GREEN

Formaldehyde/Methylene Glycol

CIR EXPERT PANEL MEETING

MARCH 3-4, 2011

Memorandum

To: CIR Expert Panel and Liaisons

From: Director, CIR Subject: Formaldehyde and Methylene Glycol Date February 10, 2011

At the December meeting, the Panel agreed to reopen the safety assessment of formaldehyde in light of some new safety data and to consider uses in hair smoothing products, and to include methylene glycol. While I am authoring this memo, it was Ivan Boyer who worked tirelessly to pull together the toxicology review sections and Bart Heldreth who prepared the chemistry section. They will be principally responsible for shepherding this safety assessment from here on out. Because the focus is on a handful of issues that need to be resolved, we have approached preparation of this draft report the same way we did with triclosan. Since EPA’s National Center for Environmental Assessment has released a lengthy, 4-volume draft toxicological review of formaldehyde for external review, we have relied heavily on this comprehensive effort. The specific issues identified in December included: (1) formaldehyde and/or methylene glycol exposure from hair smoothing products; (2) nasopharyngeal cancers dose-response; and (3) hematopoietic cancers associated with formaldehyde exposure. After reviewing the draft report, the teams should determine if there is sufficient information to reach a conclusion on the safety of formaldehyde and methylene glycol as used in cosmetics. If the available information is sufficient, then a new conclusion should be reached and a rationale developed. If the available information is not sufficient to address all three issues outlined above, then the Panel should issue an Insufficient Data Announcement and list the data needs.

Panel Book Page 1

CIR History of Formaldehyde

1984 - CIR published its original safety assessment of formaldehyde, concluding that this preservative is safe for use in cosmetics if free formaldehyde was minimized, but in no case >0.2%. The Panel also said that it can’t be concluded that formaldehyde is safe in cosmetic products intended to be aerosolized. 2003 - The Panel re-reviewed formaldehyde, confirming the original conclusion. That finding was published in the International Journal of Toxicology in 2006. 2010 - U.S. EPA National Center for Environmental Assessment (NCEA) released a lengthy, 4-volume draft toxicological review of formaldehyde for external review on 2 June 2010 - FDA asked CIR to consider the safety of formaldehyde given its detection in hair smoothing products, to consider additional data, and to address the safety of methylene glycol in cosmetics. The Personal Care Products Council and the Professional Beauty Association have supported such an effort - at the December meeting, the CIR Expert Panel agreed to reopen the safety assessment of formaldehyde to address (1) formaldehyde and/or methylene glycol exposure from hair smoothing products; (2) nasopharyngeal cancer dose-response; and (3) hematopoietic cancers associated with formaldehyde exposure

Panel Book Page 2

1

Literature Search on Formaldehyde

Studies were identified primarily from the 2 June 2010 U.S. Environmental Protection Agency (U.S. EPA) Toxicological Review of Formaldehyde – Inhalation Assessment, external review draft. Supplemental searches of PubMed, U.S. EPA’s Integrated Risk Assessment Information System (IRIS), Oak ridge National Laboratory’s Risk Assessment Information System (RAIS), and the Agency for Toxic Substances and Disease Registry (ATSDR) website were also conducted between December 3, 2010 and 4 February, 2011 to obtain the most recent information.

Panel Book Page 3

1

Full Panel discussion – December 2011

DR. MARKS: So, there was a request that we have an early second re-review of formaldehyde. And this was triggered by concerns about adverse reactions to a cosmetic product that contained formaldehyde in quantities much above what had been considered safe. That product is known as Brazilian Blowout Solution and contains formaldehyde at 42 times the recommended safe limit. And that safe limit, in 1984, was published and then was in the re-review in 2006, was reaffirmed by an independent nonprofit body of scientific and medical experts that the assessment of the safety ingredients used in cosmetics in the U.S. last reviewed the use of formaldehyde in beauty products and that conclusion was that it was safe to use formaldehyde not to exceed 0.2 percent. That body, of course, is the CIR Expert Panel. So, we have the problem of this new product creating a number of adverse effects, but also it gave us the opportunity in relooking at formaldehyde in this level to include another ingredient, methylene glycol, which just appeared in the Cosmetic Ingredient Dictionary this year. And with those facts, our team moves that we reopen formaldehyde to relook at its safety assessment and add methylene glycol. DR. BERGFELD: And that's a motion? DR. MARKS: Yes. DR. BERGFELD: Can, Paul, you speak for the Belsito team? DR. SNYDER: Yes. We were conflicted as to whether to reopen this and add in methylene glycol or to not reopen this and then just develop methylene glycol as a separate document. So, I think we can agree to that motion to proceed to reopen and see how the document becomes formulated as it transpires into basically a new document with a new ingredient and with a new use category, the hair straightener. DR. BERGFELD: Any further discussion? John Bailey? DR. BAILEY: I think as we talked about yesterday, it would be really important in reopening this, number one, that the scope of thereview be focused on these new uses. Certainly the products that seem to be on the market now are not encompassed by the current CIR safety assessment. These are new uses and that these new uses should be clearly considered. So,I think that's an appropriate step. The other is, I think that it's really important to include a thorough chemistry discussion of formaldehyde, formal- and paraformaldehyde, methelyene glycol, all the different forms of formaldehyde so that these are defined and become available for others to consider, not just within the scope of this review, but certainly outside of this review.

Panel Book Page 4

2

Because there's a lot of confusion about, you know, what's used, how they should be labeled, what the terms are, and so forth. So, I think this is a good exercise, but it should be focused pretty much on the new uses. You know, I don't see anything that would change the previous conclusion of.2 percent being a safe level. Certainly that's something you can always consider, but to my way of thinking this should be narrowly focused on those uses. DR. BERGFELD: I wonder if I could ask Linda Katz to comment and then Rachel? DR. KATZ: I would agree with John's comment, that I think it's important to go back and review formaldehyde as a new use and to extend it to include all of the rest of the chemicals that could become formaldehyde given the situation where the -- I guess as the product is actually used. And I think, depending upon what happens there, you may need to go back and reexamine whether the original conclusion was appropriate. But at this point in time, there's no reason to suggest that the original level for safe is not safe, but the new data will help to at least go back and assess whether anything along the original conclusion needs to be modified, too. DR. BERGFELD: Rachel? MS. WEINTRAUB: I support this course of action, too. I think this also illustrates something that the panel has discussed sort of as a sidebar for many years and that is what happened when there is a product that contains ingredients and is used in a way that is not described in our safety assessment, both in the way in which it's used and the level as well. So, I think this is an important issue to address. DR. BERGFELD: Thank you. So, the motion's been made to reopen and it's been seconded. Any further discussion? DR. MARKS: And with the intent to add methylene glycol, but as we work through the report, we'll decide whether or not we want to continue that. DR. BERGFELD: All right. Call for the question, all those in favor, please indicate by raising your hand? Thank you. Unanimous.

Panel Book Page 5

3

Marks Team Discussion

DR. MARKS: Next is formaldehyde rereview. This is Buff Book 1, the memo. The one that has the first is the memo. DR. SLAGA: Off the record can someone explain to me how it got its name Brazilian Blowout? DR. MARKS: That's not an off the record. The comment was how did it get the name Brazilian Blowout. What I was wondering is this manufactured in California where the company is based or is it manufactured in Brazil? MR. STEINBERG: It's manufactured in Brazil. DR. ANDERSEN: David, come on up and use the microphone, please. MR. STEINBERG: The products originated in Brazil. I believe right now about 95 percent are still manufactured in Brazil and imported into the United States. The term Brazilian Blowout is a trade name of one of the companies. At Cosmoprof in Las Vegas in July I guess of this year there were over 40 companies selling these types of products under various names. DR. MARKS: Thank you, David. That's a little diversion. We didn't get the introduction but Tom wanted to get the background on Brazilian Blowout. I'm not sure what you were imagining, Tom. I'm not going there. At any rate, this is an early second re-review of formaldehyde. In 1984 the panel published an original safety assessment of formaldehyde concluding that this preservative was safe for use in cosmetics, that free formaldehyde was minimized but in no case greater than 0.2 percent. Then it was re-reviewed in 2003 and that original conclusion was confirmed and then the findings were published in 2006. Why are we looking at this as an early second rereview? For those who are at the edge of your seats it's because of Brazilian Blowout which in Canada has been pulled off their market. In their advisory release in October 26 of this year they found the levels of formaldehyde were markedly higher, 42 times higher than the.02 percent which was the accepted level recommended by the CIR. There were a number of reports of toxicologic effects, burning eyes, nose, throat, breathing difficulties, even hair loss which was interesting. Then there was a statement by the FDA indicating that the FDA was looking into these complaints as there had been complaints received here in the United States. And in a news release by our own John Bailey indicating that an independent nonprofit body of scientific and medical experts who assesses the safety ingredients used in cosmetics and the manufacture of cosmetic products had reviewed formaldehyde and concluded the previous conclusion I mentioned. I like the way you've stated that. At any rate, I think we're tasked today with the question of reopening this formaldehyde safety assessment. In addition to that issue whether we reopen or not would be do we include methylene glycol also in the safety assessment so that that would be an add-on and it should be a no- brainer if we add that on. I'll ask the panel to comment on that. DR. SLAGA: I would say we reopen and add methylene glycol. It still would have the

Panel Book Page 6

4

same conclusion. DR. MARKS: Tom, you would reopen and add methylene glycol. Ron? DR. SHANK: I agree to reopen to add methylene glycol, but I don't think the conclusion is going to change. DR. BAILEY: I'd add something. These are products that are intended mostly for salon use by trained professionals so that there is that element. There is also an issue relative to whether or not the products are somehow offered for use or being encouraged to be used by nonprofessionals at home so that there are a number of aspects to the safety assessment. Also I think it's pretty clear that these products do not fall as they're formulated within the current CIR conclusion. In other words, it was not a part of the original assessment so that I can see a rationale for reopening and considering this use. The issue of methylene glycol and all the forms of formaldehyde if you want to call them forms is something that I would be extremely important for CIR staff and for the experts to provide a clear, concise description of the chemistry here because you've have formaldehyde gas which we all know about, you have methylene glycol which is hydrated formaldehyde, you have formalin and you have paraformaldehyde. You have a lot of terms being thrown around so I think that it would be extremely important not only for the safety assessment but for the public in general to come up with a very thorough and concise description of the chemistry definitions of terms and how those terms actually relate amongst each other to help clarify this issue. I think it would be a great service to everyone who is interested in this to do that. DR. ANDERSEN: Amen. DR. HILL: I don't know if there are any other formaldehyde-related ingredients in the dictionary, but if there were things like paraformaldehyde, for example, make sure that the dictionary gets combed for anything else that's effectively formaldehyde in disguise. 18 DR. BAILEY: I think that's exactly what it should be. That should not extend though into the so- called formaldehyde-releasing preservatives which we've already looked at, but all of the different forms or polyformaldehyde or whatever they are really need to be included in this. Again, it would be great to have this presented in a way that is very easy to understand the distinction and relevance to cosmetic formulation. DR. MARKS: I thought Scheme 1 on page 1 of the book, the 2/10 rereview of formaldehyde, was quite nice in detailing and probably the only thing I might include, I looked back and forth in terms of formalin which is the way we're really testing all of this is put the percentages there like 50 percent water, about 40 percent methylene glycol and whatever else is in formalin, Ron and Tom, when you said you expect the safety assessments to be the same, that all the formalin-based tests could be used to validate the safety of the methylene glycol. DR. SLAGA: Yes. Formalin in fixing tissue and knowing the process by which that

Panel Book Page 7

5

occurs tells me why it would straighten hair and cause hair to fall out too. DR. MARKS: I'll move that we reopen and that we add methylene glycol. Again for the minutes, methylene glycol was not included in the "Cosmetic Ingredient Dictionary and Handbook" until this year so that this is really an opportunity to move forward with that ingredient also. Somewhere in here I had noted for Alan to comment on the question of using QRAs, quantitative risk assessments, also in evaluating these compounds. DR. ANDERSEN: When we reopen it, we'll lay of the science out on the table in one full document and you'll have all of your options. DR. MARKS: Tomorrow I'm going to move that we reopen the formaldehyde assessment with the intent of adding methylene glycol and with the expectation that the conclusion is going to be similar.

Panel Book Page 8

6

Belsito Team Discussion

DR. BELSITO: Okay. So, formaldehyde. Next is the formaldehyde re-review. And, this was sort of prompted by a product -- a Brazilian hair smoothing product that states that it doesn't contain formaldehyde, but rather contains methylene glycol. And the way the product is used is, it's heated and then formaldehyde is released. And methylene glycol didn't used to be in the cosmetic dictionary, but it now is. And so the question is, based upon the information in the report on formaldehyde, do we need to reopen it? And I thought, no. Because there was really no new data on formaldehyde, is my opinion. And the restrictions that we gave before with 0.2 parts per million, I think, were more than adequate. But do we want to reopen it to add methylene glycol, which is now a cosmetic ingredient and there's an equilibrium between the formaldehyde and methylene glycol that would need to be addressed in that equation? So. MR. SNYDER: So the primary reason to reopen would be to include methylene glycol? DR. BELSITO: Glycol, correct. In my viewpoint. I mean, I didn't see anything new in the data I reviewed that changed our conclusion. MR. SNYDER: Then the new use – there is the new use straightener in which heat is applied, right? DR. BELSITO: No, no, no. But it's not labeled as formaldehyde. It's labeled as methylene glycol. So the new use is for methylene glycol. You read the MSDS sheet on the product it does not list formaldehyde as an ingredient. It lists methylene glycol. DR. BERGFELD: Even though it has 10 percent in it? DR. BELSITO: It has 10 percent. So the reason to reopen would be to add methylene glycol, in my estimation – DR. BERGFELD: It'd be timely. DR. BELSITO: Well, the product has caused problems. And my understanding is that it's going under the radar of regulation right now because it doesn't, theoretically, violate the formaldehyde regulations that we put down. While in actuality, if you read the levels being released, it does. So, it gets used. MR. LIEBLER: So we're hamstrung by a technicality in that if you put formaldehyde in water -- if you immediately have a mixture of formaldehyde with methylene glycol. However (inaudible) you have a mixture of formaldehyde and methylene glycol because of the solvencies of the (inaudible). So, to say they have methylene glycol rather than formaldehyde might be correct under -- in a theoretical circumstance. But as soon as you put this into the environment, it releases formaldehyde. DR. BELSITO: Right.

Panel Book Page 9

7

MR. LIEBLER: So. DR. ANDERSEN: I thought if we wanted to tackle the issue, that there is the rationale of adding methylene glycol, which is a perfectly reasonable thing to do given that it's a new chemical in the dictionary and is used in this product. Health Canada, the OSHA folks in Oregon --- lots of people --- have flagged that the formaldehyde in these products is of concern. And then the last piece is that there are data on formaldehyde hematopoietic cancers that aren't in the current formaldehyde document, or in the rearview. Those are new pieces of data that, as far as the world is concerned, CIR missed them. And we could fix that be covering them in a new re-review. DR. BELSITO: But those data are very -- MR. SNYDER: Weak. DR. BELSITO: Weak is a kind word. DR. EISENMANN: There're also new animal studies in them. MR. SNYDER: But we didn't get that data DR. EISENMANN: No -- MR. SNYDER: So that table is empty. MR. KLAASSEN: There's an excellent animal study by Jim Svendberg done in 2009 where he gave an isotope -- a deuterium isotope of formaldehyde and found out that all of the adducts are not from the -- the few adducts that you find "inside of the animal" and the blood cells does 16 not come from what you're exposed to. So, it doesn't appear -- which is good – that there is strong evidence to suggest – MR. SNYDER: So you have them whether you're exposed or not, basically? MR. KLAASSEN: Right. MR. SNYDER: You'd be at risk. MR. KLAASSEN: On the nose, it's a different story. DR. BELSITO: No, I mean, you know me. I think that -- I mean, it's one of those things where there's this product -- I mean, I'm fine with reopening it for methylene glycol. I didn't think, though, that we needed to reopen it for formaldehyde, even with that hematopoietic data that could be, you know, addressed and that we looked at it and it's weak, to use a kind word. And, didn't change our opinion.

Panel Book Page 10

8

But, you know, I think this new product -- it's, you know, it's creating problems. And I'm fine with reopening it for that reason. But not for the formaldehyde basis. MR. LIEBLER: Same here. DR. BERGFELD: Well, I think it's timely. No one else is covering it. The procedures that we have, we could reopen it looking at the methylene glycol -- but I would like to draw your attention to the original document, on page 178, that you have. Under discussion, last paragraph, we referred to the possible carcinogenic potential. And it's not that we weren't looking at that. DR. BELSITO: No. We just didn't look at it from hematopoietic standpoint. But, there's really.2 part per million limit, and we said it's reconfirmed by all the new data we see. I mean, the formaldehyde document is fine. It's just we're adding methylene glycol. DR. BERGFELD: Could I just ask the panel a question? So, from the get go, the methylene glycol was there. Am I saying that correctly? So we just didn't have that in our information piece on the chemistry? MR. LIEBLER: When you did formaldehyde? DR. BERGFELD: Yes. MR. LIEBLER: It would be – methylene glycol would be present if formaldehyde goes into methanol. DR. BERGFELD: Okay. MR. LIEBLER: It says (inaudible) products with methyl reacts with it, right? Oh, methylene glycol, I'm sorry. That's what you get when it reacts with water, yes. So, yes. It's always there. DR. BERGFELD: It's always there. So were we just didn't have that information -- and 8 the chemistry part of it. Because I actually think we need – DR. BELSITO: There's some information on it, but we restricted formaldehyde to such a low level that it really wasn't an issue. What's happened now is this company is using methylene glycol at 10 percent. So, it's -- and the issue isn't necessarily the methylene glycol, it's the formaldehyde that's being released when you put this on the hair and then you heat it up and it vaporizes and people's eyes burn and they choke and they're having all of these other issues. Not to mention potential sensitization to formaldehyde with such high levels of formaldehyde in their skin. DR. ANDERSEN: Sorting out whether it's actually methylene glycol that's problematic, or the conversion to formaldehyde in these products is problematic -- getting that resolved -- because I'm not sure I know what the answer to that is. But I sure would like to find out. And I think

Panel Book Page 11

9

that's part of why we need to pursue it. If one reads the original formaldehyde safety assessment carefully enough, it's hard to avoid noting that, in aqueous solution the dominant form of formaldehyde is methylene glycol. You know, we've said that back in the '80s. But methylene glycol was not a cosmetic ingredient back in the '80s and now. And it's time to add it to the safety assessment.

Panel Book Page 12

1

Draft Amended Report

Formaldehyde and Methylene Glycol

March 4, 2011

The 2011 Cosmetic Ingredient Review Expert Panel members are: Chairman, Wilma F. Bergfeld, M.D., F.A.C.P.; Donald V. Belsito, M.D.; Ronald A. Hill, Ph.D.; Curtis D. Klaassen, Ph.D.; Daniel C. Liebler, Ph.D.; James G. Marks, Jr., M.D., Ronald C. Shank, Ph.D.; Thomas J. Slaga, Ph.D.; and Paul W. Snyder, D.V.M., Ph.D. This report was prepared by Ivan J. Boyer, Ph.D., D.A.B.T, and Bart A. Heldreth, Ph.D. The CIR Director is F. Alan Andersen, Ph.D.

© Cosmetic Ingredient Review

1101 17th Street, NW, Suite 310 " Washington, DC 20036-4702 " ph 202.331.0651 " fax 202.331.0088 " cirinfo@cir-safety.org www.cir-safety.org

Panel Book Page 13

2

Introduction

In 1984, CIR published its original safety assessment of formaldehyde1, concluding that this preservative is safe for use in cosmetics if free formaldehyde was minimized, but in no case > 0.2%. The Panel also said that it can’t be concluded that formaldehyde is safe in cosmetic products intended to be aerosolized. This safety assessment acknowledged the equilibrium relationship between formaldehyde and methylene glycol, but did not address the safety of methylene glycol (which was NOT listed as a cosmetic ingredient at the time). The Panel re-reviewed formaldehyde in 2003 confirming the original conclusion.2 That finding was published in 2006.3 As best we can determine, methylene glycol was not yet listed as a cosmetic ingredient. Recently, an Oregon OSHA laboratory, following up on a salon worker complaint, measured the formaldehyde levels in Brazilian Blowout, one of the current popular salon Brazilian hair treatments. They found a range of values from 8 – 10%. Health Canada is working to stop distribution and use of Brazilian Blowout in Canada, predicated on their validated testing that found 8.4% formaldehyde in Brazilian Blowout. Recognizing that methylene glycol now is listed as a cosmetic ingredient, the U.S. Food and Drug Administration and the Professional Beauty Association asked CIR to reconsider the safety of formaldehyde and to address the safety of methylene glycol. The Personal Care Products Council supported such action. The CIR Expert Panel determined to reopen the safety assessment of formaldehyde to address issues related to new uses and to add methylene glycol. In addition to the issues related to new uses, the U.S. EPA National Center for Environmental Assessment (NCEA) released a lengthy, 4-volume draft toxicological review of formaldehyde for external review on 2 June 2010, including interagency comments on an earlier draft of the document (http://www.epa.gov/IRIS/). In particular, Volume II – Hazard Characterization – of this document provides a comprehensive summary of the toxicological literature, including both human and animal studies and all of the major exposure routes of concern (inhalation, ingestion, and skin contact). The toxicological information summarized in this report is from studies identified primarily in the external review draft document. Much of the significant new toxicology data are related to genotoxicity, carcinogenicity, and reproductive and developmental toxicity. Report structure Because this report focuses on targeted questions about the relationship between formaldehyde and methylene glycol and the use of these ingredients in new product categories, it departs from the approaches that CIR has used in the past to initiate a safety assessment. Accordingly, a brief overview of what is included in each section is provided.

Section I addresses the relevant issues for formaldehyde and methylene glycol as used in cosmetics. There is an opportunity to expand this section as appropriate.

Section II presents the highly interrelated chemistry of formaldehyde and methylene glycol. While certain data are identified as exposure to formaldehyde gas, other studies appear to involve formaldehyde in an aqueous solution (some identified as formalin) which implies the presence of methylene glycol.

Section III provides information on the extent of use of these two ingredients in cosmetics based on information provided by industry to the FDA’s Voluntary Cosmetic Ingredient Registration Program (VCRP). Use concentrations provided by the Personal Care Products Council (Council) also are provided. While there appear to be a large number of hair smoothing products on the market, relatively little information on these uses has been captured at this point.

Section IV provides an overview of data from selected reports not covered in previous reviews, including skin irritation/sensitization, genotoxicity, carcinogenicity, and reproductive/developmental toxicity. Also included in this section are the putative modes of action.

Panel Book Page 14

3

Section V discusses regulatory guidance values and other limits.

I. Issues to be resolved in safety substantiation of formaldehyde and methylene glycol as used in cosmetics.

1. Formaldehyde and/or methylene glycol exposure.

Issues: Formaldehyde functions as a cosmetic biocide, denaturant, or preservative in cosmetics. Methylene glycol functions as an artificial nail builder in cosmetics. At what concentrations are formaldehyde and/or methylene glycol used in hair smoothing products and what is the function? At what concentration is methylene glycol used in nail products? What human exposures may result from the use of formaldehyde and/or methylene glycol in hair smoothing products?

2 Nasopharyngeal cancers.

Issue: Data demonstrate that nasopharyngeal cancers are produced by formaldehyde gas, but that this is a threshold effect. Is that interpretation of the data still valid and what are the implications for the use of formaldehyde as a preservative in cosmetic formulations, for methylene glycol in nail products, and for formaldehyde and/or methylene glycol in hair smoothing products?

3. Hematopoietic cancers.

Issue: Epidemiology studies have suggested a link between exposure to formaldehyde and leukemia. Is this a linear, non-threshold effect, is there a reliable mode of action consistent with the findings, and what are the implications for the use of formaldehyde as a preservative in cosmetic formulations, for methylene glycol in nail products, and for formaldehyde and/or methylene glycol in hair smoothing products?

II. Chemistry Formaldehyde, a gas, is not commercially available, but is instead produced as a related solution called formalin.4 Formalin is industrially produced from methanol. In a first step, a mixture of vaporized methanol and steam is passed over a catalyst bed, where the methanol is oxidized to formaldehyde gas. Since this reaction is highly exothermic, the gas stream is cooled directly after passing over the catalyst to prevent thermal decomposition. In a second step, the formaldehyde is reacted with water in an absorption column, because formaldehyde in its pure, gaseous form is highly unstable. Formaldehyde quickly reacts with water to synthesize methylene glycol and, without a polymerization inhibitor (e.g., methanol), polymethylene glycols via a series of reversible reactions (Scheme 1). Scheme 1

H H

O

H

O

H

water

HO

OH

HH

methylene glycol

HO O

H

11-100

paraformaldehyde

withoutmethanol

HO O

H

2-10

multiple methylene glycols

heat pushesthese equilibriumstowards formaldehyde

formaldehyde

Methylene glycol, as a pure and separate substance, is not commercially available, but is instead produced as a related solution called formalin, as denoted above for formaldehyde. Methylene glycol is a geminal (gem) diol, or a

Panel Book Page 15

4

diol with both hydroxyl groups on the same carbon. Gem diols are typically unstable compounds. Indeed, outside of an aqueous solution, methylene glycol does not exist. The apparent good solubility of formaldehyde in water is actually the good solubility of methylene glycol in water and the capacity of the solution to accommodate small polymethylene glycols (i.e., two to ten methylene glycol units long).5 Formaldehyde itself is only sparingly soluble in water. The rate of the hydration reaction is very fast (i.e., the half-life of formaldehyde in water is 70 ms) and the equilibrium between methylene glycol and formaldehyde strongly favors methylene glycol, at room temperature.6 However, since this equilibrium is temperature and density dependent, a formulation that is of a higher density and/or is subjected to higher temperatures is likely to shift favoritism towards non-hydrated formaldehyde. The dehydration of methylene glycol to formaldehyde is also very fast. The equilibrium formation rate of the higher polymethylene glycols is much slower than the rates of hydration and dehydration, and can be inhibited by the addition of a small amount of methanol. Accordingly, an average solution of formalin consists of water (~40-60%), methylene glycol (~40%), methanol (~1-10%), small methylene glycols (e.g., dimers and trimers; ~1%), and a very small amount of formaldehyde (~0.02-0.1%). All of the components of formalin are in a series of equilibriums that favor methylene glycol at room temperature.7 However, removal of water, increase in density, the addition of heat, reduction of pH, and/or the reaction of the small amount of free formaldehyde in the solution will drive the equilibrium back towards formaldehyde.8 Accordingly, a product application process, wherein a formalin containing formulation is dried, concentrated, heated, acidified, and/or applied to a formaldehyde reactive substrate, could potentially lead to the shift of these equilibriums towards free formaldehyde.

III. Cosmetics Use As given in the International Cosmetic Ingredient Dictionary and Handbook (INCI Dictionary),9 the cosmetic functions of formaldehyde are: cosmetic biocide, denaturant, and preservative. In the FDA’s Voluntary Cosmetic Registration Program (VCRP),10 there are 58 uses of formaldehyde and 20 uses of formaldehyde solution (formalin) reported. Since these all are probably the same ingredient as added to cosmetics, they are combined in Table 1a. Formaldehyde and formalin are listed separately in Table 1b, prepared using the new use table format. From a high of 805 reported uses of formaldehyde/formalin in 1984, VCRP data from 2001/2002, 2006/2007, and 2009/2010 show that uses have leveled-off to less than 100 uses as shown in Figure 1. According to the 2010 13th Edition of the INCI Dictionary, methylene glycol is reported to function as an artificial nail builder.9 Methylene glycol is not reported to FDA to be used in cosmetics according to the VCRP database. Current usage reports of formaldehyde (336 uses), formalin (15 uses), and methylene glycol (2 uses) in Canada are given in Table 2 as a function of duration of use (leave-on vs. rinse off) and exposure type (eye area, nail, etc.).11 The MSDS provided by Brazilian Blowout for their salon product, however, does include methylene glycol. The list of ingredients provided by the manufacturer is shown in Table 3, with methylene glycol listed at <5.0%. Although the purpose and mechanism of action of formaldehyde/methylene glycol in hair relaxers/straighteners is not well documented, formaldehyde (as part of a formalin solution) is known to induce a fixative action on proteins (e.g., keratin).12 Purportedly, formaldehyde/methylene glycol hair straightening formulations, such as Brazilian or keratin based straightening products, are effective at maintaining straightened hair via an amino acid crosslinking mechanism, both intra-crosslinking within the hair strand and inter-crosslinking between the hair strand and the added keratin from the formulation.13 Formaldehyde is extremely reactive and, among a multitude of potential reactions, can react with protein residue sidechains of arginine, lysine, tyrosine, tryptophan, histidine, and cysteine/cystine.14 Additionally, the primary amides, glutamine and asparagine, are known to be capable of reaction with formaldehyde. Some of these reactions can be bi-functional as well as mono-functional. In addition to simple methylene crosslinkages (i.e., -CH2-), formaldehyde has a known propensity for self-condensation so that polymethylene glycol crosslinkages (i.e., -(OCH2)n-) are feasible. Besides proteins, formaldehyde is known to react with other biological molecules such as glycoproteins, nucleic acids, and polysaccharides.15 The action of

Panel Book Page 16

5

formaldehyde in intramolecular and intermolecular crosslinking of macromolecules can considerably alter the physical characteristics of the reacted substrates.

IV. Data from Selected Reports not Covered in Previous Reviews2 The U.S. EPA National Center for Environmental Assessment (NCEA) released a lengthy, 4-volume draft toxicological review of formaldehyde for external review on 2 June 2010, including interagency comments on an earlier draft of the document.16 U.S. EPA is conducting this assessment to support the development of new chronic inhalation toxicity values for formaldehyde. Ultimately, the final versions of these values will be incorporated into U.S. EPA’s Integrated Risk Information System (IRIS). In particular, Volume II – Hazard Characterization – of this document provides a comprehensive summary of the toxicological literature, including both human and animal studies and all of the major exposure routes of concern (inhalation, ingestion, and skin contact). The toxicological information summarized below is from studies identified primarily in the external review draft document. Much of the significant new toxicology data are related to genotoxicity, carcinogenicity, and reproductive and developmental toxicity. A few older reports addressing skin irritation/sensitization are also summarized below. In addition, several tables summarizing relevant data, largely from U.S. EPA’s draft assessment, are provided (attached) to facilitate review. Other items of note from the 2003 re-review were the two Danish product surveys in which the formaldehyde levels were measured. In the first survey of 84 shampoos and skin creams, 8 products contained formaldehyde at levels above 0.05%. In the second survey of 67 skin creams, 22 products were found to contain formaldehyde, 18 of which had levels <0.003%. The author suggested that these 18 with low levels may represent products in which formaldehyde was not intentionally used in the formulation. The author did not report whether methylene glycol was listed as an ingredient. Skin irritancy/sensitization Wahlberg (1993) applied 0.1 ml of 1%, 3%, or 10% formalin diluted in water to the shaved flanks of Hartley guinea pigs with a cotton-tipped applicator once a day for 10 days.17 The skin was not occluded. They visually scored the animals for erythema and edema, and measured skin-fold thickness using Harpenden calipers. The diluted formalin solutions induced a dose-dependent increase in skin-fold thickness, with shorter latencies at higher concentrations. For example, erythema was observed on treatment day 6 for 1%, day 5 for 3%, and day 2 for 10% formalin solution. Lee et al. (1984) exposed English smooth-haired guinea pigs topically to 100 µl 37% w/v formalin applied to shaved depilated dorsal skin once/day for 2 days (total dose = 74 µg), 25µl formalin dissolved in saline applied once to a 15-mm area of the dorsal skin (total dose not reported), or by inhalation of 6 ppm or 10 ppm formaldehyde 6 hours/day or inhalation of 10 ppm 8 hours/day for 5 days.18 They tested all of the animals for signs of contact sensitivity by applying 20 ml formalin diluted in saline (concentration not specified) over a 15-mm area of shaved dorsal skin. The sites were visually inspected for erythema 1, 6, 24, and 48 hours later, and reactions were scored. These authors reported that all of the dermally-treated guinea pigs exhibited contact sensitivity, with scores increasing in a dose-dependent manner.18 Of the animals treated via inhalation, only 2 of 4 guinea pigs tested on day 31 exhibited signs of contact sensitivity (mild) after 10 ppm formaldehyde 8 hours/day for 5 days. No contact sensitivity was observed in any of the control groups. Arts et al. (1997) applied various concentration of formaldehyde in raffinated olive oil to the dorsum of both ears of female Wistar rats (low IgE-responders) and BN rats (high IgE responders) on days 0, 1, and 2 of the study.19 They then used a local lymph node assay (LLNA) to measure response to the treatment. Briefly, they first injected (i.p.) bromo-deoxyuridine (BrdU) on day 5 and euthanized the rats. Next, ear-draining lymph nodes were collected, fixed, and sectioned. Mitotic activity was monitored following successive incubation of the sections in anti-BrdU, biotin-labeled rabbit anti-mouse antibody, peroxidase-conjugated streptavidin, and 3,3-diaminobenzidine tetrahydrochloride. The authors reported an increase in the weights of the lymph nodes and a dose-related increase in the proliferation (BrdU positive) of paracortical cells, in both rat strains in response to formaldehyde treatment. They found no statistically significant increase in serum IgE concentrations in either strain. Genotoxicity

Panel Book Page 17

6

Non-Human (in vivo) Clear evidence of mutagenicity does not emerge from animal bioassays, despite the reactivity and mutagenicity demonstrated in isolated mammalian cells (see Table 4-83,16 attached). Im et al. (2006) reported a dose-dependent increase in Olive tail moments (Olive TM) in blood lymphocytes from male Sprague-Dawley rats exposed to 0 ppm (1.24 ± 0.04), 5 ppm (1.72 ± 0.11; p=0.0019) , or 10 ppm (2.16 ± 0.14; p=0.0001) formaldehyde via inhalation 6 hours/day, 5 days per week, for two weeks, using the Comet assay.20 They reported similar results for liver cells. Using the same exposure and assay protocols, Sul et al (2007) observed a dose-dependent increase in tail moments in lung tissue in male Sprague-Dawley rats exposed to 0 ppm (0.75 ± 0.07), 5 ppm (1.11 ± 0.17; p<0.05) , or 10 ppm (1.32 ± 0.34; p<0.05).21 In a critical review, Speit (2006) noted several issues with such studies, including the observation that the formation of DNA-protein crosslinks (DPCs) and DNA-DNA crosslinks (DDCs) in the cells should have reduced, rather than increased, DNA migration in the Comet assays conducted.22 Speit et al. (2009) exposed groups of 6 F344 rats to 0, 0.5, 1, 2, 6, 10 and 15 ppm formaldehyde by whole-body inhalation 6 hours/day, 5 days/week for 4 weeks.23 They obtained peripheral blood samples from each rat at the end of the exposure period by puncturing the retro-orbital venous plexus at the end of the exposure period. They collected the blood samples in a randomized sequence, and the samples were coded by sequence number for blind evaluation. These authors conducted Comet, sister chromatid exchange (SCE), and micronucleus (MN) tests to evaluate the lymphocytes from each rat. They modified the Comet assay to include analysis both before and after irradiating the samples (2 Gy γ) to increase sensitivity for detecting DNA-protein crosslinks (DPCs). Positive controls included 6 rats treated with a single 50mg/kg dose of methyl methanesulfonate (MMS), and 6 rats treated with two 10mg/kg cyclophosphamide (CP) doses, orally, before collecting blood. These authors reported no statistically significant differences between the formaldehyde exposed and negative control groups in any of the parameters examined. In contrast, statistically significant effects were found in the positive controls (MMS and CP), demonstrating the sensitivity of the tests. Human (in vivo) Ye et al. (2005) measured formaldehyde exposures and the frequencies of MN in nasal mucosa cells and SCEs in peripheral lymphocytes, and evaluated lymphocyte subsets collected from 10 non-smoking workers at a formaldehyde manufacturing plant China (average exposure duration 8.6 years, ranging from 1 to 15 years).24 They also exposed 16 non-smoking waiters to formaldehyde for 12 weeks in a ballroom that served as an exposure chamber. A group of 23 non-smoking students with no occupational exposure to formaldehyde served as control. The average age of the workers was 29 ± 6.8 years, compared with 19 ± 2.3 years for the controls. They measured an 8h-time-weighted-average (TWA) of 0.80 ± 0.23 ppm formaldehyde, with a ceiling of 1.38 ppm, for the workers. The waiters were exposed to a 5h-TWA of 0.09 ± 0.05 ppm. The 8-hour TWA in the dormitories of the control group was 0.009 ppm. Ye et al. (2005) reported that the MN frequency was elevated in the nasal mucosa cells collected from workers (2.70 ± 1.50 per 1,000 cells; p<0.05), compared with controls (1.25 ± 0.65 per 1,000 cells). Similarly, the SCE frequency was increased in peripheral lymphocytes from workers (8.24 ± 0.89 per 1,000 cells; p<0.05), compared with controls (6.38 ± 0.41 per 1,000 cells). The frequencies of MN and SCEs in cells collected from the waiters were not different from controls. Speit et al. (2007) exposed volunteers (10 women, 11 men) 4hours/day for 10 working days to 0.15-0.5 ppm formaldehyde (specific concentration randomly assigned to each subject each day), with four 15-min 1-ppm peaks each day.25 The subjects were required to perform three 15-min bicycling exercises during each exposure. Cumulative exposure was 13.5 ppm-hour over the 10 days. Speit et al. (2007) prepared smears of exfoliated buccal mucosa cells collected from each subject 1 week before starting the study (Control 1), just before the first exposure (Control 2), immediately after the 10-day exposure period, and 7, 14 and 21 days thereafter. Each subject served as his/her own control. They analyzed 2,000 cells from each smear, and MN frequencies were determined on slides coded by an independent quality-assurance unit. These authors reported a statistically significant decrease in MN frequency 21 days after the end of the exposure period (0.44 ± 0.38 per 1,000 cells; p<0.05), compared with the

Panel Book Page 18

7

controls (Control 1: 0.95 ± 0.67 per 1,000 cells; Control 2: 0.86 ± 0.84 per 1,000 cells). MN frequencies in samples collected immediately, 7 days, or 14 days after the exposure period did not differ from controls. Yu et al. (2005) measured the workplace formaldehyde exposures and collected peripheral blood lymphocyte samples from 151 workers at two plywood factories and 112 workers at a machine manufacturing facility, which served as the control.26 They used air samplers and gas chromatograpphy (GC) to collect and analyze the air samples, a questionnaire to obtain personal information from the subjects, and a Comet assay and cytokinesis-block micronucleus (CBMN) test to identify DNA and chromosomal damage in the lymphocyte samples. The TWA concentrations for the plywood factory workers ranged from 0.08 to 6.42 ppm, compared to <0.008 ppm for the controls. The exposed workers were divided into two subgroups, including “low-exposed” and “high-exposed” workers They observed an exposure-related increase in the average Olive TM measured in the lymphocytes from controls (0.93; 0.78 - 1.10 µm), “low-exposed” (3.03; 2.49 -3.67), and “high-exposed” (3.95; 3.53 - 4.43) workers.26 The differences were statistically significant (p<0.05). Similar statistically significant differences (p<0.05) in MN frequencies (average number per 100 binucleated cells) were observed in the controls (0.27 ± 0.13), “low-exposed” (0.41 ± 0.25), and “high-exposed” (0.65 ± 0.36) workers. For controls, “low-exposed,” and “high-exposed” workers, respectively, the average Comet tail lengths were 6.78 (6.05−7.6), 11.25 (10.12−12.5), and 12.59 (11.8−13.43) µm. The authors report that the difference between “exposed” workers and the controls was statistically significant for this measure. Orsière et al. (2006) measured formaldehyde exposures using passive air-monitoring badges near the breathing zone of 59 pathology and anatomy laboratory workers for 15 minutes to 8 hours.27 Mean formaldehyde concentrations for the 59 subjects were 2.0 (range <0.1-20.4) and 0.1 (range <0.1-0.7ppm) for the 15-min and 8-h sampling times, respectively. A control group consisted of 37 individuals matched for gender, age, and smoking habits. These authors collected peripheral blood lymphocytes from 57 of the workers both before and after a 1-day exposure period. They found no increase in DNA damage in these workers after one day of exposure, using a chemiluminescence microplate assay to evaluate the lymphocytes. They used a CBMN assay combined with fluorescent in situ hybridization (FISH) and a pan-centromeric DNA probe to analyze the lymphocytes in 18 exposed and 18 control subjects randomly selected from the initial populations. Using this approach, they found statistically significant elevations in the frequencies of binucleated micronucleated cells (16.9 ± 9.3 vs. 11.1 ± 6.0 per 1,000 cells; p=0.001) and monocentromeric MN (11.0 ± 6.2 vs. 3.1 ± 2.4 per 1,000 cells; p<0.001) in pathologists/anatomists, compared to the controls. They found no statistically significant differences between these two groups in the frequencies of centromeric or acentromeric MN. Orsière et al. (2006) interpreted their results to suggest that formaldehyde genotoxicity is attributable to an aneugenic rather than clastogenic mode of action Costa et al. (2008) estimated the breathing-zone formaldehyde exposure of 30 pathology/anatomy laboratory workers at four hospitals in Portugal.28 The mean exposure concentration was 0.44 ± 0.08 ppm (range: 0.04−1.58 ppm). They selected 30 matched individuals working in administrative offices in these hospitals to serve as controls. These authors collected 10-ml venous blood samples from each participant at work, between 10 and 11 AM, and coded and analyzed the samples under blind conditions. They conducted MN, SCE, Comet and genotype analysis to evaluate the lymphocytes from each participant. Costa et al. (2008) reported statistically significant elevation in MN frequency (5.47 ± 0.76 vs. 3.27 ± 0.69 per 1,000 cells; p = 0.003), SCEs (6.13 ± 0.29 vs. 4.49 ± 0.16 per 1,000 cells; p < 0.05), and Comet tail lengths (60.00 ± 2.31 vs. 41.85 ± 1.97 µm; p < 0.05). In addition, they reported a positive correlation between formaldehyde exposure and both MN frequency (r = 0.384; p = 0.001) and Comet tail length (r = 0.333; p = 0.005).

Carcinogenicity Nasopharyngeal cancers – epidemiological studies Several reports evaluated the solid tumor mortality risks associated with formaldehyde exposures at 10 U.S. production plants in a National Cancer Institute (NCI) cohort study.29-31 The occupational histories of 25,619 workers first employed prior to 1966 in the manufacturing of formaldehyde, formaldehyde resins, molding compounds, plastic products, film or plywood, were gleaned from company records, and formaldehyde exposure was estimated for each job category.30,32 Exposure categories were defined for each of four exposure metrics,

Panel Book Page 19

8

including highest peak exposure (0, >0 to <2.0, 2.0 to <4.0, or ≥ 4.0 ppm), average intensity of exposure (0, >0 to <0.5, 0.5 to <1.0, or ≥1.0 ppm), cumulative exposure (0, >0 to <1.5, 1.5 to <5.5, or ≥5.5 ppm-years), and duration of exposure (0, >0 to <5, 5 to <15, or >15 years). Hauptmann et al. (2004) updated the cohort through 1994, reporting a 35-year median follow-up duration.31 Hauptmann et al. (2004) found 9 deaths from nasopharyngeal cancer (NPC) in this cohort, including 7 who were classified as “ever exposed” and 2 as “never exposed” to formaldehyde.31 These authors assumed a 15-year lag for NPC, used Poisson regression modeling and an internal referent group (i.e., either the unexposed or low- exposed group) to estimate the relative risks (RRs) for NPC, and regression analysis to evaluate dose-response trends, for each exposure metric (see Table 4-2,16 attached). The highest RRs were 4.14 for ≥5.5 ppm-years cumulative exposure and 4.18 for ≥15 years exposure duration, although confidence limits were not provided. Statistically significant dose-response trends were apparent for both peak exposure (p<0.001) and cumulative exposure (p=0.025). Marsh and coworkers evaluated the NCI data in a series of reports, focusing on Plant #1 (Wallingford, CT), a plastics-manufacturing plant where 5 of the 9 NPC cases evaluated by Hauptmann et al. (2004) were found.33-36 Marsh et al. (2002) conducted both a cohort and a nested case-control analysis of 7,328 workers employed in Plant #1 from 1941 to 1984, and independently evaluated the exposure assessment.33 They counted 7 NPC cases in this cohort, including 6 cases specifically identified as NPC and 1 case of pharyngeal cancer that was not identified specifically as NPC in the records. They reported that several formaldehyde exposure metrics were associated with NPC for Plant #1, including “ever exposed” (standardized mortality ratio [SMR] = 6.03; 95% CI: 2.42-2.42), exposure duration ≥10 years (SMR = 12.46; 95% CI: 1.51-45.02), and cumulative exposure ≥0.22 ppm-years (SMR = 7.51; 95% CI: 1.55-21.93) (see Table 4-2,16 attached). These authors suggested that their findings do not support a causal relationship between formaldehyde exposure and NPC mortality because elevated risks were seen in both short-term (<1 year; 4 cases) and long-term workers (3 cases), 5 NPC cases worked <5 years at the plant, the NPC cases among the long-term workers (<1 year) had relatively low average-intensity exposures (0.03-0.60 ppm), and the NPC deaths were concentrated among workers hired during 1947-1956. In a more current analysis, Marsh and Youk (2005) found that 6 of 10 NPC deaths (i.e., identified specifically as NPC) in the NCI cohort were associated specifically with employment at Plant #1, the remaining four cases distributed among four of the other nine plants studied.36 They reported a regional rate-based SMR of 10.32 (95% CI: 3.79-22.47) for formaldehyde-exposed workers at Plant #1, compared to 0.65 (95% CI: 0.08 to 2.33) for exposed workers at plants #2 through #10 combined. They found that the statistically significant peak exposure-response relationship in the NCI cohort was driven by excess NPC risk associated with the highest peak exposure category (≥4 ppm) at Plant #1. In this study, none of the exposure-response relationships for any of the four exposure metrics were statistically significant for plants #2 through #10, combined. They concluded that the suggestion by Hauptmann and colleagues31 of a causal relationship between formaldehyde exposure and NPC mortality is based entirely on the anomalous findings at Plant #1. More recently, Marsh and coworkers provided additional data from their nested case-control study, based on 7 NPC cases in the Plant #1 cohort.34 They reported a SMR of 4.43 (95% CI: 1.78-9.13; 7 deaths) for the exposed workers. However, they discovered that 5 of the 7 NPC cases also held silver-smithing and other jobs related to silver or brass or other metal work, and that this work was relatively rare in the remaining study population (OR = 7.31, 95% CI:1.08-82.1). They noted possible exposures to several suspected risk factors for upper respiratory system cancer (e.g., sulfuric acid mists, mineral acid, metal dusts and heat) associated with this type of work. Marsh and collaborators conducted additional re-analyses of the NCI cohort data, focusing on peak exposure and NPC mortality, demonstrating critical weaknesses in the model used in the Hauptmann et al. (2004) study, including instability problems related to the data from Plant #1.35 Most recently, Marsh et al. (2010) reviewed the recent finding reported by Beane Freeman et al. (2009) that Hauptmann and coworkers missed 1,006 death certificates of the NCI cohort, with proportionally greater numbers of missing deaths in the un-exposed and low-exposed groups used as internal referents in the Hauptmann et al. (2003) study.31,37-39 Marsh et al. (2010) noted that NCI has not provided corrected estimates for solid cancer deaths, including NPC deaths, for this cohort. They state that many of the recent meta-analyses, reviews, and regulatory

Panel Book Page 20

9

evaluations of the potential carcinogenicity of formaldehyde to humans should be revised to address this critical error in the crucial reports of Hauptmann and coworkers.31,38 Other cohort studies reported no association between occupational formaldehyde exposure and NPC mortalities. For example, Coggon et al. (2003) found 1 NPC case among 14,014 male British industrial workers, including 3,991 workers exposed to >2 ppm (RR = 0.5; 95% CI: 0.07-3.55; 2 cases expected),40 and Pinkerton et al. (2004) found 0 cases among 11,039 textile workers (82% female) (RR = 0; 95% CI: 0-3.00; 1 case expected)41 (see Bosetti et al., 2008).42 Further, a recent case-control study examining the potential association between formaldehyde exposure and myeloid leukemia in 6,808 deceased embalmers and funeral directors found 4 cases of NPC, only two of which had “ever embalmed” (OR = 0.1; 95% CI: 0.01-1.2).43 Exposures estimates (based on 6 different metrics) for these 2 cases were indistinguishable from controls. Nasopharyngeal cancers – mode of action Formaldehyde is highly reactive, readily forms DNA and protein adducts and crosslinks, and is a direct-acting genotoxicant. Among the potential modes of action that have been considered for the development of NPCs through the inhalation of formaldehyde in animal studies include direct mutagenesis of cells at the site of first contact and cytotoxicity-induced cell proliferation (CICP), which correlates with tumor incidence. The subchronic or chronic inhalation of formaldehyde at high concentrations (e.g., ≥6 ppm) clearly can cause NPCs in mice and rats. However, there is still considerable debate in the scientific community about whether this effect should be considered to be a non-threshold effect or a threshold effect in cancer risk assessments. For example, Monticello and colleagues exposed F344 rats via inhalation for 1, 4, 9 and 42 days (short-term) or 13, 26, 52 and 78 weeks (long-term) to 0, 0.7, 2.0, 6.0, 10.0, and 15.0 ppm formaldehyde.44-46 They reported statistically significant increases in nasal cell proliferation only at ≤6.0 ppm (short-term) and ≤10.0 ppm (long-term) in these studies. Conolly and coworkers interpreted these data to indicate that the dose-response curve is non-monotonic (i.e., highly-nonlinear), because cell proliferation was diminished at lower doses and elevated at the higher, cytotoxic doses.47-49 This view is consistent with the hypothesis that formaldehyde exposure must be sufficient to stimulate regenerative cell proliferation, thereby increasing the likelihood that mutations that would otherwise be repaired will become permanent, and could then lead to tumor formation. However, Subramaniam and Crump and their colleagues disputed this interpretation, because of the considerable uncertainty and variability in the data.50,51 Meng et al. (2010) exposed F-344 rats via inhalation to 0, 0.7, 2, 6, 10 or 15 ppm, 6h/day for 13 weeks.52 They then used allele-specific competitive blocker-PCR (ACB-PCR) to examine the nasal epithelial tissues for the presence of a K-Ras mutation and a p53 mutation previously detected in the squamous cell carcinomas produced by chronic formaldehyde exposure in a two-year bioassay.53 They also measured BrdU incorporation to monitor the proliferation of nasal mucosal cells in the rats. Meng and coworkers found that the mutation levels were not elevated above the low spontaneous background levels, even in the rats exposed to 15 ppm formaldehyde, and showed no dose-related increases.52 However, BrdU incorporation increased with dose, and was statistically significantly elevated in the rats exposed to either 10 ppm or 15 ppm formaldehyde. These results support the view that CICP plays a pivotal role in the formation of NPCs in rats and, thus, formaldehyde-induced carcinogenicity is largely a threshold effect. Lymphohematopoietic cancers - cohort studies Hauptmann et al. (2009) conducted a case-control study of lymphohematopoietic (LHP) and brain-cancer mortalities in funeral industry workers among the "professional" workers previously studied by Hayes, Walrath and their coworkers.43,54,55 They examined the death certificates (1960 to 1986) of 6,808 embalmers and funeral directors finding 168 deaths attributable to lymphohematopoietic cancers, including 99 lymphoid and 48 non-lymphoid cancers. The non-lymphoid cancers included 34 cases of myeloid leukemia. Cases were matched to control subjects (n=265) randomly selected from cohort members who died of other causes. They interviewed the next of kin and coworkers of the subjects to determine the funeral-home practices (e.g., ventilation and spill frequency) and work histories (e.g., frequency and duration of embalmings conducted for jobs ≤5 years). This information was used to estimate formaldehyde exposures for each subject. Exposure metrics included lifetime 8-hour TWA (ppm), peak

Panel Book Page 21

10

(ppm), cumulative (ppm-hours), and average intensity while embalming (ppm). Other exposure-related parameters estimated included “ever embalming,” duration working in jobs involving embalming, and number of embalmings conducted. Hauptmann et al. (2009) reported statistically significant increases in risks of lymphohematopoietic cancers of non-lymphoid origin for several of the exposure metrics, including the highest levels of exposure for cumulative, TWA, and peak exposures, as well as for subjects who embalmed for >20 years (OR = 3.5; 95% CI: 1.1-10.9; p=0.46).43 For myeloid leukemia, in particular, strong, statistically significant associations were found for exposure duration (e.g., OR = 13.6; 95% CI: 1.6−119.7 for >34 years), number of embalmings performed (e.g., OR = 2.7; 95% CI: 1.4−112.8 for >3,068 embalmings), and cumulative exposure (e.g., OR = 13.2; 95% CI: 1.5−115.4 for >9,253 ppm-hours) (see Table 4-7,16 attached). In addition, they found a statistically-significant dose-response relationship between myeloid leukemia deaths and both exposure duration (p=0.02) and peak exposure (p=0.036). Hauptmann et al. (2009) noted that there was only one case of myeloid leukemia in the reference group of non-embalmers.43 Thus they compared the subjects who performed <500 embalmings, which included 5 cases of myeloid leukemia, to the subjects with >34 years embalming (OR = 3.9; 95% CI: 1.2 -12.5; p=.024) and subjects with more than 9,253 ppm-hours cumulative exposure (OR = 3.1; 95% CI: 1.0 - 9.6; p=0.047). Several methodological issues have been identified for the Hauptmann et al. (2009) study.56 For example:

(1) Myeloid leukemia cases among the study subjects were 50% more likely than controls to have begun employment in the funeral industry before 1942. This suggests that they belonged primarily to an older and earlier population than the controls, and likely explains why they performed more embalmings.

(2) The single myeloid leukemia case in the control group yielded large, unstable confidence intervals in the Hauptmann et al. (2009) study,43 The ORs were substantially reduced when the referent group included both the controls and the subjects performing <500 embalmings.

(3) The myeloid leukemia cases and controls had nearly identical mean estimated average, 8-hour TWA, and

peak exposures. The cases had higher estimated number of embalmings and cumulative exposure than the controls, which can be explained by their earlier first employment, younger age at hire, and longer average employment in the industry, compared with controls.

Several reports evaluated the LHP cancer mortality risks associated with formaldehyde exposures for the 25,619 workers from 10 U.S. production plants in the National Cancer Institute (NCI) cohort study.30,38 The occupational histories of workers first employed before 1966 were obtained from company records, and formaldehyde exposures were estimated for each job category in each pant, based on job titles, associated tasks, and monitoring data.30,32 Beane Freeman et al. (2009) updated this cohort mortality study with follow-up through 2004, reporting a 42-year median follow-up duration.37 They discovered, and included, 1,006 death certificates that Hauptmann et al. (2003) had missed for this cohort.37,38 Proportionally greater numbers of missing deaths were among the un-exposed and low-exposed groups used as internal referents in the Hauptmann et al. (2003) paper. Beane-Freeman (2009) defined exposure categories for each of four exposure metrics, including highest peak exposure (0, >0 to <2.0, 2.0 to <4.0, or ≥ 4.0 ppm), peak exposure frequency (short-term exposures exceeding 8-hour TWA, hourly, daily, weekly, or monthly), average intensity of exposure (0, 0.1 to ,0.5, 0.5 to <1.0, or ≥1.0 ppm), and cumulative exposure (0, >0 to 1.5, 1.5 to <5.5, or ≥5.5 ppm-years).37 These authors found 319 deaths from all LHP cancers (from a total of 13,951 deaths), including 286 “exposed” and 33 “non-exposed” cases.37 Based on U.S. mortality rates, neither of these groups showed statistically significant elevations in standardized mortality ratio (SMRs) estimated for all LHP cancer (see Table 4-7,16 attached), all leukemia, lymphatic leukemia, myeloid leukemia, Hodgkin’s lymphoma, Non-Hodgkin’s lymphoma, or multiple myeloma.

Panel Book Page 22

11

Beane-Freeman (2009) assumed a 2-year lag for LHP cancers, after finding that assuming lag intervals from 2 to 25 years, by person-time, had little effect on RR estimates.37 They used Poisson regression modeling and an internal referent group (i.e., low-exposed group) to estimate the relative risks (RRs) for LHP mortalities in exposed workers, and either regression analysis or category ranks, as appropriate, to evaluate dose-response trends (see Table 4-7,16 attached). Beane Freeman and colleagues reported statistically significant elevations in RRs for all LHP cancers (RR = 1.37; 95% CI: 1.03-1.81) and Hodgkin’s lymphoma (RR = 3.96; 95% CI: 1.31-12.02) for workers with peak exposures ≥4 ppm, compared to >0 to 2.0 ppm (see Table 4-7,16 attached). In addition, they found statistically significant dose-response trends for peak exposure and all LHP (p=0.02), all leukemia (p=0.012) and Hodgkin’s lymphoma deaths (p=0.01), as well as for average exposure and Hodgkin’s lymphoma deaths (p=0.05 excluding “never-exposed” workers; p=0.03 including them). However, the RR for Hodgkin’s lymphoma in workers with the highest average exposure (≥1 ppm; RR = 2.48; 95% CI: 0.84-7.32) was lower than for workers with lower average exposure (0.5 to <1 ppm; RR = 3.62; 95% CI: 1.41-9.31). No statistically significant associations or trends were found among the LHP cancers and the other exposure metrics examined in this study, including both frequency of peak exposure and cumulative exposure37. Other recent cohort studies have reported no association between occupational formaldehyde exposure and LHP cancer mortalities. For example, Coggon et al. (2003) found 31 leukemia deaths among 14,014 male British industrial workers, including 3,991 workers exposed to >2 ppm (RR = 0.91; 95% CI: 0.64-1.29; 34 cases expected),40 and Pinkerton et al. (2004) found 59 cases among 11,039 textile workers (82% female) (RR = 1.09; 95% CI: 0.73-1.63; 61 cases expected)41 (see Bosetti et al., 2008).42 Lymphohematopoietic cancers – meta-analyses An early meta-analysis examined 18 epidemiology studies that reported leukemia rates in professional or industrial workers exposed to formaldehyde.57 These authors used fixed-effects models to evaluate both cohort and case-control studies. They found no association between leukemia and formaldehyde exposure across all of the studies (RR = 1.1; 95% CI: 1.0-1.2), across all cohort studies (RR = 1.0; 95% CI: 0.9-1.2), or across all case-control studies (RR = 2.4; 95% CI: 0.9-6.5). They reported a slightly elevated risk of leukemia among embalmers (RR = 1.6; 95% CI: 1.2-6.0) and pathologists/anatomists (RR = 1.4; 95% CI 1.0-1.9), but none for industrial workers, even those with the highest reported exposures (RR = 0.9; 95% CI: 0.8-1.0). More recently, Bosetti et al. (2008) evaluated cohort studies using either a fixed-effect or a random-effect model, depending on the heterogeneity among the cohorts.42 Using the fixed-effect model, they found a “modestly elevated” pooled RR for LHP cancers in professionals (i.e., embalmers, anatomists and pathologists) (RR = 1.31; 95% CI: 1.16-1.47; 8 studies), but not for industrial workers (RR = 0.85; 95% CI: 0.74-0.96; 4 studies). They reported similar results for leukemia. Zhang et al. (2009) reviewed many of the same studies that were included in the Bosetti et al. (2008) meta-analysis.42,58 However, they attempted to increase the statistical power of the analysis by focusing only on the highest exposure groups in each study, selecting exposure duration from some studies, and peak, average, or cumulative exposure from others. They also preferentially selected results for myeloid leukemia from studies that specifically addressed myeloid leukemia. These authors did not stratify the data to distinguish low-exposure (e.g., embalmers, pathologists, anatomists) from high-exposure (e.g., formaldehyde production) industries42,59 (see Bachand et al., 2010, for discussion).60 Zhang et al. (2009) used a fixed-effect or random-effect model, depending on the heterogeneity among the cohorts.58 Using the fixed-effect model, they calculated summary RRs (professional and industrial workers) for all LHP cancer (RR = 1.25; 95% CI: 1.09-1.43; 19 studies), all leukemias (RR = 1.54; 95% CI: 1.18-2.00; p<0.001; 15 studies), and myeloid leukemia (RR = 1.90; 95% CI: 1.31-2.76; p=0.001; 6 studies). They also report summary RRs for Hodgkin lymphoma (RR = 1.23; 95% CI: 0.67-2.29; 8 studies), non-Hodgkin’s lymphoma (RR = 1.08; 95% CI: 0.86-1.35; 11 studies) and multiple myeloma (RR = 1.31; 95% CI: 1.02-1.67; 9 studies). Bachand et al. (2010) conducted the most recent meta-analyses, which evaluated all cohort, case-control, and proportional mortality ratio (PMR) studies published through May 2009.60 Unlike earlier meta-analyses and

Panel Book Page 23

12

reviews, this study incorporated NCI cohort data updated to address the the missing the 1,006 death certificates reported by Beane Freeman and coworkers.37 In their summary risk estimates for leukemia, Bachand et al. (2010) found no statistically significant increase in the cohort (RR = 1.05; 95% CI: 0.93-1.20; 15 studies) or case-control studies (OR = 0.99; 95% CI : 0.71 - 1.37; 2 studies).60 Further, they reported no statistically significant increase in the summary leukemia RRs for embalmers and other professionsl/technical workers (RR = 1.28; 95% CI: 0.98-1.66; 7 studies) or for industrial workers (RR = 0.99; 95% CI: 0.86-1.15; 8 studies), or in the overall RR for myeloid leukemia (RR = 1.09; 95% CI: 0.84-1.40; 3 studies) calculated from the cohort studies. Although Banchand et al. (2010) found that their summary PMR for leukemia was significantly elevated (PMR = 1.44; 95% CI: 1.25- 1.67; 3 studies), they explained that PMRs are unreliable, and sugested that the inclusion of PMR studies may have caused inaccurately elevated summary risk estimates in previous meta-analyses (e.g., Collins, 2004; Zhang et al., 2009)58-60. Lymphohematopoietic cancers – mode of action There is remarkably little evidence from animal studies indicating that formaldehyde exposure can cause LHP cancer. Studies have consistently failed to find elevated levels of free formaldehyde or methylene glycol in the blood of exposed human and animal subjects, or DPCs in the bone marrow of exposed animals.61 Further, formaldehyde is a highly reactive, rapidly metabolized chemical yielding short-lived DPCs and DNA-adducts that are amenable to rapid reversal and repair.62-64 These observations are consistent with conventional wisdom, which has been that the expected sites of action of formaldehyde are limited to portals of entry (e.g., nasal epithelium), and would not likely include distal sites, such as the bone marrow, where leukemias originate.61,65 Although several posible modes of action have been postulated to explain associations between LHP cancers and formaldhyde exposure in epidemiological studies, there is little scientific evidence supporting these hypotheses, and some recent evidence against them. Thus, these proposals remain speculative, and continue to represent a highly controversial topic in the scientific community. The three proposed modes of action by which formaldehyde exposure may cause leukemia include:66

• Transport of formaldehyde/methylene glycol from the portal of entry through the blood to the bone marrow, followed by direct toxic action to hematopoietic stem cells in the marrow

• Direct toxic action of formaldehyde/methylene glycol on circulating blood stem cells and progenitors at the

portal of entry, followed by return of the damaged cells to bone marrow

• Direct toxic action of formaldehyde/methylene glycol on primitive pluripotent stem cells at the portal of entry, followed by migration of damaged cells to bone marrow

Similarly, direct toxic action of formaldehyde/methylene glycol on lymphocytes in mucosa-associated lymphoid tissues (MALT) at theportal of entry may cause lymphoid cancers (US EPA Draft Risk Assessment for Fromaldehyde). In a preliminary study, Zhang et al. (2010) measured compete blood counts and peripheral stem/progenitor cells in 43 Chinese formaldehyde-exposed workers from two factories (one producing and the other using formaldehyde-melamine resins) and 51 frequency-matched controls from three other workplaces in the same region.66 All participants wore diffusion samplers for a full shift on up to 3 working days over a three-week period to monitor formaldehyde exposures, and the factory workers wore organic vapor monitors at least twice to be analyzed for benzene and other organic solvents. The median (10th-90th percentile) formaldehyde exposure concentrations were 1.28 (0.63-2.51) ppm for the factory workers and 0.026 (0.0085-0.026) ppm for the controls. Zhang et al. (2010) reported statistically significant decreases in mean (± SD) WBC count (5,422 ± 1,529 vs. 6,269 ± 1,422 cells/µl; p=0.0016) and lymphocyte count (p=0.00002) in the subjects compared with the controls.66 Similarly, statistically significant decreases in granulocyte, platelet, and RBC counts, and increase in RBC mean

Panel Book Page 24

13

corpuscular volume (MCV) were found. No occupational co-exposures to benzene or other hemotoxic or genotoxic solvents were detected in this study. In addition, Zhang et al. (2010) conducted in vitro colony-forming unit – granulocytes, macrophages (CFU-GM) – assays on blood samples from all 94 participants cultured for 14 days.66 They reported a 20% decrease in progenitor cell colony formation in the blood samples from factory workers, compared to controls, but this result was not statistically significant. Next, Zhang et al. (2010) cultured mononuclear cells from a male volunteer of Chinese origin for 14 days after adding 0, 100, 150, or 200 µmol/l formaldehyde/methylene glycol to the culture medium on the first day.66 They found statistically significant, dose-related decreases in the number of colonies formed per plated cells. Finally, Zhang et al. (2010) analyzed metaphase spreads of the cultured CFU-GM cells from 12 of the highest exposed workers and 10 matched controls, using fluorescence in situ hybridization (FISH).66 They found statistically significant increases in the frequencies of both chromosome 7 monosomy (~2-fold; p=0.0039) and chromosome 8 trisomy (~4-fold; p=0.04) in the cells from the workers, compared with the controls. The authors indicate that both of these aneuploidies are common findings in individuals with myeloid leukemia, myelodisplastic syndromes, or benzene exposure. In a letter to the editor, Speit et al. (2010)67 indicate numerous problems in the Zhang et al. (2010)66 study, and questioned the reliability of the results by industrial hygienists. For example, they note that:

• All of the blood counts in the exposed workers were within the reference range.

• The frequencies of the aneuploidies reported were seen only after 14 days of in vitro incubation, were high for cells from both the workers and controls, and were not reported in either the factory workers or the controls in vivo.

• The most frequent chromosome aberrations associated with myeloid leukemia are translocations, but Zhang

et al. (2010)66 investigated neither translocations nor aneuploidies other than monosomy 7 and trisomy 8.

• Formaldehyde is mutagenic predominantly by a clastogenic, not an aneugenic mode of action.68-70

• Formaldehyde has been shown to damage several cell types directly exposed in vitro, an effect therefore not unique to myeloid progenitor cells.

Lu et al. (2010) exposed male Fischer rats via inhalation to 10 ppm [13CD2]-formaldehyde in a nose-only chamber, 6 hours/day for either 1 or 5 days.71 They added vaporized [13CD2]-formaldehyde to the air by thermally depolymerizing [13CD2]-paraformaldehyde. From each rat, they collected epithelial tissue samples from the right and left sides of the nose and the nasal septum, 3-5 ml of blood by cardiac puncture for lymphocyte isolation, and bone marrow from both femurs by saline extrusion. They also collected whole organs, including the spleen, thymus, lung and liver. Next, these authors isolated DNA from the each tissue samples, including ~30-50 µg DNA from nasal, 60-100 µg DNA from WBCs, and 200 µg from the other tissues, and hydrolyzed the DNA samples with DNaseI to analyze for both formaldehyde-DNA adducts and DDCs. These authors measured the formaldehyde-adducts (N2-HOCH2-dG and N6-HOCH2-dG) in the DNA samples as N2-CH3-dG and N6-CH3-dG, after reduction with NaCNBH3. DDCs were measured as dG-CH2-dG. They developed liquid chromatography - electrospray ionization - tandem mass spectrometry - selection reaction monitoring (LC-ESI-TMS-SRM) methods to quantify both the formaldehyde adducts (on-column detection limits ~240 amol and ~75 amol for N2-CH3-dG and N6-CH3-dG, respectively) and DNA-DNA crosslinks (on-column detection limits ~60 amol dG-CH2-dG). These analytical methods clearly differentiated the endogenous (N2-CH3-dG and dG-CH2-dG) from the exogenous (N2-13CD3-dG and dG-13CD2-dG) products. Using these methods, Lu et al. (2010) found exogenous products exclusively in the nasal tissues after 1 day (mean ± SD = 1.28 ± 0.49 monoadducts/107 dG; 0.14 DDCs/107 dG) or 5 days (e.g., mean ± SD = 2.43 ± 0.78 monoadducts/107 dG; 0.26 ± 0.07 DDCs/107 dG) of exposure.71 No exogenous products were detected in the DNA hydrosylates from any other tissue, even though, for example, the analytical method can detect ~3 N2-13CD3-dG

Panel Book Page 25

14

adducts/109 dG. This detection limit is ~30 times less than the endogenous N2-CH3-dG adducts/109 dG measured in WBCs. In contrast, endogenous products were found in all of the tissues examined, including blood (e.g., 1.10 ± 0.28 monoadducts/107dG, 3.66 ± 0.78 monoadducts/107dA, and 0.10 ± 0.07 crosslinks/107dG for 5-day exposure) and bone marrow (e.g., 1.17 ± 0.35 monoadducts/107dG, 2.99 ± 0.08 monoadducts/107dA, and 0.11 ± 0.03 crosslinks/107dG for 5-day exposure).71 The levels of endogenous products were comparable across all tissues examined. Lu et al. (2010) concluded that neither formaldehyde nor methylene glycol from formaldehyde reaches sites distant from the portal of entry, even when inhaled at high concentrations known to stimulate nasal epithelial cell proliferation and cause nasal tumors in rats.71 In addition, their results support the conclusion that genotoxic effects of formaldehyde/methylene glycol are not plausible at sites distant from the portal of entry. Likewise, the results demonstrate the implausibility of the idea that formaldehyde/methylene glycol transforms cells in the peripheral circulation or the nasal epithelium at the portal of entry, which can then migrate and incorporate into the bone marrow or other distant tissues to cause cancer. IARC cancer risk evaluations IARC (2006)72 concluded that there was sufficient epidemiological evidence that formaldehyde causes NPC in humans and strong but not sufficient evidence for a causal association between leukemia and occupational exposure to formaldehyde. They also elevated their evaluation of formaldehyde from probably carcinogenic to humans (Group 2A) to carcinogenic to humans (Group 1). In 2009, IARC73 updated their evaluation to conclude that there is sufficient evidence for a causal association between leukemia, particularly myeloid leukemia, and occupational exposure to formaldehyde.66,73 This conclusion was based primarily on:

• The statistically significant association between embalming and myeloid leukemia, including statistically significant trends for cumulative years embalming and peak formaldehyde exposure, reported by Hauptmann et al. (2009).43

• The levels of chromosome 7 monosomy and chromosome 8 trisomy in myeloid progenitor cells and

hematological changes in formaldehyde exposed workers reported by Zhang et al. (2010).66 The IARC Working Group was almost evenly split on the prevailing view that the evidence was sufficient for formaldehyde causing leukemia in humans (IARC, 2009)73.

Reproductive and Developmental Toxicity Non-Human Özen et al. (2005) exposed male Wistar rats (6/group) by inhalation to 0, 5, or 10 ppm formaldehyde, 8 hours/day, 5 days/week for 91 days.74 They used a chemi-luminescent enzyme immunoassay to measure serum testosterone concentrations, stained testicular tissues with Hematoxylin-Eosine (H-E) for histopathological examination, and immunohistochemical staining to estimate heat-shock protein 70 (Hsp70) levels in the tissues. They found statistically significant decreases (p<0.0001) in serum testosterone concentrations in the rats exposed to 5 ppm (244.01 ± 23.86 ng/dl) or 10 ppm (141.30 ± 23.86 ng/dl) formaldehyde, compared with controls (406.54 ± 16.82 ng/dl) (see Table 4-62,16 attached). Similarly, seminiferous tubule diameters were reduced (p<0.001) in rats exposed to 5 ppm (236.17 ± 13.09 µm) or 10 ppm (233.24 ± 10.13 µm) formaldehyde, compared with controls (259.22 ± 16.18 µm). In addition, Hsp70 levels were increased in the spermatogonia (+1 to +2), spermatocytes (+4 to +5), and spermatids (+4 to +5) of the treated rats (5 ppm or 10 ppm), compared with controls (0 to +2). Zhou et al. (2006) exposed adult male Sprague-Dawley rats (10/group) by inhalation to 8 ppm formaldehyde 12 hours/day for 2 weeks.75 Vitamin E (30 mg/kg/day) was administered by gavage to one of the groups during formaldehyde exposure. The control group consisted of rats that were not exposed to formaldehyde and received

Panel Book Page 26

15

only the physiological saline vehicle by gavage during the exposure period. They reported a statistically significant (p<0.05) reduction in testicular weight in the formaldehyde-exposed rats, compared with the controls. Histopathological examination revealed seminiferous tubule atrophy, interstitial vascular dilatation and hyperemia, disintegration and shedding of seminiferous epithelial cells into azoospermic lumina, and interstitial edema in the testes of the formaldehyde exposed rats. These authors found statistically significant (p<0.05) reductions in epididymal sperm count, percentage of motile sperm, activities of testicular SOD and glutathione peroxidase (GSH-Px), and GSH levels, and elevations in MDA levels in the formaldehyde exposed rats, compared with controls. All of these effects were markedly reduced in formaldehyde exposed rats that were also treated with Vitamin E. Golalipour et al. (2007) exposed Wistar rats (6-7 weeks old, 7/group) to 1.5 ppm formaldehyde (mean of measured concentrations), 4 hours/day 4 days/week, 2 hours/day 4 days/week, or 4 hours/day 2 days/week, by inhalation for 18 weeks.76 Testes were fixed, embedded, sectioned at 4 µm, and stained with H-E. They measured the diameter and height of 20 seminiferous tubules/testis morphometrically. They reported statistically significant reductions in both parameters in the exposed rats, compared with controls (see Table 4-63,16 attached). Further, they found severe reductions in the number of germ cells in the seminiferous tubules and evidence of arrested spermatogenesis after exposure 4 hours/day 4 days/week, decrease in the number of germ cells and increased thickness of the tubule basement membrane after exposure 2 hours/day 4 days/week, and disruption in the arrangement of Sertoli and germinal cells, with increased spacing between germ cells, after exposure for 4 hours/day 2 days/week. Xing et al. (2007) exposed male mice (12/group, strain not specified) to 0, 16.9, 33.8, or 67.6 ppm formaldehyde via inhalation 2 hours/day, 6 days/week for 13 weeks. They then evaluated the reproductive capacity of the males, using in a dominant-lethal protocol with untreated females, and examined sperm morphology. These authors reported a statistically significant increase in sperm aberration rate (p<0.05) and decrease in mean live fetuses/litter (p<0.01) after exposure to 67.6 ppm formaldehyde (see Table 4-64,16 attached). Resorption rates were statistically significantly elevated (p<0.05) for all groups of formaldehyde rats. Aslan et al. (2006) and Sarsilmaz et al. (2007) exposed neonatal male Wistar rats (10/group) to 0, 6, or 12 ppm formaldehyde in a glass chamber, 6 hours/day, 5 days/week for 30 days.77,78 After exposure, 5 rats/group were euthanized for neuropathological examination on PND30 or PND90. These authors reported lower numbers of both granular cells in the hippocampal dentate gyrus77 and pyramidal cells in the cornu ammonis of the hippocampus78 at PND90, compared to PND30. Kum et al. (2007) exposed female Sprague-Dawley rats (6 dams/group) and their offspring to 0 or 6 ppm formaldehyde 8 hours/day for 6 weeks, starting on gestation day 1(GD1), post-natal day 1 (PND1), or at 4 weeks of age. In another group, exposure was initiated during adulthood.79 These authors found statistically significant decreased mean body and liver weights in the offspring (p<0.01) when exposure began on GD1. For example, mean body weights were 20.83 ± 1.38 g (p<0.001) and 58.17 g (p<0.001) in the exposed rats on GD1 and PND1, respectively, compared with 30.33 ± 0.67 g on GD1 and 67.33 ± 1.73 g on PND1 for controls. However, liver weight was increased when exposure began at 4 weeks of age 3.83 ± 0.22 g (p<0.01), compared with controls (3.25 ± 0.11). The authors also reported a statistically significant increase in liver catalase (CAT) activity and malondialdehyde (MDA) concentration, decrease in liver gluthathione (GSH) concentration, and decrease in liver superoxide dismutase (SOD) activity in the offspring when exposure began on GD1, PND1, or at 4 weeks of age. No other significant differences in these parameters were observed among the exposed rats compared with controls. Human Taskinen et al. (1994) conducted a case-control study of spontaneous abortions in women occupationally exposed to formalin and other chemicals used in hospital laboratories in Finland.80 They identified subjects from the payrolls of state-employed laboratory workers, the laboratory workers’ union, and a register of workers occupationally exposed to carcinogens. They selected 208 women who had a single spontaneous abortion, and 329 controls who had delivered a baby without malformations, during 1973−1986. These authors used mailed questionnaires (82.4% response rate) to obtain health status, medication, contraception, pregnancy history, and exposure-related information from the subjects. Industrial hygienists developed an exposure index for each subject, based on their descriptions of work assignments, solvent use, and fume-hood use.

Panel Book Page 27

16

Taskinen et al. (1994) noted that another study (cited as Heikkilä et al., 1991) reported a mean formaldehyde concentration of 0.45 ppm (range: 0.01-7 ppm) in similar Finnish pathology/histology laboratories, and that the highest exposures occurred during emptying sample containers, washing dishes, and preparing formaldehyde solutions. Taskinen et al. (1994) reported a statistically significant association between exposure to formalin/formaldehyde 3-5 days/week and increased incidence of spontaneous abortion (OR = 3.5; 95% CI: 1.1−11.2), after adjusting for employment, smoking, alcohol consumption, parity, previous miscarriage, birth control failure, febrile disease during pregnancy, and exposure to other organic solvents in the workplace.80 Exposures to toluene (OR = 4.7; 95% CI: 1.4−15.9) and xylene (OR = 3.1; 95% CI: 1.3−7.5) were also significantly associated with the elevated incidence of spontaneous abortions. Taskinen et al. (1994) also reported no association between formalin exposure and congenital malformations in 36 laboratory workers compared with 105 controls registered in the Finnish Register of Congenital Malformations80. Taskinen et al. (1999) conducted a retrospective cohort study of fertility in women occupationally exposed to formaldehyde and other chemicals in the woodworking industry in Finland.81 They recruited subjects from the woodworkers’ union and other wood-processing businesses, and linked them to the Finnish national register of births. These authors identified 1,094 women who were born between 1946 and 1975, had a live birth at age 20−40 years during 1985−1995, worked in the wood processing industry for at least 1 month, and were first employed in the wood processing industry beginning at least 6 months before the index pregnancy (i.e., the first pregnancy that fulfilled the other criteria). They used mailed questionnaires (64% response rate) to obtain personal, pregnancy, and exposure-related information. The final sample included 602 women, after other exclusions (e.g., based on history of infertility, unknown time-to-pregnancy, or contraceptive failure). Industrial hygienists estimated the mean daily exposure during the time-to-pregnancy period for each subject, based on the proportion of the workday during which exposure occurred and either concentrations measured at the subject’s factory in the early 1990s or concentrations reported for similar industries. The subjects were divided into three categories, based on the TWA exposure estimates, including low (0.1 to 3.9 ppm), medium (4.0 to 12.9 ppm), and high (13.0 to 63 ppm).81 The authors calculated fecundability density ratios (FDRs) for each exposure category, by dividing the average pregnancy incidence density of the exposed women by that of 367 employed, unexposed women. They adjusted the FDRs for employment, smoking, alcohol consumption, parity, and menstrual irregularity. Taskinen et al. (1999) reported a statistically significant decrease in the FDR for the formaldehyde exposed women in the high exposure group (OR = 0.64; 95% CI: 0.43−0.92; p=0.02), and in the women in the high exposed group who did not wear gloves (n=17; OR = 0.51; 95% CI: 0.28−0.92).81 The reduced FDR among women in the high exposed group who wore gloves was not statistically significant (n=22; OR = 0.79; 95% CI: 0.47-1.23). Taskinen et al. (1999) also reported associations between formaldehyde exposure and spontaneous abortion and 52 women who had worked in their workplace during the year of the spontaneous abortion and at the beginning of the time-to-pregnancy period.81 The ORs were 3.2 (95% CI: 1.2−8.3), 1.8 (95% CI: 0.8−4.0), and 2.4 (95% CI: 1.2−4.8) for the low, medium, and high exposure categories, respectively. Endometriosis also appeared to be associated with formaldehyde exposure in women in the high exposure category (OR = 4.5; 95% CI: 1.0−20.0). In an earlier case-control study, John et al. (1994) investigated spontaneous abortions during 1983-1988 in cosmetologists, compared with controls who delivered a live infant during the same period.82 The subjects were identified from the 1988 North Carolina cosmetology license registry. They gathered information using mailed questionnaires (72.5% response rate). Among the full-time cosmetologists who qualified for the study, 61 cases of spontaneous abortion were selected for comparison to 315 controls. Johns et al. (1994) reported a crude OR of 2.0 (95% CI: 1.1-3.8) for use of formaldehyde-based disinfectants.82 The OR was 2.1 (95% CI: 1.0−4.3) after adjusting for maternal characteristics (e.g., age, smoking, glove use, other jobs) and other workplace exposures (e.g., chemicals used on hair, use of manicure products).

Panel Book Page 28

17

Collins and coworkers83conducted a meta-analysis of epidemiological studies that examined the potential association between spontaneous abortions and formaldehyde exposure, including John et al. (1994)82 and Taskinen (1999).81 They reported a meta-RR of 1.4 (95% CI: 0.9-2.1). However, they noted no increased risk of spontaneous abortion in formaldehyde workers after adjusting this estimate for reporting and publication biases (meta-RR = 0.7; 95% CI: 0.5-1.0).

V. Regulatory Guidance Values Standards and Guidance for Acute Inhalation Exposures The U.S. National Advisory Committee for Acute Exposure Guideline Levels (NAC AEGL Committee) for Hazardous Substances recently developed an interim acute exposure guideline level-1 (AEGL-1) of 0.9 ppm for formaldehyde.84 The AEGL-1 is defined as a concentration in air above which the general population (including susceptible individuals) could experience notable discomfort, irritation, or other adverse effects. The AEGL-1 was based on the NOAEL for eye irritation in a study in which 5 to 28 healthy subjects previously shown to be sensitive to 1.3 or 2.2 ppm formaldehyde were exposed eye-only for 6 minutes to 0, 0.35, 0.56, 0.7, 0.9, or 1.0 ppm85 (see Table 3,84 attached). Subjective eye irritation responses ranged from none to slight at 0, 0.35, 0.56, 0.7 and 0.9 ppm. The NAC AEGL Committee (2008) applied the 0.9 ppm AEGL-1 across all acute exposure durations (10-min to 8 hours) because several studies show that there is adaptation to irritation at such concentrations (see Table 10,84 attached). They also noted that, in the absence of exercise, there are no decrements in pulmonary function parameters in healthy or asthmatic subjects inhaling 3 ppm for 3 hours.86-88 Other current standards and guidance values for acute formaldehyde exposures are summarized in Table 10. U.S. EPA Risk Assessments - Non-Cancer Effects In 1990, U.S. EPA published a chronic reference dose (cRfD) of 0.2 mg/kg/day for oral exposure to “formaldehyde,” based on the results of a 2-year bioassay in rats.89 Til et al. (1989) administered “formaldehyde” (methylene glycol/formaldehyde) to Wistar rats (70/sex/dose) in drinking water, yielding mean doses of 0, 1.2, 15, or 82 mg/kg/day for males and 0, 1.8, 21, or 109 mg/kg/day for females. The NOAEL was 15 mg/kg/day in this study. U.S. EPA also noted a two-stage carcinogenesis bioassay conducted by Takahashi et al. (1986) in male Wistar rats.90 The animals were treated with 100 mg/l N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) in the drinking water for the first 8 weeks of the study, followed by 0.5% formalin (dose not specified) in the drinking water during weeks 8 through 40. Other groups of animals received only MNNG or formalin (dose not specified). Formalin alone did not produce malignant tumors, although forestomach papillomas were found in 8/10 animals. In the group receiving both MNG and formalin, forestomach papillomas were found in 15/17 rats, adenocarcinoma of the pylorus in 4/17, preneoplastic hyperplasia of the pylorus in 7/17, and adenocarcinoma of the duodenum in 1/17. U.S. EPA recently released a draft risk assessment for formaldehyde for public comment and review by the National Academy of Sciences (NAS).16 They proposed a chronic reference concentration (cRfC) of 9 ppb (900 ppt) for formaldehyde exposure by inhalation, based on three “cocritical” epidemiological studies. These studies reported associations between formaldehyde exposure and increased physician-diagnosed asthma91, increased asthma, atopy, and respiratory symptoms92, and decreased pulmonary peak expiratory flow rate93 in residential populations, including children. U.S. EPA Risk Assessments - Carcinogenicity In 1991, U.S. EPA classified formaldehyde as a B1 carcinogen (i.e., a probable human carcinogen), based on limited evidence in humans, and sufficient evidence in animals.16 They estimated an upper-bound inhalation cancer unit risk of 1.6 x 10-2 per ppm (1.3 x 10-5 per ug/m3), using a linearized multistage, additional-risk procedure to extrapolate dose-response data from a chronic bioassay on male F344 rats.12 An upper-bound 10-6 human cancer risk would be associated with continuous inhalation of 0.06 ppb (63 ppt) formaldehyde over a lifetime, based on this unit risk.

Panel Book Page 29

18

Recently, U.S. EPA proposed to identify formaldehyde as carcinogenic to humans.16 They proposed an upper-bound inhalation cancer unit risk of 8.1 x 10-2 per ppm (6.6 x 10-5 µg/m3) for NPC, Hodgkin’s lymphoma, and leukemia, combined, using log-linear modeling, extra risk procedures to extrapolate cumulative exposure estimates from the epidemiological studies of Hauptmann et al. (2004) and Beane Freeman (2009).31,37 An upper-bound 10-6 human cancer risk would be associated with continuous inhalation of 0.01 ppb (12 ppt) formaldehyde throughout adulthood, based on this unit risk. Further, they proposed an overall upper-bound inhalation cancer unit risk of 1.3 x 10-1 per ppm (1.1 x 10-4 per µg/m3) for exposure during both childhood and adulthood by applying age-dependent adjustment factors (ADAFs).16 An upper-bound 10-6 human cancer risk would be associated with continuous inhalation of 0.01 ppb (12 ppt) formaldehyde throughout childhood and adulthood, based on this unit risk. An upper-bound 10-6 human cancer risk would be associated with continuous inhalation of 0.0077 ppb (7.7 ppt) formaldehyde over a lifetime, based on this unit risk.

REFERENCES

1. Elder, R. L. Final report on the safety assessment of Formaldehyde. J Amer Coll Toxicol. 1984;3:(3):157-184.

2. Andersen, F. A. Unpublished re-review of Formaldehyde. Cosmetic Ingredient Review. 8-8-2003.

3. Andersen, F. A. Annual Review of Cosmetic Ingredient Safety Assessments - 2004/2005 - Formaldehyde. Int J Toxicol. 2006;25:(Suppl. 2):30-35.

4. Formaldehyde. 1964. ACS Monograph Series. Reinhold, New York.

5. Phenolic Resins. 1996. Chapter: 6. Resins for Coatings: Chemistry, Properties, and Applications. Dieter Stoye Werner Freitag, Günter Beuschel. Hanser Verlag.

6. Priha, E., Liesivuori, J., Santa, H., and Laatikainen, R. Reactions of Hydrated Formaldehyde in Nasal Mucus. Chemosphere. 1996;32:(6):1011-1082.

7. Burnett MG. The mechanism of the formaldehyde clock reaction. Methylene glycol dehydration. J Chem Educ. 1982;160:160.

8. Le Botlan DJ, Mechin BG, and Martin GJ. Proton and carbon-13 nuclear magnetic resonance spectrometry of formaldehyde in water. Anal Chem. 1983;55:587.

9. Gottschalck, T. E. and J.E.Bailey. International Cosmetic Ingredient Handbook and Dictionary. 13th ed.

Personal Care Products Council: Washington, DC 20036, 2010.

10. Food and Drug Adminstration. Uses of Formaldehyde and Formalin reported to the Voluntary Cosmetic Registration Program. Washington, DC;FDA, 2010.

11. Health Canada. Use of Formaldehyde, Formalin, and Methylene Glycol in Canada. Ottawa, Ontario: Product Safety Programme, 2010.

12. Kiernan, John A. Formaldehyde, formalin, paraformaldehyde and glutaraldehyde: What they are and what they do. Microscopy Today. 2000;1:8-12.

13. Drahl, Carmen. Hair Straighteners. Chemical and Engineering News. 11-8-2010. 88:(45):54-54. The American Chemical Society.

14. Helander KG. Kinetic studies of formaldehyde binding in tissue. Biotech Histochem. 1994;69:177-179.

Panel Book Page 30

19

15. Fox, C. H., Johnson, F. B., Whiting, J., and Roller, P. P. Formaldehyde Fixation. The Journal of Histochemistry and Cytochemistry. 1985;33:(8):845-853.

16. U.S. EPA. Toxicological review of formaldehyde - Inhalation Assessment - External Review Draft. 2010. http://www.epa.gov/IRIS/.

17. Wahlberg, J. E. Measurement of skin-fold thickness in the guinea pig. Assessment of edema-inducing capacity of cutting fluids, acids, alkalis, formalin and dimethyl sulfoxide. Contact Dermatitis. 1993;28:(3):141-145.

18. Lee, H. K., Alarie, Y., and Karol, M. H. Induction of formaldehyde sensitivity in guinea pigs. Toxicol Appl.Pharmacol. 1984;75:(1):147-155.

19. Arts, J. H., Droge, S. C., Spanhaak, S., Bloksma, N., Penninks, A. H., and Kuper, C. F. Local lymph node activation and IgE responses in brown Norway and Wistar rats after dermal application of sensitizing and non-sensitizing chemicals. Toxicology. 2-28-1997;117:(2-3):229-234.

20. Im, H., Oh, E., Mun, J., Khim, J. Y., Lee, E., Kang, H. S., Kim, E., Kim, H., Won, N. H., Kim, Y. H., Jung, W. W., and Sul, D. Evaluation of toxicological monitoring markers using proteomic analysis in rats exposed to formaldehyde. J Proteome.Res. 2006;5:(6):1354-1366.

21. Sul, D., Kim, H., Oh, E., Phark, S., Cho, E., Choi, S., Kang, H. S., Kim, E. M., Hwang, K. W., and Jung, W. W. Gene expression profiling in lung tissues from rats exposed to formaldehyde. Arch Toxicol. 2007;81:(8):589-597.

22. Speit, G. The implausibility of systemic genotoxic effects measured by the comet assay in rats exposed to

formaldehyde. J Proteome.Res. 2006;5:(10):2523-2524.

23. Speit, G., Zeller, J., Schmid, O., Elhajouji, A., Ma-Hock, L., and Neuss, S. Inhalation of formaldehyde does not induce systemic genotoxic effects in rats. Mutat.Res. 2009;677:(1-2):76-85.

24. Ye, X., Yan, W., Xie, H., Zhao, M., and Ying, C. Cytogenetic analysis of nasal mucosa cells and lymphocytes from high-level long-term formaldehyde exposed workers and low-level short-term exposed waiters. Mutat.Res. 12-7-2005;588:(1):22-27.

25. Speit, G., Schmid, O., Frohler-Keller, M., Lang, I., and Triebig, G. Assessment of local genotoxic effects of formaldehyde in humans measured by the micronucleus test with exfoliated buccal mucosa cells. Mutat.Res. 3-5-2007;627:(2):129-135.

26. Yu, L. Q., Jiang, S. F., Leng, S. G., He, F. S., and Zheng, Y. X. [Early genetic effects on workers occupationally exposed to formaldehyde]. Zhonghua Yu Fang Yi.Xue.Za Zhi. 2005;39:(6):392-395.

27. Orsiere, T., Sari-Minodier, I., Iarmarcovai, G., and Botta, A. Genotoxic risk assessment of pathology and anatomy laboratory workers exposed to formaldehyde by use of personal air sampling and analysis of DNA damage in peripheral lymphocytes. Mutat.Res. 6-16-2006;605:(1-2):30-41.

28. Costa, S., Coelho, P., Costa, C., Silva, S., Mayan, O., Santos, L. S., Gaspar, J., and Teixeira, J. P. Genotoxic damage in pathology anatomy laboratory workers exposed to formaldehyde. Toxicology. 10-30-2008;252:(1-3):40-48.

29. Blair, A., Stewart, P. A., Hoover, R. N., Fraumeni, J. F., Jr., Walrath, J., O'Berg, M., and Gaffey, W. Cancers of the nasopharynx and oropharynx and formaldehyde exposure. J Natl.Cancer Inst. 1987;78:(1):191-193.

Panel Book Page 31

20

30. Blair, A., Stewart, P., O'Berg, M., Gaffey, W., Walrath, J., Ward, J., Bales, R., Kaplan, S., and Cubit, D. Mortality among industrial workers exposed to formaldehyde. J Natl.Cancer Inst. 1986;76:(6):1071-1084.

31. Hauptmann, M., Lubin, J. H., Stewart, P. A., Hayes, R. B., and Blair, A. Mortality from solid cancers among workers in formaldehyde industries. Am.J Epidemiol. 6-15-2004;159:(12):1117-1130.

32. Blair, A. and Stewart, P. A. Correlation between different measures of occupational exposure to formaldehyde. Am.J Epidemiol. 1990;131:(3):510-516.

33. Marsh, G. M., Youk, A. O., Buchanich, J. M., Cassidy, L. D., Lucas, L. J., Esmen, N. A., and Gathuru, I. M. Pharyngeal cancer mortality among chemical plant workers exposed to formaldehyde. Toxicol Ind.Health. 2002;18:(6):257-268.

34. Marsh, G. M., Youk, A. O., Buchanich, J. M., Erdal, S., and Esmen, N. A. Work in the metal industry and nasopharyngeal cancer mortality among formaldehyde-exposed workers. Regul.Toxicol Pharmacol. 2007;48:(3):308-319.

35. Marsh, G. M., Youk, A. O., and Morfeld, P. Mis-specified and non-robust mortality risk models for nasopharyngeal cancer in the National Cancer Institute formaldehyde worker cohort study. Regul.Toxicol Pharmacol. 2007;47:(1):59-67.

36. Marsh, G. M. and Youk, A. O. Reevaluation of mortality risks from nasopharyngeal cancer in the formaldehyde cohort study of the National Cancer Institute. Regul.Toxicol Pharmacol. 2005;42:(3):275-283.

37. Beane Freeman, L., A.Blair, and J.H.Lubin. Mortality from lymphohematopoietic malignancies among workers in formaldehyde industries: The National Cancer Institute cohort. Journal of the National Cancer Institute. 2009;101:(10):751-761.

38. Hauptmann, M., J.H.Lubin, P.A.Stewart, R.B.Hayes, and A.Blair. Mortality from lymphohematopoietic malignancies among workers in formaldehyde industries. Journal of the National Cancer Institute. 2003;95:(21):1615-1623.

39. Marsh, G. M., Youk, A. O., Morfeld, P., Collins, J. J., and Symons, J. M. Incomplete follow-up in the National Cancer Institute's formaldehyde worker study and the impact on subsequent reanalyses and causal evaluations. Regul.Toxicol Pharmacol. 2010;58:(2):233-236.

40. Coggon, D., E.C.Harris, J.Poole, and K.T.Palmer. Extended follow-up of a cohort of British chemical workers exposed to formaldehyde. Journal of the National Cancer Institute. 2003;95:(21):1608-1615.

41. Pinkerton, L. E., M.J.Hein, and L.T.Stayner. Mortality among a cohort of garment workers exposed to formaldehyde: An update. Occupational Environmental Medicine. 2004;61:193-200.

42. Bosetti, C., McLaughlin, J. K., Tarone, R. E., Pira, E., and La, Vecchia C. Formaldehyde and cancer risk: a quantitative review of cohort studies through 2006. Ann Oncol. 2008;19:(1):29-43.

43. Hauptmann, M., Stewart, P. A., Lubin, J. H., Beane Freeman, L. E., Hornung, R. W., Herrick, R. F., Hoover, R. N., Fraumeni, J. F., Jr., Blair, A., and Hayes, R. B. Mortality from lymphohematopoietic malignancies and brain cancer among embalmers exposed to formaldehyde. J Natl.Cancer Inst. 12-16-2009;101:(24):1696-1708.

44. Monticello, T. M., Morgan, K. T., and Hurtt, M. E. Unit length as the denominator for quantitation of cell proliferation in nasal epithelia. Toxicol Pathol. 1990;18:(1 Pt 1):24-31.

Panel Book Page 32

21

45. Monticello, T. M., Miller, F. J., and Morgan, K. T. Regional increases in rat nasal epithelial cell proliferation following acute and subchronic inhalation of formaldehyde. Toxicol Appl.Pharmacol. 1991;111:(3):409-421.

46. Monticello, T. M., Swenberg, J. A., Gross, E. A., Leininger, J. R., Kimbell, J. S., Seilkop, S., Starr, T. B., Gibson, J. E., and Morgan, K. T. Correlation of regional and nonlinear formaldehyde-induced nasal cancer with proliferating populations of cells. Cancer Res. 3-1-1996;56:(5):1012-1022.

47. Conolly, R. B., Kimbell, J. S., Janszen, D. B., and Miller, F. J. Dose response for formaldehyde-induced cytotoxicity in the human respiratory tract. Regul.Toxicol Pharmacol. 2002;35:(1):32-43.

48. Conolly, R. B., Kimbell, J. S., Janszen, D., Schlosser, P. M., Kalisak, D., Preston, J., and Miller, F. J. Biologically motivated computational modeling of formaldehyde carcinogenicity in the F344 rat. Toxicol Sci. 2003;75:(2):432-447.

49. Gaylor, D. W., Lutz, W. K., and Conolly, R. B. Statistical analysis of nonmonotonic dose-response relationships: research design and analysis of nasal cell proliferation in rats exposed to formaldehyde. Toxicol Sci. 2004;77:(1):158-164.

50. Crump, K. S., Chen, C., Fox, J. F., Van, Landingham C., and Subramaniam, R. Sensitivity analysis of biologically motivated model for formaldehyde-induced respiratory cancer in humans. Ann Occup.Hyg. 2008;52:(6):481-495.

51. Subramaniam, R. P., Chen, C., Crump, K. S., Devoney, D., Fox, J. F., Portier, C. J., Schlosser, P. M., Thompson, C. M., and White, P. Uncertainties in biologically-based modeling of formaldehyde-induced respiratory cancer risk: identification of key issues. Risk Anal. 2008;28:(4):907-923.

52. Meng, F., Bermudez, E., McKinzie, P. B., Andersen, M. E., Clewell, H. J., III, and Parsons, B. L. Measurement of tumor-associated mutations in the nasal mucosa of rats exposed to varying doses of formaldehyde. Regul Toxicol Pharmacol. 2010;57:(2-3):274-283.

53. Recio, L., Sisk, S., Pluta, L., Bermudez, E., Gross, E. A., Chen, Z., Morgan, K., and Walker, C. p53

mutations in formaldehyde-induced nasal squamous cell carcinomas in rats. Cancer Res. 11-1-1992;52:(21):6113-6116.

54. Hayes, R. B., Blair, A., Stewart, P. A., Herrick, R. F., and Mahar, H. Mortality of U.S. embalmers and funeral directors. Am.J Ind.Med. 1990;18:(6):641-652.

55. Walrath, J. and Fraumeni, J. F., Jr. Mortality patterns among embalmers. Int.J Cancer. 4-15-1983;31:(4):407-411.

56. ENVIRON. Comments on the National Toxicology Program Draft Report on Carcinogens Substance Profile for formaldehyde. 2010.

57. Collins, J. J. and Lineker, G. A. A review and meta-analysis of formaldehyde exposure and leukemia. Regul Toxicol Pharmacol. 2004;40:(2):81-91.

58. Zhang, L., Steinmaus, C., Eastmond, D. A., Xin, X. K., and Smith, M. T. Formaldehyde exposure and

leukemia: a new meta-analysis and potential mechanisms. Mutat.Res. 2009;681:(2-3):150-168.

59. Collins, J. J. Formaldehyde exposure and leukaemia. Occup.Environ.Med. 2004;61:(11):875-876.

60. Bachand, A. M., Mundt, K. A., Mundt, D. J., and Montgomery, R. R. Epidemiological studies of formaldehyde exposure and risk of leukemia and nasopharyngeal cancer: a meta-analysis. Crit Rev.Toxicol. 2010;40:(2):85-100.

Panel Book Page 33

22

61. Heck, H. and Casanova, M. The implausibility of leukemia induction by formaldehyde: a critical review of the biological evidence on distant-site toxicity. Regul.Toxicol Pharmacol. 2004;40:(2):92-106.

62. Lu, K., Ye, W., Zhou, L., Collins, L. B., Chen, X., Gold, A., Ball, L. M., and Swenberg, J. A. Structural characterization of formaldehyde-induced cross-links between amino acids and deoxynucleosides and their oligomers. J Am.Chem Soc. 3-17-2010;132:(10):3388-3399.

63. Zhong, W. and Que Hee, S. S. Formaldehyde-induced DNA adducts as biomarkers of in vitro human nasal epithelial cell exposure to formaldehyde

2. Mutat Res. 9-12-2004;563:(1):13-24.

64. Zhong, W. and Que Hee, S. S. Formaldehyde-induced DNA adducts as biomarkers of in vitro human nasal epithelial cell exposure to formaldehyde

2. Mutat Res. 9-12-2004;563:(1):13-24.

65. Pyatt, D., Natelson, E., and Golden, R. Is inhalation exposure to formaldehyde a biologically plausible cause of lymphohematopoietic malignancies? Regul.Toxicol Pharmacol. 2008;51:(1):119-133.

66. Zhang, L., Tang, X., Rothman, N., Vermeulen, R., Ji, Z., Shen, M., Qiu, C., Guo, W., Liu, S., Reiss, B., Freeman, L. B., Ge, Y., Hubbard, A. E., Hua, M., Blair, A., Galvan, N., Ruan, X., Alter, B. P., Xin, K. X., Li, S., Moore, L. E., Kim, S., Xie, Y., Hayes, R. B., Azuma, M., Hauptmann, M., Xiong, J., Stewart, P., Li, L., Rappaport, S. M., Huang, H., Fraumeni, J. F., Jr., Smith, M. T., and Lan, Q. Occupational exposure to formaldehyde, hematotoxicity, and leukemia-specific chromosome changes in cultured myeloid progenitor cells. Cancer Epidemiol.Biomarkers Prev. 2010;19:(1):80-88.

67. Speit, G., Gelbke, H. P., Pallapies, D., and Morfeld, P. Occupational exposure to formaldehyde, hematotoxicity and leukemia-specific chromosome changes in cultured myeloid progenitor cells. Cancer Epidemiol.Biomarkers Prev. 2010;19:(7):1882-1884.

68. Schmid, O. and Speit, G. Genotoxic effects induced by formaldehyde in human blood and implications for the interpretation of biomonitoring studies. Mutagenesis. 2007;22:(1):69-74.

69. Speit, G., Neuss, S., Schutz, P., Frohler-Keller, M., and Schmid, O. The genotoxic potential of glutaraldehyde in mammalian cells in vitro in comparison with formaldehyde. Mutat.Res. 1-8-2008;649:(1-2):146-154.

70. Titenko-Holland, N., Levine, A. J., Smith, M. T., Quintana, P. J., Boeniger, M., Hayes, R., Suruda, A., and Schulte, P. Quantification of epithelial cell micronuclei by fluorescence in situ hybridization (FISH) in mortuary science students exposed to formaldehyde. Mutat.Res. 12-20-1996;371:(3-4):237-248.

71. Lu, K., Collins, L. B., Ru, H., Bermudez, E., and Swenberg, J. A. Distribution of DNA adducts caused by inhaled formaldehyde is consistent with induction of nasal carcinoma but not leukemia. Toxicol Sci. 2010;116:(2):441-451.

72. IARC. IARC Monographs on the Evaulation of Carcinogenic Risks to Humans - Formaldehyde, 2-Butoxyethanol and 1-tert-Butoxypropan-2-ol. World Health Organization (WHO) International Programme in Chemical Safety (IPCS). 2006.

73. Baan, R., Grosse, Y., Straif, K., Secretan, B., El Ghissassi, F., Bouvard, V., Benbramin-Tallaa, L., Guha, N., Freeman, C., Galichet, L., and Coglinano, V. A review of human carcinogens - Part F: Chemical Agents and related occupations. Lancet. 2009;10:1143-1144.

Panel Book Page 34

23

74. Ozen, O. A., Akpolat, N., Songur, A., Kus, I., Zararsiz, I., Ozacmak, V. H., and Sarsilmaz, M. Effect of formaldehyde inhalation on Hsp70 in seminiferous tubules of rat testes: an immunohistochemical study. Toxicol Ind.Health. 2005;21:(10):249-254.

75. Zhou, D. X., Qiu, S. D., Zhang, J., and Wang, Z. Y. [Reproductive toxicity of formaldehyde to adult male rats and the functional mechanism concerned]. Sichuan.Da.Xue.Xue.Bao.Yi.Xue.Ban. 2006;37:(4):566-569.

76. Golalipour, M. J., Azarhoush, R., Ghafari, S., Gharravi, A. M., Fazeli, S. A., and Davarian, A. Formaldehyde exposure induces histopathological and morphometric changes in the rat testis. Folia Morphol.(Warsz.). 2007;66:(3):167-171.

77. Aslan, H., Songur, A., Tunc, A. T., Ozen, O. A., Bas, O., Yagmurca, M., Turgut, M., Sarsilmaz, M., and Kaplan, S. Effects of formaldehyde exposure on granule cell number and volume of dentate gyrus: a histopathological and stereological study. Brain Res. 11-29-2006;1122:(1):191-200.

78. Sarsilmaz, M., Kaplan, S., Songur, A., Colakoglu, S., Aslan, H., Tunc, A. T., Ozen, O. A., Turgut, M., and Bas, O. Effects of postnatal formaldehyde exposure on pyramidal cell number, volume of cell layer in hippocampus and hemisphere in the rat: a stereological study. Brain Res. 5-11-2007;1145:157-167.

79. Kum, C., Sekkin, S., Kiral, F., and Akar, F. Effects of xylene and formaldehyde inhalations on renal oxidative stress and some serum biochemical parameters in rats. Toxicol Ind.Health. 2007;23:(2):115-120.

80. Taskinen, H., Kyyronen, P., Hemminki, K., Hoikkala, M., Lajunen, K., and Lindbohm, M. L. Laboratory work and pregnancy outcome. J Occup.Med. 1994;36:(3):311-319.

81. Taskinen, H. K., Kyyronen, P., Sallmen, M., Virtanen, S. V., Liukkonen, T. A., Huida, O., Lindbohm, M. L., and Anttila, A. Reduced fertility among female wood workers exposed to formaldehyde. Am.J Ind.Med. 1999;36:(1):206-212.

82. John, E. M., Savitz, D. A., and Shy, C. M. Spontaneous abortions among cosmetologists. Epidemiology. 1994;5:(2):147-155.

83. Collins, J. J., Ness, R., Tyl, R. W., Krivanek, N., Esmen, N. A., and Hall, T. A. A review of adverse pregnancy outcomes and formaldehyde exposure in human and animal studies. Regul Toxicol Pharmacol. 2001;34:(1):17-34.

84. NAC AEGL Committee. Interim acute exposure guideline levels (AEGLs) for formaldehyde. 2008.

85. Bender, J. R., Mullin, L. S., Graepel, G. J., and Wilson, W. E. Eye irritation response of humans to formaldehyde. Am Ind Hyg Assoc J. 1983;44:(6):463-465.

86. Green, D. J., Sauder, L. R., Kulle, T. J., and Bascom, R. Acute response to 3.0 ppm formaldehyde in

exercising healthy nonsmokers and asthmatics. Am Rev Respir.Dis. 1987;135:(6):1261-1266. 87. Sauder, L. R., Chatham, M. D., Green, D. J., and Kulle, T. J. Acute pulmonary response to formaldehyde

exposure in healthy nonsmokers. J Occup Med. 1986;28:(6):420-424. 88. Sheppard, D., Eschenbacher, W. L., and Epstein, J. Lack of bronchomotor response to up to 3 ppm

formaldehyde in subjects with asthma. Environ Res. 1984;35:(1):133-139. 89. Til, H. P., Woutersen, R. A., Feron, V. J., Hollanders, V. H., Falke, H. E., and Clary, J. J. Two-year

drinking-water study of formaldehyde in rats. Food Chem Toxicol. 1989;27:(2):77-87.

Panel Book Page 35

24

90. Takahashi, M., Hasegawa, R., Furukawa, F., Toyoda, K., Sato, H., and Hayashi, Y. Effects of ethanol, potassium metabisulfite, formaldehyde and hydrogen peroxide on gastric carcinogenesis in rats after initiation with N-methyl-N'-nitro-N-nitrosoguanidine. Jpn J Cancer Res. 1986;77:(2):118-124.

91. Rumchev, K. B., Spickett, J. T., Bulsara, M. K., Phillips, M. R., and Stick, S. M. Domestic exposure to

formaldehyde significantly increases the risk of asthma in young children. Eur Respir.J. 2002;20:(2):403-408.

92. Garrett, M. H., Hooper, M. A., Hooper, B. M., Rayment, P. R., and Abramson, M. J. Increased risk of

allergy in children due to formaldehyde exposure in homes. Allergy. 1999;54:(4):330-337. 93. Krzyzanowski, M., Quackenboss, J. J., and Lebowitz, M. D. Chronic respiratory effects of indoor

formaldehyde exposure. Environ Res. 1990;52:(2):117-125.

Figure 1. Declining use of formaldehyde in cosmetic products as reported to the FDA VCRP

(note x-axis is not linear…).

0

100

200

300

400

500

600

700

800

900

1984 2001 2002 2006 2007 2009 2010

Formaldehyde use

Formaldehyde use

Panel Book Page 36

25

Table 1a. Current and historical uses and concentrations of formaldehyde and formalin (combined) in cosmetics.

Product Category

2002 uses3

2010 uses10

2002 use concentrations3

(%)

2010 use

concentrations†

Bath Preparations Oils, tablets and salts 6 1 0.08 Bubble baths 4 1 0.08 Soaps and detergents 5 5 <0.002 – 0.08 Other bath preparations 1 -

0.08

Eye Makeup Preparations Mascara -

-

0.0002

Fragrance Preparations Other fragrance preparations -

-

0.02

Non-coloring Hair Preparations Hair conditioners 11 - - Permanent waves 2 2 - Rinses 2 - - Shampoos 59 13 <0.005 – 0.08 Hair tonics, dressings, etc. 9 6 <0.005 Wave sets 8 - - Other non-coloring hair preparations 3

7

-

Hair Coloring Preparations Shampoos 2 - - Other hair coloring preparations -

2

-

Makeup Preparations Leg and body paints - - 0.02 Other makeup preparations

- - 0.01

Nail Care Products Creams and lotions 1 - - Other nail care products 1

6

2‡

Oral Hygiene Products Dentifrices

- - 0.04

Personal Hygiene Products Other personal hygiene products 1 2 0.07 – 0.08

Shaving Preparations

Shaving cream

1 1 -

Skin Care Preparations Skin cleansing creams, lotions, liquids, and pads

1 1 0.0001 – 0.002

Depilatories - 2 - Body and hand skin care preparations 2 2 0.0001 Moisturizers 1 1 - Paste masks (mud packs) - 1 - Other skin care preparations -

5

0.06

Total uses/ranges for formaldehyde 120

78

<0.0001 – 0.08%

† 2010 survey underway

‡ product sold only in Europe and no longer marketed

Panel Book Page 37

26

Table 1b. Frequency and Concentration of Use Table Formaldehyde and Formalin

No. of Uses

(2010)10 Conc. of Use (2010) (%)

No. of Uses (2010)10

Conc. of Use (2010) (%)

Formaldehyde Formaldehyde Solution

(Formalin)

Totals 58 NS 20 NS

Duration of Use

Leave-On 29 NS 1 NS

Rinse Off 29 NS 19 NS

Exposure Type

Eye Area NR NS NR NS

Possible Ingestion NR NS NR NS

Inhalation NR NS NR NS

Dermal Contact 13 NS 11 NS

Deodorant (Underarm) NR NS NR NS

Hair - Non-Coloring 35 NS 9 NS

Hair – Coloring 2 NS NR NS

Nail 8 NS NR NS

Mucous Membrane 3 NS 6 NS

Bath Products NR NS 2 NS

Baby Products NR NS NR NS

NR = Not Reported; NS = Not Surveyed; Totals = Rinse-off + Leave-on Product Uses.

Note: Because each ingredient may be used in cosmetics with multiple exposure types, the sum of all exposure type uses may not equal the sum total uses.

Panel Book Page 38

27

Table 2. Uses of formaldehyde, formalin, and methylene glycol reported to Canada.11

Table. Use and concentration data from Health Canada.

Formaldehyde Formaldehyde

Solution (Formalin) Methylene Glycol

# of Uses Conc. of Use (%)

# of Uses

Conc. of Use (%)

# of Uses

Conc. of Use (%)

Totals 336 <0.1-30 15 <0.1-1 2 0.1-3

Duration of Use

Leave-On 146 <0.1-10 4 0.1-1 1 1-3

Rinse Off 190 <0.1-30 11 <0.1-0.3 1 0.1-0.3

Exposure Type

Eye Area 3 <0.1-0.3 NR NR NR NR

Possible Ingestion 1 <0.1 1 <0.1 NR NR

Inhalation NR NR NR NR NR NR

Dermal Contact 43 <0.1-30 11 <0.1-0.3 1 0.1-0.3 Deodorant (underarm) 1 3-10 NR NR NR NR

Hair - Non-Coloring 149 <0.1-10 NR NR NR NR

Hair-Coloring 1 0.1-0.3 NR NR NR NR

Nail 72 <0.1-10 3 0.1-1 1 1-3 Mucous Membrane 1 <0.1 1 <0.1 NR NR

Bath Products 33 <0.1-0.3 2 0.1-0.3 NR NR

Baby Products 6 <0.1, 0.3-

1 NR NR NR NR

Other* 11 <0.1-3 NR NR NR NR *Types of products not defined by Health Canada. NR = Not Reported; NS = Not Surveyed; Totals = Rinse-off + Leave-on Product Uses.

Table 3. List of ingredients in Brazilian Blowout from the Brazilian Blowout MSDS dated 10/26/10.

Ingredient Percentage Water ≤85% Methylene glycol <5% Behenyl methylammonium methosulfate/N-hexadecanol/butylene glycol ≤5% Isoparaffin ≤3% Cetrimonium chloride ≤2% Petrolatum ≤1% Hypnea musciformis extract/Gellidiela acerosa extract/Sargassum filipendula extract/sorbitol ≤1% Theobroma grandiflorum seed butter (cupuacu butter) ≤0.5% Panthenol ≤0.25% Hydrolyzed keratin ≤1% Fragrance (parfum) ≤1% Methylchloroisothiazolinone ≤0.1% Methylisothiazolinone ≤0.1%

Panel Book Page 39

Panel Book Page 40

FORMALDEHYDE NAC/Interim 1: 07/2008; Page 14 of 71

TABLE 3. Irritant Effects of Formaldehyde in Controlled Human Studies Concentration

(ppm) Time Subjects/Effect

(number of subjects) Reference 0.9, 1.0 non-responders at 1.3 or 2.2 ppm;

Eye irritation evaluated: average scores of none to slight at 0.35 to 0.9 ppm; slight to moderate at 1.0 ppm; slight adaptation with time

0, 0.10, 0.69 90 min Asthmatic nonsmoking subjects (15): No significant change in pulmonary function parameters (FEV1 and airway resistance) or in bronchial reactivity; no association of subjective ratings of asthmatic symptoms with increasing air concentrations

Harving et al. 1986; 1990

0, 0.41 2 h Healthy occupationally exposed (5) and contact dermatitis subjects (13): No effect on pulmonary parameters (VC, FEV1); immune response in subjects with contact dermatitis (increased chemiluminescense of neutrophils)

Gorski et al 1992

0, 0.41 2 h Healthy (11) and patients with skin hypersensitivity to formaldehyde (9) (all nonsmokers): No differences in response between groups; transient increase in symptoms of sneezing, rhinorrhea, or eye irritation; nasal washings showed increases in eosinophils, albumin, total protein, but not neutrophil, basophil or mononuclear cells

Pazdrak et al. 1993

0, 0.41 2 h Healthy, non-occupationally exposed (10) and occupationally exposed asthmatic subjects (10): No differences in response between groups; transient increase in symptoms of sneezing, rhinorrhea, edema, or itchy eyes; increases in leucocytes and eosinophils in nasal washings; no allergic response; no clinical symptoms of bronchial irritation or effects on pulmonary function parameters (FEV1, PEF)

Krakowiak et al. 1998

0, 0.40 1 hr 12 volunteers with intermittent asthma and allergy to pollen: No change in lung function (FEV1); no enhanced response to allergens

Ezratty et al. 2007

0, 0.17, 0.39, 0.9 5.5 h Formaldehyde exposed workers (32); controls (29): subjective symptoms (headache, tiredness) did not correlate with exposure; no clear effect of concentration on memory; some concentration-related effects in a few tests (additional speed, response time) but limitations in experimental design and control issues

Bach et al. 1990

1.0

90 min Healthy (9) and formaldehyde-sensitive (9) subjects (previously complained about non-

Day et al. 1984

Panel Book Page 41

FORMALDEHYDE NAC/Interim 1: 07/2008; Page 15 of 71

TABLE 3. Irritant Effects of Formaldehyde in Controlled Human Studies Concentration

(ppm) Time Subjects/Effect

(number of subjects) Reference respiratory effects of urea formaldehyde foam insulation): No effects on pulmonary function parameters (FVC, FEV1, max and mid-exploratory flow rate); complaints of eye irritation, nasal congestion, tearing, and throat irritation; no severity index

0, 1.0 3 h Control asthmatic subjects (4); subjects with asthma attributed to urea formaldehyde foam (23): no differences between groups in immunologic parameters, either before or after exposure; minor immunologic changes in both groups postexposure

Pross et al. 1987

0, 0.2, 0.4, 0.8, 1.6 (no concurrent control)

5 h Healthy subjects (16): No differences in nasal airway resistance or pulmonary function parameters; decrease in nasal mucus flow at all concentrations; no discomfort at 0.2 or 0.4 ppm for 2 hours (some slight discomfort reported in the 3 to 5 hours period [conjunctival irritation, dryness of nose and throat] but discomfort rated higher at 0.2 ppm than at 0.4 ppm and only 5 or fewer subjects reported any discomfort); average discomfort scored as slight during exposure to 1.6 ppm and first noted in the latter part of the first hour but decreased somewhat after three hours; no effect on performance on mathematical tests or number transfer tasks

Andersen and Molhave 1983

0, 0.15, 0.3, 0.5 ppm; 0.3 with 4 peaks to 0.6 ppm; 0.5 ppm with 4 peaks to 1.0 ppm; some exposures combined with ethyl acetate as masking agent

4 h Healthy volunteers, 11 males and 10 females: All concentrations: no significant effects on nasal flow and resistance, pulmonary function, and decision reaction time; slight to moderately increased blinking frequency and conjunctival redness at 0.5 ppm with peaks to 1.0 ppm; subjective eye and olfactory symptoms reported at 0.3 ppm (no-effect level when “negative affectively” considered); subjective nasal irritation at 0.5 ppm with peaks to 1.0 ppm

Lang et al. 2008

0, 2.0 (at rest) 0, 2.0 (exercise)

40 min Healthy (15) and asthmatic (15) non-smoking subjects: No significant decrement in pulmonary function parameters (flow-volume parameters and airway resistance) or bronchial reactivity both at rest and with exercise; subjective symptoms ranged up to severe (but not incapacitating) for odor for some individuals, but median scores for nose, throat and eye irritation were #moderate; no increase in symptomology

Witek et al. 1986; 1987; Schachter et al. 1985; 1986

Panel Book Page 42

FORMALDEHYDE NAC/Interim 1: 07/2008; Page 16 of 71

TABLE 3. Irritant Effects of Formaldehyde in Controlled Human Studies Concentration

(ppm) Time Subjects/Effect

(number of subjects) Reference with exercise

0, 0.1, 1.0, 3.0 20 min Asthmatic patients who suspected formaldehyde as the cause (13): No significant difference in pulmonary function parameters (FEV1, VC); no asthmatic response to formaldehyde challenge

Frigas et al. 1984

0, 0.5, 1.0, 2.0, 3.0 at rest; 2.0 with exercise

3 h Healthy non-smoking subjects (19): (9 exposed to 3 ppm and 10 exposed to 0.05 ppm) No significant decrements in pulmonary function parameters (FVC, FEV1, FEF25-75%, SGaw) or increases in bronchial reactivity (methacholine challenge) at any concentration; nasal flow resistance increased at 3.0 ppm; significant dose-response relationship for odor sensation and eye irritation, but eye irritation scored mild (5/9) or mild to moderate 4/9) at 3 ppm; eye irritation began at 1 ppm

Kulle et al. 1987; Kulle 1993

0, 3.0 ppm With heavy exercise (healthy subjects); moderate exercise (asthmatic subjects)

1 h Healthy (22) and asthmatic (16) non-smoking subjects: No difference in symptoms between groups; eye, nose and throat irritation scored mild to mild-moderate (group means); small decreases in some pulmonary function parameters in healthy individuals engaging in heavy exercise

Green et al. 1987

0, 3.0 With heavy exercise (15 minutes every half hour)

2 h Healthy non-smoking subjects (24): Increase in subjective symptoms of eye, nose and throat irritation, rated mild to moderate on average; small, but statistically significant increase in two (FEF25-75%, SGaw) of several pulmonary function measurements at some time intervals (no effect on FEV1, FVC, FEV3), no increase in cough

Green et al. 1989

0, 3.0 With intermittent exercise

3 h Healthy non-smoking subjects (9) non-biologically significant, transient change in some pulmonary function parameters (FEV1, FEF25-75%); increase in nose/ throat and eye irritation, rated mild to moderate by individuals; only one subject rated eye irritation as moderate

Sauder et al. 1986

0, 3.0 3 h Asthmatic non-smoking subjects (9): no significant group change in pulmonary function parameters (FEV1, FVC, FEF25-75%, SGaw, or FRC) or airway reactivity; significant increase in nose, throat (at 30 minutes), and eye irritation (at 60 minutes), rated as none to mild-moderate except for one subject who reported severe eye irritation

Sauder et al. 1987

Panel Book Page 43

FORMALDEHYDE NAC/Interim 1: 07/2008; Page 17 of 71

TABLE 3. Irritant Effects of Formaldehyde in Controlled Human Studies Concentration

(ppm) Time Subjects/Effect

(number of subjects) Reference 0, 1, 3 10 min Asthmatic non-smoking subjects (7):

Similar responses in airway resistance following exposure to 0, 1, or 3 ppm with and without exercise (exercise increased all responses)

Sheppard et al. 1984

0.03 to 3.2; 0, 1.0, 2.0, 2.9, 4.0 or 1.2, 2.1, 2.8, and 4.0

37 min (n = 33); 1.5 min (n = 48)

Healthy subjects (exposure groups of 33 and 48): Poorer air quality and greater nose irritation reported during the short exposures than during the 37-minute exposure, whereas the opposite was true for eye irritation; with increasing concentrations, both eye and nose irritation increased from none to Aa little;@ objectively-measured eye blinking not affected at 1.2 ppm, but was statistically significantly increased at 2.1 ppm

Weber-Tschopp et al. 1977

0, 1, 2, 4, 5 5 min except for 2 ppm (12 min)

Healthy students (groups of 7 to 75): Addressed eye irritation only (subjects exposed via goggles): 1 ppm considered threshold for detection; 5 ppm produced severe eye irritation

Stephens et al. 1961

8, 13 Short exposures (<15 sec)

Healthy/atopic subjects (1-6): Eye irritation for 5 of 6 subjects at 12 ppm but not at 8 ppm for 4 of 5; irritation of the throat at both concentrations; changes in airway resistance

Douglas 1974

13.8 30 min Healthy male subjects (12): Nasal and eye irritation with mild lacrimation; adaptation to the eye irritation

Sim and Pattle 1957

20 Several min Healthy subjects (2): Lacrimation (within 15-30 seconds); eye, nose, and throat irritation considered objectionable

Barnes and Speicher 1942

FVC = 1 FEV1 = 2 FEF25-75%, = 3 FRC = 4 PEF = 5 SGaw = 6 7

In an early study that addressed the threshold for eye irritation, Shuck et al. (1966) 8 reported that a linear relationship between reported eye irritation and formaldehyde 9 concentration does not hold below 0.3 ppm. Most subjects experienced the same eye irritation at 10 0.05 and 0.5 ppm. The atmospheres in this study were generated by photooxidation of propylene 11 or ethylene in order to simulate photochemical air pollution. Although formaldehyde was 12 measured in the atmospheres, additional photochemical smog irritants were present. 13 14

In a similar study by the above group (Stephens et al. 1961), the eye irritation potential of 15 both photochemical smog and its individual constituents was examined. Healthy students in 16 groups of 7 to 75 were exposed to formaldehyde via eye goggles. Eye irritation (none, medium, 17

Panel Book Page 44

This document is a draft for review purposes only and does not constitute Agency policy.

4-99 DRAFT—DO

Tab

le 4

-2.

Maj

or c

ohor

t stu

dies

of f

orm

alde

hyde

exp

osur

e an

d na

soph

aryn

geal

can

cer

(with

2 o

r m

ore

case

s)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

(num

ber

of o

bser

ved

deat

hs)

Hau

ptm

ann

et a

l. (2

004)

[E

xten

sion

of N

CI s

tudy

by

Bla

ir et

al.,

198

7,

1986

)], U

nite

d St

ates

Ret

rosp

ectiv

e co

hort

mor

talit

y st

udy

of 2

5,61

9 w

orke

rs e

mpl

oyed

at 1

0 fo

rmal

dehy

de p

lant

s in

the

U.S

. fol

low

ed fr

om e

ither

pl

ant s

tart-

up o

r firs

t em

ploy

men

t thr

ough

199

4.

The

10 p

lant

s pro

duce

d fo

rmal

dehy

de (3

pla

nts)

, m

oldi

ng c

ompo

unds

(3

pla

nts)

, pho

togr

aphi

c fil

m (2

pla

nts)

, ply

woo

d (1

pla

nt),

and

form

alde

hyde

re

sins

(6 p

lant

s).

Tim

e-de

pend

ent e

xpos

ure

estim

ates

a bas

ed o

n jo

b tit

les,

task

s, vi

sits

to p

lant

s by

stud

y in

dust

rial

hygi

enis

ts, a

nd m

onito

ring

data

mea

sure

men

ts.

Peak

ex

posu

re =

shor

t-ter

m

excu

rsio

ns >

8-ho

ur T

WA

fo

rmal

dehy

de in

tens

ity

and

know

ledg

e of

job

task

s. W

orke

rs

cont

ribut

ed p

re-e

xpos

ure

pers

on ti

me

to n

onex

pose

d ca

tego

ry.

RR

s wer

e fr

om

Pois

son

regr

essi

on

mod

els,

usin

g a

15-y

ear

lag

to a

ccou

nt fo

r tum

or

late

ncy.

Ove

rall

Non

expo

sed

SM

R

1.56

(95

% C

I: 0.

39−2

3)

(2)

Ex

pose

d SM

R

2.10

(95

% C

I: 1.

05−2

1)

(8)

Peak

exp

osur

e (p

pm)

0

RR

b 1.

00 (

refe

rent

) (2

)

>0

to <

2.0

N

/A

(0)

2

.0 to

<4.

0

N/A

(0

)

4.0

or g

reat

er

1.

83 N

ot p

rovi

ded

(7)

Tren

d p

< 0

.001

(Tre

nd o

n ca

tego

rica

l dat

a)

Ave

rage

inte

nsity

of e

xpos

ure

(ppm

)

0

RR

b 1.

00 (

refe

rent

) (2

)

≤0.5

N/A

(0

)

0.5

to <

1.0

0.38

Not

pro

vide

d (1

)

1.0

or g

reat

er

1.

67 N

ot p

rovi

ded

(6)

Tren

d p

= 0

.066

(Tre

nd o

n co

ntin

uous

dat

a am

ong

expo

sed

only

) C

umul

ativ

e ex

posu

re (

ppm

-yea

rs)

0

RR

b 2.

40 N

ot p

rovi

ded

(2)

>0

to <

1.5

1.

00 (

refe

rent

) (3

)

1.5

to <

5.5

1.19

Not

pro

vide

d (1

)

5.5

or m

ore

4.14

Not

pro

vide

d (3

) Tr

end

p =

0.0

25 (T

rend

on

cont

inuo

us d

ata

amon

g ex

pose

d on

ly)

Dur

atio

n (y

ears

)

0 R

Rb

1.77

Not

pro

vide

d (2

)

>0 to

<5

1.

00 (

refe

rent

) (4

)

5 to

<15

0.83

Not

pro

vide

d (1

)

15 o

r mor

e

4.18

Not

pro

vide

d (2

) Tr

end

p =

0.1

47 (T

rend

on

cont

inuo

us d

ata

amon

g ex

pose

d on

ly)

Panel Book Page 45

This document is a draft for review purposes only and does not constitute Agency policy.

4-100 DRAFT—DO

Tab

le 4

-2.

Maj

or c

ohor

t stu

dies

of f

orm

alde

hyde

exp

osur

e an

d na

soph

aryn

geal

can

cer

(with

2 o

r m

ore

case

s)

R

efer

ence

St

udy

desi

gn

Exp

osur

e as

sess

men

t R

esul

ts (n

umbe

r of

obs

erve

d de

aths

) M

arsh

et a

l. (2

002)

, C

onne

ctic

ut, U

nite

d St

ates

Ret

rosp

ectiv

e co

hort

mor

talit

y st

udy

of 7

,328

w

orke

rs h

ired

up to

198

4

and

follo

wed

unt

il 19

98 in

on

e pl

ant f

rom

Hau

ptm

ann

et a

l. (2

004)

. M

orta

lity

was

co

mpa

red

with

dea

th ra

tes

in tw

o C

onne

ctic

ut c

ount

ies

and

the

U.S

.

Wor

ker-

spec

ific

expo

sure

a fr

om jo

b ex

posu

re m

atrix

ba

sed

on a

vaila

ble

spor

adic

sam

plin

g da

ta

from

196

5−19

87, j

ob

desc

riptio

ns, a

nd v

erba

l jo

b de

scrip

tions

by

plan

t pe

rson

nel a

nd in

dust

rial

hygi

enis

ts.

Expo

sure

s ra

nked

on

a 7-

poin

t sca

le

with

exp

osur

e ra

nge

assi

gned

to e

ach

rank

. 17

% o

f job

s val

idat

ed w

ith

com

pany

mon

itorin

g da

ta;

rem

aini

ng 8

3% b

ased

on

prof

essi

onal

judg

men

t.

Ass

umed

pre

-196

5 ex

posu

re le

vels

sam

e as

po

st-1

965

leve

ls.

Ove

rall

U

.S. r

efer

ent

SMR

4.

94 (

95%

CI:

1.99

−10)

(7

)

Cou

nty

refe

rent

5.00

(95

% C

I: 2.

01−1

0)

(7)

Shor

t-ter

m w

orke

r (<1

ye

ar)

5.

35 (

95%

CI:

1.46

−14)

(4

) Lo

ng-te

rm w

orke

r (<1

ye

ar)

4.

59 (

95%

CI 0

.95−

13)

(3)

Fo

rmal

dehy

de e

xpos

ure

SMR

6.

03 (

95%

CI:

2.42

−12.

42)

(7)

Dur

atio

n of

form

alde

hyde

exp

osur

e (y

ears

)

0

to <

1 SM

R

5.84

(95

% C

I: 1.

59−1

5)

(4)

1−

9

3.

17 (

95%

CI:

0.08

−18)

(1

)

10+

12.5

(95

% C

I: 1.

51−4

5)

(2)

Cum

ulat

ive

expo

sure

(ppm

-yea

rs) c

ount

y

0 to

<0.

004

SMR

3.

97 (

95%

CI:

0.10

−22)

(1

)

0.00

4−0.

219

5.

89 (

95%

CI:

1.22

−17)

(3

)

0.22

+

7.51

(95

% C

I: 1.

55−2

2)

(3)

Ave

rage

inte

nsity

exp

osur

e (p

pm)

0

to <

0.03

SM

R

2.41

(95

% C

I: 0.

06−1

3)

(1)

0.

03−0

.159

15.3

(95

% C

I: 4.

16−3

9)

(4)

0.

16+

4.

13 (

95%

CI:

0.50

−15)

(2

) D

urat

ion

of e

xpos

ure

to >

0.2

ppm

(yea

rs)

U

nexp

osed

SM

R

3.01

(95

% C

I: 0.

36−1

1)

(2)

0

to <

1

4.81

(95

% C

I: 0.

58−1

7)

(2)

1−

9

4.

04 (

95%

CI:

0.10

−231

) (1

)

10+

27.6

(95

% C

I: 3.

34−1

00)

(2)

Panel Book Page 46

This document is a draft for review purposes only and does not constitute Agency policy.

4-101 DRAFT—DO

Tab

le 4

-2.

Maj

or c

ohor

t stu

dies

of f

orm

alde

hyde

exp

osur

e an

d na

soph

aryn

geal

can

cer

(with

2 o

r m

ore

case

s)

R

efer

ence

St

udy

desi

gn

Exp

osur

e as

sess

men

t R

esul

ts (n

umbe

r of

obs

erve

d de

aths

) M

arsh

et a

l. (2

002)

, C

onne

ctic

ut, U

nite

d St

ates

(con

tinue

d)

Dur

atio

n of

exp

osur

e to

≥0.

7 pp

m (y

ears

)

Une

xpos

ed

SMR

3.

64 (

95%

CI:

0.99

−9.3

1)

(4)

<1

9.51

(95

% C

I: 1.

15−3

4)

(2)

1+

11.1

(95

% C

I: 0.

28−6

2)

(1)

Hay

es e

t al.

(199

0),

Uni

ted

Stat

es

Prop

ortio

nate

mor

talit

y co

hort,

n =

4,0

46 m

ale

emba

lmer

s and

fune

ral

dire

ctor

s, di

ed 1

975−

1985

.

Expo

sure

pre

sum

ed.

Ove

rall

PMR

2.

16 (

95%

CI:

0.59

−5.5

4)

(4)

H

anse

n an

d O

lsen

(199

5),

Den

mar

k Pr

opor

tiona

te in

cide

nce

stud

y of

2,0

41 m

en w

ith

canc

er w

ho d

ied

betw

een

1970

and

198

4, id

entif

ied

from

the

Dan

ish

Can

cer

Reg

istry

and

mat

ched

with

th

e D

anis

h Su

pple

men

tary

Pe

nsio

n Fu

nd, w

hose

lo

nges

t wor

k ex

perie

nce

occu

rred

at l

east

10

year

s be

fore

the

canc

er d

iagn

osis

. Th

e SP

IR m

easu

red

the

prop

ortio

n of

cas

es in

fo

rmal

dehy

de-a

ssoc

iate

d co

mpa

nies

rela

tive

to th

e pr

opor

tion

of c

ases

am

ong

all e

mpl

oyee

s in

Den

mar

k.

Link

ed c

ompa

nies

thro

ugh

tax

reco

rds t

o th

e na

tiona

l D

anis

h Pr

oduc

t Reg

iste

r.

Ove

rall

SP

IR

1.3

(95%

CI:

0.03

−3.2

) (4

)

a Ex

posu

re e

stim

ates

by

Hau

ptm

ann

et a

l. (2

004)

wer

e 10

tim

es h

ighe

r tha

n th

ose

of M

arsh

et a

l. (2

002,

p. 2

59).

b A

djus

ted

for c

alen

dar y

ear,

age,

sex,

race

, and

pay

cat

egor

y (s

alar

ied

vers

us w

age)

. C

onfid

ence

inte

rval

s not

pro

vide

d by

aut

hors

, but

wer

e de

scrib

ed a

s in

clud

ing

1.0.

Panel Book Page 47

This document is a draft for review purposes only and does not constitute Agency policy.

4-152 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Har

ringt

on a

nd

Shan

non

(197

5)

Coh

ort m

orta

lity

stud

y of

2,0

79

path

olog

ists

and

12,

944

labo

rato

ry

tech

nici

ans f

rom

the

Roy

al

Col

lege

of P

atho

logi

sts a

nd th

e Pa

thol

ogic

al S

ocie

ty o

f Gre

at

Brit

ain

from

195

5−19

73.

The

com

paris

on p

opul

atio

n ca

me

from

na

tiona

l mor

talit

y da

ta.

Pres

umed

exp

osur

e to

fo

rmal

dehy

de ti

ssue

fix

ativ

e.

Pa

thol

ogis

ts

All

caus

e m

orta

lity

SMR

0.6

0 N

R (1

56)

LH

P ca

ncer

s SM

R 2

.0

p <

0.01

(8

)

H

odgk

in’s

dis

ease

SM

R 1

.4

NR

(1)

Le

ukem

ia

SMR

0.6

N

R (1

)

Te

chni

cian

s

All

caus

e m

orta

lity

SMR

0.6

7 N

R (1

54)

LH

P ca

ncer

s SM

R 0

.5

NR

(3)

H

odgk

in’s

dis

ease

SM

R −

N

R (0

)

Le

ukem

ia

SMR

0.5

N

R (1

)

Har

ringt

on a

nd O

akes

(1

984)

C

ohor

t mor

talit

y st

udy

of 2

,720

pa

thol

ogis

ts fr

om th

e R

oyal

C

olle

ge o

f Pat

holo

gist

s and

the

Path

olog

ical

Soc

iety

of G

reat

B

ritai

n fr

om 1

974−

1980

. V

ital

stat

us o

btai

ned

from

the

cens

us, a

na

tiona

l hea

lth re

gist

ry, a

nd o

ther

so

urce

s. S

MR

s dev

elop

ed fr

om

the

Engl

ish,

Sco

ttish

, Iris

h, a

nd

Wel

sh p

opul

atio

ns.

Pres

umed

exp

osur

e to

fo

rmal

dehy

de ti

ssue

fix

ativ

e.

All

caus

es

M

en

SMR

0.5

6 (9

0% C

I: 0.

48−0

.66)

(1

10)

W

omen

SM

R 0

.99

(90%

CI:

0.62

−1.5

0)

(16)

Leuk

emia

M

en

SMR

0.9

1 (9

0% C

I: 0.

05−4

.29)

(1

)

W

omen

SM

R 9

.26

(90%

CI:

0.47

−43.

9)

(1)

Oth

er L

HP

canc

ers

M

en

SMR

0.5

3 (9

0% C

I: 0.

03−2

.54)

(1

)

W

omen

SM

R −

−

(0)

Panel Book Page 48

This document is a draft for review purposes only and does not constitute Agency policy.

4-153 DRAFT—DO

T

able

4-7

. E

pide

mio

logi

c st

udie

s of f

orm

alde

hyde

and

lym

phoh

emat

opoi

etic

can

cers

(con

tinue

d)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Hal

l et a

l. (1

991)

C

ohor

t mor

talit

y st

udy

of 4

,512

pa

thol

ogis

ts fr

om th

e R

oyal

C

olle

ge o

f Pat

holo

gist

s and

the

Path

olog

ical

Soc

iety

of G

reat

B

ritai

n fr

om 1

974−

1987

. V

ital

stat

us o

btai

ned

from

the

cens

us, a

na

tiona

l hea

lth re

gist

ry, a

nd o

ther

so

urce

s. S

MR

s dev

elop

ed fr

om

the

Engl

ish

and

Wel

sh

popu

latio

ns.

Pres

umed

exp

osur

e to

fo

rmal

dehy

de ti

ssue

fix

ativ

e.

All

caus

e m

orta

lity

M

en

SMR

0.4

3 (9

5% C

I: 0.

37−0

.50)

(1

76)

W

omen

SM

R 0

.65

(95%

CI:

0.38

−1.0

3)

(18)

Hod

gkin

’s d

isea

se

SMR

1.2

1 (9

5% C

I: 0.

03−6

.71)

(1

)

All

canc

ers

SMR

1.4

4 (9

5% C

I: 0.

69−2

.63)

(1

0)

Leuk

emia

SM

R 1

.52

(95%

CI:

0.41

−3.8

9)

(4)

Levi

ne e

t al.

(198

4)

Coh

ort m

orta

lity

stud

y of

1,4

77

mal

e O

ntar

io u

nder

take

rs fi

rst

licen

sed

1928

−195

7, fo

llow

ed

from

195

0 to

197

7. S

MR

s de

velo

ped

from

Ont

ario

mor

talit

y ra

tes.

Pres

umed

exp

osur

e to

fo

rmal

dehy

de ti

ssue

fix

ativ

e.

All

LHP

canc

ers

SMR

1.2

4 N

R (8

)

Leuk

emia

SM

R 1

.60

NR

(4)

Stro

up e

t al.

(198

6)

Coh

ort m

orta

lity

stud

y of

2,3

17

whi

te m

ale

mem

bers

of t

he

Am

eric

an A

ssoc

iatio

n of

A

nato

mis

ts fr

om 1

888

to 1

969

who

die

d 19

25−1

979.

SM

Rs

deve

lope

d us

ing

U.S

. pop

ulat

ion

mor

talit

y ra

tes.

Pres

umed

exp

osur

e to

fo

rmal

dehy

de ti

ssue

fix

ativ

e.

All

caus

e m

orta

lity

SMR

0.6

5 (9

5% C

I: 0.

60−0

.70)

(7

38)

All

LHP

canc

ers

SMR

1.2

(9

5% C

I: 0.

7−2.

0)

(18)

Lym

phos

arco

ma

and

retic

ulos

arco

ma

SMR

0.7

(9

5% C

I: 0.

1−2.

5)

(2)

Hod

gkin

’s d

isea

se

SMR

−

− (0

)

Leuk

emia

SM

R 1

.5

(95%

CI:

0.7−

2.7)

(1

0)

Oth

er ly

mph

atic

SM

R 2

.0

(95%

CI:

0.7−

4.4)

(6

)

Panel Book Page 49

This document is a draft for review purposes only and does not constitute Agency policy.

4-154 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Logu

e et

al.

(198

6)

Coh

ort m

orta

lity

stud

y of

5,5

85

path

olog

ists

who

wer

e m

embe

rs o

f th

e C

olle

ge o

f Am

eric

an

Path

olog

ists

, 196

2−19

72,

follo

wed

for m

orta

lity

thro

ugh

1977

. SM

Rs d

evel

oped

from

U.S

. po

pula

tion

mor

talit

y ra

tes.

Pres

umed

exp

osur

e to

fo

rmal

dehy

de ti

ssue

fix

ativ

e.

LHP

canc

er o

ther

than

le

ukem

ia

SM

R 0

.48

NR

(NR

)

Leuk

emia

SM

R 1

.06

NR

(NR

)

Mat

anos

ki (1

991)

C

ohor

t mor

talit

y st

udy

of 6

,111

m

ale

path

olog

ists

from

m

embe

rshi

p ro

lls o

f the

Am

eric

an

Med

ical

Ass

ocia

tion

1912

−195

0.

Mor

talit

y w

as fo

llow

ed th

roug

h 19

78.

SMR

s dev

elop

ed fr

om U

.S.

popu

latio

n w

hite

mal

e m

orta

lity

rate

s.

Pres

umed

exp

osur

e to

fo

rmal

dehy

de ti

ssue

fix

ativ

e.

All

canc

er

SMR

0.7

8 (9

5% C

I: 0.

71−0

.85)

(5

08)

All

LHP

canc

ers

SMR

1.2

5 (9

5% C

I: 0.

95−1

.62)

(5

7)

Lym

phos

arco

ma

and

retic

ulos

arco

ma

SMR

1.3

1 (9

5% C

I: 0.

66−2

.35)

(1

1)

Hod

gkin

’s d

isea

se

SMR

0.

36

(95%

CI:

0.04

−1.3

1)

(2)

Leuk

emia

SM

R 1

.35

(95%

CI:

0.92

−1.9

2)

(31)

Oth

er ly

mph

atic

SM

R 1

.54

(95%

CI:

0.82

−2.6

3)

(13)

Bea

ne F

reem

an e

t al.

(200

9)

Prev

ious

repo

rts:

Hau

ptm

ann

et a

l. (2

003)

B

lair

et a

l. (1

986)

Ret

rosp

ectiv

e co

hort

mor

talit

y st

udy

of 2

5,61

9 w

orke

rs e

mpl

oyed

at

10

form

alde

hyde

pla

nts i

n th

e U

.S. f

ollo

wed

from

eith

er th

e pl

ant s

tart-

up o

r firs

t em

ploy

men

t th

roug

h 20

04.

SMR

s cal

cula

ted

usin

g se

x-, a

ge-,

race

-, an

d ca

lend

ar-y

ear-

spec

ific

U.S

. mor

talit

y ra

tes.

RR

s est

imat

ed u

sing

Poi

sson

re

gres

sion

stra

tifie

d by

cal

enda

r ye

ar, a

ge, s

ex, a

nd ra

ce; a

djus

ted

for p

ay c

ateg

ory.

Expo

sure

est

imat

es b

ased

on

job

title

s, ta

sks,

visi

ts

to p

lant

s by

stud

y in

dust

rial h

ygie

nist

s, an

d m

onito

ring

data

thro

ugh

1980

. Pe

ak e

xpos

ure

defin

ed a

s sho

rt-te

rm

excu

rsio

ns e

xcee

ding

the

8-ho

ur T

WA

fo

rmal

dehy

de in

tens

ity

and

know

ledg

e of

job

task

s. E

xpos

ures

to 1

1 ot

her c

ompo

unds

wer

e

All

LHP

canc

ers

Ex

pose

d SM

R 0

.94

(95%

CI:

0.84

−1.0

6)

(286

)

U

nexp

osed

SM

R 0

.86

(95%

CI:

0.61

−1.5

2)

(33)

Peak

exp

osur

e (p

pm)

0

RR

1.0

7 (9

5% C

I: 0.

70−1

.62)

(3

3)

0.

1−1.

9 R

R 1

.00

Refe

renc

e va

lue

(103

)

2.

0 to

<4.

0 R

R 1

.17

(95%

CI:

0.86

−1.5

9)

(75)

4.

0 or

gre

ater

R

R 1

.37

(95%

CI:

1.03

−1.8

1)

(108

)

Tren

d p

= 0

.02

Ave

rage

exp

osur

e (p

pm)

Panel Book Page 50

This document is a draft for review purposes only and does not constitute Agency policy.

4-155 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Bea

ne F

reem

an e

t al.

(200

9) (c

ontin

ued)

iden

tifie

d. W

orke

rs

cont

ribut

ed p

re-e

xpos

ure

pers

on-ti

me

to

none

xpos

ed c

ateg

ory.

Po

isso

n re

gres

sion

m

odel

s use

d a

2-ye

ar la

g to

acc

ount

for t

umor

la

tenc

y.

0

R

R 0

.99

(95%

CI:

0.66

−1.4

8)

(33)

0.

1−0.

4 R

R 1

.00

Refe

renc

e va

lue

(164

)

0.

5 to

<1.

0 R

R 1

.29

(95%

CI:

0.97

−1.7

3)

(67)

1.

0 or

gre

ater

R

R 1

.07

(95%

CI:

0.78

−1.4

7)

(55)

Tren

d p

> 0

.50

Cum

ulat

ive

expo

sure

(ppm

-yea

rs)

0

RR

0.8

9 (9

5% C

I: 0.

59−1

.34)

(3

3)

0.

1−1.

4 R

R 1

.00

Refe

renc

e va

lue

(168

)

1.

5 to

5.4

R

R 0

.77

(95%

CI:

0.56

−1.0

7)

(49)

5.

5 or

gre

ater

R

R 1

.07

(95%

CI:

0.80

−1.4

2)

(69)

Tren

d p

= 0

.25

Leuk

emia

Peak

exp

osur

e (p

pm)

0

RR

0.5

9 (9

5% C

I: 0.

25−1

.36)

(7

)

0.

1−1.

9 R

R 1

.00

Refe

renc

e va

lue

(41)

2

.0 to

<4.

0

RR

0.9

8 (9

5% C

I: 0.

60−1

.62)

(2

7)

4

.0 o

r gre

ater

R

R 1

.42

(95%

CI:

0.92

−2.1

8)

(48)

Tren

d p

= 0

.012

Ave

rage

exp

osur

e (p

pm)

0

ppm

R

R 0

.54

(95%

CI:

0.24

−1.2

2)

(7)

0.

1−0.

4 R

R 1

.00

Refe

renc

e va

lue

(67)

0.

5 to

<1.

0

RR

1.1

3 (9

5% C

I: 0.

71−1

.79)

(2

5)

1.

0 or

gre

ater

R

R 1

.10

(95%

CI:

0.68

−1.7

8)

(24)

Panel Book Page 51

This document is a draft for review purposes only and does not constitute Agency policy.

4-156 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Bea

ne F

reem

an e

t al.

(200

9) (c

ontin

ued)

Tr

end

p >

0.5

0

Cum

ulat

ive

expo

sure

(pp

m-y

ears

)

0

RR

0.5

3 (9

5% C

I: 0.

23−1

.21)

(7

)

0.

1−1.

4 R

R 1

.00

Refe

renc

e va

lue

(63)

1.

5−5.

4 R

R 0

.96

(95%

CI:

0.60

−1.5

6)

(24)

5.

5 or

gre

ater

R

R 1

.11

(95%

CI:

0.70

−1.7

4)

(29)

Tren

d p

= 0

.12

Hod

gkin

Lym

phom

a

Peak

exp

osur

e (p

pm)

0

RR

0.6

7 (9

5% C

I: 0.

12−3

.60)

(2

)

0.

1−1.

9 R

R 1

.00

Refe

renc

e va

lue

(6)

2.

0 to

<4.

0 R

R 3

.30

(95%

CI:

1.04

−10.

50)

(8)

4.

0 or

gre

ater

R

R 3

.96

(95%

CI:

1.31

−12.

02)

(11)

Tren

d p

= 0

.01

Ave

rage

exp

osur

e (p

pm)

0

R

R 0

.46

(95%

CI:

0.05

−3.9

3)

(2)

0.

1−0.

4 R

R 1

.00

Refe

renc

e va

lue

(10)

0.

5 to

<1.

0 R

R 3

.62

(95%

CI:

1.41

−9.3

1)

(9)

1.

0 or

gre

ater

R

R 2

.48

(95%

CI:

0.84

−7.3

2)

(6)

Tren

d p

= 0

.05

Cum

ulat

ive

(ppm

-yea

rs)

0

RR

0.4

2 (9

5% C

I: 0.

09−2

.05)

(2

)

0.

1−1.

4 R

R 1

.00

Refe

renc

e va

lue

(14)

Panel Book Page 52

This document is a draft for review purposes only and does not constitute Agency policy.

4-157 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Bea

ne F

reem

an e

t al.

(200

9) (c

ontin

ued)

1.5−

5.4

RR

1.7

1 (9

5% C

I: 0.

66−4

.38)

(7

)

5.

5 or

gre

ater

R

R 1

.30

(95%

CI:

0.40

−4.1

9)

(4)

Tren

d p

= 0

.08

Mye

loid

Leu

kem

ia

Peak

exp

osur

e (p

pm)

0

RR

0.8

2 (9

5% C

I: 0.

25−2

.67)

(4

)

0.

1 to

1.9

RR

1.0

0 Re

fere

nce

valu

e (1

4)

2.

0 to

<4.

0

RR

1.3

0 (9

5% C

I: 0.

58−2

.92)

(1

1)

4.

0 or

gre

ater

R

R 1

.78

(95%

CI:

0.87

−3.6

4)

(19)

Tren

d p

= 0

.13

Ave

rage

exp

osur

e (p

pm)

0

R

R 0

.70

(95%

CI:

0.23

−2.1

6)

(4)

0.

1 to

0.4

R

R 1

.00

Refe

renc

e va

lue

(24)

0.

5 to

<1.

0

RR

1.2

1 (9

5% C

I: 0.

56−2

.62)

(9

)

1.

0 or

gre

ater

R

R 1

.61

(95%

CI:

0.76

−3.3

9)

(11)

p =

0.4

3

Cum

ulat

ive

(ppm

-yea

rs)

0

RR

0.6

1 (9

5% C

I: 0.

20−1

.91)

(4

)

0.

1-1.

4 R

R 1

.00

Refe

renc

e va

lue

(26)

1.

5-5.

4 R

R 0

.82

(95%

CI:

0.36

−1.8

3)

(8)

5.

5 or

gre

ater

R

R 1

.02

(95%

CI:

0.48

−2.1

6)

(10)

Tren

d p

> 0

.50

Panel Book Page 53

This document is a draft for review purposes only and does not constitute Agency policy.

4-158 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Hau

ptm

ann

et a

l. (2

009)

Pr

evio

us re

ports

:

Hay

es e

t al.

(199

0)

Wal

rath

and

Fra

umen

i (1

983)

W

alra

th a

nd F

raum

eni

(198

4)

Rel

ated

re-a

naly

ses:

Mar

sh e

t al.

(200

7a)

Mar

sh e

t al.

(200

7b)

Mar

sh a

nd Y

ouk

(200

5)

Mar

sh e

t al.

(199

6)

Nes

ted

case

-con

trol s

tudy

with

in

coho

rt m

orta

lity

stud

y of

6,8

08

deat

hs fr

om 1

960

to 1

986.

Id

entif

ied

from

regi

strie

s of t

he

Nat

iona

l Fun

eral

Dire

ctor

A

ssoc

iatio

n, li

cens

ing

boar

ds a

nd

stat

e fu

nera

l dire

ctor

s’

asso

ciat

ions

, NY

Sta

te B

urea

u of

Fu

nera

l Dire

ctor

s and

CA

Fun

eral

D

irect

ors a

nd E

mba

lmer

s. O

dds r

atio

s cal

cula

ted

usin

g un

cond

ition

al lo

gist

ic re

gres

sion

Occ

upat

iona

l his

tory

ob

tain

ed b

y in

terv

iew

s w

ith n

ext o

f kin

and

co

wor

kers

usi

ng d

etai

l qu

estio

nnai

res.

Expo

sure

was

ass

esse

d by

link

ing

ques

tionn

aire

re

spon

ses t

o an

exp

osur

e as

sess

men

t exp

erim

ent.

Ex

posu

re le

vels

(pea

k,

inte

nsity

and

cum

ulat

ive)

w

ere

assi

gned

to e

ach

indi

vidu

al u

sing

a

pred

ictiv

e m

odel

bas

ed

on th

e ex

posu

re d

ata.

All

LHP

canc

ers

Emba

lmin

g

Nev

er

OR

1.0

Re

fere

nce

valu

e (2

4)

Ev

er

OR

1.4

(9

5% C

I: 0.

8−2.

6)

(144

)

Dur

atio

n of

wor

king

in jo

bs w

ith e

mba

lmin

g (y

ears

)

0 O

R 1

.0

Refe

renc

e va

lue

(24)

>

0 to

20

OR

0.8

(9

5% C

I: 0.

4−1.

8)

(28)

>

20 to

34

OR

1.5

(9

5% C

I: 0.

8−2.

8)

(50)

>

34

OR

1.8

(9

5% C

I: 1.

0−3.

4)

(66)

Tren

d p

= 0

.058

Num

ber o

f em

balm

ings

0

OR

1.0

Re

fere

nce

valu

e (2

4)

>

0 to

142

2 O

R 0

.9

(95%

CI:

0.6−

1.8)

(2

9)

>

1422

to 3

068

OR

1.9

(9

5% C

I: 1.

0−3.

6)

(62)

>

3068

O

R 1

.5

(95%

CI:

0.8−

2.9)

(5

3)

Tren

d p

= 0

.477

Cum

ulat

ive

expo

sure

(ppm

-hou

rs)

0 O

R 1

.0

Refe

renc

e va

lue

(24)

>

0 to

405

8 O

R 1

.3

(95%

CI:

0.6−

2.5)

(4

0)

>

4058

to 9

253

OR

1.4

(9

5% C

I: 0.

8−2.

8)

(49)

>

9253

O

R 1

.6

(95%

CI:

0.8−

3.0)

(5

5)

Tren

d p

= 0

.422

Panel Book Page 54

This document is a draft for review purposes only and does not constitute Agency policy.

4-159 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Hau

ptm

ann

et a

l. (2

009)

(con

tinue

d)

Ave

rage

form

alde

hyde

inte

nsity

whi

le e

mba

lmin

g (p

pm)

0 O

R 1

.0

Refe

renc

e va

lue

(24)

>

0 to

1.4

O

R 1

.6

(95%

CI:

0.9−

3.2)

(5

3)

>

1.4

to 1

.9

OR

1.4

(9

5% C

I: 0.

7−2.

7)

(47)

>

1.9

O

R 1

.3

(95%

CI:

0.7−

2.5)

(4

4)

Tren

d p

= 0

.591

Tim

e-w

eigh

ted

aver

age

expo

sure

ove

r 8 h

ours

(ppm

)

0 O

R 1

.0

Refe

renc

e va

lue

(24)

>

0 to

0.1

0 O

R 1

.3

(95%

CI:

0.7−

2.6)

(4

7)

>

0.1

to 0

.18

O

R 1

.6

(95%

CI:

0.8−

3.1)

(5

2)

>

0.18

O

R 1

.4

(95%

CI:

0.7−

2.8)

(4

5)

Tren

d p

= 0

.635

Peak

form

alde

hyde

exp

osur

e (p

pm)

0 O

R 1

.0

Refe

renc

e va

lue

(24)

>

0 to

7.0

O

R 1

.6

(95%

CI:

0.8−

3.2)

(4

8)

>

7.0

to 9

.3

OR

1.6

(9

5% C

I: 0.

9−3.

1)

(55)

>

9.3

OR

1.2

(9

5% C

I: 0.

6−2.

3)

(41)

Tren

d p

= 0

.555

Mye

loid

leuk

emia

Emba

lmin

g

Nev

er

OR

1.0

Re

fere

nce

valu

e (1

)

Ev

er

OR

11.

2 (9

5% C

I: 1.

3−95

.6)

(33)

Panel Book Page 55

This document is a draft for review purposes only and does not constitute Agency policy.

4-160 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Hau

ptm

ann

et a

l. (2

009)

(con

tinue

d)

D

urat

ion

of w

orki

ng in

jobs

with

em

balm

ing

(yea

rs)

0 O

R 1

.0

Refe

renc

e va

lue

(1)

>

0 to

20

OR

5.0

(9

5% C

I: 0.

5−51

.6)

(6)

>

20 to

34

OR

12.

9 (9

5% C

I: 1.

4−11

7.1)

(1

3)

>

34

OR

13.

6 (9

5% C

I: 1.

6−11

9.7)

(1

4)

Tren

d p

= 0

.02

N

umbe

r of e

mba

lmin

gs

0

OR

1.0

Re

fere

nce

valu

e (1

)

> 0

to 1

422

OR

7.6

(9

5% C

I: 0.

8−73

.5)

(7)

>

1422

to 3

068

OR

12.

7 (9

5% C

I: 1.

4−11

6.7)

(1

2)

>

3068

O

R 1

2.7

(95%

CI:

1.4−

112.

8)

(14)

Tr

end

p =

0.3

14

C

umul

ativ

e ex

posu

re (p

pm-h

ours

) 0

OR

1.0

Re

fere

nce

valu

e (1

)

> 0

to 4

058

OR

10.

2 (9

5% C

I: 1.

1−95

.6)

(9)

>

4058

to 9

253

OR

9.4

(9

5% C

I: 1.

0−85

.7)

(10)

> 92

53

OR

13.

2 (9

5% C

I: 1.

5−11

5.4)

(1

4)

Tren

d p

= 0

.192

Ave

rage

form

alde

hyde

inte

nsity

whi

le e

mba

lmin

g (p

pm)

0 O

R 1

.0

Refe

renc

e va

lue

(1)

>

0 to

1.4

O

R 1

1.1

(95%

CI:

1.2−

106.

3)

(10)

> 1.

4 to

1.9

O

R 1

4.8

(95%

CI:

1.6−

136.

9)

(13)

> 1.

9

OR

9.5

(9

5% C

I: 1.

1−86

.0)

(10)

Tr

end

p =

0.0

58

Panel Book Page 56

This document is a draft for review purposes only and does not constitute Agency policy.

4-161 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Hau

ptm

ann

et a

l. (2

009)

(con

tinue

d)

Tim

e-w

eigh

ted

aver

age

expo

sure

ove

r 8 h

ours

(ppm

)

0 O

R 1

.0

Refe

renc

e va

lue

(1)

>

0 to

0.1

0 O

R 8

.4

(95%

CI:

0.8−

79.3

) (8

)

>

0.1

to 0

.18

O

R 1

3.6

(95%

CI:

1.5−

125.

8)

(13)

>

0.18

O

R 1

2.0

(95%

CI:

1.3−

107.

4)

(12)

Tren

d p

= 0

.396

Peak

form

alde

hyde

exp

osur

e (p

pm)

0 O

R 1

.0

Refe

renc

e va

lue

(1)

>

0 to

7.0

O

R 1

5.2

(95%

CI:

1.6−

141.

6)

(12)

>

7.0

to 9

.3

OR

8.0

(9

5% C

I: 0.

9−74

.0)

(9)

>

9.3

OR

13.

0 (9

5% C

I: 1.

4−11

6.9)

(1

2)

Tren

d p

= 0

.036

Wan

g et

al.

(200

9)

Popu

latio

n-ba

sed

case

-con

trol

stud

y of

inci

dent

cas

es o

f non

-H

odgk

in ly

mph

oma

diag

nose

s 19

96-2

000.

Expo

sure

s cla

ssifi

ed

usin

g jo

b ex

posu

re

mat

rix b

ased

on

occu

patio

nal a

nd in

dust

ry

data

obt

aine

d fr

om

pers

onal

inte

rvie

ws.

Non

-Hod

gkin

lym

phom

a

Form

alde

hyde

Nev

er

OR

1.0

Re

fere

nce

valu

e (3

98)

Ev

er

OR

1.3

(9

5% C

I: 1.

0−1.

7)

(203

)

Ave

rage

exp

osur

e in

tens

ity

Nev

er

OR

1.0

Re

fere

nce

valu

e (3

98)

Lo

w

OR

1.4

(9

5% C

I: 1.

0−1.

8)

(129

)

M

ediu

m-H

igh

OR

1.2

(9

5% C

I: 0.

8−1.

7)

(74)

Tren

d p

= 0

.21

Panel Book Page 57

This document is a draft for review purposes only and does not constitute Agency policy.

4-162 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Wan

g et

al.

(200

9)

(con

tinue

d)

Ave

rage

exp

osur

e pr

obab

ility

Nev

er

OR

1.0

Re

fere

nce

valu

e (3

98)

Lo

w

OR

1.3

(9

5% C

I: 1.

0−1.

7)

(129

)

M

ediu

m-H

igh

OR

1.4

(9

5% C

I: 0.

9−2.

3)

(74)

Tren

d p

> 0

.50

Bot

h av

erag

e ex

posu

re in

tens

ity a

nd a

vera

ge e

xpos

ure

prob

abili

ty

Low

Inte

nsity

and

Lo

w P

roba

bilit

y O

R 1

.4

(95%

CI:

1.1−

1.9)

(1

15)

M

ed-H

igh

Inte

nsity

an

d Lo

w P

roba

bilit

y O

R 1

.0

(95%

CI:

0.7−

1.6)

(5

0)

M

ed-H

igh

Inte

nsity

an

d M

ed-H

igh

Prob

. O

R 1

.1

(95%

CI:

0.5−

2.4)

(1

4)

M

ed-H

igh

Inte

nsity

an

d M

ed-H

igh

Prob

. O

R 1

.6

(95%

CI:

0.9−

3.1)

(2

4)

Leuk

emia

Peak

exp

osur

e (p

pm)

0

RR

0.5

9 (9

5% C

I: 0.

25−1

.36)

(7

)

0.

1−1.

9 R

R 1

.00

Refe

renc

e va

lue

(41)

2

.0 to

<4.

0

RR

0.9

8 (9

5% C

I: 0.

60−1

.62)

(2

7)

4

.0 o

r gre

ater

R

R 1

.42

(95%

CI:

0.92

−2.1

8)

(48)

Tren

d p

= 0

.012

Ave

rage

exp

osur

e (p

pm)

0

ppm

R

R 0

.54

(95%

CI:

0.24

−1.2

2)

(7)

0.

1−0.

4 R

R 1

.00

Refe

renc

e va

lue

(67)

Panel Book Page 58

This document is a draft for review purposes only and does not constitute Agency policy.

4-163 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Wan

g et

al.

(200

9)

(con

tinue

d)

0.

5 to

<1.

0

RR

1.1

3 (9

5% C

I: 0.

71−1

.79)

(2

5)

1.

0 or

gre

ater

R

R 1

.10

(95%

CI:

0.68

−1.7

8)

(24)

Tren

d p

> 0

.50

Cum

ulat

ive

expo

sure

(pp

m-y

ears

)

0

RR

0.5

3 (9

5% C

I: 0.

23−1

.21)

(7

)

0.

1−1.

4 R

R 1

.00

Refe

renc

e va

lue

(63)

1.

5−5.

4 R

R 0

.96

(95%

CI:

0.60

−1.5

6)

(24)

5.

5 or

gre

ater

R

R 1

.11

(95%

CI:

0.70

−1.7

4)

(29)

Tren

d p

= 0

.12

Hod

gkin

Lym

phom

a

Peak

exp

osur

e (p

pm)

0

RR

0.6

7 (9

5% C

I: 0.

12−3

.60)

(2

)

0.

1−1.

9 R

R 1

.00

Refe

renc

e va

lue

(6)

2.

0 to

<4.

0 R

R 3

.30

(95%

CI:

1.04

−10.

50)

(8)

4.

0 or

gre

ater

R

R 3

.96

(95%

CI:

1.31

−12.

02)

(11)

Tren

d p

= 0

.01

Ave

rage

exp

osur

e (p

pm)

0

R

R 0

.46

(95%

CI:

0.05

−3.9

3)

(2)

0.

1−0.

4 R

R 1

.00

Refe

renc

e va

lue

(10)

0.

5 to

<1.

0 R

R 3

.62

(95%

CI:

1.41

−9.3

1)

(9)

1.

0 or

gre

ater

R

R 2

.48

(95%

CI:

0.84

−7.3

2)

(6)

Tren

d p

= 0

.05

Panel Book Page 59

This document is a draft for review purposes only and does not constitute Agency policy.

4-164 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Wan

g et

al.

(200

9)

(con

tinue

d)

Cum

ulat

ive

(ppm

-yea

rs)

0

RR

0.4

2 (9

5% C

I: 0.

09−2

.05)

(2

)

0.1−

1.4

RR

1.0

0 Re

fere

nce

valu

e (1

4)

1.

5−5.

4 R

R 1

.71

(95%

CI:

0.66

−4.3

8)

(7)

5.

5 or

gre

ater

R

R 1

.30

(95%

CI:

0.40

−4.1

9)

(4)

Tren

d p

= 0

.08

Mye

loid

Leu

kem

ia

Pe

ak e

xpos

ure

(ppm

)

0 R

R 0

.82

(95%

CI:

0.25

−2.6

7)

(4)

0.

1 to

1.9

RR

1.0

0 Re

fere

nce

valu

e (1

4)

2.

0 to

<4.

0

RR

1.3

0 (9

5% C

I: 0.

58−2

.92)

(1

1)

4.

0 or

gre

ater

R

R 1

.78

(95%

CI:

0.87

−3.6

4)

(19)

Tr

end

p =

0.1

3

A

vera

ge e

xpos

ure

(ppm

)

0

RR

0.7

0 (9

5% C

I: 0.

23−2

.16)

(4

)

0.1

to 0

.4

RR

1.0

0 Re

fere

nce

valu

e (2

4)

0.

5 to

<1.

0

RR

1.2

1 (9

5% C

I: 0.

56−2

.62)

(9

)

1.0

or g

reat

er

RR

1.6

1 (9

5% C

I: 0.

76−3

.39)

(1

1)

p =

0.4

3

C

umul

ativ

e (p

pm-y

ears

)

0 R

R 0

.61

(95%

CI:

0.20

−1.9

1)

(4)

0.

1-1.

4 R

R 1

.00

Refe

renc

e va

lue

(26)

1.5-

5.4

RR

0.8

2 (9

5% C

I: 0.

36−1

.83)

(8

)

5.5

or g

reat

er

RR

1.0

2 (9

5% C

I: 0.

48−2

.16)

(1

0)

Tren

d p

> 0

.50

Panel Book Page 60

This document is a draft for review purposes only and does not constitute Agency policy.

4-165 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Pink

erto

n et

al.

(200

4)

Upd

ate

of S

tayn

er e

t al

. (19

88)

Coh

ort m

orta

lity

stud

y of

11,

098

wor

kers

in 3

gar

men

t pla

nts

expo

sed

≥3 m

onth

s afte

r fo

rmal

dehy

de w

as in

trodu

ced.

W

omen

com

pris

ed 8

1.7%

of t

he

coho

rt. V

ital s

tatu

s was

follo

wed

th

roug

h 19

98.

SMR

s wer

e ca

lcul

ated

by

usin

g se

x-, a

ge-,

race

-, an

d ca

lend

ar-y

ear-

spec

ific

U.S

. mor

talit

y ra

tes.

Mul

tiple

ca

use

SMR

s wer

e de

rived

from

all

cont

ribut

ing

caus

es fr

om d

eath

ce

rtific

ates

.

Dat

a fo

r 549

rand

omly

se

lect

ed e

mpl

oyee

s in

5 de

partm

ents

in 1

981

and

1984

use

d to

est

imat

e ov

eral

l exp

osur

e le

vels

. Le

vels

pre

sum

ed to

be

0.09

−0.2

0 pp

m.

All

LHP

canc

ers

SMR

0.9

7 (9

5% C

I: 0.

74−1

.26)

(5

9)

Lym

phos

arco

ma

and

retic

ulos

arco

ma

SMR

0.8

5 (9

5% C

I: 0.

28−1

.99)

(5

)

Hod

gkin

's di

seas

e SM

R 0

.55

(95%

CI:

0.07

−1.9

8)

(2)

Oth

er ly

mph

atic

SM

R 0

.97

(95%

CI:

0.64

−1.4

0)

(28)

Leuk

emia

SM

R 1

.09

(95%

CI:

0.70

−1.6

2)

(24)

M

orta

lity

sinc

e 19

60

Lym

phoc

ytic

leuk

emia

SM

R 0

.60

(95%

CI:

0.12

−1.7

5)

(3)

ML

SM

R 1

.44

(95%

CI:

0.80

−2.3

7)

(15)

10

+ ye

ars o

f ex

posu

re

SMR

2.1

9 N

S (8

)

20

+ ye

ars s

ince

1st

expo

sure

SMR

1.9

1 p

> 0.

05

(13)

Mul

tiple

cau

se le

ukem

ia

10

+ ye

ars o

f ex

posu

re a

nd 2

0+

year

s sin

ce 1

st

expo

sure

SMR

1.9

2 (9

5% C

I: 1.

08−3

.17)

(1

5)

Mul

tiple

cau

se M

L

20

+ ye

ars s

ince

1st

expo

sure

SM

R 2

.02

(95%

CI:

1.13

−3.3

4)

(15)

10

+ ye

ars o

f ex

posu

re a

nd 2

0+

year

s sin

ce 1

st

expo

sure

SMR

2.5

5 (9

5% C

I: 1.

10−5

.03)

(8

)

Panel Book Page 61

This document is a draft for review purposes only and does not constitute Agency policy.

4-166 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Cog

gon

et a

l. (2

003)

U

pdat

e of

Gar

dner

et

al. (

1993

)

Coh

ort m

orta

lity

stud

y of

14,

014

men

em

ploy

ed in

6 fa

ctor

ies o

f the

ch

emic

al in

dust

ry in

Gre

at B

ritai

n fr

om p

erio

ds d

urin

g w

hich

fo

rmal

dehy

de w

as p

rodu

ced.

C

ohor

t mor

talit

y fo

llow

ed fr

om

1941

thro

ugh

2000

. SM

Rs b

ased

on

Eng

lish

and

Wel

sh a

ge- a

nd

cale

ndar

-yea

r-sp

ecifi

c m

orta

lity

rate

s.

Expo

sure

ass

essm

ent

base

d on

dat

a ab

stra

cted

fr

om c

ompa

ny re

cord

s.

Jobs

cat

egor

ized

as

back

grou

nd, l

ow,

mod

erat

e, h

igh,

or

unkn

own

leve

ls.

Non

-Hod

gkin

's ly

mph

oma

O

vera

ll

SMR

0.9

8 (9

5% C

I: 0.

67−1

.39)

(3

1)

H

igh

expo

sure

SM

R 0

.89

(95%

CI:

0.41

−1.7

0)

(9)

Leuk

emia

O

vera

ll

SMR

0.9

1 (9

5% C

I: 0.

62−1

.29)

(3

1)

H

igh

expo

sure

SM

R 0

.71

(95%

CI:

0.31

−1.3

9)

(8)

Mul

tiple

mye

lom

a

O

vera

ll

SMR

0.8

6 (9

5% C

I: 0.

48−1

.41)

(1

5)

H

igh

expo

sure

SM

R 1

.18

(95%

CI:

0.48

−2.4

4)

(7)

And

jelk

ovic

h et

al.

(199

5)

Coh

ort m

orta

lity

stud

y of

3,9

29

auto

mot

ive

indu

stry

iron

foun

dry

wor

kers

exp

osed

from

196

0−19

87

and

follo

wed

thro

ugh

1989

. SM

Rs c

alcu

late

d us

ing

sex-

, age

-, ra

ce-,

and

cale

ndar

-yea

r-sp

ecifi

c U

.S. m

orta

lity

rate

s.

Expo

sure

ass

essm

ent

base

d on

revi

ew o

f wor

k hi

stor

ies b

y an

indu

stria

l hy

gien

ist.

All

LHP

canc

ers

SMR

0.5

9 (9

5% C

I: 0.

23−1

.21)

(7

)

Leuk

emia

SM

R 0

.43

(95%

CI:

0.05

−1.5

7)

(2)

Ber

tazz

i et a

l. (1

986)

C

ohor

t mor

talit

y st

udy

of 1

,330

m

ale

wor

kers

in a

n Ita

lian

resi

n pl

ant.

Sub

ject

s wer

e em

ploy

ed

any

time

betw

een

1959

and

198

0 fo

r at l

east

30

days

. V

ital s

tatu

s fo

llow

ed th

roug

h 19

86.

SMR

s ca

lcul

ated

usi

ng se

x-, a

ge-,

race

-, an

d ca

lend

ar-y

ear-

spec

ific

natio

nal a

nd lo

cal m

orta

lity

rate

s.

Expo

sure

ass

essm

ent

base

d on

reco

nstru

ctio

n of

wor

k hi

stor

y.

Expo

sure

leve

ls w

ere

0.16

to 3

.1 p

pm

All

LHP

canc

ers

SMR

2.0

1

(5)

Panel Book Page 62

This document is a draft for review purposes only and does not constitute Agency policy.

4-167 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Edlin

g et

al.

(198

7)

Coh

ort m

orta

lity

and

inci

denc

e st

udy

of 5

21 S

wed

ish

wor

kers

in

an a

bras

ive

prod

uctio

n pl

ant w

ith

at le

ast 5

yea

rs o

f em

ploy

men

t be

twee

n 19

55 a

nd 1

983,

follo

wed

th

roug

h 19

83.

Expo

sure

leve

l of

0.1−

1 m

g/m

3 . Ly

mph

oma

SM

R 2

.0

(95%

CI:

0.2−

7.2)

(2

)

Mul

tiple

mye

lom

a SM

R 4

.0

(95%

CI:

0.5−

14)

(2)

Del

l and

Tet

a (1

995)

C

ohor

t mor

talit

y st

udy

of 5

,932

m

ale

empl

oyee

s of a

New

Jers

ey

plas

tics m

anuf

actu

ring,

rese

arch

an

d de

velo

pmen

t fac

ility

that

pr

oduc

ed p

heno

l-for

mal

dehy

de

resi

ns.

Exam

inat

ion

of w

ork

hist

orie

s to

iden

tify

jobs

w

here

form

alde

hyde

was

in

volv

ed.

A

ll LH

P ca

ncer

s

H

ourly

wor

kers

SM

R 0

.93

(2

8)

Sa

larie

d w

orke

rs

SMR

1.6

9

(23)

Le

ukem

ia

H

ourly

wor

kers

SM

R 0

.98

(1

2)

Sa

larie

d w

orke

rs

SMR

1.9

8

(11

)

Wal

rath

and

Fra

umen

i (1

983)

C

ohor

t stu

dy o

f 1,1

32 w

hite

mal

e em

balm

ers l

icen

sed

to p

ract

ice

betw

een

1902

and

198

0 in

New

Y

ork

who

die

d be

twee

n 19

25 a

nd

1980

iden

tifie

d fr

om re

gist

ratio

n fil

es.

Dea

ths w

ere

com

pare

d w

ith

age-

, rac

e-, a

nd c

alen

dar-

year

-ex

pect

ed n

umbe

rs o

f dea

ths f

rom

th

e U

.S. p

opul

atio

n.

No

dire

ct m

easu

rem

ents

. Pr

esum

ed e

xpos

ure

to

form

alde

hyde

tiss

ue

fixat

ive.

All

LHP

canc

ers

SM

Rb 1

.15

(2

1)

Lym

phos

arco

ma

and

retic

ulos

arco

ma

SM

Rb 1

.08

(4

)

Hod

gkin

’s d

isea

se

SM

Rb 1

.0

(2

)

Oth

er ly

mph

atic

lym

phom

a

SM

Rb 1

.18

(5

)

Leuk

emia

SM

Rb 1

.32

(1

0)

Panel Book Page 63

This document is a draft for review purposes only and does not constitute Agency policy.

4-168 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Wal

rath

and

Fra

umen

i (1

984)

C

ohor

t stu

dy o

f 1,0

07 w

hite

mal

e em

balm

ers f

rom

the

Cal

iforn

ia

Bur

eau

of F

uner

al D

irect

ing

and

Emba

lmin

g w

ho d

ied

betw

een

1925

and

198

0. D

eath

s wer

e co

mpa

red

with

age

- and

cal

enda

r-ye

ar-e

xpec

ted

num

bers

of

dea

ths f

rom

the

U.S

. po

pula

tion.

No

dire

ct m

easu

rem

ents

. Pr

esum

ed e

xpos

ure

to

form

alde

hyde

tiss

ue

fixat

ive.

All

LHP

canc

ers

SMR

b 1.2

2

(19)

Lym

phos

arco

ma

and

retic

ulos

arco

ma

SMR

b 0.9

7

(3)

Hod

gkin

’s d

isea

se

SMR

b −

(0

)

Oth

er ly

mph

atic

ly

mph

oma

SMR

b 1.3

3

(4)

Leuk

emia

SM

Rb 1

.75

p <

0.05

(1

2)

Li

cens

ed <

20 y

ears

SM

Rb 1

.24

(4

)

Li

cens

ed ≥

20 y

ears

SM

Rb 2

.21

p <

0.05

(8

)

Hay

es e

t al.

(199

0)

Prop

ortio

nate

mor

talit

y co

hort

stud

y of

3,6

49 d

ecea

sed

whi

te a

nd

397

dece

ased

non

whi

te U

.S. m

ale

emba

lmer

s and

fune

ral d

irect

ors,

deriv

ed fr

om li

cens

ing

boar

ds a

nd

fune

ral d

irect

or a

ssoc

iatio

ns in

the

32 st

ates

and

the

Dis

trict

of

Col

umbi

a. O

ccup

atio

n w

as

conf

irmed

on

deat

h ce

rtific

ate.

D

eath

s wer

e co

mpa

red

with

age

- an

d ca

lend

ar-y

ear-

expe

cted

nu

mbe

rs o

f dea

ths f

rom

the

U.S

. po

pula

tion.

No

dire

ct m

easu

rem

ents

. Pr

esum

ed e

xpos

ure

to

form

alde

hyde

tiss

ue

fixat

ive.

All

LHP

canc

ers

SM

Rb 1

.39

(95%

CI:

1.15

−1.6

7)

(115

)

Rac

e

W

hite

SM

Rb 1

.31

(95%

CI:

1.06

−1.5

9)

(100

)

N

onw

hite

SM

Rb 2

.41

(95%

CI:

1.35

−3.9

7)

(15)

Occ

upat

ion

on d

eath

cer

tific

ate

Em

balm

er

SMR

b 1.2

3 (9

5% C

I: 0.

78−1

.85)

(2

3)

Fu

nera

l dire

ctor

SM

Rb 1

.56

(95%

CI:

1.23

−1.9

4)

(78)

O

ther

SM

Rb 1

.30

(95%

CI:

0.67

−2.2

8)

(12)

Age

at d

eath

<6

0 SM

Rb 1

.35

(95%

CI:

0.88

−1.9

8)

(26)

60

−74

SMR

b 1.7

2 (9

5% C

I: 1.

33−2

.19)

(6

6)

≥7

5 SM

Rb 1

.16

(95%

CI:

0.74

−1.7

4)

(23)

Panel Book Page 64

This document is a draft for review purposes only and does not constitute Agency policy.

4-169 DRAFT—DO

Tab

le 4

-7.

Epi

dem

iolo

gic

stud

ies o

f for

mal

dehy

de a

nd ly

mph

ohem

atop

oiet

ic c

ance

rs (c

ontin

ued)

Ref

eren

ce

Stud

y de

sign

E

xpos

ure

asse

ssm

ent

Res

ults

, sta

tistic

al si

gnifi

canc

e (n

umbe

r of

obs

erve

d de

aths

for

coho

rt

stud

y)

Hay

es e

t al.

(199

0)

(con

tinue

d)

Hod

gkin

’s d

isea

se

SMR

b 0.7

2 (9

5% C

I: 0.

15−2

.10)

(3

)

Non

-Hod

gkin

’s

lym

phom

a SM

Rb 1

.26

(95%

CI:

0.87

−1.7

6)

(34)

Lym

phos

arco

ma

and

retic

ulos

arco

ma

SMR

b 1.1

2 (9

5% C

I: 0.

58−1

.96)

(1

2)

Mul

tiple

mye

lom

a SM

Rb 1

.37

(95%

CI:

0.84

−2.1

2)

(20)

Oth

er ly

mph

atic

ly

mph

oma

SMR

b 1.3

5 (9

5% C

I: 0.

85−2

.01)

(2

2)

Lym

phat

ic le

ukem

ia

SMR

b 0.7

4 (9

5% C

I: 0.

29−1

.53)

(7

)

ML

SMR

b 1.5

7 (9

5% C

I: 1.

01−2

.34)

(2

4)

Oth

er le

ukem

ia

SMR

b 2.2

8 (9

5% C

I: 1.

39−3

.52)

(2

0)

Bla

ir et

al.

(199

3)

Popu

latio

n-ba

sed

case

-con

trol

stud

y of

622

whi

te m

en w

ith L

HP

canc

ers.

Can

cers

sele

cted

from

Io

wa

and

Min

neso

ta c

ance

r su

rvei

llanc

e ne

twor

ks d

iagn

osed

be

twee

n 10

/80

and

9/82

. 1,

245

mat

ched

con

trols

for l

ivin

g ca

ses

sele

cted

by

rand

om d

igit

dial

ing

if yo

unge

r tha

n ag

e 65

and

from

M

edic

are

reco

rds i

f 65

or o

lder

. St

udy

focu

sed

on a

gric

ultu

ral

expo

sure

s.

Pers

onal

inte

rvie

ws o

f su

bjec

ts o

r nex

t of k

in

incl

uded

job

hist

orie

s, ag

ricul

tura

l exp

osur

es,

and

chem

ical

exp

osur

es.

Job

title

s use

d to

cre

ate

job

expo

sure

mat

rix.

Indu

stria

l hyg

ieni

st

estim

ated

pro

babi

lity

and

inte

nsity

of e

xpos

ures

to

larg

e nu

mbe

rs o

f su

bsta

nces

.

Non

-Hod

gkin

’s

lym

phom

a (f

orm

alde

hyde

ex

posu

re)

OR

a 1.2

(9

5% C

I: 0.

9−1.

7)

Fune

ral s

ervi

ce w

orke

r O

Ra 2

.1

(95%

CI:

0.5−

7.9)

(6

)

a Adj

uste

d fo

r age

, sta

te, s

mok

ing,

fam

ily h

isto

ry o

f mal

igna

nt p

rolif

erat

ive

dise

ase,

agr

icul

tura

l exp

osur

e to

pes

ticid

es, u

se o

f dye

, and

dire

ct/s

urro

gate

resp

onse

to

inte

rvie

w.

b Wal

rath

and

Fra

umen

i (19

83, 1

984)

. Th

ese

stud

ies a

re re

ferr

ed to

by

the

auth

ors a

s pro

porti

onat

e m

orta

lity

stud

ies a

nd re

port

prop

ortio

nal m

orta

lity

ratio

s w

hich

are

kno

wn

to b

e po

tent

ially

bia

sed.

How

ever

, rev

iew

of t

he a

ctua

l met

hods

des

crib

ed c

lear

show

s tha

t the

exp

ecte

d nu

mbe

rs o

f cau

se-s

peci

fic d

eath

s w

ere

base

d on

a st

anda

rdiz

ed g

ener

al p

opul

atio

n an

d th

eref

ore

the

repo

rted

PMR

s are

mor

e ac

cura

tely

cal

led

SMR

s.

Panel Book Page 65

This document is a draft for review purposes only and does not constitute Agency policy. DRAFT—DO NOT CITE OR QUOTE 4-170

Stayner et al. (1988) conducted a cohort study of 11,030 workers (82% female) followed 1 from 1955 or the beginning date of exposure through 1982 in three garment factories. Personnel 2 records from three garment manufacturing facilities, one in Pennsylvania and two in Georgia, 3 were used to assemble a cohort of workers who attained a minimum of 3 months of exposure 4 after the introduction of formaldehyde into these facilities. Formaldehyde resins were used to 5 treat fabrics, beginning in 1955 and 1959. Although formaldehyde levels were available on a 6 subset of the employees from monitoring data available from surveys completed in 1981 and 7 1984, they were not used in this analysis. Instead, the results were stratified by duration and 8 latency. SMRs were based on U.S. population mortality rates. Based on six cases, the SMRs for 9 leukemia were 2.43 and 3.81 among workers with 20 or more years since first exposure or at 10 least 10 years of exposure, respectively. In their conclusions, the authors suggested that, 11 although the numbers of deaths from LHP cancers were small, the risks were related to duration 12 and latency. 13

Pinkerton et al. (2004) updated the Stayner et al. (1988) study by adding 16 years of 14 follow-up. No new exposure information was added. The mean TWA exposure in 1981−1984 15 for the three plants was 0.15 ppm. No additional information regarding earlier industrial hygiene 16 data was available, although the authors stated that the levels of exposure to formaldehyde were 17 greater in the years before 1980. Stayner et al. (1988) cited independent studies of exposure 18 levels in similar garment factories in the 1960s that seemed to indicate that the formaldehyde 19 levels during that period ranged from 0.9 to 2.7 ppm (Blejer and Miller, 1966) in one garment 20 manufacturing area. Another report (Shipkovitz, 1966) of 10-minute personal exposure samples 21 indicated a range from 0.3 to 2.7 ppm in eight garment plants. In another study (Ahmad and 22 Whitson, 1973), the levels ranged from 2 to 10 ppm. Goldstein (1973) calculated that 23 concentrations in the cutting rooms of garment plants dropped from 10 ppm in 1968 to less than 24 2 ppm in 1973 because of an improvement in the resin treating process. The authors assumed 25 that exposure ceased in 1981 and 1983. This produced an underestimate of exposure based on 26 duration of employment for about 11% of the cohort who were still actively employed after those 27 dates. Stayner et al. (1988) speculated that the risks of cancer of the buccal cavity, leukemia, and 28 other LHP neoplasia may have been due to exposure to the highest potential formaldehyde levels 29 in the industry between 1955 and 1962, because the resin used to treat permanent press fabrics 30 still contained a relatively large amount of formaldehyde. 31

The SMRs were derived from age-, race-, and calendar-time-adjusted U.S. mortality 32 rates. The analysis was repeated using Georgia or Pennsylvania mortality rates. In addition to 33 the primary analysis of the underlying cause of death, the analysis used all causes listed on the 34 death certificates to evaluate multiple cause mortality. As a referent for this, the analysis relied 35

Panel Book Page 66

This document is a draft for review purposes only and does not constitute Agency policy. DRAFT—DO NOT CITE OR QUOTE 4-385

serum testosterone levels and mean seminiferous tubule diameters were significantly decreased 1 from control in a dose-responsive manner (see Table 4-62). Immunohistochemical staining of 2 testis tissues showed increased localization of heat shock protein (Hsp) 70 in the cytoplasm of 3 spermatogonia, spermatocytes, and spermatids of treated animals compared with controls (not 4 shown here). 5

6 Table 4-62. Formaldehyde effects on testosterone levels and seminiferous 7 tubule diameters in Wistar rats following 91 days of exposure 8 9

Inhalation exposurea

Testosterone levels (ng/dL)

Seminiferous tubule diameters

(μm)

n = 6 n = 100

Control 406.54 ± 16.82 259.22 ± 16.18

10 ppm 244.01 ± 23.86b 236.17 ± 13.09c

20 ppm 141.30 ± 08.56b 233.24 ± 10.13c 10

aFormaldehyde exposure was 8 hours/day, 5 days/week for 91 weeks. Values are means 11 ± SEMs. 12 bDifferent from control, p < 0.0001, by one-way ANOVA, as calculated by the authors. 13 cDifferent from control, p < 0.001, by one-way ANOVA, as calculated by the authors. 14 15 Source: Özen et al. (2005). 16 17

18 Zhou et al. (2006) investigated the effect of formaldehyde on the testes and the protective 19 effect of vitamin E against oxidative damage by formaldehyde in adult male rats. In this study, 20 adult male Sprague-Dawley rats (10/group) were treated for 2 weeks in the following groups: 21 (1) control rats were administered physiological saline by oral gavage, (2) rats were administered 22 physiological saline by gavage and exposed to 10 mg/m3 (8.05 ppm) formaldehyde by inhalation 23 for 12 hours/day, and (3) rats were administered daily gavage doses of 30 mg/kg vitamin E and 24 exposed to 10 mg/m3 (8.05 ppm) formaldehyde by inhalation for 12 hours/day. Formaldehyde 25 treatment resulted in significantly decreased (p < 0.05) mean testis weight. Histopathologic 26 findings in treated rats included atrophy of seminiferous tubules, decreased spermatogenic cells, 27 and seminiferous cells that were “disintegrated” and shed into the lumina, which was 28 azoospermic. Interstitial tissue was edematous with vascular dilatation and hyperemia. In the 29 formaldehyde-treated group, epididymal sperm count and percentage of motile sperm were 30 significantly decreased, and the percentage of abnormal sperm was increased (p < 0.05), as 31 compared with control. Evaluation of biochemical markers in testes tissue showed the activities 32 of testicular SOD, GPX, and GSH were decreased; MDA levels were significantly increased as 33

Panel Book Page 67

This document is a draft for review purposes only and does not constitute Agency policy. DRAFT—DO NOT CITE OR QUOTE 4-387

rate of resorptions (p < 0.01) (see Table 4-64). The mean number of live fetuses/litter was 1 decreased in all treated groups, with statistical significance achieved at 84 mg/m3. Although this 2 study did not assess the number of corpora lutea per dam, thereby precluding the calculation of 3 preimplantation loss, it is nevertheless indicative of formaldehyde-induced sperm morphology 4 changes and dominant lethal effects in male mice. 5

6

Table 4-63. Effects of formaldehyde exposure on seminiferous tubule 7 diameter and epithelial height in Wistar rats following 18 weeks of exposure 8 9

Inhalation exposurea

Seminiferous tubule diameters

(μm)

Seminiferous tubule height (μm)

n = 7 n = 7

Control 252.12 ± 4.82 82.77 ± 2.00

1.5 ppm, 4 h/d, 4 d/w 204.55 ± 3.29b 65.26 ± 1.43b

1.5 ppm, 2 h/d, 4 h/w 232.45 ± 2.42b 69.46 ± 1.78b

1.5 ppm, 2 h/d, 2 d/w 238.94 ± 4.37b 72.80 ± 2.03b 10

a Values are means ± SEMs. 11 b Different from control, p < 0.05, as calculated by the authors. 12 13 Source: Golalipour et al. (2007). 14

15

16 Table 4-64. Incidence of sperm abnormalities and dominant lethal effects in 17 formaldehyde-treated mice 18

19

Dose (mg/m3)

Sperm abnormalities Reproductive capacity

Total abnormal sperm heads

Aberration rate (%) Mean live fetuses/litter Resorption rate (%)

0 391 3.53 ± 0.98 11.00 ± 1.01 2.273

21 568 5.48 ± 1.45 10.67 ± 1.16 9.380 b

42 849 6.15 ± 1.36 9.63 ± 2.83 10.390 b

84 974 9.24 ± 2.13a 9.04 ± 2.98 a 12.440 b 20 aSignificantly different from controls (p < 0.05), as calculated by the authors. 21 bSignificantly different from controls (p < 0.01), as calculated by the authors. 22 23 Source: Xing et al. (2007). 24 25 26

Panel Book Page 68

This document is a draft for review purposes only and does not constitute Agency policy. DRAFT—DO NOT CITE OR QUOTE 4-437

Table 4-83. Genotoxicity in laboratory animals 1 2

Species/Strain Cells/Organ/Tumor Result References

Cytogenetic Assays

Mice/Q strain

Chromosomal aberrations (CA)

Spermatocyte - Fontignie-Houbrechts et al., 1981

Spermatogonia - Fontignie-Houbrechts et al., 1982

Mice/CBA Polychromatic erythrocytes - Natarajan et al., 1983

Spleen cells - Natarajan et al., 1983

Rats/F344 Lymphocytes - Kligerman et al 1984

Rats/Sprague-Dawley Gastric epithelial cells + Migliore et al 1989

Rats/Wistar Bone marrow + Kitaeva et al 1990

Rats/Sprague-Dawley Bone marrow - Dallas et al 1992

Rats/Sprague-Dawley Pulmonary lavage cells + Dallas et al 1992

Rats/F344 Peripheral blood cells - Speit et al 2009

Micronucleus (MN)

Mouse/NMRI Bone marrow - Gocke et al 1981

Mice/CBA Femoral polychromatic erythrocyte and spleen cell - Natarajan et al., 1983

Mice/B6C3F1 Bone marrow + Ward et al 1983

Rats/Sprague-Dawley Gastric epithelial cells + Migliore et al 1989

Sister Chromatid Exchange (SCE)

Rats/F344 Lymphocyte - Kligerman et al 1984

Rats/F344 Peripheral blood cells - Speit et al 2009 3 ‘+’ indicates a positive test result. 4 ‘-’ indicates a negative test result. 5 6 7

In a chromosomal analysis study (Fontignie-Houbrechts, 1981), formaldehyde given I.P. 8 at 50 mg/kg to male Q strain mice and analyzed 8−15 days after treatment did not induce any 9 chromosomal lesions in spermatocytes. Also, in another study from the same group (Fontignie-10 Houbrechts et al., 1982), formaldehyde (30 mg/kg) given along with hydrogen peroxide (90 11 mg/kg) as a mixture to male Q strain mice failed to produce significant increases in 12 chromosomal lesions in the spermatogonia. 13

Ward et al. (1983) studied the cytogenetic effects of commercial formalin on the bone 14 marrow of B6C3F1 mice. Since commercial formalin contains 10-15% of methanol, another 15 group was dosed with methanol (1000 mg/kg) and also included a negative control (water) and a 16

Panel Book Page 69

FORMALDEHYDE NAC/Interim 1: 07/2008; Page 53 of 71

TABLE 10. Extant Standards and Guidelines for Formaldehyde Exposure Duration

Guideline 10 min 30 min 1 h 4 h 8 h AEGL-1 0.90 ppm 0.90 ppm 0.90 ppm 0.90 ppm 0.90 ppm AEGL-2 14 ppm 14 ppm 14 ppm 14 ppm 14 ppm AEGL-3 100 ppm 70 ppm 56 ppm 35 ppm 35 ppm ERPG-1 (AIHA)a 1 ppm ERPG-2 (AIHA) 10 ppm ERPG-3 (AIHA) 25 ppm SMAC (NRC)b

0.4 ppm

EEGL (NRC)c

2 ppm

PEL-TWA PEL-STEL (OSHA)d

0.75 ppm 2 ppm (15 minute)

IDLH (NIOSH)e 20 ppm* REL-TWA REL-STEL (NIOSH)f

0.016 ppm* 0.1 ppm (15 minute)

TLV-Ceiling (ACGIH)g

0.3 ppm*

MAK Peak Limit (Germany)h

0.3 ppm* 1 ppm

MAC Peak Limit (The Netherlands)i

1 ppm 2 ppm

*potential occupational carcinogen 1 2 aERPG (Emergency Response Planning Guidelines, American Industrial Hygiene Association (AIHA 2004) 3

The ERPG-1 is the maximum airborne concentration below which it is believed nearly all individuals could be 4 exposed for up to one hour without experiencing other than mild, transient adverse health effects or without 5 perceiving a clearly defined objectionable odor. 6 The ERPG-2 is the maximum airborne concentration below which it is believed nearly all individuals could be 7 exposed for up to one hour without experiencing or developing irreversible or other serious health effects or 8 symptoms that could impair an individual=s ability to take protection action. 9 The ERPG-3 is the maximum airborne concentration below which it is believed nearly all individuals could be 10 exposed for up to one hour without experiencing or developing life-threatening health effects. 11

12 bSMAC (Spacecraft Maximum Allowable Concentration) (NRC 1994) 13

SMACs provide guidance on chemical exposures during normal operations of spacecraft as well as emergency 14 situations. The one-hour SMAC is a concentration of airborne substance that will not compromise the 15 performance of specific tasks by astronauts during emergency conditions or cause serious or permanent toxic 16 effects. Such exposure may cause reversible effects such as skin or eye irritation, but they are not expected to 17 impair judgment or interfere with proper responses to emergencies. 18

19 cEEGL (Emergency and Continuous Exposure Levels for Chemicals in Submarines) (NRC 2007) 20

EEGLs provide guidance on chemical exposures during normal operations of submarines. The one-hour EEGL 21 is a concentration that would allow up to moderate irritation in some individuals, but would not interfere with 22 critical duties. These exposures are for healthy adults. 23

24

Panel Book Page 70

FORMALDEHYDE NAC/Interim 1: 07/2008; Page 54 of 71

dOSHA PEL-TWA (Occupational Safety and Health Administration, Permissible Exposure Limits - Time 1 Weighted Average) (NIOSH 1997) is defined analogous to the ACGIH-TLV-TWA, but is for exposures of no 2 more than 10 hours/day, 40 hours/week. The OSHA PEL-STEL (Permissible Exposure Limits - Short Term 3 Exposure Limit) is defined analogous to the ACGIH-TLV-STEL. 4

5 eIDLH (Immediately Dangerous to Life and Health, National Institute of Occupational Safety and Health) 6

(NIOSH 1997) represents the maximum concentration from which one could escape within 30 minutes without 7 any escape-impairing symptoms, or any irreversible health effects. 8

9 fNIOSH REL-TWA (National Institute of Occupational Safety and Health, Recommended Exposure Limits - 10

Time Weighted Average) (NIOSH 1997) is defined analogous to the ACGIH-TLV-TWA. The NIOSH REL-11 STEL (Recommended Exposure Limits - Short Term Exposure Limit) is defined analogous to the ACGIH TLV-12 STEL. 13

14 gACGIH TLV-TWA (American Conference of Governmental Industrial Hygienists, Threshold Limit Value - 15

Time Weighted Average) (ACGIH 2004) is the time-weighted average concentration for a normal 8-hour 16 workday and a 40-hour workweek, to which nearly all workers may be repeatedly exposed, day after day, 17 without adverse effect. 18

19 hMAK (Maximale Arbeitsplatzkonzentration [Maximum Workplace Concentration]) List of MAK and BAT 20

Values 2007 (Deutsche Forschungsgemeinschaft [German Research Association] 2007) is defined analogous to 21 the ACGIH-TLV-TWA. In the case of formaldehyde a momentary value of 1 ppm should not be exceeded. 22 The MAK Spitzenbegrenzung (Peak Limit [give category]) constitutes the maximum average concentration to 23 which workers can be exposed for a period up to 30 minutes with no more than 2 exposure periods per work 24 shift; total exposure may not exceed 8-hour MAK. 25

26 iMAC (Maximaal Aanvaarde Concentratie [Maximal Accepted Concentration]) (SDU Uitgevers [under the 27

auspices of the Ministry of Social Affairs and Employment], The Hague, The Netherlands 2000) is defined 28 analogous to the ACGIH-TLV-TWA. The peak limit is defined analogous to the ACGIH ceiling. 29

30 8.3. Data Adequacy and Research Needs 31 32

Formaldehyde has a robust data set of controlled human exposures. Data from 22 well-33 conducted clinical studies involving over 500 subjects form a reliable basis for setting AEGL-1 34 and AEGL-2 values. The data base for lethality involves relatively old animal studies that lack 35 details of methodology as well as clear results. However, the data, with support from repeat-36 exposure studies with animals, can be used to set non-lethal values for humans. 37 38 9. REFERENCES 39 40 ACGIH (American Conference of Government and Industrial Hygienists). 2004. Documentation of the Threshold 41

Limit Values and Biological Exposure Indices: Formaldehyde. Cincinnati, OH: ACGIH. 42 43 Ad Hoc Panel on Health Aspects of Formaldehyde. 1988. Epidemiology of chronic occupational exposure to 44

formaldehyde. Toxicol. Ind. Health 4:77-90. 45 46 AIHA (American Industrial Hygiene Association). 2004. Emergency Response Planning Guidelines: 47

Formaldehyde. Akron, OH: AIHA. 48 49 Akbar-Khanzadeh, F. and J.S. Mlynek. 1997. Changes in respiratory function after one and three hours of exposure 50

to formaldehyde in non-smoking subjects. Occup. Environ. Med. 54:296-300. 51 52 Akbar-Khanzadeh, F., M.U. Vaquerano, Akbar-Khanzadeh, M., and M.S. Bisesi. 1994. Formaldehyde exposure, 53

acute pulmonary response, and exposure control options in a gross anatomy laboratory. Am. J. Ind. med. 54 26:61-75. 55

Panel Book Page 71