[Skip to Navigation]
Sign In
July 2016

Errors in Data

JAMA Psychiatry. 2016;73(7):759. doi:10.1001/jamapsychiatry.2016.1117

In the article titled “Multisystem Component Phenotypes of Bipolar Disorder for Genetic Investigations of Extended Pedigrees,”1 published online in JAMA Psychiatry on February 12, 2014, and in the April 2014 issue of the journal, 17 of 527 neuroimages had been misidentified. After these data were corrected, the estimates of heritability and association with bipolar I disorder have changed. Seven of 126 brain measures that had been considered significantly heritable in the previous analysis were no longer significantly heritable in the corrected data set. Of 169 phenotypes investigated, 119 (70%) were significantly heritable vs the previous erroneous report of 126 (75%); and 51 (30%) were associated with bipolar I disorder vs the previous erroneous report of 53 (31%). In addition, 82% of the neuroimaging markers were heritable (not 88% as previously reported). These errors affect some data in the Abstract; Results and Discussion sections of the text; Figures 1, 2, and 3; and eTable 1 in the Supplement. The primary conclusions of the original article are unchanged.2 This article was corrected online. This article was also corrected on April 3, 2014, for an omission in Funding/Support.

Fears  SC, Service  SK, Kremeyer  B,  et al.  Multisystem component phenotypes of bipolar disorder for genetic investigations of extended pedigrees.  JAMA Psychiatry. 2014;71(4):375-387.PubMedGoogle ScholarCrossref
Fears  SC, Freimer  NB, Bearden  CE.  Errors in identification of 17 of 527 brain images in genetic study of phenotypes associated with bipolar disorder  [published online May 18, 2016].  JAMA Psychiatry. doi:10.1001/jamapsychiatry.2016.1051.Google Scholar