Monoamine neurotransmitter interactions in drug-free and neuroleptic-treated schizophrenics.

OBJECTIVE: To study recent suggestions by a number of investigators that interactions between monoamine neurotransmitter systems play an important role in schizophrenia. It has not been clear how hypotheses about interactions might be tested in clinical data. One means for indexing interactions between monoamine neurotransmitter systems may be to compare correlations between cerebrospinal fluid (CSF) monoamine metabolite (homovanillic acid [HVA], 5-hydroxyindoleacetic acid [5-HIAA], and 3-methoxy-4-hydroxyphenylglycol [MHPG]) or ratios of these metabolites (HVA/5-HIAA and HVA/MHPG). DESIGN: We compared these putative measures of monoamine neurotransmitter interaction in 50 drug-free patients with schizophrenia (hospitalized on an inpatient ward of a tertiary care hospital) and 33 normal controls and examined the effects of neuroleptic antipsychotic treatment on these measures in 41 patients (22 of whom had antecedent drug-free CSF data). RESULTS: Drug-free patients with schizophrenia had significantly smaller correlations between CSF monoamine metabolites than normal controls. Longer drug-free time was associated with even smaller correlations between metabolites, suggesting that the difference between controls and patients was not due to acute drug withdrawal. After treatment with neuroleptic antipsychotics there were significant increases in the HVA/5-HIAA and HVA/MHPG ratios, as well as increases in correlations between monoamine metabolites. After treatment, there were no significant differences in metabolite correlations between patients and controls. Metabolite ratios and correlations did not predict subsequent treatment response, but preliminary analyses demonstrated negative relationships between HVA/5-HIAA and HVA/MHPG ratios and Brief Psychiatric Rating Scale rating at that time. CONCLUSIONS: The present findings are consistent with and support hypotheses suggesting that interactions between monoamine systems are altered in schizophrenia and that antipsychotic treatment may affect the functional balance between different monoamine neurotransmitters (although one should keep in mind factors other than interactions between monoamine systems that affect metabolite correlations and ratios.

Effects of single dose haloperidol administration on plasma homovanillic acid levels in normal subjects.

Homovanillic acid (HVA), an oxidative metabolite of dopamine, has been shown in a number of studies to reflect severity of symptoms and to predict response to neuroleptic treatment in schizophrenic patients. In several clinical studies, HVA levels have been shown to have a positive relationship with symptom severity and to decline over time upon treatment with antipsychotic agents. The magnitude of this decline appears to be related to the degree of symptom reduction in patients so treated. However, administration of dopamine postsynaptic antagonists should be expected to increase synaptic dopamine availability, thereby increasing HVA concentrations, according to traditional models of drug action. While in some studies, this appears to be the case, we saw no evidence of an early phase of HVA elevation after administration of 4- and 10-milligram doses of haloperidol to human volunteers. Rather, HVA levels declined during the period of absorption and attainment of peak haloperidol levels. Baseline HVA levels of 51.6 +/- 3.83 pmoles/ml and 56.8 +/- 5.70 pmoles/ml (after 4 mg and 10 mg., respectively) declined to minima of 35.6 +/- 1.67 pmoles/ml and 26.3 +/- 5.34 pmoles/ml respectively, at 3-4 hours after haloperidol administration. A trend was noted for the 10-mg dose to produce a greater decline than the 4-mg dose, which was most apparent at 4 hours after drug administration. The shape of both curves did not appear to be substantially different than expected on the basis of diurnal variation. These preliminary findings support the concept that dopamine turnover in humans is not increased and may be decreased by short-term administration of conventional neuroleptics.

Is the mechanism of prefrontal hypofunction in depression the same as in schizophrenia? Regional cerebral blood flow during cognitive activation.

To test the hypothesis that depressed and schizophrenic patients have a common pathophysiological mechanism for hypofunction of the prefrontal cortex ('hypofrontality'), we measured regional cortical blood flow (rCBF) in ten depressed patients, ten patients with schizophrenia, and 20 age- and sex-matched normal controls. Blood flow was measured during three different cognitive conditions: a resting state, a simple number-matching sensorimotor control task, and the Wisconsin Card Sorting test (WCS). The schizophrenic patients had lower prefrontal rCBF during the WCS. There were no differences in global or regional flow between the depressed patients and the normal subjects during any testing condition. Analysis of rCBF lateralisation showed that during the WCS normal subjects had relatively more left parietal blood flow than depressed patients, who had more right parietal blood flow. Since the testing condition that has most consistently revealed hypofrontality in schizophrenia (i.e. the WCS) was not associated with abnormal rCBF in the depressed patients, these data suggest that the pathophysiological mechanisms underlying prefrontal hypofunction in depression and schizophrenia are different.

Cerebrospinal fluid immunoreactive somatostatin concentrations in patients with Cushing's disease and major depression: relationship to indices of corticotropin-releasing hormone and cortisol secretion.

To further explore the differential effects of peripherally and centrally derived hypercortisolism on neurohormonal systems implicated in the pathophysiology of mood and cognitive disturbances, we examined the cerebrospinal fluid (CSF) concentrations of immunoreactive somatostatin (IR-SRIF) in patients with Cushing's disease and major depression and the relationship of these levels to CSF immunoreactive corticotropin-releasing hormone (CRH) concentrations and urinary free cortisol excretion. In particular, since CSF SRIF levels consistently have been shown to be reduced in depression, we wished to assess whether decreased centrally directed SRIF was more likely a primary or a secondary factor in the hypercortisolism of major depression. CSF SRIF levels were significantly reduced in 11 patients with documented Cushing's disease and in 1 patient with ectopic adrenocorticotropic hormone secretion as compared with both 41 healthy volunteers (19.4 +/- 2.9 vs. 37.4 +/- 1.5 pmol/l; p < 0.01) and 28 patients with major depression (30.2 +/- 2.4 pmol/l; p < 0.05), whose CSF SRIF levels were also significantly reduced as compared with controls (p < 0.05). CSF SRIF levels in the Cushing's disease patients correlated positively with CSF CRH (r = 0.64; p < 0.025), suggesting that either the sustained hypercortisolism in these patients and/or its suppression of central CRH secretion contributed to the reduction in SRIF. A more modest but significant correlation between CSF SRIF and CSF CRH was observed in the healthy volunteers (r = 0.37; d.f. = 37; p < 0.02); in the depressed patients, no linear relationship, but rather an inverted U-shaped relationship was found which significantly fit by a quadratic function (r2 = 0.90; d.f. = 22; p < 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)

Quantitative electroencephalographic correlates of steroid administration in man.

Elevated levels of circulating corticosteroids are frequently associated with behavioral alterations in man, although the mechanisms by which corticosteroids may affect behavior are poorly understood. To evaluate possible effects of exogenous corticosteroids on brain electrophysiological functioning and the relationship of such effects to behavioral and biochemical changes, we administered prednisone (80 mg p.o. daily for 5 days) in a double-blind manner to 11 medically healthy volunteers. Quantitative electroencephalogram analysis was performed following 4 days of prednisone administration and during the preceding and ensuing placebo administration periods. Central theta wave brain electrical activity significantly increased following prednisone administration and returned to baseline following prednisone withdrawal. This effect was directly correlated with prednisone-induced increases in subjective sadness ratings and with decreases in self-rated energy and well-being. Prednisone-induced reductions in peak alpha wave activity were also directly correlated with increases in subjective sadness and Symptom Checklist-90 ratings and with decreases in self-rated 'hypomanic' symptoms. Further, prednisone-induced increases in theta activity were significantly correlated with prednisone-induced decreases in CSF levels of somatostatin-like immunoreactivity, and prednisone-induced decreases in peak alpha activity were significantly correlated with decreases in CSF levels of norepinephrine and with relative increases (or lesser decreases) in CSF levels of beta-endorphin and beta-lipotropic hormone. This preliminary report of the concomitant development of prednisone-induced changes in brain electrical activity, neurochemistry and behavior highlights areas for future exploration in the study of corticosteroid effects on behavior in man.

Beta receptor density in human lymphocyte membranes: changing with aging?

The number of beta adrenergic binding sites (Bmax) in human lymphocyte membranes has been reported to decrease, remain the same, or increase with age. In order to address this issue, we used two highly specific beta receptor ligands with lymphocytes from healthy aged (range: 51-90 years) and young (19-39 years) subjects in two separate studies. Because depression can reduce Bmax, potential aged subjects were excluded if they had high scores on tests for depression. Bmax was higher in the aged group of each study (33% higher in the first, p less than .01, and 72.5% in the second, p less than .02); the results were similar in both studies. Antagonist affinity did not differ between young and aged groups in either study. We suggest that some of the discrepancies in the literature could be due to differences in age ranges used or to inclusion of depressed subjects in prior studies.

Cerebrospinal fluid immunoreactive corticotropin-releasing hormone and adrenocorticotropin secretion in Cushing's disease and major depression: potential clinical implications.

To explore whether possible differences in central nervous system neuromodulators contribute to the differential presentation of affective symptomatology in Cushing's disease and major depression, we examined the levels of immunoreactive CRH and ACTH in the cerebrospinal fluid (CSF) of 11 patients with Cushing's disease, a patient with ectopic ACTH secretion, 34 patients with major depression, and 60 healthy subjects. We elected to measure these peptides not only because both are classically involved in pituitary-adrenal regulation, but also because their primarily arousal-producing and anorexigenic behavioral effects in experimental animals suggest that they may play a role in the symptom complex of depressive syndromes. We also explored whether the CSF levels of these peptides were more helpful in determining the often difficult differential diagnosis between major depression and Cushing's disease than the plasma ACTH response to ovine CRH, a currently used but somewhat insensitive laboratory means of distinguishing these disorders. CSF levels of immunoreactive CRH and ACTH were significantly lower in Cushing's disease patients [21.9 +/- 2.7 and 15.4 +/- 1.8 pg/mL, (mean +/- SEM), respectively] compared to patients with major depression [38.4 +/- 2.3 pg/mL (P less than 0.01) and 24.5 +/- 1.6 pg/mL (P less than 0.01), respectively] and controls [38.4 +/- 1.6 pg/mL (P less than 0.001) and 26.3 +/- 1.1 pg/mL (P less than 0.001), respectively]. The coexistence of high plasma ACTH and low CSF ACTH in Cushing's disease yielded a CSF/plasma ACTH ratio consistently less than that in depressed patients, with only 2 of 31 subjects comprising both groups showing values that overlapped. In contrast, 9 of the combined patients showed ACTH responses to ovine CRH that overlapped. These data suggest that differences in centrally directed CRH secretion may account for the differential presentation of the dysphoric syndromes seen in major depression and Cushing's disease. Hence, the classic form of major depression (melancholia), is often associated with evidence of pathological hyperarousal, such as intense anxiety, sleeplessness, and anorexia, while that of Cushing's disease is associated with evidence of pathological hyperarousal, including hyperphagia, fatigue, and inertia. Moreover, measurement of the CSF/plasma ACTH ratio may serve as a clinically useful adjunct to the ovine CRH stimulation test and other laboratory measures in determining the differential diagnosis between major depression and Cushing's disease.

Prednisone effects on neurochemistry and behavior. Preliminary findings.

To evaluate the neurochemical, neuroendocrine, and behavioral effects of exogenous corticosteroids in humans, we administered prednisone (80 mg/d orally for 5 days) in a double-blind manner to 12 medically healthy volunteers. Behavioral measures were assessed before, during, and after prednisone administration in all 12 subjects, and cerebrospinal fluid biochemistry was assessed before and during prednisone administration in 9 of the subjects. Prednisone administration was associated with decreases in cerebrospinal fluid levels of corticotropin, norepinephrine, beta-endorphin, beta-lipotropin, and somatostatinlike immunoreactivity. No significant changes were noted in cerebrospinal fluid levels of alpha-melanocyte-stimulating hormone, corticotropin-releasing hormone, 3-methoxy-4-hydroxyphenylglycol, homovanillic acid, or 5-hydroxyindoleacetic acid. No consistent or significant group mean changes were observed in structured behavioral ratings, although 9 (75%) of the volunteers studied reported mild behavioral changes while receiving prednisone. Correlations between the neurochemical and behavioral changes are discussed.

Cerebrospinal fluid and plasma monoamine metabolites and their relation to psychosis. Implications for regional brain dysfunction in schizophrenia.

The relationship between central (cerebrospinal fluid [CSF]) and peripheral (plasma) monoaminergic metabolites and psychotic symptoms was examined in 22 drug-free schizophrenic inpatients. The CSF homovanillic acid levels did not differ significantly between patients and normal controls (n = 33). The CSF homovanillic acid levels, however, were negatively correlated with ratings of psychosis and positive symptoms, and the CSF homovanillic acid and 5-hydroxyindoleacetic acid levels correlated negatively with individual deficit symptoms. Stepwise and hierarchical multiple-regression analysis revealed that among monoaminergic measures, only the CSF and plasma homovanillic acid levels contributed significantly to the total Brief Psychiatric Rating Scale and positive symptom variance with negative and positive partial correlations, respectively. Levels of CSF 3-methoxy-4-hydroxyphenylglycol, but not of CSF norepinephrine, were significantly elevated in the schizophrenic patients compared with controls, and plasma 3-methoxy-4-hydroxyphenylglycol levels were positively correlated with negative symptoms. We discuss the potential implications of these findings for a model of dopaminergic dysfunction in schizophrenia involving distinct cortical and subcortical contributions.

Circadian variation of plasma homovanillic acid levels is attenuated by fluphenazine in patients with schizophrenia.

Plasma homovanillic acid (HVA) levels were measured hourly for a 24-hour period in 10 patients with schizophrenia during treatment with placebo and fluphenazine. Ten age- and sex-matched normal volunteers were similarly studied. Diet and activity were carefully controlled and monitored in both patients and controls. A circadian rhythm of the plasma HVA level was found in controls with a nadir in the afternoon and peak values in the early morning hours; when the patients were free from drugs, they showed a similar rhythm with lower amplitudes. Fluphenazine treatment significantly reduced the plasma concentrations of HVA and abolished the 24-hour rhythm. These data suggest that a 24-hour rhythm of the plasma HVA level exists in humans and that the amplitude of this rhythm may be less pronounced in patients with schizophrenia. Treatment with neuroleptic drugs reduces both the absolute levels and the normal circadian rhythm of the plasma HVA level.

Prednisone effects on blood-brain barrier permeability and CNS IgG synthesis in healthy humans.

Corticosteroids reportedly decrease blood-brain barrier (BBB) permeability and/or IgG synthesis in patients with multiple sclerosis or brain tumors. However, these effects have not been studied in healthy humans. We investigated the effects of prednisone, 80 mg/day for five days, on the ratio of cerebrospinal fluid (CSF) albumin/serum albumin, a measure of blood-brain barrier (BBB) permeability, and on CSF and serum IgG levels in six healthy, normal volunteers. We found significant steroid-induced decreases in serum and CSF albumin levels and in serum IgG levels. However, we found only a nonsignificant decrease in BBB permeability and no significant change in CNS IgG synthesis. These findings, based on a small number of volunteers, suggest that it may be difficult to further decrease BBB permeability and CNS IgG synthesis in medically healthy subjects.

Lymphocyte beta-adrenoreceptor density in patients with unipolar depression and normal controls.

Values of binding maximum (Bmax) and dissociation constant (Kd) of (-)3-[125I]iodocyanopindolol (ICYP) were determined in beta-adrenergic receptors of membranes of peripheral lymphocytes in 32 patients with unipolar depression (DSM-III-R) and 31 normal controls. Results were analyzed by a two-way Analysis of Covariance method. A significant difference was noted for group assignment (patient versus control, p less than 0.05). Mean Bmax (fmol ICYP bound/mg lymphocyte membrane fraction total protein) of patients was 31.9 +/- 3.84 (SE) and controls 46.3 +/- 3.92 (SE). A significant interaction was found between group membership and gender (p less than 0.05). In the female patient group (n = 14), mean Bmax was 30.5 +/- 5.79 (SE); in female controls, mean Bmax was 56.0 +/- 5.15 (SE). Differences between male patients and male controls were not significant. Mean values of Kd (pmol/liter) showed a trend for patient values to be lower than control values [69.0 +/- 13.66 (SE) versus 108.5 +/- 14.42 (SE), respectively]. A significant inverse relationship was noted between lymphocyte beta-receptor Bmax and frequency of panic attacks during the depressive episode in 18 patients (p = 0.05). No relationship was found between values of Kd and frequency of panic attacks in these patients. Thus, preliminary evidence is provided for relationships among altered beta-adrenergic receptor binding, gender, and indices of panic-anxiety in unipolar depressed patients.

Fluphenazine treatment reduces CSF somatostatin in patients with schizophrenia: correlations with CSF HVA.

CSF somatostatin and homovanillic acid (HVA) were measured in 14 schizophrenic patients while they were drug-free and during chronic fluphenazine treatment. CSF somatostatin was significantly reduced and CSF HVA was significantly elevated (p less than 0.002) during fluphenazine treatment. There was a trend toward correlation between CSF somatostatin and CSF HVA in the 14 schizophrenic patients when drug-free (r = 0.49, p less than 0.07) and fluphenazine-treated (r = 0.47, p less than 0.08). When examined in a larger group (n = 46) of drug-free schizophrenics, this relationship was highly significant (r = 0.59, p less than 0.001). These clinical data are consistent with preclinical evidence indicating a functional interaction between CNS somatostatin and dopamine systems.

Steady-state cerebrospinal fluid transfer of verapamil and metabolites in patients with schizophrenia.

Cerebrospinal fluid (CSF) and plasma levels of verapamil and its two metabolites, norverapamil and D-620, were measured in seven patients with schizophrenia under steady-state conditions. Simultaneous sampling of CSF and plasma just before the dose during week 4 of the trial showed that verapamil, norverapamil, and D-620 partition in the CSF and reflect 7%, 5%, and 12% of the corresponding levels in plasma, respectively. There was a significant decrease in the mean unbound fraction of verapamil in schizophrenic patients as compared with normal subjects (0.058 vs. 0.11; p less than 0.001). Estimates of the mean unbound fraction obtained from CSF/plasma verapamil concentrations and the pH partition hypothesis showed excellent agreement with that measured by equilibrium dialysis (0.055 vs. 0.058) in these patients. Although systemic pool protein concentrations in schizophrenic patients were within normal range, an excellent positive correlation was observed between the ratio of the bound/free verapamil concentration and alpha 1-acid glycoprotein levels (r = 0.86; p less than 0.05). Determination and development of correlations between plasma and CSF may enhance our understanding of the central nervous system effects of verapamil.

Neurobiological effects of lumbar puncture stress in psychiatric patients and healthy volunteers.

Several existing laboratory-based stress paradigms have significant shortcomings for assessing neurobiological correlates of stress in healthy volunteers and severely ill psychiatric patients. We have examined neuroendocrine effects of stress associated with the lumbar puncture (LP) procedure in drug-free depressed and schizophrenic patients and healthy volunteers. Healthy volunteers and depressed patients had significant stress-induced elevations in plasma levels of adrenocorticotropic hormone (ACTH), cortisol, and growth hormone associated with LP. In contrast, schizophrenic patients had no significant elevations in any of the neuroendocrine parameters. Depressed patients' cerebrospinal fluid norepinephrine levels were negatively correlated with stress-induced activation of the hypothalamic-pituitary-adrenal axis. In addition, depressed patients' basal plasma cortisol levels were strongly predictive of stress-related elevations in ACTH, cortisol, and growth hormone. Furthermore, stress-induced changes in schizophrenic patients' cortisol and ACTH levels were not correlated.

Cerebral structural pathology in schizophrenia: evidence for a selective prefrontal cortical defect.

To replicate earlier findings of central and cortical cerebral structural abnormalities and to further examine specific prefrontal cortical irregularities in schizophrenia, the authors examined computerized tomographic (CT) scans of 71 patients with chronic schizophrenia and 30 normal volunteer control subjects for ventricle-brain ratio (VBR), third ventricle width, and prominence of cortical markings in a generalized (parieto-occipital) distribution compared with the prefrontal area. Patients showed significantly larger VBRs and third ventricle widths than control subjects. Patients also showed significantly greater prefrontal markings in terms of both the number of individuals affected and the degree of difference. The relative differences on measures of VBR and prefrontal atrophy did not appear gender related, but the difference in third ventricle width between schizophrenic and control women was greater than that between schizophrenic and control men.

The diurnal variation in plasma homovanillic acid level persists but the variation in 3-methoxy-4-hydroxyphenylglycol level is abolished under constant conditions.

Plasma concentrations of homovanillic acid (HVA) and 3-methoxy-4-hydroxyphenylglycol (MHPG) were determined every two hours for two consecutive days in seven depressed patients and seven matched normal controls. On the first day subjects followed their regular ward routine. On the second day they were placed on a regimen in which activity, posture, diet, and wakefulness were held constant. There were significant diurnal variations in both MHPG and HVA concentrations on the baseline day, whereas on the constant routine, a diurnal variation was present only in HVA. We conclude that diurnal variations in plasma MHPG are evoked by changes in physical activity, posture, or other factors controlled on the constant routine, and that a major component of the diurnal variation in plasma HVA concentrations is regulated by a circadian oscillator that is independent of sleep or activity.

Neuroleptic responsivity of negative and positive symptoms in schizophrenia.

The authors prospectively examined the effects of double-blind, placebo-controlled neuroleptic withdrawal and administration on ratings of negative and positive symptoms in 19 young patients with chronic schizophrenia. Negative symptoms were significantly reduced by neuroleptic treatment, and negative and positive symptoms demonstrated similar patterns of reduction and exacerbation during neuroleptic treatment and withdrawal, respectively. The changes in negative and positive symptoms induced by neuroleptic treatment and withdrawal were not significantly correlated, however. The negative and positive symptom profiles of individual patients were significantly altered by neuroleptic treatment, indicating limitations to the cross-sectional classification of patients on the basis of predominance of one or the other symptom group. The authors discuss implications for the neurobiological underpinnings of negative and positive symptoms.