During peer review, a reviewer recommended that we delete the supplemental analysis because “city” is collinear with fluoride levels in municipal drinking water (which was used to derive infant fluoride intake) and therefore is a redundant variable. In the interest of full disclosure, we had included this supplemental analysis in our submission, but based on the reviewer’s recommendation, we removed the Supplemental table from the final accepted version. However, we failed to delete the sentence in the discussion that referred to this supplemental analysis. We caught this error in the proofing stage and made the correction to the final version.
For transparency, this addendum shows the Supplemental table corresponding to the sentence that was deleted between the pre-proof and final version of our manuscript. The infant fluoride intake variable is a function of city (due to the fluoridation status that differs city by city) and thus the regressors are inherently collinear. As a result, the inclusion of city in our model inflates the standard error of infant fluoride intake. We also note that our prior work used multilevel models where we accounted for city-level clustering via a random-effects model (i.e. mother-child pairs nested within cities). We showed similar effects to our multiple regression model where city was entered as a covariate (posted on OSF here: https://osf.io/7cr5e/). Thus, we found no evidence that city is a major confound on our primary effect estimates.
This addendum does not change our conclusion that the association between fluoride exposure and performance-IQ differed based on timing of exposure.
Supplementary Material
Footnotes
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.