Virginia Via Research Day Book 2026

Medical Student Research Public Health

03 ACCURACY OF FAMILY HISOTRY REPORTS OF MENTAL DISORDERS IN A COMMUNITY-BASED FAMILY STUDY

Kevin P. Conway, PhD 1 , Chaitanya S. Mishra, OMS-II 2 , Lihong Cui, MSc 1 Jackie Franco, MSc 1 , Kathleen R. Merikangas, PhD 1 Corresponding author: cmishra@vcom.edu

1 National Institutes of Health, Department of Genetic Epidemiology 2 VCOM-Virginia, Blacksburg, Virginia

examined factors associated with accuracy, adjusting for covari-ates. Results: Family informants demonstrated a high degree of agreement and accuracy for detecting mania and nic-otine dependence. Underreporting was common for most other disorders. Agreement and accuracy var ied by disorder as indicated by wide ranges for kappa (0.12 to 0.53), sensitivity (12% to 71%), specificity (73% to 99%), PPV (29% to 72%), and NPV (71% to 98%). Comorbidity in the index case, and to a lesser extent in the informant, was associated with lower agreement across disorders. Conclusions: Although the accuracy of family informants varies by mental and substance use disorder, family inform-ants can be relied on to accurately detect mania and nicotine dependence in their relatives. Findings underscore the value of informant reports, available at less cost than direct clinical interviews, for

iden-tifying individuals in need of treatment for these chronic and burdensome conditions. This research was supported by the Intramural Research Program of the National Institutes of Health (NIH), ZIA MH002953. The study was approved by the Combined Neuroscience Institutional Review Board at the NIH. All participants provided written informed consent.

Background: Obtaining psychiatric information from family informants is common in mental-health and substance-use research. Few studies include both alcohol and drug use disorders and research on nicotine de-pendence is lacking. This study examined the accuracy of, and factors associated with, family inform ants for specific mood disorders, anxiety disorders, and substance use disorders. Methods: Participants included 374 adult probands and 793 adult relatives from a community-based controlled family study of affective disorders. Direct clinical interviews assessed lifetime diagnoses for specific mood, anxiety, and substance use disorders according to Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision criteria. Informants completed family history interviews. Inform-ant accuracy was examined using kappa, sensitivity, specificity, positive predictive value (PPV), and neg-ative predictive value (NPV). Analyses

Table of Contents

18 Edward Via College of Osteopathic Medicine (VCOM)