IntroductionRecent recommendations for the more widespread prescription of statin drugs in the U.S. have generated controversy. Cholesterol is commonly thought to be the enemy of good health. Higher cholesterol has been alleged to be associated with pathologies, such as cardiovascular disease, not with higher survival.
However, there is a history of research showing correlation between low serum cholesterol and cancer in general [1,2], as well as colon cancer in 8006 men [3].
Cancer (without regard to type) has been shown to be especially prevalent in the lowest cohort of serum cholesterol in this very large study: 361,662 men aged 35 to 57 years, reported in JAMA: "Mortality follow-up revealed a significant excess of cancer in the lowest decile of serum cholesterol level during the early years of the follow-up, which attenuated over time" [4].
Cholesterol is the main known substrate in the body’s production of Vitamin D, specifically cholecalciferol, as well as steroid hormones, some of which have been shown to have anti-cancer effect.
We compared total serum cholesterol (TC) in cancer survivors vs cancer fatalities, and we assess the value of deliberately lowering TC among cancer patients.
We also examined
diet in the survivors as well as those who then died of cancer.
MethodsIn this original, previously unpublished research, we conducted a double-blind retrospective case series, in which we looked back at data from all 255 cancer patients who came to and were treated by our clinic with either current dietary information, based on a self-chosen
diet, and/or a recent (last six months) serum TC level, measured by an unaffiliated laboratory or an unaffiliated clinic over the previous seven years, comparing TC in the surviving cancer patients versus those cancer patients who died during that same time.
The 255 patients include those that survived to the present (192) and those that died from all causes (63). Some patients did not die of cancer. A few died of surgical complications. One had a myocardial infarction while hiking on a mountain. Some cases of precipitous morbidity and mortality were closely following rounds of chemotherapy. Those now deceased were more likely to have cholesterol measured than survivors, due to our ever more urgent search in their last months for any information that could possibly help them. Therefore, the deceased are over-represented in the cohort of 255 patients with the data of interest.
FindingsSurviving cancer patients had 24 points higher mean total cholesterol than the mean for deceased cancer patients (Table 1).
Some patients were found to be at the extreme ends of the distribution. Of those now deceased, outlying data was found at the extreme lower end of TC values, whereas of the survivors, there were some values at the extreme high end.
So we then decided to look at TC values of 50 < x < 300. This excluded four of the sickest patients our clinic has ever encountered, two of them with a cholesterol of 3, one with a cholesterol of 28 and one with a cholesterol of 36, the last two confined to wheelchairs. We also excluded one with a cholesterol of 428 (who frequently hikes mountains); another with TC = 397, who gardens for hours and climbs ladders at age 84, and another with TC = 308, an active person.
Limiting the field this way, we had the following averages:
We used Inductive Rule Extraction in order to observe any correlations that may exist in the data. Data analysis was performed using SPSS™ statistical analysis software. HIPAA compliance was followed to ensure patient anonymity associated with the data used in the bio-statistical analysis. Standard Bivariate Correlation Analysis was used to compare the data, and standard levels of statistical significance were determined that are typically used and reported in biological systems (e.g. p < .05). Analysis of the data is reported as Bivariate Comparison, Pearson Correlation, Significance, and the number of "n" used in the comparison. The variation in "n" is due to some "missing values" associated with the some of the bivariate comparisons (Table 2).
The following Table 3 shows the results of these comparisons:
Comparing
diet and cholesterol, there is a higher mean level of cholesterol with the omnivore (OM)
diet (189) compared to the Vegan (Vegan) and Vegetarian (Veg)
diets (168 and 167 respectively).
The results in Table 4 may be skewed by the phenomenon of cancer patients arriving to treatment at our clinic in various stages of cancer. Those who were relatively healthier may have had the luxury of choosing a vegan diet, whereas those who were worse off may have tried to obtain sustenance and calories any way they could. Therefore, it is conceivable that some of the most ill individuals may have chosen an omnivorous diet.
Comparing cholesterol with dietThere is a significant correlation of cholesterol level and diet (omnivore). One possible reason is that in the omnivore, dietary consumption of cholesterol can be significantly high. However, high consumption of carbohydrates may cause hypercholesterolemia in vegan and vegetarian
diets as well.
There is a correlation between lower cholesterol levels and higher stages of cancerThe higher the cancer stage the significantly lower the cholesterol level. This raises the question of why lower cholesterol would be correlated with lower survival rate. More detailed research has to be done to elucidate the "cholesterol factor" and cancer survival. If it is not diet specific, then there may be a uniquely physiological phenomenon in cancer pathologies.
Observations of survival and diet
We may ask: if there is a significant correlation between survival and high cholesterol, is there also a significant correlation between survival and diet? Dietary differences between cancer survivors and those who later died of cancer were also found to be notable. If we look at probability of survival, then the omnivores were significantly more likely to survive than the other two groups.
Table 5 shows that among these cancer patients omnivores were 3.6 times more likely to survive than to die. But vegetarians were only 2.1 times more likely to survive than to die, and vegans were only 1.6 times more likely to survive than to die.
Statistical Details
Shown from Table 6 to 12.
Conclusion
These results suggest that considerably more research has to be done with regard to cholesterol levels,
diet and cancer survivorship to determine if there is a cause and effect relationship that may be used by the physician to increase survival rates during the treatment process. It is possible that patients with Type II Familial Hypercholesterolemia (dyslipidemia) taking statin type medication may have diminished Vitamin D production and endocrine function reduction, which may be very problematic in cancer patient survival. In the meantime, caution should be used before prescription of statin drugs to cancer patients, or before
insisting on a vegan
diet with cancer patients.
GLOSSARY of statistical terms
Co-variance: How much do two random variables change together? Answer:
where E is the expected value or mean.
Pearson's Correlation Co-efficient: Covariance of two variables divided by the product of their standard deviations.
Null Hypothesis: Two different variables are probably not correlated, assumed to be not, till proven correlated. i.e. "Aspirin has no proven effect on headache."
Statistical significance is the probability that an effect is not likely due to chance alone. This can cast doubt on the null hypothesis.
Two-Tailed test: a way of computing the statistical significance of two extreme ends of a
bell curve P-Value: probability of obtaining a test statistic at least as extreme as the one that was actually observed.