Saying Chinese Herbs Cause Cancer Equals Bad Science

Recently the article Alternative Medicine Warning: Some Chinese Herbs May Increase Cancer Risk came to my attention. The Fox News report reviews an academic article published by the Journal of the National Cancer Institute. Having more than a passing interest in Chinese herbal medicine, I chased down the original article, Population-Based Case-Control Study of Chinese Herbal Products Containing Aristolochic Acid and Urinary Tract Cancer Risk (Lai, Wang, et. all). The authors conclude, "Consumption of aristolochic acid-containing Chinese herbal products [mu tong] is associated with an increased risk of cancer of the urinary tract in a dose-dependent manner that is independent of arsenic exposure."

A long time student of mathematics and a more recent student of public health and biostatistics, I set about critically analyzing this paper. I have two observations importance of biostatistics in public health management that cause me to personally call to question the validity of the authors' conclusion.

Which herb are they talking about?

In traditional Chinese herbal medicine, there are two different herbs that are colloquially referred to as mu tong. Guan mu tong, or aristolochia, contains aristolochic acid, a known nephrotoxin. In modern practice, this herb is rarely used and is banned from importation to the United States by the Food & Drug Administration (FDA). Chuan mu tong, or caulis mutong, on the other hand is not known to contain aristolochic acid. Chuan mu tong is effectively used to treat urinary tract infections and promote lactation in cases of blocked milk ducts.

The authors of the study make no distinction in their report if they are reviewing cases of people using chuan mu tong or guan mu tong, though considering their interest in aristolochic acid, I would guess they are studying the potentially carcinogenic effects of guan mu tong. Considering that guan mu tong is neither legal nor widely available for use, I question why Fox news took such an interest in this publication initially.

Confounding variables

In stastical analysis, a confounding variable is something that relates to both the dependent and independent variable. Study methodologies need to control for confounding variables in order to avoid making invalid inferences about the relationship of the variables being studied.

For example, suppose a study wants to determine if performing regular weight-bearing exercise is an effective way to prevent osteoporosis in post-menopausal women. Women are selected for the study and categorized as either doing the exercise or not. But suppose also that the women in the weight-bearing exercise also regularly take a calcium supplement, which is known to prevent osteoporosis. Without taking this fact into consideration and making appropriate statistical adjustments, analysis might incorrectly conclude that the weight-bearing exercise had a greater role in preventing osteoporosis than it actually did.

In the study concerning mu tong, the authors considered and accommodated for the presence of significant levels of arsenic in the subjects' drinking water (arsenic is known to cause urinary tract cancer). However, the authors site in their limitations section that smoking status was not considered during analysis. Personally, I have reviewed many articles regarding cancer and possible causative agents and I cannot think of one that was considered authoritative in which smoking tobacco products, a well established carcinogenic habit, was not considered and controlled for.

Personally, I am appalled to think that a combination of poorly designed scientific research and irresponsible journalism could lead people to the conclusion that Chinese herbs cause cancer. Herbs, like all things we ingest, need to be scrutinized for their quality and effects, but to shut the door on all Chinese herbs is a terrible mistake.

The Fox news report as well as the original study paper are both available for review online

Jeanine Adinaro is a licensed acupuncturist (Texas) and CEO of Third Coast Herb Company, Inc., a Texas company specializing in quality controlled Chinese herbal remedies, and manufacturer of Herbalogic concentrated herb drops. Jeanine also holds an MS degree in theoretical mathematics and has done graduate work in public health at the University of Texas.