Quote:
Originally Posted by NorthPeace
For a summary of the relationship between food and cancer, check out Food, Nutrition, Physical Activity, and the Prevention of Cancer:a Global Perspective. There is a chart on page 370 that summarizes the whole report.
Consumption of whole plant foods: nonstarchy vegetables; allium vegetables; fruits; and foods containing fibre, folate, or lycopene offer "probable decreased risk" of several types of cancer.
Consumption of foods or supplements that contain selenium offer probable decreased risk.
Consumption of red meat and processed meat offers convincing increased risk for colon cancer.
The only animal product that decreases cancer risk is mother's milk, and possibly milk from other animals (some increased, some decreased depending on type of cancer).
There was not enough evidence for reviewers to make conclusions about other animal sources, but there seemed to be a slight positive from fish and eggs, and no effect either way from poultry.
Calorie dense foods, including fast foods, are a risk factor for obesity, and foods with low calorie density (e.g. vegetables) reduce risk.
Obesity is a convincing risk factor for many kinds of cancer, so if low carbing helps you to lose weight, then the weight loss is helping to reduce the risk.
So if this collection of metastudies was my only source of info, and I wanted to low carb, I would take in fish, poultry and eggs, and consume a lot of nonstarchy vegetables, and some low glycemic fruit. I would eliminate processed foods, red meat, cured meat, sugary foods, doughnuts and french fries.
|
I went to the site. Didn't get into the metastudy, but did get into a monster of a report on the methodology used in the metastudy, and the criteria by which various studies were weighted:
Quoted from the SLR Specification Manual of the World Cancer Research Fund;
"For the reviews usually addressed by Cochrane 2 and NHS CRD1, the question is
generally of efficacy of interventions. In this context the currently generally accepted
hierarchy of evidence is used, which places randomised controlled trials above
observational evidence because they are
LESS OPEN TO BIAS. This is entirely appropriate.
However the questions at the heart of the systematic reviews to be commissioned by
WCRF International are aetiological – that is they are seeking to identify causes of
cancer. For aetiological questions, the inference of causation must be based upon
evidence of different types and drawn from different sources – observational,
intervention, clinical and laboratory, in order to provide a basis for considering the
conceptual frameworks of Bradford Hill 3 and others 4.
Thus the process outlined in this manual aims to conduct a comprehensive review of
all types of evidence relating to the question of relevance – using an inclusive
approach rather than a hierarchy to access the data."
In other words, mere association with cancer in observational studies is given equal weight with more solid types of evidence. I guess bias matters less when your looking for a dietary scapegoat than when you're looking for a cure.