To the Editor: We thank Drs. Levitan and Parikh for their thoughtful comments regarding our article and for giving us the opportunity to discuss the association of childhood adversity to depression subtypes. They note that our report of a significant association of childhood adversity to endogenous depression is inconsistent with their relation of early trauma to atypical but not endogenous depression (Levitan et al., 1998).

This confused us because Dr. Levitan and colleagues did not assess endogenous or atypical depression subtypes. Instead, they examined the relation of trauma to three reverse versus typical neurovegetative symptoms: increased appetite, weight gain, and hypersomnia versus decreased appetite, weight loss, and insomnia. We assume they meant to equate reverse neurovegetative symptoms with atypicality and typical neurovegetative symptoms with endogeneity. However, the relation of stress to depression symptoms versus subtypes/syndromes should be distinguished. For example, these typical symptoms are just as likely to be present in subjects with nonendogenous/nonatypical depression as in endogenous depression. Nevertheless, the comments by Drs. Levitan and Parikh highlight a crucial test that remains to be undertaken comparing endogenous, atypical, and nonendogenous/nonatypical depressive subtypes. This may reveal further specificity.

Another key difference in methods is that Dr. Levitan and colleagues simply coded abuse as present or absent, while we distinguished gradations in trauma severity. While nonsevere childhood adversity was associated with nonendogenous depression, consistent with earlier reports (Levitan et al., 1998, and Parker et al., 1997), severe childhood adversity was strongly associated with endogenous depression. Our fine-grained assessment of childhood adversity thus allowed for greater specificity in describing the relationship between early trauma and depression.

Drs. Levitan and Parikh support their contention by noting the differential association of atypicality and endogeneity to HPA axis dysregulation (Yehuda, 2001; Anisman et al., 1999; Levitan et al., 2002; Gold and Chrousos, 2002). The neural consequences of stress are complex and depend on the nature of the stressor (e.g., acute versus chronic) and the developmental period of exposure. For example, while exposure to acute traumatic stressors leads to a hyperregulated HPA axis, a history of chronic and/or uncontrollable stressors may lead to hypercortisolemia (1). Furthermore, certain forms of stress induce anhedonia in rodent models, as assessed by deficits in reward reactivity (2). These points reinforce the inference that early trauma has important implications for endogenous depression.

Drs. Levitan and Parikh note that we used a cross-sectional study design that excluded participants with chronicity and certain comorbidities. However, most of the subjects (89%) had recurrent illness. In addition, a number of comorbid disorders that are strongly linked to trauma (e.g., anxiety disorders) were not excluded (3). Most important, childhood adversity remained significantly associated with endogenous depression even after control for depression history and comorbidity.

It is unclear how our cross-sectional strategy diluted our ability to find a relation between childhood adversity and nonendogenous depression. Although it is possible that some of our participants with endogenous depression had a fluctuating course and were at one time exhibiting nonendogenous/atypical symptoms, such a course seems just as likely among the nonendogenous participants. Nevertheless, we were limited in our ability to illuminate the relation of childhood adversity to syndrome development across and within episodes. We urge researchers to pursue longitudinal studies examining this question.