You Have To Die Of Something: Cancer Deaths to Double

The morning headline is designed to make you nervous: Cancer deaths to double by 2030: study.

“My God!,” you reason to yourself before the coffee hits the bloodstream, “That means the chances are double that I’ll be stricken down by the Big ‘C’!” If you’re typical—and most of you are not—you will add: “Why doesn’t the government do something!”

Here’s the story, linked, I think appropriately, from Xinhua:

About 13.3 million people will die of cancer a year by 2030, almost double the number in 2008…The research, conducted by the International Agency for Research on Cancer (IARC), also indicated that almost 21.4 million new cases of the disease will be diagnosed annually in 2030.

In 2008, 7.6 million people died of cancer and there were an estimated 12.7 million new cases diagnosed…The IARC said the burden of cancer was shifting from wealthier to poorer nations.

There are at least five reasons why nobody should worry about this study. Coincidentally, they form the same five examples of how to cheat with statistics.

First, the obvious: it might be true that cancer rates will increase, and not just by two times, but by four, or eight, or even sixteen! Yes, everybody might die of cancer by next Tuesday. But it’s just not that probable. It’s most likely that things will continue as they have.

Here are the reasons why you should not become nervous:

  1. In 2008, there were about 6.7 billion of us prowling around the Earth. The IARC went out and counted everybody who died in that year, and placed a bean in a bag if that death was caused by cancer. At the end, they counted the beans: they told us there were 7.6 million of them. That’s a rate of 1.1 cancer deaths per 1000 people.

    They claim that by 2030 13.3 million will die of cancer. However, in 2030 there will be almost another billion of us, so we’d expect that more people would die of cancer even in the death rate stayed the same. Given the expected population of 8.3 billion, the IARC figures that the cancer rate will be 1.6 per 1000.

    This is an increase in the rate, all right, by about 1.4 times, but not a doubling. Still, any increase is bad news, right?

  2. Did you notice the plea to guilt in the story? Maybe you didn’t, especially since these pleas saturate the media and become part of the background noise. The study said, “the burden of cancer was shifting from wealthier to poorer nations.” Those poor poor people!

    When people die, they die of something. You will die of something. The doc in charge of your carcass will check a box on a form that says “Heart Disease” or “Cancer” or whatever.

    They cannot not check a box, at least here in the States. If you have multiple diseases, each of which contributed to your exit, still only one box will be checked. The doc will likely pick his favorite, or he will guess, not always correctly.

    Mistaken causes of death are especially common in “developing countries”, where cancer is often undiagnosed. In other words, we expect that as technology and bookkeeping improve in these forlorn places, certain disease rates will “increase”, but only on paper. Thus, you cannot look at the increasing trend of cancer in developing countries as direct evidence that cancer itself is becoming more common.

    Now, these facts imply that the 1.6 per 1000 rate should be adjusted down to account for increasing diagnostic accuracy. Say it comes to 1.3 to 1.4. Still higher than now. Still reason to worry?

  3. The “leading cause” of death nowadays is heart disease, not cancer. There will always be a leading cause of death, incidentally. Because of its number-one statues, a lot of money goes into curing heart disease: new cholesterol drugs come on the market weekly, it seems.

    This means that as time progresses, fewer people will die of heart disease—but these hearty folks will still die of something, including cancer. Cancer rates will thus increase, but only because people won’t be given the opportunity to keel over on the subway with massive coronaries.

    Improvements are being made in cancer treatments, too, but not as quickly as they are for heart disease. However, this evidence doesn’t allow us to lower our estimated cancer death rate, because it will be true deaths attributed to cancer will increase.

  4. We do not know what the death rate of cancer is with respect to other deaths. Our rates so far have been deaths per all people. What we’d like is deaths caused by cancer divided by all deaths.
  5. Lastly, we must ask ourselves this question: what are the chances that this United Nations organ charged with investigating cancer—its very livelihood is dependent on the disease—will, let us say, exaggerate the threat caused by cancer? To ask it is to answer it.


  1. One other factor you don’t mention is aging of the population. Even if age-specific rates of cancer remained unchanged, total cancer deaths will rise because more and more people are living to older ages. In fact, this partially explains the “shift” to poorer countries. Such a shift might actually be a marker of “progress” insofar as the elimination of childhood diseases and infectious diseases in adults ensures that a greater fraction live to an age where cancer is more common. With few exceptions, life expectancy is rising in poor and rich countries, meaning that overall mortality risk is declining. Thus, this doubling in cancer deaths simply represents a shift in cause of death rather than a signal that cancer is increasing overall mortality risk.

  2. Cancer rates go up when a population does not see a 30% mortality rate before the age of 5. Cancer rates go up when you are not dying of cholera, typhus and malaria. Cancer rates go up when you actually get a chance to live.

  3. I’d be curious to know on what that model is based.
    As with all models not describing purely mechanistic systems (i.e. free fall) what is the confidence in it’s predictions? Did you use your “skill analyzer” to rate its accuracy?
    I am always skeptic when reading these daily news about how worse/more dangerous/less happy/… our life will be in X years from now.
    And yet our life expectancy keeps increasing. And population too.
    Something doesn’t add up…

  4. Matt, you could of saved quite a bit of time by moving number 5 up earlier in the reasoning sequence. In the future should probably call that statement BS Test-1 [for Briggs Standard Test #1, not the other, more earthy term that sprang to mind].

  5. This is great news! Fewer people will die of infection, malaria, and starvation.

    Also influencing cancer rates is better means of detection. In this century, millions of people will be diagnosed with and treated for early-stage cancers that would have gone completely undetected in the last century. There will be many hand-wringing articles about the skyrocketing cancer rates. When, in fact, people are no more susceptible to cancer then they ever have been.

  6. It seems almost any study sponsored by the UN is misleading, or just plain false. I am reminded of the WHO mortality study. When you look at the numbers, they have compared dissimilar populations using data recorded under different standards, and I don’t remember any estimates of error. It makes you wonder why the study was released.

    49er is correct. This one deserves the BS (Briggs Standard) penalty flag to be thrown.

  7. Sorry, didn’t have time to read the attachment. Did the phrases: “worse than thought” or “more study is needed” show up?

  8. Here are some more mortality statistics:

    * The number of induced abortions worldwide in 2008 was ~42 million. That’s 6.2 abortion deaths per 1,000 people, about 6 times the cancer death rate.

    * For every 1,000 women of childbearing age (15–44) worldwide, ~30 were estimated to have had an induced abortion in 2008.

    Of course, there is no need to worry about this because if you are old enough to worry, you have already successfully avoided prenatal death.

    But then there’s genocide. According to the Genocide Clock, between 1917 and 2008 a total of 405 million people were murdered by governments or other organized political groups. That’s on average:

    * 5.7 million people per year, about 5 times the cancer death rate (based on current world population, probably twice that based on the population at the date of the deaths).

    You’re never too old to be murdered.

  9. My pal the epidemiologist was telling me about the fun he had, years ago, analysing some cause-of-death data from a Mediterranean country. They looked odd. On enquiry, he found that the cause of death was sometimes decided by a doctor, sometimes by a nurse and sometimes by the family. As he said, it’s food for thought when you’re toying with your spaghetti.

  10. If we live long enough, we will all die of cancer. As we age we use up the protective end of the chromosome, the telomere which than allows damage to the genetic code itself. It is this damage that leads to the uncontrolled cellular growth we call “cancer”.

    So far the ticking of this biological clock has been beyond our control and thus our quest for immortality remains but a quest.

    From and old surgeon.

  11. Cancer, aside from childhood cancer, and self induced sun cancer, generally is indication that you managed to evade, avoid and persist actual causes of death. Genetic decay will eventually cause you cancer – should you be lucky enough to live long enough to experience it. Several above posts get it right. An increase of poor country cancer means they are doing something right by lowering causes of death that kill young people. The study uses a faulty implied premise (that cancer rates increasing is necessarily an indication that humanity is failing or that the onset of cancer isn’t inevitable given a lack of other causes of death)

    * for every 100 pregnancies, 65-70 end BEFORE birth OR abortion. But yeah, let’s focus on the 3 that were aborted, some of doing so save the mother’s life and many who would be scorned, ridiculed and suffer for the genetic disabilities they were born with. If you want to save the unborn, instead of championing a cause against people who you have no information about, wouldn’t it make more sense to champion the cause to increase the likelihood of normal situation pregnancies making it to term?

    *the surest way to lower murder rates is to fund education, after school programs and the local police, yet “some” politicians are quick to cut each before anything else.

Leave a Comment

Your email address will not be published. Required fields are marked *