Prepared by the Fifty-first meeting of the Joint FAO/WHO
    Expert Committee on Food Additives (JECFA)

    World Health Organization, Geneva, 1999
    IPCS - International Programme on Chemical Safety


    First draft prepared by
    E.J. Vavasour
    Chemical Hazard Assessment Division, Bureau of Chemical Safety, Food
    Directorate, Health Protection Branch, Health Canada, Ottawa, Ontario,

          Biological data 
              Biochemical aspects 
              Toxicological studies 
                   Short-term studies of toxicity 
              Observations in humans 


         Polyglycitol syrup has not previously been evaluated by the
    Committee. Since its components are the same as those of maltitol
    syrup, differing only in the relative proportions of sorbitol,
    maltitol, and higher-order polyols, the evaluation conducted for
    maltitol syrup at the forty-ninth meeting (Annex 1, reference 131) to
    accommodate a wide range of starch hydrogenation products, would also
    apply to polyglycitol syrup.

         At the forty-ninth meeting, toxicological data in support of new
    specifications for maltitol syrup were reviewed. Since sorbitol and
    maltitol had previously been allocated ADIs 'not specified', only
    higher-order polyols required toxicological assessment. The Committee
    concluded at its forty-sixth meeting (Annex 1, reference 122) that
    proliferative lesions of the adrenal glands induced by some polyols
    (including sorbitol and maltitol) were a species-specific response and
    were not relevant to the toxicological evaluation of these substances
    for humans. Several short-term studies in which rats and dogs were
    given materials with a higher-order polyol content exceeding 80% of
    the syrup were reviewed at the forty-ninth meeting, as were data on
    their metabolic fate.

         At the present meeting, the Committee reviewed new data relating
    to digestibility  in vitro, a short-term study of toxicity in rats
    given material with a higher-order polyol content of 78%, and a study
    on the effects of hydrogenated starch hydrolysates on the glycaemic
    response in diabetic and non-diabetic individuals.


    2.1  Biochemical aspects

    2.1.1  Biotransformation

         An immobilized enzyme system of a series of carbohydrases
    (alpha-amylase, isomaltase, and maltase) was used to assess the
    relative digestibility of a group of hydrogenated starch hydrolysates,
    which included a 14:8:78 mixture, a 7:60:33 mixture, a 10.5:34:55.5
    mixture, and a 4-8:50-55:37-42 mixture. Although the polyol
    distributions of the 7:60:33 and 4-8:50-55:37-42 mixtures are similar,
    the higher-order polyols in the former tend to have a higher degree of
    polymerization than those in the latter. The hydrogenated starch
    hydrolysates produced were partially digested with the immobilized
    enzyme system, which resulted in a broadening of the peak
    corresponding to hydrogenated saccharides with a degree of
    polymerization greater than three, a decrease in the height of the
    maltotriitol peak, the appearance of glucose, and an increase in the
    height of the maltitol and sorbitol peaks for all of the hydrolysates.
    Markedly less sorbitol was produced from the hydrolysates that
    contained the highest percentage of highly polymerized saccharides,
    the 14:8:78 and 10.5:34:55.5 mixtures; the largest amount of glucose
    was produced by the hydrolysate with the highest percentage of highly
    polymerized saccharides, the 14:8:78 mixture. The overall
    digestibility of hydrogenated starch hydrolysates in this system, as
    measured by increased appearance of maltitol, sorbitol, and glucose,
    was highest for the products that contained the largest percent of
    highly polymerized saccharides, i.e. the first three mixtures. This
    was attributed to a higher affinity for or activity of the digestive
    enzymes for glucose-glucose bonds than those for glucose-sorbitol
    bonds (Allen & Watkins, 1997).

    2.2  Toxicological studies

    2.2.1  Short-term studies of toxicity

         A polyglycitol syrup, the 14:8:78 mixture, two maltitol syrups,
    7:60:33 and 7:52:41, and a metabolic control (sorbitol:glucose, 44:56)
    were fed to groups of 10 male and 10 female Crl.CD
    (Sprague-Dawley-derived) rats for 13 weeks. Two groups were fed
    untreated diet, and single groups were fed diets containing 6.7, 13.3,
    or 20% 14:8:78 mixture, 20% 7:60:33 or 7:52:41 mixture, or the
    metabolic control. The animals were checked twice daily for deaths and
    clinical signs and underwent a physical examination weekly. Body
    weights and food consumption were calculated on a weekly basis. An
    ophthalmoscopic examination was conducted before and at the end of the
    study. Blood samples were collected during the last week of the study
    for determination of a standard set of haematological and clinical
    chemical parameters. Overnight urine samples were also collected
    during the last week of the study for determination of urine volume,
    calcium ion concentration, and creatinine concentration. After
    sacrifice, all rats underwent gross necropsy, and 44 tissues and

    organs, including caecum, kidneys, and adrenal glands, were preserved
    for histopathological examination. The weights of the adrenals, brain,
    empty caecum, kidney, liver, testes, and thyroid/parathyroid were

         No deaths or clinical observations were noted that could be
    attributed to an effect of treatment. The body weights of the treated
    animals were similar to those of controls. The food consumption of the
    four groups receiving 20% liquid carbohydrate material was slightly
    increased in comparison with the control groups, but the difference
    was generally not statistically significant. The intakes of each of
    the substances in treated diets were as follows for males and females:
    4.2/4.9, 8.3/9.9, and 13/15 g/kg bw per day of the 6.7, 13, and 20%
    14:8:78 mixture; 13/15 g/kg bw per day of the 20% 7:60:33 mixture;     
    13/15 g/kg bw per day of the 20% 7:52:41 mixture; and 13/16 g/kg bw
    per day of the 20% sorbitol-glucose mixture. No effect of treatment
    was evident from haematological and clinical chemical examinations,
    including the values for serum calcium ion. Statistically
    significantly increased total calcium excretion, urinary calcium
    concentration, and urinary calcium:creatinine ratio over the values in
    both control groups were noted in both males and females. These
    parameters were also increased in a dose-related manner in males
    receiving 13% and females receiving 6.7 and 13% of the polyglycitol
    syrup. The absolute and relative caecal weights in both male and
    female rats in the groups receiving 20% liquid carbohydrate in their
    diet were increased relative to those of controls. No other
    treatment-related effect on organ weights were noted, nor was any
    effect apparent from histopathological examinations, in particular in
    the adrenals, kidneys, and caecum. Increased caecal weights have been
    observed previously in response to ingestion of polyol sugars by rats
    and are due to the accumulation of large amounts of undigested,
    unabsorbed material in the caecum. The increased fermentation by gut
    flora results in lowered pH, which promotes increased absorption of
    calcium from the gastrointestinal tract and subsequent excretion of
    the excess. Consequently, the effects observed in this study would not
    be considered adverse. The NOEL was 13 g/kg bw per day,  the highest
    dose tested (Goldenthal, 1997).

    2.3  Observations in humans

         The glycaemic availability of two hydrogenated starch
    hydrolysates, the 14:8:78 and 7:60:33 mixtures, was compared with that
    of glucose in non-diabetic persons and patients with
    non-insulin-dependent and insulin-dependent diabetes mellitus. Each
    group consisted of three men and three women, who ranged in age from
    31 to 69 years. The effects of ingestion of 50 g per 1.73 m2 of each
    substance were determined for a 5-h period, each person receiving one
    of the three oral challenges on successive days in a double-blind
    cross-over design. Peripheral venous blood samples were collected at
    0, 15, 30, 45, 90, 120, 150, 180, 240, and 300 min after ingestion of
    the test material for measurement of glucose, insulin, free fatty
    acids, and C-peptide. Alveolar breath samples for determination of H2
    were collected at 30-min intervals for 4 and 5 h after the test

    'meal'. The glycaemic response to the hydrogenated starch hydrolysate
    syrups was lower than that to glucose in both diabetic and
    non-diabetic subjects. Ingestion of the 7:60:33 mixture induced a
    lower glycaemic response than did the 14:8:78 mixture in all groups,
    but was most marked in the patients with non-insulin-dependent
    diabetes. Persons without diabetes and those with
    non-insulin-dependent diabetes released significantly less insulin in
    response to the syrups than to glucose, whereas patients with
    insulin-dependent diabetes showed no insulin response to any of the
    products. C-Peptide responses were increased in persons without
    diabetes and those with non-insulin-dependent diabetes in the order
    glucose > 14:8:78 > 7:60:33, reflecting differences in glucose
    levels resulting from absorption in the small intestinal. Reductions
    in plasma free fatty acid concentrations were also observed in
    response to the challenges with glucose, which elicited a greater
    response than the syrups. Breath H2 levels were unaffected by glucose
    challenge, but were increased by the hydrogenated starch hydrolysates
    syrups in the order 7:60:33 > 14:8:78, as a result of fermentation of
    unhydrolysed, unabsorbed material by the gut flora. These results were
    taken to indicate that the glycaemic response to hydrogenated starch
    hydrolysate syrups is lower than that to glucose in both diabetic and
    non-diabetic individuals. In addition to the lower content of glucose
    in the hydrogenated starch hydrolysate syrups, an inhibitory effect of
    sorbitol on glucose uptake was considered to affect the glycaemic
    response (Wheeler et al., 1990).


         Hydrolysis of a polyglycitol syrup composed of 14% sorbitol, 8%
    maltitol, and 78% higher-order polyols resulted in production of more
    glucose and less sorbitol than products conforming to the
    specifications for maltitol syrup, which is consistent with the
    relative proportion of glucose released on hydrolysis of each
    material. The greater digestibility of hydrogenated starch
    hydrolysates with the highest content of highly polymerized
    saccharides was considered to be the consequence of a greater activity
    or affinity of digestive enzymes for glucose-glucose than for
    glucose-sorbitol bonds. Data on the disposition of maltitol syrup
    reviewed at the forty-ninth meeting indicated that the higher-order
    polyols would be completely hydrolysed to glucose and maltitol or
    sorbitol, with a considerable portion undergoing fermentation by the
    gut flora. Bacterial fermentation was demonstrated by the detection of
    an increased amount of H2 in the breath of human subjects in a study
    reviewed at the present meeting.

         Inclusion of up to 20% of a polyglycitol syrup (containing
    14:8:78 sorbitol:maltitol:higher-order polyols) and two maltitol
    syrups (7:60:33 and 7:52:41) in the diet of rats, equal to 13 g/kg bw
    per day, for 13 weeks, was not associated with adverse effects. The
    only effects observed -- increased weight of the empty caecum and
    increased urinary calcium excretion in the absence of elevated serum
    calcium -- were considered to be the consequence of the accumulation

    of poorly absorbed material in the caecum and to be of no
    toxicological significance.

         In studies of diabetic and non-diabetic human subjects, ingestion
    of polyglycitol and maltitol syrups resulted in a lower glycaemic
    response than with glucose, in the order maltitol syrup
    < polyglycitol syrup < glucose. These results reflect the relative
    proportion of glucose released by hydrolysis of each material.


         On the basis of the data on hydrogenated oligo- and
    polysaccharides reviewed at the forty-ninth and the present meetings,
    the Committee allocated a group ADI 'not specified' to materials
    conforming to the specifications for polyglycitol syrup and maltitol


    Allen, J.C. & Watkins, R. (1997) In vitro digestibility of Hystar
    5875, Hystar 6075 and Lycasin 80/55. Unpublished report to Lonza,
    Inc., Fair Lawn, New Jersey, USA. Submitted to WHO by the
    International Diabetic Federation, Brussels, Belgium.

    Goldenthal, E.I. (1997) 13-Week dietary toxicity study in rats with
    Hystar(R) liquid carbohydrates products. Unpublished report from MPI
    Research, Mattawan, Michigan, USA, to Lonza, Inc., Fair Lawn, New
    Jersey, USA. Submitted to WHO by the International Diabetic
    Federation, Brussels, Belgium.

    Wheeler, M.L., Fineberg, S.E., Gibson, R. & Fineberg, N. (1990)
    Metabolic response to oral challenge of hydrogenated starch
    hydrolysate versus glucose in diabetics.  Diabetes Care, 13, 733-740.

    See Also:
       Toxicological Abbreviations