Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)



Research Advisor

Mona Newsome Wicks, Ph.D., RN


Kelly Askins, MD Sarah Mynatt, Ed.D., APRN, BC Muriel Rice, Ph.D., RN Caroline Riely, MD Rebecca Winsett, Ph.D., RN


Caregiving, Burden, Caregiver, Chronic Liver Disease, Depressive symptoms


This descriptive correlation study had three purposes. The study first described depressive and anxiety symptom levels, prevalence of hazardous drinking, rewards, and subjective burden reported by family caregivers of patients with chronic liver disease (CLD). Second, the investigator compared depressive and anxiety symptom levels, prevalence of hazardous drinking, rewards, and subjective burden reported by African Americans family caregivers with those reported by Caucasian family caregivers of patients with CLD. The third purpose was to determine the predictors of subjective burden and mental health status of family caregivers of persons with CLD.

A convenience sample of 73 caregivers of patients receiving care in a university-based, hepatology practice located in a large metropolitan city in the Mid-South provided data for the study. Depressive symptoms were measured with the Center for Epidemiological Studies Depression Scale (CES-D). Anxiety was measured with the Hamilton Anxiety Rating Scale (HAM-A). Hazardous drinking was measured with Alcohol Use Disorders Identification Test (AUDIT). Subjective burden was measured using Zarit’s Burden Interview (BI) and caregiver rewards were assessed with Picot Caregiver Rewards Scale (PCRS). An investigator developed demographic data form was used to obtain caregiver and patient characteristics. The Pearlin Stress Process Model (SPM) provided the conceptual basis for the study.

Cronbach alpha coefficients obtained for the CES-D, HAM-A, AUDIT, BI, and PCRS scales ranged from .72 to .93. Data were analyzed using descriptive statistics, independent sample t-tests, Mann Whitney tests, Pearson’s product-moment correlation, Spearman’s rho correlation, and stepwise multiple regression analyses. The sample consisted of more Caucasians (65.8%) than African Americans (30.1%) and more females (78%) than males (22%). Caregivers were typically married (53%), middle-aged (48.2 ± 14.7 years) high school graduates (12.4 ± 2.6 years). Most caregivers were employed full-time (41.1%).The study sample reported mild depressive symptoms, little or no anxiety or hazardous drinking, mild distress for subjective burden, and moderate caregiver rewards scores. Caucasians had a slightly higher though non-significant depressive symptom score when compared to African Americans. Both African American and Caucasian caregivers in this study reported experiencing little to no anxiety and the prevalence of hazardous drinking was low for both groups. African Americans’ mean subjective burden score reflected little to no burden while Caucasian caregivers’ scores reflected mild burden; these differences were not statistically significant. African American caregivers reported significantly higher mean rewards scores compared to Caucasian caregivers.

Stepwise regression analyses were performed to determine the degree to which selected independent variables significantly contributed to the explained variance in depressive and anxiety symptom levels, subjective burden, and caregiver rewards for the entire sample. Income decrease, worried about children, caregiver ethnicity, and caregiver gender were regressed on depressive symptoms. The variable worried about children (beta=.24, p =.05) remained in the final model and accounted for 6% of the adjusted variance in depressive symptom scores (F (1, 62) =4.768, p=.03). Worried about children was significantly and positively associated with increased depressive symptoms. Caregiver support, income decrease, worried about children, and activities of daily living were regressed on anxiety symptom levels. None of these variables significantly predicted anxiety symptom levels. In contrast, income decrease (beta=.35, p=.003) and worried about children (beta=.35, p=.003) predicted 25% of the adjusted variance in subjective burden (F (2, 56) =10.9, p=≤.000). Caregiver ethnicity (beta=.37, p=.001) and employment (beta=-.33, p=.003 accounted for 22 % of the variance in caregivers rewards (F (2, 66) =10.7, p=≤.000). Caregivers who were African American reported greater rewards and unemployed caregivers reported fewer rewards associated with the caregiver role. Findings are consistent with previous studies showing that African American caregivers experience more caregiver rewards compared to Caucasian caregivers though differences in burden did not occur. Caregivers who were worried about their children experienced more depressive symptoms and subjective burden. In contrast with published studies in other caregiver populations, neither length of care nor severity of symptoms significantly correlated with subjective burden or depressive symptoms in the study sample. Programs have been developed in three states to off-set caregiving-related income decreases, which could be an effective approach in this sample of caregivers who reported significant decreases in income. Further research could be conducted using focus groups to explore caregiver concerns related to children and this information could provide direction to reduce the burden and depressive symptoms associated with this variable. Research is also needed to determine if ethnic differences in the perception of rewards mediate or moderate the experience of burden and depressive symptoms in the study population. Interventions are particularly needed for depressive symptoms as clinically significant levels were present and clinical referrals warranted in this study sample.