JPP Student Journal Club Commentary: Opioid Exposure and Cognitive Development: Unpacking Clinical Relevance

Assessing the impact of in utero exposure to various teratogenic substances has been a central focus in the fetal and early child development literature. Much of this work has focused on the impact of exposure on cognitive development. There is evidence to suggest that males appear to have worse cognitive outcomes than females (Alaedini, Haddadi, & Asadian, 2017; Nygaard, Moe, Slinning, & Walhovd, 2015). This finding remains after accounting for other important developmental variables, such as caregiver socioeconomic status (Moe & Slinning, 2001).

Skumlien et al. (2020) report on cognitive and language outcomes of 194 females and 184 males between the ages of 1.2 and 42.8 months following in utero exposure to opioids, alcohol, tobacco, or no substances. As a unique contribution to the literature, they have conducted an analysis with an almost pure sample of pregnant women on prescription opioids (analgesic or maintenance therapy). The authors found that there was a statistically significant reduction in language and cognitive scores in opioid exposed males compared to those with no exposure. This finding was not shown in females; however, females exposed to alcohol did show significantly lower scores than those exposed to opioids (Skumlien et al., 2020).

Skumlien et al. (2020) have shown reduced cognitive and language scores in males on a gold-standard measure of cognitive and language functioning, the Bayley Scales of Infant and Toddler Development – Third Edition (Bayley, 2006). While these findings are clear, and important to the understanding of how exposure to opioids can influence child developmental outcomes, it bears consideration of the clinical relevance of these findings. The distinction between statistically significant differences and clinically relevant differences is important.

First, as reported by the authors, the standardized mean scores of both the opioid exposed group and the nonexposed group are in the “Average” range. As clinicians, despite acknowledging the margin of error embedded in our testing, scores clearly within the Average range are not cause for concern. This understanding is important when reading the final conclusions, as it could be easy for authors without an understanding of clinical categorizations and standardized scores, to take this drop in scores as clinically relevant, when that may in fact not be the case. Furthermore, the amount of the difference between mean scores was about one third of a standard deviation which did not result in the average scores of the boys to fall into another category (e.g., Low Average).

Therefore, to capture the severity of developmental implications of in utero opioid exposure, or any other exposure, one needs to employ statistics that allow for an understanding of clinically relevant changes in developmental outcomes. The authors are commended on their discussion of this concept by outlining how even with the reduced scores, the scores still remained in the “Average” range. Thus, while a statistically significant difference was found, the difference was between two mean scores that would not be considered a clinically noteworthy finding in terms of comparative performance to a normative peer group.

Another important consideration when looking at developmental outcomes is the choice of a cross-sectional design versus a longitudinal design. The term development itself implies assessing the trajectory of a particular outcome or examining multiple time points. One limitation of the discussed study is their choice of a cross-sectional design. By looking at only one time point, the authors can only discuss one snapshot of the child’s language and cognitive outcomes. Furthermore, having a sample stemming from early infancy (1.2 months) to nearing preschool age (42.8 months), the authors are collapsing over distinct developmental phases. Future studies that separate by developmental phases and assess trajectories of these outcomes over time following opioid exposure would allow for an understanding of the long-term developmental effects of opioid exposure. Conclusions could be made about whether these children catch up over time or continue to lag below same-aged peers. Skumlien et al. (2020) mention the possibility that differences in females were not seen because they may become more apparent later in childhood. There is evidence that differences between exposed and nonexposed females increase with time (Nygaard et al., 2015).

One final caveat to contextualize the author’s findings was that there was a significant amount of missing data. When the analyses were re-run with only the full dataset, the differences disappeared. While the authors took care to report potential valid explanations to the loss of significance in the reduced dataset, the interaction between opioid exposure impacting Bayley-III scores in boys alone requires replication.

In the future, it would be beneficial to look at other maternal and child factors that may influence these findings. There is evidence that caregiver factors, such as socioeconomic status, are relevant to the study of sex differences in how prenatal polysubstance exposure affects child developmental outcomes (Moe & Slinning, 2001). While Skumlien et al. (2020) addressed the issue of maternal education, there may be other caregiver factors that could be influencing their results. Moreover, the role of child temperament or other socioemotional factors as protective factors may be of interest for future research.

Several strengths of this study are noted. The authors are commended on their strong study design, particularly their use of a gold-standard outcome measure. As clinicians, these authors understand the resources that go into administering a complex battery such as the Bayley-III, and how this may influence the ability to follow participants longitudinally. Despite the above limitations, these findings may be informative toward treatment guidelines for pregnant women undergoing treatment for opioid use. These guidelines make recommendations on appropriate maintenance therapy to improve neonatal survival and other outcomes (American College of Obstetricians and Gynecologists, 2017; Canadian Paediatric Society, 2019; Wong et al., 2011), with limited focus on longer-term developmental and social needs. There is a need for more research looking at the longer-term effects of maintenance therapy on child development longitudinally (e.g., cognition, language, socioemotional outcomes), which may inform follow-up care and future interventions for children and youth spanning different developmental phases.

Funding

H. Gennis is supported by a Social Sciences and Humanities Research Council (SSHRC) Doctoral Fellowship (752-2019-2734) and is a trainee member of Pain in Child Health: A CIHR Strategic Training Initiative. R. Pillai Riddell is funded by a Natural Sciences and Engineering Research Council (NSERC) Discovery Grant (2015-06813), the Canadian Foundation for Innovation (29908), and a Collaborative Health Research Projects grant (CHRP 538853–19) funded by NSERC, SSHRC, and the Canadian Institutes of Health Research (CIHR).

Conflicts of interest: None declared.