A Haircut for the Heart?

Years ago, I attended a lecture in which the speaker used a wonderfully simple example to explain the difference between association and causation. Studies may observe that people with more gray hair also experience more cardiovascular disease and more heart attacks. The association may be perfectly genuine, but should we conclude that gray hair causes heart disease? If so, perhaps the next great advance in preventive cardiology would not be a medication, a healthier diet, or regular exercise, but a monthly appointment with the barber.

Of course, cutting or coloring gray hair does not make the heart younger. Gray hair and cardiovascular disease travel together largely because both become more common with age. The hair is not driving the cardiovascular risk. It is simply another passenger in the same vehicle, while aging is behind the wheel.

This distinction seems obvious when the example becomes slightly absurd, yet the same error appears throughout medicine. Researchers observe two things occurring together, and almost immediately, one is accused of causing the other. An association becomes a possible explanation, the possible explanation becomes a confident conclusion, and the conclusion becomes a headline, sometimes before anyone has established what is actually causing what.

Suppose a study finds that people who take more medications have worse health outcomes. We could conclude that medications make people sick and recommend throwing away the pillbox. A more sensible possibility is that people with more medical conditions usually require more medications. The number of pills may reflect the burden of illness rather than create it. Some medications can certainly cause harm, but the association alone cannot tell us which explanation is correct.

Hospital care offers another useful example. People who spend more time in hospitals generally have poorer health than those who rarely enter one. However, avoiding the hospital does not necessarily produce excellent health, just as entering it does not automatically cause illness. Hospitals collect sick people for the same reason fire engines gather around burning buildings. We would not reduce fires by banning fire engines, although the two appear together with remarkable consistency.

Even an ordinary walking cane can lead us into the same trap. If people who use canes fall more often than people who do not, we might be tempted to blame the cane. Yet the cane may simply identify people who already have impaired balance, weakness, arthritis, or neurological disease. Taking it away might not reduce their risk and could make the situation considerably worse. The visible object attracts our attention, while the underlying condition remains quietly in the background.

None of this means that associations are useless. Quite the opposite: many important medical discoveries begin with an observed association. The association raises a question, but it does not complete the answer. We must then ask whether the relationship remains after accounting for age, underlying illness, medications, lifestyle, socioeconomic conditions, and other hidden influences. We also look for a plausible biological mechanism, consistency across studies, the correct sequence in time, and, whenever possible, evidence that changing the suspected cause actually changes the outcome.

The deeper lesson is not that we should distrust medical research, but that we should interpret it with patience and proportion. An association may reveal a cause, reflect the effects of a cause, arise from a shared underlying factor, or simply accompany something else. Good science does not dismiss the observation, but neither does it ask the observation to prove more than it can.

Until then, we should continue caring for cardiovascular risk through thoughtful evaluation, healthy habits, and evidence-based medicine. Anyone who wishes to color the gray hair may certainly do so, but the heart, unfortunately, will know the difference.