Psychiatry · Mol Psychiatry · 2016
Summary: more of an enzyme in bipolar patients’ cerebellum. The paper’s own data: a third less.
More or less enzyme with bipolar disorder? What each part of the paper says
The paper says
In brains examined after death, the paper measured IDH3A, an enzyme cells use to make energy, in the cerebellum (a brain region) of 15 people with bipolar disorder and 15 without. Its summary, the part most people read, says the bipolar group had “higher” levels.
Its own data show
The paper’s own table gives the average measured amount of IDH3A in the cerebellum: 0.804 in people with the illness, 1.227 in people without, about a third less (34.5%). The sections reporting and discussing the results say the same three times: “lower”, “lower”, “reduced”. Only the summary’s one word is reversed.
A paper’s summary is often the only part people read, so readers are told the opposite of what this study found. The flipped word even contradicts the paper’s own argument: that too little of this enzyme may help cause bipolar disorder, and that the enzyme could be a target for new treatments. We found no published correction.