Main

Plasma levels of homovanillic acid (pHVA), a major metabolite of dopamine (DA), have been used as a peripheral measure of dopaminergic activity in the central nervous system (Bacopoulas et al. 1979; Kendler et al. 1981; Kendler et al. 1982). Although a relatively small part of pHVA has been indicated as central origin (Kopin et al. 1988; Maas et al. 1988; Lambert et al. 1993), there have been reports of the usefulness of pHVA as a peripheral indicator of central dopaminergic activity (Pickar et al. 1990; Stroe et al. 1997). Thus, a vast number of studies have been conducted to relate pHVA levels with some aspects of such mental disorders as schizophrenia. For example, the ability of pHVA levels to predict response to treatment with typical (Chang et al. 1993; Davilia et al. 1988; Davis et al. 1985; Pickar et al. 1984; Suzuki et al. 1994) and atypical (Green et al. 1993; Sumiyoshi et al. 1997a) neuroleptics has been tested in subjects with schizophrenia.

Diet (Kendler et al. 1983), physical activity (Kendler et al. 1983), circadian fluctuation (Zumarraga et al. 1993) or gender difference (Sumiyoshi et al. 1997b) have been suggested to affect pHVA concentrations. However, there has been little study of the effect of mental stress on pHVA levels. Previous studies report altered blood or urinary levels of catecholamines attributable to mental stress in humans (Januszewicz et al. 1979; Frankenhaeuser et al. 1985; Rauste-von Wright and Frankenhaeuser 1989). Changes in DA-dependent behaviors attributable to mental stress have been reported in laboratory animals (Friedhoff et al. 1995). These findings suggest a significant effect of mental stress on pHVA levels.

The present study was undertaken to test the hypothesis that mental stress resulting from the Kraepelin's arithmetic test would significantly decrease pHVA levels in psychiatrically well normal subjects.

SUBJECTS AND METHODS

Thirteen male psychiatrically normal healthy volunteers (mean ± SD age = 29.0 ± 3.1; range 24–34 years), who gave written consent to the study were investigated. DSM-IV Axis I disorders were ruled out for these subjects using a structured interview by experienced psychiatrists taking medical histories. Physical exercise, high monoamine diets, alcohol, and caffeine were restricted for all subjects from the start of the study (>24 h before blood sampling).

The subjects performed a modified version of Kraepelin's arithmetic test (Uchida–Kraepelin test) (Kuraishi et al. 1957), a test of continuous arithmetic addition for 30 min, which has been used in neurophysiological (Mizuki et al. 1980; Takii 1986) and neuropharmacological (Murasaki et al. 1989) studies. Pretest blood sampling was performed through an intravenous line at 9:00 A.M. on the day of investigation following an overnight fast. The subjects rested after the venous line was inserted. Immediately after the pretest blood sampling, the subjects began to add each of overlapping sequential pairs of single digit numbers (the second number of one pair becomes the first number of the following pair). The last digit of the sum of each pair was written in a space just below each pair. The numbers on the test sheet were written in such a manner that the subjects changed to a new line of numbers every 1 min. Immediately after termination of arithmetic addition for 30 lines (30 min after the pretest blood sampling), blood samples were again collected for measurement of the posttest pHVA levels. Plasma was separated by centrifugation and stored at −80°C until the free HVA was measured.

Plasma HVA analysis was carried out by high-performance liquid chromatography (HPLC) with coulometric detection using d.l.-isoproterenol as internal standard. The coefficients of variation (CV) for within day (n = 6) and between day (n = 6) analyses of pHVA were 2.4 and 3.3%, respectively.

Pre- and post-Kraepelin test pHVA levels were compared by nonparametric Wilcoxon signed rank test. Spearman's correlation coefficients were obtained relating pretest pHVA levels with percentage changes in pHVA levels as well as with changes in pHVA values (calculated by subtracting post-test pHVA levels from pretest pHVA levels). Differences were considered to be significant when p-values were less than .05.

RESULTS

Figure 1 shows pHVA levels in blood samples collected before and after the administration of the Kraepelin test. The post-test value of pHVA levels was significantly less than the pretest value (p = .0046).

Figure 1
figure 1

Plasma homovanillic acid levels (mean ± SEM ng/ml) before and after Kraepelin's arithmetic test, *p < .01.

Full size image

Figure 2 demonstrates the correlation between the pretest pHVA levels and percentage changes in pHVA levels for each subject. The percentage change in pHVA levels following the Kraepelin test correlated positively with pretest pHVA levels (rs = 0.63, p = .029). Changes in the pHVA levels themselves during the Kraepelin test also correlated with pretest pHVA levels (rs = 0.74, p = .011; data not shown).

Figure 2
figure 2

Correlation between pre-Kraepelin test pHVA levels and percentage changes in pHVA levels (ng/ml).

Full size image

DISCUSSION

The present study provided the first evidence that mental stress by use of Kraepelin's arithmetic test decreases pHVA levels in healthy subjects (Figure 1). The significant effect of mental stress on pHVA levels may be one of the factors responsible for the inconsistent results in previous studies reporting increased, decreased, or normal pHVA levels in subjects with schizophrenia as compared to normal controls (see Sumiyoshi et al. 1997b for review).

Previous studies investigating the effect of physical or mental stress generally have focused on changes in blood or urinary levels of adrenaline, noradrenaline (NE) or their metabolites (Halbrugge et al. 1988; Januszewicz et al. 1979). The observed decrease in pHVA levels caused by mental stress could be explained by both central and peripheral events. Increased urinary HVA and DA concentrations have been found after mental or emotional stress (Fibiger and Singer 1984; Frankenhaeuser et al. 1985; Rauste-von Wright and Frankenhaeuser 1989). The present version of Kraepelin's arithmetic test (Uchida–Kraepelin test) requires intellectual ability to the extent that subjects taking diazepam (20 mg/day) show marked deterioration in performance (Murasaki et al. 1989). Januszewicz et al. (1979) examined the effect of Kraepelin's arithmetic test on blood and urinary catecholamine concentrations. In that study, increased NE and 3-methoxy-4-hydroxyphenylglycol concentrations in urine as well as increased blood NE levels were observed in healthy subjects following Kraepelin's test (Januszewicz et al.). Thus, it is possible that similar mechanism(s) (e.g., changes in urinary excretion) underlying altered output of these catecholamines and metabolites may, in some way, be associated with the decrease in pHVA levels in subjects suffering from mental stress.

Changes in the central dopaminergic activity by mental stress may also be a factor to explain the observed decrease in pHVA levels. Dopaminergic activity in subcortical brain regions such as nucleus accumbens has been suggested to contribute to patterns of pHVA change (Csernansky et al. 1990). Dopamine utilization expressed as HVA/DA concentrations in nucleus accumbens is decreased in rats receiving chronic witness stress (Friedhoff et al. 1995). Also in line with the present result is a recent finding that mental stress is associated with a reduction in extracellular DA release in the rat striatum (Katoh et al. 1996). Friedhoff (1986, 1988) and Friedhoff et al. (1995) have proposed the presence of a DA-dependent protective system in the brain that maintains mental stability in the face of biological or psychological insult. They predicted that a role of the dopaminergic system is to reduce the risk of mental decompensation (disruption of mental stability in the face of changing environmental contingencies) during stress by decreasing its own activity. They proposed this occurs in normal subjects or a part of schizophrenic patients and would be reflected in pHVA levels as was observed here (Friedhoff 1986, 1988; Friedhoff et al. 1995). Thus, the reduction in pHVA levels by mental stress reported here is consistent with the hypothesis of Friedhoff et al. (1995). In this context, the positive correlation between baseline (pretest) pHVA levels and percentage changes in pHVA levels during the mental stress test (Figure 2) may suggest that subjects with higher pHVA level have a greater ability to cope with psychological stress. This finding may have relevance to the reported correlations between higher baseline (pretreatment) pHVA level and better therapeutic outcome in schizophrenic patients who receive neuroleptic treatment (Chang et al. 1993; Green et al. 1993; Sumiyoshi et al. 1997a). To further clarify the clinical significance of the present results, studies investigating the effect of mental stress on pHVA levels in subjects with schizophrenia are in progress.

The study design did not employ a control group. Because the version of Kraepelin test used here takes only 30 min, it is unlikely that significant change in pHVA would occur in the absence of stress. The stability of pHVA levels has been established in normal subjects (Kendler et al. 1983) and patients with schizophrenia (Donnelly et al. 1996) based on data from plasma sampling over a several hour period including the morning. However, inclusion of data from unstressed subjects could have ruled out a possible change in pHVA related to time.

In conclusion, we have reported a reduction in pHVA levels by mental stress in psychiatrically normal subjects. This finding may reflect some aspects of a DA-dependent restitutive system in the brain, although peripheral contribution to the decrease in pHVA levels should also be considered.