[Federal Register Volume 83, Number 98 (Monday, May 21, 2018)]
[Proposed Rules]
[Pages 23382-23392]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2018-10715]


=======================================================================
-----------------------------------------------------------------------

DEPARTMENT OF HEALTH AND HUMAN SERVICES

Food and Drug Administration

21 CFR Part 172

[Docket No. FDA-2015-F-3663]


Grocery Manufacturers Association; Denial of Food Additive 
Petition

AGENCY: Food and Drug Administration, HHS.

ACTION: Notification; denial of petition.

-----------------------------------------------------------------------

SUMMARY: The Food and Drug Administration (FDA or we) is denying a food 
additive petition (FAP 5A4811), submitted by the Grocery Manufacturers 
Association (GMA), requesting that the food additive regulations be 
amended to provide for the safe use of partially hydrogenated vegetable 
oils (PHOs) in certain food applications. We are denying the petition 
because we have determined that the petitioner did not provide 
sufficient information for us to conclude that the requested uses of 
PHOs are safe. To allow the food industry sufficient time to identify 
suitable replacement substances for the petitioned uses of PHOs, 
elsewhere in this issue of the Federal Register we have extended the 
compliance date for certain uses of PHOs, including the conditions of 
use covered by the FAP.

DATES: This document is applicable May 21, 2018. Submit either 
electronic or written objections and requests for a hearing on the 
document by June 20, 2018. Late, untimely objections will not be 
considered. See section VIII for further information on the filing of 
objections.

ADDRESSES: You may submit objections and requests for a hearing as 
follows.

Electronic Submissions

    Submit electronic objections in the following way:
     Federal eRulemaking Portal: https://www.regulations.gov. 
Follow the instructions for submitting comments. Objections submitted 
electronically, including attachments, to https://www.regulations.gov 
will be posted to the docket unchanged. Because your objection will be 
made public, you are solely responsible for ensuring that your 
objection does not include any confidential information that you or a 
third party may not wish to be posted, such as medical information, 
your or anyone else's Social Security number, or confidential business 
information, such as a manufacturing process. Please note that if you 
include your name, contact information, or other information that 
identifies you in the body of your objection, that information will be 
posted on https://www.regulations.gov.
     If you want to submit an objection with confidential 
information that you do not wish to be made available to the public, 
submit the objection as a written/paper submission and in the manner 
detailed (see ``Written/Paper Submissions'' and ``Instructions'').
     The https://www.regulations.gov electronic filing system 
will accept objections until midnight Eastern Time at the end of June 
20, 2018.

Written/Paper Submissions

    Submit written/paper submissions as follows:
     Mail/Hand delivery/Courier (for written/paper 
submissions): Dockets Management Staff (HFA-305), Food and Drug 
Administration, 5630 Fishers Lane, Rm. 1061, Rockville, MD 20852.
     For written/paper objections submitted to the Dockets 
Management Staff, FDA will post your objection, as well as any 
attachments, except for information submitted, marked and identified, 
as confidential, if submitted as detailed in ``Instructions.''
     Objections received by mail/hand delivery/courier (for 
written/paper

[[Page 23383]]

submissions) will be considered timely if they are postmarked or the 
delivery service acceptance receipt is on or before June 20, 2018.
    Instructions: All submissions received must include the Docket No. 
FDA-2015-F-3663 for ``Grocery Manufacturers Association; Denial of Food 
Additive Petition.'' Received objections, those filed in a timely 
manner (see ADDRESSES), will be placed in the docket and, except for 
those submitted as ``Confidential Submissions,'' publicly viewable at 
https://www.regulations.gov or at the Dockets Management Staff between 
9 a.m. and 4 p.m., Monday through Friday.
     Confidential Submissions--To submit an objection with 
confidential information that you do not wish to be made publicly 
available, submit your objections only as a written/paper submission. 
You should submit two copies total. One copy will include the 
information you claim to be confidential with a heading or cover note 
that states ``THIS DOCUMENT CONTAINS CONFIDENTIAL INFORMATION.'' The 
Agency will review this copy, including the claimed confidential 
information, in its consideration of comments. The second copy, which 
will have the claimed confidential information redacted/blacked out, 
will be available for public viewing and posted on https://www.regulations.gov. Submit both copies to the Dockets Management 
Staff. If you do not wish your name and contact information to be made 
publicly available, you can provide this information on the cover sheet 
and not in the body of your comments and you must identify this 
information as ``confidential.'' Any information marked as 
``confidential'' will not be disclosed except in accordance with 21 CFR 
10.20 and other applicable disclosure law. For more information about 
FDA's posting of comments to public dockets, see 80 FR 56469, September 
18, 2015, or access the information at: https://www.gpo.gov/fdsys/pkg/FR-2015-09-18/pdf/2015-23389.pdf.
    Docket: For access to the docket to read background documents or 
the electronic and written/paper comments received, go to https://www.regulations.gov and insert the docket number, found in brackets in 
the heading of this document, into the ``Search'' box and follow the 
prompts and/or go to the Dockets Management Staff, 5630 Fishers Lane, 
Rm. 1061, Rockville, MD 20852.

FOR FURTHER INFORMATION CONTACT: Ellen Anderson, Center for Food Safety 
and Applied Nutrition (HFS-265), Food and Drug Administration, 5001 
Campus Dr., College Park, MD 20740-3835, 240-402-1309.

SUPPLEMENTARY INFORMATION:

I. Introduction

    In a document published in the Federal Register on October 28, 2015 
(80 FR 65978), we announced that we filed FAP 5A4811 (``petition'') 
submitted by the Grocery Manufacturers Association, 1350 I St. NW, 
Suite 300, Washington, DC 20005 (``petitioner''). The petitioner 
requested that we amend the food additive regulations in 21 CFR part 
172 Food Additives Permitted for Direct Addition to Food for Human 
Consumption to provide for the safe use of partially hydrogenated 
vegetable oils (PHOs) in the following food applications at specified 
maximum use levels: as a carrier or component thereof for flavors or 
flavorings, as a diluent or component thereof for color additives, as 
an incidental additive or processing aid, and as a direct additive in 
approximately 60 food categories. The petition was submitted in 
response to FDA's declaratory order issued on June 17, 2015 (80 FR 
34650), announcing our final determination that there is no longer a 
consensus among qualified experts that PHOs are generally recognized as 
safe for any use in human food. In the declaratory order, we invited 
submission of food additive petitions with scientific evidence for one 
or more specific uses of PHOs for which the petitioner believes that 
safe conditions of use may be prescribed (as further discussed in 
section II).
    FAP 5A4811 was submitted by GMA to FDA on June 11, 2015. During our 
initial review, we determined that the petition did not contain an 
environmental assessment as required under 21 CFR 25.15(a); therefore, 
we informed GMA that their petition did not meet the minimum 
requirements for filing in accordance with 21 CFR 171.1(c). On 
September 18, 2015, GMA resubmitted a complete FAP 5A4811, which we 
subsequently filed on October 1, 2015. During our initial review of FAP 
5A4811, we identified several deficiencies that required resolution by 
GMA for us to continue with our review. We issued a letter to GMA on 
March 21, 2016, explaining the additional information required to 
resolve the petition's deficiencies. On May 5, 2016, GMA submitted a 
partial response to the deficiencies. The petition was then placed in 
abeyance by FDA, consistent with our procedures for food additive 
petitions.\1\ The petitioner and FDA met several times in the months 
following to discuss the deficiencies.
---------------------------------------------------------------------------

    \1\ Abeyance is an administrative category of petitions that are 
filed but non-active because of deficiencies that were identified 
during FDA's review. A petition remains in abeyance until either the 
petitioner provides FDA with the required information, requests a 
final decision based on the data currently in the petition, or 
requests withdrawal of the petition.
---------------------------------------------------------------------------

    On March 7, 2017, the petitioner submitted a substantive amendment 
to FAP 5A4811 that addressed the deficiencies identified by FDA. In 
accordance with 21 CFR 171.6, the petition was assigned a new filing 
date of March 7, 2017. The amended petition contained significant 
revisions to the proposed uses, exposure estimate, and safety 
assessment of PHOs. The revised petitioned uses of PHOs were limited to 
the following: (1) As a solvent or carrier for flavoring agents, flavor 
enhancers, and coloring agents; (2) as a processing aid, and (3) as a 
pan release agent for baked goods. Based on the revisions, the 
petitioner asserted that the amended uses of PHOs would present a de 
minimis increase in risk (in other words, a negligible increase in 
risk) and, therefore, are safe under the conditions of intended use. 
References to the ``petition'' henceforth in this document will denote 
the amended petition received on March 7, 2017.

II. Background

A. Statutory and Regulatory Requirements Regarding Food Additives

    The Federal Food, Drug, and Cosmetic Act (FD&C Act) defines ``food 
additive,'' in relevant part, as any substance, the intended use of 
which results or may reasonably be expected to result, directly or 
indirectly, in its becoming a component of food, if such substance is 
not generally recognized by experts as safe under the conditions of its 
intended use (section 201(s) of the FD&C Act (21 U.S.C. 321(s))). Food 
additives are deemed unsafe and prohibited except to the extent that 
FDA approves their use (sections 301(a) and (k) (21 U.S.C. 331(a) and 
(k)) and 409(a) (21 U.S.C. 348(a)) of the FD&C Act.)
    The FD&C Act provides a process through which persons who wish to 
use a food additive may submit a petition proposing the issuance of a 
regulation prescribing the conditions under which the additive may be 
safely used (section 409(b)(1) of the FD&C Act). When FDA concludes 
that a proposed use of a food additive is safe, we issue a regulation 
authorizing a specific use of the substance.

B. Relevant Regulatory History of PHOs

    On November 8, 2013, FDA issued a document (the tentative 
determination,

[[Page 23384]]

78 FR 67169), announcing our tentative determination that PHOs are no 
longer generally recognized as safe (GRAS) under any condition of use 
in food and therefore are food additives subject to section 409 of the 
FD&C Act. Because PHOs are the primary dietary source of industrially-
produced trans fatty acids (IP-TFA), FDA's evaluation of the GRAS 
status of PHOs centered on the trans fatty acid (TFA, also referred to 
as ``trans fat'') component of these fats and oils. The tentative 
determination cited current scientific evidence of significant human 
health risks, namely an increased risk in coronary heart disease (CHD), 
associated with the consumption of IP-TFA (78 FR 67169 at 67172). The 
scientific evidence included results from controlled feeding studies on 
trans fatty acid consumption in humans, findings from long-term 
prospective epidemiological studies, and the opinions of expert panels 
that there is no threshold intake level for IP-TFA that would not 
increase an individual's risk of CHD (78 FR 67169 at 67172). Based on 
the evidence outlined in the tentative determination, we determined 
that there is no longer a consensus among qualified experts that PHOs 
are safe for human consumption (i.e., PHOs do not meet the GRAS 
criteria.) The tentative determination also requested interested 
parties to submit comments and additional scientific data related to 
our tentative determination that PHOs are no longer GRAS (78 FR 67169 
at 67174).
    We received over 6000 comments in response to the tentative 
determination. We reviewed the comments before issuing our final 
determination as a declaratory order published on June 17, 2015 (the 
declaratory order, 80 FR 34650). The declaratory order included four 
major provisions: (1) PHOs are not GRAS for any use in human food; (2) 
for the purposes of the declaratory order, FDA defined PHOs as those 
fats and oils that have been hydrogenated, but not to complete or near 
complete saturation, and with an iodine value greater than 4 as 
determined by an appropriate method; (3) any interested party may seek 
food additive approval for one or more specific uses of PHOs with data 
demonstrating a reasonable certainty of no harm of the proposed use(s); 
and (4) FDA established a compliance date of June 18, 2018 (80 FR 34650 
at 34651).
    In our declaratory order finding that PHOs are no longer GRAS for 
any use in human food, we acknowledged that scientific knowledge 
advances and evolves over time. The declaratory order invited 
submission of scientific evidence as part of food additive petitions 
under section 409 of the FD&C Act for one or more specific uses of PHOs 
for which industry or other interested individuals believe that safe 
conditions of use may be prescribed. We also established a three-year 
delayed compliance date (compliance required no later than June 18, 
2018) to provide time for submission and review and, if applicable 
requirements are met, approval of food additive petitions for uses of 
PHOs (80 FR 34650 at 34668).

III. Evaluation of Safety

    A food additive cannot be approved for use unless the data 
presented to us establish that the food additive is safe for that use 
(section 409(c)(3)(A) of the FD&C Act). To determine whether a food 
additive is safe, the FD&C Act requires us to consider among other 
relevant factors: (1) Probable consumption of the additive; (2) 
cumulative effect of such additive in the diet of man or animals, 
taking into account any chemically or pharmacologically related 
substances in the diet; and (3) safety factors generally recognized by 
experts as appropriate for the use of animal experimentation data 
(section 409(c)(5) of the FD&C Act). Our determination that a food 
additive use is safe means that there is a ``reasonable certainty in 
the minds of competent scientists that the substance is not harmful 
under the intended conditions of use'' (Sec.  170.3(i) (21 CFR 
170.3(i))).
    FAP 5A4811 is not a typical food additive petition in that it is 
requesting food additive approval for existing uses of PHOs that 
industry, independent of FDA, had concluded were GRAS, but FDA 
subsequently determined such uses are not GRAS. Most food additive 
petitions seek premarket approval for new uses of food additives. 
Additionally, the approach that we normally use to evaluate safety of a 
direct food additive is not applicable for assessing the safety of IP-
TFA in PHOs. Food additives are typically evaluated based on 
toxicological studies in animals, as described in our guidance, 
Toxicological Principles for the Safety of Assessment of Food 
Ingredients (also known as Redbook 2000).\2\ However, key scientific 
evidence for the association of trans fat and CHD is based on human 
studies, including controlled feeding trials of trans fat intake and 
blood cholesterol levels in humans and long-term, prospective 
observational studies of trans fat intake and CHD risk in human 
populations (Ref. 1).
---------------------------------------------------------------------------

    \2\ Redbook 2000 is available at https://www.fda.gov/downloads/Food/GuidanceRegulation/UCM222779.pdf.
---------------------------------------------------------------------------

    To establish with reasonable certainty that a food additive is not 
harmful under its intended conditions of use, we typically consider the 
projected human dietary exposure to the additive, the additive's 
toxicological data provided by the petitioner, and other relevant 
information (such as published literature) available to us. FDA 
scientists use these toxicological data (usually derived from animal 
and in vitro studies) to determine a no-observed effect level or a no-
observed-adverse-effect-level, apply an appropriate safety factor to 
account for differences between animals and humans and differences in 
sensitivity among humans, and calculate the acceptable daily intake 
(ADI) for the food additive. The ADI is usually expressed in milligrams 
of food additive per kilogram body weight of humans. We compare an 
individual's estimated daily intake (EDI) of the additive from all food 
sources to the ADI established by toxicological data. The EDI is 
determined based on the amount of the additive proposed for use in 
particular foods and the amount of those foods consumed containing the 
additive, and on the amount of the additive from all other dietary 
sources. We typically use the EDI for the 90th percentile consumer of a 
food additive as a measure of high chronic dietary exposure. A food 
additive is generally considered safe for its intended uses if the EDI 
of the additive is less than the ADI. This approach assumes that a 
physiological threshold may exist below which exposure to an additive 
will not cause harm. In the case of PHOs, which contribute IP-TFA to 
the diet, the main toxicological data available to assess safety 
consists of controlled feeding trials and prospective observational 
studies in humans where the adverse health outcomes associated with the 
additive are increased CHD risk and other non-cancer risks (e.g., 
stroke). To receive approval for the petitioned uses of PHOs, the 
petitioner has the responsibility to provide scientific evidence that 
establishes that the intended uses of PHOs are safe, including the 
expected dietary exposure to trans fat resulting from the intended uses 
of PHOs.
    Our declaratory order references three safety memoranda prepared by 
FDA that document our review of the available scientific evidence 
regarding human health effects of trans fat, focusing on the adverse 
effects of trans fat on risk of CHD (Refs. 2-4). In addition, we 
previously reviewed the health effects of IP-TFA and PHOs in support of 
our tentative determination that PHOs are not GRAS in food (78 FR 
67169) and in

[[Page 23385]]

1999 and 2003 in support of our proposed and final rules requiring 
declaration of trans fat in nutrition labeling of food (64 FR 62746 and 
68 FR 41434). The safety reviews for the declaratory order, together 
with the previous safety reviews of IP-TFA and PHOs, provide important 
background scientific information for our review of FAP 5A4811.
    The petition contains a review of recent scientific literature and 
expert opinions on trans fat consumption. GMA asserted that this 
information supports the following three conclusions, which are their 
reasons why they believe the petitioned uses of PHOs are safe:
    1. ``The conservatively estimated probability of coronary heart 
disease risk falls below the probable de minimis non-cancer risk 
range.'' \3\
---------------------------------------------------------------------------

    \3\ As discussed in section E, the petitioner calculates what it 
considers to be de minimis risks for non-cancer health outcomes.
---------------------------------------------------------------------------

    2. ``iTFA \4\ exposure from the petitioned uses of PHOs (i.e., 
0.05%en [total energy intake per day]) is well below exposure levels in 
controlled feeding trials, and effects at these low iTFA exposures 
levels cannot be empirically established based on the currently 
available evidence.''
---------------------------------------------------------------------------

    \4\ The petitioner uses the abbreviation iTFA to refer to 
industrially-produced TFA in the petition.
---------------------------------------------------------------------------

    3. ``The incremental increase in iTFA intake of 0.05%en from the 
petitioned uses of PHOs is infinitesimally small and negligible in 
comparison to existing background dietary TFA exposure from intrinsic 
sources.''

(Petition, pp. 116-119)
    In this petition denial, we discuss our evaluation of the 
petitioner's request and supporting information in section IV organized 
according to the following headings: A. Chemical Identity, Intended 
Technical Effects, and Petitioned Uses of PHOs; B. Estimated Exposure 
to Trans Fat; C. Recent Scientific Literature and Expert Opinions on 
Trans Fat Consumption; D. Recent Threshold Dose-Response Research; and 
E. Risk Estimates and Safety Arguments. Each of these sections provides 
a summary of the information provided by the petitioner followed by our 
evaluation of that information, prefaced with ``FDA Assessment.'' 
Additional information regarding our evaluation of the petition can be 
found in our three review memoranda (Refs. 5-7).

IV. FDA's Review of FAP 5A4811

A. Chemical Identity, Intended Technical Effects, and Petitioned Uses 
of PHOs

    The PHOs that are the subject of FAP 5A4811 are made from the 
following vegetable oils: Soy, cottonseed, coconut, canola, palm, palm 
kernel, and sunflower oils, or blends of these oils, and consist of up 
to 60 percent trans fatty acids. As discussed in section I, GMA 
requested approval of three uses of PHOs, which are as follows:
     PHO, or a blend of PHOs, used as a solvent or carrier, or 
a component thereof, for flavoring agents, flavor enhancers, and 
coloring agents intended for food use, provided the PHOs in the solvent 
or carrier contribute no more than 150 parts per million (ppm) (150 
milligrams per kilogram (mg/kg)) IP-TFA to the finished food as 
consumed;
     PHO, or a blend of PHOs, used as a processing aid, or a 
component thereof, provided the PHOs in the processing aid contribute 
no more than 50 ppm (50 mg/kg) IP-TFA to the finished food as consumed;
     PHO, or a blend of PHOs, used as a pan release agent for 
baked goods at levels up to 0.2 grams/100 grams (0.2 g/100 g) in pan 
release spray oils, provided the PHO contributes no more than 0.14 g 
IP-TFA/100 g spray oil.
    These proposed uses excluded dietary supplements. The physical and 
technical effects of the petitioned uses of PHOs were specified as: 
Release agents, either alone or in combination with other components 
(Sec.  170.3(o)(18)); processing aids or components thereof (Sec.  
170.3(o)(24)); and as solvents, carriers and vehicles for fat soluble 
coloring agents, flavoring agents, and flavor enhancers (Sec.  
170.3(o)(27)).
FDA Assessment
    To better understand how PHOs would be used as processing aids, we 
requested that the petitioner provide specific examples. In an email 
dated May 15, 2017, the petitioner provided several examples of how 
PHOs may be used as processing aids. Many of the petitioner's examples 
involved the use of PHOs as a topical coating to prevent rancidity 
(e.g., PHO-coated almond slices or candy pieces used as ingredients in 
cookies). We view this use of PHOs as having an ongoing technical 
effect in food (e.g., to prevent rancidity and oxidation) and, 
therefore, we do not agree that this use would be considered a 
processing aid in accordance with Sec. Sec.  170.3(o)(24) and 
101.100(a)(3)(ii) (21 CFR 101.100(a)(3)(ii)). Because we are denying 
this petition, we did not need to resolve this issue regarding 
characterization of the technical or functional effect of these 
additives.

B. Estimated Exposure to Trans Fat

    The petitioner provided exposure estimates for TFA from the 
petitioned uses of PHOs and from intrinsic (i.e., naturally-occurring) 
sources such as dairy and meat from ruminant animals. To estimate 
exposure, the petitioner used food disappearance data from 2014 
compiled by the U.S. Department of Agriculture (USDA) Economic Research 
Service, food consumption data from either the 2007-2010 or 2009-2012 
National Health and Nutrition Examination Surveys (NHANES), and the 
intrinsic concentrations of TFA in the USDA National Nutrient Database 
for Standard Reference Release 27. The petitioner estimated the 
exposure to naturally-occurring TFA from intrinsic sources for the U.S. 
population (aged 2 years or more) to be 1.04 grams/person/day (g/p/d) 
at the mean and 1.91 g/p/d at the 90th percentile. If expressed as a 
percentage of total energy intake per day (%en), based on a 2000 
calorie daily diet, the exposure to TFA from intrinsic sources would be 
0.46%en at the mean and 0.75%en at the 90th percentile for the U.S 
population. The petitioner estimated the cumulative exposure to IP-TFA 
from all petitioned uses of PHOs in foods for the U.S. population aged 
2 years or more to be 0.121 g/p/d (0.05%en) at the mean and 0.122 g/p/d 
(0.05%en) at the 90th percentile.
FDA Assessment
    FDA agrees with the petitioner's estimated exposure to TFA from 
intrinsic sources, and we have no concerns regarding the general 
methodology used by the petitioner to estimate exposure to IP-TFA from 
the petitioned uses of PHOs. However, we believe the petitioner likely 
underestimated exposure to IP-TFA from the petitioned uses of PHOs for 
various reasons, such as their determination that 43 percent of the 
U.S. diet consists of processed foods, which we believe is too low, and 
not including all relevant NHANES food codes in their exposure estimate 
(Ref. 5). Although the petitioner's exposure estimate could be refined, 
we consider it sufficient for approximating exposure from the 
petitioned uses of PHOs.

C. Recent Scientific Literature and Expert Opinions on Trans Fat 
Consumption

    FAP 5A4811 included sections on dietary guidelines and expert panel 
opinions pertaining to trans fat consumption. In addition, the petition

[[Page 23386]]

presented a summary of studies assessing the effects of dietary TFA on 
intermediate biomarkers such as low-density lipoprotein cholesterol 
(LDL-C), high-density lipoprotein cholesterol (HDL-C), and other 
emerging biomarkers of CHD risk, and the association of dietary TFA 
intake with risk of CHD and risk of adverse health outcomes other than 
CHD (e.g., stroke, metabolic syndrome). Controlled feeding trials, 
prospective observational studies, and meta-analyses of these studies 
were included in the petitioner's scientific literature review.
FDA Assessment
    As discussed in our review memorandum (Ref. 7), we found that the 
petitioner provided incomplete information on certain topics or 
misinterpreted some scientific conclusions.
1. Dietary Guidelines and Expert Panel Reviews
    The petition discussed the major expert panel reports on the health 
effects of trans fat consumption from the U.S., Australia, Canada, the 
United Kingdom, the World Health Organization (WHO), the Food and 
Agriculture Organization, and the European Food Safety Authority. We 
note that while the petition provided a generally accurate summary of 
these expert reports, some important information was missing or 
understated. For example, the petition omits the expert opinions on the 
role of HDL-C as a biomarker for CHD. The petition also omits that, in 
addition to the Institute of Medicine's 2005 report (Ref. 8), many 
other expert panels have concluded that TFA has a progressive and 
linear adverse effect on blood lipids and associated CHD risk. 
Furthermore, the petition understated the recommendation from several 
expert panels that trans fat intake should be kept as low as possible 
by specifically limiting intake of IP-TFA from PHOs.
2. Effect of Changes In Trans Fat Intake on LDL-C and HDL-C
    The petition identified five meta-analysis studies (which are 
combined analyses of multiple feeding trials) that quantified the 
effect of changes in trans fat intake on LDL-C and HDL-C in the blood 
of human test subjects. The petition's summary of these studies was 
appropriate; however, we note that two available meta-analyses studies 
were not included in the petition's discussion: Zock and co-workers 
(Refs. 9-11) and Brouwer (Ref. 12). In particular, the 2016 meta-
analysis by Brouwer was an important study, commissioned by the WHO 
Nutrition Guidance Expert Advisory Group (NUGAG) Subgroup on Diet and 
Health, that affirmed the linear, progressive effect of trans fat 
intake on blood cholesterol levels (Ref. 12).
    The petition mentioned another meta-analysis of newer studies 
conducted by Hafekost et al. (2015) which reported no significant 
effect on LDL-C from a 1%en TFA intake (including both naturally-
occurring TFA and IP-TFA) in exchange for cis-monounsaturated fatty 
acids (cis-MUFA) (Ref. 13). The petition claimed that these results 
support the potential for a threshold trans fat intake below which no 
significant effect on blood lipids is observed. However, we disagree 
with the petitioner's interpretation of this study's conclusions (Ref. 
7). We note that the criteria for inclusion of feeding trials in this 
meta-analysis were not rigorous. In several of the included studies, 
the diets were not fully controlled. We also note that Hafekost et al. 
did not conclude that their results supported the potential for a safe 
threshold intake level of TFA. Rather, the authors stated, ``An 
increase in LDL was consistent with the results of Brouwer et al., who 
identified a significant increase in LDL cholesterol with a percent 
increase in the intake of industrial TFA.'' Furthermore, Hafekost et 
al. conducted an additional analysis, including the earlier Brouwer et 
al. meta-analysis results together with their analysis of newer studies 
alone. The petition did not discuss these additional analyses. The 
combined results for the newer studies alone, together with the earlier 
meta-analysis, showed a statistically significant increase in LDL-C due 
to an increase of 1%en intake from TFA. In their overall summary, 
Hafekost et al. stated, ``The results of the current review are 
consistent with previous evidence which indicates a detrimental effect 
of consumption of TFA on changes in LDL and HDL blood cholesterol'' 
(Ref. 13).
    Regarding HDL-C and CHD risk, the petition underemphasized the 
impact of trans fat intake on HDL-C. We note that the observed decrease 
in HDL-C due to TFA intake is consistently reported across the existing 
body of TFA research and that HDL-C has been recognized as a major risk 
factor for CHD (Ref. 7).
3. Prospective Observational Studies
    The petition reviewed the results of prospective observational 
studies that estimate the association of long-term, habitual TFA intake 
with CHD risk in large, free-living populations. The petition reviewed 
five meta-analysis studies (that provided combined analyses of several 
individual prospective observational studies). The petition stated that 
the results of a recent meta-analysis by de Souza et al. in 2015 (Ref. 
14) were consistent with previous meta-analyses in finding a 
statistically significant increased risk of CHD when comparing high to 
low TFA intake. Regarding individual prospective observational studies, 
the petition stated that, ``The results from these studies, while not 
able to demonstrate causality, provide supporting evidence that, 
although a relationship between increased CHD risk and high levels of 
TFA intake exists, this observed relationship is largely based on 
comparisons of differences in TFA intake above 1%en and has not been 
established at lower levels of intake.''
    We note that the overall results of the meta-analyses and recently 
published prospective observational studies were generally summarized 
accurately in the petition. However, the petition tended to understate 
the strength of the evidence from the observational studies reviewed. 
In particular, the meta-analysis by de Souza et al., a rigorously 
conducted study commissioned by WHO NUGAG, stated that the ``positive 
associations between trans fat intake and CHD and CHD mortality'' were 
``reliable and strong'' and provided supplementary analyses supporting 
a progressive and linear association of TFA intake and CHD risk (Ref. 
14). Additionally, recently published studies by Li et al. in 2015 
(Ref. 15) and Wang et al. in 2016 (Ref. 16), with long-term followup 
and increased statistical power, show significant increases in CHD or 
cardiovascular disease (CVD) risk at lower increments of TFA intake 
than the 1%en stated by the petitioner.
4. Other Health Outcomes
    The petitioner concluded, after reviewing recent scientific 
literature, that there is limited, inconsistent, and/or weak evidence 
for any effects of trans fat intake on other health outcomes including 
stroke, all-cause mortality, cancer, and metabolic syndrome. We do not 
agree with the petitioner's conclusion, in particular regarding stroke. 
In support of the declaratory order, we reviewed several well-conducted 
studies that provided a reasonable basis to conclude that TFA intake is 
associated with an increased risk of ischemic stroke (a blockage of 
blood flow to the brain) (Ref. 2). Furthermore, in our review 
memorandum for this petition, we described more recent studies that 
provide additional evidence supporting the association of TFA with 
stroke, as well as total mortality and elements of metabolic syndrome 
(Ref. 7).

[[Page 23387]]

D. Recent Threshold Dose-Response Research

    The petition acknowledged that all five of the aforementioned meta-
analyses (see section C) relied on a linear, no-threshold dose-response 
relationship between TFA intake and blood levels of LDL-C and HDL-C, 
which assumes any amount of TFA greater than 0%en causes adverse 
effects on blood cholesterol levels. The petition stated, ``Recent 
research suggests that a non-threshold linear dose-response model 
overlooks the complexities of the physiological effects of 
macronutrients and other contributing factors to LDL-C levels besides 
TFAs.'' In particular, the petition cited two recent articles to 
support the claims that a linear dose-response model is inappropriate 
for assessing the effects of TFA consumption on blood lipids, and that 
a threshold level of trans fat intake exists (Refs. 17 and 18). In the 
first publication, Reichard and Haber (Ref. 17) presented and evaluated 
a hypothesis for the biological mode of action (MOA) for the effect of 
TFA on LDL-C based on animal studies. According to the petition, ``. . 
. the authors concluded the key events in the MOA are the increased 
production of very low density-lipoprotein (VLDL) and decreased LDL-
clearance due to a reduction in the LDL-C mediated receptor activity.'' 
The authors further concluded the effect of TFA on LDL-C is non-linear 
and there is evidence that either a threshold exists or the dose-
response slope is very shallow at low dose levels (Ref. 17).
    In the second article, Allen et al. (Ref. 18) conducted a meta-
regression study of human controlled feeding trials, that considered 
both linear and nonlinear dose-response models to assess the effect of 
IP-TFA intake on LDL-C and determine which shape fit best with the MOA 
proposed by Reichard and Haber based on animal studies. (In this case, 
the meta-regression is a meta-analysis that focuses on dose-response 
relationships.) The Allen et al. meta-regression used an evidence map 
to identify additional experimental data for the effect of IP-TFA 
intake on LDL-C, particularly in the low dose region of the response 
curve where IP-TFA intake is between zero and 3%en (Ref. 19). According 
to Allen et al., an S-shaped model with an assumed threshold at low IP-
TFA doses explained more of the study-to-study variability compared to 
the linear dose-response model (Ref. 18). Using assumptions about 
intra-individual measurement variation for LDL-C and the S-shaped 
model, the authors concluded that the change in LDL-C associated with a 
change in IP-TFA intake of 2.2%en represented a biologically 
meaningless change (Ref. 18). The petition stated that this analysis 
supports the existence of a threshold level of IP-TFA intake, below 
which negligible changes in LDL-C would occur.
FDA Assessment
    We do not agree that these two studies cited by the petitioner 
provide convincing evidence to refute a linear dose-response or provide 
convincing evidence of a threshold in the effect of IP-TFA on LDL-C. In 
our review, we identified several design flaws and questionable data 
interpretations associated with these two studies (Ref. 7). One major 
concern about the MOA paper (Ref. 17) is that the authors relied 
largely on data from laboratory animal models to hypothesize an MOA 
that suggests the existence of a threshold effect of TFA on LDL-C in 
humans, despite the differences in biological response to dietary fats 
and fatty acid metabolism between humans and the animal species used in 
the study (e.g., rodents). The authors acknowledged that trans fatty 
acids such as elaidic acid do not increase serum LDL-C in hamsters, and 
suggest that animal models may underestimate the effect of TFA in 
humans (Ref. 17).\5\
---------------------------------------------------------------------------

    \5\ The scientific evidence that PHOs are no longer GRAS for use 
in food was not based on animal studies, such as those used in the 
Reichard and Haber MOA, but rather included results from controlled 
feeding studies on trans fatty acid consumption in humans, findings 
from long-term prospective epidemiological studies in human 
populations, and the opinions of expert panels that there is no 
threshold intake level for IP-TFA that would not increase an 
individual's risk of CHD (78 FR 67169 at 67172).
---------------------------------------------------------------------------

    Regarding the meta-regression paper (Ref. 18), we found that 
duplicate data points were erroneously used in the analysis; the 
validity of data points for low TFA levels below 3%en was questionable, 
and the low TFA data did not come from PHO test diets; and incorrect 
variances were applied in the weighting of the data based on the study 
designs (Ref. 7). We also question the authors' suggestion that the 
within person, day-to-day variability of blood LDL-C levels can be used 
to represent the minimum increment in LDL-C that is adverse (Ref. 7). 
Additionally, we note that the authors' proposed S-shaped dose-response 
model that levels off at high trans fat doses (above 3%en) is not 
consistent with the results of numerous controlled feeding trials of 
IP-TFA at higher doses or with prospective observational studies which 
show increases in serum LDL-C levels or CHD risk with higher intakes of 
trans fat (Ref. 7).

E. Risk Estimates and Safety Arguments

    The petition contained an estimate of ``hypothetical change'' in 
CHD risk associated with 0.05%en IP-TFA intake (the daily amount of 
energy from IP-TFA contributed by the petitioned uses of PHOs) that was 
based on FDA's four deterministic quantitative risk assessment methods 
referenced in the declaratory order (Ref. 4). The petitioner stated 
that they included this analytical approach in the petition ``for 
expediency and at the request of FDA'', although the petition 
questioned the validity of a linear-no threshold dose-response model 
for IP-TFA intake and LDL-C and HDL-C on which the FDA method is based. 
The deterministic quantitative risk assessment approach used by the 
petitioner estimated the change in CHD risk due to effects on blood 
lipoproteins from controlled feeding trials, and also estimated the 
change in CHD risk using direct observations of CHD from prospective 
studies when there is an isocaloric replacement of cis-MUFA with IP-TFA 
in the diet. The petitioner estimated that the change in CHD risk 
associated with a 0.05%en added IP-TFA intake from petitioned uses 
ranged from 0.062 percent to 0.665 percent depending on the risk method 
used. When expressed as a population-based risk estimate, the annual 
probability of CHD cases per 100,000 U.S. adults aged 35 and older 
ranged from 0.42 to 4.54. In other words, for every 100,000 U.S. 
adults, there could be up to 4.54 additional cases (fatal and non-
fatal) of CHD attributed to an intake of 0.05%en IP-TFA from the 
petitioned uses of PHOs.
    The petition asserts a standard of ``de minimis risk.'' According 
to the petitioner, a de minimis risk implies that a risk is so small 
that it should be ignored, and the petitioned use should be considered 
safe. The petitioner referenced three arguments to explain its de 
minimis risk principle: (1) The probability of a risk is below an 
acceptable cutoff (i.e., ``bright line'' or threshold); (2) there is a 
lack of scientific data to establish that the risk exists (i.e., the 
risk is non-detectable); or (3) the probability of the risk is less 
than the natural occurrence of the risk (Ref. 20). While neither the 
FD&C Act nor FDA's regulations regarding the evaluation of the safety 
of food additives in response to a food additive petition refer to de 
minimis risk, we review each of these arguments in turn.

[[Page 23388]]

1. De minimis ``Bright Line'' or Threshold Argument
    The petition referenced an article by Castorina and Woodruff (Ref. 
21) in which the authors estimated risks for non-cancer health outcomes 
from hypothetical lifetime ingestion or inhalation exposures to select 
environmental chemicals at the U.S. Environmental Protection Agency's 
(EPA) established reference doses (RfDs) or reference concentrations. 
The authors concluded that the non-cancer risk associated with RfDs 
ranged from 1 in 10,000 (1 x 10-4) to 5 in 1,000 (5 x 
10-3) using a linear dose-response relationship for the 
environmental chemicals the authors selected. The petitioner applied a 
safety factor to the authors' risk estimates associated with RfDs to 
arrive at a proposed probability of risk, ranging from 2 in 100,000 (2 
x 10-5) to 1 in 1,000 (1 x 10-3), which the 
petitioner deemed to be a de minimis risk. The petitioner compared this 
risk range to the results of their quantitative risk estimate, which 
predicted the annual probability of CHD cases attributed to 0.05%en IP-
TFA intake from the petitioned PHO uses to be in the range of 0.42 per 
100,000 adults (or 4.2 x 10-6) to 4.5 per 100,000 adults (or 
4.5 x 10-5). The petition concluded that the estimated risk 
from 0.05%en IP-TFA intake from petitioned PHO uses is de minimis 
because it is well below the probable de minimis risk ranges for non-
cancer risk calculated by applying a safety factor to the risks 
presented in the Castorina and Woodruff article.
FDA Assessment
    We will first address the petitioner's reliance on the Castorina 
and Woodruff paper to determine the concept of de minimis risk, 
followed by our comments on the petitioner's deterministic risk 
assessment. We will also include a discussion of the probabilistic risk 
assessment that we conducted as part of our review.
a. Castorina and Woodruff Study
    We disagree with the petitioner's interpretation of the Castorina 
and Woodruff article on which the petitioner's safety conclusion is 
based. The application of the Castorina and Woodruff study results has 
limitations as a basis for inferring that IP-TFA from petitioned PHO 
uses is safe because it represents de minimis risk. The study is a 
single, exploratory analysis of whether EPA reference values represent 
negligible risk levels; it is not a consensus that defines a concept of 
de minimis risk or safe exposure. In fact, the study authors themselves 
question whether the non-cancer risks associated with the EPA's 
reference values represent ``acceptable levels'' of exposure from a 
public health perspective (Ref. 21). Furthermore, we note that in the 
Castorina and Woodruff paper, the estimated risks were based on 
biochemical and physiological changes associated with several non-
cancer health outcomes that are much less serious than CHD cases or CHD 
deaths. For example, some of the biochemical and physiological changes 
the authors considered included small intestinal lesions, fatty cyst 
formation in the liver, elevated serum glutamate-pyruvate 
transaminases, chronic irritation of stomach, decreased lymphocyte 
count, changes in red blood cell volumes, decreased mean terminal body 
weights, and decreased maternal body weight gain. Therefore, we 
conclude that the petitioner's use of this single article to support 
their de minimis risk argument regarding the risk of CHD or CHD death 
associated with IP-TFA exposure is inadequate.
b. Petitioner's Quantitative Deterministic Risk Assessment
    The petitioner relied on the de minimis risk principle to conclude 
that the petitioned uses of PHOs are safe because the estimated 
probability of CHD risk associated with IP-TFA from the petitioned uses 
of PHOs falls below the probable de minimis non-cancer risk range. The 
petition included a quantitative deterministic risk assessment that 
estimated the annual probability of CHD cases that may be associated 
with IP-TFA from petitioned uses of PHOs ranged from 0.42 to 4.54 per 
100,000 U.S. adults. We note, though, that the petition did not include 
an estimated annual number of CHD cases or estimated annual number of 
CHD deaths associated with IP-TFA from the proposed uses of PHOs. Using 
the petitioner's estimated annual rate of CHD cases per 100,000 adults, 
the U.S. Census estimate of 166.7 million adults in the U.S. population 
in 2014, and a 32 percent CHD fatality rate reported by the Centers for 
Disease Control and Prevention (CDC), we expanded the petitioner's risk 
estimates associated with IP-TFA from petitioned uses of PHOs to 
estimate a range of 700 to 7,570 cases of CHD per year including 
between 224 and 2,422 deaths from CHD per year, which FDA does not 
consider to be insignificant (Ref. 7). Additionally, we conducted our 
own deterministic risk assessment to verify that the petitioner's 
methods were appropriate, and we expanded our analysis to include a 
probabilistic risk assessment to further bolster our decision that the 
estimated risks associated with the petitioned uses of PHOs cause them 
to be unsafe food additives (Ref. 6).
c. FDA's Quantitative Probabilistic Risk Assessment
    The deterministic risk assessment approach that was used by both 
the FDA in our declaratory order and by the petitioner in FAP 5A4811 to 
assess CHD risk associated with IP-TFA exposure is a risk assessment 
approach using assigned values for discrete scenarios (e.g., using most 
likely scenarios or mean values) (Ref. 6). The deterministic approach 
determines the robustness of the risk of CHD. However, it has 
limitations in that it is inadequate in applying population or other 
parameter variability information and it takes into consideration only 
a few discrete results (e.g., mean risk estimates), overlooking many 
others (e.g., probability distributions of risk estimates). The impact 
of different risk parameter values and uncertainty in risk methods 
relative to results also cannot be quantified (Ref. 6).
    The probabilistic approach allows for the analysis of human 
variability and uncertainty in the risk method to be incorporated into 
both the exposure and risk assessments, if high quality empirical data 
with the probability distribution information for key parameters are 
used in the risk assessment (Ref. 6). We considered that at the 
petitioned IP-TFA exposure of 0.05%en, there would be greater 
uncertainty in the CHD risk estimates than the IP-TFA exposure of 
0.5%en which was used in the declaratory order, and that the mean risk 
estimates alone would not be sufficient to demonstrate safety. 
Therefore, we conducted a probabilistic risk assessment for the CHD 
risk associated with an IP-TFA exposure of 0.05%en taking into 
consideration the variability and uncertainty associated with IP-TFA 
exposure and the risk parameters, and estimated both the probabilistic 
means and the uncertainty around the means.
    We used FDA's four risk methods based on a linear no-threshold 
dose-response model (Ref. 6) to estimate changes in CHD risk when 
replacing cis-MUFA or saturated fatty acids at 0.05%en, with the same 
percentage of energy from IP-TFA. The probabilistic means were in line 
with the results estimated using the deterministic approach. The 
probabilistic approach also quantified the probability distribution of 
the risk estimates (e.g., the lower and upper 95 percent statistical 
uncertainty intervals (95

[[Page 23389]]

percent UIs)). The results included estimated changes in percent CHD 
risk, increases in the rate of annual CHD cases (both fatal and non-
fatal) per 100,000 U.S. adults, and increases in the number of annual 
CHD cases, including CHD deaths, among U.S. adults. We also extended 
Method 4 (prospective observational studies) to estimate the annual 
number of CVD deaths among this same population. (CVD deaths include 
deaths from CHD, strokes, and other vascular diseases.) Our assessment 
methodology is documented in our review memorandum (Ref. 6).
    Results from our probabilistic risk assessment demonstrate that 
consuming IP-TFA at a level of 0.05%en per person per day, instead of 
cis-MUFA, can cause a mean increase in annual CHD cases per 100,000 
U.S. adults from 0.478 (95 percent UI 0.299 to 0.676) using the FDA 
risk method based on changes of LDL-C alone (Method 1) to 4.038 (95 
percent UI 2.120 to 6.280) using the FDA risk method based on 
prospective observational studies (Method 4). These increases 
correspond to a mean increase in annual CHD cases from 814 (95 percent 
UI 510 to 1,151, using Method 1) to 6,877 (95 percent UI 3,611 to 
10,694, using Method 4), which includes annual deaths from CHD from 290 
(95 percent UI 182 to 410, using Method 1) to 2,450 (95 percent UI 
1,287 to 3,811, using Method 4). The other two FDA risk methods 
produced increases in risk values from CHD that were between those 
estimated by Method 1 and Method 4.
    The same amount of IP-TFA replacing saturated fatty acids would 
result in lower estimates of annual CHD cases and CHD-related deaths 
than those estimated by replacing cis-MUFA with IP-TFA. We estimated 
the mean increase in annual CHD cases to be 170 (using Method 1) to 
5,110 (using Method 4), which includes 60 to 1,821 annual deaths from 
CHD. Using extended Method 4, the same amount of IP-TFA replacing 
either saturated fatty acids or carbohydrate could cause more than 
6,500 CVD deaths per year in U.S. adults. The results of our analyses 
are described further in our review memorandum (Ref. 6).
    Our deterministic and probabilistic quantitative risk assessments 
demonstrate that there is a probable significant health risk associated 
with 0.05%en from IP-TFA from the petitioned uses of PHOs. Our analyses 
do not support the petitioner's claims that 0.05%en from IP-TFA results 
in de minimis risk or that there is a reasonable certainty that PHOs 
are not harmful under the intended conditions of use.
2. Non-Detectability Argument
    The petitioner argued that the estimated exposure to IP-TFA from 
petitioned uses of PHOs (i.e., 0.05%en) is well below the exposure 
levels in controlled feeding studies and effects at these low IP-TFA 
levels cannot be empirically established based on the currently 
available evidence. The petition questioned the appropriateness of 
using a linear dose-response model for quantifying the effect of lower 
levels of trans fat intake (i.e., <3%en) on LDL-C and HDL-C, and 
maintained that there is a general lack of empirical evidence that 
consumption of low levels of trans fat increases CHD risk due to an 
adverse effect on blood lipoproteins. The petition highlighted one 
study (Ref. 18) suggesting that a linear dose-response model was not 
appropriate for quantifying effects of lower levels of IP-TFA intake on 
LDL-C. In addition, the petition noted that the trans fat content of 
control diets used in published feeding studies ranged from non-
detectable to 2.4%en and suggested, by example, that the non-detectable 
level of TFA in a test diet could be at 0.15%en, which is three times 
higher than IP-TFA from petitioned uses of PHOs. Moreover, the petition 
noted that overall the IP-TFA intake from petitioned uses of PHOs 
(0.05%en) is well below the intake level of diets tested in the 
controlled feeding trials that were relied upon in the meta-analyses to 
assess the effect of IP-TFA on CHD risk. Because the impact of low 
level IP-TFA intakes cannot be detected by scientific studies, the 
petition concluded that the IP-TFA intake from petitioned uses of PHOs 
could be considered de minimis.
FDA Assessment
    We will address the petitioner's non-detectability argument with a 
three-prong response. First, we will discuss the issue of statistical 
power and how it relates to detectable changes in clinical feeding 
trials. Next, we will review empirical evidence of adverse effects of 
lower IP-TFA intakes from several recent population studies. Lastly, we 
will comment on the body of evidence that supports a no-threshold, 
linear dose-response model to characterize the adverse health effects 
of trans fat intake.
a. Statistical Power of Controlled Feeding Trials
    Statistical power is the probability that a study will correctly 
detect an effect when an effect exists (Ref. 22). Larger sample sizes 
generally result in higher statistical power, increasing the likelihood 
that a study will be able to identify differences in the study 
subjects. We acknowledge that there are limits to the statistical power 
of controlled feeding trials to measure changes in LDL-C from low 
levels of TFA exposure. However, the lack of data from controlled 
feeding trials on the effect of TFA intake on blood lipids at lower TFA 
intake is not due to a potential threshold below which TFA intake has 
no effect on LDL-C and other blood lipids. Rather, the lack of data at 
lower TFA intake is due to the limited statistical power to detect 
significant changes in LDL-C at TFA intake below about 3 percent of 
energy in controlled feeding trials with feasible sample size of about 
100 participants. For example, we estimated that it would require more 
than 300,000 participants in hypothetical PHO feeding trials to detect 
statistically significant changes LDL-C at the IP-TFA dietary exposure 
of 0.05%en (Refs. 6 and 7).
b. Empirical Evidence From New Population Studies
    Recent population studies have shown empirical evidence of adverse 
effects of lower IP-TFA intake levels on CHD risk. Two recent 
prospective observational studies with long term follow-up found 
significant increases in CHD risk or CVD mortality at trans fat intake 
increments as low as 0.3%en to 0.6%en (Refs. 15 and 16). This is about 
1/10 of the approximately 3 percent of energy from TFA intake that can 
be studied in controlled feeding trials of lipid biomarkers, and is 
roughly tenfold higher than the 0.05%en IP-TFA exposure from petitioned 
PHO uses.
    Two recent studies independently examined the public health effects 
of restricting trans fat in eateries in several New York state counties 
between 2007 and 2011 (Refs. 23 and 24). In one study, the authors 
compared records of hospital admissions for heart attack and stroke in 
counties that had TFA restrictions and in control counties that had no 
restrictions (Ref. 23). They found that there was an additional 6.2 
percent decline in hospital admissions for heart attacks and strokes in 
the populations of counties with TFA restrictions. This reduction 
corresponds to 43 CVD events prevented annually per 100,000 persons. In 
another study, the authors analyzed the association of trans fat 
restrictions in certain New York state counties and annual CVD 
mortality rates (Ref. 24). They found a 4.5 percent decrease in CVD 
mortality in counties with trans fat restrictions compared with control 
counties. This reduction corresponds to 13 fewer CVD deaths annually 
per 100,000 persons. Both studies, using separate data sources, showed 
consistent results of a ``real-

[[Page 23390]]

world'' public health impact associated with the removal of trans fat 
in restaurant food.
    Four studies published in 2017 examined data on plasma trans fatty 
acid concentrations in U.S. adults from the NHANES of 1999-2000 and 
2009-2010 (Refs. 25-28). These studies showed the association between 
plasma TFA and serum lipid and lipoprotein (i.e., LDL-C and HDL-C) 
concentration before and after reductions in TFA consumption occurred 
in the U.S. population. On average, plasma TFA concentrations in U.S. 
adults were about 54 percent lower in 2009-2010 compared to 1999-2000 
(Refs. 26 and 27). Significant improvements in blood lipids (e.g., 
lower LDL-C and triglycerides, higher HDL-C) occurred over time as 
plasma TFA concentrations decreased (Refs. 25 and 26). Despite 
substantial reductions in TFA intake over time, plasma TFA 
concentrations were significantly and consistently associated with 
serum lipid and lipoprotein concentrations at both time periods (Ref. 
27). Results were similar for metabolic syndrome and most of its 
components such as large waistline, high fasting glucose, and high 
triglycerides (Ref. 28). The authors concluded that these studies do 
not support the existence of a threshold under which the association 
between plasma TFA concentration and lipid profiles might become 
undetectable (Refs. 27 and 28).
c. Consistent Support of a Progressive and Linear Dose-Response
    In response to the petitioner's argument of a non-linear dose-
response, we note that the vast majority of scientific studies have 
been consistent in their conclusions that trans fat consumption has a 
progressive and linear adverse effect on blood lipids and CHD risk 
(Ref. 7). FDA's 2015 review of the scientific evidence for human health 
effects of TFA concluded: (1) There is no evidence of a threshold below 
which TFA does not affect blood lipids and (2) both controlled feeding 
trials and prospective observational studies strongly support the 
conclusion that trans fat intake has a progressive and linear effect 
that increases CHD risk, with no evidence of a threshold (Ref. 2). 
Numerous expert panels discussed in our 2015 review and in the current 
review also support this conclusion. Additional evidence from newer 
studies also supports the conclusion that TFA has a progressive and 
linear adverse effect on blood lipids and CHD risk (Refs. 12 and 29). 
This is discussed in detail in our review memorandum (Ref. 7).
3. Natural Occurrence Argument
    The petitioner based its third argument on a ``natural occurrence'' 
theory which purports that a risk due to human activity may be de 
minimis and would not cause the activity to be considered unsafe 
provided that the risk does not exceed the natural occurrence of the 
same risk (Ref. 20). Specifically, the petitioner argued that the 
petitioned uses of PHOs are safe because the incremental increase in 
IP-TFA intake from petitioned PHO uses (i.e., 0.05%en) is 
infinitesimally small and negligible in comparison to existing 
background dietary TFA exposure from intrinsic sources. As described in 
section IV.B, the petitioner estimated the mean exposure to TFA from 
intrinsic sources (e.g., naturally-occurring TFA from meat and dairy 
foods) to be 0.46%en. The petition stated that the estimated intake of 
IP-TFA of 0.05%en from petitioned uses of PHOs equates to the 1.2th 
percentile of the TFA intake distribution from intrinsic sources. The 
petition explained further that this amount of IP-TFA intake is within 
the variability of the TFA intake from intrinsic sources and below the 
5th percentile. Thus, the petition concluded that the petitioned uses 
are safe because the incremental increase in IP-TFA exposure from the 
petitioned uses of PHOs is infinitesimally small and negligible in 
comparison to existing background dietary TFA exposure from intrinsic 
sources.
FDA Assessment
    For our safety assessment, we considered as a worst-case scenario 
the assumption that TFA from intrinsic sources is chemically and 
pharmacologically related to IP-TFA from PHOs. In general, TFA from 
intrinsic sources and IP-TFA contain the same trans fatty acid isomers, 
although in different proportions (Ref. 12). The most recent evidence 
from controlled feeding trials shows comparable effects on blood 
lipoproteins such as LDL-C and HDL-C by naturally-occurring TFA and IP-
TFA (Ref. 7). Results of prospective observational studies specifically 
of TFA from intrinsic sources (rather than total TFA) are relatively 
sparse, and generally do not show an association of naturally-occurring 
TFA with CHD risk, possibly due to limitations of the studies (Ref. 7). 
Regarding the effect of TFA from intrinsic sources on adverse health 
outcomes other than CHD (e.g., metabolic syndrome and diabetes), study 
results are divergent (Refs. 6 and 7). Although there are 
inconsistencies in the data overall, we considered for the purposes of 
our safety assessment that TFA from intrinsic sources is, in general, 
chemically and pharmacologically related to IP-TFA from PHOs.
    We disagree with the petitioner's assertion that the IP-TFA 
exposure from the petitioned uses of PHOs is safe because it is 
insignificant in comparison to existing background dietary TFA 
exposure. We note that the per capita IP-TFA intake of 0.05%en from 
petitioned uses of PHOs is approximately 10 percent of mean TFA intake 
from intrinsic sources; we do not consider this to be an 
infinitesimally small or negligible amount. The contribution of IP-TFA 
intake from petitioned uses of PHOs is not trivial, but rather will 
increase the mean population TFA exposure by 10 percent. Food sources 
of naturally-occurring TFA are widely consumed in the population, and 
therefore few members of the population consume 0.05%en TFA or less. As 
the petition indicated, 0.05%en from IP-TFA from petitioned uses of 
PHOs corresponds to about the 1.2th percentile of population TFA intake 
from intrinsic sources. We assert that this comparison is not 
particularly relevant to whether the per capita IP-TFA intake is 
significant because the contribution of IP-TFA exposure from the 
petitioned uses is in addition to, not substitutional for, exposure to 
TFA from intrinsic sources. Rather, the relevant comparison is that the 
per capita IP-TFA intake, 0.05%en, is approximately 10 percent of mean 
TFA intake from naturally-occurring sources. For these reasons, we 
disagree with the petitioner's argument that the petitioned uses of 
PHOs are safe because they are negligible in comparison to existing 
background dietary TFA exposure from intrinsic sources.
    As stated earlier, there is no explicit reference to de minimis 
risks under either the FD&C Act or FDA's regulations regarding the 
evaluation of the safety of food additives in response to a food 
additive petition. Based on the data submitted by the petitioner, FDA 
has determined that the petitioned uses present more than a de minimis 
or negligible risk. Therefore, FDA has not found it necessary as part 
of its petition response to determine how the concept of de minimis 
risk may apply to the safety analysis under section 409 of the FD&C 
Act.

V. Comments on the Filing Notification

    We received 10 comments in response to the petition's filing 
notification. Seven comments expressed opposition to the petition, one 
comment was about

[[Page 23391]]

labeling of PHOs, one comment did not pertain to the petition, and one 
comment was a duplicate submission. All of the comments opposing the 
petition cited the adverse health effects associated with the 
consumption of TFA. None of the comments provided information to 
support the petitioner's conclusion that the proposed uses of PHOs are 
safe.

VI. Conclusion

    FAP 5A4811 requested that the food additive regulations be amended 
to provide for the safe use of PHOs as a solvent or carrier for 
flavoring agents, flavor enhancers, and coloring agents; as a 
processing aid; and as a pan release agent for baked goods at specific 
use levels. After reviewing the petition, as well as additional data 
and information relevant to the petitioner's request, we determined 
that the petition does not contain convincing evidence to support the 
conclusion that the proposed uses of PHOs are safe. Therefore, FDA is 
denying FAP 5A4811 in accordance with 21 CFR 171.100(a).

VII. Compliance Date

    As discussed in section II, the declaratory order concluded that 
PHOs are no longer GRAS for any use in human food and established a 
compliance date of June 18, 2018 (80 FR 34650). In light of our denial 
of FAP 5A4811, we acknowledge that the food industry needs additional 
time to identify suitable replacement substances for the petitioned 
uses of PHOs and that the food industry has indicated that 12 months 
could be a reasonable timeframe for reformulation activities (Ref. 30). 
Therefore, elsewhere in this issue of the Federal Register, we have 
extended the compliance date to June 18, 2019, for the manufacturing of 
food with the petitioned uses of PHOs. Food manufactured with the 
petitioned uses after June 18, 2019 may be subject to enforcement 
action by FDA.
    In addition, for food manufactured with the petitioned uses before 
June 18, 2019, elsewhere in this issue of the Federal Register, we are 
extending the compliance date to January 1, 2021. This time frame will 
allow manufacturers, distributors, and retailers to exhaust product 
inventory of foods made with the petitioned uses before the 
manufacturing compliance date. All foods containing unauthorized uses 
of PHOs after January 1, 2021 may be subject to FDA enforcement action.

VIII. Objections

    Any persons that may be adversely affected by this document may 
file with the Dockets Management Staff (see ADDRESSES) either 
electronic or written objections. You must separately number each 
objection, and within each numbered objection you must specify with 
particularity the provision(s) to which you object, and the grounds for 
your objection. Within each numbered objection, you must specifically 
state whether you are requesting a hearing on the particular provision 
that you specify in that numbered objection. If you do not request a 
hearing for any particular objection, you waive the right to a hearing 
on that objection. If you request a hearing, your objection must 
include a detailed description and analysis of the specific factual 
information you intend to present in support of the objection in the 
event that a hearing is held. If you do not include such a description 
and analysis for any particular objection, you waive the right to a 
hearing on the objection.
    It is only necessary to send one set of documents. Identify 
documents with the docket number found in brackets in the heading of 
this document. Any objections received in response to the regulation 
may be seen in the Dockets Management Staff between 9 a.m. and 4 p.m., 
Monday through Friday, and will be posted to the docket at http://www.regulations.gov. We will publish notice of the objections that we 
have received or lack thereof in the Federal Register.

IX. References

    The following references are on display in the Dockets Management 
Staff (see ADDRESSES) and are available for viewing by interested 
persons between 9 a.m. and 4 p.m., Monday through Friday; they are also 
available electronically at http://www.regulations.gov. FDA has 
verified the website addresses, as of the date this document publishes 
in the Federal Register, but websites are subject to change over time.

1. Sacks, F.M., A.H. Lichtenstein, J.H.Y. Wu, et al. ``Dietary Fats 
and Cardiovascular Disease: A Presidential Advisory from the 
American Heart Association.'' Circulation 136(3): e1-e23, 2017.
2. FDA Memorandum from J. Park to M. Honigfort, Scientific Update on 
Experimental and Observational Studies of Trans Fat Intake and 
Coronary Heart Disease Risk, June 11, 2015.
3. FDA Memorandum from J. Park to M. Honigfort, Literature Review, 
June 11, 2015.
4. FDA Memorandum from J. Park to M. Honigfort, Quantitative 
Estimate of Industrial Trans Fat Intake and Coronary Heart Disease 
Risk, June 11, 2015.
5. FDA Memorandum from D. Doell to E. Anderson, April 13, 2018.
6. FDA Memorandum from J. Park to E. Anderson, Quantitative Coronary 
Heart and Cardiovascular Disease Risk Assessments of Exposure from 
Industrially-Produced Trans Fatty Acid (IP-TFA) from Proposed Uses 
of Partially Hydrogenated Vegetable Oils (PHO) in Select Foods, 
April 16, 2018.
7. FDA Memorandum from J. Park to E. Anderson, Scientific Literature 
Review Update on Trans Fats with Detailed Responses to the 
Petitioner's Safety Conclusions on the Petitioned Uses of Partially 
Hydrogenated Oils (PHOs), April 16, 2018.
8. IOM/NAS, ``Dietary Reference Intakes for Energy Carbohydrate, 
Fat, Fatty Acids, Cholesterol, and Amino Acids (Macronutrients),'' 
National Academies Press, Washington, DC, 2002/2005, Available at: 
https://www.nap.edu.
9. Katan, M.B., P.L. Zock, and R.P. Mensink, ``Trans Fatty Acids and 
Their Effects on Lipoproteins in Humans,'' Annual Review of 
Nutrition, 15:473-93, 1995.
10. Zock, P.L., M.B. Katan, and R.P. Mensink, ``Dietary Trans Fatty 
Acids and Lipoprotein Cholesterol,'' American Journal of Clinical 
Nutrition, 61(3):617, 1995.
11. Zock, P.L. and R.P. Mensink, ``Dietary Trans-Fatty Acids and 
Serum Lipoproteins in Humans,'' Current Opinion in Lipidology, 
7(1):34-7, 1996.
12. Brouwer, I.A., ``Effect of Trans-Fatty Acid Intake on Blood 
Lipids and Lipoproteins: A Systematic Review and Meta-Regression 
Analysis,'' Geneva: World Health Organization, 2016.
13. Hafekost, K., T.A. O'Sullivan, D. Lawrence, and F. Mitrou, 
``Systematic Review of the Evidence for a Relationship Between 
Trans-Fatty Acids and Blood Cholesterol,'' Canberra, Australia: On 
behalf of Food Standards Australia New Zealand, 2014, available at: 
http://www.foodstandards.gov.au/publications/Pages/Systematic-Review-of-the-evidence-for-a-relationship-between-trans-fatty-acids-and-blood-cholesterol-.aspx.
14. de Souza, R.J., A. Mente, A. Maroleanu, et al., ``Intake of 
Saturated and Trans Unsaturated Fatty Acids and Risk of All Cause 
Mortality, Cardiovascular Disease, and Type 2 Diabetes: Systematic 
Review and Meta-Analysis of Observational Studies,'' BMJ, 351:h3978, 
2015.
15. Li, Y., Hruby, A., A.M. Bernstein, et al., ``Saturated Fats 
Compared with Unsaturated Fats and Sources of Carbohydrates in 
Relation to Risk of Coronary Heart Disease: A Prospective Cohort 
Study,'' Journal of the American College of Cardiology, 66(14):1538-
48, 2015.
16. Wang, D.D., Y. Li, S.E. Chiuve, et al., ``Association of 
Specific Dietary Fats with Total and Cause-Specific Mortality,'' 
JAMA Internal Medicine, 176(8):1134-45, 2016.
17. Reichard, J.F. and L.T. Haber, ``Mode-of-Action Evaluation for 
the Effect of Trans Fatty Acids on Low-Density Lipoprotein 
Cholesterol,'' Food and Chemical Toxicology, 98(Pt B):282-94, 2016.
18. Allen, B.C., M.J. Vincent, D. Liska, and L.T. Haber, ``Meta-
Regression Analysis of

[[Page 23392]]

the Effect of Trans Fatty Acids on Low-Density Lipoprotein 
Cholesterol,'' Food and Chemical Toxicology, 98(Pt B):295-307, 2016.
19. Liska, D.J., C.M. Cook, D.D. Wang, P.C. Gaine, and D.J. Baer, 
``Trans Fatty Acids and Cholesterol Levels: An Evidence Map of the 
Available Science.'' Food and Chemical Toxicology, 98(Pt B):269-81, 
2016.
20. Peterson, M., ``What is a de minimis Risk?'' Risk Management: An 
International Journal, 4(2):47-55, 2002.
21. Castorina, R. and T.J. Woodruff, ``Assessment of Potential Risk 
Levels Associated with U.S. Environmental Protection Agency 
Reference Values,'' Environmental Health Perspectives, 111(10):1318-
25, 2003.
22. Rosner, B., Fundamentals of Biostatistics, Duxbury Press, 
Belmont, CA, 2010.
23. Brandt, E.J., R. Myerson, M.C. Perraillon, and T.S. Polonsky, 
``Hospital Admissions for Myocardial Infarction and Stroke Before 
and After the Trans-Fatty Acid Restrictions in New York,'' JAMA 
Cardiology, 2(6):627-634, 2017.
24. Restrepo, B.J. and M. Rieger, ``Trans Fat and Cardiovascular 
Disease Mortality: Evidence from Bans in Restaurants in New York,'' 
Journal of Health Economics, 45:176-96, 2016.
25. Restrepo, B.J., ``Further Decline of Trans Fatty Acids Levels 
Among US Adults Between 1999-2000 and 2009-2010,'' American Journal 
of Public Health, 107(1):156-8, 2017.
26. Vesper, H.W., S.P. Caudill, H.C. Kuiper, et al., ``Plasma Trans-
Fatty Acid Concentrations in Fasting Adults Declined from NHANES 
1999-2000 to 2009-2010,'' American Journal of Clinical Nutrition, 
105(5):1063-9, 2017.
27. Yang, Q., Z. Zhang, F. Loustalot, et al., ``Plasma Trans-Fatty 
Acid Concentrations Continue to be Associated with Serum Lipid and 
Lipoprotein Concentrations Among US adults After Reductions in 
Trans-Fatty Acid Intake,'' Journal of Nutrition, 147(5):896-907, 
2017.
28. Zhang, Z., C. Gillespie, Q. Yang, ``Plasma Trans-Fatty Acid 
Concentrations Continue to be Associated with Metabolic Syndrome 
Among US Adults After Reductions in Trans-Fatty Acid Intake,'' 
Nutrition Research, 43:51-9, 2017.
29. Mensink, R.P., ``Effects of Saturated Fatty Acids on Serum 
Lipids and Lipoproteins: A Systematic Review and Regression 
Analysis,'' Geneva: World Health Organization, 2016.
30. Letter from the American Bakers Association, et al., to Dr. 
Scott Gottlieb, Commissioner, Food and Drug Administration (April 
30, 2018) (sent by electronic mail).

    Dated: May 15, 2018.
Leslie Kux,
Associate Commissioner for Policy.
[FR Doc. 2018-10715 Filed 5-18-18; 8:45 am]
 BILLING CODE 4164-01-P