In November, 1975, the monthly figuresbased on a 10 percent sample of all deathcertificates reported to the National Center for Health Statistics were added upfor the first seven months of the year,and released to the press. These figuresshowed an apparent 5.2 percent increase in cancer mortality rates compared to the same period the year before.If true, the information would indeed bestartling, since the cancer death rate hasbeen increasing at less than one percenta year. Furthermore, when broken downby site, the trend over the past yearsshows that cancer mortality would bedeclining, but for the increase in lungcancer deaths, directly related to cigarette smoking. Concern intensified whenthe Metropolitan Life Insurance Co. issued figures that appeared to corroboratethe sharp upward rise of cancer mortalityin 1975. The result was frighteningheadlines and sensational news articles;a cancer scare was underway.

These figures were given added impact by the instant interpretations placedon them to support one or another supposition on the causes of cancer. Thus, theNovember, 1975 headlines wereprompted by an inquiry from a memberof Congress to the Director of the National Cancer Institute, as to whether thesudden “¿�increase―was due to “¿�thegrowing use of an exposure to chemicalsin our daily life.― It was in search ofsupport for this hypothesis that an aide

to the Congressman decided to add upthe monthly mortality figures of the National Center for Health Statistics, andrelease them to the press.

However, a strange thing happened.The cancer statistics for early 1976showed a decrease of cancer mortalityin the monthly reports of the NCHS,compared to 1975. And the October,1976 Statistical Bulletin of the Metropolitan Life Insurance Co. declared:“¿�CancerMortality Declines in First SixMonths of 1976.―These 1976 statistics,however, produced no headlines.

It is highly unlikely that the impact ofcarcinogens in the environment could beeliminated from one year to the next.Surely, all the food additives could nothave been miraculously removed fromour diet since Metropolitan reported thatamong their policy holders, death ratesfrom cancer of the digestive system decreased by nine percent. The effect ofthe environment on cancer is not instantaneous; 10-30 years are usually required from exposure to cancer incidence and/or mortality.

For a clear analysis of what really occurred, we are indebted to Dr. LeonardChiazze and his colleagues at the National Cancer Institute who provided asober and scientific exposition inJAMA, entitled “¿�TheCancer MortalityScare: Problems of Estimation UsingMonthly Data― (JAMA 236:2310-2312, 1976). Their analysis shows that

VOL. 27, NO. 1JANUARY/FEBRUARY 1977 63

Editorial: The Environmental Time BombThat Never Went Off

the abnormally high figure reported in1975 was probably related to an unusually high mortality from influenza inearly 1975 coupled with low rates duringthe same months in 1974. The true increase in 1975 compared to 1974 wasabout 0.7 percent, compared to a 0.2percent average annual increase in theyears 1968-1974.

But even more important, the authorscaution that early reports of the NationalCenter for Health Statistics cannot beused as a basis for extrapolating the annual trend. A warning to that effect isnow contained in the monthly reports.Furthermore, the raw death rate figuresare not standardized for age, race andsex. To show how important this is, Dr.Chiazze's study points out that “¿�nearlyall of the increase in the total cancerdeath rate from 1968 to 1974 was accounted for by the aging of the population.― Also, the provisional estimatesdo not reveal changes in mortality forindividual cancer sites. Such data are essential for proper analysis “¿�sinceconsidering only changes in total cancermortality obscures many importanttrends such as the rapid rise in lungcancer mortality and the steady declinein stomach mortality.―

The authors conclude that: “¿�Anyattempt to monitor trends in overall cancer

mortality by analyzing these provisionalmonthly estimates is a misuse of theMonthly Vital Statistics Report. Largechanges should raise our index of suspicion and provide leads for further exploration.―

Fortunately, at the time of the sensational November, 1975 reports, the Epidemiology Department of the AmericanCancer Society reviewed the situationand announced that in all probability theage-adjusted figures would show nosuch extraordinary increase. It was further pointed out that a bias may havebeen introduced by the influenza figuresin 1975 compared with 1974.

Hasty judgments are a disservice tothe serious study of environmentalcancer. It is a sad commentary that theimprobable trends of higher cancerdeaths released in 1975 received suchwide publicity, while the most destructive known carcinogen—that is ofcourse, cigarette smoking—does notget proper emphasis.

Instant analysis is easier, but Dr.Chiazze's report clearly shows the meritof thorough epidemiologic study. @

7@L4i.41/dOM.D.

Send Us Your Questions on CancerWe will refer specific questions in any area of cancer management to a leadingcancer specialist.All questions will be answered and some will be published in Ca-A Cancer Journalfor Clinicians.Please submit your questions to:

EditorCa-A Cancer Journal for Clinicians

American Cancer Society777 Third Avenue

New York, New York 10017

64 CA—A CANCER JOURNAL FOR CLINICIANS