IPCS INCHEM Home

    XYLITOL

    Explanation

         Toxicological monographs were issued in 1977 and 1978 (see Annex
    I, Refs. 44 and 49). Since the previous evaluation, additional data
    have become available and are summarized and discussed in the
    following monograph.

    BIOLOGICAL DATA

    BIOCHEMICAL ASPECTS

    Absorption, distribution and excretion

         Sixty Wistar albino rats were gradually adapted to 20% dietary
    xylitol. Fully adapted and control rats were fasted overnight and
    dosed by oral intubation with 5 Ci of U-14C xylitol, U-14C sorbitol
    or 1-14C mannitol. Each isotope dose was mixed with corresponding
    ("cold") polyol to obtain a final dose of 0.625 g/kg bw. Tail vein
    blood samples were obtained and counted. In several experiments
    0.5 mol of calcium were given together with xylitol.

         Xylitol adapted rats did not exhibit diarrhoea following the dose
    administration. There was a significant increase in peak xylitol blood
    levels as determined by radioactivity in xylitol adapted rats as
    compared to controls. Xylitol adaptation also enhanced sorbitol and
    mannitol absorption when compared to controls. Calcium caused an
    initial rise in blood levels of radioactivity but the effect was not
    statistically significant (Salminen, 1982).

    Pig

         Six male castrated pigs (39-75 kg weight range) were subjected to
    feeding trials based on a Latin square design with either a polyol
    mixture (a by-product of xylitol production containing xylane) or
    xylitol supplementations. The transition period between diets was five
    days, preliminary period seven days, and collection period seven days
    (seven days feeding per diet). The basic diet consisted of skim milk
    powder with minerals and vitamins, to which polyol mixture (levels of
    5 or 2.5% dry matter) or xylitol (2.5 or 5% dry matter) were added.
    Wheat starch (5.0 or 2.5%) served as control supplement. Faeces and
    urine were collected twice a day and frozen until analysed. Venous
    blood samples were obtained one, two, or four hours after feeding.
    Glucose, plasma insulin and various clinical chemical parameters were
    determined.

         There was a slight decrease in the nitrogen balance of diets
    supplemented with 10% of polyol mixture or 5% of xylitol. There was no
    detectable xylitol or sugar alcohol in the faeces; a small quantity of

    xylitol was found in the urine of pigs when fed polyol mixture but not
    when fed xylitol. There was a significant rise in plasma glucose
    levels in xylitol fed pigs. Urine N decreased slightly in polyol or
    xylitol fed animals. Albumin concentration was significantly raised.
    There were increases in plasma alanine and asparate transferases
    (transaminases) (ALAT and ASAT also called serum glutamic pyruvic and
    glutamic-oxaloacetic transaminases (SGPT and SGOT)).

         The ALAT and SGPT levels increased significantly in a dose-
    related manner and indicated possible liver toxicity. Only the high
    dose level was statistically significantly different from the control.

         There were increases in insulin concentrations following xylitol
    feeding, and values two hours after feeding were higher than control
    levels. This increase was also dose-related. The peak of insulin
    levels was between 40-60 minutes after feeding (Nasa & Tanhuanpaa,
    1981).

    Metabolism

         Wistar albino rats were either adapted to 20% xylitol diets or
    given a 20% xylitol diet without adaptation or given a control diet.
    These rats were placed in individual metabolism cages, and urine and
    faecal samples collected. Body weight and levels of urinary pH,
    sodium, potassium, calcium, and oxalic acid were determined. Weight
    gains of non-adapted rats were smaller than those of adapted or
    control rats. Urinary output of non-adapted rats was decreased
    initially, but when diarrhoea disappeared the urinary volume returned
    to the control range. Urinary pH was somewhat reduced. Urine
    osmolality was decreased in xylitol treated rats. Higher net stool
    weight was seen in both dietary groups.

         Excretion of Calcium was significantly increased in xylitol
    treated rats. A smaller increase was seen with sodium excretion
    (Salminen, 1982).

    Metabolism in diabetic versus normal rats

         The effect of xylitol in Long-Evans rats was compared to
    fructose, glucose or no carbohydrate supplement on levels of ascorbic
    acid, ketonic metabolites and certain serum hormones in the normal and
    streptozotocin-diabetic state. Liver glycogen levels were determined.
    The livers of rats not treated with streptozotocin, glucose or xylitol
    (4% dietary level) contained significantly smaller glycogen levels
    than fructose or control groups. In xylitol treated rats, serum
    insulin levels were slightly decreased as compared to controls and
    serum glucagon was markedly depressed as compared to the other groups.

         The liver total ascorbic acid concentrations in untreated rats
    were significantly smaller than in fructose or xylitol fed rats.
    Rendering the rats diabetic with streptozotocin produced similar
    metabolic consequences regardless of dietary carbohydrates tested. The
    above result is in contrast with previously observed antiketogenic
    effects of fructose and xylitol (Hamalainen & Makinen, 1982).

    TOXICOLOGICAL STUDIES

    Special studies on the occurrence of adrenal medullary hyper and
    neoplasia in diets containing xylitol

    Rat

         A review of the available adrenal sections from a two-year
    toxicity study in rats (Hunter et al., 1978) was carried out.
    Additional sections were also cut from stored paraffin blocks. It was
    concluded that xylitol caused a significant increase in the incidence
    of adrenal medullary hyperplasia in male and female rats in all dose
    levels tested (5%, 10% and 20%). There was an increased incidence of
    pheochromocytomas (diagnosed on the basis of morphological but not
    functional effects) in males and females in the 20% xylitol group.
    However, this increase was not statistically significant. No
    pheochromocytomas were diagnosed at the lower levels. Similar effects
    were observed in rats fed comparable levels of sorbitol (Russfield,
    1981).

         A study was carried out to determine if the diagnostic criteria
    used in human pathology were also applicable for the diagnosis of
    pheochromocytomas in rats. In this study, groups of 100 and 200 male
    Wistar rats were used. The animals were maintained on standard stock
    diet. The first group of 100 was sacrificed at 24 months, and the
    second group of 200 at 30 months. Systolic blood pressure was measured
    in rats group 1 at age 22-23 months, and in group 2 at an age of 27-28
    months. Two weeks later, three-day urine samples were taken and
    analysed from creatinine as well as 4-hydroxy-3-methoxy mandilic acid
    (VMA - the major metabolite of both adrenaline and non-adrenaline in
    man) and 4-hydroxy-3-methoxyphenylglycol (MHPG) - the principal
    catecholamine metabolite in rats. Surviving rats were sacrificed at 24
    and 30 months of age (group 1 = 70 animals, group 2 = 45 animals), and
    autopsied. Adrenals of animals dying before sacrifice were also
    examined. In addition to the normal microscopic examination, the
    adrenals were examined histochemically for the presence of
    catecholamine-containing granules. Neither elevations in the
    measurement of blood pressure nor the excretion of VMA and MHPG
    revealed the presence of pheochromocytoma of the adrenal medulla.
    Hyperplastic and neoplastic lesions showed little or no chromaffinity.
    These data show a lack of functionality of the tumours (Bosland &
    Baer, 1981).

         A recent literature review of pheochromocytomas in rats indicates
    that the tumour is not unusual for the species, and that the nature of
    the diet, cage conditions, and hormone imbalance may affect the
    prevalence, which may be quite variable even within a specific strain
    (Cheng, 1980).

    Statistical re-evaluation of a two-year dog study

         A statistical re-evaluation was carried out on xylitol versus
    control groups in the chronic feeding study in dogs. The major
    findings in the feeding study were increased liver weight, a rarefied
    appearance of periportal hepatocytes in some of the treated dogs, and
    an increased group mean values of serum enzymes including SGOT, LDH,
    SAP and SGPT, as well as serum protein, serum albumin and cholesterol
    (Heywood et al., 1981). The re-evaluation demonstrated that liver
    weights were not significantly increased in xylitol-treated animals in
    either the combined male and female data or when the sexes are
    analysed separately. A minimum effect was observed at the 20% levels
    for males. Total serum protein, serum albumin, SGOT, LDH and
    cholesterol did not show a significant difference between groups that
    could be related to treatment. Although group mean SAP and SGPT levels
    in the 20% xylitol group showed a trend towards minimal increases from
    weeks 12 to 46 and from weeks 25 to 100, respectively, individual
    increases of enzyme levels above normal range contributed to this
    observation. However, these changes were not persistent, progressive
    or systematic. The histological changes reported (rarefied appearance
    of periportal hepatocytes) in a proportion of the dogs receiving 10%
    or 20% xylitol for two years, was also reported in all other treatment
    groups (Br & Christeller, 1980).

         Comments on the statistical re-evaluation provided an additional
    analysis of the statistical method used for re-evaluation. It was
    concluded that there was a significant difference between the liver
    weights of the animals in the 20% xylitol and 20% starch group. There
    was no evidence for a treatment-related progressive shift to abnormal
    values for total serum protein and serum albumin. There was agreement
    with the analysis used for evaluating serum levels for liver enzymes,
    but it was pointed out that because of the erratic fluctuations in
    levels shown by some animals in all treatment groups it was difficult
    to provide a coherent interpretation of these data. Finally, there was
    no evidence of a relationship between the parameters considered when
    this is looked for within treatment groups, but there is some evidence
    for a relationship between SGPT and the rarefied appearance of
    periportal hepatocytes when the differences between treatment groups
    are taken into account (Chanter, 1981).

         In another report commenting on the hepatic changes observed in
    the two-year xylitol Study in the dog, it was stated that a rarefied
    appearance and occasional slight enlargement of periportal hepatocytes
    was observed in 3/12 dogs in the 10% xylitol group, and in 5/12 dogs

    in the 20% xylitol group. There was no evidence of necrosis or
    degeneration of the liver cells. It was concluded that glycogen
    accumulation as indicated by electron microscopy in slightly enlarged
    periportal hepatocytes was associated with the marginal liver weight
    increase, and that the enzyme leakage was due to altered membrane
    permeability of the distended cell (Prentice, 1980).

    Special studies on the effects of caecal flora

         Wistar albino rats were either fed a control diet or adapted to
    20% xylitol. The ability of caecal suspensions from control and
    xylitol adapted rats to metabolize xylitol was assessed by
    determination of xylitol concentration and pH changes.

         The adaptation to 20% dietary xylitol increased the ability of
    caecal flora preparations to utilize xylitol in vitro as the sole
    carbon source; this was principally seen among anaerobic bacteria.
    Aerobic incubations with xylitol as the sole carbon source did not
    appear to promote acid production whereas aerobic incubations of these
    preparations with glucose resulted in rapid acid production (Salminen,
    1982).

         Faecal samples of control, xylitol (20% diet) adapted rats or
    rats receiving 20% xylitol, without prior adaptation, were collected
    and homogenized with deaerated Ringer's medium. A 20% suspension was
    centrifuged and standard loopfuls of caecal suspensions were heat
    fixed and stained. The relative amounts of Gram positive and Gram
    negative bacteria were estimated. The dry weight and moisture content
    of faeces were determined. Anaerobic organisms were isolated in stored
    media. Xylitol was determined by high pressure liquid chromatography.

         There was an initial decrease in pH of faeces when dietary
    xylitol was fed to unadapted rats, but the pH returned to normal
    during the adaptation period. The gradual adaptation was associated
    with a gradual increase in the relative proportion of Gram positive
    bacteria from 25 to 30% of faecal bacteria count to 70% when fully
    adapted. There was a statistically significant decrease in the numbers
    of aerobic streptococci.

         To investigate the cause of xylitol induced dose-related
    diarrhoea, caecal bacterial flora were investigated. Using a rapid and
    sensitive radioisotope bioassay, in which 14CO2 production from
    i.v.-14C labelled xylitol was measured, it was possible to show that
    caecal microflora obtained from rats can metabolize xylitol. This
    activity was increased 10-, 15-, 30- and 40-fold in caecal flora taken
    from rats fed diets containing 2, 5, 5, 10 and 20% xylitol
    respectively (Krishnan et al., 1980 a,b).

    Special studies on oxalate formation and metabolism

    Oxalate formation - mice

         Twelve groups of 8 BLU-HA male weanling mice received one of six
    different dietary levels of xylitol (0, 10, 12.5, 17.5 and 20%) with
    (20%) or without (O) fructose at each xylitol level. The mice were
    adapted to the xylitol diets by daily increment of 2.5% each day until
    the specified level was reached. The mice were sacrificed after 22
    days of feeding, and liver, brain, bladder, right kidney and right
    thigh muscle were collected.

         The presence of xylitol in the diet resulted in a slight but
    significant increase in weight gain. Fructose had no effect on weight
    gain. There was a significant effect of xylitol on oxalate levels in
    brain and muscle but not in the liver. There was no consistent
    response of oxalate levels to xylitol dose. Fructose also had a
    significant effect on brain and muscle oxalate levels, but again there
    was no consistent trend. Less than half of the kidney samples had
    measurable levels of oxalate and none of the bladder samples had
    detectable levels of oxalate (1.0 g/g wet weight) (Barngrover, 1982).

    Metabolic study of oxalate formation in pyridoxine deficient mice

         Five groups of four BLV/HA male weanling mice (initial weight of
    10 g) were adapted to 10 or 20% xylitol administered in a semi-
    synthetic diet with glucose beginning at 5% xylitol level with 45%
    glucose and increasing the xylitol level by 5%, while decreasing
    glucose by 5% stepwise every other day until the desired xylitol
    dietary level was achieved. Two control groups received glucose only.
    Three of the groups received similar diets which were deficient in
    pyridoxine. The diets were fed for 25 days to establish a pyridoxine
    deficient condition; urine and faeces were collected for one week, and
    body weights were determined weekly. The mice were then sacrificed and
    the livers collected. The pyridoxine adequate group fed glucose only
    as a sugar source grew best (final mean weight 25 g). The three
    xylitol groups (10%, 20% dietary levels, pyridoxine deficient; 20%
    pyridoxine adequate) gained less weight and had approximately similar
    mean weights (20 g); the glucose, pyridoxine deficient group did not
    gain weight and had three deaths before the end of the study. The 20%
    xylitol, pyridoxine deficient group excreted the highest level of
    urinary oxalate, with the 10% xylitol deficient and 20% xylitol,
    pyridoxine adequate group excreting intermediate amounts and the
    glucose, pyridoxine adequate group the lowest levels, whereas the two
    groups fed 20% xylitol had higher urinary levels (Barngrover, 1982).

    Oxalate formation - rats

         Six groups of three Sprague-Dawley male weanling rats received
    semi-synthetic diets for 28 days; four of the groups were pyridoxine

    deficient. Diets contained either 50% glucose by weight or 30% glucose
    and 20% fructose. On day 28 the rats were injected three times at
    spaced intervals with either 15% glucose; 10% xylitol + 5% glucose;
    15% fructose; or 10% xylitol + 5% fructose. Urine was collected on
    days 28 and 29 and rats were sacrificed and livers collected 30
    minutes after a final injection on day 30.

         The final body weights for the rats show that the group with the
    best growth received fructose + xylitol injections and was pyridoxine
    adequate (mean weight 216.6 g) versus the group receiving fructose +
    fructose injection (15%) and were pyridoxine deficient (119.7 g). The
    poorest growing groups received glucose diets only and either xylitol
    or glucose and glucose injections, and were pyridoxine deficient
    (107.1 g, 103.1 g).

         The rats on the pyridoxine deficient diets tended to excrete more
    oxalate and have higher liver oxalate levels. Within the pyridoxine
    deficient group only, rats injected with xylitol tended to excrete
    more oxalate and have higher liver oxalate levels, but these
    differences were not significant. Fructose had no effect on oxalate
    excretion or liver oxalate levels in the rats injected with xylitol
    (Barngrover, 1982).

    Special studies on reproduction

         A three-generation reproduction study was conducted in NMRI mice.
    Initially groups of 12 females and three males were allocated to a
    control group and a group adapted to 20% xylitol. Diets and water were
    constantly available. Body weights were recorded weekly.

         No abnormalities of condition or behaviour were observed in any
    of the successive generations of the control or xylitol treated
    groups. However, the body weights of xylitol treated animals at birth
    were decreased as compared to controls and significantly lower weight
    gains were observed in the Fo litters of xylitol fed animals. Even
    though the growth rates were lower in xylitol dosed litters, no
    significant differences were noted in food consumption after weaning.
    There were no significant differences in mean numbers of pups per
    litter. The mean birth weights were similar in both groups and minor
    variations were observed only in relation to litter size. No treatment
    related differences in mortality figures were observed during
    lactation periods. Gross examination revealed no abnormalities or
    differences attributable to xylitol treatment. Only slight increases
    in the caecum size were observed in xylitol treated mice (Salminen,
    1982).

    Special studies on the effect of xylitol on absorption and excretion
    of oxalic acid

    Mouse

         Male (30) and female (20) CD-1 mice were either gradually adapted
    to 20% xylitol diets or fed a control diet. After a 12-hour fast the
    mice received a single oral dose of 2 Ci of U14C-oxalic acid in
    water or in an xylitol solution (a total dose of 0.625 g/kg bw). For
    five xylitol adapted male mice the oxalic acid dose was given with
    sorbitol (0.625 g/kg bw) and for another group of five with mannitol
    (0.625 g/kg bw). Urine and faeces were collected at intervals for 72
    hours to monitor the excretion of the label.

         A similar study was conducted with 40 male and 40 female NMRI
    mice. In this study, samples of the intestine, kidney, liver and brain
    were also analysed for radioactivity.

         Adaptation of male mice to 20% dietary xylitol increased the
    urinary excretion of the label fourfold (4.5 versus 20%). No major
    changes were seen in faecal excretion. Both sorbitol and mannitol
    increased the urinary excretion of the label while only sorbitol also
    affected faecal excretion of the label.

         Urinary excretion of oxalic acid was significantly higher in
    xylitol adapted mice when compared to controls receiving oxalic acid
    only.

         Even greater urinary recovery of label was observed in control
    mice receiving oxalic acid with xylitol. In female mice xylitol
    appeared to induce an even more pronounced increase in oxalic acid
    excretion (Salminen, 1982).

    Rat

         Diets containing 20% of xylitol or one of the following
    carbohydrates: glucose, fructose, sucrose, xylose, sorbitol or
    mannitol were fed to groups of five Wistar rats for seven days. The
    rats were fasted for 12 hours and given a 5 Ci dose of U14C-oxalic
    acid mixed with 0.625 g/kg of xylitol or respective carbohydrate.
    Urine and faeces were collected for 72 hours and counted for recovery
    of activity. Ten rats were gradually adapted to 20% xylitol diets.
    After a 12-hour fast these rats and 20 controls received 5 Ci of
    U14C-oxalic acid mixed with water only or together with 0.625 g/kg
    xylitol/body weight. Urine and faeces were collected from five
    rats/group; tail vein blood from the other rats at intervals up to 24
    hours. Urinary excretion of the label was virtually identical in all
    groups. The mean excretion of label in faeces of control rats
    receiving oxalic acid was significantly lower (P <0.001) than in
    control rats receiving oxalate alone, or xylitol adapted rats

    receiving oxalate with xylitol (faecal recoveries were 77.8 and 83%
    respectively). The urinary excretion of label was also significantly
    higher among control rats receiving oxalate with xylitol when compared
    to control rats receiving oxalate alone. However, xylitol adapted rats
    excreted a significantly smaller proportion of oxalate in urine
    compared to controls receiving oxalate alone.

         The mean plasma levels of radioactivity in control rats receiving
    oxalic acid with xylitol were significantly higher (P <0.05)
    immediately after the start of the study when compared to controls
    receiving oxalic acid with water only or xylitol adapted rats.

         When samples of plasma, urine and faeces were analysed by use of
    thin layer chromatography, the major part of the radioactivity was
    recovered as oxalic acid (Salminen, 1982).

    Acute oral toxicity

         The acute oral toxicity of xylitol was determined in fasted NMRI
    mice in unadapted versus fully xylitol adapted mice (five
    mice/sex/dose group).

         Toxic signs consisted of staggering gait and a prone position.
    Slight diarrhoea was noted in adapted mice as compared to extensive
    diarrhoea in controls. The median lethal doses (LD50) were between
    20.96 and 23.62 g/kg with no statistical difference between sexes and
    whether or not adapted to xylitol diets. Death occurred in one to
    three hours. Necropsy revealed reddening of intestinal mucosa, swollen
    intestines and gas formation in Caecum (Salminen, 1982).

    Subchronic toxicity

         Groups of Wistar rats (70 rats/group) were allocated to diets of
    either 20% xylitol (two groups, one group was gradually adapted
    beginning with a 5% diet to prevent diarrhoea) or control diet (one
    group). Periodically rats were killed at intervals up to 150 days and
    subjected to gross necropsy. Sections of liver, kidney, adrenals,
    stomach, caecum and bladder were prepared for histopathological
    examination. Initially non-adapted rats showed a decrease in body
    weight, and decreased food consumption. Rats receiving xylitol without
    prior adaptation exhibited caecal enlargement. This was present to
    only a slight extent in adapted rats. There were small changes in the
    relative organ weights. The bladders of many rats showed one or more
    white precipitates. Histopathologically no changes were seen in liver,
    kidney, spleen, adrenals or stomach. Histopathologically many rats
    showed focal hyperplasia of the bladder wall associated with the
    precipitates (control 2/55, xylitol adapted 2/55, xylitol unadapted
    9/55). In the non-adapted rats showing diarrhoea inflammatory changes
    were observed on the bladder at the time Of diarrhoea (Salminen,
    1982).

    OBSERVATIONS IN MAN

    Effect of xylitol on urinary oxalate excretion in humans

         Five healthy human volunteers (two males and three females)
    received an orange flavoured drink containing 30 g xylitol with
    breakfast. Before dosing, urine was collected for 24 hours and
    collections were continued 24 hours after the dose of xylitol. During
    the collection period no foods rich in oxalate were permitted. No
    significant changes in urinary oxalate excretion could be detected
    (Salminen, 1982).

    Oxalate formation in human liver tissue

         The biochemical pathways for formation of oxalate after
    intravenous injection of xylitol in humans were studied using enzymes
    derived from human liver. It was concluded that metabolic pathways
    based on a combination of the transketolase, fructokinase, and
    aldolase reactions can account for the production of glucose, lactate,
    tertronates (D-threonic and D-erythronic acids) and oxalate
    (precursors) during the metabolism of xylitol administered
    parenterally (James et al., 1982).

    Special study on xylitol loading

         A study was carried out on nine subjects who had consumed xylitol
    for 4.8 to 5.3 years. During the years 1972-1974 these individuals
    consumed amounts ranging from 376-2520 mg/kg/day, and at the end of
    1977, from 46 to 354 mg/kg/day. In 1978 the diets of these individuals
    were loaded with 82.3 to 1400 mg/kg/day (females 70 g/day, males
    100 g/day) for 14 days using a strictly controlled diet, and
    subsequently for seven days while on a normal diet. During these
    periods, and also during period of normal diet + sucrose loading the
    following plasma and urinary parameters were measured. For serum;
    alanine aminotransferase, aspartate aminotransferase, alkaline
    phosphatase, gamma-glutamyltranspeptidase, lactate dehydrogenase,
    amylase: blood acid base balance. For urine; Uric acid, oxalic
    acid, 3-methoxy-4-hydroxymandelic acid, catecholamines (adrenalin,
    noradrenalin), metanephrines (m-o-methylnorepinephrine and
    m-o-methylnorepinephrine) urine: deposits, sediments and
    microcrystals, specific gravity, pH, U.V. and visible spectrum,
    volume; acid excretion in urine; urinary electrolytes; as well as the
    usual haematological, plasma and urinary parameters. There were no
    significant changes in any of the serum or urinary parameters measured
    (Makinen et al., 1981).

    Special studies on tolerance

         In another study the tolerance of increasing amounts of dietary
    xylitol in 13 healthy children, aged seven to 16 years was
    investigated. Xylitol was administered as a supplement in addition to

    the children's regular diet. The daily dose was increased during
    successive 10-day periods from 10 to 25, 45, 65 and 80 grams.
    Gastrointestinal symptoms (flatulence, occasional abdominal pain and
    diarrhoea) were recorded daily throughout the study. Prior to xylitol
    supplementation and after 20-50 days of dietary supplement serum uric
    acid and total cholesterol were measured. Flatulence was the most
    common side effect occurring relatively infrequently in almost every
    other subject during the 45 g/day intake, and in most subjects with
    greater frequency at the 80 g/day intake. Transient diarrhoea occurred
    in four children on 65 g xylitol/day and in one child at 80 g/day.
    After 50 days of xylitol consumption, there was an increase in serum
    uric acid and Cholesterol. However, the values were within the normal
    ranges for children (Akerblon et al., 1981).

    Comments

         Additional studies have been carried out in the metabolism of
    xylitol in the rat and pig. Xylitol adaptation in the rat increased
    the absorption of xylitol. Pigs fed xylitol showed a significant
    increase in plasma glucose, as well as a sharp rise in insulin levels.
    Administration of xylitol to experimental animals and man was shown to
    cause a change in the relative proportion of different bacteria
    normally present in the gastrointestinal tract. A three generation
    reproduction study in mice adapted to a 20% xylitol showed no
    significant compound-related effects.

         Metabolic studies on oxalate formation in mice indicate that
    diets deficient in vitamin B6 contribute to oxalate formation.
    Similar effects were observed in rats. Ca excretion was also
    increased, Studies in normal humans have shown that ingestion of
    xylitol is not associated with oxalate secretion. In a study in which
    human volunteers, who had been exposed to xylitol for several years
    received a single dose of xylitol, there was no evidence of increased
    oxalate excretion related to xylitol intake. Xylitol loading of the
    exposed individuals did not result in an increase in urinary oxalate
    excretion or calcium excretion. It is not known if marginal vitamin
    B6 deficiency in individuals would result in increased oxalate
    formation.

         The occurrence of adrenal medullary hyper- and neoplasia in rats
    fed xylitol has been the subject of an additional review. It was
    concluded that xylitol caused a significant increase in the incidence
    of adrenal medullary hyperplasia in male and female rats in all dose
    levels tested (5%, 10% and 20%). At a previous meeting (the twenty-
    sixth Joint FAO/WHO Executive Committee on Food Additives), the
    production of adrenal medullary hyperplasia in rats in a feeding study
    with 20% of sorbitol in the diet was considered. It was the view of
    the committee that such a high level of sorbitol produced gross
    dietary imbalance, which may produce metabolic imbalance, and
    considered that the adrenal medullary hyperplasia produced by high

    dietary levels of sorbital and certain other nutrients might occur as
    a physiological consequence of the stress induced in the aging rat. An
    increased incidence of pheochromocytomas was only observed in the 20%
    group, and this increase was not statistically significant. In a study
    on functional effects of pheochromocytomas in aged rats, neither
    raised blood pressure nor urinary excretion of the major metabolites
    of adrenaline or non-adrenaline revealed the presence of
    pheochromocytomas of the adrenal medulla. In addition the lesions
    showed little or no chromaffinity. Thus, the normal diagnostic
    criteria used in human pathology is not applicable to the diagnosis of
    pheochromocytomas in the rat. Further, since the occurrence of
    pheochromocytomas is species specific, and of grossly variable
    incidence in untreated rats, this toxicological significance to man
    cannot be assessed.

         Clinical studies in humans who had ingested xylitol for 4.3-5.3
    years showed no abnormal urinary parameters or blood pressure
    associated with adrenal changes.

         A statistical re-evaluation of the data on serum enzyme levels
    and liver weights derived from the two-year dog study indicate that
    there is a significant increase in liver weight of the dogs in the 20%
    xylitol group. However, although group mean values of SAP and SGPT
    showed some trend towards minimal increase, interpretation is
    difficult because of erratic fluctuations in the enzyme. Although
    electron microscopy demonstrates the presence of glycogen deposits in
    the liver of the test animals, no data are available on histochemical
    or other tests for glycogen, nor is it known if this effect is
    reversible. The significance of this hepatoxic effect is not known.
    Only a transient increase in liver weight was observed in the rat.

    EVALUATION

    Estimate of an acceptable daily intake for man

         ADI not specified.*

    *    The statement "ADI not specified" means that, on the basis of the
         available data (chemical, biochemical, toxicological, and other),
         the total daily intake of the substance, arising from its use at
         the levels necessary to achieve the desired effect and from its
         acceptable background in food, does not, in the opinion of the
         Committee, represent a hazard to health. For this reason, and for
         reasons stated in the individual evaluations, the establishment
         of a numerical figure of an acceptable daily intake (ADI) is not
         deemed necessary.

    REFERENCES

    Akerblom, H. K. et al. (1981) The tolerance of increasing amounts of
         dietary xylitol in children, Int. J. Vit. Nutr. Res. (In press)

    A. & Christeller, S. (1980) Chronic feeding study in dogs. Statistical
         re-evaluation of data on xylitol versus controls. Unpublished
         internal report from F. Hoffmann-La Roche & Co. Ltd., Basle,
         Switzerland. Submitted to WHO/FAO

    Barngrover, D. A. (1982) Xylitol Metabolism, An Alternative Pathway,
         Ph.D. Thesis, Cornell U., University Microfilms Int., Ann
         Arbor, Mich. No. 82110811

    Bosland, M. C. & Baer, A. (1981) Some functional characteristics of
         spontaneous adrenal medullary tumors in aged male Wistar rats
         (Internal report of CIVO Institute of Toxicology and Nutrition
         TNO, Zeist, the Netherlands and F. Hoffmann-La Roche and Co.
         Ltd., Basle

    Boum, B. et al. (1978) Action des extraits de Carica papaya sur un
         ictere experimental Cree chez le rat par des saponosides
         provenant du Brenania Brieyi, Tox. Appl. Pharmacol, 46,
         352-362

    Chanter, C. O. (1981) Comments on "Chronic Feeding Study in Dogs:
         Statistical Re-evaluation of Data on Xylitol vs Controls". Report
         from Huntingdon Research Centre. Submitted to WHO/FAO

    Cheng, L. (1980) Pheochromocytoma in rats, incidence, etiology,
         morphology and functional activity, Journal of Environmental
         Pathology and Toxicology, 4, 219-229

    Hamalainen, M. M. & Makinen, K. O. (1982) Metabolism of glucose,
         fructose and xylitol in normal and streptozotocin-diabetic rats,
         J. Nutr., 112, 1369-1378

    Heywood, R. et al. (1981) Revised report: Xylitol toxicity study in
         the beagle dog (Report of Huntingdon Research Centre)

    Hunter, G. et al. (1978) Xylitol tumorigenicity and toxicity study in
         long-term dietary administration to rats. Unpublished report from
         Huntingdon Research Centre, Huntingdon, Cambridgeshire, England,
         for F. Hoffmann-La Roche & Co. Ltd., Basle, Switzerland.
         Submitted to WHO/FAO

    James, H. M. et al. (1982) Models for the metabolic production of
         oxalate from xylitol in humans: A role for fructokinase and
         aldolase, Austral. J, Exp. Biol. Med. Sci., 60, 117-122

    Krishnan, R. et al. (1980a) Some biochemical studies on the adaptation
         associated with xylitol ingestion in rats, Austral. J. Exp.
         Biol. Med. Sci., 58, 627-638

    Krishnan, R. et al. (1980b) The effect of dietary xylitol on the
         ability of rat cecal flora to metabolize xylitol, Austral J.
         Exp. Biol. Med. Sci., 58, 639-652

    Makinen, K. K. et al. (1981a) Turku sugar studies XXII. A re-
         examination of the subjects, Int. J. Vit. Nutr. Res. (In press)

    Makinen, K. K. et al. (1981b) Turku sugar studies XIII. Comparison of
         metabolic tolerance in human volunteers to high oral doses of
         xylitol and sucrose after long-term regular consumption of
         xylitol, Int. J. Vit. Nutr. Res. (In press)

    Nasi, M. & Tanhuanpaa, E. (1981) The effects of sugar alcohols on
         metabolism of growing pigs, Acta Vet. Scand., 22, 344-354

    Prentice, D. E. (1980) Xylitol two-year study: Comments on hepatic
         changes (letter to Huntingdon Research Centre)

    Russfield, A.D. (1981) Two-year feeding study of xylitol, sorbitol and
         sucrose in Charles River (UK) rats: Adrenal Medulla. Unpublished
         report

    Salminen, S. J. (1982) Investigations of the toxicological and
         biological properties of xylitol. A thesis submitted in
         accordance to the requirements of the University of Surrey for
         the Degree of Doctor of Philosophy, Robens Inst. of Indust.
         Environ. Hlth Safety and Dept. of Biochem. U. of Surrey,
         Guildford, Surrey, UK
    


    See Also:
       Toxicological Abbreviations
       Xylitol (WHO Food Additives Series 12)
       Xylitol (WHO Food Additives Series 13)
       XYLITOL (JECFA Evaluation)