Masarwa et al. Respond to “The Disillusionment of Developmental Origins of Health and Disease (DOHaD) Epidemiology”

We thank Gilman and Hornig (1) for their interest in our meta-analysis (2) on the association between acetaminophen use during pregnancy and the risk of attention deficit hyperactivity disorder (ADHD) and autistic spectrum disorder (ASD). Their arguments raise the possibility of an interesting scientific exchange, and although we agree with some of the arguments made, we would like to clarify some of the issues that have been raised. The first issue raised is that the summary effect of the meta-analysis has no correspondence to real-world intervention. We chose to report the associations in our analysis in terms of relative effects, because these measures are substantially more stable across risk groups than absolute differences (3). In addition, relative effect sizes provide a measure of the strength of the link between exposure and outcome, an important element of assessing causality (4). For studies reporting continuous scores, we contacted the authors to request dichotomous outcome data, and when incidence data was unavailable despite these efforts, we calculated the log risk ratios from the reported measures using established methods, as reported. These effect sizes are informative and interpretable. We are confident that the medical community and policy makers understand that a moderate increase in an uncommon outcome results in a small absolute increase in that outcome. Although the effects detected were modest and should be interpreted with caution, we believe they should not be disregarded or ignored and warrant further investigation. We acknowledge the heterogeneity of the studies included in our meta-analysis. We provided a thorough and detailed qualitative and quantitative analysis of the studies, and we highlighted the observational nature of the studies, the differences in methodology, the risk of information bias and misclassification of the exposure and outcomes, and potential sources of confounding. These limitations, and the need for cautious interpretation were also clearly communicated to the media. While the pooled estimate might not represent the precise effect of the intervention, it is representative of the current available data, which indicate a small increase in neurodevelopmental outcomes. Of note, 4 additional studies indicating a significant association between acetaminophen use during pregnancy and an increased risk of ADHD have been published following the publication of our meta-analysis (5–8). Although the effects detected were modest and should be interpreted with caution, we believe that they should not be disregarded or ignored, and they warrant further investigation. The second issue raised is that there is no point in estimating causal inference for an effect that is not well defined (9). We agree that inference from observational data without a well-defined causal effect can lead to unstable predictions, despite proper statistical analyses, and we have emphasized in our limitations that a causal link might not be established (2). The third point regards the inclusion of a wide range of diagnoses and outcomes for ADHD and ASD. We agree that the pooled analyses included a wide range of outcomes that might not be completely representative of strict definitions of ADHD or ASD diagnoses. We aimed for an inclusive approach that provided a more comprehensive assessment of the available evidence, while providing a detailed account of differences in study methodology, and quantitatively addressing heterogeneity by employing random-effect models, sensitivity analyses, and meta-regression exploration of differences between study factors. Last, due to the limitations of observational studies and meta-analysis, especially when causal association is not well established, we believe that new approaches to test the validity of the results are warranted, for example, by conducting sensitivity analyses, through quantitative bias analysis for unmeasured confounding, exposure, and outcome misclassification (10, 11). Author affiliations: Centre for Clinical Epidemiology, Lady Davis Institute, Jewish General Hospital, Montreal, Quebec, Canada (Reem Masarwa); Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, Quebec, Canada (Reem Masarwa); Division of Clinical Pharmacy, Institute of Drug Research, School of Pharmacy, the Hebrew University of Jerusalem, Israel (Reem Masarwa, Amichai Perlman, Ilan Matok); and Braun School of Public Health and Community Medicine, Hebrew University-Hadassah, Jerusalem, Israel (Hagai Levine). Conflict of interest: none declared.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.