The Truth is Slipping Away

However you define it, the truth implies a connection to reality that can be tested. It’s true that helium is lighter than air and that water boils at 212 degrees Fahrenheit. Science depends on being able to revisit those facts over and over without getting strangely ambiguous results. Yet things are shifting in stranger ways than anyone ever expected, as one discovers in an eye-opening article that appeared around Christmas in the New Yorker magazine. Everyone who is interested in how truth works should read Jonah Lehrer’s troubling “The Truth Is Wearing Off,” which can still be found online.

What Lehrer is primarily concerned with is replicability, the term scientists use for repeating an experiment and arriving at the same result. Certainly the most important findings in science have been repeated many times over. Not necessarily. Some results, particularly in medicine, are not holding up at all. This “decline effect” forms the central mystery, because no accepted reason has been found for why a treatment should suddenly begin to dwindle in its effectiveness. Lehrer cites three prominent examples: antipsychotic drugs, hormone replacement therapy for women past menopause, and the use of aspirin to prevent heart attacks. In all three, the treatments are still widely endorsed in the medical literature, ignoring the fact that the decline effect is in full swing, meaning that the original results expected from these treatments are simply not there anymore or have declined to a fraction of what they once were.

Lehrer offers detailed, cogent reasons for why this mystery can’t be solved. A big part is that drug companies don’t want to hear about bad results, which leads to the suppression of negative studies and the boosting of positive ones. Another reason is faddism: scientists are quck to jump on the bandwagon of a new discovery. But there are cases where a researcher is brave enough to go public and admit that he himself couldn’t replicate his own original findings — this takes courage when your whole career is based on that discovery.

For me, the most distressing aspect of the decline effect is how widely it is being ignored, not just in medicine but everywhere that bad news is unwelcome. Is any physicist going to welcome a well-funded complex study, recently published, that found discrepancies in gravity, of all things? but it is medicine that touches most people’s lives most closely in science. A 2005 review article in the prestigious Journal of the American Medical Association examined the forty-nine most cited articles in leading medical journals. Lehrer writes, “…of the thirty-four claims that had been subject to replication, forty-one per cent had either been directly contradicted or had their effect sizes significantly downgraded.”

If that isn’t troubling enough, there is the huge problem, also widely ignored, of results that get accepted without being replicated either enough or at all. For example, there has been a widespread fad for claiming that genetic differences between men and women account for differing risks in acquiring disorders as various as schizophrenia and high blood pressure. Yet a probe of the underlying research found serious flaws in the vast majority of the studies. And worse was to come: “…out of four hundred and thirty-two claims, only a single one was consistently replicable.” One! Yet textbooks have been written citing these studies, and just as the ordinary layman still believes in taking aspirin every day to prevent a heart attack or undergoing hormone replacement therapy, so do doctors and medical students who are relying on false, misleading, or outdated information.

I am not offering an answer about the mystery of the decline effect, and I have no ax to grind, even though it needs to be common knowledge when drugs stop working — the debunking of the science behind popular antidepressants was just one recent example. Ultimately, our addiction as a society to silver bullets makes us vulnerable to wishful thinking, so it is public demand that is fueling errant science. The decline effect, according to Lehrer, is much more widespread than anyone wants to admit. Be that as it may, science must be acknowledged for its many, vast strides. but as long as people ignore prevention and positive lifestyle changes as the best approach to health, they will be forced to fall back upon drug therapies and the glaring drawbacks involved in that.

Read more
Read more

Published by The San Francisco Chronicle