adequacy of study follow-up was not clearly reported.
Heterogeneity was high, I 2 = 74%. An examination of the funnel plot did not reveal asymmetry; Eggers regression test revealed a one-sided P = 0.46. A sensitivity analysis removing the low methodological quality studies (score < 7) did not reduce the heterogeneity.

Discussion
The most important finding of our investigation was the lack of association between the use of epidural analgesia and postpartum depression. Patients who received epidural analgesia were not less likely to develop postpartum depression compared to patients who did not received epidural analgesia. Although individual studies have demonstrated a significant association between epidural analgesia and postpartum depression, the aggregated association was not significant despite a large sample size of over 4400 patients.
Our results are clinically important since optimal labor analgesia has been recently proposed as potential modifiable risk factor for postpartum depression by multiple investigators [23,24]. This hypothesis has been based on the strong association between chronic pain conditions and the development of mood disorders [14, 25,26]. It is possible that the acute nature of labor pain is not enough to lead to the development of postpartum depression.
Epidural infusions are considered the gold standard for labor analgesia and the current use of ultrasonography and other modalities to confirm epidural catheter placement has become popular among clinicians [27][28][29]. Moreover, the use of epidural infusions has been consistently demonstrated to improve the mother's experience through labor [30,31]. Nonetheless, epidural analgesia is often not continued much longer beyond child delivery and it has yet to be investigated if the use of epidural analgesia can reduce postpartum chronic pain.
Although we did not detect an association between postpartum depression and the use of labor epidural analgesia, we cannot exclude a potential benefit of epidural analgesia on the severity of mood oscillations after labor.    Labor and postpartum pain can affect the ability of mothers to bond with their infants via attachment and caregiving during that critical developmental period. It is possible that labor analgesia can reduce the incidence of postpartum blues, and this may lead to a benefit to parturient. Future studies examining labor pain and postpartum depression should include not only a dichotomous outcome, but also the specific depression score. While the question on whether labor analgesia may reduce the likelihood of postpartum depression remains unanswered, the importance of universal maternal mental and physical health screening is well established. Women should be routinely screened for intra-labor and postpartum pain and receive proper pain management, and clinicians should be aware of the impact of pain in the development of postpartum depression. Given that many risk factors for postpartum depression cannot be modified, it is worth elucidating whether pain is one of the modifiable ones so that the risks associated with intrapartum analgesia can be better weighed against its benefits.
Our study should only be interpreted within the context of its limitations. First, no universally accepted time point for PPD screening and this may have resulted in variations on the detection of postpartum depression among the studies. Second, although epidural analgesia is the gold standard to minimize labor pain, it is possible that patients have received other less effective techniques that minimized the potential effect of an epidural. Last, observational studies cannot account for unknown confounders that can only be adequately managed by randomized clinical trials. Nonetheless, it is unlikely that a future study will be able to randomize patients to epidural or non-epidural techniques given the ethical implications.

Conclusions
In summary, we did not detect an association between epidural analgesia and the development of postpartum depression. Clinicians should not advocate for the use of an epidural as a potential strategy to reduce postpartum depression to patients. In addition, clinicians should utilize proven preventive strategies to minimize the development of postpartum depression on the mother and the newborn child.