Using item response theory to optimize measurement of chronic stress in pregnancy

Objective Utilize Rasch analysis to develop an optimized self-reported measure of stress in pregnancy and examine the association with patient demographics and biologic measures of stress. Study design Measured self-reported stress in pregnant women using 12 existing scales. Collected blood for biologic measures of stress (Epstein Barr Virus [EBV], C-Reactive Protein [CRP], Corticotropin Reactive Hormone [pCRH], and Adenocorticotropin Hormone [ACTH]). Used multidimensional scaling and Rasch analysis to produce an item reduced self-report measure. Results Enrolled 112 women. Survey items reduced to two factors: perceived stressors and buffers of stress. Women with a domestic partner had lower perceived stress (p = 0.003). Caucasian women reported higher buffers of stress (p = 0.045), as did women with private insurance (p < 0.001), a planned pregnancy (p < 0.01), and a domestic partner (p < 0.001). Women with higher buffers of stress had lower levels of pCRH (adjusted p = 0.01). Conclusion Item reduced, optimized measures of stress were associated with significant differences in patient demographics and biologic measures of stress.