Chlorine is used for the treatment of flour for special purposes,
    such as cake manufacture. Chlorine, as a flour treatment agent for
    special-purpose flour, was reviewed at the ninth meeting of the
    Committee (Annex 1, reference 11) but an acceptable level of use was
    not established. At the earlier review, it was concluded that "long-
    term studies using appropriate products made from flour treated with
    chlorine at various levels will be needed".

         Since the previous review, further studies have become available
    and are summarized and discussed in the following monograph.


    Biochemical aspects

    Interaction of chlorine with flour

         Treatment of flour with 1950 ppm of chlorine resulted in a
    lowering of the unsaturated fatty acids in the flour to 40% of
    levels in untreated flour. Oleic acid was probably converted into
    dichlorostearic acid while a range of chlorinated compounds was formed
    from linoleic and linolenic acids. Treatment of flour with up to
    120 ppm of chlorine did not materially change the major fatty acids
    (Coppock et al., 1960; Daniels, 1960).

         When soft-wheat flours were treated with chlorine, the chlorine
    content of the lipids was markedly increased, that of water-soluble
    components to a lesser extent, and of the gluten only slightly. The
    lipids and water-soluble components comprised only 5% of the flour,
    but contained more than 90% of the added chlorine. The chlorine-
    containing lipids showed a decreased iodine value (Gilles et al.,

         The chlorine content of untreated flour was found to be
    430-540 mg chlorine/kg flour and of bleached flour 1310-1890 mg
    chlorine/kg flour. Almost all of the additional chlorine was in
    water-solubles and gluten (including lipid), and treatment did not
    significantly increase the chlorine in the prime starch. At least 50%
    of the additional chlorine in the gluten was in the lipid fraction
    (Sollars, 1961).

    Absorption, distribution, and excretion

         Rats fed diets containing 4.1% lipids extracted from chlorine-
    treated flour showed a decrease in polyunsaturated fatty acids and a
    corresponding increase in palmitic, oleic, and palmitoleic acids in
    fat depots when compared with controls receiving 4.1% lipid from
    untreated flour. The chlorine content of adipose tissue was, however,
    only slightly increased. Effects of this nature could not be
    demonstrated at lower levels of dietary lipid more comparable with the
    amounts that might be ingested from treated flour (Daniels et al.,

         Oleic, linoleic, and linolenic acids, and their triglycerides,
    were chlorinated with 36Cl-chlorine and administered orally to male
    Wistar rats. Tritiated parent fatty acids acted as controls for
    comparisons of absorption, tissue distribution, and excretion of the
    compounds. Absorption of the chlorinated compounds was greatly reduced
    compared to the parent compounds and tissue deposition was generally
    greatly reduced, although deposition of chlorinated oleic acid in the
    heart was similar to that of oleic acid and deposition of chlorinated
    triolein was greater than that of unchlorinated triolein. Water-
    soluble chlorination products were readily absorbed, were not
    incorporated into body lipids or proteins, and were rapidly excreted
    in the urine (Cunningham & Lawrence, 1977a; 1977b).

         Single oral doses of tritiated oleic acid of 36Cl-chlorinated
    oleic acid were administered to groups of male rats which were then
    killed at periods of 1-28 days after administration. The chlorinated
    oleic acid was absorbed to a lesser extent than 3H-oleic acid (72.3%
    compared with 91.5%). The blood-brain barrier appeared effective
    against transfer of chlorinated oleic acid, but the compound was
    widely distributed throughout other tissues. The liver appeared to be
    the main site for dechlorination of the chlorinated oleic acid, the
    half-lives in the organs being 19.5 days in the kidney, 10.7 days in
    the liver, 10.0 days in the brain, 8.3 days in the heart, and 5.1 days
    in blood (Cunningham & Lawrence, 1976).

         Placental and mammary transfer of 36Cl-chlorinated oleic acid
    and tritiated oleic acid were studied after oral administration to
    rats. Transfer across the placenta of the chlorinated acid occurred at
    half the rate of the unchlorinated acid and 0.4% of the dose
    administered remained in the foetus after 19 days. Mammary transfer of
    chlorinated oleic acid to a 2-week-old suckling rat occurred to the
    extent of 15.9% of the parental dose in 24 hours, compared with 34.0%
    for unchlorinated oleic acid. The corresponding transfer of
    chlorinated linoleic and linolenic acids accounted for only 2.4% and
    2.7% of the dose administered, respectively (Cunningham & Lawrence,

    Toxicological studies

    Special studies on reproduction


         A multigeneration study was performed in which male and female
    Fischer rats received diets containing cake made from flours treated
    with 0, 1000, 1500, or 2500 ppm chlorine, dried to a moisture content
    of 6%, and incorporated into the diet at a level of 75%. An additional
    control group received commercial Purina chow. The study was continued
    through 3 successive generations. Selected animals of the F1a
    generation were used for long-term studies (see below) and the F3b
    litter of the cake control and top-treatment groups were used for a
    teratology study (see below). No adverse effects were seen in the F1a
    generation. In the F2a generation there were decreased numbers of
    live births in the highest-treatment group and an increased mating-
    parturition time in the mid- and highest-treatment groups (gestation
    time was not determined). Weanling weights decreased with increasing
    levels of chlorination, though weanling weights of all groups
    receiving cake were greater than those of the Purina chow control
    group. In the F3a generation, no effects attributable to treatment
    were observed; there was a non-treatment related decrease in the
    incidence of live births in cake-fed animals (Gumbmann & Gould,

         Female rats were fed a diet in which 93% of the total nutrients
    were provided by cake made from flour chlorinated at a level of
    10,718 ppm. They were mated with males receiving similar diets or
    unchlorinated cake diets. Each female produced 2 litters. The results,
    though poorly presented, indicated no adverse effects on the
    reproduction parameters monitored, viz: Number of litters, live and
    dead pups, and those surviving to weaning. However, a slight decrease
    in pup weaning-weight was noted. As a comparison, mating was performed
    between control males and females, and between males and females
    receiving 2200 ppm chlorinated cake diets. No differences were noted
    in reproduction parameters, with the exception of a slight reduction
    in pup weaning-weight (Fisher et al., 1979).

         Five groups, each containing 3 rats, were fed diets containing 0,
    0.82, or 4.1% untreated flour lipid, or 0.82 or 4.1% lipid from flour
    treated with chlorine at a level of 1950 ppm. At the 4.0% level, the
    groups receiving lipid from the treated flour had thinner and rougher
    fur, fertility was reduced, and lactation was less efficient than in
    rats from the control group. Similar depression of fertility and
    lactation was observed with the group receiving 0.82% treated lipid.
    These effects were observed consistently through 4 generations and
    were not relieved by feeding 2% linoleic acid for 8 weeks (Daniels
    et al., 1963).

    Special study on teratogenicity

         Forty-eight male and 48 female Fischer rats from the cake control
    and highest-treatment groups of the F3 generation in the
    multigeneration study (see above) were mated and the foetuses removed
    by Caesarian section on day 20 of gestation. Numbers of corpora lutea,
    implantation and resorption sites, and live foetuses were recorded.
    After gross examination, sexing and weighing, half the foetuses were
    examined for skeletal defects and half for soft tissue abnormalities.
    Pregnancy rates in both groups were low, approximately 50%, and litter
    sizes were decreased compared with F0 and F1 matings. A slight
    reduction in ossification was noted in the chlorine-treated cake
    group, but no treatment-related soft tissue malformations were
    observed (Gumbmann & Gould, 1979c).

    Acute toxicity

         No information available.

    Short-term studies


         Two groups of 11 rats were fed for 16 days on dried cake made
    from a commercial cake flour treated with chlorine at a level of 0 or
    1563 ppm. The cake constituted 90% of the final diet, to which was
    added vitamin mix in sugar and a 4% mineral mixture. No significant
    differences in food intake or growth rate were observed (FMBRA, 1968).

         Four groups of 10 male weanling rats were fed for 29 days on
    dried or undried cake flour treated with chlorine at a level of 0 or
    1563 ppm; the dried flours were supplemented with 0.15% lysine. No
    significant differences in growth rates were observed (FMBRA, 1968).

         In preliminary trials, small groups of weanling rats were fed for
    16-95 days on diets in which 93% of the total nutrients were provided
    by cake prepared by several different recipes using flour chlorinated
    at commercial levels of 1100 or 2200 ppm or at higher levels of
    chlorination, 5359 or 10,718 ppm. Food consumption and growth rates
    were recorded; post-mortem findings were limited to organ weights,
    analyses of perirenal fat, haemoglobin concentrations, and PCV.

         At chlorination levels up to 5359 ppm, growth rates were
    unaffected, but at 10,718 ppm growth rates and food consumption were
    reduced, probably due to reduced palatability. As an additional study,
    the effects on growth rate were monitored using cake from chlorinated
    flour giving the required chlorination level either by direct
    chlorination or by dilution of a more highly-chlorinated flour with
    untreated flour. The results showed that the growth rate was decreased

    proportionately after feeding increasing levels of cake made with
    "diluted" flour (and with a similar corresponding reduction in food
    consumption) whereas "direct" chlorination to an equivalent level
    resulted in little or no reduction in growth rate at chlorination
    levels up to 5000 ppm. Post-mortem findings revealed no treatment-
    related effects at a level of chlorination of 2200 ppm, but at 5359
    and 10,718 ppm absolute liver and kidney weights were significantly
    increased and perirenal fat decreased, especially in females. PCV and
    haemoglobin concentrations tended to be lower in treated animals,
    sometimes significantly, but the levels were always within the
    expected range (Fisher et al., 1979).

         Flours chlorinated to levels of 2000 ppm or 10,000 ppm were fed
    to male Wistar rats at a level of 87.4% of the diet. Severe growth
    retardation was reported after 2 weeks, accompanied by increased liver
    weights. Lipids extracted from these flours had similar effects as did
    rat chow diets containing 2000 or 6000 ppm chlorine in the form of
    chlorinated gluten. A rat chow diet containing 2000 ppm chlorine as
    chlorinated flour lipids increased absolute weights of the liver by
    40%, kidney by 20%, and heart by 10% compared with pair-fed controls.
    Histological examination of the livers showed hepatocytes with coarse,
    foamy reticulated cytoplasm which were "collapsing and rupturing".
    Animals fed chlorinated gluten showed hepatocytes "with a glassy
    appearance" (Cunningham et al., 1977).

         The ether-extracted lipids of flour were chlorinated using
    chlorine levels required to chlorinate the parent flour to 2000 ppm
    or 10,000 ppm. The lipid and the extracted-flour residue were
    incorporated into otherwise identical diets fed to male rats for 2
    weeks in the following combinations: chlorinated lipid + unchlorinated
    residue; chlorinated residue + unchlorinated lipid; or, unchlorinated
    residue + unchlorinated lipid (control). Body-weight gains were
    reduced by the chlorinated lipid, but not by the chlorinated residue.
    Relative organ weights were affected as follows: liver and heart
    weights were increased both by chlorinated lipid and by chlorinated
    residue; brain weights were increased by chlorinated residue, but not
    by chlorinated lipid. Liver lipids (% w/w) were increased by
    chlorinated residue, but not by chlorinated lipid (Cunningham &
    Lawrence, 1978).

         In a 2-week feeding study, male rats were pair-fed on diets
    containing the ether-extracted lipids obtained from a cake made from
    chlorinated flour (2000 ppm) or from unchlorinated flour. The lipid
    was incorporated into ground rat chow at a level of approximately 6%.
    The body weights of animals fed the treated diet were lower than those
    fed the control diet by about 5% (p < 0.05), liver weights were
    higher by 12% (p < 0.01), and brain weights were lower by 4.5%
    (p < 0.01). Kidney and heart weights and liver lipids (% w/w) did not
    differ significantly (Cunningham & Lawrence, 1978).

         Ten-week feeding studies in rats were carried out using flour
    chlorinated to 0, 2000, or 10,000 ppm at dietary levels of 87.4%.
    Ether-extracted wheat gluten was directly chlorinated to 0, 20,000, or
    50,000 ppm and incorporated at a level of 10% in a rat-chow diet,
    which was also used in these 10-week studies.

         In the chlorinated-flour experiment, weight gains were
    significantly reduced by the treatment (p < 0.05 and p < 0.01 for
    the 2000 and 10,000 ppm chlorination levels, respectively). Relative
    weights of the kidneys, liver, heart, and brain were significantly
    increased (p < 0.01) compared to controls at the 10,000 ppm treatment
    level; relative kidney weights were also increased at the 2000 ppm
    treatment level. At the latter level, heart and brain relative weights
    were similar to controls.

         Male rats fed diets containing gluten chlorinated to 50,000 ppm
    (5000 ppm in the diet) showed significantly-reduced weight gains
    compared with controls, but this was not the case when the dietary
    level was 2000 ppm. Relative kidney weights were increased (p < 0.01)
    at both treatment levels, and relative brain weights were increased
    (p < 0.01) only at the 5000 ppm dietary level (Cunningham & Lawrence,

         Flour was chlorinated to levels of 1250 or 2500 ppm and fed to
    groups of Wistar rats at 80% of the diet for 28 days. Body weights
    were recorded daily and, at termination, the liver, kidneys, heart and
    brain from each animal were weighed and subjected to histological
    examination. Rates of growth and food intake were only slightly
    reduced by treatment, particularly in females. Absolute organ weights
    did not differ significantly from controls, whereas treatment-related
    increases in relative liver and kidney weights were observed in both
    sexes. Histological changes observed were confined to vacuolation in
    the liver (caused by glycogen accumulation) and renal calculi in the
    collecting tubules, especially in females, but these effects were
    independent of the level of chlorination (Fisher et al., 1979).


         Two groups of 4 or 6 dogs were fed diets containing flour treated
    with 356 ppm chlorine for 21-38 days. This diet did not cause "running
    fits". Dogs which had developed "running fits" on agene-treated flour
    recovered on being switched to chlorine-treated material (Arnold,
    1949; Radomski et al., 1948; Bentley et al., 1948; Newell
    et al., 1947).

         Groups of 3 male and female beagle dogs were fed cake made from
    flour treated with 0, 1000, 1500, or 2500 ppm chlorine, dried to a
    moisture content of 6%, and incorporated into diets at a level of 75%
    for 6 months. Appearance, behaviour, and food intake were monitored

    daily; body weights were recorded weekly. Blood samples taken prior to
    treatment and at termination were subjected to haematological
    examination (RBC, total and differential white cell and platelet
    counts, haemoglobin, PCV, and prothrombin time) and serum analysis
    (SAP, SGOT, SGPT, OCT, albumin, protein, cholesterol, bilirubin, BUN,
    thyroxine, Na, K, and Cl). Urine samples taken prior to treatment and
    at termination were examined for colour, pH, specific gravity, and
    sediment. All animals were autopsied, organ weights were recorded, and
    tissues were subjected to histopathological examination.

         No treatment-related effects on appearance, behaviour, growth, or
    food intake were observed. One mid-dose female developed peritonitis
    after 5.5 months and was killed and autopsied. No adverse effects on
    haematological parameters were observed. Serum analyses revealed a
    treatment-related decrease in thyroxine levels in females, which was
    statistically significant in the high-dose group only; other
    parameters examined showed no treatment-related differences. Urine
    composition was unaffected by treatment. No treatment-related effects
    were noted in the weights of the brain, testes, spleen, adrenals,
    thyroid, or pituitary; relative kidney weights were decreased in high-
    dose males and there was a non-dose-related increase in absolute and
    relative left ventricle weight and absolute heart weight in females.
    The main pathological changes seen were mild renal medullary
    mineralization in all dogs and patchy granulomatous pneumonia, but the
    effects were not treatment-related (Gumbmann & Gould, 1979d).

    Long-term studies


         Groups of 60 male or 60 female 4-5 week-old mice, Theiller's
    original strain, were fed diets containing cake made from flour
    treated with 0 (cake controls), 1250, or 2500 ppm chlorine that had
    been dried to a moisture content of 12.6% and incorporated into the
    diets at a level of 79% w/w. Groups of 30 male or 30 female animals
    were fed commercial diet 41B (41B controls). Satellite groups
    (30 animals of each sex) were used for interim examinations at 3 and
    12 months for haematology, serum analyses, and urinalyses (12 months
    only). The intended duration of the study was 80 weeks but, due to
    poor survival, the study was terminated at month 16 for males and
    month 17 for females.

         Cage weights (5 mice/cage) and food consumption were recorded
    weekly for the first month and then monthly thereafter. Haematological
    examinations (haemoglobin, RBC, total and differential leucocyte
    counts, and platelet counts) were performed on 10 mice/sex/group
    during months 3, 12, and at termination. Serum analyses (BUN, glucose,
    SGOT, SGPT, SAP, and total protein) were performed on 5 mice/sex/group
    at months 1 and 12 and on 10 mice/sex/group at termination. Pooled
    urine samples from 10 mice/sex/group were subjected to urinalysis

    (volume, pH, SG, protein, glucose, ketones, bile pigments,
    urobilinogen, blood pigments, and renal concentration tests) at month
    12 and at termination.

         At termination, all animals were subjected to gross necropsy and
    histopathological examination, the latter being limited to abnormal
    tissues, including all masses, brain, heart, lung, liver, spleen,
    kidneys, uterus, and gonads. Renal adipose tissue samples were
    analysed for covalent chlorine.

         No adverse treatment-related effects were seen on appearance,
    behaviour, food intake, or growth, but all the cake-fed animals had a
    reduced food intake relative to 41B-diet controls during the first
    6 months of the study. Body-weight gain was markedly higher in animals
    receiving cake diets (chlorinated or not) compared with 41B-diet
    controls, and between months 6 and 12 the former groups weighed nearly
    twice as much as the latter. The energy value of the cake diet was
    3.6 kcal/g and that of the 41B diet was 3.0 kcal/g.

         The mortality rate among males fed the cake diets was similar in
    all groups, but in females the groups receiving chlorinated cake
    showed a higher mortality rate than the cake controls. However, the
    mortality rate of mice in all cake-fed groups was greater than the
    mortality rate of the 41B control mice after 12 months and much higher
    than expected for the strain of mice used.

         Statistically-significant treatment-related reductions in red
    cell counts were seen in cake-fed males at 3 and 12 months; increases
    in white cell counts were observed at month 3; a non-treatment-related
    increase in haemoglobin was observed at month 12. No significant
    differences in haematological parameters were seen among cake-fed
    females. Compared with 41B controls, however, the following
    differences were seen: at month 12, decreased haemoglobin in all male
    and female groups; decreased red cell counts in high-treatment males
    and all females, and decreased PCV in all females. At termination, all
    male values were similar, but reduced Hb, PCV and RBC persisted in
    females. Despite the observed differences, none were below normal
    ranges and were not related to the level of chlorination. Clinical
    chemical analysis at months 1 and 12 revealed similar values between
    all cake-fed groups except for a decrease in BUN in females of the
    high-treatment group at month 1. Terminal analyses revealed a
    decreased BUN in males of the high-treatment group, decreased SGPT and
    total protein in females of this group, and a dose-related decrease in
    serum alkaline phosphatase in females. Comparison with 41B controls
    revealed increases in BUN at month 1 (all cake-fed males and cake
    controls and 1250 ppm females) and in all females at month 12 and at
    termination. No treatment-related effects were seen in urinalyses nor
    in organ weights, but comparison with 41B controls revealed
    significant increases in absolute and relative weights of the heart,
    liver, kidneys, and spleen in both sexes.

         At autopsy, no macroscopic changes attributable to chlorination
    of the cake flour were observed, but all cake-fed mice showed an
    increased incidence of excess adipose tissue. The predominant
    histopathological change seen in all groups was amyloid deposition,
    principally in the spleen, liver, kidneys, heart, ovaries, and uterus,
    but the incidence was not treatment-related. In the kidneys, there
    was a higher incidence (not significant) of glomerulonephrosis in
    cake-fed mice compared to 41B controls (35-50% compared with 25%,
    respectively), a significant increase in the number of calcareous
    deposits in the medulla or low in the tubules of the pelvis in the
    high-treatment group (though the amounts deposited were minimal), and
    a high incidence of cystis glomeruli in all cake-fed mice. In the
    heart, a high incidence of calcareous deposits was noted in cake-fed
    mice only, but these deposits were not correlated with the degree of
    glomerulonephrosis. Tumour incidence was similar in all groups;
    reticulum cell neoplasms (maximum of 1/group) were observed in all
    male groups and in the female high-treatment group.

         Analysis of renal adipose tissue revealed a treatment-related
    deposition of fat containing covalently-bound chlorine.

         The changes observed in this study could mostly be attributed to
    the nature of the diets fed to the mice rather than to the level of
    chlorination (Fisher et al., 1979; Ginocchio et al., 1983).

         Groups of 40 male and 40 female CDI mice were fed diets
    containing cake made from flour treated with 0, 1000, 1500, or
    2500 ppm chlorine, dried to a moisture content of 6%, and incorporated
    into diets at a level of 75%. An additional group (Purina controls)
    was fed commercial Purina chow diet. The duration of treatment was 85
    weeks for cake-fed males, 93 weeks for Purina controls and low-
    treatment females, and 103 weeks for the remaining females. Appearance
    and behaviour were monitored daily, while cage weights and food intake
    were recorded weekly for 28 weeks and monthly thereafter. Haematology
    analyses (RBC, total and differential leucocytes, PCV, and Hb) were
    performed on 8 mice/group at 7, 12, 15, and 18 months and at
    termination. Serum analyses (albumin, total protein, SAP, and BUN)
    were carried out after 15 months and at termination. Urinalysis
    (5 mice/group) was performed at the same times as the blood samples
    were analysed.

         At 15 months, 5 mice/sex/group were sacrificed, and these animals
    and those killed at termination were autopsied. Liver, spleen, kidney,
    heart and testes weights were recorded and all animals were subjected
    to histopathological examination.

         No adverse treatment-related effects were recorded for
    appearance, behaviour, or growth rate. Survival was unrelated to the
    level of chlorination and was slightly greater in all cake-fed groups
    than in the Purina controls. No adverse effects were seen on

    haematological parameters and no differences attributable to
    chlorination were seen in serum analyses or urinalyses. At interim
    kill (15 months), increased absolute and relative kidney and heart
    weights were recorded in males of the 2500 ppm group and increased
    absolute and relative liver weights were seen in both sexes of the
    1000 and 1500 ppm groups; at termination, no differences were noted.
    The main pathological changes, none of which could be attributed to
    chlorination, were: a high incidence of amyloidosis in all groups
    (less marked in females), increased reticuloendothelial hyperplasia in
    males of the 2500 ppm group, and acute inflammation of the urinary
    tract in males of the 1500 and 2500 ppm groups.

         Although the overall tumour incidence was unaffected by
    chlorination, there was a significantly-increased incidence of
    lymphohaematopocitic tumours in mid- and high-treatment females
    compared with the Purina controls, but not compared with the cake
    controls; the tumours were classified as malignant lymphoma. Actuarial
    analyses of the lymphomas revealed that the differences between the
    numbers of tumours expected and observed were not significant. There
    did not appear to be any relationship with the level of chlorination;
    however, the apparent induction period for tumour formation was
    decreased (Gumbmann & Gould, 1979a).


         Groups of 60 male or 60 female Wistar-derived rats were fed diets
    containing cake made from flour treated with 0, 1250, or 2500 ppm
    chlorine dried to a moisture content of 12.6% and incorporated at a
    level of 79% into their diets for 104 weeks. An additional group
    received commercial diet 41B (41B controls). Satellite treatment
    groups were also included for interim studies. Appearance and
    behaviour were monitored daily, food consumption and body-weight gains
    were recorded weekly for the first 13 weeks and monthly thereafter,
    and animals were palpated for tumours when body weights were recorded.
    Water consumption was determined at monthly intervals. Blood samples
    were taken from 10 rats/sex/group for haematological examination (Hb,
    PCV, RBC, and total and differential leucocytes) at 3, 12, 18, and 24
    months and from 5 rats/sex/group for serum analyses (BUN, glucose,
    SGOT, SGPT, SAP, total protein, and albumin) at 3, 12, and 18 months;
    serum analyses were also performed on 10 rats/sex/group plus all
    female survivors at 24 months. Urinalyses (pH, SG, protein, glucose,
    ketones, blood and bile pigments, urobilinogen, NAG, and creatinine)
    were carried out on samples from 10 rats/sex/group at 6, 12, 18, and
    24 months; renal urine concentration tests were carried out in 5
    rats/sex/group at 6-month intervals, urinary cell counts at month 18,
    and urinary GOT at 18 and 24 months.

         At termination, all animals were autopsied and weights of the
    brain, heart, adrenals, liver, spleen, gonads, pituitary, kidneys,
    thyroid, and uterus were recorded. Because few females survived the
    2-year period, organ-weight analyses were performed on the 5 rats/
    group satellite-treatment animals killed at month 18 for serum
    analysis. All animals were subject to gross pathological examination
    and detailed histopathology.

         At an early stage in the study, evidence of respiratory distress
    was noted in several rats, particularly the 41B controls. Serological
    evidence of sialodacryadenitis was shown in animals that died. Because
    respiratory distress appeared to be exacerbated by the powdered nature
    of the 41B diet, this diet was fed in pellet form from week 19 onward.

         There were no differences in food intake among cake-fed groups,
    but those fed the 41B diet had a higher food intake (20-25% higher),
    especially after the diet was changed to pellet form; the energy value
    of the cake-based diets was 3.5 kcal/g and that of the 41B diet was
    3.0 kcal/g. Rate of weight gain was similar in all cake-fed rats,
    irrespective of whether the cake flour was chlorinated, and 6-7%
    higher than in the 41B controls. Water intake was unaffected by

         Mortality was similar in all cake-fed groups, including cake-fed
    controls. Initially, in male 41B controls mortality was much higher
    than in rats fed the cake-based diet (due to early loss from
    respiratory illness), but was similar to other groups by week 80. In
    females, mortality of 41B controls was lower than that of cake-fed
    animals throughout the study. After 104 weeks, mortality ranged from
    40% (41B controls) to 63% (cake controls) in males, and 85% (41B
    controls) to 98% (cake controls and 1250 ppm group) in females.

         Isolated changes relative to cake controls were seen in some
    haematological parameters, but there were no consistent treatment-
    related effects. Compared with cake-fed controls, there were
    treatment-related increases in SGOT and SGPT at month 12 in males,
    but SGPT levels were within the normal range. At month 24, total serum
    protein was decreased in high-treatment males.

         No treatment-related effects on urine composition were noted.
    Blood-stained urine was noted in the floors of a few cages from each
    male group, but it did not appear to be treatment-related; only one
    cage of females was affected. The ratio of NAG to creatinine revealed
    no treatment-related effects, even in rats with kidney lesions
    observed at post-mortem.

         No treatment-related effects on organ weights were noted at
    termination. However, taking all females (including 5 animals killed
    at month 18) into account, there was a significant reduction in spleen
    weights relative to cake-fed controls, though the spleen weights of
    all cake-fed females were higher than those of 41B controls.

         The main gross pathology finding was an increased incidence of
    enlarged kidneys in rats receiving the cake-based diets. Histology
    revealed glomerulonephrosis of varying severity in more than 90% of
    the animals, to a greater extent in those fed cake-based diets. Renal
    calculi were observed mostly in animals fed cake diets (80% of
    females, 15% of males) and few 41B rats were affected (6 females, no
    males). There was an increased incidence of haematopoiesis in the
    spleens of females fed cake-based diets.

         Pituitary chromophobe adenomas were more common in the female
    groups (approximately 65% of each group was affected) than in the male
    groups (15-33%), and probably contributed to early mortality. However,
    no differences were noted in tumour incidence that could be attributed
    to chlorination of cake flour.

         Perirenal adipose tissue from 5 rats/sex/group showed covalently-
    bound chlorine levels related to the level of flour treatment (Fisher
    et al., 1979).

         Groups of 40 male or 40 female rats from the F1 generation of
    each of the treatment groups in the multigeneration reproduction study
    (see Gumbmann & Gould, 1979c, under special studies on reproduction)
    were maintained on their respective diets until 20% survival was
    reached (the males were killed during weeks 97-99 and the females were
    killed during weeks 110-112). Satellite groups of 5 animals of each
    sex were killed on day 45 and liver samples were assayed for
    N-demethylase activity. Body-weight gains and food intake were
    recorded weekly for 28 weeks and at monthly intervals thereafter.
    Blood samples were taken from 8 rats/sex/group at 6, 12, and 18
    months and at termination and used for haematology (RBC, total and
    differential white-cell counts, PCV, and haemoglobin counts). Blood
    samples taken at 15 months and at termination were examined for serum
    albumin, protein, SAP, SGOT, SGPT, ornithine carbamoyl transferase
    (OCT), BUN, glucose, and cholesterol. Urinalyses were performed at
    6-month intervals.

         Weights of the liver, spleen, kidneys, heart, testes, adrenals,
    thyroids, and brain of all rats killed were recorded after 15 months
    (5 rats/sex/group) and at termination. All rats were subject to gross
    and histopathological examination.

         Growth rates were unaffected by chlorination, but all cake-fed
    animals had a slightly greater growth rate than Purina controls. Food
    consumption was unaffected by chlorination; rats fed Purina chow ate
    significantly more throughout the study. Mortality was not affected by
    chlorination, but survival of cake-fed animals was poor and survival
    of the Purina controls was greatest.

         No treatment-related effects of hepatic N-demethylase activity
    were noted in the satellite groups at day 45.

         Haematological examinations at termination showed reduced PCV,
    haemoglobin, and RBC counts in males, especially at the 2 higher
    levels of treatment, while females remained unaffected. No differences
    due to chlorination were seen in any of the serum analyses, but SGOT,
    SGPT, and°CT levels were lower in cake-fed animals than in Purina
    controls. No treatment-related changes were observed in urinalyses,
    but the urines of all cake-fed rats were slightly more acidic than
    those of the Purina controls.

         At autopsy no effects attributable to chlorination were observed
    on organ weights, but the absolute and relative kidney weights of rats
    fed cake-based diets were increased compared with Purina controls. The
    main pathological changes that were observed occurred in the stomach
    and kidney. Ulceration of the glandular and non-glandular stomach was
    seen in all females fed cake-based diets and males of the high-
    treatment group. There was a high incidence of nephropathy in all
    cake-fed animals, and mineralization of the cortico-medullary junction
    was noted to a greater extent in females than males. The authors
    concluded that the renal disease was the cause of the high mortality.

         Although overall tumour incidence was not increased by
    chlorination, incidences of pituitary chromophobe adenomas and mammary
    fibroadenomas were higher in females receiving chlorinated cake-based
    diets than in cake-fed controls. However, the incidences were not
    higher than in Purina controls nor were they treatment related, and
    the significance of these changes is doubtful (Gumbmann & Gould,

    Observations in man

         No information available.


         In long-term and reproduction studies in which rats and mice were
    fed diets containing 75-79% dried cakes made from flour chlorinated
    at levels up to 2500 ppm, no carcinogenic, teratogenic, or other
    toxic effects attributable to chlorination were observed. The
    glomerulonephrosis and renal calcification seen in the long-term
    studies in rats were considered to be due to nutritional imbalance and
    did not represent a toxic response to the chlorinated cake flour.


    Acceptable level of treatment of flours for cake manufacturing

         0-2500 ppm Cl2.


    Arnold, A. (1949). Effect on dogs of flours treated with various
         improving agents. Cereal Chem., 26, 46-51.

    Bentley, H.R., Booth, R.G., Greer, E.N., Heathcote, J.G., Hutchinson,
         J.B., & Moran, T. (1948). Action of nitrogen trichloride on
         proteins: Production of toxic derivative. Nature, 161,

    Coppock, J.B.M., Daniels, N.W.R., & Eggitt, P.W. (1960). Essential
         fatty acid retention in flour treatment. Chem. Ind., 1960,

    Cunningham, H.M. & Lawrence, G.A. (1976). A comparison of the
         distribution and elimination of oleic and chlorinated oleic acids
         and their metabolites in rats. Fd. Cosmet. Toxicol., 14,

    Cunningham, H.M. & Lawrence, G.A. (1977a). Absorption and metabolism
         of chlorinated fatty acids and triglycerides in rats. Fd.
         Cosmet. Toxicol., 15, 101-103.

    Cunningham, H.M. & Lawrence, G.A. (1977b). Absorption and distribution
         studies on chlorinated oleic acid and extracts of chlorinated
         lipid and protein fractions of flour in rats. Fd. Cosmet.
         Toxicol., 15, 105-108.

    Cunningham, H.M. & Lawrence, G.A. (1977c). Placental and mammary
         transfer of chlorinated fatty acids in rats. Fd. Cosmet.
         Toxicol., 15, 183-186.

    Cunningham, H.M., Lawrence, G.A., & Tryphonas, L. (1977). Toxic
         effects of chlorinated cake flour in rats. J. Toxicol. Environ.
         Hlth., 2, 1161-1171.

    Cunningham, H.M., & Lawrence, G.A. (1978). Effect of chlorinated lipid
         and protein fractions of cake flour on growth rate and organ
         weight of rats. Bull. Environ. Contam. Toxicol., 19, 73-79.

    Daniels, D.G.H. (1960). Changes in the lipides of flour induced by
         treatment with chlorine dioxide or chlorine, and on storage.
         J. Sci. Food Agric., 11, 664-670.

    Daniels, N.W.R., Frape, D.L., Eggitt, P.W., & Coppock, J.B.M. (1963).
         Lipids of Flour. II. Chemical and toxicological studies on the
         lipid of chlorine-treated cake flour. J. Sci. Food Agric.,
         14, 883-893.

    Fisher, N., Hutchinson, J.B., Berry, R., Hardy, J., & Ginocchio, A.
         (1979). FMBRA Report of toxicological trials of cake made from
         flour treated with chlorine. Collaborative studies in rats and
         mice. Unpublished report from Flour Milling and Baking Research
         Association, Chorleywood, England.

    FMBRA (1968). Submission to the Food Additives and Contaminants
         Committee, U.K. from Flour Milling and Baking Research
         Association, Chorleywood, England.

    Gilles, K.A., Kaelbe, E.F., & Young, V.L. (1964). X-ray spectographic
         analysis of chlorine in bleached flour and its fractions.
         Cereal Chem., 41, 412-424.

    Ginocchio, A.V., Fisher, N., Hutchinson, J.B., Berry, R., & Hardy, J.
         (1983). Long-term toxicity and carcinogenicity studies of cake
         made from chlorinated flour. 2. Studies in mice, Fd. Chem.
         Toxicol., 21, 435-439.

    Gumbman, M.R. & Gould, D.H. (1979a). Long-term feeding studies for
         safety evaluation of chlorine-treated flour. II. Life-span
         studies in mice. Unpublished report from the US Department of

    Gumbman, M.R. & Gould, D.H. (1979b). Long-term feeding studies for the
         safety evaluation of chlorine-treated flour. III. Life-span
         studies in rats. Unpublished report from the US Department of

    Gumbman, M.R. & Gould, D.H. (1979c). Long-term feeding studies for the
         safety evaluation of chlorine-treated flour. IV. Three generation
         reproduction and teratology studies in rats. Unpublished report
         from the US Department of Agriculture.

    Gumbman, M.R. & Gould, D.H. (1979d). Long-term feeding studies for the
         safety evaluation of chlorine-treated flour. V. Six-month study
         in dogs. Unpublished report from the US Department of

    Newell, G.W., Erickson, T.C., Gilson, W.E., Gershoff, S.N., &
         Elvejhem, C.A. (1947). Role of "agenized" flour in the production
         of running fits. J. Amer. Med. Assoc., 135, 760-763.

    Radomski, J.L., Woodard, G., & Lehman, A.J. (1948). The toxicity of
         flours treated with various "improving" agents. J. Nutr., 36,

    Sollars, W.F. (1961). Chloride content of cake flours and flour
         fractions. Cereal Chem., 38, 487-500.

    See Also:
       Toxicological Abbreviations
       Chlorine (ICSC)
       Chlorine (FAO Nutrition Meetings Report Series 40abc)
       CHLORINE (JECFA Evaluation)
       Chlorine (PIM 947)