Transactional Relations Between Maternal Anxiety and Toddler Anxiety Risk Through Toddler-Elicited Comforting Behavior

• In Event: Parental and Child Influences on the Transmission and Development of Anxiety

Abstract

The current study tests a transactional model of maternal anxiety and toddler risk for anxiety across three time-points, examining child-elicited maternal comforting behavior as a behavioral mechanism accounting for bidirectional relations as mothers and children develop across early childhood. Maternal anxiety has predicted child anxiety risk through parenting behavior (Jones et al., 2015), and theory supports child-elicited effects on anxiogenic parenting (Dadds & Roth, 2001). We examine child-elicited maternal comforting behavior (physical affection directly solicited by the child) in uncertain but low-threat situations, based on previous findings that high levels of warmth and seemingly sensitive responsiveness predicts toddlers’ anxiety outcomes (Mount et al., 2010). Comforting behavior in low-threat situations may facilitate both mothers’ and toddlers’ avoidant coping, which may then feed back into each member’s anxiety. Thus, we hypothesized that both mother-directed and child-directed effects in a bi-directional model of mother-child anxiety would be mediated through child-elicited maternal comforting.

Mother-toddler pairs (n = 174) participated in laboratory assessments at child ages 1 (T1), 2 (T2), and 3 (T3) years. Mothers self-reported anxiety symptoms via three questionnaires, with total scores aggregated into a composite. Toddler inhibited temperament was observed using an established laboratory procedure (Calkins et al., 1996; Fox et al., 2001). Mothers also reported their perceptions of toddlers’ anxious behaviors using the Infant Toddler Social-Emotional Assessment (Briggs-Gowan & Carter, 1998). Maternal anxiety, observed inhibited temperament, and mother-perceived toddler anxiety were each measured at each assessment. Child-elicited maternal comforting was observed at child age 2 across two standardized laboratory tasks validated as “low threat” (Buss, 2011; Kiel & Buss, 2010); mothers were instructed to behave naturally.

Direct and indirect bidirectional effects were tested simultaneously in longitudinal path models using FIML in MPlus 7.3. We specified autoregressive and cross-lagged paths within and between maternal anxiety and toddler anxiety risk (observed inhibited temperament [Model 1; Figure 3] or mother-perceived toddler anxiety [Model 2], tested in separate models) across the three assessments. We further specified paths from T1 and T2 maternal anxiety and toddler anxiety risk to T2 solicited comforting behavior, and paths from T2 solicited comforting behavior to T3 maternal anxiety and toddler anxiety risk. Both models demonstrated good fit to the data.

T2 solicited comforting was predicted by both T1 maternal anxiety and T1 toddler inhibited temperament indirectly through counterpart T2 variables in the Model 1 (95% CIs [0.487, 11.614] and [0.151, 3.15], respectively; see Figure 3), and directly by T1 mother-perceived toddler anxiety in Model 2 (t = 2.12, p = .034). In Model 1, T1 inhibited temperament predicted T3 maternal anxiety through T2 inhibited temperament and solicited comforting (95% CI [0.000, 0.046], 90% CI [0.001, 0.039]). Model 2 yielded similar findings, although weaker; maternal anxiety related more strongly to perceptions of inhibited temperament. Thus, although both maternal anxiety and toddler anxiety risk predicted solicited comforting behavior, comforting behavior predicted only maternal anxiety, suggesting an important child-elicited effect on mothers’ functioning above and beyond stability in anxiety symptoms.

Authors

• Elizabeth J Kiel, Miami University - Ohio
  Presenting Author

• Elizabeth Aaron, Miami University - Ohio
  Non-Presenting Author

• Sydney Marie Risley, Miami University - Ohio
  Non-Presenting Author

• Aaron M. Luebbe, Miami University - Ohio
  Non-Presenting Author