Skip to main content

Familial Risk for Bipolar Disorder is Associated With Decreased Neurocognitive Performance in Unaffected Relatives: A Preliminary Report from the Amish-Mennonite Bipolar Genetics (AMBiGen) Project

Wednesday, September 12, 2018 — Poster Session I

12:00 p.m. – 1:30 p.m.
FAES Terrace


  • EK Besançon
  • F Lopes
  • L Wille
  • M Mallmann
  • L Kassem
  • FJ McMahon


BACKGROUND: Psychiatric nosology based only on symptom measures has proven to lack strong correspondence to underlying genetic risk factors, leading to renewed interest in performance-based assessments, such as neurocognitive tests. However, performance deficits may be a consequence of psychiatric illness or its treatment. Family-based designs provide one approach to this confound, since unaffected relatives share genetic (and other) risk factors with their affected kin that may manifest as performance deficits in the absence of psychiatric symptoms. Here we report preliminary findings on neurocognitive performance assessed within large families ascertained from Amish and Mennonite founder populations METHODS: A neurocognitive battery including six assessments covering general intelligence, working memory, executive functioning, and emotion recognition was administered to an initial sample of 85 probands diagnosed with bipolar disorder or related conditions (BD) and 238 of their unaffected relatives. All probands were assessed with the Diagnostic Interview for Genetic Studies (v 4); relatives were screened for psychopathology with self-report measures. Neurocognitive performance was compared between cases and unaffected individuals, and among unaffected relatives grouped by relatedness to the proband (105 first-, 25 second-, and 108 third-degree relatives). Age, sex, and education were included as covariates. Data points lying more than 3 SD from the mean were excluded. All data were analyzed using SPSS Version 22. Heritability of the neurocognitive parameters was analyzed using SOLAR v 8.3.2. RESULTS: As expected, probands scored more poorly than unaffected individuals on several measures, including digit-symbol (p=0.009), DANVA errors (p=0.007), and Trails A (p

Category: Genetics and Genomics