do pesticides cause lymphoma? - rapal · weed killers insect sprays f a r m i n g lawn care wa t e...

Weed Killers Insect Sprays F a r m i n gLawn Care Wa t e r Ya r d w o r k G a r d e n i n gPest Control Wo r k p l a c e P a r k s F o o dWeed Killers Insect Sprays F a r m i n gLawn Care Wa t e r Ya r d w o r k G a r d e n i n gPest Control Wo r k p l a c e P a r k s F o o dWeed Killers Insect Sprays F a r m i n gLawn Care Wa t e r Ya r d w o r k G a r d e n i n gPest Control Wo r k p l a c e P a r k s F o o dWeed Killers Insect Sprays F a r m i n gLawn Care Wa t e r Ya r d w o r k G a r d e n i n gPest Control Wo r k p l a c e P a r k s F o o dWeed Killers Insect Sprays F a r m i n gLawn Care Wa t e r Ya r d w o r k G a r d e n i n gPest Control Wo r k p l a c e P a r k s F o o dWeed Killers Insect Sprays F a r m i n gLawn Care Wa t e r Ya r d w o r k G a r d e n i n gWeed Killers Insect Sprays F a r m i n gLawn Care Wa t e r Ya r d w o r k G a r d e n i n gPest Control Wo r k p l a c e P a r k s F o o dWeed Killers Insect Sprays F a r m i n gLawn Care Wa t e r Ya r d w o r k G a r d e n i n gPest Control Wo r k p l a c e P a r k s F o o dWeed Killers Insect Sprays F a r m i n gLawn Care Wa t e r Ya r d w o r k G a r d e n i n gPest Control Wo r k p l a c e P a r k s F o o dWeed Killers Insect Sprays F a r m i n gLawn Care Water Ya r d w o r k G a r d e n i n gPest Control Wo r k p l a c e P a r k s F o o d

Lymphoma Foundation of America

Do Pesticides

CauseLymphoma?

RESEARCH REPORT

Do Pesticides Cause Ly m p h o m a ?

Susan Osburn, M.T. (ASCP), M.A., Project Dire c t o r

Lymphoma Foundation of America®

A nonprofit charitable organization helping cancer survivors and families since 1986


Lymphoma Foundation of America®

1100 N. Main Street, Ann Arbor, MI 48104 Tel: (734) 222-1100 Fax: (734) 222-0044

www.lymphomahelp.org

© 2001 Lymphoma Foundation of America®

All rights reserved.

ISBN# 0-9705127-0-8

Do Pesticides Cause Lymphoma?

Michael C.R. Alavanja, Dr. P.H.Senior InvestigatorEpidemiology and Biostatistics

Program National Cancer InstituteDivision of Cancer Epidemiology

and Genetics

Aaron Blair, Ph.D.ChiefOccupational Epidemiology Branch National Cancer InstituteDivision of Cancer Epidemiology

and Genetics

Kenneth Cantor, Ph.D. EpidemiologistOccupational Epidemiology BranchNational Cancer InstituteDivision of Cancer Epidemiology

and Genetics

Brian C. Chiu, Ph.D.Assistant Professor Epidemiology SectionUniversity of Nebraska

Medical CenterDepartment of Preventive

and Societal Medicine

Patricia Duffey, R.N. Research NurseNational Institute on AgingGerontology Research Center

Stanton L. Gerson, M.D. Division ChiefHematology and OncologyCase Western Reserve University Department of Medicine

Elizabeth A. Holly, Ph.D., M.P.H.Professor of EpidemiologyUniversity of CaliforniaSan Francisco School of Medicine

David Ozonoff, M.D., M.P.H.ChairDepartment of Public HealthBoston University School

of Public Health

Paul Strickland, Ph.D. ProfessorJohns Hopkins UniversitySchool of Hygiene and Public HealthDepartment of Environmental

Health Sciences

Bruce Trock, Ph.D.Associate Professor Georgetown University School

of Medicine Lombardi Cancer CenterHuman Oncology Department

Dennis D. Weisenburger, M.D. ProfessorDirector of HematopathologyUniversity of Nebraska

Medical CenterDepartment of Pathology

Shelia Hoar Zahm, Sc.D.Deputy DirectorNational Cancer InstituteDivision of Cancer Epidemiology

and Genetics

Thank you to the many scientists, doctors, lymphoma patients, andhealth professionals who read and reviewed this research report beforeit was published. We appreciate your comments and suggestions.

Scientific Review Panel:

Thank You

www.lymphomahelp.org ■ www.lymphomaresearch.orgTel: (202) 223-6181 ■ Fax: (301) 588-5920 ■ Fax: (703) 527-4056


President

Belita H. Cowan, B.S., M.A.

Vice President

Peter Zetlin

Secretary

Jim Hecker, J.D.

Board of Directors

Ronald J. Jochim

Betsy Randall-David, R.N., Ph.D.

Counselors

Patricia C. Ford, R.N., M.P.H.Kathy E. Scouras, R.N., M.A.

Director of Research and CounselingSusan Osburn, B.S., M.A.

Advisors

Leonard Wisneski, M.D.Frank Sullivan, M.D.James O. Armitage, M.D.George Cannelos, M.D.Dan Longo, M.D.Peter Schroeder, L.C.S.W.Lynn Brallier, M.S.N., Ph.D.

Support Group CoordinatorKatrine Shorb

Regional Representatives

Dick FryKeith Hocter

Vietnam Veterans Liaison

Frank Brown



This question is often asked by people who callour Foundation. They want to know — why did Iget lymphoma? Where does it come from? Whatcan I do to keep it from coming back?

The causes of cancer are complex and not alto-gether known. There are many things to consider:our heredity, viruses, our exposures to toxic sub-stances in the environment (including chemicalsthat can cause cancer) and other factors. Anyonewho could answer this question with certaintywould surely win the Nobel Prize in Medicine.

And yet, there is so much we do know. Over the years, many researchershave done studies to see if pesticides cause lymphoma. Although they donot all agree, many scientists believe this is a serious problem.

We began our Pesticides Research Project last year. Our goal was to lookat the worldwide medical and scientific literature on the subject of lymphomaand pesticides. During that year, many of you told us that you, too, wantedto know more about these studies.

We are pleased to make available to you this Research Report, whichdescribes the most widely regarded studies on lymphoma and pesticides.We encourage you to use our Research Report as a first step in your ownresearch. Because our Pesticides Research Project is ongoing, we welcomeyour comments and suggestions.


Belita H. Cowan

Belita H. CowanPresidentLymphoma Foundation of America


Read This First Page

What “Causes“ Cancer?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Research Methods: Do They Matter?. . . . . . . . . . . . . . . . . . . 3

Who Employs the Researchersand Who Pays For the Studies?. . . . . . . . . . . . . . . . . . . . . . . . 4

Hidden Bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

Are Some Studies Better Than Others?. . . . . . . . . . . . . . . . . . 6

12 Words You Need To Know . . . . . . . . . . . . . . . . . . . . . . . . . 7

The Choice Is Yours

It’s Up To You . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

What About Drinking Water? . . . . . . . . . . . . . . . . . . . . . . . . 13

What About DDT? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Common Ways You Are Exposed To Pesticides In Daily Life . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

How Can I Minimize My ExposureTo Pesticides? 10 Easy Steps To Take. . . . . . . . . . . . . . . . . . 16

The Research Studies

The Studies on Pesticides and Lymphoma . . . . . . . . . . . . . 18

Letters, Comments, and Reports. . . . . . . . . . . . . . . . . . . . . . 41

Dog Studies: Pesticides and Lymphoma . . . . . . . . . . . . . . . 50

Contents

2

3

1

Most of us get our health information from magazines, TV news,and the Internet. We don’t usually read scientific or medical journals.Therefore, we offer you this Research Report in plain English.

It would be so much easier if “somebody”could just “do a big study” and “find theanswer” to the question of what causes lym-phoma. Both fortunately and unfortunately,that isn’t the way science works. One study

doesn’t usually “prove” anything. Instead, scientific knowledge developsby careful testing and retesting of theories through a gradual collection ofevidence.

When a scientific study is finished, the most that can be claimed isthat a statistical association between one or more factors and a certainoutcome has been found — an association, but not necessarily the cause.

Of course, important breakthroughs do occur, and some researchstudies are better designed and produce more meaningful results thanothers. Still, any study that shows an association between factors mustbe reproduced by other scientists working independently, before thetheory which is being tested gains favor in the scientific community.

What “Causes” Cancer?

All of us define “cause” differently, depending on whether we areconsidering “cause” from a medical, legal, religious, or everyday per-spective. The definition of “cause” and the whole idea of causation arerarely mentioned by cancer researchers.

Common sense tells us that the cause of an event or an effect may becomplex or unknown. In reading and evaluating research study results,we need to give some thought to the different definitions of “cause”and to the ways the author’s point of view can affect research methods.

Why does food get moldy after being left out on thekitchen counter for a few days? Ask several people and you’ll get several answers:


1

Read This First

What causes l y m p h o m a ?


2

■ Because mold is everywhere;■ Because a few days is too long to keep food at room temperature ;■ Because Annie didn’t cook it thoroughly;■ Because Jack didn’t clean the countertop;■ Because the weather was hot and humid.

Of course, there may be other theories of what “causes” mold to grow.Any of them might be involved in the spoiled food problem. Obviously,your understanding of the “true cause” will depend to some extent uponyour level of knowledge concerning mold growth, your experience withfood preparation and storage, and whether you are Annie or Jack.

In that same way, your under-standing of what causes cancer maydepend on your knowledge of theways in which cancer develops, aswell as whether you are a cancerpatient, friend of a cancer patient,cancer doctor or researcher — or amanufacturer of pesticides.

One of the most publicized storiesabout the “cause” of cancer is the ongoing drama of the cigarette companiesand their attempts to show, in extended legal battles, that people developlung cancer not from smoking, but from their own inherent problems andweaknesses.

You, the re a d e r, need to understand that when lymphoma is discussedby researchers, there are many things to consider: heredity, viruses, ourexposure to chemicals, and toxic substances in our air, water, and food.These factors and exposures may vary, depending in part on our chosenoccupations and where we live. Any or all of these may weaken a person’simmune system. These are only some of the possible — even pro b a b l e —“causes” of cancer. It appears that pesticides may be one piece of a largerlymphoma puzzle.

Two studies in our Research Report (Faustini, 1996 and Figgs, 1998) showthat exposure to 2,4-D (a commonlyused weed killer) killed large num-bers of lymphocytes in the peoplewho were studied. These people’simmune systems were weakened.

Your ideas about cancermay depend on whethery o u ’ re a cancer patient,d o c t o r, re s e a rcher orchemical company.

These people’s immunesystems were weakened.

Their lymphocyte counts improved only slowly. When the body is forc e dto form new lymphocytes after this kind of toxic injury to the immune system,mutated, abnormal cells may form in some cases, leading to the developmentof lymphoma.

As is typically the case when a question is studied for many years, thereis some disagreement among authors of lymphoma/pesticides studies.Different scientists may look at the same statistics and conclude either“significant association,” “limited evidence,” or “no clear evidence.”

You should be aware that some people are motivated to see “failure toprove cause” unless everyone, or almost everyone, who is exposed to a sub-stance later develops cancer. For example, since cigarette smoking does notcause cancer as reliably as, say, smallpox virus produces smallpox, there arethose who will say that cigarette smoking has little or no role in causingcancer. Jonathan Haar’s book, A Civil Action, shows how some companies(and their scientists and lawyers) are motivated to “disprove” that chemi-cals cause cancer.

Research Methods: Do They Matter?

There are many different ways to do research studies. When we read theresults of a study, we should look carefully at how the study was designed,carried out, and interpreted. Common sense tells us that a report whichmentions only a few cases, occurring in one region within a limited timeperiod, is only suggestive of a pattern, while a study with thousands ofpeople observed over many decades has more power to show meaningfulassociations.

Here’s what we looked for:

✔ Diagnosis of lymphoma must be accurate;✔ Method of assessing exposure to the pesticides must be valid;✔ Exposed populations and control groups must be the same except

for their exposure to the pesticide;✔ The number of people studied must be sufficiently large;✔ Analyses and interpretation of results must be appropriately done.

When scientists study people in a particular occupation, there may beunidentified factors and/or exposures that confound the results.


3


4

Differences like these may affect a study’s outcome:

■ Farmers are exposed to sunlight, dusts, machinery, fuel, fumes, anda n imals, in addition to pesticides.

■ With rare exceptions, studies fail to include any women.

■ Though the latency period for lymphoma may be very long — easily20 years after chemical exposures — some studies proclaim findingsafter only a few years.

■ Mortality studies, which include only lymphoma deaths, ignore lym-phoma survivors who are receiving treatment or who are in extendedremission, and who may have different histories of pesticide exposurethan those who have died.

■ Some re s e a rchers carefully determine a person’s exposure to potentiallyc a n c e r-causing substances or factors. Other re s e a rchers use general cat-egories, such as persons employed in a particular industry or living in acertain location.

Who Employs the Researchers? Who Pays For the Researc h ?

While scientists may strive to remain impartial and unattached, they arehuman. So we need to consider: Who employs the researchers? Who ispaying for the research? Naturally, these two questions are closely related.

This idea — that the results ofs c ientific research are likely to differdepending upon who is funding theresearch — is not just conjecture. Ateam of researchers who analyzed 70articles concerning calcium-channelblocking drugs found a very close

relationship between the authors’ financial ties and their attitudes towardthe drugs. In fact, 100% of those supporting the drugs had financial ties todrug companies, while only 43% of those critical of the drugs had such ties.There was a group who were considered “neutral” (neither for nor againstthe drugs), and of this group, 67% had drug company ties. Finally, it wasd i s c o v e red that in only 2 of the 70 articles did the authors disclose theircorporate/drug company affiliations. (“Conflict of interest in the debateover calcium-channel antagonists,” New England Journal of Medicine, Vol. 338,No. 2, pp. 101–106.)

The results are likely tod i ffer depending on whois funding the re s e a rc h .


Though some scientific and medical journals do show sources ofre s e a rch funding, this practice is not universal. In addition, there may bei n d i rect sources of financial support which are not disclosed in such listings.In our Research Report, we have listed the authors’ stated institutional/o rganizational affiliations and, when shown, funding sources. Since ourinformation comes solely from the journal articles, we have no way ofknowing all the affiliations nor funding sources of the authors, norwhether some have financial ties to the industries whose chemicals theya re studying.

Even in instances where funding is available with few strings attached,t h e re is a scientific “culture” that limits the types of re s e a rch likely to becarried out by an agency or the scientific community as a whole. A l s o ,studies with a “positive” result are more likely to be published than stud-ies that don’t show a correlation between the factors studied, even thougha lack of findings is also an important piece of scientific evidence when astudy is well-designed and well-executed.

Over the last 30 years there has been much more scientific interest incancer chemotherapy, surg e r y, transplantation, immune therapies, andradiation than in alternative approaches, despite the public’s great intere s tin them. The large role of drug companies in the funding of re s e a rch maybe part of the reason why so little money is spent on true cancer pre v e n-tion (as opposed to detection) and the study of alternative tre a t m e n t s .

5

Hidden Bias

In November, 1997, the prestigious New England Journal of Medicinepublished an extremely negative review of a book written by SandraS t e i n g r a b e r, Living Downstream: An Ecologist Looks At Cancer and theE n v i ro n m e n t. In that book, Dr. Steingraber cites evidence that the incre a s ein lymphoma may be related to pesticides e x p o s u re. Why was thereviewer so highly critical of the book? Did he have a hidden bias?

The next month, the New England Journal of Medicine a p o l o g i z e dpublicly for failing to inform readers that the author of that negativebook review was the medical director of W.R. Grace & Co., aMassachusetts corporation that had received extensive negative pub-licity for polluting ground water that may have contributed to higherthan expected rates of leukemia among nearby residents.


6

Are Some Studies Better Than Others?

Some studies really are better than others. Differences in study design areimportant. Larger studies and those involving longer time periods may bemore meaningful. The way a study is designed and carried out is critical.

Among the thousands of studies and articles which are published eachyear, not all bear the same weight or have equal influence in the scientificcommunity nor in the public understanding. Those published in morep re stigious journals are more widely read and respected.

Studies that are more highly publicized, or which catch the attention ofthe press because they address an interesting subject (like obesity or impo-tence), or are released on a day when there is little other news, get morepublicity than others. That doesn’t necessarily mean that they are moreimportant or that we should base our decisions on what we read or hearin the news.

There are also differences among studies done in various countries. Theconditions under which scientists work are not all the same. In some coun-tries, especially in Northern Europe, extensive record keeping is required ofall citizens and businesses, which gives scientists access to fairly accuraterecords showing which people worked where and under what conditions.Certain countries, such as Sweden, maintain national cancer registrieswhich provide information on all cases and all types of cancer. This infor-mation can make it much easier for scientists working in these countries toobtain complete information about the people they are studying.


Pesticide

A substance used to kill or reduce the numbers of unwanted formsof life in a particular environment. These life forms may beunwanted plants (“weeds”), fungi, insects, or animals such as rats.Most pesticides are man-made chemicals, though a few are naturallyderived from plants. Though efforts are made to choose pesticideswhich are toxic only to the intended target and not to humans orother living things, often pesticides have more general toxicity.Pesticides are widely used in schools, factories, offices, homes,public and private lands, and in farming. The use of pesticideshas g reatly increased in recent times. Many pesticides persist inthe environment and can be found in air, water, soil, plants andanimals (including the food we eat), and in our own bodies.

Herbicide

A substance used to kill plants. Some herbicides kill all plantswith which they come into contact; these are often called defo-liants. Others are more specific, affecting only certain types ofplants, such as crabgrass or plants that are not grasses. Herbicidesare used widely in home lawn care, farming, and for weed controlin public areas (parks, streets, golf courses, and highway areas).

Fungicide

A substance used to kill fungi (“funguses”). Yeasts, molds,mildew, and mushrooms are examples of fungi.

Insecticide

A substance used to kill insects. These are very widely used innearly every kind of human environment, including homes, gard e n s ,apartment buildings, schools, office buildings, restaurants, bakeries,and farms. They are also applied directly on adults, children, andpets. The term “insecticide” is sometimes used when the target isnot, strictly speaking, insects, for example, worms.

7

12 Words You Need to Know

1

2

4

3

Carcinogen

A substance that causes cancer. “Carcinogenic” means “causingcancer.”

Chlorophenol

A type of manmade chemical that contains chlorine atoms and aring-shaped carbon stru c t u re. Chlorophenols have some similar-ity to naturally-occurring substances, except for the chlorineatoms. These chemicals are toxic to living organisms, even in tinyconcentrations. Chlorophenols vary in their nature, uses, toxicity,and tendency to cause cancer.

The phenoxy herbicides,f requently mentioned inour Research Report, arec h l o rophenols. So areD D T, PCB, TCDD, andmany other chemicalsnow or formerly used in

farming, landscaping, industrial processes, and home, school,and office pest control. These chemicals are sometimes calledchlorinated (or halogenated — refers to the group of elementsthat includes chlorine) hydrocarbons, or organochlorines.

Even people who don’t use these substances in their work maybe exposed to them through drinking water, food, pesticide spraysused at home, school, work, home or business landscaping, anduse on public lands such as parks and along highways. All peopleliving in industrialized nations are exposed to these chemicals. Wehave residues of them in our body’s own body fat, where suchchemicals tend to persist. (See dioxin/TCDD, below)

Dioxin/TCDD

TCDD is 2,3,7,8-tetrachlorodibenzo-p-dioxin, also just called“ d i o x i n . ” The numbers refer to the locations on the carbon stru c-t u re where the four chlorine atoms are located. The term “dioxin”can also be applied to closely related chemicals.

One of the most toxic chemicals known, TCDD/dioxin occurs as ac o n t a m i n a n t / b y p roduct when certain pesticides are manufacture d .


8

5

6

7

These chemicals are toxicto living organisms, evenin tiny concentrations.

Do Pesticides Cause Cancer?

9

It also gets into the airf rom incineration of chem-ical wastes and from cer-tain industrial pro c e s s e s .Dioxin persists in the envi-ronment, breaks downvery slowly, and tends tobecome concentrated in the fatty tissues of living animals andh u m a n s .

Fish living in dioxin-contaminated waters contain much higher con-centrations of dioxin than the water does.

Americans get 90% of their dioxin exposure by eating dairyp roducts, meat, and fish. Since dioxin is concentrated in fattytissues, it occurs in human breast milk. A c c o rding to Dr. SandraS t e i n g r a b e r, ”a breast-fed infant receives its so-called “safe” life-time limit of dioxin in the first six months of drinking bre a s tmilk.” ( R a ff e n s p e rg e r, C. & Ti c k n e r, J., eds. P rotecting PublicHealth and the Enviro n m e n t, Island Press, Washington, DC).Dioxin is also a hormone disru p t o r, even in concentrations muchsmaller than those that can cause cancer. Other health pro b l e m swhich some experts believe are linked to widespread dioxincontamination include the international epidemic of lowere dsperm counts, the increase in breast cancer, and learning dis-abilities in children.

Case-control study

A study in which “cases” ( people who have a disease) are com-p a red with “controls” (people who don’t have the disease). Thed i ff e rences between the two groups might reveal the cause of thedisease. In our Research Report, “cases” are people with lymphoma;“ c o n t rols” are people with no lymphoma. In designing a case-c o n t rol study, scientists may consider such variables as age, gen-d e r, residence, occupation, race, military duty, etc. Sometimess e veral controls are selected for each case. Often the controls areselected at random from the same general population (hospital,state, etc.) where the cases are found. Good choice of cases andc o n t rols is a very important part of the design of all case-contro ls t u d i e s .

U n f o r t u n a t e l y, we all haveresidues of chloro p h e n o l sin our own body fat.

8


10

Cohort study

A defined group of people (cohort) is followed up over time to observetheir incidence of (or mortality from) a disease. The incidence or mor-tality in the cohort is compared either to the general population or topersons lacking the characteristic or exposure being studied.

Mortality study

A study of deaths from a particular disease (non-fatal cases are left out).A mortality study may have a cohort or case-control design. There is aweakness in the use of mortality studies for lymphoma, because manycases of lymphoma are not fatal. Since lymphoma survivors arei g n o red, the occurrence of lymphoma can appear lower than it really is.

In some populations, expo-sure to carcinogens mayincrease the incidence oflymphoma at a re l a t i v e l yyoung age, resulting in high-er than expected incidencebut few deaths, sinceyounger patients may bettertolerate cancer tre a t m e n t s .

D i ff e rences in access to treatment may affect mortality, and may diff e rf rom risk factors which affect incidence. Also, the cause of death re c o rd-ed on a certificate sometimes does not reflect underlying illness. Forinstance, a death certificate may show the cause as kidney failure, butthe kidney failure may have resulted from cancer chemotherapy. Still,mortality studies do provide useful information.

In the United States, it is often easier to conduct mortality studiesthan incidence studies because death re c o rds are easier to obtainthan information about disease incidence. Norway, Finland, andEngland have complete re c o rds of every cancer case in their coun-tries, making mortality studies less attractive as a study design (andincidence studies easier to carry out). These nations also have verycomplete data on the occupations of their citizens, which re s e a rc h e r scan fairly easily match with cancer data for individuals.

NOTE: We have not attempted to define all types of study designs, but only to explain some of those which occur frequently in our

Since lymphoma survivorsare ignored, the occurrenceof lymphoma can appearlower than it really is.

9

10


11

Research Report. For example, in a meta-analysis, a group ofsimilar studies is evaluated for possible patterns or results notobtainable from the studies separately.

Odds ratio (OR)

Simply put, a measure of the risk of cancer for a study population(people exposed to a pesticide) divided by the risk for a populationlacking the characteristic or exposure being studied (people notexposed to the pesticide). The odds ratio shows whether the gro u punder study has more, less, or about the same chance of getting thedisease as people who are not exposed to the pesticide.

Standard incidence ratio (SIR) and standard mortality ratio (SMR)

These, like the odds ratio, are ways of expressing the re l a t i o n s h i pof incidence (or mortality) from lymphoma in a study populationto that in a re f e rence group, except that in this case, the referencegroup is the whole population (of one country, usually).

Here’s an example using mortality figures: Employees in certainpesticide factories had a SMR of 3.26 for non-Hodgkin’s lymphoma.This means that 3.26 times as many people in that population diedfrom non-Hodgkin’s lymphoma as would have in a sample of thesame size taken from the general population.

Here’s an example using incidence figures: If in the population ofthe U.S., 15 cases of non-Hodgkin’s lymphoma would be expectedper 100,000 person-years, but in a study group of people exposedto a particular pesticide for a total of 100,000 person-years, 30 c a s e so c c u r red, this would result in a SIR for non-Hodgkin’s lymphomain that group of 2.0 (twice the expected incidence).

11

12


The Choice Is Yours

It’s Up To You

Every day you make choices that affect your health. Since there appears tobe a relationship between pesticides and lymphoma, then this risk factor maybe partially under your control. You can choose whether to make an effort toreduce your exposure to pesticides.

When U.S. Surgeon General Luther Terry first announced in 1964 thatsmoking causes lung cancer, his statement was considered controversial.There wasn’t scientific consensus on the issue — many studies showed astrong link between lung cancer and cigarettes, while others did not.During the decades that followed, evidence that cigarette smoke is carcino-genic mounted slowly and is now overwhelming. Today, though tobaccocompanies continue to argue otherwise, most people believe that cigarettesmoking is a major cause of disease and death.

What can we learn from this? Since it may take years to establish withsome certainty whether pesticides cause lymphoma and other cancers, whatare we, as lymphoma patients, friends, and family members, to do?

A sensible approach is to follow “theprecautionary principle,” a concept definedand developed in 1998 by an internationalgroup of scientists and environmentalactivists. This principle, meant to apply to public policy as well as to individualchoices, reads in part:

“When an activity raises threats of harm to human health or the envi-ronment, precautionary measures should be taken even if some cause andeffect relationships are not fully established scientifically.”

Since we see from the studies in this Research Report that there is someevidence that links pesticides to non-Hodgkin’s lymphoma, it makes sensefor us to reduce our exposure to pesticides.

You have an opportunity to make a conscious choice in balancing therisks of pesticides against their possible benefits. How important is it, re a lly,to have a “perfect lawn?” Do the ants that invade your kitchen each springpose any danger, or are they just plain annoying? Is there really a danger ofcontracting Lyme disease in that weedy patch in the back of your property

12

How important is it, re a l l y, to have a“perfect lawn”?


13

(and if so, might it be better to mow the patch than to spray it with pesti-cides)? Would the manager of your apartment building, condo maintenanceservice, office building, or hotel/ motel consider reducing or eliminatingtheir spraying schedules if you mention your concerns?

In the next few pages, we offer you some ideas to help you reduce yourown and your family’s exposure to pesticides.

What About Drinking Water?

You may be drinking a daily dose of pesticides without knowing it, evenif you get your water from a city, town, or suburban water system. A reportin USA Today (October 21, 1998) stated, “Powerful new pollutants imperildrinking water supply. . . f rom heartier bacteria to increasingly toxic industrialpollutants, pesticides, and fertilizers.”

While federal regulations require water system operators to test for morethan 80 substances, the rules are not enforced in many communities, andpesticide residues do appear in tap water. To learn more about the drinkingwater in your locality, you can call the EPA’s Safe Drinking Water Hotline at1 (800) 426-4791.

If you have a well on your pro p e r t y, its water, like that supplied by publicwater systems, may be contaminated by pesticides from nearby farms,lawns and gardens. The U.S. Department of the Interior states in its recentreport, The Quality of Our Nation’s Waters (p. 6): “At least one pesticidewas found…in more than one-half of shallow wells sampled in agriculturaland urban areas. Moreover, individual pesticides seldom occurred alone.”

You might consider having your well water tested by a private laborato-ry, though this is expensive. An alternative is to use tap water for washing,and bottled or filtered water for drinking. Use caution, however — somebrands of bottled water are not pure spring water, but city tap water placedinto bottles. Look for the source of the water — it’s printed on the label.

What About DDT?

D D T, a very toxic pesticide, was banned in the U.S. in 1970. Tw e n t y -five years later in 1995, re s e a rchers were still able to find it in the fatty tissuesand blood of nearly 100% of humans. So, we are all internally exposed tosmall amounts of DDT in addition to everything else we use, consume, orb reathe. We — all people who live in industrialized nations — have dioxinand other chemical residues in our bodies. It’s possible that this chemicalinsult to our immune systems is a partial or contributing cause for thee p idemic of lymphoma that we see today.


14

Common Ways You AreExposed to Pesticides in Daily Life

Home pest controlDo you use products in your home to kill fleas, cockroaches, ants, earwigs,flies, wasps, bees, termites, carpenter ants, silverfish, or spiders?

ExterminatorsHave you hired an exterminator to rid your house of pests? Has yourhome been treated for termites or carpenter ants?

ApartmentsDo you live in an apartment that is sprayed with pesticides by the buildingmanagement?

CondoDo you live in a condominium where the grounds are maintained by alandscaping service that uses weed killers and insect killers?

GardenDo you use insect killers or weed killers in your garden? Do you planttulip or daffodil bulbs with your bare hands? (Most floral bulbs havebeen drenched with fungicides before being sold.)

Lawn careDo you use lawn fertilizers that have weed killers (which are herbicides)to get rid of crabgrass and other weeds? Have you hired a lawn serviceto keep your yard and lawn free of weeds and insects? Do you allowc h i ldren or pets to play in pesticide-treated grass?


NeighborsDo your neighbors spray their shrubs, trees, or lawns? (Sprays allow pesticides to drift beyond property lines.)

WorkplaceIs your workplace treated with pesticides? (Many managers of off i c ebuildings and public buildings hire commercial exterminators to spray atregular intervals.)

Public areasDo you use a park, golf course, playing field, or other public area? (Manyparks and public areas, and virtually all golf courses, are treated with weedkillers and/or insecticides. So are the green spaces near roads and high-ways.)

Drinking waterDo you drink water from a private well or public source that contains pesti-cides? (Don’t assume that your city tap water is free of pesticides).

Hotels and motelsDo you stay in hotels or motels? (Most hotels and motels regularly spraythe rooms with pesticides.

Certain foodsDo you eat fish caught in lakes (such as the Great Lakes) or streams that arecontaminated with pesticides?

10 Easy Steps To Take

To get rid of weeds in your lawn or garden, don’t spray them —pull or dig them up, or leave them alone. Tolerate some weeds inyour lawn — after all, they are green, like grass.

Keep your use of indoor pesticides — such as insect sprays andpest strips — to a minimum, eliminate them altogether, or use lesstoxic pesticides such as pyrethrins and boric acid.

If your workplace is sprayed with pesticides, find out whetheryour office or workspace can be spared this treatment. Often, thereare no insects in office areas, especially if food is not stored oreaten there.

Try to avoid lawn fertilizers or treatments that contain pesticides.If you’re not sure whether they contain pesticides, call the storewhere you bought them, or call the manufacturer (many stores arenot familiar with all the pesticides and chemicals they sell).

If you hire a lawn care company, ask them not to use any pesti-cides. Even if you hire a “natural” lawn care company, check onall substances they plan to apply to your property. If the productkills weeds or insects, it’s a pesticide.

How Can I Minimize My Exposure to Pesticides?


16

1

2

3

4

5


17

If you live in an area where the local water supply is contaminatedby farm runoff (this includes much of the Midwest, parts ofCalifornia, and some other areas), consider drinking filtered orbottled water rather than city tap water or local well water. Becareful: not all filters effectively remove pesticides, and not a l lbottled water is free of chemical contamination. (Ask your dent i s tabout possible fluoride needs, especially for children.)

Buy organically grown fruits and vegetables from a store you cantrust. Look for the organic label. If you can’t buy organic food,wash and/or peel fruits and vegetables. These fruits and vegeta-bles are thought to have the highest pesticide residues: apples,grapes, green beans, peaches, pears, strawberries, spinach, andwinter squash (squashes like butternut, acorn, spaghetti, pumpkin,golden). But continue eating lots of fruits and vegetables!

In animals and humans, pesticide residues concentrate in fat. If you do eat dairy products, use the low-fat kinds. (It is thoughtthat young children may benefit from consuming some fat intheir diet. Check with your pediatrician.)

Reduce your consumption of animal fat, which contains morepesticide residue than the muscle (meat) portions. Remove asmuch fat as possible from meat and poultry before cooking.

Don’t eat fish caught in ponds, lakes, or rivers that are contami-nated with runoff water from nearby farms.

10

9

8

7

6

The Studies of Pesticides and Lymphoma

Alavanja, M.C., Rush, G.A., Stewart, P.,Blair, A. Proportionate mortality study ofworkers in the grain industry. Journal of theNational Cancer Institute 78(2): 247–252 (Feb.,1987).

Grain workers’ mortality rates from particularcauses of death were studied using union insur-ance records. Information concerning pesticideuse was also gathered. The authors found thatpesticides used in the grain industry includealuminum phosphide (phosphine), carbon disul -fide, carbon tetrachloride, ethylene dibromide,ethylene dichloride, Malathion, and methyl bro-mide. In this study of the deaths of 1,114 whitemales, findings indicated increased risk of lym-phoma among grain mill workers (prompting alater study — see below). Mortality rates forseveral other types of cancer were also elevated.National Cancer Institute.

Alavanja, M.C.R.; Blair, A.; & Masters,M.N. Cancer mortality in the U.S. flourindustry. Journal of the National CancerInstitute 82(10): 840–848 (1990).

Researchers looked at the causes of death of22,938 white male flour mill workers. Both acohort mortality study and a case-control studywere done. Results showed increased risk ofnon-Hodgkin’s lymphoma associated with workin flour mills, where pesticides are used fre-quently and heavily. The risk increased with thenumber of years employed in the flour industry;the OR for non-Hodgkin’s lymphoma for whitemales followed for 25 years or more was 9.4 (9.4times as many people from this group devel -oped non-Hodgkin’s lymphoma as from thecontrol groups). The standard mortality ratio(SMR) from non-Hodgkin’s lymphoma was 149in flour mill workers (this means they had a49% extra risk of dying from lymphoma com-

pared to the general population). These find-ings point to pesticide exposure as a risk factorfor non-Hodgkin’s lymphoma. The authors haveconsidered and ruled out other risk factors suchas grain dusts and solvent exposures. NationalCancer Institute.

Asp, S.; Riihimaki, V.; Hernberg, S.; &Pukkala, E. Mortality and cancer morbidityof Finnish chlorophenoxy herbicide applica-tors: an 18-year prospective follow-up.American Journal of Industrial Medicine 26:243–253 (1994).

To follow up on their earlier study (seeRiihimaki, 1982), the authors contacted 1,909pesticide sprayers they had identified and studiedpreviously, or proxies for those who had died.Though mortality figures were used in the initialstudy, in this follow-up total cancer incidencewas studied, so survivors were included also.Only one case of non-Hodgkin’s lymphomaappeared in the entire cohort, though 2.4 caseswould have been expected with a 10-year laten-cy period (elapsed time since exposure to theherbicides). The authors point out that theirstudy was not powerful enough to detect anincrease in incidence of non-Hodgkin’s lym-phoma unless it had increased to 4.5 times thelevel in the general population. Institute ofOccupational Health, Helsinki, Finland.

Baris, D.; Zahm, S.H.; Cantor, K.P.; & Blair,A. Agricultural use of DDT and risk of non-Hodgkin’s lymphoma: pooled analysis ofthree case-control studies in the UnitedStates. Occupational and EnvironmentalMedicine 55: 522–527 (March 24, 1998).

To study the incidence of non-Hodgkin’s lymphoma relative to DDT exposures, the


18


19

authors pooled data from three case-controlstudies of agricultural workers from fourMidwestern states: Kansas, Nebraska, Iowa, andMinnesota. (See Cantor et al., 1992; Hoar et al.,1986; Zahm et al., 1990). Information on pesti-cide use and other risk factors was obtained byinterview; non-farmers were used as a referencecategory. Results seemed to show a relationshipbetween DDT exposure and non-Hodgkin’slymphoma, but when corrections were made forexposure to other pesticides (most agriculturalworkers are exposed to multiple pesticides; thestudy subjects were carefully questioned aboutpesticide use) the DDT effect was not statisticallysignificant. [NOTE: Though the use of DDTwas banned in the U.S. in 1970, it persists in livingthings and in the environment, so mostAmericans have residues of DDT in their tissues.]National Cancer Institute.

Becher, H.; Flesch-Janys, D.; Kauppinen, T.;Kogevinas, M.; Steindorf, K.; Manz, A.; &Wehrendorf, J. Cancer mortality in Germanmale workers exposed to phenoxy herbicidesand dioxins. Cancer Causes and Control 7(3):312–321 (May, 1996).

These authors studied cancer deaths among2,479 workers in four German chemical plantsthat manufactured phenoxy herbicides andchlorophenols. An increase in deaths from non-Hodgkin’s lymphoma was noted (SMR of 326).These data are difficult to interpret because ofsmall numbers (6 lymphoma deaths) and rela -tively short follow-up times (or lengths of timeon the job) at some of the plants; in addition,non-fatal cases of lymphoma were not included.German Cancer Research Center, Heidelberg;Medical Center for Chemical Workers’ Health,Hamburg; Finnish Institute of OccupationalHealth; IARC, Lyon, France. Funding: GermanMinistry of Research and Technology.

Bertazzi, P.A.; Zocchetti, C.; Pesatori, A.C.;Guercilena, S.; Sanarico, M.; & Radice, L.Ten-year mortality study of the population

involved in the Seveso incident in 1976.American Journal of Epidemiology 129(6):1187–1200 (June, 1989).

The impressive figure of 1,674,743 person-yearsof observation was reached in this study. Thee n t i re population in the region of Seveso, Italywas exposed to a variety of hazardous chemicalsafter a chemical plant safety valve pre s s u re diskblew up, releasing a cloud of toxic chemicals,including chlorophenols. This report describescauses of death the re s e a rchers found after theyfollowed up the residents for ten years. Deathrates for females and males diff e red; in the mostheavily exposed region, proportionately morefemales died of all causes than did males. Onlyseven lymphoma deaths occurred, though thesew e re more than were expected. The results sug-gest, but do not prove, an increased risk of lym-phoma. Institute of Occupational Health,Milano; Istituti Clinici di Perfezionamento,Clinica del Lavoro, Milano. Funding:Government of the Region of Lombardy and theItalian National Research Council.

Betta, P.G., Crosignani, P., Russa, A.,Bottero, G., Bernardo, G., Pavesi, M.,Pastonnerlo, M. Triazine herbicides and non-Hodgkin’s lymphomas. A case-control study.Proceedings of the Annual Meeting of theAmerican Society for Clinical Oncology 13:A534 (March 13, 1994).

The authors conducted a study in Italy, usingsmall numbers (19 non-Hodgkin’s lymphomacases, 48 controls in similar rural areas) to deter-mine the relationship, if any, between exposureto triazine herbicides and later development ofnon-Hodgkin’s lymphoma. A significant rela-tionship was found; persons with long-termexposure (over 18 years) had the highest inci-dence of non-Hodgkin’s lymphoma. Theauthors concluded that their findings suggest alink between triazine exposure and non-Hodgkin’s lymphoma. Santo Spirito Hospital,Montferrato, Italy; Division of Epidemiology,INT, Milan; City Hospital, Alessandria; Clinicadel Lavoro Foundation, Pavia.


20

Blair, A.; Cantor, K.P.; & Zahm, S.H. Non-Hodgkin’s lymphoma and agricultural use ofthe insecticide lindane. American Journal ofIndustrial Medicine 33: 82–87 (1998).

As in the Baris, et al. 1998 study (above), thiswas a pooled analysis of 3 studies previouslydone in Midwestern farming areas in Iowa,Kansas, Nebraska, and Minnesota. Resultsshowed that farmers who had used lindane hada 50% higher incidence of non-Hodgkin’s lym-phoma than nonfarmers. Earlier, longer, andheavier use of lindane resulted in higher inci-dences. When the use of other pesticides, espe-cially diazinon and 2,4-D, was considered, theeffect of lindane on lymphoma incidence wasweakened (1.2 or 1.3 times the control level, asopposed to 1.5 times). Lindane is not as widelyused in agriculture today, but is still used onornamental plants, Christmas trees, pecans,a v ocados, livestock sprays, dog shampoos, andshampoos for treating head lice in humans.National Cancer Institute.

Bloemen, L.J., Mandel, J.S., Bond, G.G.,Pollock, A . F., Vitek, R.P., and Cook, R.R. A nupdate of mortality among chemical workerspotentially exposed to the herbicide 2,4-D i c h l o rophenoxyacetic acid and its deriva-tives. Journal of Occupational and Enviro n m e n t a lM e d i c i n e 35(12): 1208–12 (Dec., 1993).

This is an update of a previous study of 878chemical workers exposed to 2,4-D (a herbicide),compared with 36,804 unexposed workers fromthe same chemical factory as well as with thegeneral population. In the previous study,exposures from 1945–1983 were tracked; thisstudy provides an update through 1987. In theoriginal study, 2 deaths from lymphoma werefound with 1 expected; in the follow-up period,there were no deaths from lymphoma. Theauthors conclude that their study of workersexposed to the manufacture of 2,4-D “did notshow patterns suggestive of a causal associationwith any particular cause of death, includingcancer.” However, since only 0.3 deaths from

lymphoma would have been expected duringthe follow-up period, the study lacked sufficientpower (was too small) to detect such patterns.Dow Chemical Company (five of the sixauthors), University of Minnesota School ofPublic Health.

Buckley, J.D.; Meadows, A.T.; Kadin, M.E.;Le Beau, M.M; Siegel, S.; & Robison, L.L.Pesticide exposures in children with non-Hodgkin’s lymphoma. Cancer 2000 89: 2315-2321 (Dec. 1, 2000).

268 California children (ages 12 months – 20years) with non-Hodgkin’s lymphoma werematched with healthy controls. The children’sparents answered a telephone questionnairewhich included topics such as medications, X-rays, allergies, and others in addition to pesti-cide exposures. Parents were asked about theirpesticide exposures at work, the mothers’ expo-s u res during pre g n a n c y, and the use of pesticidesin and around the home during the child’s life-time. Results showed a significantly increasedrisk of non-Hodgkin’s lymphoma in childrenwhose mothers had greater pesticide exposuresduring pregnancy, as well as in children whowere exposed in the home. The more pesticideexposure the children had, the higher their lym-phoma rates. In children whose parents re p o r t e dpesticide use “most days,” the OR was 7.3.When professional exterminations had beendone in the home, the OR was 3.0. The authorsinclude an excellent review of other studies onpesticides and lymphoma in their report. Theynote that although in this study all pesticideswere considered, certain pesticides which aremore capable of inducing lymphoma would belikely to produce a higher odds ratio. Theinvestigators plan to conduct a larger case-contro lstudy in which the pesticides will be specificallyidentified. Children’s Cancer Group. Funding:National Cancer Institute.

Bueno de Mesquita, H.B.; Doornbos, G.; Van der Kuip, D.A.; Kogevinas, M.; &Winkelmann, R. Occupational exposure to


21

p h enoxy herbicides and chlorophenols and can-cer mortality in the Netherlands. AmericanJournal of Industrial Medicine 23(2): 289–300(Feb., 1993).

Mortality re c o rds for employees at two chemi-cal plants in the Netherlands were studied: 963men exposed to phenoxy herbicides and 1,111men who were not exposed. Time periodssince earliest known exposure to the pesticidesw e re fairly long (21 years and 30 years).Neither overall mortality nor cancer mortalitywas increased, but 2 cases of non-Hodgkin’slymphoma were observed in the exposedg roup (none in the non-exposed group), sug-gesting a possible increased risk. There werei n c reased rates of other cancers in certain sub-g roups. National Institute of Public Healthand Environmental Protection, TheNetherlands; International Agency for Researc hon Cancer, Lyons, France.

Cantor, K.P. & Fraumeni, J.F. Jr. Distributionof non-Hodgkin’s lymphoma in the UnitedStates between 1950 and 1975. Cancer Researc h40 (8 Pt. 1): 2645–2652 (Aug., 1980).

Mortality rates for non-Hodgkin’s lymphomafor all counties in the contiguous United Stateswere studied. Trends that were found:increased rates of non-Hodgkin’s lymphomaalong with increased socio-economic status(SES); increased rates of non-Hodgkin’s lym-phoma in California coastal areas; decreasedrates in the Southern United States (thoughtalso to go with lower SES in many counties);increased rates of non-Hodgkin’s lymphoma incounties with food packing industries, especiallysugar beet processing; and a continuousincrease in deaths from non-Hodgkin’s lym-phoma nationwide. The authors noted thatfood processing takes place most often in coun-ties with extensive farming, and also that manyof the death certificates listing non-Hodgkin’slymphoma as the cause of death were for per-sons employed as farmers, not as food canners.This finding helped to focus later research uponfarmers. National Cancer Institute.

Cantor, K.P. Farming and mortality fromnon-Hodgkin’s lymphoma: a case-controlstudy. International Journal of Cancer 29(3):239–247 (1982).

Death certificates of Wisconsin men with non-Hodgkin’s lymphoma listed as the cause of deathduring the years 1968–1976 were matched withmen whose death certificates showed other caus-es. The study examined whether non-Hodgkin’slymphoma occurred more frequently in farmingcommunities. A limitation of the study, noted bythe author, was that the analyses of data and thematching with controls were made according tothe county of residence at the time of death,rather than the men’s actual work histories.Results showed an increased rate of death fro mnon-Hodgkin’s lymphoma among men living incounties with higher rates of pesticide use. Thefarming group had an overall risk of non-Hodgkin’s lymphoma of approximately 1.2 timesthat of the controls. Younger farmers in generalhad higher mortality from non-Hodgkin’s lym-phoma than older farmers (over 65). Thoughe x p o s u re to chicken and animal viruses has beenp roposed as a possible cause of non-Hodgkin’slymphoma, in this study the prevalence of ani-mals or chickens in the various counties showedno correlation with lymphoma rates. NationalInstitutes of Health. Funding (partial): HealthE ffects Laboratory, Environmental Pro t e c t i o nA g e n c y, Cincinnati, Ohio.

Cantor, K.P., Blair, A., Everett, G., Gibson,R., Burmeister, L.F., Brown, L.M., Schuman,L., and Dick, F.R. Pesticides and other agri-cultural risk factors for non-Hodgkin’s lym-phoma among men in Iowa and Minnesota.Cancer Research 52(9): 2447–55 (May 1, 1992).

622 white men (over age 30) with non-Hodgkin’s lymphoma were matched with menwho did not have non-Hodgkin’s lymphoma(twice as many controls as cases) to measure therisks associated with farming and pesticideexposures. Data were obtained by interviewingthe men directly, or family members of thosewho were unable to participate or had died.


22

Detailed farming and pesticide use historieswere taken, including 23 animal insecticides, 34 crop insecticides, 38 herbicides, and 16 fungi-cides. A possible difficulty arises from theinability of the study subjects and/or their familymembers to recall exact pesticide usages, butthis would tend to weaken rather than tostrengthen the statistical association with non-Hodgkin’s lymphoma. Extensive statistical datawere generated concerning particular pesticidesand their non-Hodgkin’s lymphoma risks. “Thechemicals most strongly associated with risk ofnon-Hodgkin’s lymphoma were carbaryl, chlor-dane, DDT, diazinon, dichlorvos, lindane,malathion, nicotine, and toxaphene.” NationalCancer Institute, University of Iowa, Universityof Minn., Orlando Regional Medical Center.

Cantor, K.P., Blair, A., Brown, L.M.,Burmeister, L.F., & Everett, G.Correspondence re: K.P. Cantor et al.,Pesticides and other agricultural risk factorsfor non-Hodgkin’s lymphoma among men inIowa and Minnesota. Cancer Research 53(10):2421 (May 15, 1993).

Following publication of Hoar et al. in JAMA,1986, which found days per year of pesticideuse to be a risk factor for NHL, Cantor et al.contacted the subjects from the previously con-ducted Iowa/Minnesota study to obtain daysper year information. Many more cases thancontrols had died in the interim between theoriginal study and the re-interview, so compar-isons may have been biased. Because journalreviewers requested that these data not beincluded in the Cantor et al. 1992 paper, theauthors published the data in this letter to theeditor. No association with 2,4-D was observedbased on the second interview data. NationalCancer Institute, University of Iowa, OrlandoRegional Medical Center.

Coggon, D.; Pannett, B.; Wi n t e r, P. D . ;Acheson, E.D.; & Bonsall, J. Mortality of work-ers exposed to 2 methyl-4 chloro p h e n o x y a c e t i c

acid. Scandinavian Journal of Work andE n v i ronmental Health 12(5): 448–454 (Oct., 1986).

5,754 men who worked at a plant producing thisparticular phenoxy herbicide were included inthis epidemiological study. Mortality figureswere used primarily, with some cancer registryinformation. No increased risk of non-Hodgkin’s lymphoma was found. The numberof years of exposure, intensity of exposure, andyears of follow-up varied widely, since all menemployed during the period 1947–1975 wereincluded in the study. EnvironmentalEpidemiology Unit of the Medical ResearchCouncil, Southampton, United Kingdom; MRC Environmental Epidemiology Unit,Southampton General Hospital.

Coggon, D.; Pannett, B.; & Winter, P.Mortality and incidence of cancer at four fac-tories making phenoxyherbicides. BritishJournal of Industrial Medicine 48(3): 173–178(Mar., 1991).

Over 2,000 male workers in four chemical plantsw e re studied using mortality re c o rds and cancerregistry information. Twice the expected mortal-ity from non-Hodgkin’s lymphoma was found,but the number of non-Hodgkin’s lymphomadeaths was two. Both occurred over 10 yearsafter first exposure. The study group also had 1non-Hodgkin’s lymphoma survivor. No cases ofHodgkin’s lymphoma were found. These find-ings are not strong enough either to show or tod i s p rove increased risk of non-Hodgkin’s lym-phoma from such exposures. MRCE n v i ronmental Epidemiology Unit, University ofSouthampton, Southampton General Hospital,United Kingdom. Funding: InternationalAgency for Research on Cancer (IARC).

Corrao, G., Calleri, M., Carle, F., Russo, R.,Bosia, S. and Piccioni, P. Cancer risk in acohort of licensed pesticide users.Scandinavian Journal of Work and EnvironmentalHealth 15(3): 203–209 (June, 1989).


23

25,945 male farmers in the southern Piedmontregion of Italy, who had registered for licensesto buy and use pesticides of known toxicity,were the study subjects. The authors used hos-pital records to determine the incidence of vari-ous cancers. They also performed quality con-trol studies to determine the reliability of thecancer diagnoses. The authors found an associ-ation between cancers of the lymphatic systemand exposure to pesticides (SIR of 1.4 in thecohort as a whole; 45 cases found, 31.79 expect-ed). The authors criticize their government’spolicy of requiring licenses for those substancesthat have high acute toxicity but not for thosewhich pose long-term cancer risks. They alsocriticize the IARC working groups for failing toconclude that organic pesticides are carcino-genic. University of L’Aquila, Italy; Universityof Torino, Italy; Department of Worker HealthProtection, Region Piedmont, Asti, Italy.

Dalager, N.A., Kang, H.K., Burt, V.L., andWeatherbee, L. (1991). Non-Hodgkin’s lym-phoma among Vietnam veterans. Journal ofOccupational Medicine 33(7): 774–779 (July,1991).

This analysis of the possible relationshipbetween non-Hodgkin’s lymphoma and militaryservice in Vietnam showed no increased inci-dence in non-Hodgkin’s lymphoma amongVietnam veterans. Service records of non-Hodgkin’s lymphoma patients (men) treated inVA hospitals were compared with the servicerecords of patients from VA hospitals who hadother diagnoses. The authors note that theirsampling methods, as well as changes inVeterans’ Affairs policies concerning hospitaltreatment, might have affected the outcome oftheir study. Department of Veterans Affairs;Centers for Disease Control.

Dalager, N.A.; Kang, H.K.; Burt, V.L.; &Weatherbee, L. Hodgkin’s disease andVietnam service. Annals of Epidemiology 5(5):400–406 (Sept. 1995).

The authors examined service statistics of thoseadmitted to VA hospitals with Hodgkin’s lym-phoma, as well as controls with no cancer.They excluded from the study any veterans(both cases and controls) who were first hospi-talized before 1975. (The Vietnam era is con-s i d e red to be 1965 to March 1973, yet any casesoccurring from 1965–1975 were not included.)The overall OR for Hodgkin’s lymphomaamong Vietnam veterans vs. those servinge l s e w h e re was 1.28, but the authors do notc o nsider this significant. They conclude thattheir data do not show a correlation betweenHodgkin’s lymphoma and Vietnam service, norbetween Hodgkin’s lymphoma and particularlocations or types of service that might haveresulted in more exposure to Agent Orange.E n v i ronmental Epidemiology Service, Dept. ofVeterans A ffairs; Centers for Disease Contro l ,Natl. Center for Health Statistics; and VAMedical Center, Ann A r b o r.

Dich, J., Zahm, S.H., Hanberg, A., & Adami,H-O. Pesticides and cancer. Cancer Causesand Control 8(3): 420–443 (1997).

These authors reviewed studies that cover thewhole range of human and animal research onthe carcinogenicity of pesticides. They concludethat organochlorine and organophosphorouscompounds are associated with an increasedrisk of non-Hodgkin’s lymphoma. The variableresults obtained in studying both non-Hodgkin’sand Hodgkin’s lymphoma and other pesticidesare discussed, as well as possible reasons for thevariations in findings. Karolinska UniversityHospital, Stockholm, Sweden; U.S. NationalCancer Institute; Karolinska Institute,Stockholm; Harvard School of Public Health.

Faustini, A.; Settimi, L.; Pacifici, R.; Fano,V.; Zuccaro, P.; & Forastiere, F.Immunological changes among farmersexposed to phenoxy herbicides: preliminaryobservations. Occupational and EnvironmentalMedicine 53: 583–585 (1996).


24

Ten farmers who mixed and applied chlorophe-noxy herbicides (including 2,4-D and 4-chloro-2-methylphenoxyacetic acid [MCPA]) were thesubjects of this study. The re s e a rchers performedlymphocyte counts on the farmers’ blood, usingspecialized methods to detect specific types oflymphocytes and their activity. This testing wasperformed before the pesticides were mixed andapplied, one to 12 days after, and 50 to 70 daysafter. All ten of the farmers had reducedamounts of every type of lymphocyte in thedays following their pesticide exposures. Thegreatest reductions occurred in lymphocytetypes (CD8-DR, CTL, NK) which are thought tobe important in cell protection and immunity totumors. The ability of NK (natural killer) cellsto carry out their functions was sharplyreduced. By 50 to 70 days following pesticideexposure, all the lymphocyte test values hadreturned to normal limits, but had not reachedthe levels found before exposure. The authorsconclude that exposure to these pesticides maycause immune system suppression and thatmore study is needed to clarify the health impli-cations of this suppression, including cancerrisk. Local Health Unit, Tarquinia (VT), Italy;Istituto Superiore di Sanita’, Rome; Dept. ofOccupational and Environmental Medicine,Linköping, Sweden; Epidemiologic Unit, LazioRegional Health Authority, Rome.

Figgs, L.; Titenko-Holland, N.; Rothman, N.;Zahm, S.; Tarone, R.; Hill, R.; Smith, M.;Holmes, F.; & Blair, A. Occupational expo-sure to the herbicide 2,4-dichlorophenoxy-acetic acid is associated with increased lym-phocyte replicative index. Presented at theannual meeting of American Association forCancer Research, Mar. 28–Apr. 1, 1998.

The researchers studied the lymphocytes of peo-ple who apply herbicides. They found that thelymphocyte replicative index (RI), a measure ofthe rate at which the body is producing lympho-cytes, was significantly higher after spraying2,4-D. (This means that many lymphocytes diedafter exposure to the pesticide, and had to be

replaced rapidly. Rapid growth of lymphatictissue in response to a toxic exposure, especiallyif the genetic content of the cells is damaged,may be related to the development of lym-phoma.) The mean RI in these individuals wasalso significantly higher than that of controls(persons not exposed to the pesticide). St. LouisUniversity; National Cancer Institute; Centerfor Disease Control and Prevention, Atlanta,GA; Univ. of California at Berkeley; KansasUniv. Medical Center.

Fleming, L.E.; Bean, J.A.; Rudolph, M.; &Hamilton, K. Cancer incidence in a cohort oflicensed pesticide applicators in Florida.Journal of Occupational and EnvironmentalMedicine 41(5): 279–288 (April, 1999).

Men and women licensed to apply a wide varietyof pesticides in Florida were studied. No incre a s ein incidence of lymphoma was found, but a sig-nificant increase in genital cancers (testicular andp rostate for men, cervical for women) occurre d .The group studied was moderately large (34,211persons over 18 years, totaling 279,397 person-years), and cancer incidence rather than mortalitywas used, but having a license to apply pesticidesis only an indirect measure of pesticide exposure .University of Miami School of Medicine.Funding (partial): National Institute ofOccupational Safety and Health.

Fontana, A.; Picoco, C.; Masala, G; Prastaro,C.; & Vineis, P. Incidence rates of lymphomasand environmental measurements of phenoxyherbicides: ecological analysis and case-con-trol study. Archives of Environmental Health 53(6): 384–387 (Nov./Dec. 1998).

This is a follow-up of Vineis et al. 1991, in whichincidence of lymphoma relative to exposures tohigh levels of phenoxy herbicides in the soil wasstudied in certain regions of Italy. Besidesupdating data by locating newly-diagnosedcases, the authors conducted a case-controlstudy, identifying occupations of people withlymphoma and controls who did not have cancer.


25

The study of residents of polluted areas againshowed increased incidence of non-Hodgkin’slymphoma compared with areas that did nothave high levels of pesticides in the soil. Thecase-control study showed the highest incidenceof non-Hodgkin’s lymphoma (OR = 4.3) inwomen who worked as rice weeders—an occu-pation which involves very high exposure tosoil and the pesticides which, in that part ofItaly, contaminate it. Local Unit of PublicHealth, Novara, Italy; Istituto di AnatomiaPatologica dell’Universita Verona; IST, Genoa;Dipartimento di Scienze Biomediche eOncologia Umana, Torino. Funding (partial):Europe Against Cancer programme; U.S.National Cancer Institute.

Gambini, G.F.; Mantovani, C.; Pira, E.;Piolatto, P.G.; & Negri, E. Cancer mortalityamong rice growers in Novara Province,northern Italy. American Journal of IndustrialMedicine 31(4): 435–441 (April, 1997).

Mortality records of rice growers in a particularregion of Italy were studied, since there is aknown pattern of herbicide use in weeding ricefields (though data on individual exposures wasnot available). A slight excess of non-Hodgkin’slymphoma mortality was observed, all occur-ring among rice growers aged 65–79 who hadbeen exposed to the herbicides for over 20 years.Numbers were small (4 deaths from lymphomaobserved vs. just over 1 expected). The authorsconclude that further long-term study is needed.Occupational Health Service, OspedaleMaggiore di Norvara, Italy; University of Turin,Italy; and the Institute of PharmacologicalResearch, Milan.

Garry, V.F., Danzl, T.J., Tarone, R., Griffith,J., Cervenka, J., Krueger, L., Whorton, E.B., &Nelson, R.L. Chromosome rearrangements infumigant appliers: possible relationship tonon-Hodgkin’s lymphoma risk. CancerEpidemiology, Biomarkers & Prevention 1(4):287–291 (May/June 1992).

This highly technical study focused on chro m o-some breakage and rearrangements in circ u l a t i n gblood cells of pesticide appliers who wereexposed to phosphine, a toxic gas used to fumi-gate grain storage facilities. Six of the appliersused phosphine almost exclusively in their work;the rest used multiple pesticides. Controls werehealthy subjects not exposed to pesticides in theirwork. Men exposed mainly to phosphine andthose exposed to multiple pesticides had signifi-cantly higher rates of chromosome re a r r a n g e-ment when blood was drawn during periods ofpesticide application. Five of the six phosphineappliers chose to give up their work with thischemical, and the frequency of chro m o s o m a lrearrangements in these five individuals declinedsignificantly after one year. Specific sites ofb reakage and rearrangement suggested to there s e a rchers that this alteration of chro m o s o m e smay play a role in the increased risk of non-Hodgkin’s lymphoma in people with occupation-al exposure to pesticides. Though the number ofpeople studied was small and their occupationale x p o s u re to solvents and dusts may have beenadditional risk factors, the study provides evi-dence of chromosomal damage from pesticidee x p o s u re, possibly consistent in some cases withc h romosomal changes found in non-Hodgkin’slymphoma cells (cancer cells). Univ. of Minn.;NCI; EPA; Univ. of Texas.

Green, L.M. A cohort mortality study offorestry workers exposed to phenoxy acidherbicides. British Journal of IndustrialMedicine 48(4): 234–238 (April, 1991).

Men employed in the forestry industry inOntario, Canada were studied. Instead ofattempting to assess individual exposures, theauthor used employment in the forestry indus-try as an indication of exposure to phenoxyh e rbicides. No excess risk of non-Hodgkin’slymphoma—indeed, no mortality from non-Hodgkin’s lymphoma—was found. The authornotes that, because of the small size of the studygroup (1,222 men), “this study only had suffi -cient power to detect a sevenfold increase in


26

non-Hodgkin’s lymphoma.” The author con-cludes that the risk of non-Hodgkin’s lym-phoma from pesticide exposure is low andrequires a long latent period, necessitating long-term studies with large numbers of people.Ontario Hydro (the electricity provider for theprovince of Ontario, Canada).

Hardell, L.; Eriksson, M.; Lenner, P.; &Lundgren, E. Malignant lymphoma andexposure to chemicals, especially organic solvents, chlorophenols and phenoxy acids: a case-control study. British Journal of Cancer43(2): 169–176 (Feb., 1981).

169 men diagnosed with lymphoma werematched with men from the community, using a2:1 control:subject ratio. Pesticide exposureswere ascertained using a questionnaire and/orinterviewing relatives, as well as assessments ofpesticide use in the pertinent industries andworksites. Results showed higher correlationsbetween pesticide exposures and lymphomathan any previous studies (6-fold risk withexposure to phenoxyacetic acids or chlorophe-nols compared to non-exposure); analysis andreporting of data was very complex and some-what fragmented. The author theorized thatimmunosuppressive and/or mutagenic effectsof the pesticides and solvents might beinvolved. University Hospital, Umea, Sweden.Funding: Swedish Work Environment Fund.

Hardell, L. Relation of soft-tissue sarcoma,malignant lymphoma and colon cancer tophenoxy acids, chlorophenols and otheragents. Scandinavian Journal of Work andEnvironmental Health 7(2): 119–130 (June,1981).

In response to criticism of his earlier study,Hardell re-analyzed some of the data. A studyof colon cancer, using similar methods, wasundertaken to demonstrate that the question-naire and interview methods used were soundand introduced no undue bias. The colon can-cer study showed similar results to previous

colon cancer studies. Hardell concluded thathis original findings (significantly increasednon-Hodgkin’s lymphoma risk with exposureto phenoxy herbicides, chlorophenols, and sol-vents) were upheld. Dept. of Oncology,University Hospital, Umeå, Sweden.

Hardell, L. and Bengtsson, N.O.Epidemiological study of socioeconomicfactors and clinical findings in Hodgkin’sdisease, and reanalysis of previous dataregarding chemical exposure. British Journalof Cancer 48(2): 217–25 (Aug., 1983).

Hardell re-analyzed his data and includedsocioeconomic factors, as other studies haveshown a correlation between Hodgkin’s lym-phoma in children and small family size/highersocio-economic status. He found no correlationbetween Hodgkin’s lymphoma and social class,but did find an association between Hodgkin’slymphoma and exposure to phenoxy acids (suchas herbicides), chlorophenols, and organic sol-vents (relative risks of 1.2 to 6.6, depending onlevel of exposure). The numbers used in thisstudy were small (60 cases of Hodgkin’s lym-phoma, 335 controls). Dept. of Oncology,University Hospital, Umeå, Sweden. Funding:Research Foundation, Dept. of Oncology,University of Umeå, Sweden; Swedish WorkEnvironment Fund.

Hardell, L.; Eriksson, M.; Degerman, A.Exposure to phenoxyacetic acids, chlorophe-nols, or organic solvents in relation tohistopathology, stage, and anatomical local-ization of non-Hodgkin’s lymphoma. CancerResearch 54(9): 2386–2389 (May 1, 1994).

Hardell and colleagues analyzed their 1981 dataonce again. When analyzed by broad occupa-tional groups rather than by closer measures ofpesticide exposure, the data show no significantcorrelation between non-Hodgkin’s lymphomaand any kind of work, just as other studies doneusing occupational classifications to imply pesti-cide exposure have produced equivocal results.


27

An odds ratio of 5.5 was obtained for exposureto phenoxy herbicides, 4.8 for chlorophenolexposure, and 2.4 for solvents. Median latencyperiods ranged from 18 to 21 years. ÖrebroMedical Center, Örebro, Sweden.

Hardell, L. et al. A case-control study ofnon-Hodgkin lymphoma and exposure topesticides. Cancer 85(6): 1353–60 (Mar., 1999).

R e s e a rchers studied 404 white males (both livingand deceased) who had non-Hodgkin’s lym-phoma and 741 controls with no lymphoma,with regard to exposures to various pesticides,solvents, and other materials. Many kinds ofpesticides, alone or in combination, appeared toincrease the risk of non-Hodgkin’s lymphoma;among the highest risks were fungicides (OR =3.7) and the herbicide MCPA (a herbicide widelyused as a weed killer, somewhat chemically sim-ilar to 2,4-D) (OR = 2.7 overall; 1.7 with lowerexposures, 4.1 with higher exposures).Increased risk was also found with exposure toglyphosate, a herbicide used for home weedcontrol in the U.S. (O.R. = 2.3, but based on asmall number of cases). The highest risk of non-Hodgkin’s lymphoma was seen 10 to 20 yearsafter first exposure to pesticides, but lower riskswere seen after 20 to 30 years or more. Theauthors suggest that the carcinogenic effect ofthese chemicals may be due to a combination ofexposure to certain viruses in combination witha weakening of the immune system due to thechemical exposure. Department of Oncology,Örebro Medical Center, Sweden. Funding:Swedish Work Environment Fund, SwedishMedical Research Council, Örebro CountyCouncil Research Committee, Örebro MedicalCenter Research Foundation.

Hoar, S.K.; Blair, A.; Holmes, F.F.; Boysen,C.D.; Robel, R.J.; Hoover, R.; & Fraumeni, J.F.Jr. Agricultural herbicide use and risk oflymphoma and soft-tissue sarcoma. J o u r n a lof the American Medical Association 256 (9),1141–1147 (Sept. 5, 1986).

Cancer reporting is required by Kansas law,making it easier for researchers to find all casesof lymphoma there. The authors identified allnewly-diagnosed cases (males only) ofHodgkin’s and non-Hodgkin’s lymphoma in thestate from 1976–1982, and matched them withthree controls each. All were interviewed care-fully concerning pesticide use as well as othervariables. Results showed that exposure to her-bicides for over 20 days per year (in farm work)was associated with a sixfold increase in inci-dence of non-Hodgkin’s lymphoma. Lesserincreases were seen in farmers who had usedherbicides for fewer days per year. Insecticideswere also studied, and analyses suggested thatherbicides, especially phenoxy herbicides suchas 2,4-D, pose a greater risk for non-Hodgkin’slymphoma than do insecticides. NationalCancer Institute.

Hoar Zahm, S., We i s e n b u rg e r, D.D.,C a n t o r, K.P., Holmes, F. F., & Blair, A. Role ofthe herbicide atrazine in the development ofnon-Hodgkin’s lymphoma. S c a n d i n a v i a nJournal of Work and Environmental Health19(2): 108–114 (1993).

Atrazine is the most commonly used pesticidein the U.S. and is considered a possible carcino-gen. These researchers interviewed 993 whitemen with non-Hodgkin’s lymphoma and 2,918white men who did not have lymphoma inIowa, Kansas, Minnesota, and Nebraska.Extensive statistical analysis, including correc-tion for exposure to other pesticides, appearsnot to show a link between atrazine and non-Hodgkin’s lymphoma. National CancerInstitute.

Hooiveld, M.; Heederik, D.J.J.;Kogevinas,M.; Boffetta, P.; Needham, L. L.;Patterson, D.G.; & Bueno de Mesquita, H.B.Second follow-up of a Dutch cohort occupa-tionally exposed to phenoxy herbicides,chlorophenols, and contaminants. AmericanJournal of Epidemiology 147(9): 891–901 (1998).


28

Workers in a chemical factory (1955–1985) wereclassified according to their chemical exposures.Some had been exposed to dioxin and othertoxic chlorophenols released in an industrialexplosion. The researchers also tested the blooddioxin levels of many of the workers. Resultsshowed an increased incidence of cancer deathsin the group and an increased incidence of non-Hodgkin’s lymphoma, though the number ofcases was small (3 cases out of 140 workers).Two of these three cases of non-Hodgkin’s lym-phoma occurred in workers with high bloodlevels of dioxin. National Institute of PublicHealth and the Environment, The Netherlands;Wageningen Agricultural University;International Agency for Research on Cancer,Lyons, France; Institut Municiapald’Investigació Mèdica, Barcelona; U.S. Centersfor Disease Control. Funding: U.S. NationalInstitute of Environmental Health Sciences.

Johnson, R.A., Mandel, J.S., Gibson, R.W.,Mandel, J.H., Bender, A.P., Gunderson, P.D.,& Renier, C.M. Data on prior pesticide usecollected from self- and proxy respondents.Epidemiology 4(2): 157–164 (March, 1993).

In many studies of cancer and pesticide expo-sures, information is obtained by interviewing aspouse or family member (called a “proxy”) ifan individual under study is dead or incapaci-tated. In some studies, authors note differencesin results from proxies vs. those obtained fromactual patients. There are examples in whichthe risks from pesticides appear higher whenproxies are questioned, and others where therisks appear lower when proxies are questioned.The authors conducted a study in which theresponses of primary respondents were com-pared with the responses of proxies for the verysame individuals. For this study, proxies mightbe spouses, other relatives, or friends and neigh-bors. Differences in response to interview ques-tions were found, especially in the areas of spe-cific pesticides used, time and mode of usage.These differences did affect ORs, though invarying ways. The authors cite the results of a

study in which ORs were higher ( Zahm et alNebraska study) as evidence that the use ofproxies must be studied, while results in whichthe ORs are lower, which the authors also men-tion, do not seem to concern them. MN Dept. ofHealth; Univ. of Minn; Park Nicollet MedicalCenter. Study supported in part by the IndustryTask Force I on 2,4-D Research Data.

Keller-Byrne, J.E.; Khuder, S.A.; Schaub,E.A.; & McAfee, O. A meta–analysis of non-Hodgkin’s lymphoma among farmers in thecentral United States. American Journal ofIndustrial Medicine 31(4): 442–444 (Apr., 1997).

Six studies of the incidence of non-Hodgkin’slymphoma in farmers from Iowa, Minnesota,Missouri, Illinois, and Ohio were selected forthe meta-analysis. All studies had shown ele -vated ORs, but at varying levels (1.09 to 1.80).The OR for the six groups taken together was1.34. When the oldest and largest study wasomitted from the analysis, the OR was 1.39.The authors conclude that microorganisms orpesticides are responsible for an increase in non-Hodgkin’s lymphoma incidence amongAmerican farmers. Medical College of Ohio,Toledo.

Khuder, S.A.; Schaub, E.A.; & Keller- B y r n e ,J.E. Meta-analyses of non-Hodgkin’s lym-phoma and farming. Scandinavian Journal ofWork and Environmental Health 24(4): 255–261( 1 9 9 8 ) .

The authors selected 36 studies of non-Hodgkin’slymphoma and farming, and examined theresults and differences among the studies.Case-control studies showed a higher correla-tion between non-Hodgkin’s lymphoma andfarming than did either mortality/incidencestudies or cohort studies (relative risk of 1.19,1.10, and 0.95 respectively). U.S. studiesshowed a higher risk (relative risk of 1.26) forfarmers than foreign studies (relative risk of1.02). Overall, the meta-analysis showed a


29

small but significant risk (1.10) of non-Hodgkin’s lymphoma for farmers compared tothe general population. Female farmers had alower risk of developing non-Hodgkin’s lym-phoma (relative risk of 0.93). The authors sug-gest that multiple factors may affect farmers,including pesticides and infectious microorgan-isms; these factors may affect men more thanwomen because of traditional divisions of laboron farms, resulting in higher exposures for men.Medical College of Ohio, Toledo.

Kogevinas, M.; Kauppinen, T.;Winkelmann, R.; Becher, H.; Bertazzi, P.A.;Bueno de Mesquita, H.B.; Coggon, D.; Green,L.; Johnson, E.; Littorin, M.; et al. Soft tissuesarcoma and non-Hodgkin's lymphoma inworkers exposed to phenoxy herbicides,chlorophenols, and dioxins: two nested case-control studies. Epidemiology 6(4): 396–402(July, 1995).

Only 32 cases of non-Hodgkin’s lymphoma(found worldwide) were studied. They werematched with multiple controls. The authorsevaluated the exposures of all these individualsto a long list of chemicals by using job records,company questionnaires, and company reports.Among people who had been exposed to phe-noxy herbicides, they found an increase of softtissue sarcomas (ORs of up to 11.27), but resultsfor non-Hodgkin’s lymphoma were less clear(some elevated ORs, but to a much lesserdegree; highest OR was 1.85, and there werenon-linear results with increasing exposures;sometimes lower ORs were found with highere x p o s u res). The authors conclude that theirresults provide only weak evidence that exposedworkers are at increased risk for non-Hodgkin’slymphoma, and note that exposure to multiplechemicals complicates the analyses. IARC, Ly o n ;Intitut Municipal d’ Investigacio Medica,B a rcelona; Institute of Occupational Health,Helsinki; German Cancer Research Center,H e i d e l b e rg; Clinica del Lavoro Luigi Devoto,Milan; National Institute of Public Health andE n v i ronmental Protection, Bilthoven; MRC

E n v i ronmental Epidemiology, Univ. ofSouthampton; Ontario Hydro, To ronto; Tu l a n eUniversity Medical Center, New Orleans; LundU n i v., Sweden; Danish Cancer Registry,Copenhagen; National Institute for OccupationalSafety and Health, Cincinnati; Menzies School ofHealth Research, Australia; Univ. of Vi e n n a ,Austria; Health and Safety Executive, UK;Wellington School of Medicine, New Zealand.Funding: Grant NO1-ES-9527, U.S. NationalInstitute of Environmental Health Sciences.

Kristensen, P.; Andersen, A.; Irgens, L.M.;Bye, A.S.; & Sundheim, L. Cancer in off-spring of parents engaged in agriculturalactivities in Norway: incidence and risk factors in the farm environment. InternationalJournal of Cancer 65(1): 39–50 (Jan. 3, 1996).

Norway has a mandatory cancer case re g i s t r yand a farm census with detailed informationon farming practices. These re c o rds, consid-e red highly accurate, allowed the authors toanalyze the incidence of cancer in children offarmers and agricultural workers with re g a rdto various farm-related exposures and farmingpractices. Many cancers were studied. Theauthors found a non-significant increase forHodgkin’s lymphoma in children of farmersand farm workers from farms with fore s t r yactivities, and an association between themixed cell type of Hodgkin’s lymphoma andchicken farming. For non-Hodgkin’s lym-phoma in these children, incidence levels fol-lowed a dose-response pattern (the more pesti-cides the parents purchased, the more lym-phoma in their children). The authors state,“Our data suggest that parental use of pesti-cides in horticulture is a risk factor for non-Hodgkin’s lymphoma in childhood.” NationalInstitute of Occupational Health, Oslo; CancerRegistry of Norway; Medical Birth Registry ofNorway; Statistics Norway; Norwegian Cro pR e s e a rch Institute.

Kross, B.C.; Burmeister, L.F.; Ogilvie, L.K.;Fuortes, L.J.; & Fu, C.M. Proportionate mor-


30

tality study of golf course superintendents.American Journal of Industrial Medicine29(5): 501–506 (May, 1996).

Golf course superintendents are exposed tomany of the same chemicals and conditions asare farmers: pesticides, dusts, and years ofexposure to sunlight. They were found in thisstudy to have increased mortality from braincancer and non-Hodgkin’s lymphoma. Thep roportionate mortality ratio (PMR) for non-Hodgkin’s lymphoma was 237 (approx. 2.37times the incidence found in the general public).The authors note that their findings are consis -tent with other studies involving populationswith occupational exposures to pesticides.University of Iowa.

Lampi, P.; Hakulinen, T.; Luostarinen, T.;Pukkala, E.; & Teppo, L. Cancer incidencefollowing chlorophenol exposure in a com-munity in southern Finland. Archives ofEnvironmental Health 47(3): 167–175 (May-June 1992).

Järvelä, Finland was found to have a high con-centration of chlorophenols in its drinking waterand in a nearby lake. A local sawmill that useda fungicide on its lumber is thought to be thesource of this chemical pollution. The authorsstudied various cancers. There was a significantincrease in non-Hodgkin’s lymphoma (over sixtimes) among persons who had consumed thelocal drinking water and eaten fish from thepolluted lake. The authors note that the levelsof chlorophenols in the water were not farabove the allowable levels in some countries,such as the United States and Canada. NationalPublic Health Institute, Finland; KarolinskaUniversity Hospital, Stockholm; Finnish CancerRegistry, Helsinki.

Leiss, J.K. and Savitz, D.A. Home pesticideuse and childhood cancer: a case-controlstudy. American Journal of Public Health 85(2):249–252 (Feb., 1995).

Pesticide exposures of 252 children diagnosedwith cancer in the Denver area, plus 222 con-t rols, were evaluated. Results were analyzedseparately for exposure to home extermination,lawn treatments, and indoor pest strips, as wellas for diff e rent types of cancer. The stro n g e s tassociation for lymphoma was exposure toy a rd insecticides (ORs of 1.2 to 1.8 dependingon the age of a child at time of exposure). Thenumber of cases was small, however. Therewas a relatively strong association between theuse of pest strips and leukemias. State Centerfor Health and Environmental Statistics,Raleigh, NC; University of North Caro l i n a ,Chapel Hill.

Lynge, E. A follow-up study of canceri n c idence among workers in manufacture ofphenoxy herbicides in Denmark. BritishJournal of Cancer 52(2): 259–270 (Aug., 1985).

Lynge compiled data on men and womenemployed in two Danish chemical factorieswhich make phenoxy herbicides. All workerswere included, even those who made otherchemicals and those who worked only in officejobs. She found an excess of soft tissue sarco-mas in men (5 cases with only 1.84 expected)and no significant excess of lymphoma (7 caseswith 5.37 expected). The study is rather small(3,390 men, 1,069 women), and there were widevariations in years employed at the chemicalfirms. No attempt was made to determine theamount of actual pesticide exposure. Instituteof Cancer Epidemiology, Danish Cancer Society.Funding: Danish Work EnvironmentFoundation.

Lynge, E. Cancer in phenoxy herbicidemanufacturing workers in Denmark, 1947–87– an update. Cancer Causes and Control 4(3):261–272 (May, 1993).

Lynge added five more years of data to the pre-vious study and focused on a single chemicalfactory. One person’s disease was re-assignedas non-Hodgkin’s lymphoma based on diagnos-


31

tic review. More detailed information on exactwork sites/chemical exposures was obtained.Results were mixed: no increased incidence ofnon-Hodgkin’s lymphoma was found amongworkers possibly exposed to phenoxy herbi-cides, but there was a “puzzling 3.5-fold excessrisk” among workers employed in other manu-facturing departments, including those produc-ing dyes, pigments, and organic solvents.I n c reased risk for certain other tumors, includingsoft tissue sarcomas, was found. Lynge includesa 3-page table showing the results of studies ofexposure to phenoxy herbicides and the risk ofnon-Hodgkin’s lymphoma worldwide. DanishCancer Society.

Lynge, E. Cancer incidence in Danish phenoxy herbicide workers, 1947–1993.Environmental Health Perspectives 106 (Supp.2): 683–688 (April, 1998).

Five years later, Lynge published anotherupdate and concluded that her results showedno excess risk for non-Hodgkin’s lymphomaamong the workers, although they showed anincreased risk of soft-tissue sarcomas. Asbefore, the workers were considered “potential-ly exposed” based upon their assigned workareas in the factories. Actual exposures to thepesticides and other toxic chemicals were notdetermined. Danish Cancer Society.

McDuffie, H.H., Dosman, J.A., McLaughlin,J., Theriault, G., Pahwa, P., Choi, N.W.,Fincham, S., Robson, D., Spinelli, J.,Skinnider, L.F., & White, D. Non-Hodgkin’slymphoma (NHL) and pesticide exposure:Canada. Proceedings of the AmericanAssociation for Cancer Research 35: A1721(March, 1994). (Abstract only)

These researchers conducted a case-controlstudy in Canada, obtaining information aboutpesticide exposures from 352 men with non-Hodgkin’s lymphoma and 704 men who did nothave lymphoma. A significantly higher inci-

dence of non-Hodgkin’s lymphoma was foundin men with occupational exposures to any com-bination of herbicides, insecticides, fumigants,or fungicides. The risk of non-Hodgkin’s lym-phoma increased with increasing number ofexposure hours per year. University ofSaskatchewan, Canada.

Morrison, H.I., S e m e n c i w, R.M., Wilkins, K.,Yang Mao, & Wigle, D.T. Non-Hodgkin's lymphoma and agricultural practices in theprairie provinces of Canada. ScandinavianJournal of Work and Environmental Health 20(1):42–47 (1994).

This is a follow-up to Wigle, et al. (see below).Records from the Census of Agriculture(Canada) were cross-referenced by computerwith records from the Census of Population tolook for any relationship between farm size,h e rbicide use, and non-Hodgkin’s lymphoma.Results varied according to farm size, province,farming practices, and fuel and oil expenditures.The authors conclude that their findings showan association between herbicides and the riskof fatal non-Hodgkin’s lymphoma. LaboratoryCentre for Disease Control (Canada).

Mulder, Y.M., Drijver, M., & Kreis, I.A.Case-control study on the associationbetween a cluster of childhood haemopoieticmalignancies and local environmental factorsin Allsmeer, The Netherlands. Journal ofEpidemiology and Community Health 48(2):161–165 (April, 1995).

The authors did a case-control study of a clusterof leukemia and lymphoma cases in a horticul-tural community. The number of subjects wassmall: 7 lymphoma cases and 7 leukemia cases.Age at diagnosis with lymphoma varied from 12to 39. A questionnaire was used to identify riskfactors. Several associations emerged; mostnotable was a history of swimming in a pondknown to be polluted with DDT, polycyclic aromatic hydrocarbons, and benzene chemicals


32

associated with farming activities and with anearby airport. There was a strong associationbetween these cancers and the fathers’ exposureto pesticides. Community Health ServiceAmstelland De Meerlanden; Community HealthService Zuid Kennemerland; National Institutefor Public Health and Environmental Protection,The Netherlands.

Nanni, O., Amadori, D., Lugaresi, C.,Falcini, F., Scarpi, E., Saragoni, A., andBuiatti, E. Chronic lymphocytic leukaemiasand non-Hodgkin’s lymphomas by histologi-cal type in farming-animal breeding workers:a population case-control study based on an apriori exposure matrices. Occupational andEnvironmental Medicine 53(10): 652–657 (Oct.,1996.)

In an earlier study, the authors noted a pro b a b l ec o r relation between pesticide exposures andincidence of chronic lymphoctyic leukemia(CLL) and non-Hodgkin’s lymphoma. In thep resent study, they analyzed numerous vari-ables for 187 cases and nearly a thousand con-t rols in an agricultural region of Italy. Themethods were complex, involving ten diff e re n tq u e s t i o n n a i res depending upon whichc rops/animals were raised, plus estimates oftypes and amounts of exposures. The re s u l t sw e re also complex. Increased ORs for CLL a n dnon-Hodgkin’s lymphoma were found withe x p o s u re to stannates (tin-containing com-pounds; OR of 2.44), arsenates (OR = 1.83), andto a lesser degree, for most other pesticides.When occupational titles alone were studied(for example, farmer vs. non-farmer), no corre-lations were found. Oncological Institute ofRomagna, Italy; Medical Oncology Unit, Forli;Phytopathology Institute, Cesena; Departmentof Pathology, Forli; Epidemiology Unit, Centerfor Study and Prevention of Cancer, Fire n z e .Funding: Province of Forli; Emilia-RomagnaRegion; Italian League Against Cancer; ItalianAssociation for Cancer Research; Ministry ofEmployment, Health, and Social Services,R o m e .

O’Brien, T.R., Decouflé, P., & Boyle, C.A.Non-Hodgkin’s lymphoma in a cohort ofVietnam veterans. American Journal of PublicHealth 81(6): 758–760 (June, 1991).

By examining the hospital records of a largecohort (18,313) of male Vietnam-era veteranswho had died after discharge, the authors deter-mined that Vietnam veterans had an increasedrisk of non-Hodgkin’s lymphoma (7 fatal casesin Vietnam veterans vs. only 1 case in non-Vietnam veterans; similar numbers studied).None of the non-Hodgkin’s lymphoma casesoccurred in veterans who had served in areasthought to have Agent Orange exposure. Thecause of the increased risk was unclear. Non-fatal cases of non-Hodgkin’s lymphoma werenot considered; neither were differences insocio-economic status, education, race, orwartime living conditions between Vietnam vet-erans and non-Vietnam veterans. U.S. Centersfor Disease Control.

Olsson, H. and Brandt, L. Non-Hodgkin’slymphoma of the skin and occupationalexposure to herbicides [letter]. Lancet 2(8246):579 (Sep. 12, 1981).

P rompted by reports of a relationship betweenpesticides and cutaneous (skin) lesions as the siteof non-Hodgkin’s lymphoma, the authors re p o r ton their own lymphoma patients, from whomthey obtained occupational histories. Of 5 malepatients with skin lesions as the only detectablemanifestation of non-Hodgkin’s lymphoma, 4(80%) had been actively spraying pesticides for18 to 20 years. Of 118 male patients whose lym-phomas had begun at other sites, only 7 (6%)had similar exposures. Though the numbers aresmall, this finding suggests a possible re l a t i o n-ship between pesticide exposures and cutaneousnon-Hodgkin’s lymphoma. Dept. of Oncology,University Hospital, Lund, Sweden.

Pearce, N.E., Smith, A.H., Howard, J.K.,Sheppard, R.A., Giles, H.J., & Teague, C.A.


33

Non-Hodgkin’s lymphoma and exposure tophenoxyherbicides, chlorophenols, fencingwork, and meat works employment: a case-control study. British Journal of IndustrialMedicine 43(2): 75–83 (Feb., 1986).

Using the New Zealand Cancer Registry, theauthors compared occupational exposures of 83men who had non-Hodgkin’s lymphoma withmen having other cancers as well as withhealthy men. A wide range of occupationaland exposure variables were included, and theresults were complex. Exposure to phenoxyherbicides did not result in significantly elevat-ed odds ratios. Results in relation toc h l o rophenol exposure were unclear, sincemany of the men had been exposed to otherchemicals as well. There was an increase innon-Hodgkin’s lymphoma in men who hadworked both in fencing (building and main-taining fences in rural areas) and in a meat pro-cessing plant (they had exposure both toarsenic compounds used in the treatment offence posts, and exposure to various chemicals,animal viruses, etc. while processing meat andpelts). The authors noted possible synerg i s t i ce ffects of various exposures, and the diff i c u l t yin assessing multiple exposures. We l l i n g t o nClinical School of Medicine, We l l i n g t o nHospital, New Zealand; School of PublicHealth, Univ. of North Carolina, Chapel Hill;School of Public Health, Univ. of California,B e r k e l e y. Funding: New Zealand Health Dept.,War Pensions Medical Research Trust Board ,Medical Research Council, Northern CaliforniaOccupational Health Center.

Pearce, N.E., Sheppard, R.A., Smith, A.H.,and Teague, C.A. Non-Hodgkin’s lymphomaand farming: an expanded case-controlstudy. International Journal of Cancer 39(2):155–61 (1987).

In this follow-up study, the authors nowi n c l u ded lymphosarcoma and reticulosarcomapatients, previously excluded, to facilitate com-parisons with studies done by Hardell, 1981 and

Hoar et al., 1986. No increased risk from expo-s u re to phenoxy herbicides was found, but twog roups – persons who had been engaged inboth fencing work and meat works employ-ment (OR 3.8), and orc h a rd workers, (OR 3.7)showed rather sharply increased risks. Theauthors were puzzled as to the reasons whytheir results showed no increased risk fro mphenoxy herbicides, in contrast to earlier stud-ies (such as Hardell, et al.). They felt that thehigh rates found in the fencing/meat workersand orc h a rd workers might be a result of syn-e rgistic effects of co-carcinogens, includingarsenic compounds which are used as pesti-cides in orc h a rd work and in fencing. Univ. ofNorth Carolina; Wellington Clinical School ofMedicine, New Zealand; Univ. of Cal.,Berkeley; Princess Alexandra Hospital,A u s t r a l i a .

Pearce, N. Phenoxy herbicides and non-Hodgkins lymphoma in New Zealand: frequency and duration of herbicide use.British Journal of Industrial Medicine 46:143–144 (1989).

In this short update, Pearce added more subjects(all male, as before), and focused specifically onphenoxy herbicides. He found little increasedrisk of non-Hodgkin’s lymphoma withincreased numbers of years of exposure. A dose-response relationship between non-Hodgkin’s lymphoma and numbers of days peryear of using phenoxy herbicides was observed,but this finding was not consistent at the high-est use level (over 20 days of use per year).Pearce’s findings differed from those in theSwedish study (Hardell et al.), in which a strongcorrelation was found. He hypothesizes thatdifferences in the spraying schedules mayaccount for these differences. In Sweden herbi-cide use is concentrated in a 2 or 3 month period,but in New Zealand spraying is more intermit-tent. [An additional factor not mentioned is thatin New Zealand, the workers were usuallyexposed to 2,4,5-T rather than to 2,4-D.]Wellington School of Medicine, New Zealand.


34

Persson, B., Dahlander, A., Fredriksson, M.,Brage, H.N., Ohlson, C-G., & Axelson, O.Malignant lymphomas and occupationalexposures. British Journal of IndustrialMedicine 46: 516–520 (1989).

The authors found increased risks for bothHodgkin’s lymphoma and non-Hodgkin’s lym-phoma in relation to exposure to phenoxy herbi -cides (a type of weed killer). However, thenumber of subjects was quite small; of all per-sons with lymphoma and controls (persons withno lymphoma) studied, only nine in total hadbeen exposed to phenoxy herbicides. Increasedrisks were also found with welding work, expo-sure to solvents, exposure to wood (creosotesuspected as a cause), and hairdressing as acareer (hairdressers are exposed to manmadechemicals, including hair dyes). UniversityHospital and Örebro Medical Centre Hospital,Sweden.

Persson, B., F redriksson, M., Olsen, K.,Boeryd, B., & Axelson, O. Some occupation-al exposures as risk factors for malignantlymphomas. C a n c e r 72(5): 1773–8 (Sept. 1,1993).

124 cases (a relatively small number) includingboth Hodgkin’s lymphoma and non-Hodgkin’slymphoma, men only, were matched with menwho did not have cancer. The number of occu-pational and other risk factors studied was larg e .The highest non-Hodgkin’s lymphoma risksw e re found among lumberjacks (crude OR 7.0),workers exposed to raw wood (various occupa-tions including sawmill workers, lumberjacks,and paper-pulp workers; OR 2.9) and personsexposed to phenoxy herbicides (crude OR 2.6).The authors note that their findings concerningelevated risks of non-Hodgkin’s lymphoma inassociation with exposures to raw, untre a t e dwood raise the question of whether naturally-occurring fungicides which are present in woodmay be carcinogenic. Univ. Hosp., Linköping,Sweden. Funding: Local Cancer Fund, Countyof Östergötland; Swedish Cancer Society.

Riihimaki, V., Asp. S, & Hernberg, S.Mortality of 2,4-dichlorophenoxyacetic acidand 2,4,5-trichlorophenoxyacetic acid herbi-cide applicators in Finland. First report ofan ongoing prospective cohort study.Scandinavian Journal of Work andE n v i ronmental Health 8(1): 37–42 (1982).

Using personnel records, the authors located1,926 pesticide sprayers who had probably beenexposed to 2,4-D and 2,4, 5-T. Causes of deathwere determined using the Finnish CentralStatistical Office. Results showed no cases ofnon-Hodgkin’s lymphoma in the group. Theauthors point out that their study cannot beused to show that phenoxy herbicides are or arenot implicated in non-Hodgkin’s lymphoma,because the number of people studied is toosmall, the exposures were brief and low, and thefollow-up period too short. Institute ofOccupational Health, Helsinki, Finland.

Ritter, L., Wigle, D.T., Semenciw, R.M.,Wilkins, K., Riedel, D., & Mao, Y. Mortalitystudy of Canadian male farm operators:c a ncer mortality and agricultural practices inSaskatchewan. La Medicina del Lavoro 81(6):499–505 (1990).

73,538 male farm operators (the primary person who owns and operates a farm) inSaskatchewan, Canada were studied using census data and questionnaires. Mortality andcause-of-death data were obtained from theCanadian Mortality Data Base (non-fatal cases oflymphoma were not included). Although farmoperators overall showed no increased risk ofdeath from non-Hodgkin’s lymphoma, certains u b g roups had significant increases. Riski n c reased with the number of acres sprayed withpesticide in 1970 (a year for which data wereavailable), and also with increasing annual fuele x p e n d i t u re (up to a relative risk of 2.29 forfarmers spending over $900 for fuel in 1970). A nexception occurred with farms of over 1,000a c res sprayed with pesticide—risks were lowerfor this group than for farm operators with less


35

a c reage. The authors speculate that when farmsa re quite large, spraying may be done by hire dworkers rather than by the farm operator. Fuele x p e n d i t u res are possibly indicative of (1)i n c reased exposure to pesticides due to extensivespraying using tractors and (2) exposure todiesel fumes, which could be significant alone orin conjunction with pesticide exposure s .E n v i ronmental Health Directorate, HealthP rotection Branch, Laboratory Centre for DiseaseC o n t rol, Health and We l f a re Canada, Ottawa.

Rothman, N., Cantor, K.P., Blair, A., Bush,D., Brock, J.W., Heizlsouer, K., Zaham, S.H.,Needham, L.L., Pearson, G.R., Hoover, R.N.,Comstock, G.W., & Strickland, P.T. A nestedcase-control study of non-Hodgkin lym-phoma and serum organochlorine residues.Lancet 350: 240–244 (1997).

Using very sophisticated methods, these scien-tists compared the amounts of DDT and PCBs inthe blood of persons who developed non-Hodgkin’s lymphoma with the blood of healthycontrols. An important factor in this studydesign is that the blood was obtained andfrozen in 1974, before rather than after the dis-ease had developed. Although increased inci-dence of non-Hodgkin’s lymphoma withincreased serum DDT was found, the resultswere not considered statistically significant. Astronger, significant relationship was detectedbetween serum PCB levels and later develop-ment of non-Hodgkin’s lymphoma. (Note:PCBs are mainly associated with electricalequipment rather than pesticides, but are simi-lar in chemical nature to some organochlorinepesticides, and were sometimes mixed with pes-ticides as adhering or extending agents.) Therewas also a relationship between Epstein-Barrvirus (in combination with PCBs) and non-Hodgkin’s lymphoma. National CancerInstitute; Johns Hopkins University School ofHygiene and Public Health; National Center forEnviron- mental Health, Centers for DiseaseControl and Prevention; and GeorgetownUniversity Medical School. Funding: Dept. of

Health and Human Services grants CA60754and ES03819; Research Career Award HL21670.

Saracci, R., Kogevinas, M., Bertazzi, P.A., de Mesquita, B.H.B., Coggon, D., Green, L.M.,Kauppinen, T., L’Abbe, K.A., Littorin, M.,Lynge, E., Mathews, J.D., Neuberger, M.,Osman, J., Pearce, N., & Winkelmann, R.Cancer mortality in workers exposed tochlorophenoxy herbicides and chlorophenols.Lancet 338(8774): 1027–32 (October 26, 1991).

In this very large study (17,372 men, 1,537women, and one person of unknown gender),20 groups of workers employed in pro d u c t i o nor spraying of chlorophenoxy herbicides fro m10 countries were included. Mortality wasstudied rather than incidence, so non-fatalcases of lymphoma were not included. Nos i gnificant increase in death from lymphomawas found. The authors note that interpre t a-tion of the findings will be easier when theg roups of workers have been followed for alonger time and more deaths have occurre d .The authors are involved in epidemiologicalre s e a rch and are employed by medical schoolsor government agencies. The primary author(Saracci) is associated with the Unit ofAnalytical Epidemiology, International A g e n c yfor Research on Cancer, Lyon, France.Funding: U.S. National Institute ofE n v i ronmental Health Sciences.

Sathiakumar, N., Delzell, E., & Cole, P.Mortality among workers at two triazineh e rbicide manufacturing plants. AmericanJournal of Industrial Medicine 29(2): 143–151(Feb., 1996).

A slight increase in mortality from non-Hodgkin’s lymphoma was found among workersdirectly exposed to triazines (triazine herbicidesare widely used on corn crops). There weremore deaths (3) among workers whose expo-sures were “definite or probable” than amongthose who had “possible” exposure (2 deaths).


36

The authors note two factors which might haveweakened the study’s ability to detect anincrease in lymphoma: insufficient follow-uptime and the omission of non-fatal cases. Dept.of Epidemiology, School of Public Health,University of Alabama at Birmingham.

Sathiakumar, N., & Delzell, E. A review ofepidemiologic studies of triazine herbicidesand cancer. Critical Reviews in Toxicology27(6): 599–613 (1997).

The authors reviewed eleven studies on the rela-tionship between triazine herbicides and cancer.They note that none of the studies includedassessments of actual herbicide exposures;instead, the potential for exposure was assumedbased on later questioning or on employmentinformation. Most of the people who had usedtriazine herbicides had also been exposed toother pesticides. Although all the studies onnon-Hodgkin’s lymphoma and triazine herbi-cides showed excess incidence of the disease,and some showed increased incidence withincreased years of exposure, these authors con-clude that the associations are weak. No studiesshowed a relationship between triazines andHodgkin’s lymphoma. Dept. of Epidemiology,School of Public Health, University of Alabama,Birmingham.

Scherr, P.A.; Hutchison, G.B.; & Neiman,R.S. Non-Hodgkin’s lymphoma and occupa-tional exposure. Cancer Research 52(19 Suppl):5503s–5509s (Oct. 1, 1992).

303 cases of non-Hodgkin’s lymphoma (includ-ing patients, proxies, and parents) were inter-viewed, along with 303 controls, concerningtheir occupations, possible risk factors, and awide range of general questions. Data were ana-lyzed with re g a rd to types of lymphomas. Oneunusual finding was a relative risk of 0.4 forP rotestants as compared to all other re l i g i o n sand a relative risk of 3.1 for Jews (the risk in thetotal population would be expressed as 1.0, soP rotestants only had 40% of the expected inci-

dence of lymphoma, whereas Jews had overt h ree times the expected incidence). [Ed. note:No theory or explanation is given for this find-ing. The authors did not report results forRoman Catholics or other religions.] For expo-s u re to pesticides and insecticides, the authorsfound an increased risk for all lymphomas and a large risk for diffuse, large cell lymphoma inp a r t i c u l a r. There was a relative risk of 8.0 ford i ffuse large cell lymphoma for those havingever been exposed in their work to particles suchas dust, sawdust, or fibers. This risk rose to 8.2when adjusted for religion. Relative risk of 11 . 0was found when all types of non-Hodgkin’slymphoma were correlated with a history ofwork as a carpenter or plumber; this risk rose to13.2 when adjusted for religion. Mathematiciansand programmers had a sharply decreased riskof lymphoma. The authors conclude that they“have found an increased risk of lymphomaamong persons exposed to insecticides, pesti-cides, and particles such as dust” as well as peo-ple employed in the agriculture, fore s t r y, fishing,c o n s t ruction, paper, wood, leather, and variousbuilding-material trades, such as carpenters,stone masons, etc. They note that in each ofthese occupations, workers had the possibility of exposure to chlorophenols or phenoxyacids(pesticides also associated with lymphoma inmany other studies). Harvard School of PublicHealth; Boston University School of Medicine.

Smith, J.G. and Christophers, A.J. Phenoxyherbicides and chlorophenols: a case controlstudy on soft tissue sarcoma and malignantlymphoma. British Journal of Cancer 65(3):442–448 (March, 1992).

52 Australian men with lymphoma werematched with non-cancer and cancer controls,and interviews were conducted to determinepesticide exposures. The study took more thanten years to complete because of a poorresponse rate from prospective participants.Although results show increased risk of non-Hodgkin’s lymphoma among those exposed topesticides, with a dose-response effect, theauthors conclude that with such a small sample,


37

their results do not support the hypothesis thatexposure to chlorinated phenoxy herbicides orchlorophenols causes lymphoma. They found acorrelation between smoking and lymphoma(though not between alcohol and lymphoma),but did not consider this to be significant either.Peter MacCallum Cancer Institute, Melbourne;University of Melbourne. Funding: HealthDept. of the State of Victoria, Australia.

Torchio, P.; Lepore, A.R.; Corrao, G.; Comba,P.; Settimi, L.; Belli, S.; Magnani, C.; & diOrio, F. Mortality study on a cohort of Italianlicensed pesticide users. Science of the TotalEnvironment 149: 183–191 (1994).

Using re c o rds of people who obtained licenses toapply pesticides and correlating these with deathrecords, the authors studied death rates fromvarious causes for a large number of people in aparticular region of Italy. Overall, death ratesw e re unusually low, due to what is called the“healthy worker effect.” A slight increase in mor-tality from both Hodgkin’s and non-Hodgkin’slymphoma was found in pesticide appliers (typi-cally farmers) whose homes were in farmlanda reas, as opposed to woodland or village are a s .The authors note that the mere possession of alicense to apply pesticides is not indicative ofe x p o s u re. They state that had they been able tolimit the study to those who actually applied pes-ticides, the results might have showed a stro n g e rassociation. University of L’Aquila, Italy; IstitutoS u p e r i o re di Sanità, Roma; University Service ofCancer Epidemiology, Torino. Funding:Consiglio Nazionale delle Richerche [Italy] and the Piedmont Region [Italy].

Viel, J.F. and Richardson, S.T. Lymphoma,multiple myeloma and leukemia amongFrench farmers in relation to pesticide expo-sure. Social Science and Medicine 37(6):771–777 (1993).

By correlating mortality records for Frenchfarmers with amounts of pesticides used in 89

geographical areas (“départements”), theauthors examined the effects of pesticide expo-sure on incidence of non-Hodgkin’s lymphoma,leukemia, and multiple myeloma. Data wereinterpreted using complex statistical analyses.Results showed increased mortality withincreased pesticide use for leukemia and formyeloma, but not for non-Hodgkin’s lymphoma.The paper includes discussion of similar, earlierresearch worldwide. Public Health Department,Villejuif, France.

Vineis, P., Faggiano, F., Tedeschi, M., &Ciccone, G. Incidence of lymphomas andsoft-tissue sarcomas and enviro n m e n t a lm e asurements of phenoxy herbicides. Journalof the National Cancer Institute 83(5): 362–363(March 6, 1991).

This study is unusual in that the authors hadinformation concerning the levels of water andsoil contamination with 2,4-D and 2,4,5-T (herbi-cides widely used in agriculture) in variousregions of the Italian rice-growing provinceswhere their subjects live. A limited number ofhospitals serves the area, so all cases of non-Hodgkin’s lymphoma, Hodgkin’s lymphoma,and soft-tissue sarcoma could be identified.Quite significant increases in non-Hodgkin’slymphoma incidence were found: (age-adjustedincidence rates of 5.7 to 18.2, compared to therate of non-Hodgkin’s lymphoma in the worldpopulation), with the highest rates of non-Hodgkin’s lymphoma found in the “category B”area, which had the highest levels of pesticidesin water and soil. Both males and females wereincluded; males had a higher incidence of non-Hodgkin’s lymphoma. The authors concludethat the excess of non-Hodgkin’s lymphoma inthis group is a result of the workers’ exposuresto the contaminated soil and water (in manynations, rice workers often walk bareleggedthrough the wet fields). The authors plan to doa case-control study also (see Fontana et al.).Università di Torino, Italy. Funding: EuropeanEconomic Community Europe Against CancerCampaign, Associazione Italiana per le


38

Richerche sul Cancro, and Lega Italiana per laLotta Contro i Tumori, Sezione Provinciale diAlessandria.

Weininger, R.B., Davis, G., & Hawks, C.D.Herbicides and cancer. Journal of the AmericanMedical Association 257(17): 2292 (May 1, 1987).

This is a brief report on two neighboring ruralcounties in upstate New York. Using the localhospital tumor registry, the authors found thatnon-Hodgkin’s lymphoma incidence was 15.3new cases per 100,000 people per year, about 11/2

times that for the general U.S. population at thattime. 2,4-dichloro-phenoxyacetic acid herbi-cides are widely used in their communities.The impetus for this study was Hoar et al. 1986.Columbia Memorial Hospital, Hudson, NY.

Weisenburger, D.D. Environmental epi-demiology of non-Hodgkin’s lymphoma ineastern Nebraska. American Journal ofIndustrial Medicine 18(3): 303–305 (1990).

The author, having noted an increase in non-Hodgkin’s lymphoma (especially aggressive,diffuse large-cell subtype) in eastern Nebraska,conducted a study, using telephone interviews,of 201 men with non-Hodgkin’s lymphoma and725 controls. Results showed a 50% increasedrisk of non-Hodgkin’s lymphoma with use of2,4-D herbicide, with a threefold risk for thosewho were exposed to 2,4-D for 20 or more daysper year. Increased risk of non-Hodgkin’s lym-phoma was also found for those exposed toorganophosphates, carbamates, atrazine, orchlorinated hydrocarbons (all of these are pesti-cides). The paper also includes a discussion ofcontamination of the area’s groundwater withpesticides and fertilizers. University ofNebraska Medical Center.

White, D., McDuffie, H.H., Pahwa, P.,Gomez, S., Parker, R., & Dosman, J.A. Ruralresidence: a risk factor for non-Hodgkin’slymphoma in women? Proceedings of the

American Society for Clinical Oncology 10: A390(p. 131) (March, 1991). (Abstract only)

Since most researchers study men only, theauthors began a study of women living in ruralareas and also a case-control study of non-Hodgkin’s lymphoma. Preliminary resultsshow that rural women have significant expo-sure to grain dust, diesel exhaust, and pesti-cides, and that more non-Hodgkin’s lymphomacases occur in women living in rural areas thanin the general population, with the most casesoccurring among women spending the mostyears as rural dwellers. Saskatoon CancerClinic and Dept. of Medicine, University ofSaskatchewan.

Wigle, D.T., Semenciw, R.M., Wilkins, K.,Riedel, D., Ritter, L., Morrison, H.I., & Mao, Y.Mortality study of Canadian male farm oper-ators: non-Hodgkin’s lymphoma mortalityand agricultural practices in Saskatchewan.Journal of the National Cancer Institute 82(7):575–582 (April 4, 1990).

By linking government mortality records withfarming census records, the authors studied therelationship between non-Hodgkin’s lymphomaand variables which might be associated withdifferences in farmers’ health, including acressprayed with herbicides as well as many otherfactors. Overall death rate in this group offarmers was low, and individual exposures topesticides could only be determined generallyand indirectly. Still, increased risks of non-Hodgkin’s lymphoma were associated withincreased number of acres sprayed with herbi-cides in a “dose-response” relationship. Theauthors conclude that their results are consistentwith an increased risk of non-Hodgkin’s lym-phoma in association with the spraying of herbi-cides. Laboratory Centre for Disease Control,Ottawa, Canada.

Wiklund, K., Dich, J. and Holm, L-E. Riskof malignant lymphoma in Swedish pesticide


39

appliers. British Journal of Cancer 56(4):505–508 (1987).

Since 1965, Sweden has re q u i red pesticideappliers to complete a 4-day training courseand obtain a license before handling toxic pes-ticides in agriculture. Records kept for thispurpose allowed the authors to study 20,425pesticide appliers. However, the mean follow-up time was only 12.2 years. The authorsfound a very small, non-significant incre a s e drisk of Hodgkin’s lymphoma and no incre a s e drisk of non-Hodgkin’s lymphoma. A s l i g h ti n c reasing trend in risk for both non-Hodgkin’s lymphoma and Hodgkin’s lym-phoma was found with increased time of hold-ing a license to apply pesticides. It was notedthat the majority of pesticide appliers werefarmers. The authors note that persons withpesticide-applying licenses have higher expo-s u res to phenoxy herbicides than do otherfarmers. They suggest that a longer time offollow-up would be of value. Depts. of CancerEpidemiology and General Oncology,K a rolinska Hospital and Institute, Stockholm,S w e d e n .

Wiklund, K., Lindefors, B.-M., & Holm, L.-E.Risk of malignant lymphoma in Swedishagricultural and forestry workers. BritishJournal of Industrial Medicine 45: 19–24 (1988).

Over 350,000 Swedish men employed in agricul-t u re or forestry were followed for 18 yearsusing a national cancer re g i s t r y. Overall, theg roup showed no increased incidence of non-Hodgkin’s lymphoma, nor did any of the sub-g roups. Increased incidence of Hodgkin’s lym-phoma was found in the “silviculture ”( f o restry maintenance) group and also amongmink farmers and poultry breeders. Theauthors note that veterinarians have ani n c reased incidence of Hodgkin’s lymphoma,raising questions about a possible viral cause.Depts. of Cancer Epidemiology and GeneralO n c o l o g y, Karolinska Hospital and Institute,Stockholm, Sweden.

Woods, J.S., Polissar, L., Severson, R.K.,Heuser, L.S., and Kulander, B.G. Soft tissuesarcoma and non-Hodgkin’s lymphoma inrelation to phenoxyherbicide and chlorinatedphenol exposure in western Washington.Journal of the National Cancer Institute 78(5):899–910 (May, 1987).

In this case-control study, men with non-Hodgkin’s lymphoma were identified thro u g ha state tumor registry and matched with contro l sof similar age who did not have cancer. Bothg roups were asked about their occupationalb a c k g rounds, health histories, and pesticidee x p o s u res. The authors performed an extensivearray of statistical analyses and various bre a k-downs of the study groups. An increased risk ofnon-Hodgkin’s lymphoma was found amongmen with certain occupational exposures top h enoxy herbicides (a type of weed killer),o rganochlorine insecticides (certain insectsprays), organic solvents (chemicals used to dis-solve other substances), and other chemicals usedin agricultural, fore s t r y, or wood product work.Relationship of non-Hodgkin’s lymphoma toimmune system diseases, to past treatment withi m m u n o s u p p ressive drugs, to chloracne, and toe x p o s u re to routine weed spraying was alsoexamined. The authors conclude that a compro-mised immune system is a very significant riskfactor for non-Hodgkin’s lymphoma, and specu-late that prolonged exposure to phenoxy herbi-cides may contribute to the development of non-Hodgkin’s lymphoma in individuals who havesome additional challenge to their immune sys-tems. Battelle Human A ffairs Research Centers,Seattle; Fred Hutchinson Cancer Research Center,Seattle; Swedish Hospital Medical Center.Funding: National Cancer Institute. [Ed. note:Dow Chemical Company re s e a rchers Bloeman et al. (above), cite this study (Woods et al.) as anexample of studies which do not support a possi-ble association between non-Hodgkin’s lym-phoma and phenoxy herbicides].

Zahm, S.H., We i s e n b u rg e r, D.D., Babbitt,P.A., Saal, R.C., Vaught, J.B., Cantor, K.P., &


40

B l a i r, A. A c a s e - c o n t rol study of non-Hodgkin’s lymphoma and the herbicide 2,4-d i c h l o ro-phenoxyacetic acid (2,4–D) in easternNebraska. E p i d e m i o l o g y 1(5): 349–356 (1990).

By using cases of non-Hodgkin’s lymphomarather than mortality data, and by questioningcases and controls closely concerning use of ande x p o s u re to pesticides, these re s e a rchers eliminat-ed some of the weaknesses affecting some otherstudies. 201 white men with non-Hodgkin’s lym-phoma were matched with three times as manycontrols. No correlation was found betweensimply working on a farm and non-Hodgkin’slymphoma, but men who mixed or applied 2,4-Dhad an elevated risk (1.5 times the risk of lym-phoma as the controls). Farmers who failed tochange into clean work clothes after using 2,4-Dhad higher risk of lymphoma than those who didchange clothes; this is another example of howpesticide exposures may vary among individualsin ways that are not detected by measures suchas occupational category or number of daysworking with the substance. Exposure too rganophosphates (a type of pesticide used infarming) had an independent effect of incre a s e dnon-Hodgkin’s lymphoma risk. National CancerInstitute. Funding: National Cancer Institute andState of Nebraska Department of Health.

Zahm, S.H., Babbitt, P.A., Weisenburger,D.D., Blair, A., Saal, R.C., & Vaught, J.B. Therole of agricultural pesticide use in the devel-opment of non-Hodgkin’s lymphoma inwomen. Archives of Environmental Health48(5): 353–358 (Sept./Oct. 1993).

This is one of only a few studies on pesticidesand lymphoma that is focused on women.Many variables were examined, and the findingswere complex. Merely living on a farm did notincrease the risk of non-Hodgkin’s lymphoma,but personally handling several types of pesti-cides (chlorinated hydrocarbons, organophos-phates, and others) did. The highest rates ofnon-Hodgkin’s lymphoma were found amongwomen who had personally handled/applied

pesticides and also had close relatives (mother,father, sister, brother) with histories of lymphat-ic or blood cancers. National Cancer Institute,University of Nebraska, Westat, Inc.

Zahm, S.H. Mortality study of pesticideapplicators and other employees of a lawn careservice company. Journal of Occupational andE n v i ronmental Medicine 3 9 ( 11): 1055–1067 (1997).

Following publication of Dr. Zahm’s 1986 studyof Kansas farmers (see Hoar, 1986), ChemLawnServices asked the National Cancer Institute tostudy the mortality of their employees, many ofwhom are exposed to multiple pesticides intheir work. Zahm examined records for about32,600 employees. Preliminary results indicate aslight increase in deaths from non-Hodgkin’slymphoma over the expected numbers for thisgroup. There were four deaths from non-Hodgkin’s lymphoma, three of them men aged24-35 who applied pesticides to lawns; two ofthese had been employed for three or mor eyears. Zahm points out that, since the averageage at diagnosis of non-Hodgkin’s lymphoma is64 years old in other studies, the young age ofChemLawn employees being studied may nothave allowed the study to adequately assess therisks. As the study continues, the aging of theemployee group will increase the study’s futurepower to detect any excess of non-Hodgkin’slymphoma. National Cancer Institute.


41

Letters, Comments and Reports

Axelson, O. Seveso: disentangling thedioxin enigma? Epidemiology 4(5): 389–392(Sept.1993).

These editorial remarks by a Swedish researcheroffer a clear, open-minded overview of theongoing issues concerning cancer and pesticideexposures.

Blair, A. and Zahm, S.H. Methodologicissues in exposure assessment for case-controlstudies of cancer and herbicides. AmericanJournal of Industrial Medicine 18(3): 285–293(1990).

This article is useful reading for anyone wondering why there are variations in resultsamong the many case-control studies of pesti-cides and lymphoma. The authors discuss thefactors and research challenges which may pro-duce differing results, and note that in mostinstances, errors will fall in the direction of dis-covering less exposure to pesticides, and lesscorrelation with cancer, than actually exist.National Cancer Institute.

Blair, A. Herbicides and non-Hodgkin’slymphoma: new evidence from a study ofSaskatchewan farmers. Journal of the NationalCancer Institute 82(7): 544–545 1990). [This isan editorial comment on Wigle et al.]

Commenting on the discrepancies among studies, as well as the “dilemma to the scientificcommunity in how to draw conclusions regard-ing carcinogenicity of a substance when thee p idemiologic and experimental data do notagree,” Blair points out that exposure to a pesti-cide is difficult to measure (i.e., pesticides areabsorbed through the skin, so studying amountssprayed in the neighborhood may not tell the

story), and that the cancer-causing mechanismof DNAdamage, studied experimentally, maynot be the same mechanism by which phenoxyherbicides cause lymphoma. Yet, correlationsbetween exposure to pesticides and lymphomaincidence and/or mortality are found. NationalCancer Institute.

Blair, A., and Zahm, S.H. Overinterpre-tation of small numbers in the Dow 2,4-Dcohort study. Journal of Occupational andEnvironmental Medicine 37(2): 126–127 (1995).

Blair and Zahm criticize Bloeman et al. (authorsof a Dow Chemical Co. study) for drawing con-clusions from the absence of deaths from lym -phoma in their study. The Bloeman et al. studygroup was so small (878 people) that only 0.3deaths from lymphoma would have beenexpected in a random population of the samesize. Therefore, a doubling of deaths from lym-phoma (for instance) would still possiblyamount to no deaths at all, and it is misleadingto note their absence as being meaningful.Bloeman et al. respond in turn that their articlehad not implied anything inappropriate.National Institutes of Health.

Coggon, D. and Acheson, E.D. Do phenoxyherbicides cause cancer in man? Lancet 1(8280): 1057-1059 (May 8, 1982).

The authors review much of the worldwide scientific literature up to 1982, including thework of Hardell, who found a stronglyincreased risk of lymphoma among peopleexposed to certain pesticides. The main conclu-sion: some researchers have found an increasedrisk, so more studies are needed. University of Southampton, MRC Epidemiology Unit,Southampton, England.


42

Colton, T. Herbicide exposure and cancer.Journal of the American Medical Association 256:9, 1176–1178 (Sept. 5, 1986).

This article is an editorial comment on there s e a rch to that date re g a rding the possiblere l ationship between phenoxyacetic herbicides(such as 2,4-D) and various forms of cancer,including lymphoma. Colton discusses thework of Hardell in contrast to that of Hoar (astudy headed by Hoar is published in the sameissue), as well as the confusion regarding AgentOrange and Vietnam veterans. He gives a clearand useful explanation of the various studymethods (such as case-control studies vs. pro-portional mortality studies), and discusses otherfactors related to study size and the difficulty offinding a large enough cohort to detect anincreased incidence of non-Hodgkin’s lym-phoma, a disease which has a relatively lowincidence in the general population. While hepraises the methods used by Hoar et al., Coltonstates, “Perhaps the question will never beanswered entirely satisfactorily.” BostonUniversity School of Public Health.

Cozen, W., and Bernstein, L. [untitled]Comments on Palackdharry [see below].Oncology 8(8): 77–78 (August, 1994).

The authors agree with both Palackdharry andLongo concerning possible causes of the recentincrease in the incidence of non-Hodgkin’s lym-phoma, with some exceptions and differingemphases. They note an increase in lymphomaswhich originate at sites other than lymph nodes,and the variability of results in studies of pesti-cide exposures and non-Hodgkin’s lymphoma.They question whether exposure to specific ani-mal viruses might also be a factor in some farmpopulations, and suggest immunosuppressionresulting from exposure to ultraviolet radiation(sunlight) as a possible factor in non-Hodgkin’slymphoma. University of Southern California.

Crosignani, P., & Berrino, F. Re: “Role ofthe herbicide atrazine in the development of

non-Hodgkin’s lymphoma” by S. HoarZahm, D.D. Weisburger, K.P. Kantor, F.F.Holmes, & A. Blair. Scandinavian Journal of Work and Environmental Health 1993;19:108–14. Scandinavian Journal of Work andEnvironmental Health 20(3): 223–226.

This article is a response to Hoar Zahm et al.,same journal, 1993. The authors believe that theconclusions of the Hoar Zahm group were tooconservative, and that the data could have beenanalyzed differently (more indicative of a causalrelationship between atrazine and non-Hodgkin’s lymphoma). Issues raised includeconfounding effects due to exposure to multiplepesticides and the likelihood of toxicity/muta-genicity based upon laboratory-based knowl-edge of atrazine. The authors also point outthat measuring the farmers’ exposure to atrazineby counting the number of days they spendapplying it may be misleading, since atrazinestays in the fields for a long time and the farm-ers are also exposed to it while they are cultivat-ing, harvesting, cleaning equipment, etc. TheHoar Zahm group responded that they had con-sidered all of these issues. For the public, thiskind of open discussion serves mainly to illus-trate the difficulties involved in conducting andevaluating research. Epidemiology Unit,National Cancer Institute, Milan, Italy.

Gough, M. Human health effects: what thedata indicate. The Science of the TotalEnvironment 104(1–2): 129–158 (1991).

Mr. Gough reviewed much of the researchthrough 1990 concerning pesticide-cancer links,and decided that there wasn’t a significantcausal relationship. He was particularly criticalof Hardell’s early work in Sweden, which wasnot replicated elsewhere for some years. Basedupon the studies he cited, Gough questioned thelinks between dioxin and cancer. He did indi-cate (p. 152) a link between 2,4-D and lym-phoma. His main concern is the near impossi-bility, in his view, of establishing a clear causallink to pesticide exposure in any particular caseor cases of disease, failing stronger associations


43

and more consistent study results. Center ForRisk Management, Washington, D.C.; later,Office of Technology Assessment, U.S. Congress.

Hardell, L. Malignant lymphoma of histio-cytic type and exposure to phenoxyaceticacids or chlorophenols [letter]. Lancet1(8106): 55–56 (Jan. 6, 1979).

This letter describes a study of 17 male patientswith histiocytic lymphoma. 14 of the 17 workedin occupations involving pesticides. A tableshows 11 patients with lymphoma (mostly inunusual sites), all of whom had been exposedto pesticides in the past, about 15 years beforediagnosis. [Hardell published on similar sub-jects in 1977 — ”describing several patients withsoft-tissue sarcomas and previous exposure tophenoxyacetic acids” (Hardell, L.,Läkartidningen 74: 2753, 1977) and “Malignantmesenchymal tumors and exposure to phenoxyacids—a clinical observation” (Hardell, L.,Läkartidningen 74: 2853).] University Hospital,Umeá, Sweden.

Hardell, L. and Axelson, O. Phenoxyherbi-cides and other pesticides in the etiology ofcancer: some comments on Swedish experi-ences. In: Cancer Prevention. Strategies in theWorkplace. Becker, C.E. & Coye, M.J., eds.Washington, D.C.: Hemisphere PublishingCorporation: 107–119 (1986).

The Swedish authors discuss their research find-ings and the international debate concerningtheir validity. Included is a discussion of themethods used and whether they might havebeen subject to error. They note the longerlatency periods in the Swedish studies com-pared to some of the American studies. Theauthors explain that the unusually completeinformation on cancer cases and pesticide usewhich is available in Sweden allows researchersto perform their studies using actual pesticideexposures rather than broad occupational cate-gories. They suggest that some scientists whohave been critical of the Swedish studies or who

have questioned the results may have failed tounderstand these differences. The authors con-clude that they have indeed uncovered signifi-cant cancer risks. University Hospital, Umeå,Sweden.

Hardell, L. and Eriksson, M. Non-Hodgkin’slymphoma and previous exposure to hexa-c h l o rophene: a case report [letter]. J o u r n a lo f Occupational Medicine 34(8): 849–850 (Aug., 1992).

The authors found a cluster of 8 cases of non-Hodgkin’s lymphoma in which all had beenexposed to hexachlorophene through workexposure in the healthcare field.(Hexachlorophene was widely used as a skincleanser before its toxicity was recognized.)Latency periods varied from 15 to 28 years.This finding may be relevant to the pesticide/non-Hodgkin’s lymphoma issue, since hexa-chlorophene (formerly widely used as a skincleanser) is an organochlorine and, like somepesticides, contains small amounts of dioxin(which occurs as a contaminant during thechemical manufacturing process) and also hasdirect toxic effects aside from the effects of thedioxin. Örebro Medical Center Hospital,Sweden.

Hardell, L. Phenoxy herbicides, chlorophe-nols, soft-tissue sarcoma (STS) and malignantlymphoma [letter/comment]. British Journalof Cancer 67(5): 1154–56 (1993).

This is a response to Smith and Christopher’sstudy (1992). Hardell says that their conclusionsshould be regarded with caution because Smithand Christopher had very few cases; theresponse rate was low (only 56%–70% of thoseapproached were willing to respond); andbecause Smith and Christopher did findincreased risk among those who were exposedto phenoxyacetic acids or chlorophenols for over30 days (compared to those who were exposedfor fewer days). Referring to three other studiesrecently published in leading journals, Hardell


44

notes that there have been few significantresults from studies which estimate exposuresusing occupational categories rather than assess-ing of actual pesticide exposures. ÖrebroMedical Center, Sweden.

Hardell, L.; Liljegren, G.; Lindstrom, G.; & van Bavel, B. Polychlorinated biphenyls,chlordanes, and the etiology of non-Hodgkin’s lymphoma. Epidemiology 8(6): 689(Nov., 1997).

Many pesticides and similar organic toxins arestored in fatty tissues. This letter elaborates onthe results of a case-control study done earlierby this group of scientists in which the levels ofsome organochlorines (this chemical groupincludes several pesticides) in fatty tissues werecompared in people with non-Hodgkin’s lym-phoma and in people who had no lymphoma.They found that people with non-Hodgkin’slymphoma had higher level of PCBs and chlor-danes than the control group. PCBs (bichlori -nated phenyls) are not pesticides; they areindustrial chemicals formerly used in electricalequipment (such as transformers) and ascoolants, etc. They may cause impairment ofimmune functioning. Their use was banned in1977, but these chemicals have a strong abilityto persist and are still found in the environment,in food, and in human tissues. Occupationalexposures were not the cause of the increasedlevels of PCBs in the subjects of this study.Ö rebro Medical Center, Örebro, Sweden.

Hardell, L., Lindstrom, G., van Bavel, B.,Fredrikson, M.; & Liljegren, G. Some aspectsof the etiology of non-Hodgkin’s lymphoma.Environmental Health Perspectives 106 (Supp.2): 679–681 (April, 1998).

The researchers discuss various individual andenvironmental factors which have been found inassociation with increased incidence of non-Hodgkin’s lymphoma, including exposures tocertain pesticides; higher than usual levels ofPCB’s, dioxin, and chlordane in fatty tissues;

blood transfusions; viruses; and various medicalconditions which depress immune functioning.They suggest that impairment of immuneresponse is a common thread linking all thesepossible causes of lymphoma, and that interac-tions between immune response and other fac-tors should be further studied. Örebro MedicalCenter, Sweden.

Hoffman, W. Organochlorine compounds:risk of non-Hodgkin’s lymphoma and breastcancer? Archives of Environmental Health 51(3):189–192 (1996).

This is an overview of 42 epidemiologic studiesand basic scientific studies on the possible rela-tionship between organochlorine compounds(often used as pesticides) and cancer. (Ed. note:Basic scientific studies include a wide range oflab studies involving cell growth using eithertest tubes or animals, chromosome studies, etc.,whereas epidemiologic studies focus on theactual occurrence of the disease amonghumans.) Mechanisms of cancer causationand/or promotion are also discussed. Hoffmannotes the variations in findings when differentstudy methods are used, including the findingsof higher risks when more detailed measures ofpesticide exposure are used. He concludes thatorganochlorines, in particular herbicides such as2,4-D, pose a significant risk of non-Hodgkin’slymphoma as well as breast cancer. He recom-mends a ban on their use. School of PublicHealth/Dept. of Epidemiology, University ofNorth Carolina at Chapel Hill. Funding:German Academic Exchange Organization inthe “Sonderprogramm Epidemiologie”.

IARC working group on the evaluation of car-cinogenic risks. Occupational exposures inspraying and application of insecticides. I A R Cmonographs on the evaluation of the carc i n o g e n i crisk of chemicals to humans 53: 45–92 (1991).

This report on people’s occupational exposuresto insecticides describes what pesticides areused for—worldwide—their various application


45

methods, and the different amounts of insecti -cides that people are exposed to dependingupon application methods used. Some studiesof pesticides and various cancers are mentioned,including several on lymphoma and pesticides.The authors conclude that “there is limited evi-dence that occupational exposures in sprayingand application of nonarsenical insecticidesentail a carcinogenic risk.” (Arsenic compoundsare known to be carcinogenic.) InternationalAgency for Research on Cancer (World HealthOrganization).

IARC working group on the evaluation ofcarcinogenic risks. Occupational exposuresto chlorophenols. IARC Monographs EvalCarcinog.Risk Chem. Hum.41: 319–356 (1986).

This is a long and detailed overview of thekinds of exposures that workers may experiencein the manufacture and use of chlorophenols,which are used as pesticides in many settings.(Chlorophenol compounds include 2,4-dichlorophenoxyacetic acid [2,4-D], 2,4,5-T,2,4,6-T, and at least 14 others which are avail-able commercially.) The authors reviewed someof the literature on lymphoma. Their final eval-uation: “There is limited evidence for the car -cinogenicity of occupational exposure tochlorophenols to humans.” InternationalAgency for Research on Cancer (World HealthOrganization).

IARC working group on the evaluation ofcarcinogenic risks. Occupational exposuresto chlorophenoxy herbicides. IARCMongraphs Eval Carcinog Risk Chem Hum 41:357–406 (1986).

Covering much of the same ground as themonograph above (but with an emphasis onherbicides rather than chlorophenols in general),the authors discuss the structure, manufacture,and occupational exposures to phenoxy herbi-cides, including the use of Agent Orange inVietnam. Urinary concentrations found in vari-ous categories of workers are included, as are

the results of experiments in which volunteersingested phenoxy herbicides or allowed theseherbicides to be applied to their skin, and thenclearance from the body was evaluated. Thematerial on lymphoma is nearly identical to thatin the above monograph, as is the final opinion.International Agency for Research on Cancer(World Health Organization).

Ibrahim, M.A., Bond, G.G., Burke, Cole, P.,Dost, F.N., Enterline, P.E., Gough, M.,Greenberg, R.S., Halperin, W.E., McConnell,E., Munro, I.C., Swenberg, J.A., Zahm, S.H., &Graham, J.D. Weight of the evidence on thehuman carcinogenicity of 2,4-D. E n v i ro n m e n t a lHealth Perspectives 96: 213–222 (1991).

This group examined the published research on2,4-D (a phenoxy herbicide) and cancer. Theyreviewed issues and controversies, case-controlstudies (in which each person with cancer iscompared with one or more people who do nothave cancer), and cohort studies (in which agroup of people sharing some characteristic orcircumstance are studied for their incidence ofcancer compared to those not exposed). Theyfound that, generally, evidence for a cause-effectrelationship between 2,4-D and cancer wasweak, but that some studies which focused onnon-Hodgkin’s lymphoma tended to show astronger link. Eleven of the thirteen scientistsconcluded that 2,4-D is a possible carcinogen.Various universities; National Cancer Institute;Dow Chemical Co.; Office of TechnologyAssessment; NIOSH. Panel funded through theNational Wheat Growers Foundation by a grantfrom the Industry Task Force II on 2,4-DResearch Data; meetings were held at anAmerican Chemical Society facility.

James, W.H. Further evidence that pesti-cides or herbicides cause non-Hodgkins lym-phoma. Journal of the National Cancer Institute83(19): 1420–1 (Oct. 2, 1991).

James poses this question: are alterations in sexhormones in men exposed to certain pesticides


46

related to findings concerning the altered ratioof male/female offspring born to men with non-Hodgkin’s lymphoma (more female children)?He notes that at least one pesticide (1,2-dibro-mo-3-chloropropane/DBCP) has been shown tocause high gonadotropin levels and sterility inexposed males. He suggests that both the sexhormone alterations and the lymphoma couldbe caused by the pesticide exposures. (DBCP, asoil fumigant for killing worms that harm crops,was formerly used on tomatoes and grapes,especially in California, but in 1979 the U.S. EPAbanned all uses except for pineapples inHawaii.) University College of London.

Johnson, E.S. Association between softt i ssue sarcomas, malignant lymphomas, andphenoxy herbicides/chlorophenols: evidencefrom occupational cohort studies.Fundamental and Applied Toxicology 14(2):219–234 (Feb., 1990).

Johnson reviews many studies and, believingthat workers in pesticide manufacturing plantsexperience higher exposure levels than do thosewho spray pesticides, compares these twogroups of studies (pesticide factory workers vs.pesticide appliers). He finds no convincing evi-dence that the presumed higher exposures pro-duce a higher incidence of lymphoma than dolower exposures, but remarks that the factorycohorts are small and the latency periods tooshort. He notes that all groups studied hadmultiple exposures to many potential carcino-gens. He believes the evidence of a pesticide/lymphoma link is weak. National Institute ofEnviromental Health Sciences, EpidemiologyBranch, North Carolina.

Levin, P.H. and Hoover, R. The emergingepidemic of non-Hodgkin’s lymphoma: cur-rent knowledge regarding etiological factors.Cancer Epidemiolgy, Biomarkers, and Prevention1: 515–517 (1992).

This is a report of a National Institutes of Healthconference on research directions and findings

concerning the rapid increase in incidence ofnon-Hodgkin’s lymphoma. By the time of theconference (1991), non-Hodgkin’s lymphomahad become the sixth most common malignancyfor both men and women in the U.S. A varietyof theories and research results were presentedand discussed, including studies on viruses,pesticide exposures, HIV, immune diseases, thetyping of lymphomas, relation of cell type topossible cause, genetic factors, etc. This techni-cal report is an overview of research directionsand scientific thought on non-Hodgkin’s lym-phoma. National Cancer Institute.

Longo, D.L. [Untitled] Comments onPalackdharry. Oncology 8(8): 73–77 (August,1994).

Longo agrees with the Palackdharry article link-ing pesticides and lymphoma, and adds com-ments from his own perspective. He re g a rd s“ c h ronic exposure to man-made chemicalsdesigned to control the environment” as a majorcause of the increase in non-Hodgkin’s lym-phoma. He discusses the particular vulnerabilityof lymphocytes to chemicals as well as the eff e c t sof HIV infection, aging, medical tre a t m e n t s(including chemotherapy for other cancers),v i ruses, and autoimmune disorders, noting, “Itappears that either too much or too little immunereactivity is bad for you.” Longo mentions thework of Scherr, et al. (1992) in which certainoccupations such as the agricultural, fore s t r y, andfishing industries were associated with incre a s e dincidence of lymphoma. After briefly re v i e w i n gthe re s e a rch concerning possible causes and theways in which lymphoma develops within cells,he points out that more knowledge is neededabout all the various causes of non-Hodgkin’slymphoma, especially in view of the re c e n ti n c rease in incidence. National Cancer Institute.

Markovitz, A. and Crosby, W.H. Chemicalcarcinogenesis. A soil fumigant, 1,3-dichloro-propene, as possible cause of hematologicmalignancies. Archives of Internal Medicine144(7): 1409–1411 (1984).


47

This is an unusual case report. In April, 1973,nine firemen who cleaned up a truck spill of 1,3-dichloropropene (a pesticide used to kill wormsin soil) were treated for toxic symptoms. Of thenine firemen, two became ill in October, 1979.Both were found to have developed histiocyticnon-Hodgkin’s lymphoma. Both died in early1980. In a third case, a farmer was exposed to1,3-dichloropropene for 30 days (it leaked ontohim continually from a broken hose on hisequipment). He developed acute myelomono-cytic leukemia. A year later, when he againapplied the same chemical to his fields and wasagain exposed, the leukemia became very agres-sive and he died. The authors note that a casereport of this kind can only serve as an alert.UCLASchool of Medicine and Medical Corps;Walter Reed Army Institute of Resarch.

Morrison, H.I.; Wilkins, K.; Semenciw, R.;Mao, Y.; Wigle, D. Herbicides and cancer.Journal of the National Cancer Institute 84(24):1866–1874 (Dec. 16, 1992).

This well-organized review of the studies onherbicides and cancer includes tables that showthe types of studies done, locations, dates, andtypes of cancers studied. There is a discussionof study design and the ways it influencesresults. The authors note that “the most con-vincing evidence that herbicides may be humancarcinogens arises from studies of non-Hodgkin’s lymphoma” and “there is reasonableevidence to suggest that phenoxy herbicideexposure results in an increased risk of develop-ing non-Hodgkin’s lymphoma.” Bureau ofChronic Disease Epidemiology, Health andWelfare Canada.

Palackdharry, C.S. The epidemiology of non-Hodgkin’s lymphoma: why theincreased incidence? Oncology 8(8): 67–78(August, 1994).

In light of the dramatic increase in incidence of non-Hodgkin’s lymphoma in the U.S. (called an “epidemic” by some observers),

Palackdharry provides an overview of the literature to date and discusses possible causesof lymphoma. Concerning pesticides, theauthor notes the varying results obtained byresearchers studying occupationally-exposedpopulations, and also the high level of pesticideuse among private homeowners. She discussesthe hypothesis that chemical exposures maypredispose to non-Hodgkin’s lymphoma byweakening the immune system. MedicalCollege of Ohio.

Roberts, H.J. Effects of pentachlorophenolexposure. Lancet 349(9069): 1917 (June 28,1997).

Roberts points out that pentachlorophenol(PCP) exposure, which is very widespreadbecause PCP is used to treat wood, has beenassociated with Hodgkin’s and non-Hodgkin’slymphoma as well as many other serious healthproblems. [PCP is a manmade chemical used asa wood preservative, preventing growth of bac-teria and fungi which cause wood to rot. Peoplemay be exposed to PCP by touching or workingwith wood in construction projects, breathingcontaminated air at wood treatment facilities orlumber mills, breathing contaminated air nearwork sites or waste sites, drinking contaminatedwater near waste sites, etc.] Palm Beach Institutefor Medical Research Inc., Florida.

Smith, A.H., and Bates, M.N.Epidemiological studies of cancer and pesti-cide exposure. American Chemical Society(414): 207–22 (1989).

After conducting an overview of the scientificliterature on pesticides and cancer—not justlymphoma—the authors conclude that onlyarsenic-containing pesticides appear to pose arisk. They consider fairly low ORs and varyingresults in various studies to be signs of failure todetect any causal relationship. Also, the authorsconsider immediate toxic effects of chemicals,such as skin rashes, to be significant signs ofpotential carcinogenicity, and the lack of such


48

effects to indicate safety. Univ. of Calif,Berkeley.

Weisenburger, D.D. Epidemiology of non-Hodgkin’s lymphoma: recent findingsregarding an emerging epidemic. FifthInternational Conference on MalignantLymphoma, Lugano, Switzerland, 1993 (p. 29of the proceedings).

This report offers a brief but thorough overviewof risk factors which have been found to beassociated with non-Hodgkin’s lymphoma inepidemiological studies. As regards pesticides,the author reviews the various studies and con-cludes that pesticides may play a significant rolein the recent increase in incidence of non-Hodgkin’s lymphoma. University of NebraskaMedical Center.

Youness, E., Ahearn, M.J., & Drewinko, B.Simultaneous occurrence of non-Hodgkin’slymphoma and spontaneous acute granulo-cytic leukemia. American Journal of ClinicalPathology 70(3): 415–420 (Sept., 1978).

This is a report of a single case of non-Hodgkin’s lymphoma and acute granulocyticleukemia occurring simultaneously in the sameperson—which is very rare. The patient hadbeen chronically exposed to an insecticide (typeunknown), and the authors conjecture that thisexposure may have been a causative factor.University of Texas System Cancer Center,Houston.

Zahm, S.H. Geographical variation in lym-phoma incidence. British Journal of Cancer57(4):443 (1988).

This is a brief letter in response to a report byBarnes et al., in which a greater incidence ofnon-Hodgkin’s lymphoma, but not Hodgkin’slymphoma, was found in rural areas ofYorkshire, England, than in urban areas. Shepoints out that the difference, also found in the

U.S., may be due to greater pesticide exposuresin rural areas. National Cancer Institute.

Zahm, S.H., and Blair, A. (1992). Pesticidesand non-Hodgkin’s lymphoma. C a n c e rR e s e a rc h 52(19): 5485s–5488s (October 1,1992).

This concise, but information-packed articlep rovides an excellent overview of the re s e a rc hfindings to date. Zahm and Blair review themany studies, methods, and results, and sum-marize the findings, their meaning, reasons ford i s c repancies in results, and re c o m m e n d a t i o n sfor future re s e a rch (i.e., better evaluation ofpesticide exposures in study populations).Noting that the incidence of NHL rose over50% in the fifteen years preceding 1992, theyconclude that 2,4-D in particular may wellhave been a factor in this increase, since studiesin which length of time of exposure was con-s i d e red show increasing risk of non-Hodgkin’slymphoma with increasing time of exposure ,even though less specific studies have shownmixed results. National Cancer Institute.

Zahm, S.H. Epidemiologic research onVietnam veterans. Difficulties and lessonslearned. Annals of Epidemiology 5(5): 414–416(Sept., 1995).

Zahm’s article is a preface to two epidemiolog-ical studies of Vietnam veterans, Watanabe etal. and Dalager et al. She says that re c o rd -keeping on the 9 million members of the armedservices during the Vietnam era was not donein a way that makes good epidemiologicalre s e a rch possible. No well-organized militaryre c o rds exist concerning either the pesticidee x p o s u res or the Vietnam service of individu-als; re s e a rchers would have to comb thro u g hall the re c o rds separately to discover whichpersons served in Vietnam. The lesson learnedis that military re c o rds should be kept in a waythat allows systematic retrieval of information.National Cancer Institute.


49

Zahm, S.H., & Ward, M.H. Pesticides andchildhood cancer. Environmental HealthPerspectives 106 (Supp. 3): 893–908 (June,1998).

This overview of the scientific literature con-cerning children, pesticides, and cancer givesinformation on the scope and nature of chil-d ren’s exposures to pesticides. Children aref requently exposed to pesticides in their homes,and are more vulnerable than adults to theireffects. The authors summarize the results ofeight studies of non-Hodgkin’s lymphoma. Inseveral of these, the incidence of non-Hodgkin’slymphoma was greater when the time oramount of exposure to pesticides was greater.This dose-response effect may indicate acausative relationship. In one study, childrenwhose mothers used household insecticidesdaily showed 5.2 times the risk of non-Hodgkin’s lymphoma as children whose moth-ers did not use them. Children whose mothersused these insecticides 1 or 2 times a week had2.2 times the risk of children whose mothersused them less than once a week. However,many of the studies were small and/or reportedcancers without distinguishing among types ofcancer. The authors state that much more inves -tigation of the relationship between pesticidesand childhood cancer is needed. NationalCancer Institute.


Dog Studies: Pesticides and Lymphoma

Hayes, H.M.; Tarone, R.E.; & Cantor, K.P.On the association between canine malignantlymphoma and opportunity for exposure to2,4-dichlorophenoxyacetic acid.Environmental Research 70(2): 119–125 (Aug.,1995).

In 1991, the authors published the results of ac a s e - c o n t rol study of lymphoma in dogs in re l a-tion to their exposure to 2,4-D, a chemical herbi-cide frequently used in lawn and garden care as aweed killer. They had found a small but signifi-cant increase in incidence and also a dose-response effect . The present report is a detailedanalysis of the data in response to critical objec-tions by industry-sponsored reviewers. Criticscomplained that there could have been “re c a l lbias” (greater tendency to remember using pesti-cides) among the owners of dogs with cancer.H o w e v e r, Hayes et al. had used separate contro lg roups: dogs with cancers other than lymphoma,and dogs without cancer. There was no signifi-cant diff e rence between the two control gro u p s ,showing that recall bias re g a rding cancer in gen-eral was not an issue. In this report, data show-ing the geographic location of cases and contro l s ,use of pesticides other than 2,4-D, multiple pesti-cide use, and owner application vs. application ofthe pesticides by a lawn care company or pro f e s-sional landscaper were all included. The authorsconclude that there is no evidence of inappro p r i-ate bias in their work, and that while no one epi-demiological study can prove causation, “untiladditional studies are undertaken…the higherrisk we found in dogs…would, at the very least,a rgue for prudence in the pursuit of a perfectlawn.” National Cancer Institute.

Sternberg, S.S. Canine malignant lym-phoma and 2,4-dichlorophenoxyacetic acidherbicides. Journal of the National CancerInstitute 84(4): 271 (Feb. 19, 1992).

In response to Hayes’ 1991 findings, Sternbergwrote a letter expressing doubt whether theincreased incidence of lymphoma in dogs couldbe attributed to 2,4-D, since blood levels of 2,4-D in the dogs had not been measured and it wasnot even known to what extent, if any, dogsabsorb 2,4-D from lawn environments. He saidthat dogs are more likely to lick their backsidesthan their paws, making ingestion of lawn-careproducts unlikely. Memorial Sloan-KetteringCancer Center.

Reynolds, P.M.; Reif, J.S.; Ramsdell, H.S.;Tessari, J.D. Canine exposure to herbicide-treated lawns and urinary excretion of 2,4-dichlorophenoxyacetic acid. Cancer EpidemiolBiomarkers Prev 3(3): 233–237 (Apr.–May1994).

To determine whether dogs absorb 2,4-D andexcrete it through their urine after their ownershave applied lawn herbicides (weed killers), theauthors tested the urine of 44 dogs which hadrecently been exposed to herbicides on theirhome lawns and 15 dogs which had not.Findings showed a clear correlation betweenrecent application of lawn herbicides and thepresence of 2,4-D in the dogs’ urine. Some dogswhose owners had not recently applied herbi-cides had small amounts of 2,4-D in the urine,indicating that they might have been exposed tothe chemical at parks or by contact with pesti-cides applied to neighbors’ yards. These find-ings appear to support those of Hayes et al. con-cerning canine lymphoma and to contradict theSternberg letter. College of Veterinary Medicine,Colorado State University.

50

LYMPHOMA FOUNDATION OF AMERICA

Helping You Get Better

Lymphoma Foundation of America is the national organizationdevoted to helping lymphoma patients and families.

We were founded in 1986 by people with lymphoma who cametogether to help each other recover from cancer.

We organized the first lymphoma support group in the UnitedStates. We advocate your right to know the truth about therisks and benefits of cancer treatments — whether standard,experimental, or alternative.

Lymphoma Foundation of A m e r i c a is an independent, nonprofitc h a r i t y. We offer support, experience, advice, and a helpinghand. All programs and services are free.

Lymphoma Foundation of America is listed with the U.S. government’s cancer information hotline, 1-800-4-CANCER, a service of the National Cancer Institute.

Need to get rid of an old car? Want a great tax benefit?

Click on the tow truck to find out howyou can donate a car. We’ll tow it for free.

Lymphoma Foundation of America 1100 N. Main St. • Ann Arbor, MI 48104

Tel: (734) 222-1100 • Fax: (734) 222-0044www.lymphomahelp.org

This site does not contain any drug company advertisements or industry promotions.

Combined FederalCampaign #1163 Donor Option

do pesticides cause lymphoma? - rapal · weed killers insect sprays f a r m i n g lawn care wa t e...

Documents