Discussion: Studying a community sample of young adults, we found that exposure to peer verbal abuse was associated with increased drug use and elevated psychiatric symptom ratings. Substantial exposure was associated with a greater than twofold increase in clinically significant ratings of depression, a threefold to fourfold increase in anxiety and "limbic irritability," and 10-fold increase in dissociation. This level of peer verbal abuse was reported by 9.2% of participants who had no exposure to childhood sexual abuse, witnessing of domestic violence, or parental physical or verbal abuse and by 17.9% of the entire community sample. Hence, exposure to substantial levels of peer verbal abuse is a relatively common occurrence. Moreover, the effects of childhood exposure to peer verbal abuse on risk of psychopathology in early adulthood mirror results we previously reported for parental verbal abuse (5). Thus, verbal aggression from peers is an important and potent childhood stressor.
Middle school was the peak period of exposure to peer verbal abuse, with 9.8% of our community sample newly exposed. This finding fits with previous observations that peer physical aggression declines over the period from ages 8 to 18 while peer verbal abuse increases from ages 8 to 11, plateaus, and then declines from ages 15 to 18 (22, 23).
More importantly, the timing of exposure appears to shape its impact. Path analysis suggests that exposure during the middle school years (ages 11—14) was the most consequential and was associated with symptoms of anxiety, depression, dissociation, "limbic irritability," and degree of drug use. Overall, there were no significant associations between these symptoms and degree of exposure during elementary or high school when degree of middle school exposure was excluded. However, exposure at early and later ages amplified the association between symptom ratings and middle school exposure, more than doubling the amount of variance explained. This suggests that exposure during elementary and high school may sensitize or reinforce the effects of exposure during middle school. These findings are consistent with previous reports indicating that exposure to peer verbal abuse in secondary school is more serious than peer verbal abuse during primary school (23, 24). This may be because children in primary school predominantly engage in dyadic relationships, which can attenuate the perceived impact of bullying outside the dyad (23).
Another perspective is also possible. We recently published data indicating that there are sensitive periods when brain regions are most susceptible to the effects of childhood sexual abuse (20). The hippocampus was most vulnerable to childhood sexual abuse occurring at ages 3—5 years and 11—13 years. It is possible that the hippo-campus is also susceptible to other forms of abuse occurring during these years. Anxiety, depression, dissociation, and temporal lobe epilepsy-like symptoms have all been associated with aspects of hippocampal function (12, 25—29). Hippocampal volume was not assessed in this study.
Diffusion tensor imaging, however, revealed an association between degree of exposure to peer verbal abuse and measures of mean diffusivity, radial diffusivity, and fractional anisotropy in the splenium of the corpus callosum and the overlying corona radiata. The corpus callosum is a massive fiber tract interconnecting the left and right hemispheres. The corona radiata contains both descending and ascending axons that carry nearly all of the neural traffic to and from the cerebral cortex. Many of these ax ons pass through the corpus callosum. Studies suggest that alterations in radial diffusivity but not axial diffusivity, as observed, result from effects on myelin rather than axon numbers (30, 31). Corpus callosum alterations appear to be the most consistent finding in maltreated children (32—34), and it is perhaps remarkable that they emerged in a sample of comparison subjects with no axis I or II disorders. The sensitive period for the splenium (the most caudal portion of the corpus callosum) likely occurs during the middle school years, given the rostral-caudal progression of corpus callosum myelination (35) and our finding that the rostral body of the corpus callosum had a sensitive period between ages 9 and 10 (20).
It is interesting to speculate on how white matter alterations in the splenium might be related to elevated risk for depression, dissociation, or substance abuse. Fibers passing through the splenium interconnect the right and left occipital and inferior temporal cortices. Together these regions comprise the ventral visual processing stream, which has reciprocal connections with the hippocampus. The visual cortex is a plastic structure that is extensively modified by early experience. We previously reported (36) that exposure to childhood sexual abuse was associated with a 12%—18% reduction in gray matter volume in the right and left primary and secondary visual cortex. We have also found similar alterations in witnessing domestic violence (unpublished data). While the visual cortex plays a critical role in sensory perception, it may have additional functions. A reproducible finding in major depression is a substantial reduction in occipital cortex g-aminobutyric acid (GABA), which is restored following treatment with antidepressants or ECT (37). Exposure to early stress may target GABA-ergic interneurons or fiber pathways of the visual cortex and increase risk for the development of mood disorders. We and others have also speculated that alterations in the corpus callosum may set the stage for dissociative phenomena by diminishing intrahemispheric integration (38). It is also possible that lack of integration between right and left hemispheric processing of visual cues may lead to greater cue-induced craving in substance users and enhanced risk for abuse and dependence.
This study is unique for a number of reasons. First, it assessed and controlled for exposure to other forms of mal-treatment, such as childhood sexual abuse and parental verbal abuse. Second, it focused entirely on peer verbal abuse as a specific form of childhood trauma distinct from peer abuse involving physical assaults. Third, effects of exposure during different developmental stages were assessed based on our finding of "sensitive periods" when brain regions are particularly susceptible to abuse.
However, we must acknowledge that these analyses are retrospective and correlative and hence are subject to problems associated with faulty recall and spurious association. It is possible that individuals who show a tendency toward psychopathology in childhood are targeted as "odd" by peers and subjected to verbal abuse (39). It is also conceivable that a preexisting abnormality in the corpus callosum enhances risk for psychopathology and peer abuse. Path analysis delineates a mathematical solution to a series of equations, not a causal pathway. Causality cannot be inferred from this retrospective experimental design. Prospective studies are needed to explore these possibilities.
Psychiatric disorders probably arise from the temporal intersection of genetic susceptibility and adverse early experiences during critical developmental stages (26). Theorists have often focused on the importance of very early experiences and parent-child interactions. This study and other recent reports (9, 40) should strengthen interest in the importance of peer interactions and the vulnerability of peripubertal children. These findings further enhance concern that exposure to ridicule, disdain, and humiliation from parents, partners, or peers is emotionally toxic and may adversely affect trajectories of brain development.
Comments (2)