[Federal Register Volume 60, Number 139 (Thursday, July 20, 1995)]
[Proposed Rules]
[Pages 37507-37549]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 95-17505]




  Federal Register / Vol. 60, No. 139 / Thursday, July 20, 1995 / 
Proposed Rules  
=======================================================================
-----------------------------------------------------------------------

[[Page 37507]]


DEPARTMENT OF HEALTH AND HUMAN SERVICES

Food and Drug Administration

21 CFR PART 101

[Docket No. 95P-0003]


Food Labeling: Health Claims; Sugar Alcohols and Dental Caries

AGENCY: Food and Drug Administration, HHS.

ACTION: Proposed rule.

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

SUMMARY: The Food and Drug Administration (FDA) is proposing to 
authorize the use, on food labels and in food labeling, of health 
claims on the association between sugar alcohols and the nonpromotion 
of dental caries. In addition, FDA is proposing to exempt sugar 
alcohol-containing foods from certain provisions of the health claims 
general requirements regulation. FDA is proposing these actions in 
response to a petition filed by the National Association of Chewing Gum 
Manufacturers, Inc., and an ad hoc working group of sugar alcohol 
manufacturers (hereinafter referred to as the petitioners).

DATES: Written comments by October 3, 1995. The agency is proposing 
that any final rule that may issue based upon this proposal become 
effective 30 days following its publication.

ADDRESSES: Written comments to the Dockets Management Branch (HFA-305), 
Food and Drug Administration, rm. 1-23, 12420 Parklawn Dr., Rockville, 
MD 20857.

FOR FURTHER INFORMATION CONTACT: Joyce J. Saltsman, Center for Food 
Safety and Applied Nutrition (HFS-165), Food and Drug Administration, 
200 C St. SW., Washington, DC 20204, 202-205-5916.

SUPPLEMENTARY INFORMATION:

I. Background

A. The Nutrition Labeling and Education Act of 1990

    On November 8, 1990, the President signed into law the Nutrition 
Labeling and Education Act of 1990 (the 1990 amendments) (Pub. L. 101-
535). This new law amended the Federal Food, Drug, and Cosmetic Act 
(the act) in a number of important ways. One of the most notable 
aspects of the 1990 amendments was that they confirmed FDA's authority 
to regulate health claims on food labels and in food labeling. As 
amended by the 1990 amendments, section 403(r)(1)(B) of the act (21 
U.S.C. 343(r)(1)(B)) provides that a product is misbranded if it bears 
a claim that characterizes the relationship of a nutrient to a disease 
or health-related condition, unless the claim is made in accordance 
with the procedures and standards contained in regulations adopted by 
FDA.
    Under section 403(r)(3)(B)(i) of the act, the Secretary of Health 
and Human Services (and, by delegation, FDA) shall promulgate 
regulations authorizing such claims only if he or she determines, based 
on the totality of publicly available scientific evidence (including 
evidence from well-designed studies conducted in a manner which is 
consistent with generally recognized scientific procedures and 
principles), that there is significant scientific agreement, among 
experts qualified by scientific training and experience to evaluate 
such claims, that the claim is supported by such evidence.
    Section 403(r)(3)(B)(ii) and (r)(3)(B)(iii) of the act describes 
the information that must be included in any claim authorized under the 
act. The act provides that the claim shall be an accurate 
representation of the significance of the substance in affecting the 
disease or health-related condition, and that it shall enable the 
public to comprehend the information and understand its significance in 
the context of the total daily diet. Finally, section 403(r)(4)(A)(i) 
of the act provides that any person may petition FDA to issue a 
regulation authorizing a health claim.
    The 1990 amendments, in addition to amending the act, directed FDA 
to consider 10 substance-disease relationships as possible subjects of 
health claims.

B. FDA's Response

    In the Federal Register of January 6, 1993 (58 FR 2478), FDA 
adopted a final rule that implemented the health claim provisions of 
the act. In that final rule, FDA adopted Sec. 101.14 (21 CFR 101.14). 
The regulation sets out the circumstances in which a substance is 
eligible to be the subject of a health claim (Sec. 101.14(b)), adopts 
the standard in section 403(r)(3)(B)(i) of the act as the standard that 
the agency will apply in deciding whether to authorize a claim about a 
substance-disease relationship (Sec. 101.14(c)), sets forth general 
rules on how authorized claims are to be made in food labeling 
(Sec. 101.14(d)), and establishes limitations on the circumstances in 
which claims can be made (Sec. 101.14(e)). The agency also adopted 
Sec. 101.70 (21 CFR 101.70), which establishes a process for 
petitioning the agency to authorize health claims about a substance-
disease relationship (Sec. 101.70(a)) and sets out the types of 
information that any such petition must include (Sec. 101.70(d)). These 
regulations became effective on May 8, 1993.
    In addition, FDA conducted an extensive review of the evidence on 
the 10 substance-disease relationships listed in the 1990 amendments. 
FDA has authorized claims that relate to 8 of these 10 relationships.
    The present rulemaking on sugar alcohols and dental caries 
represents the first rulemaking that FDA has conducted in response to a 
health claim petition.

C. History of Sugar Alcohol Labeling

    In a set of findings of fact and a tentative order on label 
statements for special dietary foods that the agency issued on July 19, 
1977 (42 FR 37166), FDA addressed the issue of the use of the terms 
``sugar free,'' ``sugarless,'' and ``no sugar.'' The agency stated that 
consumers may associate the absence of sugar in a product with weight 
control and with foods that are low calorie or that have been altered 
to reduce calories significantly. At that time, FDA viewed foods 
intended to be useful in maintaining or reducing calorie intake or body 
weight as foods for special dietary use, that is, as foods intended for 
supplying particular dietary needs that exist by reason of a physical, 
physiological, pathological, or other condition.
    Evidence had been introduced at a public hearing in the 1977 
rulemaking to show that the ``sugarless'' claim is useful to identify 
foods like chewing gum, which is in sustained contact with the teeth, 
in which the use of a sweetener other than a fermentable or cariogenic 
carbohydrate may not promote tooth decay. The secretary of the American 
Dental Association's Council on Dental Therapeutics supported the 
importance of advertising and labeling sugarless chewing gum and mints 
as noncariogenic, in the sense that they did not contribute to the 
development of dental caries (Ref. 80).
    In the final rule on label statements for special dietary foods 
published in the Federal Register of September 22, 1978 (43 FR 43248), 
FDA required a statement that a food is not low calorie or calorie 
reduced (unless it is in fact, a low or reduced calorie food) when a 
``sugar free,'' ``sugarless,'' or ``no sugar'' claim is made for the 
food. The agency decided to allow ``useful only in not promoting tooth 
decay'' as an alternative statement to accompany such claims. The 
agency stated that the statements that the food is not low calorie or 
not useful for weight control, as well as ``useful only in not 
promoting tooth decay,'' were needed because the 

[[Page 37508]]
term ``sugar free'' meant only that the food was sucrose free. A 
``sugar free'' food could contain other fermentable carbohydrates. 
Thus, the information about the effect of sugar alcohol-containing 
foods on the risk of developing dental caries was originally placed on 
the food label primarily to clarify that the product was not 
necessarily useful in weight control, not to highlight the effect of 
sugar alcohol on dental caries production.
    In the Federal Register of November 27, 1991 (56 FR 60421), in 
response to the 1990 amendments, FDA published a proposed rule entitled 
``Food Labeling: Nutrient Content Claims, General Principles, 
Petitions, Definition of Terms'' (the nutrition labeling general 
principles proposal). In that document, FDA recognized that 
developments in nutrition science had established that the focus of 
nutrient content claims for providing dietary guidance had shifted from 
special populations with particular conditions to the general 
population (see 56 FR 60421). Therefore, in the nutrition labeling 
general principles proposal, FDA proposed to treat several claims that 
had been subject to regulation in Sec. 105.66 (21 CFR 105.66) as 
special dietary use claims as nutrient content claims for the general 
population. To eliminate redundancy in the regulations and to conform 
Sec. 105.66 to the 1990 amendments, FDA proposed to define these claims 
in part 101 (21 CFR part 101) and to remove them from part 105 (21 CFR 
part 105). Specifically, FDA proposed to adopt definitions for terms 
such as ``low calorie'' and ``reduced calorie,'' for other comparative 
calorie claims, and for sugar claims under section 403(r)(2) of the act 
and to codify them in Sec. 101.60. It also proposed to delete these 
claims from Sec. 105.66.
    In the Federal Register of January 6, 1993 (58 FR 2302), FDA 
published its final rules on nutrient content claims. FDA adopted 
definitions for claims for the calorie content of foods in Sec. 101.60 
(58 FR 2302 at 2415). FDA defined claims regarding the sugars content 
of a food, e.g., ``sugar free,'' ``free of sugar,'' ``no sugar,'' in 
Sec. 101.60(c). In addition, FDA published a final rule that deleted 
these claims from Sec. 105.66 (58 FR 2427).
    However, based on its consideration of comments on the use of the 
statement ``useful only in not promoting tooth decay'' to qualify the 
``sugarless'' claim, FDA concluded that the statement was actually an 
unauthorized health claim (58 FR 2302 at 2326). The claim is a health 
claim because it characterizes the relationship of a substance (sugar 
alcohols) to a disease (dental caries).
    In the nutrient content claim general principles proposal (56 FR 
60421 at 60437), the agency stated that it intended to reevaluate the 
usefulness of chewing gums sweetened with sugar alcohols in not 
promoting tooth decay. The agency stated that the data supporting the 
claim were over 20 years old and requested that new data be submitted 
in accordance with the final rule on health messages. In the nutrient 
content claim final rule, FDA stated that it had received data on the 
validity of a claim about this nutrient-disease relationship, and that 
it would make a determination on whether to authorize a claim in 
accordance with the final rule on health claims (58 FR 2302 at 2326).
    On February 5, 1993, under the procedure established in section 
701(e) of the act (21 U.S.C. 371(e)), a group of sugar alcohol 
manufacturers submitted an objection to the revocation of 
Sec. 105.66(f) (Ref. 2) and asked for a hearing on their objection. At 
the same time, the group petitioned for reconsideration of the agency's 
decision and for a stay of any administrative action by FDA pursuant to 
the determination announced in the preamble of the nutrient content 
claims rules that ``useful only in not promoting tooth decay'' is an 
unauthorized health claim.
    Filing objections to the revocation of Sec. 105.66(f) stayed the 
effect of the final rule as a matter of law. FDA's response to these 
objections and to the petitions is set forth elsewhere in this issue of 
the Federal Register.
    In the Federal Register of August 18, 1993 (58 FR 44036), FDA 
published technical amendments to the health claim regulations in 
response to comments that the agency received on the implementation 
final rule that was published with the other final rules that responded 
to the 1990 amendments in January of 1993 (see 58 FR 2066, August 18, 
1993). One of the comments stated that if a petition were submitted for 
the claim ``Useful Only in Not Promoting Tooth Decay,'' virtually none 
of the sugar-free products on the market would be eligible to bear the 
claim based on the requirements of a subsection of health claims 
general principles regulation, Sec. 101.14(e)(6). FDA acknowledged that 
certain food products of limited nutritional value that have been 
specially formulated relative to a specific disease condition, such as 
dental caries, may be determined to be appropriate foods to bear a 
health claim (58 FR at 44036). The agency commented that it was its 
intention to deal with such situations within the regulations 
authorizing specific health claims. Therefore, FDA amended 
Sec. 101.14(e)(6) to state that:

    Except for dietary supplements or where provided for in other 
regulations in part 101, subpart E, the food contains 10 percent or 
more of the Reference Daily Intake or the Daily Reference Value for 
vitamin A, vitamin C, iron, calcium, protein, or fiber per reference 
amount customarily consumed prior to any nutrient addition.

II. Petition for the Noncariogenicity of Sugarless Food Products 
Sweetened With Sugar Alcohol

A. Background

    On August 31, 1994, the petitioners submitted a health claim 
petition to FDA requesting that the agency authorize a health claim on 
the relationship of sugar alcohols (i.e., xylitol, sorbitol, mannitol, 
maltitol, lactitol, isomalt, hydrogenated starch hydrolysates, and 
hydrogenated glucose syrups) in sugarless foods to dental caries (Ref. 
1). On September 15, 1994, FDA sent the petitioners a letter stating 
that study reports that are needed to support the petition, and that 
are required for a health claim petition under Sec. 101.70, were not 
included in the petitioners' submission. The agency stated that no 
further action would be taken until that information was received (Ref. 
3).
    On September 27, 1994, the petitioners filed an amendment to their 
petition submitting the required information. On October 7, 1994, the 
agency sent the petitioners a letter acknowledging receipt of the 
additional information and stating that the agency had begun its 
scientific review of the petition (Ref. 4).
    In this document, the agency will consider whether a health claim 
on the relationship between sugar alcohols and dental caries is 
justified under the standard in section 403(r)(3)(B)(i) of the act and 
Sec. 101.14(c) of FDA's regulations. In addition, the agency will 
consider the petitioners' request that the agency provide in any 
regulation authorizing a claim that foods sweetened with sugar alcohols 
be exempt from the requirement in Sec. 101.14(e)(6). The following is a 
review of the health claim petition.

B. Preliminary Requirements

1. The Substances That Are the Subjects of the Petition
    Sugar alcohols are a class of organic compounds that contain chains 
of carbon atoms that bear two or more hydroxyl groups and have only 
hydroxyl functional groups (Ref. 1). The hydroxyl groups replace ketone 
or aldehyde groups that are found in sugars (Sec. 101.9(c)(6)(iii)). 
The specific sugar alcohols that are the subject of this petition are 
xylitol, sorbitol, mannitol, 

[[Page 37509]]
maltitol, maltitol syrup, maltitol solution, isomalt, lactitol, and 
mixtures of sugar alcohol substances, i.e., hydrogenated glucose syrup 
(HGS) and hydrogenated starch hydrolysate (HSH) products.
    Xylitol is a monosaccharide polyhydric alcohol with a 5-carbon 
backbone. It occurs naturally in fruits (e.g., plums, strawberries, and 
raspberries) and vegetables (e.g., cauliflower and endive) (Refs. 82 
and 83). Xylitol is made commercially by the hydrogenation of D-xylose.
    Sorbitol is a monosaccharide polyhydric alcohol with a 6-carbon 
backbone. It is found naturally in many types of berries and fruits and 
in seaweeds and algae (Ref. 82). Sorbitol is made by hydrogenation of 
glucose.
    Mannitol is also a 6-carbon, monosaccharide polyhydric alcohol. It 
occurs widely in nature in plants (e.g., pumpkins, mushrooms, onions, 
beets, celery, and olives), algae, and fungi. Like sorbitol, mannitol 
is made commercially by the hydrogenation of glucose.
    Maltitol is a disaccharide alcohol (4-D-glucopyranosyl-D-sorbitol) 
with a 12-carbon backbone. It is produced commercially by hydrogenation 
of maltose.
    Lactitol is also a disaccharide alcohol (-D-
galactopyranosyl D-sorbitol) with a 12-carbon backbone. It is produced 
by hydrogenation of lactose (Ref. 84).
    HSH and HGS are mixtures of sugar alcohols manufactured by 
hydrogenation of corn starch or glucose syrups. The composition of the 
sugar alcohols in the final product will depend on the manufacturing 
process. Therefore, HSH and HGS products from different manufacturers 
may contain different proportions of the same sugar alcohols. One HSH 
product, under the trade name ``Lycasin,'' was first produced in Sweden 
by hydrogenation of potato starch. The Swedish product contained a 
mixture of sorbitol, maltitol, maltotrititol, and hydrogenated 
dextrines of various molecular weights. When the manufacturing process 
was moved to France in the 1970's, the production process was also 
changed (Ref. 85). The French product, ``Lycasin 80/55,'' was made from 
the hydrogenation of corn starch and contained 6 to 8 percent sorbitol, 
50 to 55 percent hydrogenated disaccharides, 20 to 25 percent 
trisaccharides, and 10 to 20 percent hydrogenated polysaccharides (Ref. 
75). Lycasin 80/55, or HSH 80/55, is less fermentable and produces less 
acid than the Swedish product (Ref. 85).
    Isomalt, also known by the commercial name ``Palatinit,'' is an 
equimolar mixture of the disaccharide alcohols of -D-
glucopyranosyl-D-sorbitol and -D-glucopyranosyl-D-mannitol. 
It is produced by treating sucrose with enzymes, followed by 
hydrogenation of the resulting mixture.
2. The Substances are Associated With a Disease for Which the U.S. 
Population is at Risk
    Dental caries is recognized in The Surgeon General's Report on 
Nutrition and Health (Surgeon General's report) as a disease or health-
related condition for which the United States population is at risk 
(Ref. 7). The overall prevalence of dental caries imposes a substantial 
burden on Americans. Of the 13 leading health problems in the United 
States, dental diseases rank second in direct costs (Ref. 7).
    Based on this fact, FDA tentatively concludes that sugar alcohols 
meet the requirement in Sec. 101.14(b)(1).
3. The Substances Are Food
    Sugar alcohols are used as replacements for simple and complex 
sugars as sweeteners and bulking agents in foods (Ref. 1). Thus sugar 
alcohols are consumed for their taste and for their effect as a 
stabilizer and thickener (21 CFR 170.3(o)(28)). Therefore, FDA 
tentatively concludes that these substances satisfy the preliminary 
requirements of Sec. 101.14(b)(3)(i).
4. The Substances Are Safe and Lawful
    Several of the sugar alcohols that are the subject of this 
proceeding are currently listed in FDA's food additive and generally 
recognized as safe (GRAS) regulations, i.e., xylitol (21 CFR 172.395), 
mannitol (Sec. 180.25 (21 CFR 180.25)), and sorbitol (Sec. 184.1835 (21 
CFR 184.1835)). Moreover, GRAS affirmation petitions have been 
submitted for each of the remaining substances, i.e., maltitol (GRASP 
6G0319), maltitol syrups (HGS syrups) (GRASP 3G0286), isomalt (GRASP 
6G0321), lactitol (GRASP 2G0391), HSH (GRASP 5G0304) and HSH syrups 
(GRASP 1G0375).
    The agency notes that these GRAS affirmation petitions are under 
consideration and that any positive action resulting from this proposed 
rule should not be interpreted as an indication that the agency has 
affirmed those uses of the sugar alcohols as GRAS. Such determinations 
can only be made after the agency has completed its review of the GRAS 
petitions. A preliminary review of the GRAS affirmation petitions 
reveals that they contain significant evidence supporting the safety of 
these substances.
    The agency also points out, however, that some concerns about the 
safety of sugar alcohols do exist. For example, in a filing notice for 
the affirmation of the GRAS status of lactitol (58 FR 47746, September 
10, 1993), FDA stated that ``the agency's notice of filing of GRASP 
2G0391 should not be interpreted either as a determination, preliminary 
or otherwise, that the issue of Leydig cell tumors has been resolved or 
that lactitol qualifies for GRAS affirmation.'' Also, by notice in the 
Federal Register of December 13, 1994 (59 FR 64207), the agency 
announced the filing of a food additive petition (FAP 4A4412) to amend 
the interim food additive status of mannitol to permit an alternate 
method of manufacture. In this notice, the agency pointed out concerns 
about data from studies on mannitol that demonstrate a significant 
incidence of benign thymomas, and an abnormal growth of thymus gland 
tissue, in female rats fed mannitol. In addition, the safety of sugar 
alcohols has been examined by the Federation of American Societies for 
Experimental Biology (FASEB) (Ref. 90), as well as internationally by 
the Joint Expert Committee on Food Additives (Ref. 91). The agency also 
notes that two of the sugar alcohols that are listed in FDA's food 
additive and GRAS regulations, i.e., mannitol (Sec. 180.25) and 
sorbitol (Sec. 184.1835), require a warning label regarding laxation if 
daily consumption of these sugar alcohols is expected to exceed 20 
grams (g) per day for mannitol and 50 g per day for sorbitol. Nothing 
in this proposal alters these requirements.
    Based on the totality of the evidence, the agency is not 
challenging, at this time, the petitioner's position that the use of 
sugar alcohols is safe and lawful. Although FDA tentatively concludes 
that the petitioner has satisfied the requirements of 
Sec. 101.14(b)(3)(ii), the agency requests comments on its tentative 
conclusion.
III. Review of Scientific Evidence

A. Introduction

    The development of dental caries is the result of an interaction 
between sugars (and other fermentable carbohydrates, such as refined 
flour) and oral bacteria in a suitable environment (Ref. 71). 
Microorganisms, and Streptococcus mutans (S. mutans) in particular, in 
dental plaque metabolize available dietary sugars, producing acid and 
sticky polysaccharides that adhere to the tooth as plaque. Acid 
produced from rapid and complete fermentation of sugars creates an acid 
environment within the 

[[Page 37510]]
plaque, characterized by a pH of usually less than 5.0, that is capable 
of demineralizing tooth enamel and causing a carious lesion.
    Studies designed to measure the cariogenicity of a food assess the 
potential to cause caries if it is consumed in a standard way by a 
highly susceptible subject (Ref. 8). The methods used to measure 
cariogenic potential include long-term controlled human caries trials, 
in vivo and in vitro plaque pH measurement, demineralization and 
remineralization techniques, and rat caries models (Refs. 8 through 
11). Because long-term clinical caries trials are difficult to conduct, 
an integration of the plaque pH, animal caries, and demineralization 
methodologies has been recommended as the best measure for establishing 
the cariogenic potential of a food (Ref. 12). Experts recommend, 
however, that these methods be used with appropriate controls, such as 
sucrose, to assess experimental results (Ref. 13).
    Plaque acidity studies are useful in providing evidence on the 
effects of many microbial and physiological factors on the cariogenic 
potential of foods (Ref. 78). An acidic plaque environment at the tooth 
surface, specifically a pH of less than 5.5, suggests microbial 
fermentation of a substrate resulting in microbial growth, plaque and 
acid production, and promotion of carious lesions from enamel 
decalcification. Factors that can modify these effects include the 
presence of promoters or inhibitors in food products that affect 
bacteria growth, the nature of the acids produced as a result of 
bacterial metabolism of food carbohydrates (Ref. 78), intraplaque 
buffering, and the pH of mixed saliva (Ref. 74).

B. Review of Scientific Evidence

1. Evidence Considered in Reaching the Decision
    The petitioners submitted scientific evidence on the various sugar 
alcohols and their effects on plaque, plaque pH, and dental caries. 
This evidence included human (in vivo and epidemiological), animal, and 
in vitro studies regarding the association between consumption of sugar 
alcohols from chewing gum and other foods and plaque pH, acid 
production, plaque quantity and quality, bacteria levels, and the 
incidence of caries. The petition included four tables that summarized 
the information for: (1) Human plaque and demineralization, (2) 
bacteriological studies, (3) animal experiments, and (4) human 
longitudinal and field studies. A fifth table provided a summary of 
review articles.
    In addition to the information submitted by the petitioner, the 
agency considered other studies and reviews, such as the reports on 
health aspects of sugar alcohols by the Life Sciences Research Office 
(LSRO) and the FASEB (Refs. 14 through 16). The agency also considered 
the results of additional human epidemiological studies on caries 
incidence and demineralization; studies of animal caries; and in vitro 
plaque pH studies.
2. Criteria for Selection of Human Studies
    The criteria that the agency used to select pertinent studies were 
that the studies: (1) Present data and adequate descriptions of study 
design and methods; (2) be available in English; (3) provide daily 
intakes of the sugar alcohol or enough information to estimate their 
daily intakes; (4) include in vivo or in vitro assessment of the 
changes in plaque pH or plaque acid production; (5) for intervention 
studies on caries development, be of no less than 2 years (yr) in 
duration; and (6) be conducted in persons who generally represent the 
healthy United States' population (adults or children).
    In selecting human studies for review, the agency decided that only 
those studies investigating the use of sugar alcohols in chewing gums 
and other foods, including mouth rinses that would be representative of 
beverages, were appropriate for review. The agency excluded studies 
that were published in abstract form because they lacked sufficient 
detail on study design and methodologies, and because they lacked 
necessary primary data. In selecting animal and in vitro studies for 
review, the agency chose those studies that measured caries 
development, plaque pH, or acid production from plaque bacteria.
3. Criteria for Evaluating the Relationship Between Sugar Alcohols and 
Human Dental Caries
    The subject of the petitioned health claim is the nonpromotion of 
dental caries by sugar alcohol-containing foods, especially chewing gum 
and confectioneries. To support this claim, there needs to be 
significant scientific evidence to show that the sugar alcohol or sugar 
alcohol mixture, e.g., HSH, makes no contribution to the progression of 
dental carious lesions in humans. It would be difficult, if not 
impossible, to design and execute a study that would directly address 
this issue because such a study would require a control group that 
consumed foods containing no sugars, fermentable carbohydrates, or 
sugar alcohols.
    In the absence of studies that directly evaluate the nonpromotion 
of dental caries by sugar alcohol-containing foods, the agency gave the 
greatest weight to those studies that evaluated in vivo the acidogenic 
potential of plaque and plaque pH of sugar alcohols and sucrose in 
representative food systems (e.g., confectioneries and solutions). 
These in vivo measures can provide specific information about the 
effect of sugar alcohols in the oral environment and, more 
specifically, about the effect of sugar alcohols on pH at the interface 
between dental plaque and tooth surfaces. The more acidic the 
environment on the tooth surface, the greater the chance for enamel 
demineralization and caries formation.
    The agency also considered in vitro studies that measured plaque pH 
and acid production of sugar alcohols in solution, and long-term caries 
trials that evaluated caries development in a population using foods 
containing sugar alcohols and sucrose. Studies investigating in situ 
the demineralization or remineralization of enamel as a result of the 
action of sugar alcohols on human dental plaque were considered as 
supporting evidence by the agency.
C. Human Studies

1. Evaluation of Human Studies
    FDA evaluated the results of studies against general criteria for 
good experimental design, execution, and analysis. The criteria that 
the agency used in evaluating these studies included appropriateness of 
subject selection criteria; adequacy of the description of the 
subject's oral health before intervention; extent of evaluation of 
subject's type of dental plaque (i.e., sticky or nonsticky, thick or 
thin); methods of plaque collection; adequacy of methods used to assess 
study endpoints (e.g., in vivo versus in vitro assessment of plaque 
pH); and other study design characteristics, including randomization of 
subjects, appropriateness of controls, report of attrition rates 
(including reasons for attrition), frequency of snack or substance 
consumption, recognition and control of confounding factors (for 
example, the subject's use of fluoride during the test period), and 
appropriateness of statistical tests and comparisons. The agency also 
considered it desirable if information on treatment and control diets, 
the sugar alcohol content of the test substance, and daily sugar 
alcohol and nutrient intakes was available. 

[[Page 37511]]

    A review of the studies evaluating the effect of sugar alcohols on 
plaque pH and acid production and of the in vitro microbiological 
studies is provided in Table 1. Table 2 provides a review of 
epidemiological studies evaluating the incidence of dental caries and 
studies on demineralization and remineralization.
2. Summary of Evidence Relating Sugar Alcohol and Plaque pH or Acid 
Production
    Bibby and Fu (Ref. 38) measured human plaque pH in vitro using 0.1-
, 1.0-, or 10-percent solutions of the following sweeteners: Sucrose, 
HSH, mannitol, isomalt, xylitol, isomaltulose, sorbose, saccharin, and 
aspartame. Results showed the lowest plaque pH was attained with 
sucrose (1- and 10- percent solution: pH less than 5.0). Plaque pH 
decreased with increasing concentrations of isomalt, sorbitol, 
mannitol, and HSH. The lowest pH attained for isomalt was about 5.6, 
for sorbitol 5.82, for mannitol 5.22, and for HSH about 5.0. Negligible 
acid production was measured from aspartame, saccharin, and xylitol. 
Solution mixtures of xylitol (5 to 20 percent) and sucrose (10 percent) 
were fermented to the same low pH as sucrose alone. Thus, the presence 
of xylitol in a sucrose and xylitol mixture did not affect acid 
production in plaque from sucrose.
    The results of this study support the contention that xylitol does 
not promote dental caries by lowering plaque pH below 5.5. However, the 
results for sorbitol, mannitol, isomalt, and HSH do not support a 
``nonpromotion'' claim. The results suggest that when higher 
concentrations of these sweeteners are present in food, the plaque pH 
may reach a level that will promote decalcification of dental enamel.
    Birkhed and Edwardsson (Ref. 39) measured plaque pH and acid 
production of human plaque samples in solutions of mannitol, xylitol, 
maltitol, sorbitol, French HSH, Swedish HSH, fructose, and glucose 
syrups. Results showed that plaque pH in the presence of xylitol, 
maltitol, mannitol, and French HSH increased or slightly decreased from 
baseline (pH remaining at 6.8 or above). Sorbitol showed a slight 
decrease in plaque pH, but the final pH attained was about 6.0. The 
other sweeteners, including Swedish HSH, depressed plaque pH below pH 6 
over the 30-min (min) test period. The results of this study showed 
that mannitol and xylitol produced no plaque acid compared to sucrose. 
Maltitol and sorbitol produced plaque acid at rates that were 10 to 30 
percent of that of sucrose. French HSH produced 20 to 40 percent and 
Swedish HSH 50 to 70 percent of the acid produced by sucrose.
    Birkhed et al. (Ref. 40) measured acid production in vitro and 
plaque pH changes in vivo over a 30-min period following a 30-second(s) 
mouth rinse with 10-percent glucose or sorbitol solutions. To determine 
whether plaque microorganisms can adapt to the presence of sorbitol, 
i.e., use it as a source of energy like sucrose, with repeated exposure 
to the sugar alcohol, investigators measured plaque pH and acid 
production at the end of a 6-week (wk) period. During the 6-wk period, 
each subject rinsed their mouth six times per day for approximately 2 
min at a time with a 10-percent sorbitol solution. At the end of 6 wk, 
plaque pH was again measured for a 30-min period following a mouth 
rinse with glucose and sorbitol. The study results showed acid 
production in the presence of sorbitol, before adaptation, to be 11.3 
percent of that from glucose. After the adaptation period, plaque acid 
production from sorbitol increased to 30 percent of the glucose rate. 
After the adaptation period to sorbitol, the glucose rinse produced 
mean plaque pH values that were higher than before the adaptation 
period. The differences in plaque pH, however, were only significant at 
2 and 5 min following the rinse.
    Overall results of this study suggest that sorbitol produces very 
little plaque acid. Mean plaque pH values after sorbitol adaptation in 
the presence of the 10-percent sorbitol rinse showed only a slight 
decrease from the baseline value. The differences in mean plaque pH, 
compared to baseline, at 5, 10, 20, and 30 min following the rinse were 
significant. The authors noted that the fermentability of sorbitol was 
more pronounced after the adaptation period than before.
    Birkhed et al. (Ref. 41) studied the effects on in vivo plaque pH 
and in vitro acid production from HSH (Swedish HSH), maltitol, 
sorbitol, and xylitol. Subjects in each group sucked on two lozenges a 
day, containing 0.5 g of one of the four sweeteners and 0.5 g of gum 
arabic, four times daily between meals (total of eight lozenges per 
day) for 3 months (mo). Changes in plaque pH over a 30-min period were 
measured in each of the sugar alcohol groups after a 30-s mouth rinse 
with a 50-percent solution containing the same sweetener as the 
lozenge. The rinse was used 1 wk before and 1 wk after the lozenge 
period. A control group consumed no lozenges but rinsed with each of 
the four sweeteners. At least 1 wk separated each mouth rinse 
experiment. Acid production activity (APA) from dental plaque suspended 
in glucose and each of the four sugar alcohols was determined 1 wk 
before and 1 wk after the 3-mo consumption period.
    The results with HSH showed that although plaque pH values differed 
before and after the lozenge period, differences were not statistically 
significant, and that the lowest plaque pH attained was above 6.0. In 
the maltitol group, plaque pH before the lozenge period was higher than 
the pH following the lozenge period. Differences at 2, 10, and 30 min 
were statistically significant. However, there were no significant 
differences in plaque pH at any time compared to baseline. The lowest 
plaque pH recorded was about 6.9. Plaque pH in the xylitol group 
changed very slightly, remaining around pH 7. Plaque pH in the sorbitol 
group was higher before than after the lozenge period. Differences in 
pH at times 0 to 20 min and 0 to 30 min before compared with after the 
test period were statistically significant (p<0.05). Final plaque pH 
values after the 30-min test period were between 6.7 and 7.0. There 
were no significant differences in plaque pH between the test and 
control groups using any of the test rinses.
    Comparing the APA results for each sweetener with those for glucose 
showed that HSH was 56 percent of that of glucose 1 wk before the 
lozenge period and 59 percent of that of glucose 1 wk after the lozenge 
period. The APA for maltitol compared to glucose was 26 percent 
(before) and 32 percent (after), sorbitol was 15 percent (before) and 
18 percent (after), and xylitol was 0 percent at both time periods. 
Differences before and after each 3-mo lozenge period were not 
statistically significant for any of the sugar alcohols.
    The results of this study suggest that even though there is some 
acid production from HSH, maltitol, and sorbitol, the effect on plaque 
pH in vivo is not detrimental to tooth enamel.
    Frostell (Ref. 42) evaluated the effect on plaque pH of sugar 
solutions and different types of candy and foods. Although the focus of 
this study was not sugar alcohols, the investigators used sorbitol and 
HSH as a comparison to sucrose in some of the experiments. Plaque was 
collected prior to the test period, and its pH was determined. Subjects 
then rinsed with a test solution or ate a piece of candy or other food 
being tested. Plaque was collected after 2, 5, 10, 20, and 30 min and 
its pH was again measured. Sweeteners tested included a sucrose rinse 
(concentrations from 0.05 to 50 percent), sorbitol tablets (2 g 
sorbitol), sugar tablets (containing 

[[Page 37512]]
glucose and sucrose), HSH candy, sugar candies (with sucrose, dextrose, 
and maltose), marmalades (60-percent HSH or sucrose), and sugar-
sweetened sponge cakes, ginger cakes, marshmallows, and chocolates. 
Results with the sucrose rinses showed that plaque pH decreased with 
increasing concentrations of sucrose.
    Comparing the effects on plaque pH between the sorbitol and sucrose 
candies results showed that in the sorbitol group's plaque pH increased 
from about 6.5 (baseline) to 6.9 before returning to baseline. Plaque 
pH decreased in the sucrose group from 6.5 (baseline) to about 6.0. 
After 10 min, the pH in the sucrose group slowly increased to about 
6.3. Differences in plaque pH between the sorbitol candy and sucrose 
candy groups were significant at all time periods. In the HSH candy 
group, plaque pH was significantly higher than that in the group 
consuming sucrose candy. Differences were significant at all time 
periods. The lowest plaque pH in the HSH group was above pH 6.3. The 
group consuming marmalade with HSH experienced a drop in plaque pH to 
about 6.0 (from 7.0) after 5 min, followed by a gradual increase to a 
final pH of about 6.5. The group consuming sucrose marmalade 
experienced a plaque pH of about 5.3 after 5 min, followed by a gradual 
increase in pH to about 6.0.
    Toors and Herczog (Ref. 43) evaluated in vivo plaque pH and in 
vitro fermentability of an experimental (nonsucrose) licorice in a 
pooled plaque-saliva mixture. Fermentability (i.e., acid production) of 
the test substrates was expressed as a percentage of the sucrose 
licorice. Plaque was collected from 12 volunteers on the day after they 
consumed 10 pieces of the candy. In vivo plaque pH was measured during 
and after consumption of licorice by means of pH telemetry. Substrates 
used in the above tests included sucrose licorice, the experimental 
licorice, components of the experimental licorice (including sorbitol, 
potato starch derivative, soy flour, and others), xylitol, hydrogenated 
potato starch (HPS) (a type of HSH), and a white bread suspension. 
Results showed the fermentability of the test substrates to be as 
follows: Potato starch derivative (82 percent), soy flour (75 percent), 
sorbitol (12 percent), experimental licorice (68 percent), xylitol (5 
percent), HPS (60 percent), and white bread suspension (79 percent). In 
vivo plaque pH results showed sucrose licorice with a minimum plaque pH 
of about 5.0, experimental licorice with a minimum plaque pH of about 
5.5, and a sucrose rinse with a plaque pH of about 4.5.
    The results of this study show that food ingredients like soy flour 
can contribute to the cariogenicity of a food regardless of the 
presence of a sugar alcohol.
    Gallagher and Fussell (Ref. 44) compared the in vitro 
fermentability of xylitol and other sugar alcohols with sucrose in 
dental plaque. Plaque collected from adults and children of different 
ages was incubated in broth culture. Acid production was measured as 
pH. The control media contained no added carbohydrates.
    The results of acid production measurements showed that sucrose was 
significantly more acidogenic compared to the control and xylitol. 
Differences were significant. There was no significant difference in 
acid production between the control groups and the xylitol groups.
    Gehring and Hufnagel (Ref. 45) described intra- and extraoral pH 
measurements of dental plaque. Six adult men and women rinsed for 2 min 
using one of seven test substances followed by intraoral plaque pH 
measurements after 3, 4, 5, 7, 9, 13, 17, 21, 27, and 32 min. For the 
extraoral test, visible plaque was removed, suspended in distilled 
water, and the pH measured at 3, 5, 7, 9, 11, 15, and 25 min after 
subjects rinsed with test substances. Test substances included 20 
percent solutions of glucose, sucrose, fructose, HSH, mannitol, 
isomalt, sorbitol, sorbose, or xylitol.
    The results of the intraoral plaque pH measurements showed only 
slight pH decreases within 5 min after administration of xylitol and 
mannitol, with a return to baseline measures at the end of the 32-min 
test period. Sorbitol, HSH, isomalt, and sorbose reached a minimum pH 
just below 6.0 after 5 min followed by a slight increase to about pH 
6.1 to 6.4 at the end of the test period. Sucrose, glucose, and 
fructose showed a minimum pH value of about 4.6 to 4.7 (after 5 min) 
with an increase to about pH 5.3 to 5.5 at the end of 32 min. Minimum 
plaque pH by extraoral measurements were higher than the pH according 
to intraoral measurements. Sucrose, glucose, and fructose minimum pH 
values ranged from about 5.0 to 5.7 after 5 min and increased to about 
5.6 to 6.0 after 32 min. Other pH values were not given. The authors 
attribute the differences in intra- and extraoral plaque pH 
measurements to methods in handling plaque removal and the influence of 
saliva substances.
    Havenaar et al. (Ref. 46) evaluated in vitro acid formation from 
oral bacteria in the presence of sugar substitutes and the influence of 
xylitol on glucose in growing cultures of S. mutans. Fresh isolates of 
Streptococci and other strains were obtained from caries free and 
caries active subjects. Acid production in 1-percent solutions of 
glucose (control), sorbose, sorbitol, xylitol, lactitol, maltitol, and 
HSH was determined by incubating the sweetener in phenol red broth 
containing oral bacteria. A color change indicated acid formation. 
Changes in pH was measured after subculturing S. mutans in each of the 
sweeteners, after frequent subculturing in each sweetener to obtain 
adapted strains of S. mutans, and after subculturing the adapted 
strains once in glucose and resubculturing in the sweetener. Growth of 
S. mutans and pH measurements were also measured in a glucose broth 
with and without added xylitol.
    The results showed no acid production from xylitol or sorbose and 
acid production from sorbitol, lactitol, and HSH. The authors stated 
that S. mutans slowly fermented maltitol. Results also showed no change 
in pH with xylitol and a moderate drop in pH to about 6 to 6.5 (actual 
values not given) with maltitol, sorbitol, lactitol, and HSH after 120 
min. Adaptation by S. mutans to the sweeteners resulted in a marked 
increase in fermentation, with final pH values dropping to about 4.5 to 
5.5. After one subculturing of the adapted strain in glucose, S. mutans 
lost most of its ability to ferment the sweeteners. The addition of 
small amounts of xylitol to glucose broth somewhat inhibited acid 
production from S. mutans, but it had no effect on final pH attained.
    Jensen (Ref. 47) measured interproximal plaque pH in subjects using 
five different HSH's and sorbitol and sucrose as controls. Four 
subjects rinsed with a 5 milliliter (mL) portion of the test solution 
for 60 min. Plaque pH was then monitored for 30 min. Following the pH 
measurements, the subject rinsed their mouth with distilled water and 
chewed paraffin for about 5 min to bring oral pH back to resting 
levels. The test was repeated with each subject using each of the four 
test solutions.
    The results showed that plaque pH for all test substances remained 
above pH 6.0 over the 30-min test period. Plaque pH using the sorbitol 
rinse was similar to that using the test substances. Using the sucrose 
rinse resulted in plaque pH measurements of approximately 4.0 to 4.1. 
The identity of the test substances was not provided in this 
unpublished study. Results indicate that the HSH solutions used in this 
study were 

[[Page 37513]]
significantly less acidogenic than sucrose and no different than 
sorbitol.
    Maki et al. (Ref. 48) compared acid production in vivo from 
isomaltulose, sorbitol, xylitol, and sucrose (control) in human dental 
plaque. Dental plaque was collected from 12 individuals and incubated 
with phosphate buffer. After endogenous acid production was measured, a 
1-percent solution of the test substance in the same buffer was added, 
and acid production measured again.
    The results showed no acid production in the presence of xylitol. 
Compared to sucrose (100-percent acid production), acid production from 
sorbitol was 1 percent. The authors noted that the percent acid 
production from sorbitol may vary considerably among individuals and 
with the amount of exposure to sorbitol.
    Park et al. (Ref. 49) measured interproximal plaque pH in five 
subjects after consuming one of three snacks alone or one of three 
snacks followed by a single mint containing sorbitol (94 percent) or a 
sorbitol and xylitol blend (79 percent and 15 percent, respectively). 
When mints were used, they were consumed 3 min following ingestion of 
the sweet snack. Snacks tested included a sandwich cookie, cupcake, and 
granola bar. A randomized block design was used to administer the test 
products and mints (see Table 2 for further details). The lowest plaque 
pH attained after consuming the three test products without mints 
ranged from 4.02 to 4.16. When the sorbitol mint was consumed following 
the test product, mean plaque pH values increased and ranged from 4.68 
to 5.04. When the sorbitol and xylitol mint was consumed following 
consumption of the test products, mean plaque pH increased to a range 
of 5.32 to 5.60. Differences in mean plaque pH values between the mint 
products differed significantly when the mints were used after the 
granola bar and cupcake challenges. There was no significant difference 
in mean plaque pH between the sorbitol (5.04) and the sorbitol and 
xylitol mint (5.60) products when these products were used after the 
sandwich cookie challenge.
    The results show that consumption of a sugarless mint reduced the 
acidogenicity of the test snacks, although final pH values remained 
below pH 5.5 in all but one test. The authors attributed the results of 
this study to the stimulatory effects on salivary flow by sugar 
alcohols. Increasing salivary flow increases the buffering capacity of 
saliva, thus reducing the acidogenic potential of a variety of snack 
foods. The authors also attributed the additional buffering effects of 
the sorbitol and xylitol mint to the presence of xylitol and its 
potential benefits in reducing plaque microbial activity. Without a 
sucrose-containing mint as a comparison, however, the influence of 
sugar alcohols on saliva production cannot be adequately assessed.
    Soderling and coworkers (Ref. 50) investigated the effect on dental 
plaque of chewing gums that contained either xylitol, sorbitol, or a 
mixture of xylitol and sorbitol and compared the results with those 
obtained with subjects who used sucrose gums. Twenty-one subjects 
(adults, ages 19 to 35 yr) who were not habitual gum chewers were 
randomly assigned to chew gum containing either xylitol, sorbitol, or a 
blend of the two sugar alcohols for 2 wk. Subjects chewed 10 pieces of 
gum per day for an intake of either 10.9 g xylitol, 10.9 g sorbitol, or 
10.9 g xylitol and sorbitol (8.5 g xylitol and 2.4 g sorbitol). The 
control group was made up of seven habitual sucrose gum users. Subjects 
maintained their usual diets and oral hygiene except just before to 
clinic visits. Interdental plaque pH was collected, and the resting 
plaque pH determined. Plaque pH was measured at 2, 5, 10, 15, and 20 
min after an oral rinse containing the same sugar alcohols as used in 
the gum. Afterward, subjects rinsed with water and chewed a piece of 
paraffin for 1 min to expedite removal of sugar alcohols from the 
mouth. Baseline pH was again measured, followed by a mouth rinse with 
10 mL of 10-percent sucrose. Plaque pH was again determined.
    The results from using gum for 2 wk showed no significant changes 
in resting plaque pH in the xylitol and xylitol and sorbitol groups, 
whereas the use of sorbitol gum was associated with a lower pH. Final 
plaque pH values after use of sorbitol gum were significantly lower 
than baseline values, but all final values remained above pH 6.0.
    Birkhed and Skude (Ref. 51) evaluated, among other tests, the APA 
from glucose, soluble starch, and Swedish HSH in dental plaque. Eleven 
adults were instructed to avoid oral hygienic procedures for 2 days. No 
dietary changes were required. At the end of 2 days, plaque was 
collected. The APA was determined from 3-percent solutions of glucose, 
boiled soluble starch, and HSH. The APA was also determined in 
increasing concentrations (0.003 to 12 percent weight per volume (w/v)) 
of starch and HSH.
    The results showed significantly lower (p<0.001) APA from soluble 
starch (75.7 percent) and HSH (61.5 percent) compared to glucose (99.7 
percent). The APA from HSH was also significantly lower (p<0.01) than 
that from soluble starch. The range of optimum acid production for both 
substrates was 0.03 to 6 percent. The authors noted that Swedish HSH is 
more fermentable than French HSH, which contains less high molecular 
weight hydrogenated saccharides than Swedish HSH.
    Grenby et al. (Ref. 76) evaluated the dental properties of lactitol 
compared to five other bulk sweeteners, i.e., sucrose, glucose, 
sorbitol, mannitol, and xylitol, in vitro using a standardized mixed 
culture of dental plaque microorganisms. Sweeteners were incubated for 
24 hours (h) in media containing a 1-percent solution of one of the six 
sweeteners. Plaque microorganisms were also incubated in media 
containing the sweeteners with segments of intact surfaces or with 
segments of pulverized dental enamel. The demineralization action of 
the acid produced by microbial fermentation was assayed by calcium and 
phosphorous analyses.
    The greatest amount of acid production and lowest pH (significantly 
different than the sugar alcohols) were reported with sucrose and 
glucose (pH of 4.0 to 4.3). Lactitol and xylitol showed only slight 
changes in pH and acid production over the 24 h (final pH of 6.1 to 
6.3); whereas sorbitol and mannitol showed slight changes in pH during 
the first 12 h (pH6), then gradually decreased to a final pH 
of 4.6 to 5.1 after 24 h.
    The results of the demineralization test showed highly significant 
differences (p<0.001) between sucrose and glucose and the sugar 
alcohols. The reductions in calcium and phosphorous dissolving in 
sorbitol was approximately 80 to 85 percent, mannitol 63 to 69 percent, 
and lactitol and xylitol 94 to 98 percent compared to mineral loss in 
the presence of glucose.
3. Summary of Evidence Relating Sugar Alcohol and Dental Caries: Long-
Term Studies
    Moller and Poulsen (Ref. 20) determined the effect of long-term 
chewing of sorbitol chewing gum on the incidence of dental caries, 
plaque, and gingivitis. The sorbitol chewing gum contained calcium 
phosphate which acts as a buffer in saliva to help maintain pH and aid 
remineralization. Two groups of children, ages 8 to 12 yr of age, from 
two different schools in Denmark took part in this 2-yr study. Group 1 
chewed one piece of sorbitol-containing gum three times a day, after 
meals. Group 2 chewed no gum and 

[[Page 37514]]
served as the control. At the start of the study, subjects in group 1 
had more decayed and filled toothsurfaces than the control group; 
however, the differences were not statistically significant.
    The results showed that the sorbitol group had a significantly 
lower incidence of dental caries compared to the control after 2 yr. 
The control group, which did not chew gum, did not experience the same 
salivary stimulation from the chewing of gum, nor did they have an 
equivalent source of calcium phosphate. These are large confounders in 
this study. The authors noted a number of factors that could contribute 
to the observed results, such as the sorbitol content of the chewing 
gum, reduced consumption of sugar-containing sweets, intra-examiner 
variability, and other unknown conditions.
    Banozcy et al. (Ref. 21) evaluated the effect of sorbitol-
containing sweets on the caries increment of children aged 3 to 12 yr, 
in a clinical longitudinal study planned for 3 yr. The test group 
consumed 8 g of sorbitol per day between meals, while the control group 
consumed a similar amount of sucrose-containing sweets.
    The results showed that mean decayed, missing, or filled (DMF) 
values for teeth in the sorbitol group were 1.09, 0.90, and 1.18 in the 
first, second, and third yr, respectively. The sucrose group had mean 
DMF values of 2.61, 1.86, and 1.13 for the first, second, and third yr, 
respectively. The differences in caries increments were significant 
(p<0.001) in the first and second yr but not in the third yr. The 
authors noted that the lack of significance in the third yr may be 
attributed somewhat to a lack of subject compliance since the children 
in the sorbitol group traded sweets with the sucrose group, in addition 
to other factors. Results of this study indicate that sorbitol is less 
cariogenic than sucrose.
    Kandelman and Gagnon (Ref. 22) reported on the incidence and 
progression of dental caries in school children after 12 mo of a 2-yr 
study using xylitol in chewing gum. The subjects were 433 children, 
ages 8 to 9 yr old, from 13 elementary schools, and were from low 
socioeconomic areas with a high caries rate. The children were assigned 
to one of three groups: A control group that received no chewing gum 
and chewed no gum while at school, a test group that received gum 
containing 15-percent xylitol and 50-percent sorbitol (XYL15), and a 
second test group that received gum containing 65-percent xylitol 
(XYL65). Students were not randomly assigned to groups. Rather, an 
entire class was assigned to one of the three groups. The XYL65 group 
consumed 3.4 g xylitol per day, and the XYL15 group consumed 0.8 g per 
day.
    The results showed significantly lower net progression of decay 
(NPD) (i.e., the difference in the number of reversals from the 
progressions of decay for each child) in the XYL65 group (1.25) than in 
XYL15 group (1.87) (p< 0.05), and each xylitol group had significantly 
(p<0.001) lower NPD than the control. The decayed, missing, filled 
surfaces (DMFS) increment was also significantly lower in the xylitol 
groups compared to the control. There was no significant difference in 
DMFS between the gum groups. Results of this study suggest that chewing 
gum containing xylitol or a blend of xylitol and sorbitol provided more 
benefits for teeth than not chewing gum at all.
    Rekola (Ref. 23) compared the progression of incipient carious 
lesions on buccal smooth surfaces in subjects participating in the 2-yr 
Turku sugar study (Ref. 24). Subjects consumed either a diet containing 
sucrose or one with almost complete replacement of sucrose products 
with xylitol-containing products. The progression of carious lesions 
were assessed by use of color dental photographs of the right and left 
sides and of the front of maxillary and mandibular teeth.
    The results showed that the sucrose group had a significant 
tendency for increased size of carious lesions over the 2-yr period 
compared to the group consuming xylitol (p<0.01). The white spot 
lesions in the xylitol group were significantly smaller than those in 
the sucrose group.
    Rekola (Ref. 25) quantified changes in the size of approximal 
carious lesions in subjects after 2 yr of almost complete substitution 
of dietary sucrose with xylitol (Ref. 23). Bitewing radiographs were 
taken during the 2-yr study. In this study, the radiographs were 
projected onto a planimetry plate so that the area of the lesions could 
be determined. The sizes of the lesions at the different time periods 
were compared, and the rate of caries progression was also compared. At 
the beginning of the study, there was no difference in the mean size of 
carious lesions between groups. The size of the approximal lesions, 
i.e., lesions that were neither filler nor overlapping at 0 and 24 mo, 
in the sucrose group increased significantly (p<0.001) over 2 yr 
compared to the lesions in the xylitol group. The lesion size in the 
xylitol group remained virtually unchanged.
    The authors reported a trend towards decreasing lesion size in 
canines and first molars compared to molars and second premolars in the 
xylitol group. This trend was not observed in the sucrose group. 
Results of these studies support the observation that xylitol is less 
cariogenic than sucrose.
    In a World Health Organization (WHO) field trial in Hungary (Ref. 
26), the effects of a partial substitution of sucrose for xylitol in 
the diets of 689 institutionalized children, ages 6 to 11 yr, were 
examined. The xylitol group used fluoride dentifrice and consumed no 
more than 20 g of xylitol per day in chewing gum, chocolate, hard 
candy, and wafers. The fluoride group received fluoride in dentifrice, 
water, and milk, but consumed no xylitol products. The control group 
received no fluoride treatment and consumed no xylitol-containing 
products. After 3 yr, the xylitol group had a statistically significant 
(p<0.001) lower incidence of caries compared to the control and 
fluoride groups. The authors noted that results from this study were 
obtained under conditions where caries prevalence and incidence were 
still high. Results of this study support the observation that xylitol-
containing products are less cariogenic than sucrose-containing 
products.
    In a 2-yr substudy (Ref. 28) of the WHO xylitol field studies in 
Hungary (Ref. 26), Scheinin and coworkers assessed the caries increment 
with systemic fluoride (fluoride group) and restorative treatment only 
(control group). This study differed from the 3-yr study primarily in 
baseline differences. Children entering the institutions during the 
first yr of the 3-yr study were included in this substudy.
    The substudy showed similar favorable results with xylitol compared 
to the control. The caries increment was 3.8 in the xylitol group, 4.8 
in the fluoride group, and 6.0 in the control group. The differences in 
caries increment between the xylitol group and the other two groups 
were significant (p<0.001). Results again supported a lower incidence 
of caries when xylitol is substituted for sucrose in the diet.
    In a WHO field trial in Thailand and French Polynesia (Ref. 29), 
the usefulness of a fluoride rinse, fluoridated sucrose chewing gum, 
and fluoridated xylitol (51 percent) and sorbitol gum in controlling 
dental caries was evaluated in children over a 3-yr period. In French 
Polynesia, a fourth group used nonfluoridated chewing gum sweetened 
with xylitol (51 percent) and sorbitol. Approximately 250 children at 
each of the ages 6 to 7 yr, 9 to 10 yr, and 12 to 13 yr were examined. 
The 12- to 13-yr age group was intended to provide data for 

[[Page 37515]]
comparison with the 9- to 10-yr old group, who would be ages 12 to 13 
yr at the end of the study.
    The results from the Thailand study showed that the fluoridated 
xylitol and sorbitol gum group had lower decayed, missing, and filled 
teeth (DMFT) and DMFS scores than either the fluoride rinse group or 
the fluoridated sucrose gum group. Results from the French Polynesia 
study showed that the subjects started with much higher DMFT and DMFS 
mean scores initially than the subjects in Thailand. Although the 
results with the fluoride gum sweetened with the sugar alcohols were 
better than any of the other treatments, the overall caries incidence 
in this population is very high. The presence of fluoride in the 
chewing gums confounds the results of the sugar alcohols. The authors 
describe this study population as a community experiencing an increase 
in the prevalence of the disease. This study group does not reflect the 
general population of the United States.
    In another WHO field trial, Kandelman and coworkers (Ref. 30) 
evaluated the effects of xylitol intervention on dental caries in 
French Polynesian children, ages 7 to 12 yr. Of 746 subjects enrolled 
in this 32-mo study, 468 completed the study. Subjects in the xylitol 
groups consumed 20 g of xylitol daily in various food products, such as 
chewing gum, hard candy, chocolate, and gumdrops. The control group 
received no xylitol-containing products.
    The results showed significantly reduced caries increment rate by 
37 percent to 39 percent in the xylitol groups compared to the 
controls. This study was neither randomized nor blinded. Results 
support the observation that xylitol-containing products are less 
cariogenic than the sucrose-containing products.
    Frostell and coworkers (Ref. 31) determined the effect on caries 
increment in children, ages from 2\1/2\ to 4 yr, of substituting HSH 
for sucrose in candy. During this 1\1/2\- to 2\1/2\-yr study, subjects 
in the test group consumed candies made with HSH and chewing gum made 
with sorbitol. The control group consumed sucrose candies and gum. 
Investigators monitored the intake of candies by use of coupons which 
the parents used at local stores to buy the candy. An analysis of the 
coupons used showed that parents of the children in the test group used 
a smaller number of coupons than the parents of the children in the 
control group. Based on inquiries, the investigators discovered that 
the parents of the subjects in the HSH group had also given the 
children other candy in addition to HSH candy. The consumption of HSH 
candy was reported from 50 to 75 percent of the total candy 
consumption.
    The results showed no significant differences in caries scores 
after 1\1/2\ to 2\1/2\ yr with HSH candy consumption compared to 
sucrose candy consumption. When investigators analyzed the data of 
those children whose parents consumed the correct candy for their 
group, the differences in caries increment between the groups were 
still not significant but showed a trend towards a lower incidence of 
caries in the HSH group. The results of this study were confounded by 
poor compliance, inter-examiner variability, lack of blinding, and 
inconsistent results and do not support significant dental benefits 
from the use of HSH.
    Glass (Ref. 32) evaluated the cariogenicity of sorbitol chewing gum 
with regular use by children, ages 7 to 11 yr old, living in a 
nonfluoride area. In this 2-yr study subjects were randomly assigned to 
either a no-chewing group (control) or to the one which chewed gum 
twice daily. Subjects in the gum group were provided two sticks of gum 
daily for use at school and four sticks of gum for use at home when 
school was not in session.
    The results showed that over the 2-yr study period, mean caries 
increments were 4.6 new decayed and filled (DF) surfaces for the 
sorbitol gum group (n=269) and 4.7 new DF surfaces for the no-gum group 
(n=271). The difference between the groups was not statistically 
significant. Although the results of this study suggest that adding 
sorbitol-containing gum to the diet did not result in any additional 
dental caries, the effect of chewing gum per se on the incidence of 
dental caries was not considered.
4. Summary of Evidence Relating Sugar Alcohol and Dental Caries: Short-
Term Studies
    Ikeda et al. (Ref. 33) evaluated the cariogenicity of maltitol and 
a polysaccharide alcohol using an intraoral cariogenicity test (ICT) 
and rat tests. Most of the details of the methods used in the ICT were 
not provided, making the results difficult to interpret. Bovine enamel 
fragments were extraorally dipped in 3-percent solutions of sucrose 
(control), maltitol, or the polysaccharide alcohol for 1 min every day. 
After 1 wk, hardness was measured. The higher the value for hardness 
means a softer enamel and a greater loss of enamel.
    The results showed a decalcification score for maltitol of 1.66 
compared to a score of 2.70 for sucrose. These differences were 
significant. In the animal study, one group was provided a feed with 
26-percent maltitol and 30-percent starch, a second group was provided 
a feed with sucrose instead of maltitol, and a third group consumed a 
diet without sucrose. Results showed a caries score of 45.8 for the 
sucrose group, 3.2 for the maltitol group, and 5.2 for the no-sucrose 
group. Differences between the sucrose group and the other groups were 
statistically significant.
    Yagi (Ref. 34) evaluated the effects of maltitol on changes in 
enamel hardness. Enamel decalcification was measured using an ICT with 
a denture containing two bovine enamel slabs. Four subjects wore the 
dentures for 7 days. Each day, one enamel slab was exposed to a 3-
percent maltitol solution and the other to a 3-percent sucrose 
solution. Enamel hardness was measured at the end of the wk.
    The results showed that the average change in hardness compared to 
pretreatment levels for the enamel in maltitol was 1.47 micrometers 
compared to 3.35 micrometers for the enamel in sucrose. Differences 
between the two measurements were significant. The authors noted that 
there were considerable differences in individual responses to sucrose 
and maltitol. They attributed these differences to the oral environment 
(e.g., plaque bacteria and quality and quantity of saliva). However, 
general observations were that sucrose causes significant loss of 
enamel, as evidenced by changes in enamel hardness, compared to the 
effect of maltitol on tooth enamel.
    Leach et al. (Ref. 35) evaluated in situ the effect on 
remineralization of artificial caries-like lesions in human enamel with 
sorbitol. Ten adult subjects wore cast bands containing enamel on one 
lower first molar tooth for two 3-wk periods during which they 
continued to use normal oral hygiene procedures. Artificial caries 
lesions were made in each enamel slab and covered with gauze to 
encourage the formation and accumulation of plaque on the enamel 
surface. Subjects were given snack foods (chocolate bar, raisins, 
cream-filled wafers, and cream-filled, iced cupcake) and instructed to 
consume one each morning and afternoon between meals. During the first 
experimental period, subjects chewed, for 20 min each, five sticks per 
day of commercial sugarless gum after meals and snacks. The gum was 
sweetened primarily with sorbitol and small amounts of mannitol, HGS, 
and aspartame. During the second experimental period, snacks were 
consumed but without chewing gum (control). 

[[Page 37516]]

    The results showed statistically significant (p<0.001) 
remineralization during both experimental periods compared to the 
original lesion. The difference between the remineralization with and 
without gum was also significantly different (p<0.01), indicating 
overall promotion of remineralization by gum chewing. The authors 
attributed the remineralization during the nongum period to the 
presence of gauze used with the intraoral device to collect plaque. The 
gauze could have concentrated calcium and phosphates from the diet in 
plaque and fluoride from dentifrice. It is not known what effects the 
duration and timing of the gum chewing had on the results. Without a 
comparison to sucrose-containing gum and a nonsweetened gum, it is not 
possible to evaluate the effect of chewing gum for 20 min.
    Rundegren et al. (Ref. 36) evaluated in situ the effect on 
demineralization of sucrose substitutes in a 4-wk test. Intraoral 
devices containing bovine enamel mounted on acrylic blocks were used 
with group 1. Partial dentures with enamel slabs were used with group 
2. Sweeteners tested included 10 percent solutions of sucrose, 
maltitol, and HSH. Sucrose was used as the positive control, and 0.9-
percent solution of sodium chloride was used as a negative control. 
Subjects immersed the test sites of their appliances in the test 
sweetener four times a day for a 10-min period. Plaque was collected at 
the end of 4 wk and plated to determine the content of S. mutans. The 
degree of demineralization was measured by evaluating changes in 
microhardness of the enamel. The buffering capacity of whole saliva was 
evaluated weekly by measuring final pH in a mixture of 1 mL of saliva 
and 3 mL of sodium chloride.
    The results showed a higher degree of demineralization overall in 
the adults (ages 56 to 59 yr) using the partial dentures compared to 
students (age 19 yr) using an intraoral device. Results from the test 
(n=4) of enamel microhardness in HSH versus sodium chloride suggest 
that HSH does not contribute to demineralization, and that measured 
changes in microhardness reflected the background of fermentable 
carbohydrates in the diet. Comparing the differences in microhardness 
of enamel slabs between the sucrose and HSH diets and the sucrose and 
maltitol diets showed that sucrose results in significant 
demineralization compared to the sugar alcohols.
    Creanor et al. (Ref. 37) evaluated the effect of chewing gum for 20 
min on in situ enamel lesion remineralization compared to a fluoridated 
dentifrice. Artificial enamel lesions were created in vitro in sound 
human enamel and mounted for wearing just opposite the lower first and 
second molars. Baseline mineral contents were measured. Subjects used a 
fluoridated dentifrice twice daily and maintained their regular diets. 
Six subjects chewed five sticks of chewing gum containing sorbitol and 
some HGS and aspartame after each meal and snack. The gum was chewed 
for 20 min in order to minimize any deleterious effects of sucrose. Six 
other subjects received no gum and served as the control. At the end of 
7 wk, the test subjects became the control group, and the control 
subjects became the new test group. The new test group then chewed 
sucrose-containing gum for 7 wk.
    The results showed that after using sugar-free gum for 7 wk, the 
degree of mineral loss for the enamel corresponded to a 
remineralization value of 18.2 percent. After 7 wk of chewing sucrose 
gum, the percent remineralization was calculated to be 18.3 percent. 
The difference between the sorbitol and sucrose gum groups was not 
significant. Results of this study suggest that chewing gum for 20 min, 
regardless of the sweetener, can be beneficial to dental health.
    A common problem in studies evaluating the dental health benefits 
of sugar alcohol-containing chewing gum is the absence of an 
appropriate control group. Most of the studies that have been done use 
a control group that does not chew gum. Ideally, to evaluate the 
relationship of sugar alcohol-sweetened chewing gum in not promoting 
dental caries, the control group would chew an unsweetened gum product. 
Such a group is needed to take into consideration the effects of 
chewing gum itself on the endpoint measure, e.g., plaque pH or plaque 
acid production. Chewing gum is known to stimulate saliva, which can 
help neutralize oral acids, raise plaque pH, and help to promote enamel 
remineralization in some circumstances. It would be considered 
unethical by standards in the United States to use a control group that 
chews sucrose-containing gum and, as a consequence, puts the subjects 
at risk of dental disease, in order to compare the incidence of dental 
caries to that from a sugar alcohol-containing gum.
    The few long-term caries field trials that were submitted with this 
petition show how multiple problems in the execution of clinical 
studies can easily confound the results. Problems often include subject 
compliance, reporting and control of dietary intake, selection of 
appropriate control foods, inter- and intraexaminer variability, 
subject attrition, and inability to blind the study. The majority of 
these trials compared sucrose consumers to individuals who had partial 
or complete substitution of sugar alcohols for sucrose. The results 
consistently demonstrated significantly fewer caries in the group 
consuming sugar alcohols than in the group consuming sucrose.
    Although the relationship between some of the sugar alcohols and 
promotion of dental caries has not been well studied in humans, it is 
becoming increasingly evident that sugar alcohols, when substituted for 
sucrose and other fermentable carbohydrates, may provide important 
dental health benefits for the consumers of those products.

D. Animal Studies

    FDA reviewed over 20 animal studies investigating the effects of 
sugar alcohol consumption on the incidence of dental caries or on the 
acidogenic potential of dental, S. mutans, or mixed oral 
microorganisms. Most of the animal studies that have been done to test 
the effect of sugar alcohols on the incidence of caries were programmed 
feeding studies using weanling rats. The animals were usually divided 
into groups and fed diets containing different test sweeteners. The 
control diets were either a basal diet with no carbohydrate sweeteners 
or sugar substitutes or a basal diet with added sucrose. The test diets 
were administered over a period of weeks, increasing the sugar 
substitute concentration slowly to allow the animals time to adapt to 
the specific sweetener and to minimize the severity of diarrhea, a side 
effect of sugar alcohol consumption that increases with increasing 
concentration of the sugar alcohol.
    Investigators also evaluated the general health and growth of the 
animals during the experimental period. Many animals, and rats in 
particular, do not like the taste of sugar alcohols and, therefore, 
will eat less of the test diet and increase their intake of water. Most 
investigators monitored the animals' total dietary intake to ensure 
that consumption patterns were similar between the control and test 
animals.
    A potential confounding factor in these studies is the effect of 
total food and water intake on caries development. If animals consume 
less of a sugar alcohol diet compared to the control animals consuming 
a sucrose diet, any significant differences in caries incidence may 
actually be attributable to the differences in food and water 
consumption and not to an effect of the sugar substitute. Some studies 
reported a lower survival rate in animals on the 

[[Page 37517]]
sugar alcohol diets. This finding made interpretation of the results 
more difficult because of uneven group sizes.
    In order to promote the cariogenic process, the animals were 
inoculated with either mixed strains of plaque bacteria or purified 
strains of S. mutans and other microorganisms found in dental plaque. 
Experimental periods lasted, on the whole, for 60 to 70 days. These 
periods included the time given for the animals to adapt to the test 
diets.
    Havenaar et al. (Ref. 52) fed S. mutans inoculated rats one of six 
diets 18 times a day: The basal diet plus 50-percent starch, or the 
basal diet plus 30-percent starch and 20 percent of either sucrose, 
HSH's, xylitol, sorbitol, or L-sorbose. In a second experiment, the 
rats were fed the same diets 14 times a day and alternated with the 
basal diet containing 20-percent sucrose and 10-percent glucose (four 
times a day). In both experiments, the starch, HSH, xylitol, and L-
sorbose groups showed significantly less fissure lesions than the 
sorbitol and sucrose groups. The sorbitol group showed significantly 
less fissure caries in the mandibular molars with respect to the 
severity of the lesions compared to the sucrose group.
    Havenaar et al. (Ref. 53) in five successive experiments, fed rats 
ad libitum on diets containing sucrose or HSH 80/55. In each 
experiment, the rats were inoculated with plaque from rats in the 
previous experiment (Ref. 52). Results showed that compared to sucrose, 
HSH was relatively noncariogenic. The incidence of fissure caries in 
the mandibular molars for rats consuming 20-percent sucrose was 13.1, 
whereas the fissure caries incidence in rats consuming 20-percent HSH 
was 1.5 to 2.5 (p<0.001).
    Havenaar et al. (Ref. 54) evaluated the usefulness of diets for 
testing the caries promoting or inhibiting properties of sugar 
substitutes. The investigators fed two groups of rats experimental diet 
2000 containing 50-percent sucrose and 14-percent starch or 50-percent 
sucrose, 9-percent starch, and 5-percent xylitol for a period of 42 
days. Results showed no significant differences in caries incidence 
between the sucrose starch, the xylitol group and the sucrose and 
starch group. In another experiment animals were fed diet SSP 20/5 
containing 20-percent sucrose, 5-percent glucose, and 25-percent starch 
or 20-percent sucrose, 5-percent glucose, 20-percent starch, and 5-
percent xylitol for a period of 66 days. Results showed the xylitol, 
sucrose, and starch group to have significantly fewer caries (12.3 
caries versus 14.8) compared to the sucrose, starch, and glucose group.
    Havenaar and coworkers (Ref. 55) fed one group of rats a basal diet 
containing 20-percent sucrose, 5-percent glucose, and 25-percent 
starch. The test group received the basal diet with 20-percent starch 
and 5-percent xylitol and fluoride. After 54, 75, or 96 days, rats were 
crossed over to the other diet for an additional 21 to 42 days. Results 
showed that the xylitol group had significantly fewer fissure caries 
than the sucrose group. The authors also reported that the longer the 
experimental period, the more severe the caries, irrespective of the 
presence of xylitol. After crossover, total numbers of caries did not 
change, but the xylitol group showed significantly fewer initial 
lesions compared with the mean caries incidence in the sucrose group on 
day 54.
    Grenby and Colley (Ref. 56) fed a control group of rats a 
cariogenic diet containing 46-percent sucrose and fed two test groups 
the same cariogenic diet either with 20 percent of the sucrose replaced 
with xylitol, sorbitol, mannitol, or wheat starch (experiment A). The 
animals consuming sorbitol and mannitol did not remain healthy during 
the experiment, so this part of the experiment was terminated. The 
animals consuming xylitol also experienced difficult health effects at 
first but later improved and were returned to the 20-percent xylitol 
diet. In experiment B there were only two diets: A cariogenic diet with 
46-percent sucrose and an experimental diet with 10 percent of the 
sucrose in the diet replaced with xylitol.
    In experiment A, significantly fewer caries were experienced only 
in the group consuming the sucrose and xylitol diet compared to the 
control group. In experiment B, the level of caries was high for both 
the sucrose group and the sucrose and xylitol group. The overall caries 
scores were not significantly different.
    Karle and Gehring (Ref. 57) evaluated the effect of sugar alcohols 
and sucrose on both xerostomized (salivary glands removed) and 
nonxerostomized rats. The control group consumed a basal diet without 
sweetener. Test groups received the basal diet plus sucrose, xylitol, 
isomalt, or other sweeteners. Sweetener concentrations were increased 
over a 3-wk period to a level of 30 percent of the diet. The 
xerostomized rats had more caries with all substances than the 
nonxerostomized rats. Sucrose was shown to be the most cariogenic 
sweetener, and xylitol the least cariogenic, in the nonxerostomized 
rats. Both the xylitol and isomalt groups had significantly fewer 
caries than the sucrose group.
    Muhlemann and coworkers (Ref. 58) compared the cariogenicity of 
diet 2000 (containing 64-percent wheat flour) to the same diet 
containing xylitol or sorbitol (15 percent and 25 percent of the flour 
replaced) or sucrose (15 percent and 25 percent of the flour replaced). 
Sweetener mixtures containing 15-percent sucrose and 15-percent xylitol 
or sorbitol and 25-percent sucrose and 25-percent xylitol or sorbitol 
were also substituted for the flour ingredient of the basal diet. The 
rats consuming diets with 15- and 25-percent sucrose experienced 17.3 
and 17.8 smooth surface caries, respectively. Rats consuming animal 
chow with 15-percent xylitol or sorbitol experienced 0.0 and 1.9 smooth 
surface caries, respectively. The caries score for the control group 
was 4.9. The highest number of fissure caries (11.3) occurred in the 
25-percent sucrose group. The control group had 5.1 lesions. 
Substituting xylitol (25 percent) in the diet resulted in fewer caries 
(0.2) compared to the control, but differences were not significant. 
Twenty-five percent sorbitol in the diet produced a caries score of 
2.8.
    Shyu and Hsu (Ref. 59) evaluated the cariogenicity of 10-percent 
xylitol, mannitol, sorbitol, and sucrose in rats fed a plain basal 
diet. A control group was fed the basal diet without sweetener. Caries 
evaluations were made on the 45th and 90th days of feeding. The xylitol 
group had 86 percent fewer caries (significant) compared to the sucrose 
group and 76 percent fewer caries than the control. The mannitol group 
experienced 70 and 51 percent fewer caries than the sucrose and control 
groups, respectively. The sorbitol group experienced 48 and 14 percent 
fewer caries than the sucrose and control groups, respectively.
    Bramstedt et al. (Ref. 60) evaluated the cariogenicity of isomalt, 
xylitol, and sucrose in 60 rats divided into five groups. The control 
diet was a basic diet containing half synthetic feed. Another control 
group received a special basic diet containing no low molecular weight 
carbohydrates. The test groups received the basic diet with increasing 
doses of sweetener up to 30 percent of the diet. The sucrose group had 
a significantly higher number of caries than either of the sugar 
alcohol groups. The group consuming the special basic diet had the 
lowest incidence of caries. There were no significant differences in 
the number of caries between the basic diet, xylitol, and isomalt 
groups, although the isomalt group showed a slightly higher incidence 
of caries.
    Izumiya et al. (Ref. 61) fed rats 10 or 20 percent by weight of 
sweeteners in 

[[Page 37518]]
feed. Rats consuming a dietary feed containing 10-percent maltitol had 
significantly fewer caries than the sucrose group. Details of this 
study and the results were not given in this reference.
    Gehring and Karle (Ref. 62) evaluated the cariogenic properties of 
isomalt, in comparison to those of sucrose and xylitol in the basal 
diet of conventional and gnotobiotic (i.e., specially reared laboratory 
animals in which the microflora are specifically known) rats. The final 
concentration of sweetener in the feed was 30 percent. A second 
experiment was performed using isomalt, xylitol, sorbitol, and sucrose 
in chocolate. The basal diet constituted 40 percent of the total diet, 
and the chocolate constituted 60 percent. The isomalt group had 
significantly fewer caries than the sucrose group, and the xylitol 
group had significantly fewer caries than the isomalt group. The second 
experiment showed significant differences in caries experience after 
the T (initial caries lesions) and B (advanced caries) stages between 
the sucrose and sorbitol chocolate groups, the sorbitol and isomalt 
chocolate groups, and also between the isomalt and xylitol chocolate 
group. The order of cariogenicity of the test substances was sucrose 
greater than (>) sorbitol > isomalt > xylitol > control. An in vitro 
microbiological experiment was performed to test acid production 
capacity of plaque microorganisms in 10 percent solutions of isomalt, 
glucopyranosido mannitol (GPM), glucopyranosido sorbitol (GPS), 
sorbitol, mannitol, sucrose, and fructose. GPS and GPM are the two 
components that make up isomalt. Sucrose produced acid rapidly and had 
the greatest acid formation. Sorbitol and mannitol produced acid 
slowly, and isomalt and its two components had practically no acid 
production in vitro.
    Karle and Gehring (Ref. 63) evaluated the cariogenicity of isomalt 
in rats. Six groups of rats received the basic diet without low 
molecular weight carbohydrates in addition to xylitol, sorbose, 
isomalt, lactose, and sucrose. The control group received only the 
basic diet. Sweetener concentrations were increased slowly up to 30 
percent by weight of the basic feed. The highest number of fissure 
caries were caused by sucrose (about 33) followed by lactose (25), 
isomalt (about 13), sorbose (about 12), xylitol (about 7) and the 
control (5). Differences in caries incidence between the sucrose and 
the other groups were significant.
    Larje and Larson (Ref. 64) fed rats a caries diet, diet 2000, to 
which various sweeteners were added. The caries diet, containing 56 
percent sucrose, was used as a control ration. Sucrose substitutes used 
in at least one of the experiments included glucose, fructose, 
mannitol, sorbitol, potato starch, starch/sucrose mixtures, or HPS 
(contains sorbitol and hydrogenated dextrins). In the first experiment 
each group was fed diet 2000 for a few days, then they were changed to 
one of the diets containing a sucrose substitute. Each test diet was 
fed for 7 out of every 14 days followed by rotation back to the control 
diet. The diets were changed every 2 or 3 days according to a 
predetermined schedule. A second experiment was designed to determine 
the effect of feeding the sucrose diet after the period of bacterial 
implantation on diets containing sucrose substitutes. The animals 
consumed one of the test diets the first week while being inoculated 
with S. mutans, followed in the final 7 wk by the control diet 
containing sucrose. A third experiment was designed to determine the 
effect of feeding sucrose and sucrose-substitute diets intermittently 
after the period of bacterial implantation on the sucrose diet. The 
animals consumed diet 2000 the first wk, followed in the final 7 wk by 
diets containing the sugar substitutes.
    The results of the first experiment showed significantly (p<0.001) 
fewer smooth surface caries with all sugar alcohols, potato starch, 
dextrose, and hydrogenated starch compared to the sucrose group. 
Significantly (p<0.05) fewer sulcal caries were experienced in the 
groups receiving mannitol, sorbitol plus starch, potato starch, and HPS 
compared to the sucrose group. The authors observed that in all of the 
experiments, every group in which sucrose was restricted, whether by 
dietary substitution or by shortened feeding periods, developed 
significantly fewer caries on smooth surfaces compared to the sucrose 
control animals. The animals in the mannitol, sorbitol plus starch, and 
sorbitol groups consumed less food during the test period compared to 
the sucrose controls. The authors stated that food consumption and 
weight gains were directly related to the incidence of caries.
    The results of experiment 2 showed significantly (p<0.001) fewer 
smooth surface caries in groups fed hydrogenated starch, potato starch, 
dextrose, fructose, sorbitol plus starch, dextrose plus fructose 
compared to the sucrose group. Groups receiving HPS, fructose, and 
sorbitol plus starch experienced significantly (p<0.001) fewer sulcal 
caries compared to the sucrose group.
    The results of experiment 3 showed significantly (p<0.001) fewer 
smooth surface caries in groups receiving potato starch, fructose, 
sorbitol plus starch, dextrose plus fructose, dextrose, and 
hydrogenated starch compared to the sucrose group. The overall results 
showed that reducing the exposure to sucrose results in fewer carious 
lesions.
    Muhlemann (Ref. 65) tested the effects of topical applications of 
sugar substitutes on caries incidence and bacterial agglomerate 
formation in rats receiving a cariogenic diet containing 20-percent 
sucrose. Sweeteners tested (50 percent w/v) included the following: 
Sucrose, mannitol, GPS, GPM, isomalt, sorbitol, maltitol, and French 
HSH. Three control groups were used: (1) One group received the 
cariogenic diet (20-percent sucrose) and no topical applications, (2) 
the second group received a topical application of water with the 
cariogenic diet, and (3) the third group was treated topically with 
chlorhexidin digluconate (0.5 percent) as a positive control. Topical 
solutions were applied five times a day for 23 days.
    Among the carbohydrates treatments, the isomalt, GPS, and GPM 
groups had the lowest incidence of fissure and smooth surface caries. 
The differences, however, between the caries incidence in these three 
groups and the other test groups were not statistically significant. 
The incidence of caries in the chlorhexidine control group was 
statistically significantly lower than all treatment groups. The 
control groups receiving no application and water both experienced 
slightly more caries than the sugar alcohol groups. Results of these 
studies suggest that in the presence of a cariogenic diet, topical 
application of mannitol, isomalt, sorbitol, maltitol, or HSH does not 
affect the promotion by sucrose of dental caries in rats.
    Ooshima et al. (Ref. 66) evaluated the cariogenicity of maltitol in 
rats infected with S. mutans. Animals were divided into 12 groups. 
Group A received a control diet containing 56-percent wheat flour. 
Groups B through L received the same diet as the control group but had 
portions of the wheat flour replaced with one of the test substances. 
The sweeteners tested were as follows: 10-percent maltitol plus 46-
percent wheat flour (group B), 20-percent maltitol plus 36-percent 
wheat flour (group C), 10-percent sucrose plus 46-percent wheat flour 
(group D), 10-percent sucrose plus 10-percent maltitol plus 36-percent 
wheat flour (group E), 20-percent sucrose plus 36-percent wheat flour 
(group F), 20-percent sucrose plus 20-percent maltitol plus 

[[Page 37519]]
16-percent wheat flour (group G), 24-percent sucrose plus 32-percent 
wheat flour (group H), 24-percent sucrose plus 16-percent maltitol plus 
16-percent wheat flour (group I), 28-percent sucrose plus 28-percent 
wheat flour (group J), 28-percent sucrose plus 12-percent maltitol plus 
16-percent wheat flour (group K), or 40-percent sucrose plus 12-percent 
wheat flour (group L).
    The results of this study showed that the maltitol did not induce 
dental caries in groups B and C compared to the wheat flour alone 
(group A). Groups A, B, and C experienced significantly (p<0.001) fewer 
caries than the sucrose group (group L). Groups D through I and K 
reported significantly (p<0.001 and p<0.01, respectively) fewer caries 
than group L. There was no significant difference in caries score 
between group J (equal parts sucrose and wheat flour) and group L. 
Thus, this study suggests that replacing sucrose with less cariogenic 
sweeteners or wheat flour results in fewer dental caries in rats.
    Tate et. al. (Ref. 67) reported on the correlations between 
progressive caries and sugar intake in hamsters inoculated with S. 
mutans. Animals were fed a diet with 10-percent sucrose (group 1), 20-
percent sucrose (group 2), 10-percent sucrose plus 10-percent maltitol 
(group 3), 10-percent sucrose plus 10-percent coupling sugar (group 4), 
10-percent maltitol (group 5), or 10-percent coupling sugar (group 6). 
Group 2 experienced the most caries. There was no significant 
difference in caries score between group 1 and groups 3 and 4. Groups 5 
and 6 had significantly (p<0.01) fewer caries than groups 1 or 2. This 
reference did not provide sufficient details regarding the methodology 
and analysis of results for purposes of evaluating the weight of the 
results.
    Leach and Green (Ref. 68) fed two groups of rats a basal diet 
supplemented with sucrose plus 3-percent xylitol or 6-percent xylitol. 
The control group consumed the basal diet with sucrose. In experiment 
1, rats were continuously fed the same diet during the experimental 
period. In experiment 2, rats were fed diets alternating between the 
control diet one day and the test diet the next day. In experiment 1, 
rats fed the sucrose and 6-percent xylitol mixture had significantly 
(p<0.02) fewer fissure caries than the control. There were no 
significant differences in the xylitol mixture groups. In experiment 2, 
both xylitol mixture diet groups had significantly (p<0.001) fewer 
fissure caries than the control. There were no significant differences 
among the xylitol mixture groups.
    Mukasa (Ref. 69) evaluated the cariogenicity of maltitol and SE58 
in rats. Product SE58 is a highly purified corn starch treated with 
enzyme and hydrogenated. It contains 20- to 25-percent sorbitol, 20- to 
30-percent maltitol, 15- to 25-percent maltotrititol, and 30- to 40-
percent maltopentaitol. In experiment one, three groups of rats were 
fed diet 2000 containing either 56-percent sucrose, maltitol, or SE58, 
among other ingredients. Because the rats consuming the maltitol and 
SE58 diets experienced serious growth problems, experiment one was 
discontinued. In experiment two, the level of all sweeteners in diet 
2000 was reduced to 26 percent, with the remaining 30 percent as added 
corn starch. The sucrose group had a mean fissure caries score of 31.5 
and a smooth surface caries score of 14.1. The maltitol group had 3.1 
fissure caries and no smooth surface caries. The SE58 group had 4.6 
fissure caries and 0.5 smooth surface caries. Differences between the 
sucrose group and each sugar alcohol group were significant.
    Van der Hoeven (Ref. 70) evaluated the cariogenicity of isomalt in 
rats. Test diets consisted of a base diet containing 16-percent sucrose 
and 44-percent wheat flour and a base diet with 16-percent isomalt and 
44-percent wheat flour. The control diet consisted of 60-percent wheat 
flour and no added sweetener. Diets were offered ad libitum over a 
period of 14 wk. Results showed increasing incidence of dentinal 
fissure lesions in the sucrose group (wk 2 = 4; wk 14 = 14 lesions) and 
almost no caries in the isomalt group (wk 2 = 0; wk 8 = 4; wk 14 = 1 
lesion). There was no difference in the incidence of caries between the 
isomalt and the control groups.
    Van der Hoeven (Ref. 73) evaluated the cariogenicity of lactitol in 
program-fed rats. The sweetener was incorporated into a powdered diet, 
described by Havenaar et al. (Ref. 54), consisting of a basic part (50 
percent), wheat flour (25-percent), and test substance (25-percent). 
Lactitol was compared with sorbitol, xylitol, sucrose, and a control 
with wheat flour in addition to the basic part. The animals received 9 
g of diet divided into 18 portions of 0.5 g each per day. The animals 
on the xylitol and sorbitol diets were reported to experience reduced 
weight gains and a reduced appearance of the fur. None of the animals 
suffered from diarrhea.
    There were significantly fewer caries in the xylitol, lactitol, 
sorbitol, and wheat flour groups compared to the sucrose group. The 
incidence of caries in the lactitol and sorbitol groups was slightly, 
but not significantly, higher than in the wheat flour group. The 
incidence of caries was lowest in the xylitol group.
    In a twofold experiment using caries-active rats, Grenby and 
Phillips (Ref. 77) evaluated: (1) The cariogenicity of lactitol, 
sucrose, and xylitol at a level of 160 g per (/) kilogram (kg), a level 
stated to approximate the average sucrose content of the diet in 
developed countries, and (2) the cariogenicity of lactitol in a sweet 
biscuit compared to a sucrose-sweetened biscuit. In the first 
experiment, the sweetener was incorporated into a laboratory chow 
containing white flour, skim milk powder, liver powder, and a vitamin-
mineral supplement. In the second part of the experiment, biscuits, 
containing 166 g of lactitol/kg, were incorporated into the animal chow 
for a final concentration of lactitol of 110 g/kg. Animals were fed the 
diets for a period of 8 wk. Experiment 1 showed highly significant 
differences in caries score, total number of lesions, and severity of 
lesions in the sugar alcohol groups compared to the sucrose controls. 
The sugar alcohol groups had very few caries, and differences between 
groups were not significant. The animals in both the xylitol and 
lactitol groups required several weeks to adapt to the diets, showing 
increased water intake and decreased food intake. Because of poor 
physical condition, only 11 of the 22 rats in the xylitol group 
completed the full 8-wk test. Animals on the sucrose diet were 
significantly heavier than the sugar alcohol animals.
    Results of the second test showed highly significant differences 
between the lactitol- and sucrose-biscuits groups in all caries 
parameters. The average caries score for the lactitol group was less 
than one per animal. Weight gains, however, were consistently lower, 
and water intake increased in the lactitol group.
    The results of the above animal studies show that animals fed sugar 
alcohols in animal chow had fewer and less extensive caries than 
animals fed sucrose. The studies also show that, in general, rats do 
not eat as much of a sugar alcohol-containing diet as a sucrose-
containing diet and, therefore, tend to gain less weight and have more 
physiological problems.

E. Summary of Human and Animal Studies

1. Xylitol
    In its 1978 review of the studies on xylitol, FASEB concluded that 
xylitol appeared to be noncariogenic in studies evaluating the effect 
of sucrose 

[[Page 37520]]
replacement with xylitol and in studies evaluating the effect of 
partial replacement of sucrose with xylitol in chewing gum (Ref. 14). 
However, FASEB concluded that it was essential that these studies be 
replicated by other workers in order to confirm the observations and 
conclusions.
    Rekola (Refs. 23 and 25) conducted a followup assessment of results 
from the 2-yr Turku sugar study evaluating the progression of incipient 
carious lesions and lesion sizes on buccal smooth surfaces with dietary 
substitution of xylitol for sucrose. In the 2-yr Turku sugar study, 
dietary xylitol was almost completely substituted for sucrose. Subjects 
were assigned to groups based on individual preference. Rekola examined 
color dental photographs, taken during the 2-yr study, of 33 subjects 
in the sucrose group and 47 subjects in the xylitol group. The xylitol 
group showed significantly smaller white spot lesions and had a 
significantly lower caries score compared to the sucrose group.
    Results of several more recent human caries studies (Refs. 22, 26, 
and 28 through 30) reported significantly fewer caries in the xylitol 
group compared to the sucrose group. Kandelman and Gagnon (Ref. 22) 
reported significantly less NPD and incidence of DMFT in school 
children chewing three sticks per day of xylitol gum (3.4 g) or xylitol 
and sorbitol gum (0.9 g xylitol and 2.4 g sorbitol) compared to the 
nongum control group. Results of xylitol field studies in Hungary 
(Refs. 26 and 28), French Polynesia (Refs. 29 and 30), and Thailand 
(Ref. 29) conducted by WHO showed lower caries incidence and caries 
increment rate in children consuming xylitol and sorbitol in chewing 
gum (Ref. 29) and xylitol in other snack foods (Ref. 30) compared to a 
nonsugar alcohol group. However, results of the gum study in French 
Polynesia and Thailand (Ref. 29) were confounded by the presence of 
fluoride in the gums tested. In addition, the prevalence and incidence 
of dental caries in these population groups were high and increasing 
and do not reflect the general healthy population of the United States.
    The effect of xylitol on acid production or plaque pH was studied 
in ten studies (Refs. 38, 39, 41, 43 through 46, 48, 50, and 76). In 
nine of these (Refs. 38, 39, 41, 43 through 46, 48, and 50), xylitol 
was found to result in negligible to no acid production with little to 
no change in plaque pH. Similarly, results showed no significant effect 
of xylitol on resting plaque pH. Plaque pH from exposure to xylitol was 
always significantly higher than that of sucrose or glucose.
    Twelve animal studies (Refs. 52, 54, 56 through 60, 62, 63, 68, 73, 
and 77) evaluated the effects of xylitol on dental caries in rats or 
hamsters. Eight of these (Refs. 52, 57 through 60, 62, 63, and 77) used 
a test diet that contained only one sweetener, either sucrose or 
xylitol. In all of these studies, there were significantly fewer caries 
reported in animals consuming the basal diet with xylitol compared to 
sucrose controls. The incidence of caries was also significantly less 
in the xylitol group compared to animals consuming isomalt (Ref. 63) 
and sorbitol (Ref. 52). The concentrations of xylitol in the test diets 
ranged from 10 percent up to 30 percent by weight.
    Results of the animal studies evaluating the effect of xylitol in 
diets containing sucrose (Refs. 54, 56, 68, and 73) showed mixed 
results depending on the concentrations of sucrose and xylitol in the 
test diets. Havenaar et al. (Ref 54) showed no significant difference 
in caries in animals consuming a diet with sucrose and 5-percent 
xylitol, but a significant difference in caries when the sucrose was 
lowered to 20-percent of the diet and xylitol 5-percent. Grenby and 
Colley (Ref. 56) reported a high caries level in animals consuming 
either a diet containing 46-percent sucrose or 36-percent sucrose and 
10-percent xylitol. The caries score was significantly lower in rats 
consuming a diet with 26-percent sucrose and 20-percent xylitol 
compared to the 46-percent sucrose diet. An in vitro microbiological 
test showed no acid production by S. mutans from xylitol. Van der 
Hoeven (Ref. 73) reported significantly fewer caries in rats consuming 
a diet with 25-percent xylitol compared to the rats consuming a basic 
diet with 25-percent sucrose. The xylitol group also had fewer caries 
than the wheat flour control group.
2. Sorbitol
    In its March 1979, review of sorbitol in health and disease (Ref. 
15), FASEB reviewed available animal and human studies regarding the 
cariogenicity of sorbitol. FASEB concluded that the weight of evidence 
from animal studies suggests that sorbitol is less cariogenic than 
sucrose, fructose, glucose, and dextrin. Based on the human studies 
published in the early to mid-1970's, FASEB noted that the results do 
not provide definitive data on the effect of sorbitol on the caries 
process. It noted that the results of studies on plaque pH suggest that 
sorbitol is slowly fermented to plaque pH levels of about 6. It also 
said that some studies have provided evidence of adaptation of oral 
flora after long-term use of sorbitol-containing products. FASEB noted 
that a human population that regularly consumes sorbitol-containing 
foods, such as jams and jellies, baked goods, or other food products, 
has not been identified and studied to establish whether sorbitol 
significantly alters the carious process.
    Two studies submitted with the petition evaluated the cariogenicity 
of sorbitol in chewing gum (Refs. 20 and 32), and one study (Ref. 35) 
evaluated the effect of sorbitol in chewing gum on demineralization of 
enamel. Moller and Poulsen (Ref. 20) reported an increased number of 
sound tooth surfaces and a smaller caries increment rate in children 
consuming sorbitol gum containing calcium phosphate compared to the 
control group that did not consume chewing gum. However, the presence 
of calcium phosphate, which acts as a buffer in saliva to help reduce 
its acidity, and the absence of gum chewing in the control group, 
confound these observations.
    Glass (Ref. 32) reported no significant differences in the number 
of DF surfaces or teeth in children using sorbitol chewing gum for 2 yr 
compared to a no-gum group. This study, however, did not consider the 
effect of chewing gum per se on dental caries.
    Leach et al. (Ref. 35) conducted an intraoral test in subjects 
fitted with bands containing human enamel with artificial white spot 
lesions. The subjects consumed sucrose-containing snacks. During one of 
the test periods, the subjects chewed gum containing sorbitol with 
small amounts of mannitol, HGS, and aspartame, for 20 min at a time 
after each meal and snack. The study showed significantly more 
remineralization during the sorbitol gum period compared to baseline 
and the no-gum (sucrose) period. Results of this study are confounded, 
however, because of the duration (i.e., 20 min) and timing (i.e., 
immediately after meals and snacks) of the gum chewing. In addition, 
the effect of sorbitol alone cannot be determined because of the 
presence of other sugar alcohols and aspartame in the test gum.
    Banoczy et al. (Ref. 21) reported a significantly lower caries 
increment in children consuming sorbitol-containing sweets between 
meals compared to children consuming sucrose-containing sweets between 
meals over a 2-yr period. Differences between groups were not 
significant during the third yr of this study, however, the authors 
attributed the lack of significance during the third yr to the trading 
of sweets between groups. 

[[Page 37521]]

    Twelve studies evaluated changes in plaque pH after exposure to 
sorbitol-sweetened mouth rinses (Refs. 39 through 41, 45, and 47), 
solutions (Refs. 38, 46, and 76), tablets (Ref. 42), mints (Ref. 49), 
chewing gum (Ref. 50), and licorice (Ref. 43). Plaque pH changes in the 
presence of sorbitol decreased from baseline pH but remained 
approximately at or above a pH of 6.0 (Refs. 39 through 42, 45 through 
47, and 50). Bibby and Fu (Ref. 38) reported progressively decreasing 
plaque pH values in vitro with increasing concentrations of sorbitol in 
a concentrated plaque suspension. Only slight decreases in pH were 
reported in 0.1- to 1.0-percent solutions. In the presence of a 10-
percent sorbitol solution, plaque pH dropped to about 5.8. Grenby et 
al. (Ref. 76) reported a pH of about 6.0 after 12 h and a final pH in 
vitro of about 4.6 after 24 h of incubating concentrated plaque with 
10-percent sorbitol. The results of these studies suggest that higher 
concentrations of sorbitol may lead to further decreases in plaque pH 
to a level that may become detrimental to tooth enamel (i.e., at or 
below pH 5.5).
    Park et al. (Ref. 49) found that use of sorbitol mints or mints 
with a blend of sorbitol and xylitol helped reduce the acidogenic 
potential of certain snack foods, although final pH values remained 
low. Toors and Herczog (Ref. 43) showed that plaque pH is affected by 
more than the sweetener component of a food. Results of plaque pH in 
vivo with an experimental licorice, containing sorbitol, soy flour, and 
potato starch derivative among other ingredients, showed a minimum pH 
of about 5.5. A sucrose-containing licorice used in this study lowered 
plaque pH to about 5.0. The fermentability of both the potato starch 
derivative (82 percent) and soy flour (75 percent) contributed to the 
observed changes in plaque pH in the experimental licorice. The 
fermentability of sorbitol in the experimental licorice was 12 percent.
    Five studies (Refs. 39 through 41, 43, and 48) measured the APA of 
plaque with sorbitol. In all cases, sorbitol was fermented slowly with 
a reported range of acid production of 10 to 30 percent compared to 
sucrose or glucose. The higher acid production rate (i.e., 30 percent) 
was attributed to adaptation to sorbitol by S. mutans and other plaque 
microorganisms capable of fermenting carbohydrates. Havenaar et al. 
(Ref. 46) also reported a marked increase in fermentation of sorbitol 
and other sugar alcohols after multiple subculturing of plaque 
microorganisms with the sugar alcohol. However, the investigators 
reported that adaptation to sorbitol and other sugar alcohols was lost 
after subculturing once in glucose.
    Results of animal studies evaluating sorbitol (Refs. 35, 52, 58, 
59, 62, 64, and 73) showed significantly fewer caries in the sorbitol 
group than in the sucrose group. However, use of sorbitol resulted in 
more caries compared to animals consuming other sugar alcohols, such as 
xylitol and HSH (Refs. 52, 64, and 73). The concentration of sorbitol 
in these studies ranged from 10 percent up to 56 percent.
3. Mannitol
    In its August 1979, review of mannitol in health and disease, FASEB 
(Ref. 16) reviewed available animal and human studies regarding the 
effect of mannitol on acid production, plaque pH changes, and changes 
in microhardness of bovine enamel in an ICT. It noted that human plaque 
studies in vivo or in vitro found that plaque pH decreases from 0 up to 
1.0 units over a 30-min test period. FASEB concluded that the results 
were consistent with the results of animal experiments showing that 
mannitol, in the absence of adaptation of the oral microflora, is less 
cariogenic than sucrose.
    Bibby and Fu (Ref. 38) measured in vitro plaque pH changes, over a 
20-min incubation period, in the presence of increasing concentrations 
of mannitol (0.1-, 1.0-, and 10-percent concentrations) in a 
concentrated plaque suspension. Results showed that plaque pH decreased 
with increasing concentrations of mannitol. Final plaque pH values were 
5.67, 5.54, and 5.22, respectively. Similar plaque pH values were 
reported by Grenby et. al. (Ref. 76). Results of the Grenby study 
showed that a 1-percent solution of mannitol, when incubated for 24 h 
with concentrated plaque and pieces of a human molar tooth, resulted in 
slight acid production and pH decrease over a 12-h period, but that 
after 24 h, the final pH was about 5.1. However, results from an in 
vitro demineralization test showed very little loss of calcium and 
phosphorus, significantly less than the loss of minerals with glucose.
    Results of other studies, however, show that mannitol results in 
little change to plaque pH. Birkhed and Edwardsson (Ref. 39) reported 
only slight changes in plaque pH following use of a mouth rinse with a 
concentrated solution of mannitol. In addition, they reported an acid 
production rate from mannitol in dental plaque suspension of 0 percent 
compared to sucrose (100 percent). Gehring and Hufnagel (Ref. 45) used 
intraoral measurements to evaluate the effect of sugar alcohols on 
plaque pH. Results of plaque exposed to a 20-percent mannitol solution 
showed the minimum pH obtained was slightly above 6.0. The plaque 
samples in these two studies were not concentrated as they were in the 
study by Bibby and Fu (Ref. 38) or by Grenby et al. (Ref. 76), which 
may account for the differences in plaque pH values reported for 
mannitol solutions. The results of one other in vitro microbiological 
study, with 10-percent mannitol and an incubation time of 48 h (Ref. 
62), support the observation that mannitol is fermented very slowly, 
resulting in little acid production and small pH changes.
    Animals fed mannitol (Refs. 59 and 64) or maltitol (Refs. 66, 67, 
and 69) showed significantly fewer caries compared to animals fed 
sucrose diets. The concentrations of the sugar alcohols in these 
studies ranged from 10 to 56 percent. An in vitro microbiological study 
(Ref. 62) showed that a 10-percent solution of mannitol was fermented 
very slowly.
4. Maltitol
    Three studies (Refs. 33, 34, and 36) measured the effects on enamel 
decalcification of maltitol and sucrose solutions using an ICT with 
bovine enamel fragments adhered to a partial denture. Ikeda and 
coworkers (Ref. 33) showed significantly more decalcification in the 
presence of sucrose as compared to maltitol. Additional rat caries 
tests were in agreement with the results of the ICT. Rats fed a diet 
with maltitol had significantly fewer caries than the sucrose group. In 
this study maltitol was almost noncariogenic. Yagi (Ref. 34) reported 
significantly harder enamel after exposure to maltitol than after 
exposure to sucrose. Lack of details in this study, however, make it 
difficult to completely interpret the results. Rundegren (Ref. 36) 
reported significantly less enamel demineralization with maltitol 
compared to sucrose. The authors associated the changes that they 
observed in enamel hardness in the maltitol group with the effects of 
other dietary carbohydrates and not maltitol. Sucrose was found to 
exert an effect on enamel hardness that is not related to the effects 
of other dietary carbohydrates.
    Three studies (Refs. 39, 41, and 46) evaluated plaque pH or acid 
production in maltitol. Birkhed and Edwardsson (Ref. 39) measured in 
vitro acid production and pH changes in human dental plaque following 
the use of various sweeteners in a mouth rinse. The results with 
maltitol showed an acid production rate of 10 to 30 percent 

[[Page 37522]]
of that of sucrose. Changes in plaque pH in the presence of maltitol 
showed only a slight decrease from baseline pH (about pH 6.9).
    Birkhed et al. (Ref. 41) measured in vivo pH changes in human 
dental plaque after subjects consumed lozenges sweetened with various 
sweeteners for 3 mo and then rinsed with a mouth rinse sweetened with 
the same sweetener as in the lozenge. A sucrose mouth rinse was also 
used by each sweetener group. Results with maltitol showed small, but 
some significant, changes in plaque pH compared to baseline pH (about 
pH 7.0) over the 30-min test period. The lowest plaque pH recorded, 
however, was about pH 6.8. In vitro acid production with maltitol was 
found to be about 26 to 32 percent of glucose.
    Havenaar et al. (Ref. 46) measured changes in pH and acid 
production in vitro in growing cultures of oral bacteria obtained from 
caries active and caries free subjects. Results showed that a 1 percent 
solution of maltitol was slowly fermented to acid by plaque bacteria. 
Cell suspensions of S. mutans in maltitol showed pH decreased from a 
baseline of about pH 7.0 to about pH 6.5. Adaptation of S. mutans by 
frequent subculturing in maltitol showed a marked increase in 
fermentation by S. mutans. However, the ability to ferment the sugar 
alcohol was lost after one subculturing of the adapted strain in 
glucose.
5. Lactitol
    Havenaar et al. (Ref. 46) showed that a 1 percent solution of 
lactitol was fermented by S. mutans and Actinomyces. Cell suspensions 
of S. mutans in lactitol showed pH decreased from a baseline of about 
pH 7.0 to about pH 6.5 or above after a 2-h incubation period. 
Adaptation of S. mutans by frequent subculturing in lactitol showed a 
marked increase in fermentation by S. mutans to give a plaque pH of 
about 5.0. However, the ability to ferment the sugar alcohol was lost 
after one subculturing of the adapted strain in glucose. Grenby et al. 
(Ref. 76) showed that a 1-percent solution of lactitol, when incubated 
for 24 h with human plaque and pieces of a human molar tooth, resulted 
in slight acid production and a final pH of about 6.3 and almost no 
loss of calcium and phosphorus from tooth enamel.
    Results of two animal studies (Refs. 73 and 77) showed that 
substitution of lactitol for sucrose in laboratory chow resulted in 
significantly fewer caries in the lactitol group compared to the 
sucrose group. The lactitol group (Ref. 73) experienced slightly, but 
not significantly, more caries than the xylitol group and the wheat 
flour control group and fewer caries than the sorbitol group. There was 
no significant difference between the caries score in animals fed 
lactitol-containing or xylitol-containing chow (Ref. 77). There were 
significantly fewer caries in animals fed lactitol-containing biscuits 
compared to the sucrose biscuit group (Ref. 77). The average caries 
score in the lactitol biscuit group was less than one per animal.
6. Isomalt
    Two studies investigated the effects on plaque pH with isomalt 
(Refs. 38 and 45). Bibby and Fu (Ref. 38) measured pH changes in fresh 
plaque from adult volunteers with increasing concentrations of isomalt. 
Results showed that as the concentration of the sugar alcohol 
increased, the pH of the plaque decreased. The range of plaque pH 
values reported for isomalt was from 6.6 (0.1 percent solution) to 
approximately 5.7 (10-percent solution). Gehring and Hufnagel (Ref. 45) 
reported a minimum plaque pH of about 6.0 after 5 min with isomalt. 
This value increased gradually over the next 27 min to about pH 6.3. As 
discussed above, the methods and type of dental plaque must be 
considered when comparing the results of these studies.
    Results of animal studies with concentrations of isomalt from 16 to 
30 percent of the rat diet showed significantly fewer caries compared 
to sucrose diets (Refs. 57, 60, 62, 63, 65, and 70). The caries 
incidence was high in xerostomized rats consuming either sucrose or 
isomalt (Ref. 57). The isomalt group of nonxerostomized rats, however, 
had significantly fewer caries than the sucrose group.
7. HGS and HSH
    Frostell et al. (Ref. 31) studied the effect on caries increment in 
children of substitution of HSH for sucrose in candy. The results of 
this study are confounded for a number of reasons (see Table 2) and do 
not support a significant dental benefit from the use of HSH candies in 
place of sucrose-containing candies.
    Rundegren et al. (Ref. 36) measured enamel hardness in the presence 
of sucrose, sodium chloride, or HSH using an ICT. The investigators 
reported significantly less enamel demineralization with HSH. The 
results of the study were that only sucrose promoted demineralization 
over and above the effect of dietary carbohydrates. The authors 
attributed the demineralization measured in the presence of HSH to the 
effect of dietary carbohydrates.
    Eight studies measured plaque pH changes from exposure to HSH in 
solutions (Refs. 38 and 46), rinses (Refs. 39, 41, 45, and 47), and 
candy (Refs. 42 and 43). Bibby and Fu (Ref. 38) showed that as the 
concentration of HSH increased, plaque pH decreased. The lowest plaque 
pH value (10-percent solution of HSH) obtained was about 5.0. Havenaar 
et al. (Ref. 46) showed that a 1-percent solution of HSH was fermented 
by S. mutans and Actinomyces. Cell suspensions of S. mutans in HSH 
showed a pH decrease from a baseline of about pH 7.0 to about pH 6.5. 
Adaptation of S. mutans by frequent subculturing in HSH showed a marked 
increase in fermentation by S. mutans to give a plaque pH of slightly 
below 6.0. However, the ability to ferment the sugar alcohol was lost 
after one subculturing of the adapted strain in glucose.
    Birkhed and Edwardsson (Ref. 39) measured plaque pH in vitro 
following the use of a mouth rinse containing Swedish or French HSH. 
French HSH appeared to have little effect on plaque pH. Plaque pH 
values remained slightly below or at 7.0. Swedish HSH showed a decrease 
in plaque pH within 5 to 10 min to just less than pH 6.0. Over the 
remaining 20 min, the pH increased to just over 6.0. Birkhed et al. 
(Ref. 41) measured pH changes in human dental plaque after subjects 
consumed lozenges sweetened with Swedish HSH for 3 mo and then rinsed 
with a mouth rinse sweetened with Swedish HSH. Plaque pH was also 
measured after a sucrose mouth rinse. The results of the study showed 
that HSH resulted in a drop in plaque pH in all tests; however, the 
minimum pH values reached were above 6.0. Gehring and Hufnagel (Ref. 
45) reported an intraoral plaque pH change with a HSH rinse (20 percent 
solution) from about pH 6.6 to about 5.6.
    Jensen (Ref. 47) showed interproximal plaque pH values from five 
different HGS rinses were statistically significantly different 
compared to the sucrose control. Differences between the HGS test 
solutions and a sorbitol control were not significantly different. The 
minimum pH values obtained with the HGS solutions were above pH 6.0. 
Composition of the HGS test substances was not provided.
    Frostell (Ref. 42) reported a slight decrease in vitro plaque pH 
(from about 6.7 to about 6.5) after subjects consumed HSH candy. After 
consuming a sucrose lozenge, plaque pH decreased to about 5.8. A 
sucrose solution resulted in a minimum plaque pH of about 5.3. Toors 
and Herczog (Ref. 43) showed that plaque pH is affected by more than 
the sweetener component of a food. Results 

[[Page 37523]]
of plaque pH in vivo with an experimental licorice, containing soy 
flour, HPS, and potato starch derivative among other ingredients, 
showed a minimum pH of about 5.5. The fermentability of the HPS (60 
percent), potato starch derivative (82 percent) and soy flour (75 
percent) contributed to the observed changes in plaque pH in the 
experimental licorice.
    Acid production in vitro was reported in two studies (Refs. 39 and 
51). Birkhed and Edwardsson (Ref. 39) reported an acid production rate 
from French HSH of 20 to 40 percent and from Swedish HSH of 50 to 70 
percent compared to glucose syrups. Birkhed and Skude (Ref. 51) 
reported significantly lower acid production rates (i.e., slower rate 
of fermentation) from a 3 percent solution of Swedish HSH (61.5 
percent) compared to glucose (99.7 percent). The investigators also 
reported that HSH was metabolized significantly more slowly than 
soluble starch.
    Results of animal studies evaluating the effect of HSH showed the 
sweetener to be relatively noncariogenic compared to sucrose (Refs. 52, 
53, 64, and 69). Differences in the incidence of caries between the 
sucrose and HSH groups were significant.
IV. Decision To Propose a Health Claim Relating Sugar Alcohols To the 
Nonpromotion of Dental Caries

    FDA limited its review of the scientific evidence relating sugar 
alcohols and dental caries to those studies evaluating changes in 
plaque pH, plaque acid production, decalcification or remineralization 
of tooth enamel, and the incidence of dental caries with sugar 
alcohols. FDA considered these limitations to be appropriate because 
previous Federal government and other authoritative reviews had focused 
on these areas (Refs. 14 through 16), and the majority of research 
efforts to date have focused on these areas.
    FDA tentatively concludes that, based on the totality of publicly 
available scientific evidence regarding the relationship among sugar 
alcohols, plaque pH, and dental caries, there is significant scientific 
agreement to support the relationship between the use of xylitol, 
sorbitol, mannitol, maltitol, isomalt, lactitol, HSH, HGS, or a 
combination of these sugar alcohols and the nonpromotion of dental 
caries. Thus, it appears that use of a health claim relating the use of 
sugar-alcohol containing products to dental caries will be useful in 
helping consumers identify food products consumption of which will not 
promote the development of dental caries.

A. Xylitol

    In its 1978 review of the xylitol studies, FASEB concluded that 
xylitol appeared to be noncariogenic in studies evaluating the effect 
of sucrose replacement with xylitol and in studies evaluating the 
effect of partial replacement of sucrose with xylitol in chewing gum 
(Ref. 14).
    The agency reviewed over 15 studies published since the FASEB 
report that evaluated the relationship between xylitol and dental 
caries, plaque pH, and acid production. Overall results from the human 
caries field trials (Refs. 26 and 28) suggest that substitution of 
xylitol-containing foods and chewing gums for sucrose-containing foods 
and chewing gums is associated with a lower incidence of dental caries. 
Plaque pH and acid production studies further support this result. In 
both in vivo and in vitro studies, xylitol had negligible to no effect 
on plaque pH or plaque acid production. In some instances, xylitol 
increased plaque pH above the mean baseline value, suggesting that 
xylitol may truly be nonpromotional of dental caries. The results of 
over 10 animal studies confirm the observations from clinical and in 
vitro studies. Substituting xylitol (from 10 to 30 percent) for sucrose 
in a basic laboratory chow resulted in significantly fewer dental 
caries. FDA tentatively concludes that the overall results from human 
and animal studies strongly support the observation that xylitol does 
not promote acid production in plaque and, therefore, does not promote 
dental caries.

B. Sorbitol

    In its 1979 report on sorbitol, FASEB concluded that the weight of 
evidence from animal studies suggests that sorbitol is less cariogenic 
than sucrose and other fermentable sugars (Ref. 15). The report noted 
that the results of human plaque studies show that sorbitol does not 
lower plaque pH below 5.5, the pH of plaque where decalcification may 
begin. FASEB concluded that it could be assumed that sorbitol may have 
similar relative cariogenic properties in humans as observed in 
animals.
    The agency reviewed over 10 clinical studies with sorbitol 
published since the FASEB report. Subjects consuming sorbitol-
containing sweets between meals experienced fewer dental caries than 
those consuming sucrose-containing sweets. Plaque pH and acid 
production studies consistently show that sorbitol is slowly fermented 
by plaque microflora and by S. mutans in particular. However, results 
show that plaque acid did not decrease pH to levels associated with 
incipient enamel decalcification (i.e., approximately at pH 5.5 or 
below). There is some evidence that suggests that long-term, 
uninterrupted use of sorbitol results in adaptation by S. mutans and 
other plaque microorganisms and, therefore, in more acid production. 
However, there are no human caries trials to show whether such 
adaptation results in a change in the incidence of dental caries. There 
is some evidence to show that adaptation may be lost in the presence of 
other sugars.
    The results of six animal studies confirmed the observations from 
human studies. The incidence of caries in animals consuming diets 
containing sorbitol was significantly less than the caries incidence in 
animals consuming diets containing sucrose. FDA tentatively concludes 
that the overall results from human and animal studies show that oral 
bacteria cannot be sustained in the presence of sorbitol, and that 
changes in acidity are within a range that is safe for tooth enamel.

C. Mannitol

    In its 1979 report on mannitol, FASEB concluded that results of 
acid production, plaque pH changes, and changes in microhardness of 
bovine enamel were consistent with the results of animal experiments 
indicating that mannitol, in the absence of adaptation of the oral 
microflora, is less cariogenic than sucrose (Ref. 16). One study 
evaluated plaque pH with mannitol in a concentrated plaque suspension 
in vitro (Ref. 38). One and ten percent solutions of mannitol resulted 
in a plaque pH of 5.5 or below. Contrary to these results, however, 
three studies showed only slight acid production and small changes in 
plaque pH to a value not below pH 6.0 from mannitol (Refs. 39, 45, and 
76). Likewise, there was little evidence of demineralization from 
mannitol in vitro (Ref. 76). Two rat studies, in which mannitol was 
substituted for sucrose in animal chow, showed significantly fewer 
caries with the mannitol diet (Refs. 59 and 64). FDA tentatively 
concludes that the overall results from both human and animal studies 
support the claim that mannitol does not promote dental caries.

D. Maltitol

    Results of three ICT's showed significantly less decalcification 
with maltitol than sucrose. Additional plaque pH studies showed that 
maltitol is fermented very slowly (acid production of 10 to 30 percent) 
compared to sucrose and is associated with small plaque pH changes from 
resting baseline values. 

[[Page 37524]]
Four animal studies confirmed that maltitol was significantly less 
cariogenic than sucrose. FDA tentatively concludes that the overall 
results from both human and animal studies support the claim that 
maltitol does not promote dental caries.

E. Isomalt

    The agency reviewed two plaque pH studies evaluating the acidogenic 
potential of isomalt. Results with 10 percent isomalt showed a minimum 
in vitro plaque pH of 5.7. An intraoral test with a 20 percent solution 
of isomalt reported a minimum pH of about 6.0. Results of five animal 
studies consistently showed that isomalt was significantly less 
cariogenic than sucrose. FDA tentatively concludes that the overall 
results show that isomalt does not lower plaque pH below 5.5 and does 
not promote dental caries.

F. Lactitol

    Two in vitro plaque pH studies showed that lactitol produced little 
acid and only slight changes in plaque pH from resting baseline values. 
Results of two animal studies are consistent with these results and 
showed lactitol to be significantly less cariogenic than sucrose. The 
cariogenicity of lactitol was not significantly different than xylitol. 
FDA tentatively concludes that the overall results support the claim 
that lactitol does not promote dental caries.
G. Hydrogenated Starch Hydrolysates and Hydrogenated Glucose Syrups

    In an ICT, a solution of HSH resulted in significantly less 
demineralization than sucrose. The investigators attributed the 
observed demineralization with HSH to an effect of other dietary 
components. The effects of sucrose on enamel demineralization, however, 
were noted to be over and above the effect of other dietary components.
    Seven studies evaluating the effect of HSH on plaque pH showed 
inconsistent results in final pH values reported. The differences in 
results are attributed to the source of the HSH. HSH is manufactured by 
hydrolyzing a source of food grade starch (usually potato or corn 
starch) with acid or an enzyme to a mixture of sugars and dextrins of 
various glucose lengths (i.e., glucose syrups). The hydrogenated 
mixture contains sorbitol, maltitol, maltitriol, maltotrititol, and 
hydrogenated dextrins of various molecular weights (Ref. 79). The 
percentage of each component sugar alcohol in the final substance 
depends on the manufacturing process and controls. The two major forms 
of HSH (i.e., one manufactured in Sweden and the other in France) used 
in the studies reviewed gave dramatically different results in plaque 
pH and acid production tests. The Swedish version, which has a higher 
percentage of higher molecular weight, fermentable polysaccharides than 
the French version, produced plaque pH values of 5.5 to 6.0 and an acid 
production of 50 to 70 percent compared to sucrose. The French version 
produced final plaque pH values above 6.0 and an acid production rate 
of 20 to 40 percent of sucrose. Results with HGS of unidentified 
composition showed minimum plaque pH values all above 6.0. Results of 4 
rat studies support the observations that HSH (source not identified) 
is significantly less cariogenic than sucrose. FDA tentatively 
concludes that the overall results support the claim that HSH and HGS 
do not promote dental caries.
    Based on its review of the scientific evidence, the agency noted 
that the HSH and HGS sugar alcohol mixtures may vary in their 
acidogenic response in dental plaque. For example, HSH manufactured in 
Sweden usually gave a lower plaque pH response than the French version 
of HSH. This variation in acidogenic response has been attributed to 
the differences in the chemical composition of these substances. HSH 
and HGS are not well defined chemical substances as are xylitol and 
sorbitol. Instead, the sugar alcohol compositions of these substances 
will vary depending on the manufacturing process. Therefore, the agency 
is asking for comments on how to determine whether sugar alcohol 
mixtures, such as HSH, when used in a food whose label bears a dental 
caries health claim, are in compliance with any final rule resulting 
from this proposal.

V. Decision To Propose An Exemption From Sec. 101.14(E)(6) For Chewing 
Gum and Confectioneries

    Section 101.14(e)(6) provides, as stated above, that except for 
dietary supplements or where provided for in other regulations in part 
101, subpart E, to be eligible to bear a health claim, a food must 
contain 10 percent or more of the reference daily intake or the daily 
reference value for vitamin A, vitamin C, iron, calcium, protein, or 
fiber per reference amount customarily consumed before there is any 
nutrient addition.
    The petition states that products containing sugar alcohols often 
will not be able to satisfy the requirement of Sec. 101.14(e)(6) 
because the products utilizing sugar alcohols are largely chewing gum 
and confectioneries, none of which are a significant source of any 
nutrients. The petition states that the use of these products in lieu 
of traditional sugar-based confectionery would be consistent with 
public health recommendations, and that the health claim statement, 
``useful only in not promoting tooth decay,'' is an important and 
useful message for consumers in making decisions on which foods to 
purchase.
    FDA has tentatively determined that there is significant public 
health evidence to support providing an exemption to Sec. 101.14(e)(6) 
for sugar alcohol-containing foods, e.g., chewing gums, hard candies, 
and mints. In the Surgeon General's Report (Ref. 7), dental caries is 
recognized as an important and widespread public health problem in the 
United States. Although dental caries among children are declining, the 
overall prevalence of the condition imposes a substantial economic 
burden on American health care costs. The Surgeon General's report 
states that of the 13 leading health problems in the United States, 
dental disorders rank second in direct costs (Ref. 7).
    The role of sugars, and of sucrose in particular, in the etiology 
of dental caries is well established. Caries-producing bacteria can 
readily metabolize a range of simple sugars (e.g., sucrose, glucose, 
fructose) to acids that can demineralize teeth. The unique role of 
sucrose, however, is related to its ability to be used by S. mutans, 
the primary etiologic agent in coronal caries, and other oral bacteria 
to form extracellular polymers of glucose or fructose that adhere 
firmly to tooth surfaces (Ref. 7).
    The Surgeon General's report recommends several types of 
intervention to help reduce the risk of dental caries. The diet-related 
factors include the use of fluoridated drinking water and control of 
sugars consumption. In this regard, the Surgeon General's report 
recommends that those who are particularly vulnerable to dental caries, 
especially children, should limit their consumption and frequency of 
use of foods containing relatively high levels of sugars.
    FDA agrees that limiting the amount of sugars in the diet is one 
important approach to help reduce the risk of dental caries. Sugar 
alcohols can be used to replace dietary sugars in food by providing 
sweetness and usefulness as bulking agents. Sugar alcohol-containing 
chewing gum and confectioneries, such as hard candies and mints, are 
specifically formulated without dietary sugars. Although these foods 
have little or no nutritional value, 

[[Page 37525]]
they are an important alternative to sucrose-containing snacks. 
Therefore, FDA tentatively finds that the use of health claims on the 
label of sugar alcohol-containing products will facilitate compliance 
with dietary guidelines that recommend a reduced intake of dietary 
sugars to reduce the risk of dental caries. Moreover, the sugar alcohol 
and dental caries health claim, if authorized, will apply in large 
measure, although not entirely, to snack foods that do not play a 
fundamental role in structuring a healthy diet.
    Section 101.14(e)(6) was included in FDA's regulations to ensure 
that those foods that bear a health claim are useful in structuring a 
healthy diet. Usually usefulness in structuring a healthy diet derives 
from the vitamin, mineral, protein, or fiber content of the food. In 
this case, however, FDA tentatively finds that the replacement of 
dietary sugars with sugar alcohols will help reduce the risk of dental 
caries and thus will help to facilitate compliance with the dietary 
guidelines. In recognition of the special character of the foods 
involved, FDA tentatively concludes that it is appropriate to exempt 
these food products from Sec. 101.14(e)(6). Therefore, in new 
Sec. 101.80(c)(1), FDA is proposing to exempt sugar alcohol-containing 
food products from the provisions of paragraph 101.14(e)(6).

VI. Description And Rationale For Components Of Health Claim

A. Relationship Between Sugar Alcohols and Dental Caries

    In proposed Sec. 101.80(a), FDA describes the relationship between 
sugar alcohols and dental caries. Dental caries is a multifactorial 
disease, characterized by the demineralization of the surface of tooth 
enamel by acid-forming organisms in dental plaque. It is well 
established that the relationship between sugars consumption and dental 
caries is one of cause and effect within the multifactorial context 
(Refs. 71 and 72). The role of sucrose in the etiology of dental caries 
is related to its ability to be metabolized by oral bacteria into 
extracellular polymers that adhere firmly to the tooth surfaces, at the 
same time forming acids that can demineralize tooth enamel (Ref. 7). 
The extracellular polymers that adhere to tooth surfaces (i.e., plaque) 
facilitate the further attachment of additional plaque to teeth and the 
proliferation of bacteria. Although saliva can help neutralize plaque 
acids and influence the attachment of oral bacteria to the tooth 
surface (Ref. 7), it has limited access to the acids generated at the 
tooth surface beneath the plaque.
    Diets in the United States tend to be high in sugars. Although 
there has been a decline in the prevalence of dental caries in the 
United States, there has been no decline in the consumption of sugars. 
Furthermore, the incidence of dental caries is still widespread (Ref. 
7).
    Sugar alcohols are used as sweeteners and bulking agents to replace 
dietary sugars in foods. Because of their composition, sugar alcohols 
are not as fermentable by plaque bacteria as sucrose and are, 
therefore, less cariogenic than dietary sugars. Replacing dietary 
sugars with sugar alcohols helps to maintain dental health.
B. Significance of Sugar Alcohols in the Caries Process

    As explained in section IV of this document, based on the totality 
of the publicly available evidence, FDA has tentatively concluded that 
there is significant scientific agreement among experts qualified by 
training and experience to evaluate such claims that there is adequate 
scientific evidence to conclude that the sugar alcohols xylitol, 
sorbitol, mannitol, maltitol, isomalt, lactitol, HSH, and HGS are less 
cariogenic than sucrose and do not promote dental caries. In proposed 
Sec. 101.80(b), FDA discusses the significance of the relationship 
between sugar alcohols and dental caries.
    Sugar alcohols have been shown in human and animal studies to be 
nonfermentable (i.e., xylitol) or slowly fermentable (i.e., sorbitol, 
maltitol, mannitol, isomalt, lactitol, HSH, and HGS) by S. mutans and 
other acid-forming microorganisms in dental plaque. Human studies have 
shown a reduced rate of acid production in plaque and, in some studies, 
a reduced incidence of dental caries from the use of sugar alcohol-
containing products.

C. Nature of the Claim

    In new Sec. 101.80(c)(1), FDA is proposing that all requirements of 
Sec. 101.14 be met except, as explained above, that sugar alcohol-
containing foods are exempt from Sec. 101.14(e)(6).
    Under Sec. 101.14(d)(3), nutrition labeling in accordance with 
Sec. 101.9 must be provided on the label or labeling of any food for 
which a health claim is made. Therefore, if FDA adopts this proposed 
regulation, the labeling of the amount of sugar alcohol in a serving 
will have to be declared on the nutrition label in accordance with 
Sec. 101.9(c)(6)(iii) when a claim is made on the label or in labeling 
about sugar alcohols and dental caries.
    In new Sec. 101.80(c)(2)(i), FDA is proposing to authorize a health 
claim on the relationship between sugar alcohols and the nonpromotion 
of dental caries. This action is consistent with the agency's review of 
the scientific evidence, which showed that, although sugar alcohols are 
slowly fermented by S. mutans and can form some acid, they do not 
contribute to the promotion of dental caries.
    In new Sec. 101.80(c)(2)(i)(A), the agency is proposing to require 
that in describing the relationship between sugar alcohols and dental 
caries, the claim states ``does not promote,'' ``useful in not 
promoting,'' or ``expressly for not promoting'' dental caries. FDA 
finds that these terms accurately describe the relationship between 
sugar alcohol consumption and dental caries.
    In new Sec. 101.80(c)(2)(i)(B), the agency is proposing to require 
that the terms ``dental caries'' or ``tooth decay'' be used in 
specifying the disease. These terms are commonly used in dental and 
dietary guidance materials and are familiar to consumers.
    Under Sec. 101.14(d), a health claim must be complete, truthful, 
and not misleading. It must enable the public to comprehend the 
information provided and to understand the relative significance of 
such information in the context of a total daily diet. In addition, a 
health claim may not attribute any specific degree of reduction in risk 
of disease from consumption of the product.
    In recognition of these general requirements, and in light of the 
fact that both environmental and genetic factors, as well as eating 
behaviors, all affect a person's risk of developing dental caries (see 
proposed Sec. 101.80(a)(1)), FDA is proposing in 
Sec. 101.80(c)(2)(i)(C) that for packages that have a total surface 
area available for labeling of 15 or more square inches, the claim must 
state that dental caries depends on many factors.
    FDA is aware that many sugar alcohol-containing chewing gum and 
confectionery products have a total surface area available for labeling 
of less than 15 square inches, however. Such a small area would 
preclude the use of a health claim that included all of the required 
elements. Many of these products, packaged in small packages, have used 
the claim ``useful only in not promoting dental caries'' on their 
labels for more than 15 years. Because of the potential dental health 
benefits to consumers resulting from a positive action on this proposal 
and given the unique history of this claim, the agency tentatively 
finds that continued use of an abbreviated claim on packages with less 
than 15 square inches of surface area will not be misleading or 
confusing 

[[Page 37526]]
to consumers of these products. However, the agency continues to 
believe that the fact that dental caries are multifactorial in their 
etiology is fundamental to an understanding of the claim. Therefore, 
the agency tentatively concludes that this fact is a material fact, and 
that it must be disclosed on packages with space available for labeling 
of 15 or more square inches. In Sec. 101.80(c)(2)(i)(D), given the 
unique circumstances surrounding this claim, FDA is proposing to exempt 
packages with a total surface area available for labeling of less than 
15 square inches from the provisions of Sec. 101.80(c)(2)(i)(C).
    In proposed Sec. 101.80(c)(2)(i)(E), FDA states that the claim must 
not attribute any degree of nonpromotion of dental caries to the use of 
the sugar alcohol-containing food. Based on the agency's review of 
human and animal studies in this document, none of the studies provide 
a basis for determining the percent reduction in risk of dental caries 
from consuming sugar alcohol-containing foods. This requirement is also 
consistent with the general requirements for health claims in 
Sec. 101.14(d), and those health claims authorized under part 101, 
subpart E.

D. Nature of the Food

    In Sec. 101.80(c)(2)(ii)(A), FDA is proposing to require that the 
food bearing this health claim meet the requirement in 
Sec. 101.60(c)(1)(i) with respect to sugars content, that is, qualify 
to bear the claim ``sugar free.'' This requirement is consistent with 
the scientific evidence showing that foods with a mixture of sugar 
alcohols and sugars are still acidogenic (Ref. 38) and cariogenic 
(Refs. 52, 55, and 56, for examples).
    In new Sec. 101.80(c)(2)(ii)(B), the agency is proposing that the 
sugar alcohols be limited to xylitol, sorbitol, mannitol, maltitol, 
isomalt, lactitol, HSH, HGS, or a combination of these. This 
requirement reflects the available scientific evidence on the sugar 
alcohols and their effects on the promotion of dental caries.
    Sugar alcohols in combination with high intensity sugar 
substitutes, such as aspartame and saccharin, are also used to replace 
sucrose. The agency notes that under proposed Sec. 101.80(c)(2)(ii)(A) 
and (c)(2)(ii)(B), a sugar alcohol and dental caries claim could appear 
on a food that contains a combination of sugar alcohols and high 
intensity sweeteners but no sugars. The agency notes that high 
intensity sweeteners are not considered fermentable by oral bacteria 
(Ref. 75).
    The agency is not specifying a level of sugar alcohols in the food 
product because these ingredients are being used as a substitute for 
sugars. Therefore, the amount of the substance required is that needed 
to achieve a desired level of sweetness.
    In new Sec. 101.80(c)(2)(ii)(C), the agency is proposing that to 
qualify to bear a claim, the sugar alcohol-containing food, when tested 
for its effects on plaque pH using in vivo methods, must not lower 
plaque pH below 5.7. Based on the agency's review of the scientific 
evidence, foods that lowered plaque pH below 5.5 were contributing to 
an acidic environment in the mouth that is detrimental to tooth enamel. 
Although a ``critical'' plaque pH has not been defined, changes in pH 
to a minimum that is above 5.5 are generally considered above the level 
where enamel decalcification would be promoted (Refs. 8, 75, 86, and 
87).
    In its review of the scientific evidence, the agency noted that 
sugar alcohol-containing chewing gum and confectioneries, such as 
mints, that do not contain fermentable carbohydrates, did not lower 
plaque pH below 5.5. However, in one study that evaluated the 
cariogenic potential of an experimental licorice that contained soy 
flour, the soy flour was shown to be highly fermentable and dropped 
plaque pH to below 5.5 (Ref. 43). The agency is concerned that use of 
sugar alcohols in a food product that contains an ingredient, such as 
refined flour, that would cause plaque pH to drop below 5.5 would thus 
cause the food to be cariogenic.
    In the Swiss ``zahnschonend'' program, if a food does not promote a 
drop in plaque pH, using intraoral plaque pH telemetric tests, below 
5.7 by bacterial fermentation either during consumption or up to 30 min 
later, the food is considered ``safe for teeth'' and may be labeled as 
such (Ref. 75). The intraoral plaque pH telemetric test is an in vivo 
method that measures the acidogenicity of foods and dietary patterns. 
Based on experience and experimentation, foods judged by the Swiss 
program to be safe for teeth are those that have been shown not to 
promote dental decay in animal or human model systems (Ref. 75).
    In this proposed rule, FDA is proposing to require in 
Sec. 101.80(c)(2)(ii)(C) that to be eligible to bear the claim, the 
food product not lower plaque pH below 5.7, based on in vivo 
measurements, during the time food is consumed and for up to 30 min 
after the food is consumed. The agency is proposing a more conservative 
value than pH 5.5 because such a value gives assurance that, consistent 
with the health claim, the food will not promote dental caries.
    The methods that have been described as the most suitable for 
assessing plaque acidity of dietary constituents in humans are 
indwelling electrode systems, such as the intraoral plaque pH 
telemetric test used in the Swiss program (Refs. 8 and 75). ICT's (Ref. 
88), which incorporate enamel blocks into dental appliances for the 
production of carious lesions when used in combination with intraoral 
plaque pH telemetry, are also good methods for assessing changes in 
plaque pH in response to food. The agency is asking for comments on 
whether establishing a minimum plaque pH that is measured in vivo 
during consumption and up to 30 min following consumption is a 
reasonable approach to use to determine whether a sugar alcohol-
containing food, other than sugar alcohol-containing chewing gum and 
confectioneries, that contains other carbohydrate ingredients is in 
compliance with any final rule resulting from this proposal.

E. Optional Information

    FDA is proposing in new Sec. 101.80(d)(1), consistent with the 
regulations that have authorized other health claims, that health 
claims about the relationship between sugar alcohols and dental caries 
may provide additional information that is drawn from proposed 
Sec. 101.80 (a) and (b).
    In new Sec. 101.80(d)(2), the agency is proposing that when 
referring to sucrose, the claim may use the term ``sucrose'' or 
``sugar.'' The use of either of these terms is consistent with FDA's 
regulation that affirms that use of this substance is GRAS 
(Sec. 184.1854).
    FDA is proposing in Sec. 101.80(d)(3), consistent with the health 
claims that it has already authorized under part 101, subpart E, to 
allow manufacturers to provide additional information about risk 
factors associated with the development of dental caries. Although 
sugars consumption and infection with S. mutans are often identified as 
the cause of dental caries, there are several risk factors that play 
significant roles in the etiology of this disease (Ref. 71). These 
factors include frequent consumption of sucrose or other fermentable 
carbohydrates, presence of oral bacteria capable of fermenting sugars, 
length of time sugars are in contact with the teeth, lack of exposure 
to fluoride, individual susceptibility, socioeconomic and cultural 
factors, and characteristics of tooth enamel, saliva, and plaque (Refs. 
7, 71, and 89).

[[Page 37527]]


F. Model Health Claims

    In proposed Sec. 101.80(e), FDA is providing model health claims to 
illustrate the requirements of new Sec. 101.80. FDA emphasizes that 
these model health claims are illustrative only. If the agency 
authorizes claims about the relationship between sugar alcohols and 
dental caries, manufacturers will be free to design their own claim so 
long as it is consistent with Sec. 101.80(c).

VII. Environmental Impact

    The agency has determined under 21 CFR 25.24 (a)(11) that this 
action is of a type that does not individually or cumulatively have a 
significant effect on the human environment. Therefore, neither an 
environmental assessment nor an environmental impact statement is 
required.

VIII. Analysis of Impacts

    FDA has examined the impacts of the proposed rule under Executive 
Order 12866 and the Regulatory Flexibility Act (Pub. L. 96-354). 
Executive Order 12866 directs agencies to assess all costs and benefits 
of available regulatory alternatives and, when regulation is necessary, 
to select regulatory approaches that maximize net benefits (including 
potential economic, environmental, public health and safety, and other 
advantages; distributive impacts; and equity). The agency believes that 
this proposed rule is consistent with the regulatory philosophy and 
principles identified in the Executive Order. In addition, the proposed 
rule is not a significant regulatory action as defined by the Executive 
Order and so is not subject to review under the Executive Order.
    The Regulatory Flexibility Act requires agencies to analyze 
regulatory options that would minimize any significant impact of a rule 
on small entities. Because it enables firms to make claims that they 
would otherwise be prohibited from making, the agency certifies that 
the proposed rule will not have a significant economic impact on a 
substantial number of small entities. Therefore, under the Regulatory 
Flexibility Act, no further analysis is required.

IX. Effective Date

    FDA is proposing to make these regulations effective 30 days after 
the publication of a final rule based on this proposal.

X. Comments

    Interested persons may, on or before October 3, 1995, submit to the 
Dockets Management Branch (HFA-305), Food and Drug Administration, rm. 
1-23, 12420 Parklawn Dr., Rockville, MD 20857, written comments 
regarding this proposal. Two copies of any comments are to be 
submitted, except that individuals may submit one copy. Comments are to 
be identified with the docket number found in brackets in the heading 
of this document. Received comments may be seen in the office above 
between 9 a.m. and 4 p.m., Monday through Friday.
XI. References

    The following references have been placed on display in the Dockets 
Management Branch (address above) and may be seen by interested persons 
between 9 a. m. and 4 p. m., Monday through Friday.
    1. Drozen, Melvin S., ``Health claim petition regarding the 
noncariogenicity of sugar alcohols,'' August 31, 1994.
    2. Drozen, Melvin S., ``Objections and request for a hearing by 
Working Group of sugar alcohol manufacturers to the revocation of 21 
C.F.R. section 105.66(f),'' Docket No. 91N-384L, Dockets Management 
Branch, FDA, Rockville, MD.
    3. Saltsman, Joyce J., CFSAN, FDA, Letter to Melvin S. Drozen, 
September 15, 1994.
    4. Saltsman, Joyce J., CFSAN, FDA, Letter to Melvin S. Drozen, 
October 7, 1994.
    5. Drozen, Melvin S., Letter to FDA, November 15, 1994.
    6. Saltsman, Joyce J., CFSAN, FDA, Memorandum of telephone 
conversation, December 8, 1994.
    7. DHHS, Public Health Service (PHS), ``The Surgeon General's 
Report on Nutrition and Health,'' U.S. Government Printing Office, 
Washington, DC, 1988.
    8. Harper, D. S., D. C. Abelson, and M. E. Jensen, ``Human plaque 
acidity models,'' Journal of Dental Research, 65 (Special Issue):1503-
1510, 1986.
    9. Ten Cate, J. M., ``Demineralization models: Mechanistic aspects 
of the caries process with special emphasis on the possible role of 
foods,'' Journal of Dental Research, 65 (Special Issue):1511-1515, 
1986.
    10. Curzon, M. E. J., ``Integration of methods for determining the 
cariogenic potential of foods: Is this possible with present 
technologies?,'' Journal of Dental Research, 65 (Special Issue):1520-
1524, 1986.
    11. Stookey, G. K., ``Considerations in determining the cariogenic 
potential of foods: How should existing knowledge be combined?,'' 
Journal of Dental Research, 65(Special Issue):1525-1527, 1986.
    12. Working Group Consensus Report, ``Integration of methods,'' 
Journal of Dental Research, 65(Special Issue):1537-1539, 1986.
    13. DePaola, D. P., ``Executive summary,'' Scientific Consensus 
Conference on Methods for Assessment of the Cariogenic Potential of 
Foods, Journal of Dental Research, 65(Special Issue)1540-1543, 1986.
    14. LSRO, FASEB, ``Dietary Sugars in Health and Disease, II. 
Xylitol,'' Bethesda, MD, July, 1978.
    15. LSRO, FASEB, ``Dietary Sugars in Health and Disease, III. 
Sorbitol,'' Bethesda, MD, July, 1978.
    16. LSRO, FASEB, ``Dietary Sugars in Health and Disease, IV. 
Mannitol,'' Bethesda, MD, July, 1978.
    17. Working Group Consensus Report, ``Animal caries,'' Journal of 
Dental Research, 65:1528-1529, 1986.
    18. Working Group Consensus Report, ``Human plaque acidity,'' 
Journal of Dental Research, 65:1530-1531, 1986.
    19. Working Group Consensus Report, ``Demineralization/
remineralization,'' Journal of Dental Research, 65:1532-1536, 1986.
    20. Moller, I. J., and S. Poulsen, ``The effect of sorbitol-
containing chewing gum on the incidence of dental caries, plaque and 
gingivitis,'' Community Dental and Oral Epidemiology, 1:58-67, 1973.
    21. Banozcy, J., E. Hadas, I. Esztary, I. Marosi, and J. Nemes, 
``Three-year results with sorbitol in clinical longitudinal 
experiments,'' Journal of the International Association of Dentistry in 
Children, 12:59-63, 1981.
    22. Kandelman, D., and G. Gagnon, ``Clinical results after 12 
months from a study of the incidence and progression of dental caries 
in relation to consumption of chewing-gum containing xylitol in school 
preventive programs,'' Journal of Dental Research, 66:1407-1411, 1987.
    23. Rekola, M., ``Changes in buccal white spots during two-year 
total substitution of dietary sucrose with xylitol,'' Acta Odontologica 
Scandinavica, 44:285-290, 1986.
    24. Makinen, K. K., and A. Scheinin, ``Turku sugar studies. VI. The 
administration of the trial and the control of the dietary regimen,'' 
Acta Odontologica Scandanavia, 33:105-127, 1975.
    25. Rekola, M., ``Approximal caries development during 2-year total 
substitution of dietary sucrose with xylitol,'' Caries Research, 21:87-
94, 1987.
    26. Scheinin, A., J. Banoczy, J. Szoke, I. Esztari, K. 
Pienihakkinen, U. Scheinin, J. Tiekso, P. Zimmerman, and E. Hadas, 
``Collaborative WHO xylitol 

[[Page 37528]]
field studies in Hungary. I. Three-year caries activity in 
institutionalized children,'' Acta Odontologica Scandinavica, 43:327-
347, 1985.
    27. Banoczy, J., A. Scheinin, R. Pados, G. Ember, P. Kertesz, and 
K. Pienihakkinen, ``Collaborative WHO xylitol field studies in Hungary. 
II. General background and control of the dietary regimen,'' Acta 
Odontologica Scandinavica, 43:349-357, 1985.
    28. Scheinin, A., K. Pienihakkinen, J. Tiekso, J. Banoczy, J. 
Szoke, I. Esztari, P. Zimmerman, and E. Hadas, ``Collaborative WHO 
xylitol field studies in Hungary. VII. Two-year caries incidence in 976 
institutionalized children,'' Acta Odontologica Scandinavica, 43:381-
387, 1985.
    29. Barmes, D., J. Barnaud, S. Khambonanda, and J. Sardo Infirri, 
``Field trials of preventive regimes in Thailand and French 
Polynesia,'' International Dental Journal, 35:66-72, 1985.
    30. Kandelman, D., A. Bar, and A. Hefti, ``Collaborative WHO 
xylitol field study in French Polynesia. I. Baseline Prevalence and 32-
month caries increment,'' Caries Research, 22:1-10, 1988.
    31. Frostell, G., L. Blomlof, I. Blomqvist, G. M. Dahl, S. Edward, 
A. Fjellstrom, C. O. Henrikson, O. Larje, C. E. Nord, and K. J. 
Nordenvall, ``Substitution of sucrose by Lycasin in candy. 
`The Roslagen study','' Acta Odontologica Scandinavica, 32:235-253, 
1974.
    32. Glass, R. L., ``A two year clinical trial of sorbitol gum,'' 
Caries Research, 17:365-368, 1983.
    33. Ikeda, T., K. Ochiai, Y. Doi, T. Mukasa, and S. Yagi, 
``Maltitol and SE58 in rats and decalcification as human intraoral 
substrate, Nihon University Journal of Oral Science, 25:1-5, 1975.
    34. Yagi, S., ``Effects of maltitol on insoluble glucan synthesis 
by S. mutans and change of enamel hardness,'' Nihon University Journal 
of Oral Science, 4:136-144, 1978.
    35. Leach, S. A., G. T. R. Lee, and W. M. Edgar, ``Remineralization 
of artificial caries-like lesions in human enamel in situ by chewing 
sorbitol gum,'' Journal of Dental Research, 68:1064-1068, 1989.
    36. Rundegren, J., T. Koulourides, and T. Ericson, ``Contribution 
of maltitol and Lycasin to experimental enamel demineralized 
in the human mouth,'' Caries Research, 14:67-74, 1980.
    37. Creanor, S. L., R. Strang, W. H. Gilmour, R. H. Foye, J. Brown, 
D. A. M. Geddes, and A. F. Hall, ``The effect of chewing gum use on in 
situ enamel lesion remineralization,'' Journal of Dental Research, 
71:1895-1900, 1992.
    38. Bibby, B. G., and J. Fu, ``Changes in plaque pH in vitro by 
sweeteners,'' Journal of Dental Research, 64:1130-1133, 1985.
    39. Birkhed, D., and S. Edwardsson, ``Acid production from sucrose 
substitutes in human dental plaque,'' Proceedings of ERGOB Conference, 
pp. 211-217, 1978.
    40. Birkhed, D., S. Edwardsson, B. Svensson, F. Moskovitz, and G. 
Frostell, ``Acid production from sorbitol in human dental plaque,'' 
Archives of Oral Biology, 23:971-975, 1978.
    41. Birkhed, D., S. Edwardsson, M. L. Ahlden, and G. Frostell, 
``Effects of 3 mo frequent consumption of hydrogenated starch 
hydrolysate (Lycasin), maltitol, sorbitol and xylitol on human dental 
plaque,'' Acta Odontologica Scandinavica, 37:103-115, 1979.
    42. Frostell, G., ``Dental plaque pH in relation to intake of 
carbohydrate products,'' Acta Odontologica Scandanavia, 27:3-29, 1969.
    43. Toors, F. A., and J. I. B. Herczog, ``Acid production from a 
nonsugar licorice and different sugar substitutes in Streptococcus 
mutans monoculture and pooled plaque-saliva mixtures,'' Caries 
Research, 12:60-68, 1978.
    44. Gallagher, I. H., and S. J. Fussell, ``Acidogenic fermentation 
of pentose alcohols by human dental plaque microorganisms,'' Archives 
of Oral Biology, 24:673-679, 1979.
    45. Gehring, F., and H. D. Hufnagel, ``Intra- and extraoral pH 
measurements on human dental plaque after rinsing with some sugar and 
sucrose substitute solutions,'' Oralprophylaxe, 5:13-19, 1983.
    46. Havenaar, R., J. H. J. Huis In't Veld, O. Backer Dirks, and J. 
D. de Stoppelaar, ``Some bacteriological aspects of sugar 
substitutes,'' Proceedings from ERGOB Conference, pp. 192-196, 1978.
    47. Jensen, M. E., ``Human plaque acidogenicity studies with 
hydrogenated starch hydrolysates,'' unpublished.
    48. Maki, Y., K. Ohta, I. Takazoe, Y. Matsukubo, Y. Takaesu, V. 
Topitsoglou, and G. Frostell, ``Acid production from isomaltulose, 
sucrose, sorbitol, xylitol in suspensions of human dental plaque,'' 
Caries Research, 17:335-339, 1983.
    49. Park, K. K., B. R. Schemehorn, J. W. Bolton, and G. K. Stookey, 
``Comparative effect of sorbitol and xylitol mints on plaque 
acidogenicity,'' presented at the International Association for Dental 
Research, April 17-21, 1991.
    50. Soderling, E., K. K. Makinen, C.-Y. Chen, H. R. Pape, and P.-L. 
Makinen, ``Effect of sorbitol, xylitol and xylitol/sorbitol gums on 
dental plaque,'' Caries Research, 23:378-384, 1989.
    51. Birkhed, D., and G. Skude, ``Relation of amylase to starch and 
Lycasin metabolism in human dental plaque in vitro,'' 
Scandinavian Journal of Dental Research, 86:248-258, 1978.
    52. Havenaar, R., J. S. Drost, J. D. de Stoppelaar, J. H. J. Huis 
in't Veld, and O. Backer Dirks, ``Potential cariogenicity of Lycasin 
80/55 in comparison to starch, sucrose, xylitol, sorbitol and L-sorbose 
in rats,'' Caries Research, 18:375-384, 1984.
    53. Havenaar, R., J. S. Drost, J. H. J. Huis in't Veld, O. Backer 
Dirks, and J. D. de Stoppelaar,, ``Potential cariogenicity of Lycasin 
80/55 before and after repeated transmissions of the dental plaque 
flora in rats,'' Archives of Oral Biology, 29:993-999, 1984.
    54. Havenaar, R., J. H. J. Huis In't Veld, J. D. de Stoppelaar, and 
O. Backer Dirks, ``A purified cariogenic diet for rats to test sugar 
substitutes with special emphasis on general health,'' Caries Research, 
17:340-352, 1983.
    55. Havenaar, R., J. D. Huis in't Veld, J. D. J. de Stoppelaar, and 
O. B. Dirks, ``Anti-cariogenic and remineralizing properties of xylitol 
in combination with sucrose in rats inoculated with Streptococcus 
mutans,'' Caries Research, 18:269-277, 1984.
    56. Grenby, T. H., and J. Colley, ``Dental effects of xylitol 
compared with other carbohydrates and polyols in the diet of laboratory 
rats,'' Archives of Oral Biology, 28:745-758, 1983.
    57. Karle, E. J., and F. Gehring, ``Kariogenitatsuntersuchungen von 
zuckeraustauschstoffen an xerostomierlen ratten. (Studies on the 
cariogenesis of sugar substitutes in xerostomized rats),'' Deutsche 
Zahnarztliche Zeitschrift, 34:551-554, 1979.
    58. Muhlemann, H. R., R. Schmid, T. Noguchi, T. Imfeld, and R. S. 
Hirsch, ``Some dental effects of xylitol under laboratory and in vivo 
conditions,'' Caries Research, 11:263-276, 1977.
    59. Shyu, K.-W., and M.-Y Hsu, ``The cariogenicity of xylitol, 
mannitol, sorbitol and sucrose,'' Proceedings of the National Science 
Council ROC, 4:21-26, 1980.
    60. Bramstedt, F., F. Gehring, and E. J. Karle, ``Comparative study 
of the cariogenic effects of Palatinit, xylitol and 
saccharose in animals,'' unpublished, 1976.
    61. Izumiya, A., T. Ohshima, and S. Sofue, ``Caries inducibility of 
various sweeteners,'' Academy of Pedodontia, p. 65, May 1984.
    62. Gehring, F., and E. J. Karle, ``The sugar substitute 
Palatinit with special emphasis on microbiological and 
caries-preventing aspects,'' Zeitschrift Ernahrungswiss, 20:96-106, 
1981. 

[[Page 37529]]

    63. Karle, E. J., and F. Gehring, ``Palatinit-A New Sugar 
Substitute and its Carioprophylactic Assessment,'' Deutsche 
Zalnarztliche Zeitschrift 33:189-191, 1978.
    64. Larje, O., and R. H. Larson, ``Reduction of dental caries in 
rats by intermittent feeding with sucrose substitutes, Archives of Oral 
Biology, 15:805-816, 1970.
    65. Muhlemann, H. R., ``Effect of topical application of sugar 
substitutes on bacterial agglomerate formation, caries incidence and 
solution rates of molars in the rat,'' unpublished, 1978.
    66. Ooshima, T., A. Izumitani, T. Minami, T. Yoshida, S. Sobue, T. 
Fujiwari, and S. Hamada, ``Non-cariogenicity of maltitol in SPF rats 
infected with mutans streptococci,'' submitted for publication.
    67. Tate, N., S. Wada, H. Tani, and K. Oikawa, ``Experimental 
studies on correlations between progressive caries and sugar intake,'' 
unpublished.
    68. Leach, S. A., and R. M. Green, ``Effect of xylitol-supplemented 
diets on the progression and regression of fissure caries in the albino 
rat,'' Caries Research, 14:16-23, 1980.
    69. Mukasa, T., ``The possibility of maltitol and SE 58 as non-
cariogenic sweeteners: their utilization by Streptococcus mutans for 
insoluble glucan synthesis and experimental dental caries in rats,'' 
Nihon University Journal of Oral Science, 3:266-275, 1977.
    70. Hoeven, J. S. van der, ``Cariogenicity of disaccharide alcohols 
in rats,'' Caries Research, 14:61-66, 1980.
    71. Burt, B. A., and A. I. Ismail, ``Diet, nutrition, and food 
cariogenicity,'' Journal of Dental Research, 65 (Special Issue): 1475-
1484, 1986.
    72. National Research Council, National Academy of Sciences, ``Diet 
and Health,'' National Academy Press, Washington, DC, 1989.
    73. Hoeven, J.S. van der, ``Carigenicity of lactitol in program-fed 
rats,'' Caries Research, 20:441-443, 1986.
    74. Imfeld, T., and H. R. Muhlemann, ``Cariogenicity and 
acidogenicity of food, confectionery and beverages,'' Pharmacology and 
Therapeutic Dentistry, 3:53-68, 1978.
    75. Imfeld, T., ``Identification of Low Caries Risk Dietary 
Components,'' Monographs in Oral Science, vol. 11, Karger, Basel, 
Switzerland, pp. 1-8 and 117-144, 1983.
    76. Grenby, T. H., A. Phillips, and M. Mistry, ``Studies of the 
dental properties of lactitol compared with five other bulk sweeteners 
in vitro,'' Caries Research, 23:315-319, 1989.
    77. Grenby, T. H., and A. Phillips, ``Dental and metabolic effects 
of lactitol in the diet of laboratory rats,'' British Journal of 
Nutrition, 61:17-24, 1989.
    78. Edgar, W. M., and D. A. M. Geddes, ``Plaque acidity models for 
cariogenicity testing--some theoretical and practical observations,'' 
Journal of Dental Research, 65 (Special Issue): 1498-1502, 1986.
    79. Birkhed, D., S. Kalfas, G. Svensater, and S. Edwardsson, 
``Microbiological aspects of some caloric sugar substitutes,'' 
International Dental Journal, 35:9-17, 1985.
    80. Schrotenboer, G. H., ``In the Matter of Revising the Regulation 
for Foods for Special Dietary Uses,'' Docket No. FDC-78, March 4, 1970 
at 6-7.
    81. Saltsman, Joyce J., CFSAN, FDA, Memorandum to file--
Environmental Assessment of Health Claim Petition, December 23, 1994.
    82. Ayers, C. S., and R. A. Abrams, ``Noncariogenic sweeteners, 
sugar substitutes for caries control,'' Dental Hygiene, April, 162-
167:1987.
    83. Rugg-Gunn, A. J., and W. M. Edgar, ``Sweeteners and dental 
health,'' Community Dental Health, 2:213-223, 1985.
    84. Grenby, T. H., ``Nutritive sucrose substitutes and dental 
health,'' In: Developments in Sweeteners, editors: T. H. Grenby, K. J. 
Parker, and M. G. Lindley, Elsevier Science, Inc., 2:51-88, 1983.
    85. Rugg-Gunn, A. J., ``Lycasin and the prevention of 
dental caries,'' In: Progress in Sweeteners, editor: T. H. Grenby, 
Elsevier Science, Inc., pp. 311-328, 1989.
    86. Loesche, W. J., ``The rationale for caries prevention through 
use of sugar substitutes,'' International Dental Journal, 35:1-8, 1985.
    87. Mandel, I. D., ``Dental caries,'' American Scientist, 67:680-
688, 1979.
    88. Koulourides, T., R. Bodden, S. Keller, L. Manson-Hing, J. 
Lastra, and T. Housch, ``Cariogenicity of nine sugars tested with an 
intraoral device in man,'' Caries Research 10:427-441, 1976.
    89. Baer, A., ``Significance and promotion of sugar substitution 
for the prevention of dental caries,'' Lben.-Wiss U. Technology, 
Academic Press, 22:46-53, 1989.
    90. LSRO, FASEB ``Health Aspect of Sugar Alcohols and Lactose,'' 
Bethesda, MD, December 1986.
    91. Joint FAO/WHO Expert Committee on Food Additives, ``Evaluation 
of Certain Food Additives and Contaminants,'' Geneva, Switzerland, pp. 
16-17, 1993.

List of Subjects in 21 CFR Part 101

    Food labeling, Nutrition, Reporting and recordkeeping requirements.

    Therefore, under the Federal Food, Drug, and Cosmetic Act and under 
authority delegated to the Commissioner of Food and Drugs, it is 
proposed that 21 CFR part 101 be amended as follows:

PART 101--FOOD LABELING

    1. The authority citation for 21 CFR part 101 is revised to read as 
follows:

    Authority: Secs. 4, 5, 6 of the Fair Packaging and Labeling Act 
(15 U.S.C. 1453, 1454, 1455); secs. 201, 301, 402, 403, 409, 701 of 
the Federal Food, Drug, and Cosmetic Act (21 U.S.C. 321, 331, 342, 
343, 348, 371).

    2. New Sec. 101.80 is added to subpart E to read as follows:
Sec. 101.80  Health claims: dietary sugar alcohols and dental caries.

    (a) Relationship between dietary sugar alcohols and dental caries. 
(1) Dental caries, or tooth decay, is a disease caused by many factors. 
Both environmental and genetic factors can affect the development of 
dental caries. Risk factors include tooth enamel crystal structure and 
mineral content, plaque quantity and quality, saliva quantity and 
quality, individual immune response, types and physical characteristics 
of foods consumed, eating behaviors, presence of acid producing oral 
bacteria, and cultural influences.
    (2) The relationship between dietary sugars consumption and tooth 
decay is well established. Sucrose is one of the most, but not the 
only, cariogenic sugar in the diet. Bacteria found in the mouth are 
able to metabolize sugars producing acid and forming dental plaque. 
Prolonged exposure of the tooth enamel to acids from dental plaque 
causes tooth enamel to demineralize, or decay. Frequent between-meal 
consumption of sugary foods, particularly foods that easily stick to 
the teeth, can cause tooth decay.
    (3) U.S. diets tend to be high in sugars consumption. Although 
there has been a decline in the prevalence of dental caries in the 
United States, per capita consumption of sugars has not declined, and 
the disease remains widespread throughout the population. Federal 
government agencies and nationally recognized health professional 
organizations recommend decreased consumption of sugars.
    (4) Dietary sugar alcohols can be used to replace dietary sugars in 
food. Sugar alcohols are significantly less cariogenic than dietary 
sugars. Thus, replacing dietary sugars with sugar alcohols helps to 
maintain dental health.
    (b) Significance of the relationship between sugar alcohols and 
dental caries. Sugar alcohols do not promote 

[[Page 37530]]
dental caries because they are slowly metabolized by bacteria to form 
some acid. The rate and amount of acid production is significantly less 
than that from sucrose and does not cause the loss of important 
minerals from tooth enamel.
    (c) Requirements. (1) All requirements set forth in Sec. 101.14 
shall be met, except that sugar alcohol-containing foods are exempt 
from section Sec. 101.14(e)(6).
    (2) Specific requirements. (i) Nature of the claim. A health claim 
relating sugar alcohols and the nonpromotion of dental caries may be 
made on the label or labeling of a food described in (c)(2)(ii) of this 
section, provided that:
    (A) The claim shall state ``does not promote,'' ``useful in not 
promoting,'' or ``expressly for not promoting'' dental caries.
    (B) In specifying the disease, the claim uses the following terms: 
``dental caries'' or ``tooth decay.''
    (C) For packages with a total surface area available for labeling 
of 15 or more square inches, the claim shall indicate that dental 
caries depends on many factors.
    (D) Packages with a total surface area available for labeling of 
less than 15 square inches are exempt from paragraph (C) of this 
section.
    (E) The claim shall not attribute any degree of nonpromotion of 
dental caries to the use of the sugar alcohol-containing food.
    (ii) Nature of the food. (A) The food shall meet the requirement in 
Sec. 101.60(c)(1)(i) with respect to sugars content.
    (B) The sugar alcohol in the food shall be xylitol, sorbitol, 
mannitol, maltitol, isomalt, lactitol, hydrogenated starch 
hydrolysates, hydrogenated glucose syrups, or a combination of these.
    (C) The sugar alcohol-containing food shall not lower plaque pH 
below 5.7 by bacterial fermentation either during consumption or up to 
30 minutes after consumption, as measured by in vivo tests.
    (d) Optional information. (1) The claim may include information 
from paragraphs (a) and (b) of this section, which describe the 
relationship between diets containing sugar alcohols and dental caries.
    (2) In referring to sucrose, the claim may use the term ``sucrose'' 
or ``sugar.''
    (3) The claim may identify one or more of the following risk 
factors for dental caries: Frequent consumption of sucrose or other 
fermentable carbohydrates; presence of oral bacteria capable of 
fermenting sugars; length of time sugars are in contact with the teeth; 
lack of exposure to fluoride; individual susceptibility; socioeconomic 
and cultural factors; and characteristics of tooth enamel, saliva, and 
plaque.
    (e) Model health claim. The following model health claims may be 
used in food labeling to describe the relationship between sugar 
alcohol and dental caries.
    (1) For packages with total surface area available for labeling of 
less than 15 square inches:
    (i) Useful only in not promoting tooth decay;
    (ii) Does not promote tooth decay; and
    (iii) [This product] does not promote tooth decay.
    (2) For packages with total surface area available for labeling of 
15 or more square inches:
    (i) Tooth decay is a disease caused by many factors including 
frequent between meal consumption of sugary foods. [Name of sugar 
alcohol] does not promote tooth decay.
    (ii) [Reserved].

    Dated: July 7, 1995.
William B. Schultz,
Deputy Commissioner for Policy.

    Note: The following tables will not appear in the annual Code of 
Federal Regulations.

BILLING CODE 4160-01-P

[[Page 37531]]
[GRAPHIC][TIFF OMITTED]TP20JY95.000



[[Page 37532]]
[GRAPHIC][TIFF OMITTED]TP20JY95.001



[[Page 37533]]
[GRAPHIC][TIFF OMITTED]TP20JY95.002



[[Page 37534]]
[GRAPHIC][TIFF OMITTED]TP20JY95.003



[[Page 37535]]
[GRAPHIC][TIFF OMITTED]TP20JY95.004



[[Page 37536]]
[GRAPHIC][TIFF OMITTED]TP20JY95.005



[[Page 37537]]
[GRAPHIC][TIFF OMITTED]TP20JY95.006



[[Page 37538]]
[GRAPHIC][TIFF OMITTED]TP20JY95.007



[[Page 37539]]
[GRAPHIC][TIFF OMITTED]TP20JY95.008



[[Page 37540]]
[GRAPHIC][TIFF OMITTED]TP20JY95.009



[[Page 37541]]
[GRAPHIC][TIFF OMITTED]TP20JY95.010



[[Page 37542]]
[GRAPHIC][TIFF OMITTED]TP20JY95.011



[[Page 37543]]
[GRAPHIC][TIFF OMITTED]TP20JY95.012



[[Page 37544]]
[GRAPHIC][TIFF OMITTED]TP20JY95.013



[[Page 37545]]
[GRAPHIC][TIFF OMITTED]TP20JY95.014



[[Page 37546]]
[GRAPHIC][TIFF OMITTED]TP20JY95.015



[[Page 37547]]
[GRAPHIC][TIFF OMITTED]TP20JY95.016



[[Page 37548]]
[GRAPHIC][TIFF OMITTED]TP20JY95.017



[[Page 37549]]
[GRAPHIC][TIFF OMITTED]TP20JY95.018



[FR Doc. 95-17505 Filed 7-19-95; 8:45 am]
BILLING CODE 4160-01-C