First draft prepared by
    Dr B. Priestly
    Department of Pharmacology
    University of Adelaide


         Thermally oxidized soya bean oil (TOSO) is made by oxidation of
    refined soyabean oil in air at 190-250 C. It is used as a
    tin-greasing emulsion in baking. It is also interacted with food
    fatty acids under vacuum at 130 C to form thermally oxidized soya
    bean oil interacted with mono- and diglycerides of fatty acids
    (TOSOM) which has been used as an emulsifier in oleomargarine for
    many years. TOSOM also imparts anti-spattering properties to
    oleomargarine when used for frying. TOSO and TOSOM were considered
    at the fifteenth, seventeenth, twentieth, and twenty-fourth meetings
    of the FAO/WHO Joint Expert Committee on Food Additives. (Annex 1,
    references 26, 32, 41, 53) under the name "esters of glycerol and
    thermally oxidized soy bean fatty acids". Specifications were
    tentatively assigned (Annex 1, reference 28) at the fifteenth
    meeting, but withdrawn at the seventeenth meeting when it appeared
    that the compound was a mixture and not a true ester. A
    toxicological monograph was prepared at the seventeenth meeting
    (Annex 1, reference 33) from data describing both the esterified and
    unesterified oxidized soyabean oil, but an ADI was not established,
    pending the evaluation of short-term and long-term studies on a
    material of well-defined composition.

         Since the previous evaluation, new data on metabolic
    disposition and short-term and long-term studies in the rat have
    become available and have been summarized and discussed in the
    following monograph. Relevant material from the previously published
    monograph has been included.


    2.1  Biochemical aspects

    2.1.1  Absorption, distribution, and excretion

         Groups of five male and five female rats received in their diet
    either 20% soybean oil or thermally oxidized material. Highly
    modified material reduced the absorption of dietary fat in
    proportion to the degree of modification introduced. Delay in
    absorption of modified oil material from the gastrointestinal tract
    was measured by residuals found in the gut. Similar delaying effects
    have been demonstrated on the presence of chylomicrons and fat in
    the intestinal lymph fluid. The passage of intestinal contents is
    delayed if modified oil is administered. Compared with normal oil
    there is an early increase followed by only slight decrease in bile
    flow following oral administration of modified oil. Intraperitoneal
    administration of modified oil increases the diuretic effect of
    intraperitoneal normal soybean oil. Study of the liver function
    after eight weeks feeding of modified oil showed retention of the
    bromo-sulphophthalein used as indicator compared with normal oil.
    The  in vitro effect of modified oil on the kinetics of various
    cellular enzyme systems showed generally no inhibition of oxidative
    metabolism (Kieckebusch  et al., 1962).

         Wistar rats, Tuck TO mice and Dunkin-Hartley guinea-pigs were
    dosed orally with 20 mg/kg bw 14C-labelled refined soyabean oil,
    200 mg/kg bw 14C-labelled Homodan O (a mixture of 2 parts soyabean
    oil and 1 part TOSO), or 200 mg/kg bw 14C-labelled Emulsifier MO
    (previously called Homodan MO and of equivalent specifications to
    the emulsifier TOSOM). Urine, faeces and expired CO2 were
    collected for up to three days.

         After dosing with the control soyabean oil, the radiolabel
    disappeared rapidly from the GI tract, there was a transient
    accumulation in the tissues (mainly liver, kidney, lung, fat and
    muscle) and most of the dose was recovered in the excreta within
    24 h. In rats and mice, recovery of 14CO2 accounted for most of
    the dose (50-70%), while in guinea-pigs, only 10-13% was recovered
    as 14CO2. Urinary excretion accounted for 2% in rats and mice
    and 4-8% in guinea-pigs. Faecal recovery accounted for the remainder
    (5-15% rats and mice; 70-80% guinea-pigs). There were no sex
    differences in disposition, and 7 days repeated administration of
    unlabelled material prior to dosing did not alter the excretory
    pattern. Only the liver, fat muscle and GI tract retained >1% of
    the dose after three days.

         The patterns of excretion of the 14C-labelled Emulsifiers O
    and MO (TOSO and TOSOM) were comparable to that of soyabean oil in
    the guinea-pig, but both rats and mice expired less 14CO2
    (11-26%) and excreted more urinary metabolites (6-12%, 7-9%,

    respectively). Faecal recoveries of radioactivity were 3-4-fold
    greater for each emulsifier when compared with refined soyabean oil.
    Tissue levels 1-4 h after dosing with TOSOM were approximately 3x
    those observed after comparable dosing with unesterified TOSO. The
    authors concluded that TOSOM is absorbed slightly better than the
    unesterified TOSO, and that thermal oxidation of the soya bean fatty
    acids results in retarded absorption in comparison with refined
    soyabean oil (Phillips  et al., 1978 a,b). The effect of thermal
    oxidation on absorption is consistent with that seen with other food
    oils. (Perkins  et al., 1970).

    2.1.2  Biotransformation

         No data available.

    2.1.3  Effects on enzymes and other biochemical

         No data available.

    2.2  Toxicological studies

    2.2.1  Acute toxicity

         No data available

    2.2.2  Short-term studies  Rats

         For 12 to 16 weeks groups of 20 male and 20 female rats
    received in their diets either 20% untreated soybean oil or 5%
    variously modified soybean oil plus 15% olive oil or 2.5% modified
    oil plus 17.5% olive oil. Only highly modified oils caused
    significant reduction in growth and food intake, and increased
    mortality. No definite effect was noted at the 2.5% level. Motor
    activity of animals receiving high doses of highly modified oil was
    increased compared with controls. The weights of major organs were
    similarly increased for the groups on the more highly modified oils.
    Gross and histo-pathology showed some pathological changes in the
    thyroid and kidney of the group receiving the most highly modified
    material (Kieckebusch  et al., 1962).

         Three groups each of six rats were treated for 17 weeks with an
    esterified product at 0, 0.084 and 0.84% of their basic diet. There
    was no significant effect on weight gain nor macroscopic appearance.
    Liver and stomach were examined histologically and found to be
    normal (Dam, 1952).

         Two groups of nine male rats received for 36 weeks feeds
    containing oleomargarine made with an esterified product at 0.3 and
    3% levels. No controls were used. No weight differences nor gross
    pathological changes were noted. The histology of kidneys, liver and
    small intestine was normal (Aaes-Jorgensen  et al., 1954).

         Groups of 20 Wistar rats of each sex were fed diets containing
    1, 5 or 10% of a commercial tin-greasing emulsion, for 91 days. The
    emulsion was specified only with regard to its content (1%) of TOSO.
    Intake of the emulsion was determined to be 0.88-0.97, 4.6-4.88, or
    9.38-10.16 g/kg bw/day in the three dosed groups, making the upper
    level of TOSO intake in this study approximately 100 mg/kg bw/day.
    There were no effects noted on growth, food intake, urinalysis,
    haematology or blood biochemistry. There were no effects on organ
    weight or histopathology. The Committee considered this study to be
    of minimal usefulness. The dose levels were too low to produce any
    toxicity, and only 10/20 rats from each group were subjected to
    haematological analysis, blood biochemistry or autopsy (Kemper
    1981).  Pigs

         Four groups of 4 female pigs received an esterified product for
    98 days at dietary levels of 0, 0.4, 2 and 10%. No significant
    effects were noted on growth rate, food consumption, blood
    parameters, liver and kidney function, serum chemistry, organ
    weights, nor gross and histopathology due to administration of the
    test compound (Gyrd-Hansen & Rasmussen, 1968).

    2.2.3  Long-term/carcinogenicity studies  Rats

         A three-generation study was performed using an average level
    of 15% esterified product in the diet. The parent generation (57
    females, 15 males) was observed for at least 24 months. Growth, body
    weight gain and appearance were similar to controls (seven males,
    seven females) receiving 15% soybean oil. Five animals in the test
    group developed tumours but none in the control group. There were 16
    survivors in the test group and two in the control group.

         The F1 generation (37 males, 37 females) was also observed
    for 24 months. Thirty-seven males and females received soybean oil
    as controls. Twenty-four test and 10 control animals survived two
    years. Two test animals and three controls developed tumours. The
    F2 generation (57 females, 27 males) was observed for over two
    years. Seventy males and females received soybean oil as controls.

    Four test and 19 control animals survived 24 months. No animals in
    the test group but three in the control group developed tumours.
    Neither the F1 nor the F2 generation showed significant
    differences from controls as regards growth, body weight gain, nor
    gross and histopathology (Harmsen, 1959, 1960).

         In another experiment an esterified product was fed orally at
    the rate of 3 g/day and injected s.c. to 29 rats at 1 ml weekly for
    five months and 2 ml bi-weekly for a further three months. Animals
    were observed for 27 months. No tumours developed. A control series
    of 30 rats treated similarly with oral esterified product and
    injected s.c. with refined soybean oil showed four tumours after 24
    months observation, none at the site of injection. The F1 test
    generation (nine males, nine females) received 3 g esterified
    product orally and 1 ml i.p. for eight weeks, followed by a further
    1 ml i.p. for four months. After 11 months six survived without any
    tumours being noted. The control group (14 males, 4 females)
    received 1 ml soybean oil s.c. for five months and 2 ml s.c. for
    three months. During the 29-month observation period, one rat
    developed a tumour at the site of injection (Harmsen, 1961).

         Rats fed with 20% of thermally oxidized soybean oil in the diet
    had a significantly longer life-span (807  32 days) than those fed
    with 20% of fresh soybean oil (673  42 days) (mean  S.E.): a
    similar difference was observed between other oxidized and fresh
    fats and oils (Kaunitz  et al., 1966).

         Groups of 60 Wistar rats of each sex were fed diets containing
    1.2% TOSO for 132-137 weeks. The test substance for these groups,
    designated TOSO(G), was provided by Grindstedvaerket A/S, Denmark
    and met official Danish specifications. Further groups of 60 rats
    were fed a diet containing 0.3 or 1.2% of TOSO(N), a material of
    similar specification, but obtained from Nexus ApS, Holland. Average
    intake of TOSO was 0.125 and 0.5 g/kg bw/day. Other groups were fed
    diets containing 3, 6 or 12% TOSOM for 132-137 weeks. The test
    substance was provided by Grindstedvaerket A/S, Denmark and met
    official Danish specifications. Average intake of TOSOM was 1.3, 2.7
    and 5.4 g/kg bw/day in males and 1.8, 3.6 and 7.4 g/kg bw/day in
    females. Wistar rats (120 of each sex) served as controls. All diets
    were adjusted to a total caloric value of 12% edible fatty acids by
    supplementation with mono- and diglycerides. There were minor
    variations in growth, food consumption and blood bio-chemistry
    throughout the study, but there was no discernible treatment-related
    effect. Survival was >50% in all groups at 2 years. A decrease
    (p <0.05) in total white cell count and percentage of lymphocytes
    and an increase in percentage of monocytes was observed at the
    highest TOSOM dose rate and in all groups receiving TOSO, but only
    in some of the blood samplings. However, there was no consistent
    dose or time pattern and, with the exception of an increased
    percentage of monocytes at weeks 10 and 38 in rats treated with
    TOSO, the parameters were within the normal range for Wistar rats.

         In rats treated with TOSO, the number of rats bearing malignant
    tumours was increased slightly (p <0.05) at the 1.2% dose level in
    male rats given TOSO(G) and in female rats given TOSO(N) 16/60 vs
    24/119; 23/60 vs 24/118 respectively), although there were no
    effects on the overall number of tumour-bearing rats, or in the
    number of rats with multiple tumours. The incidence of tumours at
    selected sites achieved statistical significance in the case of
    cortical adenomas in males which had received 1.2% TOSO(G) (5/60 vs
    1/119); adenocarcinoma of the uterus in females which had received
    1.2% TOSO(N) (7/60 vs 3/118) and subcutis fibrosarcoma in males
    which had received 1.2% TOSO(N) (5/60 vs 3/119).

         In rats treated with TOSOM, there were no treatment-related
    effects on the overall number of tumour-bearing rats, or on the
    number of rats with malignancies or multiple tumours. The incidence
    of mammary adenoma (7/60 vs 4/118) in females and adrenal medullary
    adenoma (4/59 vs 1/119) in males were the only sites at which a
    significant (p <0.05) increase compared to the controls occurred.

         There was no consistency across treatments, no evidence of
    reduced tumour latency, and the observed incidences were within the
    range of historical controls. The authors concluded that the effects
    were not related to dosing with either TOSO or TOSOM. There were no
    compound-related changes in the incidence of non-neoplastic lesions.

         In summarizing both the neoplastic and non-neoplastic data, the
    authors excluded the data from the oral cavity, then summarized it
    separately. It was suggested that a relatively high incidence of
    chronic inflammation and carcinomas  in situ or squamous-cell
    carcinomas at these sites in control and treated rats (14 of 119 in
    controls and 4 to 10 of 60 in treated groups) was not associated
    with the emulsifiers, but was apparently caused by penetrative
    lodgement of dietary oat and barley chaff particles in the nasal and
    pharyngeal cavities. The Committee was not convinced by this
    explanation but could see no reason to cast doubt on the authors'
    conclusion that neither TOSO nor TOSOM was responsible for the
    effect (Gry  et al., 1987).

    2.2.4  Reproduction studies

         No data available.

    2.3  Observations in humans

         No data available.


         The Committee considered studies showing that after oral
    administration of radiolabelled material to rats and mice, tissue
    levels of TOSOM were slightly greater than those of TOSO. The
    absorption of both thermally oxidized substances was lower than that
    of refined soya bean oil.

         A 2.5 year study in rats used dietary levels of 3, 6, or 12%
    TOSOM (equal to 1.3, 2.7 and 5.4 g/kg bw/day in males; 1.8, 3.6 and
    7.4 g/kg bw/day in females), and 0.3 or 1.2% thermally oxidized soya
    bean oil (from two sources) (equal to 130 and 540 mg/kg bw/day in
    males; 180 and 740 mg/kg bw/day in females). The only effects noted
    were transient minor variations in total leukocyte counts and
    percentages of lymphocytes and monocytes, which were not considered
    to be of toxicological significance. For some sites, the incidence
    of tumours in treated animals was slightly higher (P<0.05) than in
    controls. However, the incidences were within the range of those of
    historical controls for Wistar rats, there was no dose-response
    relationship, and there was no consistency of effect for the three
    compounds tested. The Committee concluded that neither TOSO nor
    TOSOM is carcinogenic in rats.


         Using a safety factor of 200, the Committee allocated an ADI of
    0-3 mg/kg bw for TOSO and 0-30 mg/kg bw for TOSOM, based on the
    observation that the highest doses in the 2.5 year rat study
    (approximately 600 and 6000 mg/kg bw/day respectively) produced no
    adverse effects.


    AAES-JORGENSEN, E.  et al. (1954) Unpublished report supplied by
    Grindstedvaerket Laboratoriet.

    DAM, H. (1952) Unpublished report submitted to WHO by
    Grindstedvaerket Laboratoriet.

    P., ROSWALL, K., THORUP, I. & WURTZEN, G. (1987). Thermally oxidized
    soya-bean oil interacted with mono- and diglycerides of food fatty
    acids (Esters of glycerol and thermally oxidized soybean fatty
    acids). A long-term study in rats. Report of the Institute of
    Toxicology of the National Food Agency of Denmark.

    GYRD-HANSEN, N. & RASMUSSEN, F. (1968). Short-term feeding study of
    the emulsifier Homodan MO in pigs.  Fd. Cosmet. Toxicol., 6: 163.

    KEMPER, F. (1981). Report on the investigation of the subchronic
    (91 day) oral toxicity of "CARLO" tingreasing emulsion in rats.
    Unpublished report of the Institute of Pharmacology and Toxicology,
    University of Munster. Submitted to WHO through the National Food
    Agency, Denmark, by Nexus ApS, Palsgaard, Denmark.

    KAUNITZ, H. , JOHNSON, R. E. & PEGUS, L. (1967). Longer survival
    time of rats fed oxidized vegetable oils.  Proc. Soc. exptl. Biol.
     Med., 123: 204.

    KIECKEBUSCH, K.  et al. (1962)  Fette, Seifen, Anstrichmittel, 64:

    PERKINS, E.G., VACHHA, S.M. & KUMMEROW, F.A. (1970). Absorption by
    the rat of nonvolatile oxidation products of labeled randomized corn
    oil.  J. Nutrition, 100: 725-731.

    PHILLIPS, J.C., TOPP, C.E., COOK, M & GANGOLLI, S.D. (1978a). The
    metabolic disposition of 14C-labelled soyabean oil and
    14C-labelled HOMODAN O in the rat, guinea-pig and mouse.
    Unpublished Report No. 185/2 from BIBRA. Submitted to WHO through
    the National Food Agency, Denmark, by Grindstedvaerket A/S,
    Brabrend, Denmark.

    PHILLIPS, J.C., TOPP, C.E., COOK, M. & GANGOLLI, S.D. (1978b). The
    metabolic disposition of 14C-labelled soyabean oil and
    14C-labelled EMULSIFIER MO in the rat, guinea-pig and mouse.
    Unpublished Report No. 185/1 from BIBRA. Submitted to WHO through
    the National Food Agency, Denmark, by Grindstedvaerket A/S,
    Brabrend, Denmark.

    See Also:
       Toxicological Abbreviations