Lung cancer: One joint = 20 cigarettes?

The lung cancer study was the scariest. Since cigarettes are a known lung cancer risk, it seems plausible that marijuana might carry similar risks. In fact, most of the scientific evidence tends in the opposite direction - though one would never know it from reading either the study or the Reuters wire story that got the heaviest circulation.

Conducted in New Zealand, this was what is called a "case-control" study, in which researchers looked at a group of patients who had lung cancer and compared them to a group without cancer - the controls - matched for age and other demographics. All were asked about various factors that might increase their lung cancer risk, including smoking cigarettes or marijuana. After running the data on 79 cancer cases and 324 controls through myriad equations and mathematical analyses, the researchers proclaimed that one joint packed a cancer risk roughly equal to 20 cigarettes - an assertion that became Reuters' lead.

What was downplayed in the study, published in the European Respiratory Journal, and missing entirely from most media reports was context - context that strongly suggests that its alarming conclusion is wrong.

For one thing, the new conflicts with other, much larger studies. In a study published in 1997, Kaiser-Permanente researchers followed 65,000 patients for 10 years and saw no sign of marijuana use increasing the risk of lung cancer or other smoking-related cancers. And a UCLA study similar in design to this one, published in 2006, found a trend toward lower lung cancer rates among marijuana smokers. Instead of 79 cancer cases, the UCLA team looked at 1,212. The result was so striking that they speculated that it "may reflect a protective effect of marijuana."

That's right: Marijuana might protect from cancer. Piles of published studies going back to the mid-1970s document the cancer-fighting properties of marijuana's active components, THC and other chemicals called cannabinoids. Anticancer activity has been shown in many types of malignant cells, including lung cancer cells. So even though marijuana smoke contains tars and other potentially carcinogenic compounds, it is entirely plausible that cannabinoids counter any harmful effects.

But even without such context, a closer look at the New Zealand data raises questions that should have been asked by reporters. For example, most marijuana smokers in the study actually didn't show an increased risk of cancer. The only group that did was those whose marijuana use equaled at least 10.5 "joint-years" (one joint-year equals smoking a joint every day for one year). That group constituted a whopping 14 people. All those complicated mathematical models leading to the "20 times the risk" assertion, and contradicting reams of published research, rest on exactly 14 people.