Anxious predisposition

Twin and adoption studies demonstrate that childhood anxiety is moderately heritable (Gregory and Eley, 2007), although the magnitude of effects varies according to child sex, age, information source (i.e., parent versus child rated), and the form of anxiety. The genetic substrate of anxiety is poorly understood (e.g., Battaglia et al., 2005), yet it is likely to differ in its effects in response to interactions with environmental characteristics (e.g., Fox et al., 2005). The psychological trait most widely investigated as a potential vulnerability factor for the development of anxiety is behavioral inhibition (BI) – that is, the temperamental pattern characterized by fear and withdrawal in unfamiliar situations. BI is considered to be an early temperamental precursor of ADs, particularly social AD (e.g., Hirshfeld-Becker et al., 2008).

Studies for the degree to which parental ADs constitute a risk of BI among offspring have mostly been limited to those of parents with panic disorder and agoraphobia, and the results have been mixed, some studies suggesting that high rates of BI are accounted for by comorbidity with parental depression (Rosenbaum et al., 2000). In the Reading longitudinal study, we found no significant difference in rates of high infant irritability at 10 weeks (Murray et al., 2007) or (nonsocial) BI at 14 months of age (Murray et al., 2008) between infants of anxious and nonanxious mothers. Together, these findings fail to show a clear link between early BI and parental AD. Thus, reported associations between later BI and parental ADs may reflect offspring responses (e.g., to environmental influences associated with the parental disorder), or may occur as a result of comorbid parental conditions.

It has been suggested that biases of attention and interpretation may create a predisposition to the development of anxiety in children (Hudson and Rapee, 2004). The question of whether parental anxiety creates a risk of such biased information processing has received little examination, although there is some suggestion that offspring (aged 9–19 years) of parents with panic disorder show distinctive attentional patterns to emotional stimuli that are not accounted for by child anxiety (Pine et al., 2005). It is not clear whether these findings reflect genetic or environmental factors; however, our finding that 10-week-old infants of mothers with social AD avoid high-intensity fear faces (compared to offspring of mothers with generalized anxiety disorder (GAD) and nonanxious mothers) does indicate that biases in the processing of emotional stimuli may be present from a very early age (Creswell et al., 2011).

Together, these lines of evidence highlight that, while children of parents with ADs are at an inflated risk of ADs, underlying vulnerability factors have not been clearly identified.

Socialization experiences

Exposure to negative life events may play a causal role in the development of ADs (e.g., Goodyer et al., 1988). Our model suggested that parental anxiety may be associated with increased exposure to particular negative life events, and this could increase the risk of child anxiety, either directly or by diminishing the child’s sense of control (Murray et al., 2009). There has been little research examining this proposition. However, life events that are known to be more common among adults with ADs (e.g., Kessler et al., 1998; Merikangas et al., 1998), such as marital breakdown, have been found to be associated with ADs in offspring (Shanahan et al., 2008).

Furthermore, in line with suggestions by Rapee and Heimberg (1997), we suggested that anxious parents may, as a function of their own disorder, place limits on their children’s wider experiences, a limitation that could inflate apprehension, reduce opportunities to learn skills and develop independence, and create an overreliance on a limited number of adults. Again, little research has directly examined this proposal. In the Reading longitudinal study, the infants of mothers with social AD experienced less caregiving by people other than the mothers (including the father), and they were less likely to have attended nonparental group care at 10 and 14 months of age; this caregiving pattern was associated with infant sleep problems at 2 years, a potential marker of later emotional problems (Creswell et al., 2010). These preliminary findings suggest that the socialization practices of parents with AD may present a risk to child adjustment.

Anxiogenic modeling and information transfer

Learning accounts have emphasized two further, probably overlapping, routes for the development of anxiety: observation of others’ anxiety (modeling or a vicarious acquisition), and receiving information regarding the threatening nature of the environment (i.e., informational acquisition) (e.g., Rachman, 1990).

Reinforcing responses to child anxious behaviors

The parenting dimensions that have most commonly been implicated in the development of child ADs are parental control and negativity. Specifically, it has been proposed that a parent’s excessive regulation of child behavior and discouragement of independence are likely to (1) communicate to the child that the world is a dangerous and uncontrollable place (e.g., Hudson and Rapee, 2004), and (2) militate against the child’s developing a sense of competence and mastery, reinforcing avoidance of challenge (Chorpita and Barlow, 1998). Cross-sectional and experimental studies have provided evidence to support this hypothesis (e.g., McLeod et al., 2007; Thirlwall and Creswell, 2010). Furthermore, parental negativity, or lack of warmth, has been held to lead the child to believe that the environment is fundamentally hostile and threatening and that outcomes will be negative, thereby contributing to the development of a sense of low self-worth and competence (Parker, 1983). Findings have been inconsistent with regard to this hypothesis (e.g., McLeod et al., 2007).

A number of theorists have suggested that parental control and, possibly, negativity are likely mechanisms by which parental AD poses a risk to offspring adjustment. Specifically, highly anxious parents may be more likely to exert excessive control over their child because of their own heightened perception of threat in their own and their child’s environment and their increased sensitivity to their child’s distress (e.g., Hudson and Rapee, 2004). Parental anxiety has also been proposed to lead to increased negativity to the child as a function of parents’ limited sense of control over their children’s anxious responses (e.g., Hudson and Rapee, 2004). There is currently little support for these proposals. In a small study by Whaley et al. (1999), maternal AD was associated with increased negativity and reduced autonomy granting; however, in an expanded version of this study, the effects appeared to be accounted for by the presence of child, rather than parental, anxiety (Moore et al., 2004). Other studies have similarly failed to find associations between parental AD and overcontrol or negativity (e.g., Gar and Hudson, 2008; Ginsburg et al., 2005); some, instead, found a lack of engagement among parents with AD (Turner et al., 2003). Overall, it appears that more involved or controlling behaviors are common parental responses to having an anxious child, possibly reinforcing child anxiety. Indeed, a recent meta-analysis revealed a weak association between parental anxiety and observations of control (d = 0.08; van der Bruggen et al., 2008), although notably stronger associations were found depending on the aspect of parental control assessed and the methodological features of the studies (d = 0.25–.49).

It is, nonetheless, possible that infant or child inhibition may provoke particular parenting styles more readily among anxious than nonanxious parents. For example, Hirshfeld et al. (1997) found that maternal criticism of the child was a function of a significant interaction between child BI and maternal status: within the group of anxious mothers, 65% of those with inhibited children were critical compared to 18% of those with noninhibited children. Similarly, Murray et al. (2008) found that mothers with social AD showed low levels of autonomy promotion (encouragement) to their infants to engage with a stranger only where the infant was behaviorally inhibited. Recent findings from our research clinic have provided consistent results. Eighty-eight children with ADs (7–12 years) were observed interacting with their mothers in a range of challenging tasks. Despite no observable differences in child anxiety and avoidance during the tasks, mothers who had an AD reported more negative expectations of their child, showed higher levels of intrusive and anxious behaviors and lower levels of warmth, and exhibited a more negative interactive tone than mothers without ADs, particularly when their children appeared anxious during the task (Creswell et al., 2013).

Together, these findings suggest that overcontrolling and, possibly, negative responses may reflect a common response to parenting an anxious or inhibited child, and that these behaviors may be especially likely to arise in parents who are themselves prone to anxiety when their child is confronted by a challenge. Only a few studies have examined whether these problematic parenting behaviors arise in particular ADs and in particular contexts. Challacombe and Salkovskis (2009) found that both mothers with OCD (obsessive compulsive disorder) and those with panic disorder were less warm and autonomy promoting than nonanxious mothers during discussion tasks, with no significant differences between the two disorders. However, disorder-specific effects have been found in the case of parental social AD, with socially anxious mothers exhibiting less warmth, more criticism, and more doubts about their child’s competency than mothers with other (nonsocial) ADs specifically in situations that posed a specific challenge in relation to the anxious mothers’ disorder (Murray et al., 2012). The context-dependency of potentially anxiogenic parental responses of anxious parents has also been supported by a recent study by Stein et al. (2012). Specifically, mothers with GAD, in comparison to a healthy control group, and, to some extent, a group of mothers with major depressive disorder, were found to exhibit reduced responsiveness and vocalization with their 10-month-old infants after being primed to worry relative to their responses to a neutral prime. These findings highlight the need for carefully selected paradigms when assessing parent–child interactions in the context of parental ADs.