Colorectal cancer and cholesterol

Because I'm quite interested in colorectal cancer I've had this abstract lying around on my desktop for some time, awaiting full reading before "drag and drop" in to that enormous and chaotic folder labeled "Diet stuff".

OK, the usual caveats apply; it's observational and it's a bit of a data trawl from the Cardiovascular Health Study. Hmmm, this means that cardiologists are involved in the data acquisition. Bad.

On the plus side it's prospective and no drug company is involved. US government funded. It's also available in full text, which is nice because it means I can work through their results and see if their conclusions are derived from their findings.

You know how it is, you browse down these tables slotting yourself in to the various groups and... Ooooh, they measured cholesterol levels! Now there's a surprise. Did you notice any mention of cholesterol levels in the abstract? Particularly that killer LDL cholesterol?

No, neither did I.

LDL-C got its little section as the penultimate parameter in table 2. Let's read the table, the bit we want is down at the bottom. And the risk of colorectal cancer is:

LDL-C quartile and risk of colorectal cancer:

Lowest.......1.0 (reference value)
2nd............0.8
3rd.............0.6
Highest.......0.5

The absolute numbers were, from lowest to highest, 35/1447, 26/1416, 23/1467 and 16/1386 incident cancers.

The relationship is remarkably linear, p = 0.01

None of the other parameters (fasting glucose or glucose 2h post OGTT) did any better than p = 0.02. But they got in to the abstract. And in to the title too!

The LDL-C association did get a brief mention in the discussion:

"Surprisingly, a strong association between increased LDL levels and decreased risk of colorectal cancer was identified. The explanation for this finding is unclear."

That's it.

That's the total discussion of their most statistically significant finding. Possibly their most biologically significant finding.

The lowest quartile for men had LDL-C below 100mg/dl, women below 110mg/dl. Those were the ones most likely to get bowel cancer. I think it's worth pointing out that the average cardiologist would consider an LDL-C of 100mg/dl or above as a peracute statin deficiency. That lowest quartile is where your cardiologist wants you to be.

BTW I'm in the highest LDL-C group. Just as well, with my family history.

Just assuming, for a split second, that there is causality between low LDL-C and colorectal cancer, would you expect statins to trade heart attacks for cancer? Pravastatin perhaps? I posted the body counts here.

Anyone chosen between heart attack and bowel cancer yet?


The last author on this paper is Savage. He's surprised at the association. Why should he be? He helped write the guide to treating metabolic syndrome for the US government. He would be expected to have read the literature. Maybe.

Obviously he wouldn't read an Irish paper:

"A group of 114 Irish patients with primary adenocarcinoma of the large bowel had significantly lower serum cholesterol concentrations than an age and sex matched group drawn from the general population"

or a Brazilian paper:

"Our study suggests an association between low blood cholesterol and colorectal cancer"


or an Austrian paper:

"low cholesterol was significantly associated with all-cause mortality, showing significant associations with death through cancer, liver diseases, and mental diseases"


or a Japanese paper:

"The J-curve association was observed between average TC or LDL-C concentrations and total mortality. Malignancy was the most prevalent cause of death. The health of patients should be monitored closely when there is a remarkable decrease in TC and LDL-C concentrations with low-dose statin"

or even stuff from Honalulu:

"Only the group with low cholesterol concentration at both examinations had a significant association with mortality (risk ratio 1.64, 95% CI 1.13-2.36). INTERPRETATION: We have been unable to explain our results. These data cast doubt on the scientific justification for lowering cholesterol to very low concentrations (<4.65 mmol/L, 181mg/dl) in elderly people"

But, for crying out loud, you would have thought he would have memorised chunks of Keys as part of his cardiology training!

Keys:

"Among 477 cancer deaths five years after cholesterol measurement, there was a significant excess of lung cancer deaths in the bottom 20% of the cholesterol distributions in the populations"

How about NHANES1:

"The inverse cholesterol-cancer relation in men was present for cholesterol determinations made 6 or more years before diagnosis of cancer"

or even Framingham:

"Although the Framingham data are not conclusive, they do suggest that in some cancer cases where the serum cholesterol level was lower than that expected at as much as 16--18 years before cancer diagnosis, the depressed level was likely to be a precursor to the tumor growth"

Surprised is he?

That's enough. I was looking for information about glucose and cancer and yet again ended up grinding my teeth on the stupid cholesterol hypothesis and its followers.

I'll calm down now.

Peter