Science as a Social Process

I recently posted about a study that corrects our (scientists') understanding of the effects of lithium treatment.  

 Cousins et al. Lithium, Gray Matter, and Magnetic Resonance Imaging Signal. Biological Psychiatry. 73(7):652-657, 2013.

One of the interesting things about this study is that it invalidates a number of previous findings that appeared to show increased gray matter in the brains of people with bipolar who take lithium.  This new study invalidates those findings by giving an alternative explanation for the apparent increase in gray matter size---that lithium ions an water together appear to have more volume when imaged---that is more parsimonious.  This alternative explanation not only explains the increased imaged volumes for people with bipolar who are taking lithium while also explaining why lithium takers without bipolar also showed an increase in volume.

While I have already discussed how this study influences our understanding of lithium, what I have not discussed is the scientific process that led to both the original erroneous hypotheses about lithium and also the recent correction to out understanding.  For lay people this may appear to be a scientific scandal when in fact this simply a part of the scientific method. 

Science is a social process.  What this means is that science is conducted by fallible people in a social environment.  We as scientists subscribe to some version of the scientific method which is a loose collection of techniques that help us design studies, collect data, analyze the data, and draw conclusions.  Every step of this process has the potential to introduce error, so the broader scientific community engages in publishing whereby the authors of a scientific study have their work vetted by reviewers, published, and then discussed both at conferences and sometimes with a series of follow-up articles if the community is interested enough in the original paper.

In this particular case, the original hypothesis that lithium acts to increase the gray matter of people with bipolar was sufficiently interesting to the community that many papers were generated on the subject.  This is the scientific method at work.

For example, one of the original volume studies discussed their findings with precise scientific language that hedges in precisely the ways that are needed for accuracy

Bearden et al. Greater Cortical Gray Matter Density in Lithium-Treated Patients with Bipolar Disorder. 62(1):7-16. 2007.

These brain maps are consistent with previous voxel-based morphometry reports of greater GMD in portions of the anterior limbic network in bipolar patients and suggest neurotrophic effects of lithium as a possible etiology of these neuroanatomic differences.

The emphasis is mine.  The reason that these authors use the emphasized words is because they do not actually know a mechanism that might link lithium use to changes in brain volume.  Their studies simply showed correlation, not causation.  Because of the precise language used by studies such as this one, their findings are not actually invalidated by the new Cousins et al. study---mearly reinterpreted in light of a more parsimonious hypothesis.

The lay public also learned about this thread of scientific inquiry, because there were news articles on the topic.  Unfortunately news articles tend to sensationalize by relating extreme conclusions even in the titles of the articles:

Lithium Builds Gray Matter in Bipolar Brains, UCLA Study Shows

This illustrates the danger of taking a scientific exchange out of context.  The hedging of Bearden et al. means their work is not invalidated but reinterpreted by Cousins et al.  However, the news article referenced above is entirely invalidated by the result of Cousins et al.  If any part of this story is scandalous, it the sensationalized reporting.

I hope everyone reading my account of this scientific exchange will understand that the hedging used by scientists is a necessary both for correctness and for future scientific inquiry.  I hope more people will respect the scientific method, will understand that there is no scientific scandal in this case.  And I fervently hope that people will refrain from sensationalizing the results of studies.