Breast Milk and PFASs - Part 8

The posting for a conclusion to all of this is taking me longer than I anticipated.  I'll explain why in a bit if you keep reading.

Focus: Should a new mother forgo the benefits of breastfeeding based on the findings based on the findings present in this paper:

Breastfeeding as an Exposure Pathway for Perfluorinated Alkylates

This paper made my Google Newsfeed show me news articles with titles such as this:

Breast milk may be tainted with toxic chemicals – says new research

The authors of this paper, which from here on I will call the "Breastfeeding Paper," tell us in the abstract that:

The duration of exclusive breastfeeding was associated with increases of most PFAS concentrations by up to 30% per month, with lower increases during partial breast-feeding.

That "30% per month" statement is the result of their modeling:

We first included only the three concentrations obtained at birth and at age 11 and 18 months. The model was estimated using a linear mixed model with breastfeeding variables included as covariates.

What is important to understand at this point, is that the amount of PFASs in the breast milk was never measured:

However, neither the newborn baby’s PFAS concentrations nor the concentrations in milk were measured in this study, which instead relied on the maternal serum concentrations measured at a specified point during pregnancy/

The concentration used in their model for month 0 is this:

Serum-PFAS concentrations at birth were calculated from maternal concentrations.

Month 0 is critical for their model, which uses three points for their prediction, 0 month, 11 months, and 18 months.  This is where the statement "increases of most PFAS concentrations by up to 30% per month" comes from and is shown in their prediction graphs.

Source: Breastfeeding Paper

Here is where I started to have troubles and why it is taking me a while to get these posts out.

I need to make sure that what I see is correct before I put it out for the whole world to see.  I understand I could be wrong in anything I post.  I admit that, and I will correct my mistakes if they are pointed out.  I want to be objective and fair.

I am trying to have my facts straight and my understanding of what I see as sound as I can get it before I post it for all the world to see.  So here goes.  If I am mistaken from what I post, I will apologize and I will correct it.

Let's get back to month 0 in their model.

Serum-PFAS concentrations at birth were calculated from maternal concentrations.

Where did this data come from?

A birth cohort of 656 children born in the Faroe Islands was formed during 1997−2000 and followed prospectively. A serum sample and informed consent were obtained at week 32 of pregnancy.

Hmmmm.  That's the same cohort used in the paper they cite for "vulnerability of the immune system during early development":

Serum Vaccine Antibody Concentrations in Children Exposed to Perfluorinated Compounds.

From this point on I will call this the "Antibody Paper."

The Breastfeeding Paper's authors used the same data from the same people used in the Antibody Paper.

The Breastfeeding Paper authors took from those 656 children studied this data:

During a 12-month period of the follow-up period, a subgroup of mothers was invited to bring in their children for an examination and blood tests at ages 11 months and 18 months

Remember this:

The prenatal exposure was assessed from the mother’s serum-PFAS concentrations at pregnancy week 32

The data in the Breastfeeding Paper came from 81 mothers of the cohort of 656 used in the Antibody Paper.

Can we agree that the Mother's serum PFAS concentration detected in the Antibody Paper are what is being used in the Breastfeeding Paper for month 0?  There was no additional analysis performed by the Breastfeeding Paper's authors.  The 81 mothers used in the Breastfeeding Paper were part of the 656 mothers used in the Antibody Paper.

I am pretty sure I am correct on this and I am not missing something or misreading their paper.

Here is where I started to have my doubts on the model's findings they reported.  I know better than to accept the conclusion of a paper without first looking at the data.  Statistics and squishy wording are used a lot in peer reviewed journal papers to support or elude to a finding of positive or negative outcomes.

With this in mind, what does their data say?

Let's look at Table 1 in the Breastfeeding Paper:

Source: Breastfeeding Paper

Let's look at the data for "0 month: aka: birth - in particular the serum concentrations for 80 mothers who had their serum concentrations for PFAS's measured at pregnancy week 32.

Source: Breastfeeding Paper - Excerpt from Table 1

We now need to agree that the serum concentration in Table 1 is in ng/mL.  Which means that for 80 mothers pulled from the 656 mothers used in the Antibody Paper, these 80 mothers had a geometric mean of 6.0 ng/mL of PFOS.

The authors of this paper - see note at bottom of Table 1 - use "median" for "geometric mean."  They are not the same, but that point is for another paper someone else can write about statistics.

What is important is that we agree that the Breastfeeding Paper authors are telling us that the geometric mean for PFOS in these 80 mothers was 6.0 ng/mL.

We need to also agree that the IQR reported - 5.2 and 7.2 - means that 75% of the PFOS concentrations found in these 80 mothers were no higher than 7.2 ng/mL

Source

 Still with me?

According to the Breastfeeding Paper's authors, the serum concentration for PFOS used in their model for each of the 80 mothers had a geometric mean of 6.0 ng/mL and an IQR of 5.2, 7.2 which means 75% of the PFOS found in these 80 mothers were below 7.2 ng/mL.

Those 80 mothers were pulled from the 656 mothers with serum concentrations of PFOS collected at week 32 of pregnancy in the Antibody Paper.

In order for me to be correct in what I am about to show, my observation of their data must show that a geometric mean of 6.0 ng/mL and an IQR of 5.2, 7.2 was used in their model that produced a result that allowed them to make this statement:

The duration of exclusive breastfeeding was associated with increases of most PFAS concentrations by up to 30% per month, with lower increases during partial breast-feeding.

Next post: Part 9