Detection of illness worry in the general population: A specific item on illness rumination improves the Whiteley Index

Publikation: Bidrag til tidsskrift/Konferencebidrag i tidsskrift /Bidrag til avisTidsskriftartikelForskningpeer review

Objective: The Whiteley Index (WI) is the most widely used screening tool for health anxiety/illness worry. Diverse versions (different number of items and factors) have been used. We aimed to examine psychometric properties of 7 items of the WI besides adding a new item on obsessive illness rumination for better future detection of health anxiety. Methods: Data from a large population-based study in Denmark (N = 9656). Construct validity was examined by exploratory (EFA) and confirmatory factor analysis (CFA) plus hypothesis testing. Criterion validity was evaluated via Receiver Operating Characteristic curves and area under the curve (AUC) using a diagnostic criterion as gold standard. Results: Factor loadings of EFA revealed viable one-factor models (6, 7, or 8 items) and two-factor models (7 or 8 items). Factor one indicated a dimension of illness worry. Factor two indicated a somatic symptoms dimension. The new item on obsessive illness rumination merged well with the existing items. EFA of two-factor models and one-factor 6-item model showed good fit. CFA resembles these findings. A one-factor 6-item model (including the item on obsessive illness rumination and excluding two items concerning somatic symptoms) was chosen as the optimal model and presented good criterion validity: AUC 0.88 (95%CI(0.84;0.92)). Main hypotheses concerning associations with somatic symptoms, anxiety, and depression were met. Conclusions: We found good psychometric properties for a new one-factor 6-item version of the WI. Through elimination of items concerning somatic symptoms and inclusion of obsessive illness rumination, we propose a clear, unidimensional and improved measure of illness worry: Whiteley-6-R.

OriginalsprogEngelsk
Artikelnummer110245
TidsskriftJournal of Psychosomatic Research
Vol/bind138
NummerNovember
Antal sider8
ISSN0022-3999
DOI
StatusUdgivet - 2020

Se relationer på Aarhus Universitet Citationsformater

ID: 197612563