Food Safety: Regulation

Center for Science in the Public Interest

Regulatory Comments and Petitions


Petition for Regulatory Action to Revise the ) 

Labeling Requirements for Foods Containing )

   Docket No. ______________
Sorbitol ) 

Submitted by the
Center for Science in the Public Interest
September 27, 1999

Michael F. Jacobson, Ph.D.
Executive Director
1875 Connecticut Ave., N.W.
Suite 300
Washington, D.C. 20009
September 27, 1999

Dockets Management Branch
Food and Drug Administration
Department of Health and Human Services
5630 Fishers Lane, Room 1061
Rockville, Maryland 10857


The Center for Science in the Public Interest (CSPI) submits this petition pursuant to section 4(d) of the Administrative Procedure Act (5 U.S.C. § 553(e)) and sections 402(a)(1), 402(a)(2)(A), 406, and 701(a) of the federal Food, Drug, and Cosmetic Act (FFDCA), (21 U.S.C. §§ 342(a)(1), 342(a)(2)(A), 346, and 371(a)). We request that the Food and Drug Administration (FDA) take regulatory action to revise the labeling requirements for foods that contain sorbitol.


   Sorbitol is a naturally-occurring hexahydric alcohol that is found in various fruits and plants. Because it is sweet-tasting, non-cariogenic, and less caloric than sugars, sorbitol is produced commercially and commonly used as a sugar substitute in such dietetic food products as sugar-free candies, breakfast syrups, and cake mixes. Such products are popular among diabetics and others who seek to limit their consumption of sugar.

   Unfortunately, ingestion of sorbitol can cause a range of gastrointestinal problems, including diarrhea, abdominal pain, and bloating. At sufficiently high doses, the substance can produce osmotic diarrhea, a property that has been exploited by clinicians to induce catharsis. Children are especially susceptible to sorbitol-related gastrointestinal problems. In fact, at least one outbreak of diarrhea among youngsters has been attributed to consumption of sorbitol-sweetened dietetic candies.

   FDA regulations require a small subset of sorbitol-containing products to bear a label alerting consumers about the potential for gastrointestinal problems. The labeling requirement applies only to those food products whose "reasonably foreseeable consumption may result in a daily ingestion of 50 grams of sorbitol."(1) The labels of such products must state: "Excess consumption may have a laxative effect."(2)

   The current labeling requirement does not adequately protect the health of many consumers. Numerous clinical studies show that the 50-gram threshold that triggers the label notice is too high, because susceptible adults can experience diarrhea and other symptoms at doses as low as 10 grams. The current requirement therefore exempts many products that, even with moderate use, could cause gastrointestinal problems in some people. In addition, the prescribed notice statement is too vague. Without more precise information about what constitutes "excess consumption," consumers cannot determine how to limit their consumption of a sorbitol-containing food to avoid gastrointestinal problems. Moreover, the statement’s potency is diminished by use of the term "laxative effect," which is likely to be perceived as a mild effect (such as slight stool softening). That term does not reflect the actual gastrointestinal symptoms that sorbitol can cause, which include diarrhea, bloating, and abdominal pain.

   Nor does the required statement alert consumers to the fact that it is sorbitol, and not some other ingredient, that is responsible for a product’s potential to cause gastrointestinal problems. Susceptible consumers need that specific information if they are to learn that they must limit the amount of sorbitol in their diets to avoid its ill effects. Also, the current label notice does not inform consumers that children are especially susceptible to sorbitol’s gastrointestinal effects, despite the fact that children are prone to binge on many sorbitol-sweetened products, such as gums and candies.

   CSPI urges FDA to correct the numerous shortcomings in the existing regulations. The agency should revise the sorbitol labeling provision so that it applies to all products whose consumption reasonably could lead to gastrointestinal problems in susceptible consumers. The label should also be revised to help consumers understand the potential adverse health effects of eating too much sorbitol, and to alert them to the special danger posed to children.

   To achieve those important public-health objectives, FDA should amend the existing regulation to require a label notice for all products containing one or more grams of sorbitol per serving. The revised regulation should prescribe the following (or similar) language for such products: "NOTICE: This product contains sorbitol, which may cause diarrhea, bloating, and abdominal pain. Not suitable for consumption by children. To protect yourself, start by eating no more than one serving at a time."

   FDA also should revise its regulations governing sorbitol-containing confectioneries sold in bulk. If the individual wrappers of such products do not bear the required statement, the statement should be placed on the retail bulk containers in which the products are sold so that consumers will be alerted to the potential health risk at the time of purchase.

   Finally, the agency should revise the existing label-notice requirement for foods that contain other diarrhea-inducing sugar alcohols, including mannitol, maltitol, isomalt, and hydrogenated starch hydrolysate. The revisions should be analogous to those proposed for sorbitol.


   CSPI requests that FDA take regulatory action to revise the current labeling requirement for foods that contain sorbitol, found at 21 C.F.R. § 184.1835(e). Specifically, CSPI asks the agency to replace the existing requirement with the following: "The label and labeling of food containing one or more grams of sorbitol per serving shall bear the statement: ’NOTICE: This product contains sorbitol, which may cause diarrhea, bloating, and abdominal pain. Not suitable for consumption by children. To protect yourself, start by eating no more than one serving at a time.’" CSPI also asks the agency to modify 21 C.F.R. § 1.24(a)(4), pertaining to individually wrapped penny candies and other confectioneries, so that such products are not exempted from the sorbitol label-notice requirement unless the retail bulk container in which they are sold bears the required statement.

   In addition, CSPI requests that FDA revise the existing labeling requirements for mannitol-containing food products, found at 21 C.F.R. § 180.25(e), to read as follows: "The label and labeling of food containing one or more grams of mannitol per serving shall bear the statement: ’NOTICE: This product contains mannitol, which may cause diarrhea, bloating, and abdominal pain. Not suitable for consumption by children. To protect yourself, start by eating no more than one serving at a time.’" CSPI also asks FDA to establish similar requirements for foods that contain maltitol, isomalt, hydrogenated starch hydrolysate, or other sugar alcohols that cause gastrointestinal problems in humans.

   FDA is authorized to take all of the requested actions under sections 201(n), 403(a)(1), and 701 of the FFDCA.


A. Factual Grounds

1. Clinical Studies Show That Sorbitol Can Cause Gastrointestinal Effects at Doses Far Lower than 50 Grams Per Day

   The laxative threshold for sorbitol was reported in 1941 to be approximately 50 grams (833 mg per kg body weight) in healthy adults,(3) and the substance has been used for many years by clinicians to induce catharsis.(4)

   More recently, clinical researchers have discovered that far less than 50 grams of sorbitol can produce gastrointestinal symptoms, ranging from mild discomfort to severe diarrhea, in healthy individuals. For instance, in a 1990 study involving 12 diabetics and 23 nondiabetics, a 10-gram dose of sorbitol produced diarrhea and other symptoms in many subjects.(5) Although none of the 12 diabetics developed diarrhea after sorbitol ingestion, six complained of bloating and two of abdominal pain. Of the 23 nondiabetic subjects, 13 developed bloating, nine developed abdominal pain, and three suffered diarrhea. Those results echoed the findings of a 1985 study, in which 20 of 42 volunteers complained of gastrointestinal symptoms after ingesting 10 grams of sorbitol, with nine suffering from bloating, four from bloating and abdominal pain, and seven from those symptoms as well as diarrhea.(6) A smaller study conducted in 1983 yielded similar results, with five of seven young-adult subjects complaining of gas and bloating after consuming as little as 10 grams of sorbitol and one subject experiencing cramps and/or diarrhea.(7) That study found that a dose of five grams caused gas and/or bloating in three subjects.

   At doses of 20 grams and higher, sorbitol is even more likely to induce diarrhea and cramps. Four of seven subjects in the 1983 study exhibited those symptoms after ingesting 20 grams of sorbitol,(8) and in a study reported in 1995 five of 22 subjects who ingested 20 grams of the substance reported abdominal pain, while three others suffered from mild diarrhea.(9)

   At a dose of 40 grams, sorbitol causes severe gastrointestinal problems in many individuals, as found in a recent study conducted by Procter and Gamble that compared the effects of sorbitol to the effects of the indigestible fat substitute olestra.(10) That study, in which sorbitol was used as a positive control, was designed to measure the chemicals’ effects on subjects’ stool consistency and composition, as well as on the occurrence of gastrointestinal symptoms, including cramping, bloating, flatulence, heartburn, nausea, and urgency. The 40-gram daily dose of sorbitol, derived from sorbitol-sweetened candies, induced the following symptoms, as summarized by an FDA scientist: (1) rapid-onset liquid/rice-water stools coupled with an increase in mean stool-water output and in bowel-movement frequency (all signaling diarrhea); (2) a considerable increase in electrolyte output in stools (also signaling diarrhea); and (3) a statistically significant increase (over placebo) in the severity of cramping, nausea, and urgency.(11)

   Children, because of their relatively low body weight, are even more susceptible than adults to sorbitol-induced diarrhea and other gastrointestinal symptoms. It has been reported that diarrhea can result from a young child’s consumption of as little as 0.5 grams of sorbitol (in a liquid vitamin supplement) per kilogram of body weight.(12) A three-year-old child weighing 15 kg (33 pounds) therefore could develop diarrhea after consuming only 7.5 grams of the substance.(13)

   In 1984, epidemiologists reported an outbreak of diarrhea linked to sorbitol-containing candies in New Hampshire. Eight children between ages 5 and 13 years who had eaten as few as three candies (providing a total of nine grams of sorbitol) suffered abdominal cramps, urgency in defecation, and multiple loose bowel movements.(14)

   A recent retrospective study involving children ages one to five provides additional indirect evidence that sorbitol consumption can cause diarrhea in youngsters.(15) In that study, researchers surveyed the children’s parents to ascertain the amount of sorbitol-sweetened products that the children had eaten over a three-month period, as well as the number of days on which the children had suffered from diarrhea unaccompanied by a fever. For the three-year-old children in the study, a positive correlation existed between the number of reported days of diarrhea and sorbitol intake from sugar-free gum, breath mints, and candies. In fact, all five three-year-olds who had consumed 0.5 grams or more of sorbitol per kilogram of body weight suffered from some diarrhea during the three-month period, while only four of 15 three-year-olds who had not consumed any sorbitol had suffered diarrhea.(16) No statistically significant correlation between sorbitol consumption and afebrile diarrhea existed for the other age groups in the study, a fact that the researchers noted may be attributable to flaws in the study design.(17)

2. Consumption of Sorbitol is Widespread

   Sorbitol is used as a sugar substitute in a wide range of dietetic foods. Examples of such foods include sugar-free maple syrup, brownies, cookies, and pancake and cake mixes, as well as gums, breath mints, licorice, and other candies.(18) The amount of sorbitol per serving in those products ranges from one or two grams in gums and candies to 10 or more grams in some syrups and cake mixes.

   Processed foods in which sorbitol serves as a sugar substitute are not the only source of the substance in many people’s diets. Minimal amounts of sorbitol naturally occur in several fruits, including apples, sweet cherries, plums, pears, prunes, and peaches.(19) The substance also is used in a wide variety of medicinal products. A 1994 survey showed that many liquid medications listed in Physicians’ Desk Reference contained large quantities of sorbitol per normal dosage unit.(20) For instance, a single dose of Roxane Laboratories’ Milk of Magnesia contains six grams of sorbitol, and a single dose of some versions of theophylline contain as many as 26 grams of the substance.(21) Obviously, consumers who eat sorbitol-sweetened foods while taking those medications would ingest far more sorbitol than that provided by the foods alone.

   Because sorbitol is found in a range of foods -- both dietetic and nondietetic -- and in many liquid medicinal products, the amount of sorbitol derived from sorbitol-sweetened foods may constitute just a fraction of the daily sorbitol intake for some people. As explained more fully below, the fact that consumers can ingest sorbitol from numerous food and medicinal sources supports a requirement that all food products containing more than a minimum amount of sorbitol bear a label notice stating that the sorbitol in the product can contribute to gastrointestinal problems.(22)

3. FDA’s Current Labeling Requirement for Sorbitol-Containing Foods Does Not Adequately Protect Consumers and Should Be Revised

   As previously stated, under existing FDA regulations sorbitol-containing foods whose "reasonably foreseeable consumption may result in a daily ingestion of 50 grams of sorbitol" must bear a label notice stating that "Excess consumption may have a laxative effect." That requirement is insufficiently protective of public health, because the 50-gram threshold is far too high and the prescribed language is too vague to alert consumers to the potential dangers of sorbitol consumption.

a.   All Products Containing One or
 More Grams of Sorbitol Per
Serving Should Bear the Required
Label Notice

   Data from the clinical studies summarized above demonstrate that although sorbitol’s laxative dose in adults was measured in 1941 to be approximately 50 grams, consumption of as little as 10 grams can cause gastrointestinal symptoms ranging from mild discomfort to osmotic diarrhea in some people. Even in those studies in which only mild effects were seen at a 10-gram dose, osmotic diarrhea was observed when subjects consumed 20 grams of sorbitol. Young children appear to be especially susceptible to sorbitol-induced diarrhea, which may occur after ingestion of as few as 7.5 grams of the substance.

   While the study results described above indicate that sensitivity to sorbitol varies from person to person, and not everyone suffers from gastrointestinal problems even at the 20-gram dose, there is no question that the current 50-gram threshold for triggering the sorbitol label-notice requirement does not protect the health of many consumers. That threshold is based on older, more limited data on laxative dose and does not take into account the newer studies described above.(23) Moreover, it completely ignores the concept of a safety margin.

   CSPI urges FDA to revise its current sorbitol labeling regulation to reflect the latest scientific evidence about sorbitol’s gastrointestinal effects. Specifically, the agency should abandon the current 50-gram threshold and instead require a label notice on all food products that contain one or more grams of sorbitol per serving.

   Several considerations favor adoption of that requirement. First, such a requirement is necessary to ensure that the notice appears on the labels of candies, breath mints, and other confectioneries that contain only one or two grams of sorbitol, but which may be consumed in large numbers in a single sitting. Thus, although a single "serving," as indicated on the label of such a product, may comprise just one or two individual pieces (containing one or two grams of sorbitol), some consumers may eat far more than that amount of sorbitol in a short period of time, ingesting a substantial amount in the process.

   Second, because sorbitol is used in many food and medicinal products and also occurs naturally in some commonly eaten fruits and fruit juices, even foods that contain only one gram of the substance can contribute to gastrointestinal problems if they are eaten around the same time that other sorbitol-containing foods or medicines are consumed. A label notice on all foods containing one or more grams of the substance per serving would alert consumers to the fact that the sorbitol present could cause problems and would discourage over consumption of sorbitol-sweetened foods.

   Third, the greater vulnerability of children to sorbitol-related gastrointestinal problems provides additional support for the proposed change. Consumption of sorbitol-sweetened foods appears to be significant among children.(24) As already explained, as little as 0.5 grams of sorbitol per kilogram body weight (or about 7.5 grams of sorbitol for a 15-kilogram child) can produce diarrhea in small children. In fact, young children may ingest more than that modest amount of sorbitol in a single sitting by consuming just a few sorbitol-sweetened gums and candies. Requiring the label notice on all gums and candies that contain one or more grams of sorbitol would alert parents to the health risk those products pose to their children.

   For the foregoing reasons, all products containing one or more grams of sorbitol per serving should bear a label notice. Until such a requirement is in place, products that contain enough sorbitol to cause gastrointestinal products will continue to be marketed without any indication of the risk to consumers’ health.

   Predictably, the current 50-gram threshold has led some manufacturers to avoid labeling products that we believe should bear the label notice. In a survey of sorbitol-containing products in Washington, D.C.-area stores, we found several products whose labels disclose that they contain from two grams (sugar-free breath mints) to 11 grams (sugar-free syrup) of sorbitol per serving but lack the label notice.(25) Those products should bear the notice. For instance, for children and other susceptible people, consumption of just a single serving of the syrup would result in the ingestion of enough sorbitol to cause gastrointestinal problems, as would as few as three or four servings of the breath mints (9 to 12 pieces).

b.   The Sorbitol Label Notice Should
 Be Prominent and Conspicuous

   Our product survey also revealed that even when the label notice is printed on food packages it often is difficult to see. Manufacturers rarely use a type size, color, and location for the statement that stand out from the marketing and nutritional information on the package.(26) Instead, the label notice frequently is relegated to fine print on the back of the food package.

   Obviously, a label notice that is not seen by consumers is worthless. To ensure that the sorbitol label notice is read and understood by consumers, FDA should amend the current regulations to require that either the statement itself be placed on the principal display panel (PDP) or that the PDP contain a brief note advising consumers to read the health notice on the side or back panel. In addition, FDA should require that the statement be enclosed in a graphic box and satisfy other prescribed standards of legibility.

   All but two of the sugar-free, sorbitol-containing foods in CSPI’s survey use the PDP to advertise the fact that they do not contain added sugar.(27) Consumers should be informed, also by means of a statement on the PDP itself or by a note directing them to the side or back panel, that those foods pose a risk of gastrointestinal problems. FDA does a disservice to consumers by permitting food manufacturers to promote the lack of sugar in their products on the front of the package, while relegating all information about the health risks posed by the sugar substitute to fine print on the side or back of the package.

   In addition to requiring that the PDP contain the label-notice statement itself or a note advising consumers where it may be found, FDA should mandate that the statement be enclosed by a graphic box, use an easy-to-read type size, and be printed on a contrasting background to enhance visibility. FDA should model those aspects of the revised regulation on the rule governing the labeling of foods containing olestra, found at 21 C.F.R. § 172.867(e)(2). Specifically, the statement should: (1) be enclosed in a 0.5 point box rule with at least 2.5 points of space around the statement; (2) utilize at least one point leading; (3) have a type that is kerned so the letters do not touch; (4) be all black or one color type and printed on a white or other neutral contrasting background; (5) use a single easy-to-read type style and upper and lowercase letters; and (6) be in type size no smaller than eight point (with the minimum size to increase as the package size increases). Also, the first sentence of the proposed statement should be printed in bold type, as is required under the olestra labeling rule.

c.   The Statement Should Indicate
 That Sorbitol is the Ingredient
That May Induce Gastrointestinal
Problems, Describe the Symptoms
That May Result, and State That
Children Should Not Consume
Sorbitol-Containing Products

   CSPI’s product survey revealed other shortcomings in the existing labeling requirements. Because it is not currently required, many of the products that contain the label notice do not indicate that the potential laxative effect is caused by sorbitol and not by some other ingredient.(28) In addition, none of the labels from the sorbitol-containing products we examined explains how much sorbitol or how many servings of the food constitute "excess consumption" that may cause gastrointestinal problems. Nor do the products provide a more precise description of the symptoms caused by sorbitol’s "laxative effect." Finally, none of the products warn of the increased risk of gastrointestinal problems that sorbitol poses to children.

   To address those shortcomings, FDA should revise the text of the required statement to read as follows: "NOTICE: This product contains sorbitol, which may cause diarrhea, bloating, and abdominal pain. Not suitable for consumption by children. To protect yourself, start by eating no more than one serving at a time."

   That language would address all four problems discussed above. By explicitly referring to sorbitol, the proposed statement would alert consumers to the fact that sorbitol is the ingredient that poses the risk of gastrointestinal problems. Also, by advising consumers to start by eating no more than a single serving of the product, the statement would help ensure that consumers exercise caution in using the product and do not inadvertently ingest a potentially harmful amount of sorbitol.(29) In addition, CSPI’s proposed statement would replace the vague term "laxative effect" with a more accurate and readily understood description of the actual symptoms caused by sorbitol.

   The proposed language stating that children should not consume the product is warranted by the findings that less than 10 grams of sorbitol can induce diarrhea in children, as well as the tendency of children to eat multiple servings of sweet products in a short period of time.

   Among other benefits, the language proposed by CSPI would have the salutary effect of increasing consumer knowledge about the presence of sorbitol in dietetic foods and the potential health effects of sorbitol consumption. Several researchers have noted that greater consumer education in this area would be beneficial because people generally are unaware that many of the sugar-free products they consume contain sorbitol, and that sorbitol can cause gastrointestinal problems.(30)

d.   Retail Containers of Bulk
 Confectioneries Containing
Sorbitol Should Bear the Required
Label Notice

   Another problem faced by consumers is that some sorbitol-containing candies are sold in bulk without proper labeling to indicate that sorbitol is an ingredient or that the candies may cause gastrointestinal problems. For instance, the hard candies implicated in the 1984 outbreak of diarrhea among children in New Hampshire, discussed above, had been purchased in bulk and were individually packaged in wrappers that contained no ingredient information or label notice concerning the potential adverse effects due to sorbitol.(31) In another case reported in the literature, a man suffered two bouts of severe diarrhea after eating sorbitol-containing sugar-free gummy bears that had been sold in bulk without an ingredients list or sorbitol label notice.(32)

   The products involved in those cases were unlabeled because FDA exempts from its labeling requirements all individually wrapped pieces of penny candy and other confectionery weighing less than one-half ounce per piece, provided the candy’s shipping container satisfies all labeling requirements.(33)

   FDA should revise its regulations to make sorbitol-containing candies ineligible for that exemption unless the required sorbitol label notice is clearly and prominently placed on the retail bulk container in which the candies are sold. Because bulk candies can contain as many as three grams of sorbitol(34) per piece and they commonly are consumed in large amounts in a single sitting, such products pose a significant diarrhea risk to consumers. The gummy-bear case discussed above, in which the patient consumed eight ounces of sugar-free candies in one hour, attests to that fact.

   Because the small wrapper size may make it impossible to include the required label notice on bulk candies, FDA instead should require that the retail bins in which the products are sold bear the label notice. That would allow consumers to read the statement at the time of purchase.

e.   Products Containing Mannitol and
 Other Diarrhea-Inducing Sugar
Alcohols Should Be Subject To the
Same Labeling Requirements as
Those Proposed in This Petition

   Mannitol, like sorbitol, is a sugar alcohol used to sweeten dietetic foods. Also like sorbitol, mannitol causes diarrhea. According to the same 1941 study in which sorbitol’s laxative threshold was determined, mannitol has a laxative threshold of 10 to 20 grams.(35) That finding led FDA to mandate a label notice stating that "[e]xcess consumption may have a laxative effect" on all products whose reasonably foreseeable daily consumption would provide more than 20 grams of the substance.(36)

   The 20-gram threshold is too high. The very study that the agency relied upon to set the threshold found that some subjects suffered loose stools after administration of as little as 10 grams of mannitol.(37) To protect children and other sensitive consumers, FDA should require that all products containing one or more grams of mannitol per serving bear a label notice analogous to that proposed for sorbitol.(38)

   CSPI’s survey of dietetic foods indicates that other sugar alcohols also are being used to sweeten dietetic foods. Specifically, we found dietetic candies and snack bars that contain maltitol, isomalt, and hydrogenated starch hydrolysate (HSH) in amounts ranging from 8 to 31 grams per serving.(39) Although the packages of each of those products contained some version of the label notice required for sorbitol and mannitol, FDA’s regulations do not currently mandate use of the notice on foods containing maltitol, isomalt, or HSH.(40) Nor did the statements on the products in our survey stand out from the marketing or nutritional information on the food packages.

   While there is little data on the ability of such substances to cause gastrointestinal effects, FDA should take regulatory action to require a label notice for products containing maltitol, isomalt, HSH, or any other sugar alcohol that is likely to induce gastrointestinal problems in humans. The text of the statement should be similar to that proposed by CSPI for sorbitol-containing products. The label notice should be required for all products that contain sufficient amounts of the sugar alcohols to induce gastrointestinal problems in children and other sensitive people.

B. Legal Grounds

    FDA is authorized under sections 201(n), 403(a)(1), and 701(a) of the FFDCA (21 U.S.C. §§ 321(n), 343(a)(1), and 371(a)) to require label notices on food products. Section 403(a)(1) states that a food is misbranded if its labeling is false or misleading in any particular. Under section 201(n), FDA determines whether labeling is misleading by examining, among other things, the extent to which the labeling fails to reveal facts material as to consequences that may result from use of the product under conditions of use prescribed in the labeling or under customary or usual conditions of use. Section 701(a) generally authorizes FDA to issue regulations for the efficient enforcement of the FFDCA. FDA has relied upon its authority under those sections of the FFDCA to require label notices that alert consumers to the potential health hazards posed by certain foods and food ingredients.(41)

   Of course, FDA already has exercised that authority to promulgate the existing label-notice requirements for sorbitol- and mannitol-containing products. FDA could revise those requirements as requested by CSPI without exercising any new or additional authority under the FFDCA. As already explained, CSPI only asks that FDA revise the current requirement to take more recent scientific evidence into account and to provide consumers with a more meaningful label notice. The statutory authority supporting the existing label-notice requirement also would support the requested changes.

   That authority also provides clear support for the establishment of such requirements for products that contain maltitol, isomalt, and HSH.


   This petition is categorically excluded from the requirement for an environmental assessment under 21 C.F.R. § 25.32(k), because it requests the "[e]stablishment or repeal by regulation of labeling requirements for marketed articles" for which "there will be no increase in the existing levels of use or change in the intended uses of the product or its substitutes." In any event, CSPI does not believe that the actions requested in this petition would have any environmental impact.


   The existing labeling requirements for sorbitol- and mannitol-containing products do not adequately protect the health of those consumers, including children, who are susceptible to the potentially severe gastrointestinal problems caused by the substances. By adopting the changes urged by CSPI, FDA would help ensure that such consumers do not inadvertently consume products that contain sufficient sorbitol or mannitol to cause diarrhea or other problems. FDA should fulfill its responsibility to revise its outdated labeling requirements to reflect more recent clinical science.

   FDA should also take regulatory action to ensure that other sugar alcohols, including maltitol, isomalt, and HSH, can be safely used by consumers. The agency already has developed regulations that allow manufacturers to tout the potential health benefits of those substances (as well as sorbitol and mannitol); it should now protect susceptible consumers by requiring an appropriate statement concerning the substances’ gastrointestinal effects.


   The undersigned party certifies that, to the best knowledge and belief of the undersigned, this petition includes all information and views on which the petition relies, and that it includes representative data and information known to the petitioner which are unfavorable to the petition.

Respectfully submitted,

Darren Mitchell
Staff Attorney, Food Safety Program*

*Diana Birkett, CSPI Research Assistant, provided significant assistance in the preparation of this petition.


1. 21 C.F.R. § 184.1835(e).

2. Ibid.

3. F.W. Ellis & J.C. Krantz, "Sugar Alcohols XXII. Metabolism and Toxicity Studies with Mannitol and Sorbitol in Man and Animals," J. Biol. Chem., Vol. 141, (1941), pp. 147-154 [hereinafter cited as 1941 Mannitol and Sorbitol Study]. That study was the basis for the 50-gram trigger for the sorbitol label-notice requirement. U.S. Department of Health and Human Services, Food and Drug Administration, "Mannitol and Sorbitol: Affirmation of GRAS Status of Direct Human Food Ingredients," Federal Register, Vol. 38, No. 143 (1973), p. 20047 [hereinafter cited as Mannitol and Sorbitol Gras Affirmation].

4. N.K. Jain, D.B. Rosenberg, et al., "Sorbitol Intolerance in Adults," American Journal of Gastroenterology, Vol. 80, No. 9, (1985), pp. 678-681 [hereinafter cited as 1985 Sorbitol Study].

5. M.S. Badiga, N.K. Jain, et al., "Diarrhea in Diabetics: The Role of Sorbitol," J. Am. College Nutrition, Vol. 9, No. 6, (1990), pp. 578-582 [hereinafter cited as 1990 Sorbitol Study]. This clinical study, like the other sorbitol studies discussed in this petition, did not include a control group fed a placebo solution. Presumably, researchers do not use control groups in studying the effects of sorbitol because of the very low likelihood that subjects will spontaneously develop gastrointestinal problems absent exposure to a known laxative during the short-duration experiment.

6. 1985 Sorbitol Study, pp. 678-679.

7. Jeffrey S. Hyams, "Sorbitol Intolerance: An Unappreciated Cause of Functional Gastrointestinal Complaints," Gastroenterology, Vol. 84, No. 1, (1983), pp. 30-33 [hereinafter cited as 1983 Sorbitol Study].

8. Ibid.

9. P. Vernia, C. Frandina, et al., "Sorbitol Malabsorption and Nonspecific Abdominal Symptoms in Type II Diabetes," Metabolism, Vol. 44, No. 6, (1995), pp. 796-799. Subjects in that study exhibited fewer symptoms of sorbitol intolerance at 10 and 20 gram doses than did those in the other studies discussed above. The authors suggest that the discrepancy may be due to, among other things, the failure of researchers in the earlier studies to maintain isomolar solutions of sorbitol, and the fact that those earlier studies involved more nonwhite subjects, who exhibit a greater frequency of clinically severe sorbitol intolerance. Ibid., p. 798.

10. U.S. Department of Health and Human Services, Food and Drug Administration, Medical Officer’s Consult Review, Olestra: Stool Composition Study No. FP148, (Feb. 10, 1998).

11. Ibid., p. 35.

12. Richard E. Hill and K. Ramananda Kamath, "Pink Diarrhoea," Medical Journal of Australia, (May 1, 1982), pp. 387-389.

13. Margaret L. Payne, Winston J. Craig, et al., "Sorbitol is a Possible Risk Factor for Diarrhea in Young Children," Journal of the American Dietetic Association, Vol. 97, No. 5, (1997), pp. 532-534 [hereinafter cited as Young Children Sorbitol Study]. In fact, the results from the clinical studies on adults discussed above suggest that even lower amounts of sorbitol may cause gastrointestinal problems in children. In those studies, from 10 to 20 grams of the substance was found to induce gastrointestinal symptoms in adults. Assuming that the adult subjects averaged 60 kg in weight (no weight data were provided in the articles), the dose necessary to induce diarrhea is approximately 0.17 to 0.33 g per kg body weight. Applying those results to a 15 kg child would suggest that a 2.5 to 5 gram dose of sorbitol could cause gastrointestinal problems.

14. Centers for Disease Control and Prevention, "Outbreak of Diarrhea Linked to Dietetic Candies -- New Hampshire," Morbidity and Mortality Weekly Report, Vol. 33, No. 35 (1984), pp. 494-495 [hereinafter cited as New Hampshire Outbreak Report].

15. Young Children Sorbitol Study.

16. Ibid., p. 533.

17. Ibid.

18. A list of sorbitol-containing food products, as well as the amount of sorbitol per serving for each product, is provided in Appendix A to this petition. The information in Appendix A was gathered by CSPI from product packages and nutritional information found on the Internet.

19. 1985 Sorbitol Study, p. 680; 1983 Sorbitol Study, p. 32.

20. K.R. Johnston, L.A. Govel, et al., "Gastrointestinal Effects of Sorbitol as an Additive in Liquid Medications," Am. J. Med., Vol. 97, (1994), pp. 185-191.

21. Ibid., p. 187.

22. We urge CFSAN to coordinate with FDA’s Center for Drug Evaluation and Research to ensure that all products containing more than one gram of sorbitol per serving or daily dosage bear a notice about gastrointestinal effects.

23. As explained above, the 50-gram threshold was based on data from a 1941 study of 12 healthy adults in which the laxative threshold for sorbitol was found to be 50 grams per day (833 mg per kg of body weight), as well as another undated study in which a 10 gram daily dose of sorbitol for one month did not affect human subjects. Mannitol and Sorbitol GRAS Affirmation, p. 20047. In the sorbitol GRAS affirmation, FDA also cited a study in which healthy infants and children fed 9.3 grams of the substance "remained unaffected except for the appearance of diarrheal stools in the younger group." Id.

24. Young Children Sorbitol Study (finding that nine percent of one-year-olds, 50 percent of two-year-olds, 64 percent of three-year-olds, 71 percent of four-year-olds, and 81 percent of five-year-olds had consumed sorbitol-containing products during the three-month period surveyed). Based upon those findings, and the positive correlation between sorbitol consumption and the occurrence of afebrile diarrhea in three-year-old children, the researchers concluded that monitoring the consumption of sorbitol-containing products by young children may be useful in some cases, and that the public could benefit from more education regarding sorbitol use. Id., p. 533.

25. Those products are: Dentyne Ice Spearmint Gum (2 grams of sugar alcohols (sorbitol, maltitol, and mannitol) per serving), Brown & Haley Extra-Strength Cinnamons (2 grams of sorbitol per serving), Fifty 50 Low Calorie Strawberry Spread (3 grams sorbitol per serving), Smuckers Sugar Free Hot Fudge Topping (5 grams sorbitol per serving), and Cary’s Sugar Free Maple Flavor Syrup (11 grams sorbitol per serving). Copies of the package labeling for those products is provided in Appendix B to this petition.

26. Copies of the package labels for those products are provided in Appendix C to this petition. Examples of product packages with difficult-to-see sorbitol label notices include: Fifty 50 Sugar Free Hard Candy, which places the statement at the end of a section labeled "Persons with Diabetes," where it may not be read by nondiabetics; Frutay Peppermint Drops, which places the statement in small print on the back of the package; Murray Sugar Free Lemon Sandwich Cookies, which places the statement in small (albeit bold) print under the ingredients list; and Ricola Pearls Mountain Herb Breath Mints, which also places the statement in small print under the ingredients list.

27. The two exceptions are Certs Powerful Mints with Retsyn Crystals and Frutay Peppermint Drops, both of which contain sorbitol but are not labeled as sugar-free. No nutritional information, including the amount of sorbitol per serving, is provided on the Certs package.

28. A small number of products do indicate that sorbitol is the ingredient that may cause the laxative effect by linking the word "sorbitol" in the ingredients list to the label notice by means of asterisks. An example of such a label is provided in Appendix D to this petition.

29. Some sugar-alcohol-sweetened products already contain such advice on their packaging. For instance, packages of Estee Smart Treats Sugar Free Rice Crunchy Bars, which contain the sugar alcohol maltitol, provide the following label notice: "Maltitol may cause a laxative effect in children and other sensitive individuals when excess amounts are eaten at a time. We recommend starting with no more than one serving of rice crunchy bars at one time" [emphasis added].

30. H.K. Oberrieder & E.B. Fryer, "College Students’ Knowledge and Consumption of Sorbitol," J. Am. Diet. Assoc., Vol. 91, No. 6, (1991), pp. 715-717 (calling for better product labeling and education regarding sorbitol, after finding that less than half of surveyed college students had heard of sorbitol, and over 80 percent of those who had heard of the substance were unaware of its side effects); Young Children Sorbitol Study; 1990 Sorbitol Study, p. 581; Ray A. Breitenbach, "’Halloween Diarrhea’: An Unexpected Trick of Sorbitol-Containing Candy," Postgraduate Medicine, Vol. 92, No. 5, (1992), pp. 63-66 [hereinafter cited as Halloween Diarrhea]; New Hampshire Outbreak Report.

31. New Hampshire Outbreak Report.

32. Halloween Diarrhea.

33. 21 C.F.R. § 1.24(a)(4).

34. New Hampshire Outbreak Report.

35. 1941 Mannitol and Sorbitol Study, p. 151.

36. 21 C.F.R. § 180.25(e); Mannitol and Sorbitol GRAS Affirmation, pp. 20046, 20047.

37. 1941 Mannitol and Sorbitol Study, p. 151.

38. There is less clinical evidence available concerning mannitol’s gastrointestinal effects at lower doses than there is for sorbitol. In the absence of such evidence, it seems reasonable to apply the same conservative standard to mannitol-containing products that CSPI proposes for products that contain sorbitol.

39. Copies of those product packages are provided in Appendix E to this petition.

40. GRAS affirmation petitions have been submitted to FDA for maltitol, maltitol syrups, isomalt, HSH, and HSH syrups. U.S. Department of Health and Human Services, Food and Drug Administration, "Food Labeling: Health Claims; Sugar Alcohols and Dental Caries; Final Rule," Federal Register, Vol. 61, No. 165 (1996), p. 43436. Those petitions apparently are still under consideration by the agency. Id. Although FDA has promulgated regulations allowing the use of health claims about the relationship between dental caries and those substances (as well as sorbitol and mannitol), the agency has not required a label notice stating that the substances may cause gastrointestinal problems.

41. U.S. Department of Health and Human Services, Food and Drug Administration, "Food Labeling: Warning and Notice Statements; Labeling of Juice Products; Proposed Rule," Federal Register, Vol. 63, No. 79, (1998), p. 20487.