It’s a very seductive message designed to undermine trust in standard methods of cancer treatment. I don’t think most consumers are likely to recognize why the message is misleading. There really has been a “cancer holocaust,” as I’ll explain, but its cause is neither mysterious nor unavoidable. The real “cancer holocaust” has no relevance to Goldberg’s advocacy for “alternative” cures for cancer.
There’s a great deal to unpack regarding Goldberg’s opening pitch. Please bear with me as I attempt to explain the insidious nature of the cancer holocaust ploy.
I’m not sure where Goldberg came up the 1 in 33 figure, which, given the phrasing of his first sentence would represent an estimate of prevalence (how common a condition is). I am unaware of any system that was in place in the U.S. in 1900 to provide cancer prevalence data. Even today, reports on cancer trends present incidence data, which tell us about people who are newly diagnosed with cancer, and mortality data, which tells about people dying of cancer.
So let’s see if I can help Goldberg provide some substance to his argument by considering real data on mortality. In 1900
, cancers were the eighth leading cause of death in the United States and accounted for one out of every 25 deaths. On average, for every 100,000 people in the U.S. population, 64 people died of cancer in 1900. In 2011
, cancers were the second leading cause of death in the United States and accounted for one out of every 4 deaths. On average, for every 100,000 people in the U.S. population, 185.1 people died of cancer in 2011.
Now, didn’t I just make the case for a cancer holocaust stronger than Goldberg did? Perhaps, but I did so by using accurate statistics presented with subtle statistical sleight-of-hand and misdirection. To explain, I need to discuss two important reasons for the increased importance of cancers as contributor to the burden of mortality.
The main reason that cancers are now a more important cause of death in the U.S. than they were in 1900 is that today a far greater proportion of the U.S. population is fortunate to live long enough to develop the types of cancer that cause the most deaths. You can’t die of cancer if you have already succumbed to other diseases or injuries.
In 1900, life expectancy at birth was less than 50 years; by 2009 it was over 78 years
. In 1900, it was common for people to die young because there hadn’t yet been much public health progress in preventing deaths from infectious diseases, malnutrition, and injuries. The top seven leading causes of death were: (1) pneumonia & influenza, (2) tuberculosis, (3) diarrhea, enteritis, and ulceration of the intestines, (4) diseases of the heart, (5) intracranial lesions of vascular origin, (6) nephritis (inflammation of the kidneys), and (7) accidents.
Deadly cancers are diseases that develop predominantly among middle-aged and older adults. The American Cancer Society estimated
that 585,720 Americans would die of cancer in 2014 with 1,960 of those deaths (0.3%) occurring in the 1 to 19-year age group. The types of cancer that often kill middle-aged and older adults are very different from the much less common cancers that are responsible for deaths of children and young adults.
The American Cancer Society’s report on “Cancer Statistics, 2014”
contradicts Goldberg’s claim that cancer is the leading cause of death among children. Cancer is the leading cause of death among children through disease,
but in the U.S. in 2010, for those who were 1 to 19 years of age, accidents (or unintentional injuries, in public health parlance) were the leading cause of death, taking the lives of more than 7,500 children. For young people (ages 1 to 19), cancers were the fourth leading cause of death among males and the second leading cause of death among females. Even in the 20 to 39-year age group, accidents were the leading cause of death while cancer was the fifth leading cause of death among males and the second leading cause of death among females.
In the U.S., the incidence rate for childhood cancers increased at an annual rate of 0.6% per year
from 1975 to 2010. Although it isn’t clear what caused this slight increase, it is likely due to a combination of unidentified environmental influences, improved diagnostic efforts including increased use of magnetic resonance image, and increased access to medical care. But in the same period, death rates have declined by an average of 2.1% per year. This decline is attributable to improved treatment and supportive care. Trends in childhood cancer incidence and mortality do not suggest that there has been a “cancer holocaust.”
I don’t know whether Goldberg is correct about cancer being only the tenth leading cause of childhood cancer in 1926. I suspect he’s not far off. But that low ranking doesn’t mean that cancer rates have gotten dramatically worse. It reflects how major causes of death among children in 1926 are much less important today (largely due to improvements in sanitation and vaccination).
In the United States, cancers are now the second leading cause of death (behind heart disease) for the population overall instead of the eighth leading cause of death (as they were in 1900). The higher ranking of cancers as a cause of death today is due mainly to the greater progress made against other diseases and injuries that killed so many young people in 1900. Rankings for a disease category as a cause of death are affected by mortality trends for other diseases.
By analogy, a sports team can have the same won-loss record two years in a row, but the team’s ranking in league standings can be very different in those two years. Rankings are affected by what competitor teams do.
There has been clear progress in reducing death rates from all cancers combined
, but the amount of progress has been disappointing compared to the dramatic progress we’ve seen in reducing heart disease death rates in the United States in recent decades.
Cancers are already the leading cause of death in the United States for people under age 85
. If the trend of dramatic decline in heart disease death rates continues, cancers will eventually become the leading cause of death in the United States when all age groups are considered.
Moving up in rank as a cause of death doesn’t mean that the risk of dying of cancer is increasing. The mortality rates for cancer that I presented above for 1900 and 2011 are the correct crude (unadjusted) mortality rates. But it would be a mistake to argue that the numbers mean that from 1900 to 2011, there has been almost a tripling in the cancer mortality rate (from 64.0 per 100,000 to 185.1 per 100,000).
The problem is that the numbers are crude when they should be properly adjusted in order to avoid making an “apples to oranges” comparison. The necessary adjustment in this case must take into account how the population in 2011 includes a significantly larger proportion of older people at high risk for cancers than did the population in 1900.
In other words, comparing mortality rates for two very different time periods requires adjustment for age. Age-standardized rates rather than crude rates are needed to compare rates from populations from different places or times.
When I teach introductory epidemiology, I use an analogy to illustrate the importance of adjusting statistics in order to make fair comparisons. I ask students to consider changes over decades in the entry-level salary of public school teachers in the U.S. My students are surprised when I tell them the salary I was expecting when I was pursuing a public school teaching job in 1980. I wound up with a different kind of job that gave me an annual salary that was close to what I could have expected as a high school science teacher: $10,500. Contrast that with the average annual entry-level salary of public school teachers in 2007: $35,284
Is it correct to say, based on these data, that entry-level public school teachers are paid more than three times better in 2007 than they were in 1980? It might look that way, but most of my students are savvy enough to realize that dollars in 1980 are not equivalent to dollars in 2007. In order to make an unbiased comparison between salaries in 1980 and 2007, it is necessary to adjust the salary figures for inflation.
The crude salary data, just like the crude cancer mortality data, simply aren’t appropriate for comparison. Age-standardization
(age-adjustment) of mortality rates from different years is needed to make fair comparisons just as inflation-adjustment is needed for comparisons of salaries for different years.
Making comparisons using crude, non-standardized mortality rates is a misleading way of making the case that there has been a “cancer holocaust.” But it’s quite clear that there has been a very real, tragic “cancer holocaust” that has nothing to do with how effective standard or “alternative” cancer treatments are. The real cancer holocaust that we continue to face is due to cigarette smoking
, the second most important reason for high rates of cancer mortality today.
According to the National Cancer Institute: