Shahar Madjar, MD

How do scientists and doctors know that smoking is a bad habit?

In my previous two articles, I examined the question: Is beef good for you? My goal wasn’t only to tell you whether red meat is good or bad (it turned out that beef isn’t as bad as many people tend to believe) but also to describe how scientists conduct scientific research. I wanted to tell you how scientists think.

The conclusion I presented in my last article regarding red meat relied on observational studies. I wrote that scientists are careful when they interpret the results of observational studies. Why?

An observational study isn’t an experiment. The researchers do not interfere with the study population. They do not propose or suggest any behavioral change or a new medication. They just observe and report what they have learned. A good observational study typically relies on a large group of participants. The data may be collected at one point in time, or over a long period. The goal is to find a correlation–a type of discernible association between a specific exposure and a health outcome, say beef consumption and mortality, or smoking and lung cancer. But does this correlation indicate a cause-and-effect relationship? Does beef consumption lead to premature death? Does smoking cause cancer?

The problem with observational studies is in the interpretation given to them. People, even those who are supposed to understand how science works, tend to translate ‘correlation’ to ‘causation.’ Taken far enough, the tendency of the human mind to look for explanations can result in ridiculous claims. One attempt to demonstrate how one should be cautious when interpreting observational studies is a publication that showed a correlation between chocolate consumption and the number of Nobel laureates in different countries. The correlation was strong–the more cocoa-flavanols a nation consumed, the more Nobel laureates it produced. Should you feed your kids more chocolate to increase their chances of winning a Nobel prize? Of course not.

This isn’t to say that observational studies are worthless. Today, I will give you an example of how scientists and doctors use observational studies to draw almost definite conclusions with high levels of confidence. A case in point is the association between cigarette smoking and lung cancer.

The evidence linking smoking to lung cancer began to emerge in the 1950s. A study in the British Medical Journal showed that the incidence of lung cancer began to dramatically and consistently increase 15 years after an increase in the consumption of tobacco and cigarettes. Similar observations were repeatedly made in different geographical locations and at different points in time. Later, researchers found that there is a clear association between both the duration and the severity of smoking and the rate of death from lung cancer. In 1984, one group of scientists wrote: “…. we can say that the death rate of cigarette smokers from cancer of the lung has been thirteen times the rate of non-smokers, and that the death rate of heavy cigarette smokers has been over thirty times the rate of non-smokers … “

As evidence from different observational studies continues to accumulate, there comes a point where researchers start to consider the evidence to be convincing enough to indicate a cause-and-effect relationship. They often use a set of 9 different criteria (called Bradford Hill criteria for causation) to think about such y. When the information from different observational studies is robust, coherent, and consistent, a cause-and-effect state of affairs can be declared. This process may take years, sometimes decades to complete.

Such was the case with the correlation between smoking and lung cancer. Finally, a point in time arrived when several factors played together in harmony: Scientists indicated that death from lung cancer comes after exposure to tobacco and smoking (and not at the same time); that the association between smoking and lung cancer is strong and consistent; and that the risk of death increases with higher exposure to tobacco and cigarettes. Scientists were then able to conclude with utmost (although not absolute) certainty that smoking causes lung cancer.

The evidence was strong enough that aiming at absolute scientific proof, one obtained by conducting a controlled study where one group of participants is instructed to smoke heavily while the other group is advised to avoid smoking, would prove highly unethical.

When faced with a new research study that suggests you should make a change in your behavior–a new fad diet, a new exercise routine, a secret for longevity and happiness–try to weigh the quality of the information and whether the conclusion is indeed based on consistent evidence.

EDITOR’S NOTE: Shahar Madjar, MD, MBA, is a urologist and an author. He practices at Schoolcraft Memorial Hospital in Manistique, and in Baraga County Memorial Hospital in L’Anse. Find his books on Amazon or contact him at smadjar@yahoo.com.