Joint Committee on Dietary Supplements

Joint Committee on Dietary Supplements

Meeting Summary

May 15, 2012

This document is part of the NSFInternationalStandards process and is for NSF Committee uses only. It shall not be reproduced, or circulated, or quoted, in whole or in part, outside of NSF activities, except with the approval of NSF.

IADMINISTRATIVE

Joan Hoffman welcomed Joint Committee members and observers.Kevin Cox mentioned when he joined the teleconference that he would provide the update for the Metals Task Groupin place ofCliff McLellan.Joan recorded the attendance, read the NSF Antitrust Statement, and turned the meeting over to Staci Eisner, Vice Chair.

Review of Draft Meeting Summary – October 13, 2011

Staci asked for comments and/or changes to the October 13, 2011draft Meeting Summary.MichaelMcGuffin moved to approve the Meeting Summary as written;Heather Arnold seconded. All were in favor.

IINSF/ANSI 173-2011 REVISIONS

Joan reminded participants that Standard 173 is a living document and that suggestions for improvement are always welcome. She gave a brief summary of the revisions included in the 2011 update, which included some general editorial corrections as well as the following:

Issue 28/29 – Methods/QC Update

Updated: (1) section 6, Test methods used by testing laboratories for identification and quantification of ingredients – raw materials and finished products; (2) section 7.4, Test methods for chemical contaminants; (3) the quality assurance sections related to verification testing performed to evaluate compliance with the Standard; and (4) Tables 3 and 4 have been removed.

Issue 30 – Aristolochic Acid

Updated: (1) section 5.3.4, Natural toxins; (2)section7.4, Test methods for chemical contaminants; (3) Table A1, Botanicals known or suspected to contain Aristolochic Acid.

Issue 36 – Lead

Updated the lead limit for finished products from 20 to 10 mg/d.

Issue 37 – Cadmium

Updated the cadmium limit for finished products from 6 to 4.1 mg/d.

Issue 38 – Normative References and Definition

Updated section 2 and added a definition for “qualified individual” in section 3.

Issue 39 – Macroscopic Test Methods

Updated section 6.1.1.1.

Issue 40 – 7.4 Test methods for chemical contaminants

Eliminated discrepancies with section 7.4 language; achieved consistency with language in paragraphs related to method selection and development; ensured that all language meets ANSI requirements.

Issue 41 – Mercury

Corrected the mercury limit for finished products to 0.002 mg/d.

Annex B – Reference information for contaminant level acceptance criteria wasupdated.

IIIREVIEW OF RECENT BALLOTS

Issue 43 – Scope, Labeling, Production Requirements

Updates 1.2 Scope, 4 Labeling and literature requirements, 5 Product requirements, and 5.1.2 Finished product in NSF/ANSI 173, allowing NSF International increased flexibility in selecting finished product claims for analysis based on the number of finished product claims and ingredients present on the product label.

Issue 44 – Finished Products

Updates 5.2.2 Finished products in NSF/ANSI 173.

Issue 45 - Pesticides

Proposes updates to5.3.2 Pesticides and 7.2.1 Multi-residue method. Also proposed is a modification to the requirements specific to Panax ginseng … read moreor Panax quinquefolius to only apply the zero tolerance limits to products sold and/or distributed within the United States.

Joan mentioned that both Issue 43 and 44 have been approved by the Joint Committee, the Technical Committee, and the Council of Public Health Consultants. Having completed the NSF standards development process, they are in the queue to be published in NSF/ANSI 173-2012 in the near future. She also mentioned one Negative vote and comment had been submitted on the 173i45r1 ballot (Issue 45), and wondered if the issue of pesticides could be discussed and further clarified at the meeting. M. McGuffin indicatedthat he, as the negative voter, was fine with that suggestion. Michael asked if participants could be reminded of what sections of Standard 173 would be affected by the Issue 45 proposal.

Kerri explained that the ballot proposed revisions to 5.3.2 in section 5 which deals with product requirements, as well as revisions to 7.2.1 in section 7 which is the description of the methods used to verify compliance if done through testing.

As background, Kerri clarified that prior to balloting, the Task Group on Pesticides had met to review and discuss proposedlanguage revisions to these sections of the Standard. She indicated that one key part of the current language which has been in the Standard for a very long time is the requirement that compliance with FDA and EPA regulated limits needs to be verified. The Task Group had discussed this as being an impossible requirement since very few limits are established for herbs and that it starts to become a zero tolerance policy. Also discussed by the Task Group was the task of determining exactly what pesticides are to be screened for in a given product.

When the language in the Standard was originally adopted, Kerri said it was her understanding that the term “banned pesticides” was included as a way to focus on pesticides known globally to be unsafeand which should not be in dietary supplement products. During the TG meeting, Staci had suggested several websites to use for lists of banned pesticides, and Kerri pointed out that the current NSF pesticide screen doesn’t exactly match anypublished list.

There was general agreement among participants that one solution would be to add a note and/or footnote referencing the Rotterdam Convention in section 5.3.2 for clarification of “banned pesticides” since the Rotterdam list is an international convention with a broad basis of acceptance. Michael endorsed this solutionand said it would address his first concern completely. He added that the EPA does not post a list of banned pesticides; however, their website links to the Rotterdam list.

Kerri brought up an additional potential concern about pesticidesthat may not be on the list, but depending on their detection levels, could be considered unsafe. In terms of NSF trying to be protective of public health, if a pesticide is not on the list but seems to exceed safety, NSF could prevent the product from being approved. Staci did not recall a Task Group discussion of this concern. Although she agreed there needs to be a mechanism for dealing with it, she advised having the TG address it as a new issue, separatefrom the current ballot proposal.

Regardingdetection of pesticides,Kevin explained that NSF may identify a substance in a product that is not on the list of “banned pesticides” and may not be at an unsafe level but still present. Although NSF would perform a risk assessment, if there is no regulatory limit, the client may not know what to do with that information.Michael supported Staci’s decision and also indicated Kevin had raised an excellent issue. He conceptually supported both Kerri and Kevin’s explanations but thought it best for the Task Group on Pesticides to address this issue first.

Michael’s second concern relates to the proposed change to the first paragraph of section 5.3.2:

Unless a manufacturer has controls in place to screen for pesticides or use certified organic ingredients as demonstrated in the GMP audit, aA broad pesticide screen shall be performed to confirm compliance with FDA and EPAregulated limits and the absence of banned pesticides in botanical products.

According to Michael, the current language in the Standard (“Unless a manufacturer has controls in place…”)provides the option that a manufacturer can have controls in place to screen for pesticides. Removing this language seems to imply that the manufacturer would need to test every single lot ("A broad pesticide screen shall be performed..."). Michael wondered if this was the intention.

Kerri said that NSF/ANSI 173 is often hard to interpret and that section 5 describes the requirements to comply with the Standard, while section 8 describes what manufacturers need to do. Further, the language in section 5.3.2 specifies a product should be free of banned pesticides, and section 7 describes how the certification organization would verify this. The change in language in section 5 does not dictate action to be taken by the manufacturer; and as required in 21CFR111 GMPs, it remains up to the manufacturer to decide how to manage theircontaminant risk including pesticides.

Staci indicated it was her understanding the proposed changes donot impose any specific testing burden on the manufacturer. Rather, she said, the manufacturer is only required to have appropriate controls. The ballot proposal relates to the testing NSF does, and that NSF will need to pull a sample and test it according to schedule to make sure that it meets the requirements. The proposal does not require manufacturers to test every lot. Kerri confirmed this andexplained that an auditor’s checklist points are based on section 8 (GMP) only. Michael shared that in the attempt to resolve his ballot comment, this had been explained to him. And although a JC member for many, many years, Michael acknowledged he never before understood that companies who want to comply with the Standard need only comply with section 8 and are not required to do everything in sections 5, 6, and 7. Staci emphasized that as someone who has been NSF certified in the manufacturing facility, she was aware of the differences in the requirements of different sections in the Standard, and that she did understand this. She also recalled that although it may not be clearly spelled out in the Standard that perhaps it does need to be addressed; however, this distinction between the partswas discussed in the past at both the Joint Committee and Task Group levels. Michael confirmed his acceptance, for now, that although it may not be clear in the Standard, it becomes clear when a manufacturer seeks NSF certification to the Standard.

S. Dentali thought it unfortunate that NSF is in the position of testing something that is not likely or certain to contaminate the product, but Kerri indicated it is NSF’s charge to prove that products are safe. In fact, in a way, if more products are tested,she explained that batching could be used and that NSFwould be able to test more cost effectively.

Discussion circled back to section 5.3.2, and Michael indicated he was very close to being willing to withdraw his concern. When Staci added that it is very clear to a manufacturer coming into the program that NSF will pull a sample annually, not test every lot, Michael said his concerns had been addressed.

NSF will determine the need for reballoting Issue 45 and also look into including an explanation of the differences in parties responsible for executing the testing requirements for the different sections in NSF/ANSI 173. Additionally, JC member and observers who would like to join the Pesticides Task Group were encouraged to let Joan know.

IVTASK GROUP UPDATES

Metals Task Group

In the absence of Clif, Kevin pointed out that as previously mentionedthe Task Group on Heavy Metalsput forth three ballot proposals establishing new limits for Lead, Cadmium, and Mercury in finished products, and these were approved and published in NSF/ANSI 173-2011. Kevin addedthere will be no proposal to change the current limit for Chromium in the Standard, and that Arsenic is still under discussion.

Ingredients are also under discussion. Kevin explained the direction of the proposal being discussed establishes a new methodology for metal limits in raw materials. Although still in preliminary discussion, some progress has been made regarding setting maximum use levels for ingredients. The Task Group hopes to complete its charge in the near future.

Kerri asked if the Task Group would be looking at consistency between raw material and finished product levels, and Kevin confirmed this.

G. Cumberford wondered about the nature of the arsenic debate within the Task Group. According to Kevin, the debate relates to NSF looking at a limit for inorganic arsenic in finished products that is significantly lower than other groups like AHPA and USP. He added that Greg and any others are welcome to join the Heavy Metals Task Group.

Michael pointed out that although “arsenic” is listed in 5.3.1.1 Raw materials, “inorganic arsenic” is listed in 5.3.1.2 Finished products. He would like to see some precision there, and Kevin said this would be addressed.

Task Group on Oxidation and Rancidity

H. Rice first explained that the Task Group on Oxidation and Rancidity was formed in 2011 and charged with making a recommendation about requirements to ensure that oils were not rancid. Harry said the recommendation was to rely more directly on data from manufacturers; however, this had resulted in a conflict with certification requirements. Since there were no specifics on what should be audited, NSF auditors objected to it.

Kerri explained the two drivers for creating this Task Group: 1) it is not always possible to test rancidity in the finished product, and 2) the proposed recommendation for putting more responsibility on the manufacturer and using manufacturer test data did not meet the ANSI audit requirement that for making certification decisions test data has to be developed under ISO 17025. The proposal for reviewing manufacturing data was not viable since manufacturer data may not have been developed under ISO 17025.

In the revised proposal, Kerri said,the recommendation was to movethis to an audit requirement, since it is the responsibility of the manufacturer to verify that an ingredient is not rancid before use and that this should be done on every lot. As a responsibility of the manufacturer, it seemed to make sense to have this not be an NSF testing requirement but to verify it through the audit and therefore have a subsection added to section 8. However, NSF auditors thought the languageproposed by the Rancidity Task Group was too vague and that it needed to specify exactly what manufacturers have to demonstrate. Also, it was suggested having a table of oils to be tested and the relevant action(s) to be taken.

Currently at the Task Group level, there is an action item for NSF to create a list of oils using the Food Chemical Codex (FCC) as a guide. NSF will also work on revising language for the Standard to be reviewed by auditors, and possibly including a table of customized limits based on the oil.

Steven wondered why a listing of oils would be needed and thought an auditor should be able to figure out the difference between fixed and volatile oilsbased on their definitions. Staci advised that based on TG discussion, auditors are not chemists, and if having a list of oils resolves the issue, it is worth doing.

K. Sharpless agreed with Steven that defining fixed and volatile oils makes more sense than an ever increasing list. Relative to rancidity, Kerri asked if there is concern about an ever increasing list or if listing the fixed oils would be enough. She stressed the purpose of differentiating the oils and the goal for defining a list that will work with the test methodology. Further, she pointed out that volatile plant compounds make the method(s) invalid.

Staci informed participants the Task Group has been wrestling with this for quite a while and have tried many different ideas that turned out not to work for one reason or another. Most recently, a decision was made to start with what works today and to then determine the possibility of expanding beyond that in the future. Starting with a list of common fixed oils now used in commerce for which we know the test methods work, and which we know will be understood, should address 90% of the problem. Then, efforts can be focused on addressing the other 10%. Steven suspected the 10% are plants where there may be volatile oils mixed in with the fixed oils that would invalidate the test, and Kerri confirmed this. J. Betz asserted there are fixed oils and volatile oils and very few cases where there are mixes of the two. He suggested checking the American Oil Chemists’ Society for a list of fixed oils. Kerri stated the plan is to create a list of fixed oils. As far as volatile oils, they will be allowed to not meet the rancidity requirement because the tests do not work, and they do not go rancid. In the end, Staci explained that this issue will be addressed by placing a specific burden on manufacturers which will be verified through audit.

Task Group on Pesticides

Although previously covered in the discussion of Issue 45 – Pesticide testing (above), Kerri restated the action items: 1) NSF will create a modified list of pesticides based on the Rotterdam Convention;2) the inclusion of pesticides beyond that list will be discussed by the Task Group in the future, and also 3) to ballotrevised language for the Standard as necessary.

VNEW BUSINESS

DS-2012-2 – Upper-Lower Acceptance Limits

Kerri pointed that both she and Kevin are involved with this issue, which resulted from an ANSI auditwhich concerns the handling of results that may be significantly above claim – i.e., 200 - 300%. Currently, NSF does verify to make sure product limits are not near or above Institute of Medicine (IOM)upper safety limits. According to Kevin, however, part of the issue relates to the current language in section 5.2.2, and specifically paragraphs two and three: