I attended the California Breast Cancer Research Program (CBCRP) conference on Monday in San Francisco.  The first presenter, Robert Hiatt, presented a model of breast cancer  (BC) causation that they had created, funded by CBCRP. The model looks at factors affecting causation of postmenopausal BC.  They are now working on a similar model for premenopausal BC.

Lets take a closer look at their model, shown above.

The model highlights the multilevel nature of causation of BC.  It focuses on 4 domains: biological (yellow), behavioral (green), sociocultural (orange), and physical/environmental (blue).

You can see that the lines on the model vary – some are solid, some are dashes, some are thick, some are thin.  All these indicate the strength of the evidence for causation and also the quality of the data. Check the key on the left hand side for an explanation.

It is an interactive model, so I recommend you go to New Model of Breast Cancer Causation and try clicking on different factors and looking at their connections and the strength/quality of evidence.

You can see at the top right hand side, it gives definitions of the factors, and if you click on one of the actual lines, it will give the reference paper citation that was used for evidence of this link/association.  In the example below, I clicked on just ‘endogenous hormones’ to investigate their effect on causation, and then clicked on the thick yellow line from ‘endogenous hormones’ to ‘age at menarche’, which gave the reference for that connection.


The model is based on only research from human studies.  In the premenopausal BC model, the researchers are looking at how to also incorporate animal study data into the model. 

If you are interested in reading how they developed this model, check the link at the bottom of this post for their paper “A multilevel model of post menopausal breast cancer” [1]. In it, you can read about how they utilized a transdisciplinary team of experts from many fields to come up with the model.

I think that this model gives a good overall picture of causation and while some factors may not be included, it clearly demonstrates the multifactorial and complex nature of BC causation.  For researchers, policy makers, and practitioners, it illustrates how focusing on just one area, may be missing the bigger picture. The use of a transdisciplinary approach may be more appropriate – with cell biologists,  psychologists, social scientists, environmental scientists, etc all involved – rather than the usual single discipline approach of most research. It is also useful in that it highlights gaps in knowledge for further research areas where there is little or weak evidence.

A key factor in this type of model is keeping it current, with results from new research. Its value will depend on its being a continuously updated resource that people can rely on.

What do you think about it?  Click on some of the factors and see what links surprise you.

[1]. A multilevel model of post menopausal breast cancer.


Print Friendly