Don’t be fooled by the study which found lower cancer rates in vegetarians

I saw this morning reports on a study which has found links between vegetarianism and a reduced risk of certain cancers. The study, which assess nutritional habits and cancer risk in more than 61,000 men and women in the UK found that compared to meat-eaters, vegetarians had a 53, 45 and 74 per cent reduced risk in bladder, blood and lymph system (eg. leukaemia and lymphoma) and stomach cancers respectively [1]. Looking at all cancers combined, risk reduction was found to be 12 per cent. Something tells me that those keen on the vegetarian way will wave this study around as yet more ‘proof’ that the vegetarian diet is healthiest for us.

I do not dismiss this evidence out of hand, but I think it is important to bear in mind that this study was epidemiological in nature, and can only then tell us about associations between diet and cancer. But just because two things are associated does not mean one is causing the other. The usual assumption with studies of this nature is that there’s something bad about meat that ups cancer risk. But it may be vegetables are cancer protective and vegetarians eat more of these. Or maybe it’s neither of these things. Or maybe, it’s nothing more than an association and eating meat/fewer vegetables does not cause cancer at all. We just don’t know.

One major problem with epidemiological studies is what are known as ‘confounding’ factors. So let’s say we find from a study less physically active people turn out to have a higher risk of lung cancer. We have found an association between two things that might cause people to conclude that a sedentary lifestyle causes lung cancer. But, imagine that sedentary people are more likely to smoke. Smoking is a potential confounding factor here, and may be the real reason for why sedentary individuals are at heightened risk of lung cancer. In epidemiological studies often an attempt is made to ‘control for’ potential confounding factors. It’s an imprecise science, for sure, but generally better than nothing.

Controlling for potential confounding factors is particularly important when comparing meat-eaters and vegetarians because, generally speaking, they are likely to be more health-conscious than meat-eaters. Meat-eating has an unhealthy reputation, right? And vegetarianism generally as a ‘healthy’ image too. So someone inclined to do what they can to protect or improve their health may be drawn towards eschewing meat. But these individuals may also be drawn to other habits too like smoking less and exercising more.

The study focused on here did control for a variety of potential confounding factors including age, smoking, alcohol intake, body mass index and physical activity level. But another way to level the playing field might be to attempt to compare vegetarians and non-vegetarians who have been matched for a similar level of health consciousness.

In one study, researchers attempted to counteract any confounding factors by focusing only on individuals who shopped in health food stores. The idea here is that all of these individuals are generally ‘health-conscious’, whether they are vegetarian or not. This allows, some would argue, a fairer appraisal of the impact of vegetarian or non-vegetarian eating.

This study did not focus on risk of individual conditions, but a much better marker for overall health: overall risk of death. This study found that compared to the general population, death rates in vegetarians and non-vegetarians were significantly lower than in the general population (which supports the notion that health food shoppers are a generally health-conscious bunch). Crucially, though, overall risk of death in vegetarians and non-vegetarians the same [2].

In another study, vegetarians were asked to recruit their friends and family into the study. Doing this was thought to help ensure that all individuals in the study were similarly health-conscious. Again, death rates for vegetarians and non-vegetarians were essentially the same [3].

In May this year, a study was published which assessed total mortality in a large group of vegetarians and non-vegetarians [4]. Confounding factors controlled for were age, sex, smoking, and alcohol consumption. The result? No difference in risk of death between vegerarians and non-vegetarians. Risk of death from heart disease was the same too.

I read here that Professor Tim Key, the lead author of the study being reported on today said it was impossible to draw strong conclusions from this one single study. He is quoted as saying: “At the moment these findings are not strong enough to ask for particularly large changes in the diets of people following an average balanced diet.”

Even the study’s lead author seems to have poured cold water on its findings. Perhaps Professor Key is aware of the evidence which shows vegetarianism does not offer distinct health advantages? He should be, seeing as he’s the lead author of the studies mentioned above [2,4] that show that vegetarians have mortality rates the same as non-vegetarians.


1. Key TJ, et al. Cancer incidence in British vegetarians. Br J Cancer. 2009 Jun 16. [Epub ahead of print]

2. Key TJA, et al. Dietary habits and mortality in 11000 vegetarians and health conscious people: results of a 17 year follow up. Br Med J 1996;313:775″9

3. Thorogood M, et al. Risk of death from cancer and ischemic heart disease in meat and non-meat eaters. BMJ 1994;308:1667″70

4. Key TJ, et al. Mortality in British vegetarians: results from the European Prospective Investigation into Cancer and Nutrition (EPIC-Oxford). Am J Clin Nutr. 2009;89(5):1613S-1619S

13 Responses to Don’t be fooled by the study which found lower cancer rates in vegetarians

  1. Marsha 1 July 2009 at 11:04 pm #

    It’s impossible to know what the research means without reading the primary research reports. You are right that researchers need to control for confounding or covarying variables. But your suggestion to only look at people who are “health conscious” risks restricting the range of overall variability – and you won’t get any relationships at all.

    Also the classifications of “meat eaters” and “vegetarians” are artificial. Better to measure either the amount of meat (or fish) the subjects eat or the amount of animal protein the subjects eat. Furthermore, vegetarians or vegans have usually not been so since birth – so they may have lived many years as meat-eaters.

    This does not invalidate all the research; it only makes it difficult to assess by looking at secondary research reports.

  2. Ted Hutchinson 2 July 2009 at 9:28 am #

    My main concern with this report is that the classification of “meat eater” contains a lot of different types of meat eater.
    There is a huge difference between those relying on processed meat from industrially made sausage, pies, ready meals, and those buying meat only from free range, grass finished animals direct from the farm also eating organ meat regularly.
    We may be just looking at the effect of a high omega 6 intake compared with a high omega 3 intake.
    In the same way the classification of vegetarian includes the full range of those existing solely on industrially processed vegetarian meals to those making their own meals from organically grown fresh vegetables straight from the garden.

  3. Sherry 3 July 2009 at 9:54 am #

    Surely we all have exactly the same Risk of Mortality in that we are all going to die, so what exactly does that mean? Some people die young and others linger on for many sad years in nursing homes.
    Virtually no-one will have followed the same dietary plan for the whole of their lives.
    These studies make headlines, but seem to have very little substance, other than an axe to grind.
    I think that a vegetarian lifestyle must be so boring and fraught with flatulence and prohibitions,that it would just seem to be longer. Give me a nice steak any day, I`ll take my chances!

  4. Dr John Briffa 3 July 2009 at 10:00 am #


    ‘Risk of mortality’ refers to risk of death over a given period of time.
    I’m with you on your sentiments regarding steak.

  5. Liz 4 July 2009 at 1:49 pm #

    ‘Meat eater’ is too general a term. Wild/organic lean meat and oily fish is a different kettle of fish (sorry) than pies and sausage rolls.
    Also if people eat meat they may eat fewer vegetables. Is the change in risk associated with the type of meat? The quantity of meat? Or is it that meat eaters consume far fewer vegetables. So if you could get the meat eaters to eat the same quantity of veggies as the vegetarians would this change the results?
    We should probably all eat more veg and don’t need to eat as much meat, but I don’t think this makes the case for giving up meat altogether.

  6. Paul Anderson 8 July 2009 at 5:28 pm #

    With reductions of 53, 45 and 74 per cent for bladder, blood and lymph cancers respectively, yet an ovreall reduction in cancer risk of only 12 per cent, it follows that figures for some cancers must inevitably be much lower. In fact there must be some cancers where the risk must be higher in vegetarians than in meat eaters. The break down of the figures for all forms of cnacer would make interesting reading.

    My suspicion is that meat per se isn’t the problem but that processes foods in some shape or form are – not all meats are equal just as with carbs and fats. It might also not be the inclusion of meats in the diet that is the problem but the inclusion or exclusion of something else.

    The study has highlighted the biggest reduction in risks – fair enough, but some explanation of the disparity between those high percentages and the much lower reduced risk for all forms of cancer requires an explanation. The report might also have said ” vegetarians show reduced risks for x,y and z cancers and increased risk for a, b and c. This presumably might help to identify other factors at play – one of which may be statins or low cholesterol levels for example.


  7. Trinkwasser 12 July 2009 at 8:51 pm #

    Agree with the others. meat can be an indicator for a crap diet full of processed foodlike substances or an indicator for a primal diet full of healthy Real Foods, which will undoubtedly have differing health consequences.

    I’m all for a vegetarian diet so long as you add plenty of meat and fish . . .

  8. Hilda Glickman 17 July 2009 at 8:19 am #

    It all depends on what the vegetarian eats. A vegetarian can live on doughnuts and coca cola. I knew one who never ate fruit or veg while many meat eaters eat these in abundance. Hilda Glickman

  9. James H 19 July 2009 at 8:15 pm #

    It wasn’t long ago that most people thought vegetarians would die young or be unhealthy.


  1. New cancer research shows vegetarians at less risk than meat eaters - 3 July 2009

    […] Correlation is not the same as causation. That’s basically what Dr. Briffa said on his blog: Don’t be fooled by the study which found lower cancer rates in vegetarians | Dr Briffa’s BlogI do not dismiss this evidence out of hand, but I think it is important to bear in mind that this […]

  2. Vegetarian Diet Can ‘Cut Risk Of Cancer by 45%’ « purehealthclinic news - 8 July 2009

    […] Tim Key of Cancer Research UK at Oxford University also sounded caution and said they are not sure whether it is the protective effect of eating more vegetables or something about the meat & called for more research. John Briffa also gave an interesting response about whether it was the fact that people didn’t eat meat, or that non-meat eaters tend to look after themselves more in terms of not smoking, drinking too much and getting exercise and fluids.  Here’s a link to what he had to say:… […]

  3. Canibais e Reis » Blog Archive » Gorduras saturadas associadas a menor mortalidade cardiovascular, pelo menos é o que se verifica na Europa - 11 July 2009

    […] não tenha muita crença nestes estudos, que são, por regra, resultado de epidemiologia muito rebuscada e tendenciosa. Quanto à associação entre gorduras totais e mortalidade cardiovascular, ela […]

  4. Canibais e Reis » Blog Archive » Conferência da NUTRISCIENCE com o Dr. Frits Muskiet foi espectacular, mais um momento de alta ciência em Portugal - 20 March 2010

    […] tem menos cancro que carnívoros e vegetarianos, e o mais provável é que vegetarianos só têm menos cancro que carnívoros pelo estilo de vida que seguem, e não propriamente por não consumirem carne, e/ou então pelas carnes […]

  5. Canibais e Reis » Blog Archive » Pingo Doce propõe-nos este prato tão tradicional e típico da Dieta Mediterrânica, “caçarola de seitan com legumes”, dá para acreditar nisto, o vegetarianismo/ seitanismo já chegou à Dieta Me - 26 May 2010

    […] que não é excessivo dizer que não servem para nada. Sobre este estudo, veja o comentário do Dr. Briffa. E não existe nenhum estudo de intervenção mostrando superioridade do […]

  6. Canibais e Reis » Blog Archive » RAW FOOD SOS sobre o China Study, o estudo epidemiológico de suporte “científico” às dietas vegetarianas/ veganas debaixo de fortes críticas - 10 July 2010

    […] myth (Dr. Eades) Meat and mortality (Dr. Eades) Vegetarians AGE faster (Dr. Eades) Don’t be fooled by the study which found lower cancer rates in vegetarians (Dr. Briffa) Does eating meat really increase our risk of colon cancer? (Dr. […]

Leave a Reply