Total Pageviews

Sunday, May 20, 2012

Embracing uncertainty is the defining feature of Medicine and of all good Science

A front-page headline in yesterday's New York Times announces an apology to the gay community authored by an aging giant in the field of Psychiatry.

Dr. Robert L. Spitzer has retracted the findings of a study that he published in 2003 which evaluated the effect of "reparative therapy," a technique aimed at "reorienting" gay individuals to heterosexuality.  Although the methodology of this retrospective inquiry was criticized as deeply flawed by his peers, the key finding - that reparative therapy seemed to be effective in "curing" homosexuality in some cases - influenced the national conversation on gay rights in ways that were offensive and damaging to  the gay community.

For me, this retraction is important and relevant.  Most obviously, because it is welcome news to a minority that, despite progress on many fronts, still faces societal discrimination, especially from the religious right-wing. At another level, however, it says something about the way good Science, and good scientific Medicine work.

Despite the widespread use of the term "scientific certainty" to denote significant surety,  this use of language reflects a misunderstanding of the scientific method and the nature of scientific knowledge.  The opposite of "Science" is superstition, while the sine qua non of Science is what philosopher Carl Popper called "falsifiability."  (More on what this means next time.)  Suffice it to say that falsifiability means, among other things, that a scientific claim is, by definition, one which can be overturned with evidence or, in many cases, with a more careful analysis of existing evidence.

Which in turn means that the process is self-correcting in a way that dogmatic thinking is not. This may unsettle individuals who perceive that the house of Medicine is like a pinball, ricocheting incessantly from one position to another.  The better metaphor, however, is probably that of a seagoing vessel guided by a skilled but invisible captain who counters the currents and shifting winds, bringing her direction through small but frequent corrections closer and closer to the desired course.  That course refers, of course, to verifiable truth; in the present case it means a clearer understanding of sexual orientation, not a concept retrofitted to accommodate someone's ideology but one arrived at through the collaborative scientific method, and for the benefit of all, regardless of health status, race, gender or sexual orientation.


Saturday, May 5, 2012

Autism, Obesity and the Metabolic Syndrome

AUTISM.  No childhood condition is more complicated or less understood.  We're not even sure how to talk about it.  Is it a "condition" at all, or more like a symptom pointing to other underlying problems?  And if the latter, what kind of problems? And perhaps most vexing of all: Is the prevalence of autism increasing?  The is very good reason to believe that some of the apparent increase is the result of better detection among mildly affected children; reclassification, or what some call "diagnostic substitution" seems to be at work as well.  But is there more to the story? We don't know.

One thing is clear.  We will never understand autism until we understand its causes.  And such causes appear to be numerous and diverse. Researchers are reasonably certain that autism has a strong genetic basis, but that "environmental" factors play an important role as well.  Does this kind of thinking make sense?  Skeptics argue that this is just a lot of mumbling among scientists, and a sophisticated way of avoiding the difficult admission that we are in the dark.  Before tuberculosis was found to be the result of infection with a special bacterium, "experts" blamed it on multiple causes, like climate and crowded conditions.  

However, autism is probably more like cancer than tuberculosis.  Meaning that it simply isn't a single condition, but a whole host of conditions, with separate causation, prevention, symptoms, outlooks and treatments - hence the "spectrum."  And these conditions may have very different causes.  This isn't good news of course - its much easier to tackle a single condition than a "category" of disorders.  

In the May issue of PEDIATRICS,  a team from UC Davis and Vanderbilt University report on their evaluation of more than five hundred children on the autistic spectrum between 2 and 5 years of age, comparing them with 172 others with developmental disorders and 315 age-matched controls.  A major purpose of the study was to see whether the parents of affected children were any more likely to suffer from "metabolic" conditions, including obesity, diabetes and high blood pressure.  They were.  For families with a child on the spectrum, the odds of a mother having one of these conditions was 1.6 times higher than for a "control" family.  What does this mean?

On the one hand, a case-control study like this needs to be interpreted cautiously.  Did the mother's diabetes or obesity cause autism?  Or are these conditions all the result of the same common (genetic) factor?  Other kinds of investigation will be needed to confirm the association and better work out the matter of causation.  On the other hand, these results may turn out to be very significant, especially given the rising incidence of type 2 diabetes and hypertension, which are both consequences of the obesity epidemic.  Additionally, some experts have suggested that an association between these conditions and autism may be explained as the adverse effect of elevated blood sugar, or "hyperglycemia."  We know that excess glucose causes a range of disturbances at the molecular level and that some of these could plausibly impact nervous system development in utero. This realization proves nothing, but "biological plausibility" is an important consideration for an candidate theory about autism. 

Finally, we must keep in mind that even if a causal relationship between metabolic conditions and autism is found, the majority of mothers of affected children are not hypertensive, diabetic or obese, suggesting that important as this association may be, it accounts at best for a fraction of children and families afflicted with this challenging developmental problem.