HealthDay has published a news article about a recent study that demonstrated an inverse association between high-density lipoprotein (HDL) and colon cancer. The brief article can be found here:
HDL/Colon Cancer article
The study used data from the EPIC trial, which is a European trial investigating the relationship between cancer and lifestyle, environment, and nutrition. The current study is known as a nested case-control study, which means previously collected cohort data (the EPIC trial) is used to generate a cohort-of-interest (certain blood levels of HDL, insulin, etc) and a matched cohort (control group). A relationship between the variable of interest and cancer is determined by an incidence rate ratio (RR). The RR is the ratio of the rate of developing the disease (colon cancer) in the experimental group (elevated HDL) to the rate in the control group (normal HDL); an RR greater than 1 indicates an increased rate of disease, less than 1 indicates a decreased rate, and 0 indicates no change in rate. Statistics are then used to quantify the relationship between the variables and ensure that the RRs are not simply an artifact of random chance.
The HealthDay article cites that "for each 16.6 milligrams per deciliter (mg/dL) increase in HDL and 32 mg/dL increase in apoA, the risk of colon cancer was cut by 22 percent and 18 percent, respectively." And this risk reduction was independent of other risk factors for colon cancer.
The authors of the study propose that the HDL particle itself may directly reduce the risk of colon cancer, possibly through an anti-inflammatory mechanism. However, this is speculation. It could be that high HDL is simply an indicator of an anti-colon cancer lifestyle or metabolic profile outside of what was controlled for.
HDL concentration is generally elevated in healthier people, and is manipulated by obesity, smoking, exercise, and other factors, including diet. The authors controlled for meat, fruit, vegetables, calories, and fiber. However, they didn't control for carbohydrate consumption, nor the most offensive carbohydrate - sugar. This is particularly interesting because HDL has been described as a marker for dietary carbohydrate (Sacks 2009), that is, the less carbohydrate that is consumed, the higher the concentration of HDL. Therefore, this data could suggest that decreased dietary carbohydrate, and therefore replacement by more protein, fat, or both, could have decreased the rate of colon cancer.
However, suggesting that carbohydrate reduction would reduce the risk of colon cancer is as speculative as suggesting that the HDL per se reduces the risk of colon cancer. This is just a good exercise to demonstrate how observational data can easily be used to speculate or support a theory. And since there is essentially no end to what one can speculate, perhaps observational data should be reported sans speculation.