CSF corticotropin-releasing factor in personality disorder: relationship with self-reported parental care

This cross-sectional investigation tests the relationship between the level of self-reported childhood parental care and cerebrospinal fluid (CSF) corticotropin-releasing factor (CRF) concentration in adults with and without personality disorder (PD). Based on preclinical models of the lasting effect of post-natal parental care on central CRF function, the primary hypothesis was that childhood parental care, as reflected by the parental bonding inventory (PBI) care and involvement subscale, is inversely correlated with adult CSF CRF levels. The sample includes cerebrospinal fluid CRF samples from 19 subjects who were included in a previously published report on the relationship between CRF level and Childhood Trauma Questionnaire score. Parental bonding was measured retrospectively with the PBI in 54 medication-free male and female subjects, 37 of whom were diagnosed with a DSM-IV PD, 17 of whom were normal controls free of Axis I and II psychopathology. CSF CRF level as measured by lumbar puncture was entered into a model as the dependent variable, with the independent variables of PBI Parental Care and Involvement, diagnostic category, age, and gender. The model predicting CSF CRF level was significant, with PBI parental care and involvement negatively correlated with CSF CRF level. PD subjects with higher than median PBI parental care and involvement score had significantly lower CSF CRF levels than both normal controls and PD subjects with lower than median PBI parental care and involvement. Higher levels of self-report parental care predict lower CSF CRF levels in PD subjects, consistent with a beneficial effect of parental care on decreased stress reactivity, and consistent with previous reports in humans. The cross-sectional design of the study, however, limits causal inferences.