People in industrialized countries are dying from cancer at a greater rate than ever, and the increase cannot be explained by the aging of the population, suggesting environmental causes are playing a role, according to a report being published by the New York Academy of Sciences.

About 2.3 million of the 11 million deaths that occur annually in industrialized countries are caused by cancer. The report found that when lung cancer, which is most often smoking-related, is excluded, the cancer death rate for men in the industrialized world has increased 9 percent since 1950.

Among women the pattern is more complex, with cancer death rates rising in some countries and falling in others.

"Cancer is increasing above and beyond what we would expect due to aging or cigarette smoking alone," said Devra Lee Davis of the National Academy of Sciences, who edited the 345-page report with David Hoel, acting director of the National Institute of Environmental Health Sciences.

The report, compiled by 26 scientists from 13 countries, greatly expands upon the findings of a study by Davis and other epidemiologists published last August, which reported striking increases in death rates from several kinds of cancer in industrialized countries.

That study has generated a scientific controversy. Some critics contend the increase in death rates reflects little more than an aging population coupled with improvements in diagnosis and record-keeping.

In an interview yesterday, Davis said the new report establishes that those factors, while important, cannot fully explain the changes in death rates. She compared the current situation to that seen in the 1940s, when rising lung cancer death rates gave scientists the first clue that cigarette smoking might be linked to the disease.

"I think that it would be a real tragedy if we continued to insist that these {changes} are nothing more than improvements in diagnosis," she said.

Death rates for lung cancer have been increasing dramatically worldwide in recent decades because of smoking. But death rates for a variety of other tumors are also rising among older people in industrialized countries -- including brain tumors, breast cancer, non-Hodgkins lymphoma, multiple myeloma (a tumor of white blood cells), and malignant melanoma (a skin cancer).

In an analysis of the international data, epidemiologist John Bailar of Montreal's McGill University wrote, "Taken together, these sets of trends indicate that the worldwide effort to control cancer has failed to obtain its primary objective -- substantial reduction of the overall cancer death rate -- despite some 40 years of intense effort."

Davis said environmental factors are suggested as a source of the increased cancer death rates because the highest rates are in the most industrialized regions of countries and the rates for men are rising faster than for women, suggesting a possible role for occupational exposure.

For example, the study found that between 1968 and 1986, with the exception of lung cancer, men were dying from cancer at increasingly higher rates than women in the United States, Italy, Japan, Czechoslovakia and West Germany.

For some tumors, such as malignant melanoma and colon cancer, the changing gender difference in death rates was particularly striking. For instance, in 1970 among Americans over 65, the death rate from melanoma was 40 percent higher for men than for women. By 1982, it was 70 percent higher for men than for women. Although melanoma is increasing in both sexes because of increased sun exposure, Davis said the male-female difference suggested that other environmental factors may also be important.

The report also offers evidence, in a chapter by Davis and four other researchers, that lung cancer rates in nonsmokers have increased in the United States, Japan and Italy. Using a complicated mathematical model, the researchers calculated that for an American woman who does not smoke, the risk of developing lung cancer now is higher than it was for a female smoker in 1955.

But that view was challenged in a chapter written by statisticians from the American Cancer Society, who used data from a voluntary survey to conclude there was no evidence of an increase in lung cancer among nonsmokers.

The most striking increases in death rates have occurred among older adults, suggesting that part of the rise in cancer deaths may be occurring because people are living longer and because death rates from heart disease -- the biggest cause of death in developed countries -- are dropping.

For instance, deaths from multiple myeloma are increasing slowly among people in their late 50s, but are rising by more than 4 percent per year in people over 85. Brain tumor death rates among the elderly have also increased substantially in many industrialized countries, the report found.

But some tumors also appear to be increasing in younger people. For example, Davis reports in a study that will be published in the American Journal of Industrial Medicine that from 1973 to 1987 the frequency of brain tumors increased 2 percent per year in Americans under 45.