Risk: The Science and Politics of Fear – Chapter 10, The Chemistry of Fear

In Chapter 10 of “Risk: The Science and Politics of Fear” Dan Gardner sets out the commonly underestimated risk factors for cancer and the commonly overestimated risk factors. Many of the analyses were confounded by other factors of risk of cancer. When they are adjusted, some factors remain strong (e.g., individual habits) and others disappear (e.g., pollutants).

The same fact take the sting out of the statement Carson thought was so important she put it in italics: ‘Today, more American school children die of cancer than from any other disease.’ By 1963, traditional child killers such as diphteria had been wiped out. More children were dying of cancer than any other disease not because huge numbers of children were dying of cancer but because huge numbers of children were not dying of other diseases. (page 266)

But as age is the primary risk factor for cancer the fact that far more people were surviving childhood and living old age would inevitably mean that more people would get cancer – mostly when they were old – and so the ‘lifetime risk’ would rise. … Data on cancer were still sketchy in that era but in the previous two decades there was an apparent 200 per cent rise in the incidence of cancer among women and a 600 pe cent rise among men, which was mostly the result of a rise in only one type of cancer. Lung cancer ‘is the only form of cancer which shows so definite a tendency,’ the report noted. (page 266)

‘Exposure to pollutants in occupational, community, and other settings is thought to account for a relatively small percentage of cancer deaths,’ says the American Cancer Society in Cancer Facts and Figures, 2006. Of those, occupational exposures – workers in aluminum smelters, miners who dug asbestos under the unsafe conditions of the past – are by far the biggest category, responsible for perhaps four per cent of all cancer. The ACS estimates that only 2 per cent of all cancers are the result of exposure to ‘man-made and naturally occurring’ environmental pollutants – a massive category that includes everything from naturally occurring radon gas to industrial emissions to car exhaust. (page 269)

It’s critical to understand that not all carcinogenic chemicals in the environment are man-made. Far from it. To take just one example, countless plants produce carcinogenic chemicals as defences against insects and other predators and so our food is positively riddled with natural carcinogens. They are in coffee, carrots, celery, nuts, and a long, long list of other produce. Bruce Ames, a leading cancer scientist at the University of California at Berkeley, estimates that ‘of all dietary pesticides people eat, 99.99 per cent are natural’ and half of all chemicals tested – synthetic and natural – cause cancer in high-dose lab animal experiments. (page 270)

Major health organizations agree that traces of synthetic chemicals in the environment are not a large risk factor. What is hugely important is lifestyle. Smoking, drinking, diet, obesity, and exercise: These things make an enormous difference – by most estimates, accounting for 65 per cent of all cancers. (page 270)

‘You’re talking about studying humans and there are a million confounders. A study will say this and another will say the opposite’ … Ames cites the example of a controversy in California’s Contra Costa county. ‘There are a lot of refineries and there is more lung cancer. Ah, the refineries are causing the lung cancer. But who lives around refineries? Poor people. And who smokes more? Poor people. And when you correct for smoking, there’s no extra risk in the county.’ (page 275)

The media, in pursuit of the dramatic story, are another contributor to prevailing fears about chemicals. Robert Lichter and Stanley Rothman scoured stories about cancer appearing in the American media between 1972 and 1992 and found that tobacco was only the second-most mentioned cause of cancer – and it was a distant second. Man-made chemicals came first. Third was food additives. Number 6 was pollution, 7 radiation, 9 pesticides and 12 was dietary choices. Natural chemicals came 16th. Dead last on the list of 25 – mentioned in only nine stories – was the most important factor: aging. … The [American Institute of Cancer Research] noted with regret that only 49 per cent of Americans identified a diet low in fruits and vegetables as a cause of cancer; 46 per cent said the same of obesity; 37 per cent, alcohol; and 36 per cent, diets high in red meat. But 71 per cent said pesticides residues on food cause cancer. ‘There’s a disconnect between public fears and scientific fact,’ said an AICR spokesperson. (pages 277-278)

… cancer is primarily a disease of aging, a fact which has a profound effect on cancer statistics. The rate of cancer deaths in Florida, for example, is almost three times higher than in Alaska, which looks extremely important until you factor in Florida’s much older population. ‘When the cancer death rates for Florida and Alaska are age-adjusted,’ notes a report from the American Cancer Society, ‘they are almost identical.’ (page 281)

Consider chlorine. Treat drinking water with it and it creates by-products that have been shown to cause cancer in lab animals in high doses and may increase the cancer risk of people who drink the water. There’s even some epidemiological evidence that suggests the risk is more than hypothetical. So the precautionary principle would suggest we stop putting chlorine in drinking water. But what happens if we do that? ‘If you take the chlorine out of the drinking water, as was done in South America, you end up with an epidemic of 2,000 cases of cholera,’ says Daniel Krewski. And cholera is far from the only threat. There are many other water-borne diseases, including typhoid fever, a common killer until the addition of chlorine to drinking water all but wiped it out in the developed world early in the 20th century. (page 287)

If pesticides were banned, agricultural yields would decline. Fruits and vegetables can reduce the risk of cancer if we eat enough of them, which most people do not do even now. And so banning pesticides in order to reduce exposure to carcinogens could potentially result in more people getting cancer. (page 288)

It is undoubtedly true that, as Dan Gardner frequently noted throughout his book, media tend to exaggerate the severity of events, especially when the promotion of fear helps to make huge profits. But the current situation is certainly bleaker than what Gardner would have us believe. In Chapter 7, “Fear Inc.”, the author notes a crucial detail :

On current trends, more people would get cancer than ever, and more would die from it. But what the ad didn’t mention is that this is because the population is growing – more people means more cancer – and it is aging, which means more cancer because aging is by far the biggest risk factor for cancer. The ad also failed to note that the death rate from cancer is falling and expected to fall further, nor did it mention that the incidence rates of most types of cancer – after taking population aging into account – are flat or falling. (pages 174-175)

Here’s the problem. Obesity is expanding worldwide, and obesity is easily responsible for 30% of cancers. And not to mention other health risks caused by obesity (Adams et al., 2006, Figure 1 & 2 , pp. 776-777). Nevertheless, the discovery by Adams et al. (2006) that a BMI of 25, relative to a BMI of 20, reduces the risk of mortality was strongly contested by Zajacova et al. (2011, figure 1) since they found a relationship between BMI and mortality that is monotonous and not U-shaped. The reason for finding curvilinear relationship could be due to confounding owing to preexisting disease (sometimes referred to as reverse causality) or by sarcopenia (loss of lean body mass). For instance, low BMI may be a consequence of weight loss from disease processes leading to death.

Obesity Trends in the U.S. [1985 – 2006]*


Finucane et al. (2011) show that most countries have an upward trend in obesity.




The prospective study of Calle et al. (2003), involving more than 900,000 people, showed that obesity has been longitudinally and linearly associated with death for all cancers (esophageal cancer, stomach cancer, colorectal cancer, liver cancer, gallbladder cancer, pancreatic cancer, breast, cervical and ovarian cancers, prostate cancer, kidney cancer, non-Hodgkin’s lymphoma, multiple myeloma, and leukemia). To deal with confoundings, they controlled for other risk factors such as age, race, smoking status, education, physical activity, marital status, current aspirin use, fat consumption, vegetable consumption and status with respect to estrogen-replacement therapy in women. Restricting the sample to those who had never smoked makes the relative risk even stronger for some cancers. For men and women aged 50, the proportion of all death from cancer attributable to overweight and obesity in the U.S. is 14% and 20% respectively.

Obesity accounts for 21 percent of medical care costs

The Cornell study reports that an obese person incurs medical costs that are $2,741 higher (in 2005 dollars) than if they were not obese. Nationwide, that translates into $190.2 billion per year, or 20.6 percent of national health expenditures. The study appeared in the January issue of the Journal of Health Economics (31:1). Previous estimates had pegged the cost of obesity at $85.7 billion, or 9.1 percent of national health expenditures.

Obesity represents a big cost for medical care and, interestingly, it seems to affect more people of lower classes, and consequently, of lower IQ, even if the link between low-SES (or medium-SES) and obesity still seems valid only among white populations, while the link between low-SES and obesity among black american populations has now moved from positive to negative; and it seems to vary significantly over the decades (Wang, 2011, p. 27; Wang & Zhang, 2006, Figure 3; Wang & Beydoun, 2007, Figure 3).

Trend over time in obesity is not necessarily inconsistent with genetic factors associated with obesity. Yang et al. (2007) reviewed the literature on the heritability of obesity. Twin, adoption, and family studies establish that the heritability is high, 16-85 percent for BMI, 37-81 percent for waist circumference (WC), 6-30 percent for WHR, 35-63 percent for percentage body fat.

At first glance, it seems strange that Dan Gardner says “the incidence rates of most types of cancer – after taking population aging into account – are flat or falling”. It takes years, decades for healthy cells to evolve into cancerous cells, proliferate and lead to a formation of a tumor, due to poor hygiene of life for a very long time : lack of fruits and vegetables, alcohol, tobacco, BMI, etc. It is also said that inactivity is a risk factor for cancer, but there are surely confounding factors such as obesity or malnutrition (not to be confused with under-nutrition). Most cancers detected involve the elderly. If we exclude the elderly, we effectively remove the prevalence of cancer. But this is fallacious procedure. However, if we adjust by regression, the result can be interpreted as meaning that, at any given age, the cancer rate is constant or falling. What the above sentence implies is that dying at younger age reduces the detection of cancer. At the same time, it is difficult to believe that the (age-adjusted) cancer rate really has diminished, given that some indicators of bad habits (smoking, obesity, bad food) showed a great upward trend over time.


  1. Adams, K. F., Schatzkin, A., Harris, T. B., Kipnis, V., Mouw, T., Ballard-Barbash, R., … & Leitzmann, M. F. (2006). Overweight, obesity, and mortality in a large prospective cohort of persons 50 to 71 years old. New England Journal of Medicine, 355(8), 763-778.
  2. Calle, E. E., Rodriguez, C., Walker-Thurmond, K., & Thun, M. J. (2003). Overweight, obesity, and mortality from cancer in a prospectively studied cohort of US adults. New England Journal of Medicine, 348(17), 1625-1638.
  3. Finucane, M. M., Stevens, G. A., Cowan, M. J., Danaei, G., Lin, J. K., Paciorek, C. J., … & Ezzati, M. (2011). National, regional, and global trends in body-mass index since 1980: systematic analysis of health examination surveys and epidemiological studies with 960 country-years and 9· 1 million participants. The Lancet, 377(9765), 557-567.
  4. Wang, Y. (2011). Disparities in pediatric obesity in the United States. Advances in Nutrition: An International Review Journal, 2(1), 23-31.
  5. Wang, Y., & Beydoun, M. A. (2007). The obesity epidemic in the United States—gender, age, socioeconomic, racial/ethnic, and geographic characteristics: a systematic review and meta-regression analysis. Epidemiologic reviews, 29(1), 6-28.
  6. Wang, Y., & Zhang, Q. (2006). Are American children and adolescents of low socioeconomic status at increased risk of obesity? Changes in the association between overweight and family income between 1971 and 2002. The American journal of clinical nutrition, 84(4), 707-716.
  7. Yang, W., Kelly, T., & He, J. (2007). Genetic epidemiology of obesity. Epidemiologic reviews, 29(1), 49-61.
  8. Zajacova, A., Dowd, J. B., & Burgard, S. A. (2011). Overweight adults may have the lowest mortality—do they have the best health?. American journal of epidemiology, kwq382.
This entry was posted in Society and tagged . Bookmark the permalink.