Neurobiologic function and temperament in subjects with personality disorders.

BACKGROUND

Personality traits have been hypothesized to involve specific neurotransmitter systems. In order to test this model, the relationship between the responses to serotonergic and noradrenergic probes, central cerebrospinal fluid (CSF) measures of monoamine neurotransmitters and the Tridimensional Personality Questionnaire (TPQ) were evaluated in a cohort of personality disorder subjects.

METHODS

A total of 142 patients meeting at least one personality disorder (meeting Diagnostic and Statistical Manual of Mental Disorders, Third Edition-Revised criteria) participated in these studies. The prolactin response to fenfluramine (a measure of serotonin function) was obtained for 110 subjects; growth hormone response to clonidine (a measure of noradrenergic function) was obtained for 77 subjects, while homovanillic acid (HVA) at baseline, an index of dopaminergic function, was available for 103 subjects. Measures of central neurotransmitter function (dopaminergic, serotonergic, and noradrenergic: HVA, 5-hydroxyindolacetic acid, and 3-methoxy-4-hydroxyphenylglycol, respectively) were available for 36 subjects. Separate regression analyses utilizing each of the hypothesized associations, where the TPQ total scores were used as the dependent measures and the biologic indices were the independent measures were conducted. Exploratory correlational analyses between these biologic measures and the four dimensions of the TPQ (and its subscales) were also conducted. (Correlations are reported if they would remain significant at P<.01 level after Bonferroni correction for multiple comparisons across the six neuroendocrine measures).

RESULTS

In the regression analyses, there was a trend association between CSF and plasma HVA in predicting novelty-seeking (P<.07). No other significant associations were found in the other three measures. Regarding the individual correlational analyses, the persistence scale of the TPQ was significantly positively correlated with the growth hormone response to clonidine (r=.30, P<.008). The sentimentality subscale (reward dependence) was positively correlated with CSF 5-hydroxyindolacetic acid (r=0.45, P<.001), while the attachment subscale (also reward dependence) was correlated with CSF 3-methoxy-4-hydroxyphenylglycol (r=0.49, P<.002).

CONCLUSION

Limited support was provided for a relationship between monoamines, particularly dopamine and novelty-seeking as well as norepinephrine and reward dependence but other hypothesized relationships were not supported by these measures.