| Poster | Poster Session 05 Program Schedule
02/15/2024
02:30 pm - 03:45 pm
Room: Shubert Complex (Posters 1-60)
Poster Session 05: Neuropsychiatry | Addiction/Dependence | Stress/Coping | Emotional/Social Processes
Final Abstract #53
Confirmatory Factor Analysis of the Neurobehavioral Symptom Inventory in Women with and without Brain Injuries due to Intimate Partner Violence
Justin Karr, University of Kentucky, Lexington, United States
Agnes White, University of Kentucky, Lexington, United States
Sharon Leong, University of Kentucky, Lexington, United States
TK Logan, University of Kentucky, Lexington, United States
Category: Acquired Brain Injury (TBI/Cerebrovascular Injury and Disease - Adult)
Keyword 1: brain injury
Keyword 2: traumatic brain injury
Keyword 3: psychometrics
Objective:
No standardized questionnaire of neurobehavioral symptoms has been psychometrically examined among women survivors of intimate partner violence (IPV) with traumatic and anoxic-hypoxic brain injury. This study examined the latent structure and reliability of the Neurobehavioral Symptom Inventory (NSI) in women survivors of IPV and compared symptom severity across women with and without IPV-related brain injuries.
Participants and Methods:
Cisgender women were recruited online from Prolific.co (N=205, M=39.8 years-old, SD=11.9; 83.9% non-Hispanic White; 42.4% college-educated). All participants reported a prior history of physical or sexual IPV, with 112 reporting a prior traumatic or anoxic-hypoxic brain injury (e.g., due to nonfatal strangulation) and 93 reporting no prior history of brain injury. All participants completed the 22-item NSI, with confirmatory factor analyses conducted to assess model fit for one-factor, three-factor (i.e., cognitive, somatosensory, affective), and four-factor models (i.e., cognitive, vestibular, somatosensory, affective). A comparative fit index (CFI) ≥.95 and root mean square error of approximation (RMSEA) ≤.08 were indicative of acceptable model fit. The NSI total score and subscales from the best-fitting model were assessed for reliability using omega (ω). NSI scores were compared between women with and without IPV-related brain injury using Mann-Whitney U tests, with rank-biserial correlations (r) reported as effect sizes.
Results:
The one-factor model did not fit well (CFI=0.921, RMSEA=0.102). The three-factor model fit well (CFI=0.951, RMSEA=0.081), but the four-factor model showed the best fit (CFI=0.970, RMSEA=0.064). The NSI total score had strong reliability (ω=.93 [95% confidence interval: .92, .95]) and each of the subscale scores had acceptable-to-strong reliability: vestibular (ω=.77 [.72, .83]), somatosensory (ω=.72 [.66, .78]), cognitive (ω=.89 [.86, .91]), and affective (ω=.89 [.87, .92]). Women with brain injuries reported higher symptoms on the NSI somatosensory (p=.041, r=.17 [.01, .32]) and cognitive (p=.011, r=.21 [.05, .35]) subscales and the NSI total score (p=.029, r=.18 [.02, .33]), but not the affective (p=.163, r=.11 [-.05, .27]) or vestibular (p=.266, r=.09 [-.07, .24]) subscales.
Conclusions:
The confirmatory factor analyses supported a four-factor model of neurobehavioral symptoms among women survivors of IPV. The factors had acceptable-to-strong reliability and showed group differences based on prior IPV-related brain injury in terms of overall symptom severity and somatosensory and cognitive symptom severity. Among women with IPV-related brain injuries, cognitive and somatosensory symptoms may be uniquely elevated and serve as targets of intervention for this community.
|
|---|
