Abstract
OBJECTIVE: This study aimed to characterize mothers' communication with their children in a sample of families with a new or newly relapsed pediatric cancer diagnosis, first using factor analysis and second using structural equation modeling to examine relations between self-reported maternal distress (anxiety, depression, and posttraumatic stress) and maternal communication in prospective analyses. A hierarchical model of communication was proposed, based on a theoretical framework of warmth and control. METHODS: The sample included 115 children (age 5-17 years) with new or newly relapsed cancer (41% leukemia, 18% lymphoma, 6% brain tumor, and 35% other) and their mothers. Mothers reported distress (Beck Anxiety Inventory, Beck Depression Inventory-II, and Impact of Events Scale-Revised) 2 months after diagnosis (Time 1). Three months later (Time 2), mother-child dyads were video-recorded discussing cancer. Maternal communication was coded with the Iowa Family Interaction Ratings Scales. RESULTS: Confirmatory factor analysis demonstrated poor fit. Exploratory factor analysis suggested a six-factor model (root mean square error of approximation = .04) with one factor reflecting Positive Communication, four factors reflecting Negative Communication (Hostile/Intrusive, Lecturing, Withdrawn, and Inconsistent), and one factor reflecting Expression of Negative Affect. Maternal distress symptoms at Time 1 were all significantly, negatively related to Positive Communication and differentially related to Negative Communication factors at Time 2. Maternal posttraumatic stress and depressive symptoms each predicted Expression of Negative Affect. CONCLUSIONS: Findings provide a nuanced understanding of maternal communication in pediatric cancer and identify prospective pathways of risk between maternal distress and communication that can be targeted in intervention.