Validation of screening questions and symptom coherence of night eating in the Swedish Twin Registry

Abstract

BackgroundScreening criteria have been used to estimate the prevalence of night eating syndrome (NES), but no validation studies have been conducted. MethodWe examined the validity of two screening questions for NES using a structured interview with adults enrolled in the Swedish Twin Study of Adults: Genes and Environment (STAGE) study. We also examined the coherence of the proposed diagnostic criteria for NES. A total of 416 participants (men=179; women=237) completed the interview and reported complete data for study inclusion. ResultsThe following values were calculated for the screening items after confirmation by interview for men and women, respectively: positive predictive value=.66 and .67, negative predictive value=.52 and .45, sensitivity=.62 and .63, and specificity=.56 and .50. As increasingly stringent diagnostic criteria were applied to the sample, prevalence of NES dropped. Nocturnal ingestions were more likely to co-occur with other NES symptoms than evening hyperphagia, which occurred frequently, but often in isolation; women were more likely to report a co-occurrence of symptoms than men, who were not likely to report distress related to NES symptoms. ConclusionIn sum, almost two-thirds of participants were correctly identified as having NES (without considering distress or impairment) if they answered positively on the screening questions, and about half of the participants were correctly identified as not having NES when answering negatively on the screening questions. Although self-report questions are somewhat informative, a structured interview remains the gold standard for diagnosing NES.