Development and preliminary validation of a modified Food Craving Inventory for Pregnancy (FCI-P) in U.S. military active-duty Service women
Katherine Thompson,Marney A White,Ruby Schrag,Jason M Lavender,Viviana Bauman,Autumn Mains,Stephanie Rioux,Holly Spinner,Isabel Thorstad,David Klein,Mark Haigney,Marian Tanofsky-Kraff
DOI: https://doi.org/10.1016/j.appet.2024.107677
IF: 5.016
2024-12-01
Appetite
Abstract:Food cravings during pregnancy are highly common, yet no measure of cravings has been validated among pregnant women. The current study evaluated the psychometric properties of the Food Craving Inventory (FCI) for use during pregnancy. U.S. military active-duty Service women (N = 192; 29.5 ± 3.8 years old; 44% Army, 36% Air Force, 15% Navy, and 4% Marine Corps) were recruited from the community at 12-27 weeks' gestation. Participants completed a modified version of the FCI validated for adults with binge-eating disorder, which included 13 additional items assessing cravings for foods that women commonly report experiencing during pregnancy (e.g., pickles, sour cream, hot or spicy wings). Additional measures also assessed disinhibited eating behaviors (i.e., loss of control eating and emotional eating). A series of confirmatory factor analyses were conducted to examine model fit for a four-factor structure of: (1) the FCI validated for binge-eating disorder (excluding the pregnancy-oriented food items) and (2) the FCI modified for pregnancy (with the pregnancy-oriented food items added). The previously validated four-factor structure of the FCI for binge-eating disorder demonstrated poor model fit in the current sample of pregnant women. After examining the structure of the FCI modified for pregnancy, several items were removed due to high cross-loading across multiple subscales. The resulting 16-item, four-factor (Fats, Sweets, Carbohydrates, Spicy/Strong foods) FCI for pregnancy (FCI-P) demonstrated generally good model fit (CFI = .95, TLI = .94, SRMR = .04, RMSEA = .09) and good-to-excellent internal consistency (Cronbach's alphas: .83-.96). Convergent validity was supported by significant correlations between the FCI-P scores and the disinhibited eating behavior scores (ps < .001). Results highlight the importance of psychometrically evaluating eating-related measures for use during pregnancy to appropriately capture the potentially unique experiences of the perinatal period.