risk factors for gliomas and meningiomas in males in los ... · risk factors for gliomas and...

[CANCER RESEARCH 49, 6137-6143. November 1, 1989]

Risk Factors for Gliomas and Meningiomas in Males in Los Angeles County1

Susan Preston-Martin,2 Wendy Mack, and Brian E. Henderson

Department of Preventive Medicine, University of Southern California School of Medicine, Los Angeles, California 90033

ABSTRACT

Detailed job histories and information about other suspected riskfactors were obtained during interviews with 272 men aged 25-69 with aprimary brain tumor first diagnosed during 1980-1984 and with 272individually matched neighbor controls. Separate analyses were conducted for the 202 glioma pairs and the 70 meningioma pairs. Meningi-oma, but not glioma, was related to having a serious head injury 20 ormore years before diagnosis (odds ratio (OR) = 2.3; 95% confidenceinterval (CI) = 1.1-5.4), and a clear dose-response effect was observedrelating meningioma risk to number of serious head injuries (/' for trend= 0.01; OR for > 3 injuries = 6.2; CI = 1.2-31.7). Frequency of full-mouth dental X-ray examinations after age 25 related to both glioma (/'for trend = 0.04) and meningioma risk (/' for trend = 0.06). Glioma, but

not meningioma risk, related to duration of prior employment in jobslikely to involve high exposure to electric and magnetic fields (P for trend= 0.05). This risk was greatest for astrocytoma (OR for employment insuch jobs for > 5 years = 4.3; CI = 1.2-15.6). More glioma cases hadworked in the rubber industry (discordant pairs 6/1) and more worked inhot processes using plastics (9/1). More meningioma cases had jobs thatinvolved exposure to metal dusts and fumes (discordant pairs 13/5), andsix of these cases and two controls worked as machinists. Finally, therewas a protective effect among glioma pairs relating to frequency of useof vitamin C and other vitamin supplements (P for trend = 0.004); theOR for use at least twice a day was 0.4 (CI = 0.2-0.8).

INTRODUCTION

The age-adjusted incidence rate for primary tumors of thebrain and cranial meninges (called BT3) in Los Angeles County

is about 8/100,000/year. Therefore, more than 600 BT arediagnosed each year among Los Angeles County residents.Eighty-six % of all primary BT are gliomas or meningiomas.Studies of the etiology of brain tumors in adults have suggestedthe importance of exposure to ionizing radiation (1-5), physicaltrauma (4-7), farm residence (8-10), familial aggregation (11-12), and various occupational exposures such as vinyl chloridemonomer (13-15), rubber (16-19), formaldehyde (20-22), andelectric and magnetic fields (23-25).

This paper describes a case-control study of gliomas andmeningiomas diagnosed in men in Los Angeles County from1980 to 1984. The purpose of this study was to investigate theimportance of occupational exposures and of other suggestedrisk factors for this disease.

MATERIALS AND METHODS

Cases. The patients were black and white men with a glioma ormeningioma first diagnosed during 1980-1984. Any man who was aresident of Los Angeles County and 25-69 years of age at the time hisBT was diagnosed was eligible for inclusion if he was alive and able tobe interviewed. We decided to only include personal interviews withthe subjects themselves because of the level of detail we sought on jobhistory and other experiences. Because we decided not to accept inter-

Received5/9/89; revised8/1/89; accepted8/7/89.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

' The work was supported by Public Health Service grant POI CA 17054 from

the National Cancer Institute.2To whom requests for reprints should be addressed.3The abbreviations used are: BT. brain tumor; CI. confidence interval; OR.

odds ratio; RR. relative risk.

views with proxy respondents, we were unable to include a largeproportion of otherwise eligible cases because they were deceased orwere too ill or impaired to participate in an interview. The Los AngelesCounty Cancer Surveillance Program identified the cases (26). Alldiagnoses had been microscopically confirmed.

A total of 478 patients were identified. The hospital and attendingphysician granted us permission to contact 396 (83%) patients. Wewere unable to locate 22 patients, 38 chose not to participate, and 60were aphasie or too ill to complete the interview. We interviewed 277patients (74% of the 374 patients contacted about the study or 58% ofthe initial 478 patients). An additional 161 patients who were otherwiseeligible were deceased, only 10 of these were meningioma patients.

Controls. We sought a neighborhood control for each of these 277patients by use of a procedure that defines a sequence of houses onspecified neighborhood blocks. Our goal was to interview the first maleresident in the sequence who corresponded to the patient in race andage (birth year within 5 years of birth year of the patient). We wouldnot interview a match until we ascertained that there were no availablematches at houses visited earlier in the sequence. If no one was homeat the time of the visit, we left a return envelope, an explanatory letterand a brief questionnaire about the age, sex and race of householdmembers at the residence and made a follow-up visit after several days.In 210 instances, the first appropriate person agreed to be interviewed.In 62 neighborhoods, the first match refused, but we were able to locateand interview another matched control in the sequence. For any patient,we visited 40 housing units and made three return visits before weconceded failure to secure a matched control. We were unable to obtaina control in five neighborhoods. In all, we identified and interviewed272 controls. These 272 controls and the corresponding 272 cases wereincluded in the analysis.

The Interview. A questionnaire sought information on various lifeexperiences that had occurred 2 years or more prior to the year ofdiagnosis of the case. The first and longest section obtained a detailedjob history including information on specific job tasks and materialsused. After the work history was obtained, specific questions were askedabout exposures, on any job, to each of: 45 substances; five processes(e.g., welding); radiation or radioactive materials; and farming andvarious related exposures. Questions were also asked about occupational illness or injury. There were also questions about head trauma;head X-rays; relatives with nervous system tumors or cancer; andconsumption of tobacco, alcohol, vitamin supplements, and certainfoods. It was not feasible for interviews to be conducted blindly, but allquestions were asked in a standard manner. Both members (case andcontrol) of each matched pair were interviewed by the same interviewerand both were interviewed by the same method, either in person or bytelephone. Interviews were conducted from August 1980 through December 1985.

Statistical Analysis. In the analysis of dichotomous exposure variables, we used matched OR (the ratio of discordant pairs) to estimaterelative risks and used the exact binomial test to calculate associated Pvalues and CIs. Conditional logistic regression models were used forthe dose-response analysis of a single variable considered at more thantwo levels, for tests of trend, and for multivariate analyses (27). If, forany variable, the information for either the patient or the control wasnot known, we excluded the pair from the relevant analysis. All statistical significance levels (P values) cited are based on a likelihood ratiotest and are two-sided for occupational factors and for tobacco andalcohol use. P Values are one-sided for other variables which wereasked to test one-directional hypotheses relating to variables such asX-rays, head trauma, and vitamin use. However, the 95% confidenceintervals (CI) shown are two-sided. In tests for trend, factors areconsidered as continuous variables whenever possible.

Analyses of occupational data compared the distributions of casesand controls by blue versus white collar and by major occupation (10

6137

Research. on November 17, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

RISK FACTORS FOR GLIOMAS AND MEN1NGIOMAS

groups) and industry groups (11 groups). Comparisons were also madeof the numbers who worked in specific occupations or industries orspecific groups of occupations or industries that had been identified inthe literature as having apparent excesses of brain tumors. Similarly,comparisons were made for each of the 45 substances, five processes,and other exposures queried separately as well as the hundreds ofsubstance exposures reported during the initial job history.

RESULTS

Analyses used data for 202 glioma pairs and 70 meningiomapairs. Only a few pairs were black (11 glioma and five meningioma pairs). Glioma cases and controls were similar withregard to median birthyear (1938 for both); mean age at dateof diagnosis of the case (45.2 years for cases; 45.1 for controls);and mean number of years of education (13.3 versus 13.7).Meningioma pairs were also similar with regard to these variables: median birthyear (1924 versus 1926); mean age (55.3versus 54.5); and mean years of education (13.1 for both). Thedistribution by histolÃ³gica! type is shown in Table 1.

Serious Head Trauma. Table 2 compares cases and controlson prior serious head injury. "Serious head injury" was defined

Table 1 HistolÃ³gica!types of primary brain tumors (gliomas and meningiomas)in men aged 25-69, Los Angeles County, 1980-1984

HistolÃ³gica! type Number

GliomasAstrocytoma. NOS"

Astrocytoma, anaplastic typeProtoplasmic, astrocytomaGemistocytic astrocytomaFibrillary astrocytomaGlioblastoma, NOSGiant cell glioblastomaMalignant gliomaMixed gliomaSubependymal gliomaChoroid plexus papillomaEpendymomaPapillary ependymomaOligodendrogliomaTotal gliomas

MeningiomasMeningioma. NOSMeningotheliomatous meningiomaFibrous meningiomaPsammomatous meningiomaHemangioblaslic meningiomaHemangiopericytic meningiomaTransitional meningiomasTotal meningiomas

982I

121

431561I41

20202

163372615

70a NOS. not otherwise specified.

Table 2 Comparison of male brain tumor cases (glioma and meningioma) andcontrols on prior serious head injury,0 Los Angeles County, 1980-1984

Gliomas Meningiomas

Discordant Discordantpairs OR (95% Cl) pairs OR (95% CI)

Had serious head injury," 20 or more

years before year ofdiagnosis41/49

0.8 (0.5-1.3) (23/10) 2.3 (1.1-5.4)

Dose-response relationship, meningiomas

Number of serious head Number of Number ofinjuries* cases controls OR (95% CI) P trend

012>330191293921g21.01.32.16.2(0.6-2.9)(0.8-5.9)(1.2-31.7)0.01

" "Serious injury" was any injury that resulted in a medical visit, loss of

consciousness, or dizziness.* Includes all serious injuries up to reference year (i.e., 2 years before the year

of diagnosis of the case).

to include all head injuries occurring before reference year (2years before the year of diagnosis of the case) that led to lossof consciousness, dizziness, or a medical consultation. Menwho had a serious head injury 20 or more years before the yearof diagnosis of the case had a significantly increased risk ofmeningioma, but not of glioma. Head injury appeared not tobe associated with glioma risk no matter what the assumedlatent period (interval between first head injury and diagnosis).The increase in meningioma risk was greater after an intervalof 20 years than after intervals of 5, 10, or 30 years. In a dose-response analysis, the meningioma risk increased with an increase in the number of serious head injuries. (ORs for 0, 1, 2,and 3+ head injuries 10 or more years prior to diagnosis = 1.0,1.6, 2.3, and 4.6; P trend = 0.01).

Exposure to X-Rays. Table 3 compares cases and controls onprior dental radiography. Questions were asked about frequencyof any type of dental X-ray examination and frequency of full-mouth X-ray series in particular for two periods (up to age 25;after age 25). Whether or not subjects had ever been exposedto dental X-rays was not a good discriminator because about90% of all controls had such an exposure. Exposure to dentalX-rays before age 25 was less common than exposure after age25, and exposure to full-mouth X-ray examinations (which usean average of 18 intraoral films) was less common than exposure to dental X-rays of any type. Few men had full-mouthexams before age 25. Glioma risk increased with an increase inthe frequency (from never to yearly) of full-mouth examinationsafter age 25; a similar trend was seen for meningiomas. Bothglioma and meningioma risk also increased with an increase inthe frequency of exposure to dental X-rays of any type beforeage 25, although these trends did not reach statistical significance. The way these questions were asked did not allow for ananalysis by calendar year of exposure; the per film dose fromdental radiography had declined dramatically since 1920 andas recently as the 1960s was many times higher than it is today(28). Cases and controls were similar with regard to medicalradiography of the head before reference year. Pairs in whicheither the case or the control had exposure to the brain orcranial meninges from radiation treatment to the head or neckprior to reference year (3 cases; 3 controls) were excluded fromall X-ray analyses.

Occupational History. Some significant case-control differences related to major occupational categories. Compared totheir controls, more glioma cases had spent the majority oftheir working years in blue collar jobs [discordant pairs, 46/31;OR = 1.5; P (two-sided) = 0.09]. This difference was attributable mainly to more cases working as transport or equipmentoperatives (discordant pairs 7/3; OR = 2.3; P = 0.22) or aslaborers (11/3; OR = 3.7; P = 0.05). Of those who spent themajority of their working years as laborers, nine cases and nocontrols worked as warehousemen, freight and material handlers, or stock handlers and they held such jobs for an averageof 12 years. Seven of these nine said they were exposed tovarious types of dust (two to chlorine), three to acids, and oneeach to methylethylketone, formaldehyde, petroleum products,plastics, and rubber.

Similar proportions of meningioma cases and controls spentthe majority of their working years in blue versus white collarjobs, but more meningioma cases worked in service jobs formore than 5 years (discordant pairs, 10/4; OR = 2.5; P = 0.12).Five cases versus six controls worked as policemen, securityguards, or detectives; the mean years at these jobs was 16.5 forcases versus 9.0 for controls. Three cases and no controls wereprofessional cooks. Three cases worked an average of 17 years

6138



RISK FACTORS FOR GLIOMAS AND MENINGIOMAS

Table 3 Comparison of male brain tumor cases (glioma and meningioma) and controls on frequency of exposure to dental radiography, Los Angeles County,1980-1984

Gliomas0No.

Ca No. Co OR P" (95%Cl)Frequency

of dentalX-rays up to age 25Never or < every 5 67 84 1.0

yearsEvery 2-5 years 87 74 1.5 0.04 (1.0-2.4)Once a year 40 36 1.4 0.12(0.8-2.4)Frequency

of full-mouthX-rays after age 25

Never or < every 5 168 177 1.0years

Every 2-5 years 20 15 1.6 0.15 (0.7-3.6)Once a year 7 3 3.0 0.09(0.6-14.9)"

Pairs in which either the case or control had exposure to the brain or cranial*One-sided.as

janitors compared to 1 control who worked as a janitor for4 years.

In addition, more meningioma cases worked as a manager oradministrator for more than 5 years (17/6; OR = 2.8; P =0.03). Nothing striking emerged, however, in an analysis byspecific jobs held.

Employment in occupations likely to involve high exposureto electric and magnetic fields appeared to increase risk ofgliomas but not meningiomas (Table 4). The association wasstrongest with astrocytomas (OR for working in such jobs > 5years = 4.3; CI = 1.2-15.6). Risk increased with an increase inthe number of years working in these occupations (P for trend= 0.008). The occupations in this rubric are listed in a footnoteto Table 4.

No subjects worked in the manufacture of Polyvinylchloride,but 14 glioma cases versus seven controls had jobs that involvedcontact with other plastics (i.e., other than Polyvinylchloride,polystyrene, polyethylene, or polyurethane; discordant pairs,13/6; Table 5). A comparison by job type showed that ninecases and one control worked in the manufacture of plasticproducts or heated plastic to melt it down, whereas similarnumbers of cases and controls cut or shaped plastic (four cases;four controls) or used plastic materials for bonding or covering(one case; two controls). Most subjects did not know the nameof the plastic compound they used.

More cases than controls answered yes when asked if theyhad ever worked in rubber processing (discordant pairs: glioma,6/1; meningioma, 3/2; Table 5). In addition, when giving theirjob histories, more cases had volunteered previously (intheTable

4 Comparison of male brain tumor cases (glioma and meningioma) andcontrols on prior employment in occupations likely to involve high exposure to

electric and magnetic fields' Los Angeles County,1980-1984Gliomas

MeningiomasDiscordant

Discordantpairs OR (95% CI) pairs OR (95%Cl)14/8

1.8 (0.7.4.8) 2/3 0.7(0.1-5.8)Dose-response

analysis,gliomasYears

worked No. Ca No. Co OR (95% CI) Ptrend0

years 172 182 1.0 0.05>0- 5 years 16 12 1.4 (0.7,3.1)>5 years 14 8 1.8(0.8.4.3)"

Variable recorded as binary based on subject's working more than 5 years in

any jobs coded to the following 1970 U. S. Census occupation codes: 012, 171,384. 430. 43 1. 475, 482. 484, 485, and 554. These jobs include electrical engineers,radio operators, telegraph operators, electricians, electrician apprentices, dataprocessing machine repairmen, household appliance and accessory installers andmechanics, office machine mechanics and repairmen, radio and television repairmen, and telephone linemen and splicers.Meningiomas"P

trend* No. Ca No. Co OR P" (95% Cl) Ptrend*0.07

36 38 1.00.2621

22 1.1 0.43 (0.5-2.2)11 8 1.5 0.23(0.5-4.4)0.04

51 58 1.00.0615

9 2.0 0.08 (0.8-5.0)2 1 2.5 0.23(0.2-29.5)meninges

from prior radiation treatment were excluded from theseanalyses.Table

5 Other job exposures positively associated with occurrence of brain tumors(gliomas and meningiomas) in men, Los Angeles County,1980-1984Gliomas+

-/- + OR 95%CLExposures

asked about explicitly (exposed at leastweekly)

Rubber processing 6/1 6.0 0.7, 276.0Used any other type of pÃas- 13/6 2.2 0.8,6.9

tic?"

Synthetics (nylons, rayons. 6/1 6.0 0.7,276.0etc.)Volunteered

information onexposures (ever exposed)*

Rubber or rubber dust 7/2 3.5 0.7, 34.5Rayons 3/0 Inf 0.6.InfMeningiomas+

-/- + OR 95%CLExposure

asked about explicitly (exposed at leastweekly)

Any other petroleum prod- 18/5 3.6 1.3. 12.4uct such as gasoline orkerosene'

An> other metal dusts or 13/5 2.6 0.9.9.3fumes''Volunteered

information onexposures (ever exposed)*

Rubber or rubber dust 4/0 Inf 0.9. InfAmmonia 7/2 3.5 0.7. 34.5Kerosene 9/4 2.3 0.6.10.0"

Plastics other than Polyvinylchloride, polystyrene, polyethylene, and polyure

thane.* Information on frequency of exposure not available.' Petroleum products other than paints, cooling, cutting, or lubricating oils,

coal tar. soot, pitch, or smudge.a Metals other than aluminum, arsenic, beryllium, cadmium, lead, mercury,

and nickel.' Inf,infinity.occupational

history section of the questionnaire) that theywere exposed to rubber or rubber dust when asked about substances they came in contact with (discordant pairs: glioma, 7/2; meningioma, 4/0; Table 5). Four subjects worked in tireplants: two glioma cases worked 24 and 9 years, and onemeningioma case and one meningioma control each worked 2years in the manufacture of rubber tires.

Six glioma cases and one control had jobs working withsynthetics (e.g., nylon, rayon, and other polyesters), but onlyone of these seven (one case) worked in the manufacture ofsynthetics and the other six worked only with the finishedproduct. Three of the cases (and no controls) had volunteeredpreviously that their jobs involved working with rayon, and one

6139



RISK FACTORS FOR CHOMAS AND MEN1NGIOMAS

of these worked with rayon for 12 years.There were no remarkable differences between cases and

controls in job exposure to specific petroleum products such aspaints, lacquers, or coal tar, soot, or pitch; but more meningi-oma cases than controls said they had worked with otherpetroleum products such as gasoline or kerosene at least weekly(18/5, Table 5). Ten of these cases and three controls usedgasoline or another petroleum-based solvent to clean theirhands or clean items they worked with, and seven cases andfour controls used it daily as a fuel and inhaled the fumes.Three cases and three controls used airplane fuel. Ten menin-gioma cases and five controls used cooling, cutting, or lubricating oils, and most used these daily (eight cases; four controls).Seven meningioma cases and two controls had daily exposureto inks and most mixed inks or worked with wet inks (fourcases; one control).

Information volunteered about specific job exposures in theoccupational history section of the questionnaire suggested thatkerosene might be an exposure of interest (nine meningiomacases; four controls), but most exposures appeared relativelytrivial (e.g., "occasionally filled a kerosene lamp"), and only

two of the nine cases used kerosene at least weekly.More meningioma cases had jobs that involved exposure to

metal dust and fumes (discordant pairs 13/5; Table 5). Amongthose with metal dust exposure, six cases and two controlsworked as machinists and both cases and controls worked asmachinists for an average of 5 years. Similar numbers of casesand controls were exposed to metals as welders (two cases, twocontrols), inspectors (two cases, one control) or as laborers ofvarious sorts (four cases, two controls).

No other significant case-control differences were apparentin the analysis of occupational factors. This analysis includedcomparisons on the 37 specific substances and five specific jobsqueried as well as comparisons by job titles, industry of employment, and information given about exposures in the jobhistory section of the questionnaire. Some comparisons weremade by grouping job titles or exposures in an attempt toconfirm associations reported in the literature. A similar number of cases and controls worked in the aircraft industry (glioma,34 cases and 34 controls; meningioma, 13 cases and 14 controls). Also, cases and controls were similar with respect to jobexposure to ionizing radiation.

Other Exposures. Table 6 compares cases and controls onsmoking status and consumption of alcoholic beverages. Therewere no remarkable differences between cases and controls withregard to history of cigarette smoking when we considered agestarted smoking, amount smoked, years smoked, and pack-years of exposure ( 1 pack-year is equivalent to smoking a packa day for a year, i.e., 7300 cigarettes). The only significantdifference in relation to consumption of beer, wine, and hardliquor was that fewer meningioma cases than controls were beer

Table 6 Comparisons of male brain tumor cases (gliomas and meningiomas) andcontrols by cigarette smoking and alcohol consumption, Los Angles County,

1980-1984

GliomasDiscor

dantpairsEver

smokedEver drank at least

once a month:BeerWineHard liquor+

-/3532

3955-

+524453

44OR0.70.70.71.3MeningiomasDiscor

dantpairs(95%

CI) + -/ -+(0.4-1(0.5-1

(0.5-1(0.8-1.0).2)

.1)

.9)1771415141821

21OR1.20.40.70.7(95%

CI)(0.6-2.7)(0.1-0.9)

(0.3-1.4)(0.3-1.4)

drinkers (i.e., drank beer at least once a month). No differenceswere apparent in maximum amount of alcohol drunk regularly.

There were no significant differences between cases and controls in their consumption of various cured meats or of citrusfruits. Among glioma pairs, a significant protective effect wasassociated with the use of vitamin supplements, and increasingprotection related to increasing frequency of use (one-sided Pfor trend = 0.004). The OR for use at least twice a day was 0.4(95% CI = 0.24-0.77). The protective effect was somewhatstronger for use of vitamin C tablets than for other vitaminpreparations. Among meningioma pairs, no protective effect ofvitamin supplements was seen but a protective effect was seenfor consumption of citrus fruit (OR for ate citrus five or moretimes a week = 0.4; P = 0.07; P for trend = 0.12). Also,meningioma risk appeared to increase with increasing consumption of all types of cured meats combined, but this findingwas not statistically significant.

Fewer cases than controls had ever lived on a farm, andamong the glioma pairs this difference was statistically significant (OR = 0.5; P = 0.016). When just those subjects who livedon a farm were considered, fewer cases reported exposure tovarious farm animals, crops, pesticides, and fertilizers. In ananalysis by decade when they worked on a farm, the onlysignificant finding was that more glioma cases did farm workin the 1950s (discordant pairs, 6/1; OR = 6.0; P = 0.05).

The same number of cases and controls had blood relativeswith cancer (discordant pairs: glioma, 42/42; meningioma, 12/12). More meningioma cases had two or more relatives withcancer (10/3; OR = 3.3; P = 0.02), but there was nothingremarkable about the site distribution of these cancers. A similar number of cases and controls had relatives with nervoussystem tumors. Two meningioma cases had a close blood relative with a meningioma (one had a son diagnosed at age 17, theother a sister diagnosed at age 61). No cases had a prior primarybrain or other nervous system tumor, and no controls hadnervous system tumors.

DISCUSSION

The present study has limited generalizability because itincludes only living subjects who were able to be interviewed.Although interviewers could not be blinded as to case or controlstatus, they were blinded as to the study hypotheses, and everyeffort was made to conduct interviews with cases and controlsin a like manner.

The study does support various associations reported previously such as the association of brain tumors with variousoccupational factors, (e.g., working in the rubber industry) andwith frequent full mouth dental X-rays; however, not all suchpreviously reported differences are statistically significant here.This study does not support other previously suggested associations. For example, we see no increase in risk related tocigarette smoking or wine drinking. The hypothesis that exposure to jV-nitroso compounds causes brain tumors has gainedlimited support from the association with employment in therubber industry and from the clearly protective effect seen forgliomas of frequent use of vitamin C and other vitamin supplements (both discussed below). Finally, this study, by comparingdata for the two major histolÃ³gica! groups of primary braintumors, has helped to clarify the relative importance of suspected risk factors such as head trauma in the etiology of each.

Serious Head Trauma. The possibility of recall bias must beconsidered in all such case-control interview studies, but it is ofparticular concern when investigating factors such as trauma

6140



RISK FACTORS FOR GLIOMAS AND MENING1OMAS

which among lay persons is often thought to relate to tumordevelopment. The concern is that direct questions about headinjuries may be answered more completely by brain tumorpatients than controls because patients may already havethought an injury caused their tumor; this phenomenon is called"differential recall." For this reason, we limited our analysis ofthis variable to "serious head injuries" (i.e., those injuries that

resulted in a medical visit, loss of consciousness, or dizziness)that occurred 2 or more years before the year of diagnosis. Thefact that head trauma was found to be a risk factor for menin-gioma but not for glioma provides further reassurance that thisfinding is not attributable to differential recall. Also, it supportsfindings from other recent case-control studies which did notfind an association of head trauma with gliomas (29) but didfind an association with meningiomas (4, 5). In the presentstudy, the association was strongest with serious head injury 20or more years prior to the year of diagnosis, and risk increasedwith an increase in the number of serious head injuries. A fewdramatic case reports suggest that trauma can lead to thedevelopment of meningioma. For example, a boiler explosiondrove a wire 1 cm long into the brain of a man who 20 yearslater was found to have a large meningioma with the wire at itscenter (30). Another man, who struck his head when he wasthrown from a stretcher by a bomb explosion, developed a smalllump on his head at the site of the injury. Twenty years later,the lump began to gradually increase in size and after 5 yearswas surgically removed and histologically diagnosed as a meningioma (7). The present study further suggests that headtrauma is an important factor in the development of meningiomas, and that it is unlikely to be an important factor in thedevelopment of gliomas.

Exposure to X-Rays. Both major tumor types, on the otherhand, appear to be related to the frequency of full-mouth dentalX-rays after age 25. There was also the suggestion of a relationship with the frequency of exposure to dental X-rays (of anytype) before age 25. Meningiomas have been related previouslyto frequency of full-mouth dental X-ray series and to young ageat first full-mouth series (4, 5). In addition, the risk for braintumors of all histological types in young people age 15-24 wasincreased among those who had five or more full-mouth examinations starting at least 10 years before diagnosis (31). Arecent study of gliomas in adults, however, did not find anassociation with having ever had dental X-rays (29). This variable (ever versus never had dental X-rays) is likely, however, tobe a poor discriminator. It does not correlate well with totalexposure from dental radiography (28), and, in the present dataset, about 90% of controls ever had their teeth X-rayed. A betterindicator of exposure from dental radiography is total numberor frequency of dental X-ray examinations, in particular full-mouth examinations. Dental X-rays account for most of theexposure of the head from diagnostic radiography, and cumulative exposure from dental X-rays can be quite high (in a recentstudy of cases and controls of similar ages to those in thepresent study, 37% had >5 rads and 4% >50 rads exposure tothe parotid gland, 28).

Since dental X-ray examinations are not particularly memorable events, recall of them may be poor. We have asked similarquestions about dental radiography in several studies includingone that validated interview information by a review of dentalcharts. Results of this validation study suggest that any mis-classification in interview data on dental X-rays is similar forcases and controls and that these data are good enough to usealone in the analysis of case-control studies (28).

Ionizing radiation can cause brain tumors, and a recent

follow-up of the Israeli tinea capitis cohort suggests that thisassociation is stronger for meningiomas (RR = 9.5) than forgliomas (RR = 2.6; Ref. 32). We were unable to evaluate thispotential difference in our study because our data on exposureto X-rays and other sources of ionizing radiation was limited.However, a test for interaction between tumor type and fullmouth X-ray frequency after age 25 was not statistically significant implying no differential risk by histological type for thisX-ray variable.

Occupational History. We confirmed the association ofglioma with previous employment in the rubber industry (OR= 6.0). Several earlier studies have found a similar association(29, 33-35) but others have not (36, 37). Four of the fivepositive studies found the association to be specific to workingin the manufacture of tires, and the fifth study gave no information on type of job held in the rubber industry (29). Thethree negative studies, on the other hand, looked at employmentin other branches of the industry or in the industry as a whole.In our study, two glioma cases, one meningioma case, and onemeningioma control were employed in the manufacture of tires.Tire plants have been found to have high levels of contamination with various /V-nitroso compounds (A'-nitrosodimethyla-mine, Â¿V-nitrosodiethylamine, /V-nitrosomorpholine, and N-ni-trosodiphenylamine) in the air, soil, floor scrapings, steamcondensate, and compounding chemicals (38). Hot processessuch as tire curing had the highest levels of volatile A'-nitrosocompounds, in particular 7V-nitrosomorpholine. Various other./V-nitroso compounds (in particular the nitrosoureas) have beenshown experimentally to cause brain tumors in a variety ofspecies (39, 40).

The association with exposure to vinyl chloride reportedpreviously (14, 41) was not found in our study. The possibleassociation we found of gliomas with working in hot processesusing other plastics needs to be investigated in other studies.

We also confirmed an association of gliomas, in particularastrocytomas, with duration of employment in jobs involvingmore than the usual exposure to electric and magnetic fields.This association with gliomas (23, 25) and with astrocytomasin particular (24) has been reported previously. Although it istempting to assume that this increase is attributable to increasedexposure to extremely low frequency electromagnetic radiation,the mechanism for such an effect is not understood and thepossible association of brain tumor risk with other exposurescommon to this industry (such as organic solvents) must beconsidered (24).

More meningioma cases than controls reported exposure tometal dust and fumes and this difference was attributable tomore cases working as machinists. Machinists and other precision metal workers have been shown to have an elevated braintumor risk (43). In addition, more of our meningioma casesused metal working fluids such as cooling and cutting oils. Suchfluids have been shown to contain A'-nitroso compounds (38).

Descriptive studies have shown that glioma incidence andmortality rates are highest among those in the highest socialclasses regardless of whether social class is defined by censustract of residence or by occupation (43-45). The present study,in contrast, found an increase in glioma risk associated withemployment (for most of one's working years) in blue collar

jobs. The most striking difference in this group was that ninecases and no controls had worked as warehousemen or freighthandlers. The most common exposures mentioned in relationto these jobs were dusts of various sorts (seven cases) and acids(three cases).

Compared to controls, more of our meningioma cases worked6141




as managers and administrators, and more worked in the serviceindustry. Similar numbers of cases and controls in serviceindustry jobs worked as policemen, watchmen, and firemen, butmore cases worked as cooks or janitors. Our study did not findother previously reported associations of brain tumors withoccupational factors such as employment in the petrochemicalindustry or the aircraft industry.

Other Factors. We failed to confirm recently suggested associations of brain tumors with smoking nonfilter cigarettesand with wine drinking (29), but our study had more limiteddata on tobacco and alcohol use and did not ask explicitly aboutnonfilter cigarettes. We also found no increase in risk relatedto farm residence or exposure to farm animals or agriculturalchemicals although such a relationship has been reported previously (46). Cases and controls were not different in theirpersonal or family histories of previous nervous system tumorsor cancers.

The hypothesis that exposure to jY-nitroso compounds causeshuman brain tumors received only limited support from thepresent study. Glioma risk was elevated among those who hadworked in rubber processing, an industry heavily contaminatedby N-nitroso compounds. Also, among glioma pairs we observed a protective effect for consumption of vitamin C andother vitamin supplements. Vitamins C and E have been shownto be effective in blocking the endogenous formation of N-nitroso compounds in the gut from precursors ingested in foodand water (47). The protective effect was strongest among thosewho took vitamins at least twice a day. It seems likely that thisgroup took the supplements around meal time when they wouldbe most effective to act as nitrosation inhibitors. Compared totheir controls, meningioma cases did not use more vitaminsupplements but they did consume more citrus fruit; citrus fruitcontains high levels of vitamin C. Fruit consumption was alsofound to have a protective effect in a recent brain cancer study(29); most fruits contain vitamin C and/or E. Although thefinding of a protective effect for vitamins and fruit can beinterpreted as being consistent with the hypothesis that braintumor risk relates to exposure to ./V-nitroso compounds theremay be some entirely different explanation. Our study is limited, however, in its ability to test this hypothesis. We failed tofind an association with the only other relevant factor queried:cured meats, which contain nitrite, a precursor of TV-nitrosocompounds. We have no information on other dietary sourcesof ./V-nitroso compounds, precursors, or modulators of theirmetabolism (either inhibitors or catalysts). We also have nodata on water supply, drug use, or other potential sources of N-nitroso exposure.

ACKNOWLEDGMENTS

We thank Dr. William Wright for assistance in coding substanceexposures, John Mouzakis for conducting the interviews, AureliaChang for programming assistance, and Camilla Turner for preparationof the manuscript.

REFERENCES

1. Modan. It.. Baidalz. I).. Mart. 11.. and Steinitz, R. Radiation induced headand neck tumors. Lancet, /: 277-279, 1974.

2. Shore, R. E., Albert, R. E., and Pasternack, B. S. Follow-up study of patientstreated by X-ray epilation for tinea capitis. Arch. Environ. Health, 31: 21-28, 1976.

3. Schneider, A. B., Shore-Freedman, E., Ryo, U. Y., Bekerman. C., Favus, M.,and Pinsky, S. Radiation-induced tumors of the head and neck followingchildhood irradiation. Medicine, 64: 1-15, 1985.

4. Preston-Martin, S., Paganini-Hill, A., Henderson, B. E., Pike, M. C., andWood, C. Case-control study of intracranial meningiomas in women in Los

6.

7.

8.

9.

10.

11.12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

Angeles county. J. Nati. Cancer Inst., 75:67-73, 1980.Preston-Martin, S., Yu, M. C., Henderson, B. E., and Roberts, C. Riskfactors for meningiomas in men in Los Angeles County. J. Nati. CancerInst., 70: 863-866, 1983.Parker, H. L., and Kernohan, J. W. The relation of injury and glioma of thebrain. J. Am. Med. Assoc., 97: 535-540, 1931.Walsh, J., Gye, R., and Connelly, T. J. Meningioma: a late complication ofhead injury. Med. J. Australia, /.- 906-908, 1969.Haenszel, W., Marcus, S. S., and Zimmerer, E. G. Cancer morbidity in urbanand rural Iowa. Public Service Health Monograph No. 37, U. S. Dept. HealthEd. Welfare. Washington, DC: U. S. GPO, 1956.Choi, N. W., Schuman, L. M., and GÃ¼llen,W. H. Epidemiology of primarycentral nervous system neoplasms. I. Mortality from primary central nervoussystem neoplasms in Minnesota. Am. J. Epidemici., 91: 238-259, 1970.Cancer Registry of Norway: Cancer Registration in Norway. The incidenceof cancer in Norway, 1953-1958. Oslo, The Norwegian Cancer Society,1961.Van der Wiel, J. J. Inheritance of glioma. Amsterdam: Elsevier, 1960.Joynt, R. J., and Perret, G. E. Meningiomas in a mother and daughter. Caseswithout evidence of neurofibromatosis. Neurology, //: 164-165, 1961.Tabershaw, I. R., and Gaffey, W. R. Mortality study of workers in themanufacture of vinyl chloride and its polymers. J. Occup. Med., 17: 509-518, 1974.Monson, R. R., and Peters, J. M. Proportional mortality among vinylchlorideworkers. Lancet, 2: 397-398, 1974.Waxweiler, R. J., Stringer, W., Wagoner, J. K., Jones, J., Falk, H., andCarter, C. Neoplastic risk among workers exposed to vinyl chloride. Ann. N.Y. Acad. Sci., 271: 40-48, 1976.McMichael, A. J., Spirtas, R., and Kupper, L. L. An epidemiologie study ofmortality within a cohort of rubber workers, 1964-72. J. Occup. Med., 16:458-464, 1974.McMichael, A. J., Spinar, R., Gamble, J. F., and Tousey, P. M. Mortalityamong rubber workers: relationship to specific jobs. J. Occup. Med., 18:178-185, 1976.Monson, R. R., and Fine, L. J. Cancer mortality and morbidity among rubberworkers. J. Nati. Cancer Inst., 61: 1047-1053, 1978.Hakama, M., and Kilpikari, I. Cancer risk among rubber workers. J. Toxicol.Environ. Health, 6: 1211-1218, 1980.Harrington, J. M., and Oakes, D. Mortality study of British pathologists1974-1980. Br. J. Int. Med., 41: 188-191, 1984.Walrath, J., and Fraumeni, J. F. Cancer and other causes of death amongembalmers. Cancer Res., 44:4638-4641, 1984.Stroup, N. E., Blair, A., and Erikson, G. E. Brain cancer and other causes ofdeath in anatomists. J. Nati. Cancer Inst., 77: 1217-1224, 1986.Lin, R. S., Dischinger, P. C., Conde, J., and Farrell, K. P. Occupationalexposure to electromagnetic fields and the occurrence of brain tumors. J.Occup. Med., 27: 413-419, 1985.Thomas, L., Stewart, P. A., Stemhagen, A., Correa, P., Norman, S. A.,Bleecker, M. L., and Hoover, R. N. Risk of astrocytic brain tumors associatedwith occupational chemical exposures. A case-referent study. Scand. J. WorkEnviron. Health, 13: 417-423, 1987.Speers, M. A., Dobbins, J. G., and Miller, V. S. Occupational exposures andbrain cancer mortality: a preliminary study of East Texas residents. Am. J.Indus. Med., 13: 629-638, 1988.Hisserich, J. C., Preston-Martin, S., and Henderson, B. E. An areawidereporting network. Pubi. Health Rep., 90: 15-17, 1975.Breslow, N. E., and Day, N. E. Statistical methods in cancer research. In:W. Davis, (ed.). The Analysis of Case-Control Studies, Vol. 1, p. 211,1 ARCScientific Publication No. 32. Lyon, France: International Agency for Research on Cancer, 1980.Preston-Martin, S., Thomas, D. C., White, S. C., and Cohen, D. Priorexposure to medical and dental X-rays related to tumors of the parotid gland.J. Nati. Cancer Inst., SO:943-949, 1988.Burch, J. D., Craib, K. J. P., Choi, B. C. K., Miller. A. B., Risch, H. A., andHowe, G. R. An exploratory case-control study of brain tumors in adults. J.Nati. Cancer Inst., 78: 601-609, 1987.Reinhardt, G. Trauma-Fremdkorper-Hirngeschwulst. Munich Med. Woch-enschr., 75:399-401, 1928.Preston-Martin, S., Yu, M. C., Benton, B., and Henderson, B. E. /V-Nitrosocompounds and childhood brain tumors: a case-control study. Cancer Res.,42:5240-5245,1982.Ron, E., Modan, B., Boice, J. D., Alfandary, E., Stovall, M., Chetrit, A., andKatz, L. Tumors of the brain and nervous system following radiotherapy inchildhood. New Engl. J. Med., 319: 1033-1039, 1988.Mancuso, T. F. Tumors of the central nervous system. Industrial considerations. Acta UniÃ³Int. Contra Cancrum, 19:488-489, 1963.Mancuso, T. F., Ciocco, A., and El-Attar, A. A. An epidemiologie approachto the rubber industry. A study based on departmental experience. J. Occup.Med., 10: 213-232, 1968.Monson, R. R., and Nakano, K. K. Mortality among rubber workers. I.White male union employees in Akron, Ohio. Am. J. Epidemici., 103: 284-296, 1976.Fox, A. J., Lindars, D. C., and Owen, R. A. Survey of occupational cancerin the rubber and cablemaking industries: results of a five-year analysis,1967-71. Br. J. Ind. Med., 31: 140-151, 1974.Andjelkovic, D., Taulbee, J., and Symons, M. Mortality experience of acohort of rubber workers, 1964-1973. J. Occup. Med., 18: 387-394, 1976.Rounbehler, D. P., and Fajen, J. M. A'-Nitroso compounds in the factory

6142




environment (Contract No. 210-77-0100). Cincinnati: Nail. Inst. Occup. R. N. Occupational risk factors for brain tumors. A case-referrent death-Safety Health, 1983. certificate analysis. Scand. J. Work Environ. Health, 12: 121-127, 1986.

39. Magee, P. N., Montesano, R.. and Preusmann, R. /V-Nitroso compounds and 44. Preston-Martin. S., and Mack. W. Nervous system. In: D. Schottenfeld andrelated carcinogens. In: E. D. Searle, (ed.). Chemical Carcinogens. Washing- J. F. Fraumeni (eds.). Cancer Epidemiology and Prevention. New York:ton, DC: American Chemical Society, 1976. Oxford University Press, in press. 1989.

40. Bogovski, P., and Bogovski, S. Animal species in which Â¿V-nitrosocompounds 45. Buell, P., Dunn, J. E., and Breslow, L. The occupational-social class risks ofinduce cancer. Int. J. Cancer, 27:471-474. 1981. cancer mortality in men. Cancer (Phila.). 12: 600-621, 1960.

41. Cooper, W. C. Epidemiologie study of vinyl chloride workers: mortality 46. Musicco, M., Filipini, G., Bordo, B. M., Melotto, A., Berrino, F., andthrough December 31, 1972. Environ. Health Persp., 41: 101-106, 1981. Morello, G. Gliomas and occupational exposure to carcinogens: case-control

42. Modan, B. Exposure to electromagnetic fields and brain malignancy: A newly study. Am. J. Epidemiol., 116: 782-790, 1982.discovered menace? Am. J. Indus. Med., 13: 625-627, 1988. 47. Tannenbaum, S. R., and Mergens, W. Reaction of nitrite with vitamins C

43. Thomas, T. L., Fontham, E. T., Norman, S. A., Stemhagen. A., and Hoover, and E. Ann. NY Acad. Sci., 355: 267-277, 1980.

6143



1989;49:6137-6143. Cancer Res Susan Preston-Martin, Wendy Mack and Brian E. Henderson Angeles CountyRisk Factors for Gliomas and Meningiomas in Males in Los

Updated version

http://cancerres.aacrjournals.org/content/49/21/6137

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/49/21/6137To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



risk factors for gliomas and meningiomas in males in los ... · risk factors for gliomas and...

Documents