Results from the final multivariable logistic regression demonstrated EDA lacked association with PPD scores at 6 weeks postpartum, following confounder adjustment. However, a mere 3 variables were adjusted for, and though they were carefully chosen, this small number itself cannot guarantee an elimination of cofounding. Other variables with known associations for PPD, such as a lifetime history of smoking, preterm birth, and fetal demise should have also been considered. Researchers often assess the “total” effect of exposure on outcome, while ignoring possible intermediary variables and explanatory factors.
Studies such as these, hindered by the limitations of their respective data sets, often overlook complex interrelationships among pain, antenatal depression, social support, and fear of childbirth. In addition, the multifaceted interrelationships of these topics can be disregarded by analyses. In this particular Swedish cohort, the cultural implications of using EDA imply a strong fear of childbirth, indicating that the findings may not apply to other settings such as the United States, where EDA is a common form of pain management during childbirth and does not necessarily indicate fear of labor. Chinese studies which indicate strong protective relationships between EDA and PPD can also overestimate the benefits of this relationship due to third party factors which cannot be measured (ie, a higher educated cohort aware of the process of labor and subsequent risks of PPD).
Eckerdal’s study also failed to assess pain before, during, or after labor. Furthermore, as not every epidural catheter placement is successful, EDA should not be viewed as an all-or-nothing variable in terms of directly causing or preventing PPD. Other preliminary risk factors for PPD may include adverse early life events such as negative childhood events, sexual abuse, or physical abuse. These early events cause a physical sensitization of the central nervous system to later chronic pain, depression, and anxiety. Such sensitization leads to the question of whether or not EDA, a relatively brief procedure, could alter the effect of these early events on postpartum health at all.
Despite this conundrum, it does remain possible for the traumatic event of labor to serve as either a primary nidus of sensitization or as a secondary injury, which could nudge a previously susceptible individual over to depression. In this theory, the pain involved with the labor and delivery itself could be the trigger for PPD, indicating that EDA prevents the psychologic stress response and cortisol release that would otherwise initiate PPD.
Unfortunately, randomized trials are unfeasible and impractical, as randomization to no EDA or “placebo” epidurals would be unethical. Continued observation is therefore the means of gathering evidence for the relationship between EDAs and PPD. However, the most significant change should be to focus on questions regarding the role of pain itself rather than whether or not to utilize EDAs. The scientific community should focus on disseminating the results of exploratory investigations, to establish research priorities.