By Robert Verkerk PhD

Founder, executive & scientific director

A US study by two scientists at Johns Hopkins University just published in the journal Science says that two-thirds of the variation in tissue cancer risk (cancer types, not cancer cases) is down to ‘bad luck’. Only one-third, the scientists claim, is attributable to lifestyle and genes.

Unsurprisingly, this finding has turned into something of a media circus, with newspapers, radio and TV broadcasts around the world suggesting to the public that there is little it can do to reduce their risk of cancer – the number one or two killer disease in most Western societies.

It’s been big news to some extent because this view flies in the face of a huge body of existing research that suggests the exact opposite: that there is a lot an individual can do through the modification of diet, lifestyle and behaviour to reduce his or her risk of cancer — and even that of the next generation.

The authors of the paper, biostatistician Dr Cristian Tomasetti and oncologist Professor Bert Vogelstein, suggest that more emphasis should be placed on early detection and secondary prevention, to increase the chance of successful cancer treatment.

While there are merits in the study, we think there’s a major problem with the way it’s been interpreted, especially by the media. We also question some of the conclusions and highlight reasons why the study findings might represent a significant over-estimate of the role of chance in the development of many cases of cancer.

The key findings of the study

Among the key take-home points from Professor Vogelstein and Dr Cristian Tomasetti study are:

  • All cancers are the result of 3 factors: chance (or bad luck), lifestyle (including environment) and genes
  • Of 31 cancers studied, the authors found that two-thirds were caused by bad luck i.e., random mutations arising during DNA replication in normal, noncancerous stem cells
  • The authors found that some cancers e.g. bone, small intestine and some types of leukaemia, appeared to be more affected by chance than others. Conversely, lung cancer was very strongly related to lifestyle, smoking in particular.
  • The authors stress that if chance is the most important factor affecting risk of most cancers, more effort should be spent on secondary prevention or early detection, on the basis that this will increase the chance of successful treatment

Media skew

The implications of the findings in the paper have been widely misreported in the media. The UK’s Daily Telegraph headline declared “Most cancers are caused by bad luck not genes or lifestyle, say scientists”. India Today announced “Cancer caused by bad luck, not unhealthy lifestyle or inherited genes: Study”. The New York Times lead with “Cancer’s Random Assault”. And io9 pronounced “Math Suggests Most Cancers Are Caused By "Bad Luck.”

The resounding message being issued is that Tomasetti and Vogelstein’s findings now turn the tables on previous wisdom. Diet and lifestyle choices are being claimed as having a relatively insignificant affect on whether or not someone develops cancer, and secondary prevention (early diagnosis and screening) is the way forward. Previously, it has been argued that primary prevention, not secondary prevention, should be prioritised. This argument was forcefully made by Vineis and Wild in the Lancet just last year, who state “primary prevention is the most effective way of fighting cancer.”

Accepting Tomasetti and Vogelstein’s findings at face value could have untoward consequences by causing the public to continue high-risk behaviours including sedentary behaviour, smoking, excess alcohol consumption, diets low in fruit and vegetables and over-eating.

In the face of the balance of evidence, we also think the conclusions are wrong! Here’s why.

How the role of chance has been over-estimated

There are a number of reasons that suggest that the importance of genetic make-up, lifestyle, environment — and interactions between all of these — may have been under-estimated in terms of their combined role in cancer risk, including the risk of dying from cancer. Some of the authors’ findings have already been challenged, including by relatively conservative bodies, such as Cancer Research UK.

Among the most important reasons why we feel Tomasetti and Vogelstein’s findings might represent a significant over-estimate of the role of chance in cancer risk are:

  • Tomasetti and Vogelstein looked at 31 cancers and their mathematical analysis does not reflect the number of cancer cases, but rather the types of cancer. For example, in the UK, over half the incidence of cancer is linked to just 4 types of cancer, namely breast, lung, prostate and bowel.
  • Even cancers that have not been clearly linked to environment or lifestyle and where genetic risk factors have been identified, such as breast and prostate, have been shown to be influenced significantly by environment (e.g., exposure to xenoestrogens, even prenatally, diet) or lifestyle. In many cases, emerging epidemiological and (epi-)genetic research is showing stronger, not weaker, linkages between cancer, genes and lifestyle/environment. As this is a rapidly emerging field, we must be very careful about how we interpret older data, including those incorporated into Tomasetti and Vogelstein’s analysis.

  • The authors are looking at factors that affect the development and incidence of the 31 types of cancer studied. They are not looking at the factors that affect the survival of patients with those cancers. It is well-known among integrative medicine practitioners that changes in diets and lifestyles following cancer diagnosis can significantly influence survival rates, as well as reducing adverse effects including death) of chemo- or radio-therapy.
  • Tomasetti and Vogelstein’s conclusions are based, obviously, only on data available to them. In most cases, evaluation of environmental exposures from, for example, occupational or industrial chemicals through to foodstuffs or chemicals formed as a result of high temperature cooking, has been extremely limited. We are also learning that it may be the exposure to chemical mixtures that is more important than exposure to individual chemical agents, and this is rarely taken into account in studies.
  • The same can be said for analysis of genetic influences on cancer. Studies of genetic variations (single nucleotide polymorphisms) have only ramped up in the last decade (following the completion of the Human Genome project in 2003); by and large, most studies on polymorphisms affecting cancer risk have focused on a limited numbers of polymorphisms and have yet to address complex interactions between these polymorphisms.
  • Adding further to this complexity is the fact that the environment alters the expression of genes, primarily through DNA methylation and histone modifications. This all relates to the key and rapidly emerging area of epigenetics. We are just beginning to understand the role of environment on gene expression, including affects that pre-date birth and ones that are inherited from previous generations. While we lack a proper and complete understanding of these interactions, cancer risk will appear more likely to be the result of chance. Let’s not forget that chance, by definition, relates to occurrences that are unexpected in the absence of any obvious cause. It follows, then, if or when an explanation for that event is found, it would no longer be viewed as having occurred by chance.

Screening – yes or no?

We then feel it necessary to apply some major words of caution around one of the key recommendations made by the Tomasetti and Vogelstein: the need to prioritise secondary prevention or earlier detection of cancer (screening). The idea is that screening of cancers increases the chance of them being nipped in the bud early; cancer found at an earlier stage is deemed more treatable. This is certainly the case with some cancers, but it isn’t with others, including the predominant cancer in women, breast cancer. Breast cancer screening using mammography may lead to more harm than good through over-diagnosis and over-treatment.

You can find out more about this important issue by watching "The Promise" documentary.

Equally, there are not dissimilar problems with over-treatment and over-diagnosis with prostate cancer, the most common cancer in men. This is admirably shown in Peter Starr’s documentary, which includes 56 interviews with doctors, entitled Surviving Prostate Cancer without Surgery, Drugs or Radiation.

The current reality

Coming back to the main thrust of Tomasetti and Vogelstein’s findings, it is crucial to appreciate that where data on environmental or genetic influences on cancer are deficient, incomplete or missing — which is in the vast majority of studies — variations in cancer risk will appear to be down to chance, or bad luck. The reality is, until more complete studies are available, we cannot be sure if luck or complex gene-environment interactions are involved.

As time moves forward, our understanding of the epigenetics of cancer will continue to increase. Given recent trends, we speculate that we will find, in years to come, that the currently held view that estimates that up to 50% of cancers are preventable, will be shown to be a significant under-estimate.

Changes in our understanding about the mechanisms of cancer in the body will also likely change our approach to dealing with it. Currently, malignant tumours are viewed somewhat like deadly parasites that need to be cut out of our bodies or poisoned with deadly chemicals or radiation. Less commonly tumour development is seen as a disruption of the immune system. Even less commonly is it appreciated that individuals have a great capacity to influence how our immune, neurological and endocrine systems, in concert with our microbiome, with appropriate support, are able to manage or overcome cancer in the body.

We also tend to make many assumptions that affect our approach to cancer. Some of these assumptions may be shown to be invalid in due course. One of them is that people who have never been diagnosed as cancer patients have always been free of malignant cancers. This is a big assumption. It’s perhaps one of the reasons oncologists get such a surprise when they witness so-called “spontaneous regression”, sometimes after a patient modifies one or more aspects of his or her diet or lifestyle.

We uphold that one of the fundamental keys to improving our chances of surviving cancer is in developing a better understanding both of our genetic and metabolic background and ways in which our environment affects gene expression, cell metabolism and function and complexity of self-regulating (homeostatic) mechanisms required for healthy function. We have the capacity to alter our behaviour in ways that modify this gene-environment interaction to our benefit. From what we know this far, modifying our diets and our lifestyles represents the most profound tool available to us for this purpose.

While I've argued here that Tomasetti and Vogelstein may have over-estimated the role of chance in most cases of cancer, this doesn't devalue the authors’ findings. What's important is that the findings are interpreted with extreme caution, especially when offering clinical treatments or dietary and lifestyle advice, or when recommending screening.

More information:

Preventive Medicine Research Institute:

Cancer Prevention Coalition:

Environmental Working Group:

Cancer prevention guidelines, Cancer Research UK:


ANH-Europe Homepage