Thread by @DBrozeLiveFree

Day 3 of phase 2 of the #FluorideLawsuit resumes with the EPA questioning of Dr. Grandjean, FAN's 3rd witness.

EPA asks Dr. Grandjean about a forest plot which was previously mentioned.

EPA has him look at the mention of Bashash 2017 study, and the confidence interval assigned.

EPA has him look at Green 2019 and the confidence interval.

EPA asks Grandjean to confirm that this means these studies conclusions were not statistically significant.

Grandjean confirms. EPA has no further questions.

FAN Connett begins redirect of expert witness Dr. Grandjean.

Connett has Grandjean look at the NTP report, specifically the risk-of-bias subgroup analysis.

Connett asks Grandjean to confirm that the studies with a high-risk-of-bias did find an association between lower IQ and fluoride.

Connett asks Grandjean to confirm that the studies with low-risk-of-bias studies also show an association between lower IQ and fluoride.

Grandjean confirms that both high risk and low risk found an association.

Connett: One of the costs/limitations of doing a means-effect meta analysis is that in some studies you may lose some of the power that the individual studies have, would you agree with that?

Grandjean: It can.

Judge Chen asks for some clarification about how some studies might "lose their power" in a means-effect analysis.

Grandjean says its bc you group all the data with elevated levels of fluoride.

Grandjean: "you lose information if you assign a group level to all of the children who have exposures over a wide range."

Judge Chen: if you have a narrow exposure range, what happens in the meta analysis?

Grandjean: you lose some of the power/knowledge.

Judge Chen is going back and forth with Dr. Grandjean to get clarity on his statements about the benefits and drawbacks of a means-effect analysis.

FAN Connett resumes direct. Is going to read a statement from an exhibit from the Board of Scientific Counselors Working Group (BSCWG).

Connett: Dr. Grandjean I am going to read you a statement from the NTP. Asks Grandjean if he affirms with the NTP's belief on group analysis vs individual analysis. He does.

Connett: can you explain the strength of the Bashash study, the strength of relying on individual urine samples vs the group samples?

Grandjean: the problem with using other fluoride excretion data is that they may not reflect the most vulnerable time of exposure.

Grandjean: if you look at the children's sample you will have a risk of imprecision involved.

Connett: so is there anything that we see in this data that conflicts Bashash 2017 regarding maternal fluoride and IQ?

Grandjean: I dont think the presentation of the fluoride data is appropriate. Connett moves on.

Connett asks Grandjean about the BMCL he derived from the square approach.

Connett: If you wanted to know the precision of what the distribution of urinary fluoride levels is in the MIREC study, would you look to the Green study, or the 2018 Till study of the entire cohort?

Grandjean: I would go with the more recent study.

Connett wants to show the last page of the EPA BMD guidance, a section titled "human data".

Connett so the first 63 pages are dealing with animal data, not human data, correct?

Grandjean: correct.

Connett reads the EPA guidance which says that "modeling human data often involves adjusting for covariants".

Connett: one of the studies EPA references for how to do a BMCL is one of your studies on mercury, correct?

Grandjean: we did that on contract with the EPA.

Connett ends his questioning.

EPA picks up for final redirect.

EPA: you have been sitting here the last 3 days listening to the trial, correct?

Grandjean: I havent heard every word but I've been here.

EPA is looking at the overall low risk-of-bias markers from the NTP meta analysis.

EPA asks Grandjean to confirm that the risk of bias is close to zero.

EPA ends questioning. Connett picks back up to say that the data point the EPA just showed him doesnt include other relevant and more recent data.

Connett ends questioning. EPA is done as well.

Judge Chen has his own questions.

Judge Chen: is it common to find a contrary studies after studies have shown toxicity from a chemical?

Grandjean says lead is the best example. Says there were studies coming out claiming that lead could not be responsible for harm to people. Says there was industry influence.

Judge Chen: but it is possible that there are high quality studies that make an opposite conclusion?

Grandjean: its entirely possible bc its all about how you consider the risk of bias.

EPA and FAN have no further questions.

Judge Chen asks for next witness. FAN Connett says it might be best to start on Monday. Judge Chen says no, lets not give up 30 minutes.

FAN Connett calls Dr. Stanley Barone, risk assessment scientist at the EPA, with the Office of Pollution Prevention and Toxics, as their next witness.

Connett: were not going to talk about fluoride today. We are going to talk about how the EPA evaluates risk evaluations under the Toxic Substances Control Act (TSCA).

Connett: its fair to say, of all the risk assessment analysts at the EPA you have the broadest knowledge of the first 10 assessments by the EPA, correct?

Barone: that sounds about right.

Connett is establishing Barone's experience and credentials as it relates to assessing neurotoxic hazards, and how this relates to EPA's TSCA assessment.

Connett: Dr. Barone, what I want to do is walk through the various steps of a hazard assessment. You agree that hazard and risk are two separate things?

Barone: Yes, they are related but hazard is part of risk, but risk takes into account other aspects.

Connett: under EPA's guidelines and risk assessment, theres no need to prove causation to prove a hazard, is that correct?

Barone: that is correct.

Connett: so when I refer to the first 10 risk analysis that were completed, I am referring to all of them but asbestos, does that make sense?

Barone: correct.

Connett: Dr. Barone, if we look through all the EPA risk evaluations, we will not see a single sentence where EPA provides a separate confidence assessment for the high dose data vs the low dose data, correct?

Barone: I am a little confused.

Connett: let's take it step by step

Judge Chen says he might be confused and asks Dr. Barone to elaborate on how the EPA makes their hazard ID and assessment.

Connett continues, asks Barone to further explain the EPA's process for assessing hazard.

Barone: generally speaking its the medium and high quality studies that are taken to dose-response.

Connett: so just to be clear, if EPA has a medium study that presents a hazard they will then go to dose-response analysis, correct?

Barone: yes.

Connett: Dr. Barone I am going to refer you to several undisputed facts. "EPA has made unreasonable risk determinations where it did not have high confidence in the data", correct?

Barone: that is correct.

Connett is showing a document relating to EPA risk assessment of Dioxane which Dr. Barone has previously seen.

Connett gives Barone a physical copy of the doc. Asks Barone if he remembers the doc.

Connett asks Barone if he remembers making the checkmarks on the document during deposition.

Connett doesnt want to put words in Barone's mouth, gives him a chance to add any contextual statements to make sure its accurate.

Barone says this doc only focused on part of the evaluation of routes of exposure.

Connett wants to get further into the doc, EPA still objects.

Connett: Let's go to the exposure assessment...

Judge Chen says this is a good time to stop.

Court adjourns for the day, will resume on Monday morning with FAN continuing to question Dr. Barone.

Thanks to everyone following along!

If you want to donate to our crowdfunding to cover costs, please go here: givesendgo.com/fluoridetrial