Risk of behavioral inhibition from infancy to age seven
This is a longer than usual post because I find behavioral inhibition fascinating and this study is very well done. Anaya et al. (2023) published “Developmental Trajectories of Behavioral Inhibition from Infancy to Age Seven: The role of genetic and environmental risk for psychopathology” in Child Development. Here’s the edited abstract and article with some key information in bold:
The present study leveraged data from a longitudinal adoption study of 361 families recruited between 2003 and 2010 in the United States. We investigated how psychopathology symptoms in birth parents (BP; Mage = 24.1 years; 50.5–62.9% completed high school) and adoptive parents (AP; Mage = 37.8 years; 80.9% completed college; 94% mother–father couples) influenced children's behavioral inhibition (BI) trajectories.
Behavioral inhibition (BI) is a temperament profile first used by Kagan (1999) to describe a group of 21-month-old infants who were overly cautious of, and withdrew from, unfamiliar stimuli. The BI profile is characterized by shy and reticent behaviors, including clinging to the caregiver, vigilance, and avoidance of novel social situations, even in non-threatening contexts.
Decades of research suggest that childhood BI is one of the single, most reliable predictors of anxiety symptoms and disorders. In a meta-analysis, Clauss and Blackford (2012) reported that behaviorally inhibited children were three to four times more likely to develop social anxiety disorder (SAD) by age 15, with a significant odds ratio of 7.59. Similarly, toddlers who show continuity of BI into middle childhood and adolescence are also at a greater risk for social anxiety.
Despite strong links between high levels of BI and anxiety, roughly 60% of behaviorally inhibited children do not grow up to develop clinical levels of anxiety (Fox et al., 2021), attesting to the heterogeneity of BI trajectories. This heterogeneity also suggests the potential for endogenous and exogenous factors that may canalize trajectories toward risk or resilience. In line with Bronfenbrenner's Bioecological theory Henderson et al. highlighted that behaviorally inhibited tendencies contribute to and color children's social experiences given their sensitivity to novel and unpredictable environmental inputs. In addition, individuals in the proximal environment (e.g., parents, peers, siblings, teachers) may respond to inhibited tendencies in systematic ways, shaping the behaviorally inhibited child's idiosyncratic experience of the social world. For example, parents may respond with increased over-solicitous behaviors, teachers' biases about shyness may influence their evaluation of inhibited children, and peers may respond with greater disengagement and rejection. The cumulative effect of parental, teacher, and peer responses may potentiate children's inhibited tendencies over time.
Many studies suggest that individual differences in reactive cognitive control, attention, and neural correlates of overcontrol may all contribute to an underlying neurobiological profile of BI that predicts continuity of inhibited tendencies and increased risk for social reticence and anxiety.
Links between parent psychopathology symptoms and child BI seem to be domain specific to internalizing and not externalizing tendencies (Essex et al., 2010), despite the two being modestly correlated in childhood (Willner et al., 2016). Children of parents with higher levels of depressive and anxiety symptoms show higher rates of BI and increased inhibition in novel contexts. Parents with higher internalizing symptom levels may show lower sensitivity and be less effective when emotionally scaffolding their children, which may contribute to the stability of child BI. Children are not passive recipients of caregiving; thus, children's behaviorally inhibited tendencies can also influence the rearing environment. Furthermore, many genetically informed studies show that children's genetically influenced characteristics can elicit distinct parenting behaviors and parent psychopathology symptoms to influence the rearing environment.
Parents with higher levels of internalizing symptoms may be more likely to avoid social interactions, hindering children's access to a more diverse range of contexts and limiting most of their interactions to the home environment. Concomitantly, children's genetic propensity for psychopathology may also lead to a profile of behaviors (e.g., dysregulated fear in non-threatening contexts, frequent negative affect in social settings, and difficulty soothing) that increases parent's own reticence in social interactions. As a result of limited social stimulation, children may struggle to habituate to novel interactions and situations, which may reinforce BI over time. For example, behaviorally inhibited children who receive home care are more likely to retain an inhibited profile compared to behaviorally inhibited children who attend daycare (Fox et al., 2001). Daycare settings may provide more diverse social opportunities for behaviorally inhibited children to experience and solve social problems with peers.
An early environment characterized by limited social interactions may be particularly detrimental to behaviorally inhibited children. Indeed, children of mothers with higher avoidant behaviors, specifically, seem to show higher distress to novelty. Parents with higher levels of internalizing symptoms may themselves be affected by novel situations and by their own child's genetic predispositions for inhibition, hindering parents' ability to scaffold their child when navigating settings that elicit reticent and avoidant tendencies. Furthermore, these parents may also directly model avoidant and inhibited strategies to their children as normative coping mechanisms. The incidence of BI is higher in children of parents who report higher levels of anxiety, panic disorder, and comorbidity with major depression. Although correlational, these studies support Henderson et al.'s transactional model, suggesting that repeated exposure to parent internalizing symptoms may amplify BI over time.
In contrast, parents with low levels of internalizing symptoms may provide emotional scaffolding to their inhibited child more efficiently, modeling how to engage in novel situations, and socializing their child to pursue more social interactions. For example, toddlers characterized as shy and inhibited display less social wariness when their mothers verbally encourage them to play and interact with unfamiliar peers. Furthermore, parents with low levels of internalizing symptoms may have better self-regulation skills, which may provide them with greater behavioral and emotional flexibility to parent an inhibited child, and to foster the child's own development of self-regulation. In turn, better self-regulation in behaviorally inhibited children may reduce risk for anxiety.
Twin studies report moderate to high heritability of BI and genetic influences on the relation between temperamental inhibition and psychopathology, suggesting that links between parent psychopathology and child BI may be explained by shared genetic variance. Together, these findings suggest that continuity and change in children's BI can be accounted for by both environmental and genetic influences.
In summary, genetically informed studies of BI converge on one key finding: neither higher genetic propensity for internalizing nor higher internalizing symptoms in the child's environment are strongly predictive of child BI during infancy or early childhood anxiety risk.
To address these research gaps, the present study used data from 361 children adopted at or near birth, their BPs, and their APs, and examined the influences of BP psychopathology and AP internalizing symptoms on the developmental trajectories of child BI, observed at 18 months, 27 months, 4.5 years, and 7 years of age. We hypothesized that: (a) across the sample, BI trajectories would decrease from 18 months to 7 years in line with noted declines in BI as regulation strategies become more effective, (b) higher genetic risk for psychopathology and higher environmental risk for internalizing, but not externalizing, symptoms would independently predict higher initial levels of BI and stable or increasing BI trajectories over time, and (c) higher genetic risk coupled with higher AP internalizing symptoms would lead to increasing BI trajectories, suggesting a dual risk model of gene by environment (GxE) interaction on the stability and continuity of BI.
We include Cohort I, a set of 361 triads of adopted children, APs (357 mothers; 362 fathers), and BPs (359 mothers; 119 fathers), who provided behavioral data.
BP psychopathology significantly predicted the BI latent slope, indicating that higher BP psychopathology scores predicted lesser decreases or flatter slopes in BI trajectories (β = .033, p = .044). Analyses suggested that at low and mean levels of AP internalizing symptoms, BI trajectories were stable over time and did not systematically vary as a function of BP psychopathology. Interestingly, differences in BI trajectories across levels of BP psychopathology only emerged in the context of higher levels of AP internalizing symptoms.
While BI generally decreased over time, there were notable individual differences in trajectories. Our results did not support any direct genetic or environmental effects on BI trajectories. Rather, we noted a GxE interaction effect on the continued high stability of inhibited tendencies over time that was specific to a higher genetic propensity for psychopathology and higher internalizing levels in the adoptive family. Consistent with our initial hypothesis, BI trajectories across the sample significantly decreased from 18 months to 7 years.
At the level of the sample, there is substantial variability in the developmental patterns of inhibited tendencies, indicating that infants exhibit varying levels of inhibition and varying decreasing patterns of inhibition. This decrease over time is in line with the transition to school, a period when peer interactions gain influence. Reinforcement from peers may disfavor inhibited tendencies and reward social engagement, leading to decreases in normative levels of inhibition. This pattern of feedback from peers and adults also reflects Western cultural norms that view shyness as a negative trait.
BP, but not AP, psychopathology symptoms independently predicted change in BI, partially in line with our hypotheses. Maternal psychopathology (e.g., anxiety and depression) is associated with higher levels of BI and with greater stability of BI trajectories in genetically related families. Parental modeling and socialization may contribute to these patterns, as parents' beliefs and behavioral transactions with their children may mediate the link between parents' psychopathology and child BI beyond heritable influences on early child temperament. However, our results did not fully support this environmental effect, as we did not find a significant independent association between AP internalizing and BI slope that would support a direct role of parental socialization.
We found a significant independent association between BP psychopathology symptoms and the BI slope, supporting direct genetic influences on trajectories of inhibition. This effect was no longer significant when G×E interaction effects were incorporated in the model. However, our results join at least one older study and a more recent body of literature in which direct heritable influences on inhibition and social fear have been absent when G×E interactions are also considered. We noted a significant interaction between BP psychopathology and AP internalizing, but not externalizing, levels. The specific effect of internalizing symptoms in the rearing environment may be surprising in the context of comorbidity between internalizing and externalizing symptoms. However, it should be noted that we examined AP internalizing and externalizing symptoms in the same model, controlling for their overlap.
Children of BPs with lower levels of psychopathology showed BI trajectories that decreased at a faster rate than the decreasing trajectory seen on average for the full sample. It is possible that children with a lower genetic propensity to psychopathology may simply fit a developmental profile of inhibition that will generally decrease over time regardless of rearing environment. However, this interpretation is not fully supported by our simple slopes analysis, which shows that at low and average levels of AP internalizing symptoms, BI trajectories of children with the same lower genetic propensity were not significantly different from the rest of the sample. Instead, our findings suggest that this pattern of accelerated decrease in inhibition was specific to a rearing environment characterized by relatively higher levels of internalizing symptoms. Potentiated decreases in BI may still seem counterintuitive in this context because higher caregivers' internalizing levels are usually interpreted as risk factors and low inhibition and shyness are usually seen as positive. However, some research suggests the latter need not always be true.
The second pattern that emerged from our simple slope analysis was in line with our hypothesis, showing that within the context of higher levels of internalizing symptoms in the APs, children of BPs with higher psychopathology symptoms exhibited significant increases in inhibition from 18 months to 7 years compared to the overall sample's decreasing trajectory. These children exhibit trajectories of inhibition and shyness that are reinforced over time, a pattern that is consistently associated with greater risk for anxiety. We report this effect with caution, as the size of this subgroup of children in our sample was rather small (n = 20), and thus, future replication is warranted.
In summary, the present study expands the BI literature by modeling for the first time, trajectories of inhibited tendencies in the context of genetically unrelated families and suggesting that in line with sociocultural perspectives, inhibition seems to, on average, decrease over time. We also show that BI trajectories systematically differed as a function of genetic influence only in the context of higher internalizing levels in the child's rearing environment.
What I love about this study is its complexity and its suggestion of important questions to ask about early experience when children present with internalizing symptoms. I find it especially fascinating that children in a setting with internalizing symptoms lose their behavioral inhibition faster than others.