Document bavaoRrdZpGronavjM5dZG4G6

c pf a poisoning caused by rice oil contaminated with polychlorinated biphenyls MASANORI KURATSUNE, M.D,, TAKESUMI YOSHIMURA, M.D, JUN1CHI MATSUZAKA, M.D,, and ATSUKO YAMAGUCHI, M.D. QPORAOIC outbreak* of a peculiar akin dis- department of hygiene of our faculty of medicine O cose were reported in Fukuoka-Ken (Fuku and the departments of public health of the Pre oka Prefecture). Japan, in early October 1968. It fecture of Fukuoka and of the cities of Kilakyu* was characterized by such symptoms as follicular Shu, Fukuoka, and Ohmuta. accentuation, acneform eruption, pigmentation of The epidemic spread not only over Fukuoka- c the skin and nails, and hypersecretion of the Mei Ken but also over 20 other prefectures in the bomian gland Staff members of the department of western part of Japan it produced 1,001 patients \ ' dermatology of the Faculty of Medicine, Kyushu --502 males and 499 females--according to the ~ University, Fukuoka, who examined some of the latest tabulation by the section of food hygiene, early patients suspected they had chloracnc. A Ministry of Welfare. Although our study of the possible causal relationship between the disease epidemic was confined to Fukuoka-Ken, and the and ingestion of a certain brand of rice oil was epidemic proved to be caused by accidentally con also suspected in view of the distinct familial ag taminated oil, we believe that publication of our gregation of the patients and their common use of results will help prevent similar food poisonings. the oil (/). This relationship was subsequently These intidents can occur anywhere in the world proved, and the disease was called Yusho, or rice if sufficient care is not exercised. oil disease. . . To clarify the cause of the epidemic, a study The authors ore with the deportment of public group headed by Prof. S. Katsuki of the Kyushu health. Faculty of Medicine, Kyushu Universtty, University Faculty of Medicine was organized by Fukuoka, Japan' Dr, Kuratsune is a professor the staff of that university, the School of Medicine of public health, Dr. Yoshimura and Dr. Matsu- of Kurume University, and local health depart taka ore working toward doctoral degrees in pub `' ments. An epidemiologic study subgroup was set lic health, and Dr. Yamaguehi is a research stu up within the study group on October 19, 1968, dent. The study on Yusho was supported by a which was directed by Kiiratsune and also in special research grant from the Japanese Govern cluded the other three authors. This subgroup im ment. Tearsheet requests to Masanori Kuratsune, mediately designed and carried out a series of M.D., Faculty of Medicine, Kyushu University, extensive surveys in dose cooperation with the Fukuoka, Japan. i Pacamaw lan, V*. s. Na. ta iocs MONS 201832 Most of the rcsulu of our itudy have already b^cn published in Japanese (2-5). We refer also, whcn necessary, lo some important observations made by a clinical study subgroup headed by Prof. K. Higuchi of the department of dcrmatol- 0y at our university and also to studies by a chemical study subgroup headed by Prof. H. TsuIcanioio of the Faculty of Pharmaceutical Sciences. Clinical Symptoms . The subjective symptoms of the 89 male and the 100 female patients whose cases of Yusho had been diagnosed by the end of October 1968 an listed in table l. Most of these symptoms related to the skin and eyes. Detailed clinical examina tions of these 189 persons by the clinical atudy subgroup revealed many of the clinical character istics of Yusho (1,6-9). Dermatological observations. Acncform erup tion--marked enlargement and elevation of the follicular opening; pigmentation of the skin, lips, gingiva, the mucous membrane of the oral cavity, and nails; formation of hyperkcratotic plaque on the soles of the feet and palms of the hands; hypcrliidrOMv; hypertrichosis; swelling of Mont gomery's gltittd in the breast; cyst formation of the sebaceous gland in the genital region; dry skin; and, in children, milia near the joints. Table 1. Percent distribution of symptoms of Yusho reported by 189 patients examined be fore October 31, 1968 Symptoms Mites Female* (N~9)(N-I00> Dart, brown pigmentation of nail*. Diiiidciivc har fofltctes,. -........a. * Incrcavif iwcating at palm*. .** Aenelike skin eruption*.................. Red pliiquet on limbi............ I*ignicniaiion of skin....................... Swelling of limbi..................\ Stiffened aolct in feel and palms ef hainlt................................. . ***w I'lKiiKiiicil nmcmii membrane.... lm.TcuM.il eyediMtlwtrgc..........y* v* 1 typcrcmi* of conjunct**............ ta * a Transient vnual diilurbtMS....... Jaundice.......................... ................. Swelling of upper cyellUu............... peeling of wcBkncu......................... Numbnct* in btnbt.*.......... Fever....................... ........................ Hearing difficult ic*........................... Spaim of limb*....................... . Headache......................................... Vomiting............... ;............... Diarrhea.................. 1........... .......... }.( 64.0 $0.6 17.6 20 1 41.7 7J.J 10.2 >4.7 $6.1 II.1 70.1 $6.1 11.1 71.9 $1.4 $1.6 16.9 It.O 7.9 30.) 13.6' 19.1 71.0 41.0 29.0 47.0 13.0 71.0 $s.o 11.0 74.0 52.0 39.0 19.0 19.0 1.0 39.0 21.0 17.0 bbbobb Ophthalmic observations. Hypersecretion of the meibomian gland and abnormal pigmentauon of I he conjunctiva. Neurological observations. Numbness, pain, hypocsthcsia, and lowered sensory conduction by the radial and sural nerves. Other observations. Elevated scrum triglycer ide. Epidemiologic Data Available When the epidemiologic study subgroup hegnn to act, it had several pieces of information that appeared to be important but whose implications were uncertain, such as the following: t. The familial aggregation of patients was not able, and the disease showed no distinct age and sex preference. 2. All the patients seemed to have used one brand of rice oil (abbreviated at K rice oil be cause it was produced by K Company). An emer gency survey undertaken by the health department Of the city of Fukuoka hag disclosed that tome of the patients in a certain area had, as a group, pur chased a can of K rice oil and divided it among the group members. Those who did not buy the oil seemed to be unaffected even though they lived next door to the patients. 3. The patients seemed to be broadly distrib uted throughout Fukuoka-Ken. These data of course were valuable, but we knew that if we relied on them alone, they might mislead us. We exercised particular caution, there fore. to avoid such a risk, always seeking possible antitheses when interpreting any result. Descriptive Epidemiologic Studies The epidemiologic atudy subgroup first tried lo leant how the 323 patienta were distributed by place of residence end time of onset of the disease (2). Those persons who thought they were svf* -fering from Yusho were asked to report to their local health departments. From October 1968 until the epidemic practically terminated in Janu ary 1969, a total of 6,61 i`persons in FukuokaKen had reported that they were affected. Among them, 325 cases of Yusho were diagnosed by the clinical study subgroup. An early analysis of a portion of the 325 persons with confirmed cases, carried out in late October 1968, revealed several peculiar features in the geographic distribution of cases, features which did not basically change dur ing the epidemic. Therefore, only the results of 10S4 HSMHA Kaaim ftaporu MONS 201633 analysis of ihcsc 32.5 cases will be presented to illustrate these peculiarities. Fomihul aggregation. The 325 patients be longed to 112 families, a result indicating a dis tinct familial aggregation, which is considered one of the most important characteristics of Yusho. Time of occurrence. Ninety-nine percent of the patients were affected during 1968. and 55 percent of the eases were concentrated in the 3 months from `June to August {table 2). The monthly distribution of the eases was alike for both sexes These results indicated that exposure to the causal agent or agents was probably tem poral and indifferent to sex. Sex and age. Distribution of the 325 patients by sex and age revealed no atgnificant sex differ ence (table 3). Age-spccific incidence rales per 100,000, calculated by using the census popula tion of 1965, indicated lower risks for both males and females in the age group over 60 years, but again no marked sex difference (table 4). Geographic dirtrihutiem. We calculated crude incidence rates of Yusho for three large cities fXitnfcyuthu, Fukuoka, and Ohmuta) and for the ju risdictional areas of the 22 local health depart ments of Fukuoka-Ken. A very marked differencc was noted in their rates. Ten health depart ments reported no eases of Yusho while extremely high incidence rates per 100,000 were noted for the area covered by the health department of Soeda (58.9) and that covered by (he health de partment of Tegawa (44.5); intermediately high rates, from 7 to 14, were observed for (he three large cities and the area of the Kasuya health department. There arc many coal mines, meat of which have recently been abandoned. In the jurisdictional areas of the Soeda and Tagawa health depart ment*. The city of Kitakyushu hat big steel and chemical industries, Fukuoka is a scat of com merce, and Ohmuta has big coal mines and chem ical industries. In contrast, the Kasuya health ds- Tablc 2. Distribution of 325 patients with Yusho diagnosed by January 20, 1949, by sex and month when symptoms appeared t' Before 1961. Jtmjjry,**,, Fehruary..,, March......... April-------Msy............. July............. Aucusl., , September.. October....... November.. December... Total... !9U Mate* Number Females Number Pareaat ' ( Total ^Cumber Percent 2 i.) . 2 t.l " . 4 0 .0 4 2.5 5.1 15 9.5 l> 11.4 2) 14.4 17 17.1 34 .21.5 17 `Ifi.l a 5,1 i . i .4 .0 6 12 10 15 . 55 32 30 12 14 1 0 .0 3.4 7.2 4.0 9.0 19.1 19.2 II.O 7.2 >.4 .0 . 9 10 29 25 54 59 44 2V 22 2 1 151 100.9 147 100.0 325 1.2 .0 3.1 4.2 7.7 10.2 17.2 ll.l 19.7 4.9 4.4 .4 .3 100.0 Table 3. Distribution of 325 patients with Yusho diagnosed by January 20, 1969, by sex and age group A|t group (ymn)> 0-9 .... 10-19,... 20-:9,..., JO-39,... *0-49,,., J0-J9.... SO-69,.,.. 70-79....... Tots! *< I 1 z Maim Pereaat FemaJat ' Number Percent Total Number PNeeni 34 24 30 11 9 14 1 134 23.4 24.1 17.7 19.0 7.9 5.7 2.5 .5 100.0 27 ir 34 39 , 23 * 11 3 0 147 14.3 14.4 21.4 23.4 13.4 4.4 1.4 .0 IQO.O 44 19.7 ' 44 20.3 44 19.7 49 21.2 34 10.5 20 4.2 7 3.2 1 .3 323 100,0 / r Pwwwbw ten. vat. m. n*. 11 sets w W" HOMS 201834 parimcnt cover* an area that it primarily agricullural. Mott area* where few Yusho patients were teen were agricultural. No other common *ociocconomic or environmental factor was observed that might explain the geographic distribution of patients Oil Used by Patients As stated, K rice oil had been suspected as the possible cause of the epidemic. K nee od it i high-quality salad oil, made from rice bran by the procedure shown in the chan. It was popular in Fukuoka-Ken, even though it was slightly more expensive than other oils on the market. A thor ough investigation was undertaken to determine whether the patients with Yusho had actually used this brand of oil before they were affected and if so, whether there were any common characteris tics in regard to the container of the oil (bottle or can), date of production, datfc of purchase, or route of marketing. To seek information on these points, an investigation was conducted in coopera tion with the local health departments. All pa tients and their wives were visited end interviewed in their homes. In addition, all the available ship ping records of K Company and all available rec ords of purchases and sales of K rice oil at whole sale oil dealers' offices and at retail stores were carefully examined. These followup surveys soon disclosed an as tonishing fact. AIJ the persons with Yusho had used, for varying periods, K rice oil which came cither from 16.5 kg. cans or 1.65 kg. bottles. One hundred seventy of the 325 patients--49 house holds--had used only canned oil, while the re maining 155 patients--63 households--had used only bottled oil {table 5). We were able to con firm the shipping dales for the oil used by all but four of these 170 patients. No matter where the persons lived or what their age was, 166 of the 170 had apparently used rice oil that was shipped by the K Company'on February 5 and 6, 1968, according to the lot numbers on the few remaining cans. Even the four patients for whom we could not confirm shipping dales of the oil had alio undoubtedly used a canned rice oil produced by the K company. ' / The shipping or production .dates could not be confirmed, however, for the bottled K oil that 155 patients had used exclusively, because they had no old hollies at home. Purchasers of oil in bottles return the empty ones to retail stores when they buy new oil. Nevertheless, all the shipping, pur- . chase, or selling records at K Company, at (he m< dealers, and at the retail stores were examined to learn whether any bottled K nee oil had been shipped to. and had reached, the retail store* from which the 155 patients bought their oil. All ship ments of oil to the stores used by the patients were further investigated to determine whether any of them had been made on February 5 or 6,. 1968, or immediately before or after these dates By this method, we discovered that 143 of the, 155 patients might have bought the bottled oil shipped from the K Company in the period Feb ruary 6-15, 1968. Whether the remaining 12 pa tients might have also purchased oil shipped in this period remains in doubt because of lack of records. Thus, nearly all of the patients had a distinguishing characteristic--they all had a com mon experience of use. or possible use. of rice oil shipped by one company in a specific period. Such a coincidence could hardly be expected to happen by chance. Table 4. Agt-sperifio Incidence rates for Yusho, by sex of patient Age group (yews) Incidence rata per 100.000 ;Haie* Fcmsln 9-9....................................... ............................ 79-19..................................... 30-39..................................... 49-49..................................... 30-59..................................... 40-49..................................... 70 and over........................... A# ages............ 1.6 6.6 10 3 II V 9.3 7.2 2.4 .0 l.l ' Table 5. Distribution of 325 patients and 112 households Ihnt used K rice nil, by type of oil container and shipping data of oil Container end hlinu HouiehoWi hipping detu -------------------- -------------------- Number Percent Number Percent Users of canoed oil... Shipped February 3 and S.............. Shipping dates unknown.............. Users of bottled od... Probably used oil shipped February 5--15...................... Shipping dates unknown.............. Total..................... 179 52.) 144 Sl.l 4 1.2 155 47.7 14) 44.0 12 3.7 325 100.0 4* 43.0 47 42.0 2 l.l 43 54.) 34 50.0 7 4.3 112 > 100.1 * Docs not add lo MOJO percent became of rounding. IOM HSMHA HhIUi Jtaporte `- . v MOMS 201835 r FLOW (HEFT OF RISC JWL PRODUCTION Ikt tail -r--; kna N-Jixa# utid tt xtr*c* CraSssil After nmmi if satvaut, ktrtateri Mfe wRInm umpawato I 1------MfOH Layer N*utr*tU< wild H*SO. -OirtT ait \ Tttote* with MlOH 1 OH -WatbM wttb tot water *i Ftttertl t | ' Ct with RHtataia clay Datotrliae to toatmi al *nt 300* C. at 3 4 mm. Hi FUat frttfacte Our survey also demonstrated a high attack rate for ail persona who consumod the oil. Alto gether, 266 persons in 64 households had eon* turned canned K nee oil shipped in the period it) . question, and 170 had confirmed eases of Yusho. Thus, the attack rate was as high as 63.9 percent (i>. . K oil From Other Shipments ^ An additional survey seemed necessary to de termine whether those persons who regularly used K rice oi! but had not used any from shipments made in the period in question were free of the disease. For this purpose, a group of persons was needed who had reliable records of their health tutus id of their purchases of K rice oil. A group of 113 persons of 29 households living in one apartment house seemed to be appropriate subjects for this study. According to the records of this group and the oil dealer who supplied them, these 113 persons had purchased canned K rice oil as a unit from one dealer and divided it among themselves f.urly regularly from December 1967 to September 1968 except for the period from January to April 1968, when this bulk purchasing was not carried out. Their disease experience, particularly any skin disease that occurred during 1968, was care fully explored through interviews and, if necessary, through investigation of their medical records at hospitals. No cases of Yusho were found among baeiwtor sen, tot as. n^ ix seer "TV MOWS 201836 ( tii cm tied of Ihts group. Thus, we demonstrated ih;u |K.VM>n* who had used K rice oil regularly for more ih;> a half year but did not use the oil front the specific shipments did not suffer from Yusho 12). Table 7, Distribution of 69 households with pa tients and 207 control households, liy use of various fats or oils, based on case-control study Fat or oil Households with patients Control households Retrospective Cast-Control Studios All these results suggested that Yusho was cau sally connected with use of the X rice oil that was shipped from K Company on Februory 5 and 6. iOfiB. or soon thereafter. Nevertheless, some other factors or agents might have been the primary or secondary cause of the epidemic. Therefore two Number Percent Number Percent Butler,............ Margarine. Sesame oil............... Rape-teed oil......... .. Rtee-bran oit....... Lard............ Other oil*.... ..... 30.7 105 30.7 63.1 127 1.4 30.5 03 41.1 t.3 77 37.2 93.7 04 30.9 n, 31 IS 4 It.l t IT 34.3 rcsc-contfol studies were done. In one, the per sona! backgrounds of the patients and matched controls were compared. In the other, the use of oil and fats in the households of the patients and the controls were compared. To discover whether persons with Yusho eases and those without h;id different personal back grounds, home interviews were conducted with all persons whose cases bad been reported by Octo ber 31, 1968. and with healthy controls matched by age, sex, and place of residence. One hundred twenty-one patients and 121 controls (S3 males and 68 females) were selected in this manner and asked 60 questions concerning their occupations. medical background, general health status, habits, customs, diet, pets, and other characteristics of their lives. The patients and the controls differed signifi cantly in respect to only two of the personal fac tors tested (table 6). Persons who ate fried foods or tempura nearly every day were more numerous among the patients than among the controls, while more of the controls than the patients had pets. To compare the use rdf oil and fats, 69 house holds with Yusho patients were matched by place of residence with 207 households without such patients--three control households for each housi ehold with cases. a As to regular use of fats and oils, a distinct Table 6. Ferccnt of 121 patients and 121 controls difference was noted between the two groups only having various allergies and dietary and per* xoaal habits, based on case-control study hunt investigated Piikau Con trols for rice-bran oil. About 96 percent of the patients' households and 31 percent of the controls' house holds reportedly used rice-bran oil regularly (table 7). When asked about the use of rice oil from the Allergic to fish....................................... 7.3 Allergic l npirin.................... . - ,0 4.2 Allergic to drugs other than aspirin.,. . Bathing facilities at home... 1.5 Bath every day..................... . 73.0 70.4 Pen at home.................................... 134.3 Flow space of house smaller than 44 square meters ......................... .. . *4.9 *4.1 Handling agricultural chemicals........... 4.4 Taking coOlivcr oil* . to.a . M Taking vitamin pith.. 11.3 Taking fiber rettoraliu* dreg*............. 7.5 Witter sup|ily avmlshleMhome.......... 74.7 timing out occaioisa%..... . 31.t 30.4 Tiik ing all meals with the family.\,... . Green vegetablesdefer................... , 63.1 9.6 it.9 Milk nearly every day......................... .. . 40.0 29.0 Butter nearly every day... * * , 32.4 ' AY4.9 Eggs nearly every day. . 64.7 39.1 r, >22.4 * 11.6 ' 4J.7 35.7 Ptili nearly <very day,,,......................... A 21,* Mayonnaise nearly every day............. . . 'to.a 29.1 to.a Instant "rahmen" or Chinese noodles nearly every day, , iO.S 10.0 'f <0.03. ;K Company, all of the households with patients affirmed its use in comparison with only 32.4 per cent of (he control households; 61.3 percent rdf the control households denied using it; the remaining 6.3 percent were uncertain. Thus, the case-control studies dearly indicated that none of the (acton *' -tested except use of K rice oil could account for the disease (2), Doso-Rotponta Ralalioftship To prove that a causal relationship existed, a distinct dose-response relationship needed to be demonstrated. Estimating the quantity of contami nated rice oil consumed by each patient and mem bers of bis family was not feasible. Nevertheless, If we disregarded age and sex, we could estimate roughly the average amount of oil each person had consumed by dividing the amount used in a 10M MSMHA HsnHti naparta 'k _ v HONS 201837 I household by (he number of persons in it, taking into account the frequency with which these per sons ate at home, but not the possible loss of oil during and sifter cooking, nor the virying teste* end food intake of the family members. Dr. Masayasu Goto, a dermatologist, with no knowledge of the amount of oil used by the 146 consumers of the contaminated oil, examined these persons and then graded the eases of Yusho he diagnosed according to their- clinical severity. Only later did we link the severity of the eases, as determined by the dermatologist, to the amount of contaminated oil used by the patient Eighty of the 146 users of the contaminated oil were believed to have consumed, individually, leu than 720 ml. (table 8). Of these 80 light users, 12 percent were not affected, 49 percent had light cases, and 39 percent had severe cases; the attack rate was 88 percent The 45 persons who had used 720-1,440 ml- were alt affected; 69 percent of this group had severe cases. A still larger pro portion of the.javcre cases was found among the persons who had used more than 1,440 ml, of the contaminated oil; this group had an attack rate of 100 percent. Since the clinical severity of the disease was not found to differ significantly between the sexes but differed considerably according to age, a standard ization for age seemed necessary to achieve a more accurate interpretation of the observed rela tionship between use of (he oil and the clinical seventy of the disease. With the age composition of the 146 users of the ml as the standard, the factot of age was standardized few each of the three levels of oil intake. The figures, however, hardly changed from those shown in table 8. Thus a very clear dose-response relationship could be demonstrated, even though the estimates of the dose were inaccurate (3). ; Toxic Agent ... In view of all these epidemiologic observations, we concluded that K rice oil from specific ship ments had caused the Yusho, Now the question arose as to why tlut aH was toxic. To answer this question, we of course had to rely on other ap proaches than epidemiology. In mid-Novdmber1968, the chemical study subgroup discovered, mainly by gas chromatographic analysis, that the canned K ncc oil shipped on February 5, 1968, portions of which hod been collected from soma t of the patients, was contaminated with about 1 2,000 ppm of Kanechlor. Kanechlor is a commer- dal brand of polychlorinated biphenyls containing 48 percent chlorine. Skin fat collected from the patients also contained Kanechlor (10). In this discovery. Prof. K. lnagami of the de partment of food technology, Faculty of Agricul ture, Kyushu University, and his associates played a leading role. A technological inspection of the K Company plant revealed that Kanechlor was used as an agent to transfer heat to the processed oil at a reduced pressure so that odorous substances could be removed (see chart). The Kanechlor must have leaked from the heating pipe in whtch it was being conveyed and contaminated the oil, since small openings were discovered in the old P^ To discover whether only the K rice oil shipped in February 1968 was contaminated with Kancchlor, the chemical study subgroup analyzed 109 samples of the bottled K rice oil which, according to the tot numbers, had been shipped between October 1967 and October 1968. This analysis revealed that only the samples shipped on Febru ary 7. 10, 12. 15, 17. 18. and 19 and on March It, 18, and 21. 1968, were contaminated with Kanechlor. This contamination was most marked for the oil shipped on February 7; the contamina tion was rapidly reduced to nil after February 10 (/<?). No analysis could be made of bottled K rice oil shipped February 5 and 6 because of the lack of sample*. Similarly, 479 random samples of bottled rice oil were analyzed for chlorine content by using the X-ray fluorescence method with a count meter from a U.S. instrument manufacturer (the Ap plied Research Laboratory). Again, only those samples shipped from February 7 to 10 contained a large amount of chlorine (maximum 462 ppm). None of the oils shipped in other months were contaminated with more than a trace amount of chlorine. Thus, the results of chemical analysis coincided with those achieved by epidemiologic approaches. Other Epidemiologic Results Babies bom to patients. In latt October 1968 e pregnant patient with Yusho had a stillbirth in which the fetus was unusual; the skin, nails, and gingiva were a peculiar dark brown and stained. Similar fetuses were delivered by other patients with Yusho. Since no such birth or stillbirth was known to have been reported previously in Japan, the study group investigated these occurrences with particular care. The epidemiologic study r 1 } aeciwmr ien, auttau sees / HONS 201030 subgroup questioned all female patients with Yii'hn and .ill unaffected women in the house hold*. of patients with YiinIio about any preg nancies and dvltvcrwif in the period from Febru ary 15, 1968, to the end of the year. Thirteen women--10 with Yusho, two without, and one with status unspecified--had given birth to 11 live born and two stillborn infants during the specified period. (Eight of the live born in fants were males and three were females; one of the stillborn infants was a female and the sex of the other was unspecified ) Among these infanta and fetuses, 10 had the characteristic grayish dark-brown stained skin at birth, five the darkcolored nails and gingiva, and nine an increased eye discharge. Women with Yusho accounted for all these deliveries, except for the delivery of one of the nine babies with eye symptoms. The high incidence of these unusual phenomena among pregnant women with Yusho strongly sug gested that the phenomena were caused by inges tion of the contaminated oil. The fact that nothing unusual was seen in the women's medical histo ries, health conditions, or use of drugs during pregnancy also supported such au interpretation. All these facts, together with the results of chemical assays which showed that the placenta and the subcutaneous adipose tissue of the still born fetuses contained components of Kanechlor, prove that babies and fetuses were affected be cause of the women's ingestion of this particular complex compound (4). The majority of the babies were small for their periods of gestation. As they grew older, their stained skin gradually faded. Postnatal growth curves for the affected babies were found to be simitar in shape to the national standards when both sexes combined were compared. When only males were compared, the curves tor the effected male infants were plainly lower than the national standard curves. A clear comparison between the curves for girls could not be made because of the . limited number of female infants who were af fected by Yusho.- The amount contaminated rice oil that nine of the 13 mothers consumed during pregnancy was estimated w be 03 to 2.6 liters. No distinct dose-effect relationship could be demonstrated be tween oil consumption and the babies' symptoms because of the limited number of cases available for analysis. Nor could evidence be Obtained in regard to any possible physical and mental retar dation of the babies (4). Growth of affected children. To discover whether Yusho disturb; children's growth, the af fected school children, 23 boys and 19 girls, were compared in 1967. 1968, and 1969 with 719 healthy children matched in height and body weight. The gain in height or in body weight which each sick child showed before the poison ing. namely from 1967 to 1968. and Ihc gam after the poisoning, from 1968 to 1969, were compared with the distribution of the corresponding gains seen in the corresponding control group. The gains of the affected boys in both height and weight decreased significantly after the poi soning, white the affected girls showed no definite change in this respect. Because of the limited number of cases available for analysis, no signifi cant relationship could be found between the clin ical severity of the eases and the decrease in gains in height and weight. Discussion This unfortunate episode of rice oil poisoning taught us some important lessons. In the food industries of many countries, no effective adminis trative control and supervision h apparently exer cised over the use of toxic substances like Ka nechlor. These products do not contaminate the final products In normal operations, but they do when the operation is accidentally disturbed or carelessly conducted. In contrast to the case with food additives, too little care has been exercised to avoid the hidden risks in the use of such tubstances. A greatly increased effort is needed to im prove this situation. The polychlorinated biphenyls seem to be par ' tteularly hazardous In this respect because their use Is still increasing in many industries, including food industries. If these substances are absorbed by the body, they wilt easily accumulate, gradually reaching a level high enough to exert adverse ef fects. since they have an extremely slow rate of, mctabolizatioo or excretion. Many patients with Yusho are still tortured by the disease even after more than 2 years of treatment A 1971 report has indicated that the subcutaneous fat of the patients still contains a large amount of polychlorinated biphenyls although their dermal conditions are slowly improving {13), Recently, alto, it has been demonstrated that wide environmental pollution with polychlorinated biphenyls b quietly growing in many countries (//). To prevent another epidemic of Yusho, 10M HSMHA Health SwMto HONS 201639 \ which will break out on a global scale if the pollu tion worsens, radical efforts are urgently needed. Our success in identifying the cause of the epi demic scents to have been due to many factors: (a) common epidemiologic principles were hon estly followed, (A) local health departments and the patients themselves cooperated fully, (c) the causal agent was not multiple, but single, (rf) contamination of the oil with Kancchlor was tem porary and marked, (e) the latent period of the epidemic was several months, and the epidemic was relatively short, (/) Yusho had characteristic symptoms, allowing easy differentiation from other diseases, (g) some of the oil used by the patients, along with the original containers, were still available, and (A) some persons who bought the oil as a group had contracted the disease. AQ these factors facilitated our study. Epidemiologic study of an illness or of adverse effects which were more chronic, less peculiar, which progressed less noticeably than Yusho, end were caused by multiple agents would present greater difficulties. We need to develop more sensitive and more efficient epidemiologic methods. We must, more over, exert constant efforts to make our epide miologic information more accurate, quantitatively and qualitatively, by using the methodology of other disciplines when it seems necessary for our purposes. REFERENCES (/) Cmd. M., and Higuehi, K.; The symptomatology of Yuiho (chlorobiphenyl poisoning) in dermatol ogy. Fukuoka Acta Medica CO: 409-4)1 (14*9). (J) Kuraiwne, M., el a).: An ep*demiotogk study on "Yusho" or chloroMphenyf pobening. Fukuoka Acta Medica CO: )l)-))2 (1969). (J) Yothimura, T.: Epidemiological analysis of "Yusho" patients with special reference to sea. ape. dinical trades end oil consumption. Fukuoka Ada Medica C2: 104-101(1971). (() Yamagoehi, A,, Yoshimwa, T, and Koratsrftie. M.: A survey of preensm women having con sumed rice oil conlcminaled with ctvkwobiphenyls, and of their babies. Fukuoka Acta Medica 62: 117-122(1971). (J) Yothimura. T.: A case control study on growh of school children with "Yutbo." Fukuoka Acta Medica 62: 109-11C (1971). (d) Ikui, H,, Sufi. K, and Up, Ocular sian* of chronic chlorobiphenyl poisontnf ("Yusho"). Fu kuoka Acta Medica CO: 4)2-419 (I9C9). (7) Okumura, M., and Katauki, Clinical observa tion on Yusho (chlorobiphenyl porsoninc). Fuku. oka Acta Medica CO: 440-441 (t9C9). (f) Uxawa. H, Ito, Y Noiomi. A., and Kalsuki. S.: Hyperflyctridemia resuliinf from intake of nee oil contaminated with chlorinated biphenyls. Fukuoka Ada Medica CO: 449-4)4 (I9C9). (9) Kuroiwa, Y., Mbrat, Y_ and Santa. T.: Naurotogical and nerve conduction velocity studies on 2) patients with eblorobiphenyl poisoning. Fukuoka Ada Medica CO; 462-46) (IMS). * (70) Tsukamoto. H,, at al.; fki cbamkal studies on detection of loaie compounds in Ihc nee bran oib used by patients with yusho. Fukuoka Acta Medica CO; 496-912(1969). (//) Risebrough, It. W,, at aL: Poiycblorinsird biphen. yb in the global ecosystem. Nature (London) 220; I09C-1102 (I9CI). (12) Kofimt, T.: Chlorobiphenyh in the iputc end tis sues. Fukuoka Ada Medica C2: 25-29 (1971). (tJ) Taki, I, Hiaaoaga, S., and Amagase, Y.-. Report on Yusho (chlorobiphenyl poisoning) in pregnant women and their fetuses. Fukuoka Ada Medica CO: 471-474 (19C9). KURATSUNE, MASANORI (Kyushu University), YOSIlfMURA, TAKESUMI, MATSU7.AKA, JUMCIJI, ami YAiMAGUClIJ, ATSUKOi Yu*ho, t poitonitig causer/ by rice oil conlmmintteA trii/i polychlorinated 'j^iphenyh, USMllA Health Reports, VoL 86, December I972t pp. 1083--I0VJ. A peculiar skin disease, Yusho, was observed in residents of Fukuoka-Ken, Japan, in Octo ber 196S, wMdt Inter- spread over the wmtwn part of the country. By m* of basic epide miologic methods, the cause of the disease was shown to be a specific brand of rice oil that had become contaminated with a heat transfer agent during processing. This agent, Kancchlor, is a com- picx mixture of polychlorinated disease are still tortured by it biphenyls. even 2 or more years later, de Women who consumed the spite all possible treatment meth contaminated oil delivered babies ods. or fetuses with peculiar darkbrown -'stained skin, nails, and gingiva. The poisoning signifi cantly decreased the growth of the boys who were afTccted, while In view of the stow recovery from the poisoning and the in creasing pollution of the environ ment with these compounds on a the girls who were affected global Kale, radical efforts must showed no such change. Many of ' be made to avoid the hidden the persons who contracted the risks of contamination. I /, '* ' Dmernew sari, Vw. at, n*. u ioi I.-.I. p, i --I I .,11,1 I.I |,,niMpyi HHI t, i v--ivi'Vi-- MON 5 201840 j