Cancer study has major limitations

Photo Credit:  U.S. Department of Agriculture Photo Credit: U.S. Department of Agriculture

When you look at all the factors that may be confounding the study published by Bradbury et al. in the British Journal of Cancer, broad generalizations about cancer incidence decreases due to organic food consumption are not possible. The study compares middle age women who self-report that they eat organic foods and those who don’t on incidence of various cancers.  While the authors conclude that, of the cancers they examined only the risk of developing non-Hodgkin lymphoma was decreased by consuming organic food, the study itself has some shortcomings that make these conclusions unsound.  For example, the study survey answers were self-reported and not controlled for accuracy or consistency, the construction of the study did not include many important external factors that could play a role in cancer development, and the short timescale of the study does not allow cancer associations to be examined.

Because this is a cohort or prospective study, rather than a clinical trial, it means that the findings gathered were self-reported by the subjects answering the surveys circulated to them, and not controlled.  The general categories that study participants reported on included eating organic “Never, Sometimes, Usually, or Always.”  Study participants were asked to self-identify how often they ate organic food, but their eating habits were never validated by a researcher so it is impossible to determine the accuracy of the reporting.  Additionally, these categories are subjective and could mean different things to different people.  There is no way to quantify the difference between “Usually” and “Sometimes”, so different people who eat the same amount of organic may end up answering the survey differently.  This lack of consistency would make finding significant results showing associations between amount of organic food consumption and cancer risk impossible.

Photo Credit:  Anthony Albright Photo Credit: Anthony Albright

This difficulty in finding significant results is also inherent in longitudinal epidemiological methods.  Without controlling for all the external variables it is not possible to look at a single factor’s effect on health outcomes.  For example, this study did not take into account the types of organic food that consumers were choosing.  This could alter the results of the study because each conventional food category may expose consumers to different types of pesticides with variable toxicity levels.  Additionally, researchers did not examine the levels of pesticide residue present on the conventional food consumed by study participants.  If some consumers were diligent about washing their food, or chose conventional products that by chance had fewer pesticide residues than others it would confound the study results.  There are also many external factors such as maternal age, air pollution exposure, possible water contamination, and stress level that were not taken into consideration which could play an important part in cancer development. 

One of the major areas that was missed by this study is the interplay between exposure to environmental pollutants, such as pesticides, and people’s individual genes.  This topic is extremely important, because consumers may have different health reactions to pesticide exposure based on their genetic predispositions.  For example, some people may be more susceptible to developing cancer from certain environmental contaminants than others.  Without taking these hereditary factors into account it would be difficult to draw general conclusions about how contaminants affect public health.

The short timescale of the research is another limitation that puts this study at a disadvantage when it comes to examining how organic food consumption affects cancer risk.  Authors admit they did not collect information on how long the women in the study who said they consume organic foods had been consuming organic food before baseline (Year Three of the study). How long one consumes organic foods could be an important consideration in cancer risk, especially because exposure to toxins such as pesticides at a young age, or even prenatally, could have a disproportionate effect on long-term health.  To make an accurate estimate of how organic food can alter the risk of developing cancer, data would need to be taken starting when the individuals were in utero.  In fact, new studies are showing that pesticides may have negative health effects on an even longer timescale.  Scientists have been examining a phenomenon called disease inheritance, where the effects of environmental chemical exposure may affect the health of future descendants, even if they are never directly exposed to pollutants. An example of this was recently described in a study that showed DDT exposure could increase rates of obesity several generations down the line. Researchers in the study exposed rats to DDT, and found that it caused an increase in dramatic fat accumulation and weight gain three generations later. 

Photo Credit:  U.S. Department of Agriculture Photo Credit: U.S. Department of Agriculture

Because of the many limitations of the study, it is no surprise that researchers did not find associations between organic consumption and most cancer risks.  What is interesting is that, despite all the study construct shortcomings, researchers were still able to find a link between consuming organic food and reducing one’s risk of developing non-Hodgkin lymphoma.  To show significance in the face of so much statistical noise means that the connection between avoiding this type of cancer and eating organic must be extremely strong.  I suggest that future studies should focus on the specific pesticides that are causing the risk of non-Hodgkin lymphoma to increase when consumers eat conventional food.  To look at the link between using organic food to decrease the risk of developing other forms of cancer, future studies should be constructed in ways that control for external, confounding factors and take long-term approaches to examining pesticide exposures, including in-utero and youth exposures. 

One example of a study that has been taking a long-term perspective on the effects pesticide exposures have on health is the CHAMACOS project by the Center for Environmental Research and Children’s Health (CERCH) at the University of California, Berkeley.  This project started in 1999 when CERCH enrolled 601 pregnant women living in the Salinas Valley into the CHAMACOS study.  They have continued to follow the children, and in 2010 enrolled 300 additional 9-year olds into the cohort.  Their findings have been quite poignant, and often find significant links between health issues and exposure to environmental contaminants.  For example, they found that children of mothers with high exposure to organophosphate pesticides are at greater risk for neurodevelopmental problems at every stage that CHAMACOS researchers checked on the children.  Keep an eye open for more of their studies as the children in their study continue to grow.

In conclusion, if you are interested in decreasing your cancer risk it is still a good idea to avoid pesticide exposure by choosing organic.  The President’s Cancer Panel released in May 2010 exhorted consumers to choose food grown without pesticides or chemical fertilizers, antibiotics, and growth hormones to help decrease their exposure to environmental chemicals that can increase their risk of contracting cancer. That advice still holds.           



health, pesticides, cancer