Attention deficit disorder symptoms and urine catecholamines

Psychiatry Research. 21,241-25 I

241

Elsevier

Attention Deficit Disorder Symptoms Catecholamines

and Urine

Graham A. Rogeness, James W. Maas. Martin A. Javors, Carlos A. Macedo, Charles Fischer, and William R. Harris Received June 26, 1987; revised version received October received May 13, 1988; accepted June 30, 1988.

23, 1987; second

revised version

Abstract. The symptoms of hyperactivity, impulsivity, and concentration deficits associated with attention deficit disorder (ADD) may be related, in part, to alterations in dopaminergic and noradrenergic functioning. In this study we correlate the above symptoms with 24-hour urinary catecholamines and their metabolites in emotionally disturbed boys divided into two groups based on their plasma dopamine-&hydroxylase (DBH) activities and also divided into the following diagnostic groups: conduct disorder, undersocialized; conduct disorder, socialized; and subjects without conduct disorder. Boys in the low DBH group showed significant correlations between the ADD symptoms and the biochemical measures. The low DBH group may be more genetically homogeneous with regard to catecholamine function, making relationships between catecholamine function and behavior more visible. The group of boys with conduct disorder, socialized had higher 24-hour urinary norepinephrine and vanillylmandelic acid output. The relationship between monoamines and their metabolites appeared to differ among diagnostic groups. Key Words. Attention dopamine-&hydroxylase,

deficit disorder, norepinephrine.

catecholamines,

conduct

disorder,

Biological research in psychiatry has generally been approached by attempting to define “pure” diagnostic groups of subjects and then looking for biological differences between the diagnostic group and a comparison group. Many diagnostic groups are heterogeneous, and actual differences may be obscured by this heterogeneity. The classification of depressive disorders into subtypes has revealed biological differences between subtypes that would not be evident if all depressive disorders were grouped together. Buchsbaum and Rieder (1979) have suggested an alternative strategy-the “biochemical high risk strategy”-using a biochemical variable as an independent variable and psychopathology as the dependent variable. We have used this approach and have found an increase in the diagnosis of conduct

Graham A. Roaeness, M.D., is Clinical Professor; James W. Maas, M.D., is Professor; Martin A. Javors, Ph.D., is Assistant Professor; and Carlos A. Macedo, M.D., Charles Fischer, M.D., and William R. Harris. M.D.. are Clinical Assistant Professors, Department of Psychiatry, The University of Texas Health Science Center at San Antonio. Drs. Rogeness, Macedo, Fisher,.and Harris are-also Child Psychiatrists at Southwest Neuropsychiatric Institute. (Reprint requests to Dr. G.A. Rogeness, Dept. of Psychiatry, The University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Dr., San Antonio, TX 78284-7792, USA.) 01651781/89/$03.50

@ 1989 Elsevier Scientific

Publishers Ireland

Ltd.

242 disorder, undersocialized (CDU) in emotionally disturbed boys with very low plasma dopamine-fi-hydroxylase (DBH) activities (Rogeness et al., 1984, 1986, 1987, 1988). In this article, we examine biochemical-behavioral relationships on the basis of both biochemical and clinical subtyping in an attempt to gain increased understanding of biochemical-behavioral relationships. DBH is an enzyme involved in the conversion of dopamine to norepinephrine. Its concentration in plasma is genetically determined and relatively constant in an individual over time from about age 6 or 8 years (Weinshilboum et al., 1973; Weinshilboum, 1983). However, there is marked interindividual variation in plasma DBH activity. Ciaranello and Boehme (198 I) hypothesize that one gene controls the phenotypic expression of the three catecholamine-synthesizing enzymes. It is possible, then, that a low DBH group could be more homogeneous in their catecholamine function and that this homogeneity could lead to different or more visible relationships between catecholamine measures and behavioral measures. The symptoms of attention deficit disorder (ADD)-hyperactivity, impulsivity, and concentration deficits-may be related, in part, to alterations in dopaminergic and noradrenergic functioning (Shekim et al., 1983). Biochemical studies comparing boys with ADD to those without have shown inconsistent results. Shekim et al. ( 1979, 1983) found lower 24-hour urinary 3-methoxy-4-hydroxyphenyglycol (MHPG) and homovanillic acid (HVA) in boys with ADD. Other investigators have found no differences in these metabolites (Wender et al., 1971; Rapoport et al., 1978) or higher levels of MHPG (Kahn and Dekirmenjian, 1981). Rapoport et al. found higher 24-hour urinary norepinephrine (NE) in boys with ADD in one study (Rapoport et al., 1970) but not in another (Rapoport et al., 1978). It is likely that ADD has a heterogeneous etiology. A variety of neurotransmitter dysfunctions and a variety of neuropsychological deficits could contribute to the symptoms of the disorder. Also, many children with ADD symptoms have other psychopathological disorders such as conduct disorders, depressive disorders, and psychotic disorders. The usual strategy to find associations between ADD and catecholamine function is to compare a diagnostically “pure” group of ADD subjects with normal subjects. An alternative strategy would be to consider the symptoms of hyperactivity, impulsivity, and concentration deficits as having a similar biochemical etiology, regardless of the diagnosis, and look for relationships between the symptoms and measures of catecholamine function. If one had a sample of subjects that was more genetically homogeneous, one might increase the likelihood of finding relationships between catecholamines and ADD symptoms. In this preliminary study we show correlations between 24-hour urinary catecholamines and their metabolites and measures of ADD symptoms in boys hospitalized in a psychiatric hospital. The specificity or visibility of the correlations is examined by grouping by plasma DBH activity and then by diagnostic group. The clinical sample is divided into three groups: CDU; conduct disorder, socialized (CDS); and neither CDU nor CDS (Neither). Conduct disorder is likely a heterogeneous disorder with a variety of different etiologies. CDU and CDS are DSM-III(American Psychiatric Association, 1980) subtypes of conduct disorder that have shown biological and clinical

243 differences in our studies. Quay (1986) has hypothesized decreased noradrenergic functioning in the CDU type. Biochemical differences between groups are examined in two ways: (1) The 24hour urinary values are compared among the CDU, CDS, and Neither groups. (2) The correlation matrixes for the low and high DBH groups and for the CDU, CDS, and Neither groups are calculated and compared. Roy et al. (1988) found high correlations between catecholamine measures in middle-aged depressed patients and controls. Koslow et al. (1983) found less robust correlations in a similar population with variability in the correlations when patients were broken down by diagnostic group. The variation between diagnostic groups could be by chance or could reflect differences in metabolism between diagnostic groups that become visible when relationships between amines and their metabolites are studied that are not apparent when a specific amine or metabolite is compared between groups. One can make hypotheses about possible differences in amine relationships in the low DBH group compared to the high DBH group, and if the relationships are theoretically consistent, it may suggest that this approach has some validity. Differences in relationships between the CDU and CDS groups may then suggest biochemical differences between the two groups, and hypotheses about the meaning of these differences can be proposed for further study.

Methods Subjects. Patients studied were boys hospitalized at a private nonprofit children’s psychiatric hospital. The primary purpose of the study was to compare boys with low plasma DBH activity (G 6 puM/min/l) with boys with higher DBH activity (2 I5 PM/ min/ I) to determine if there were biochemical and clinical differences between the two groups. The study population therefore consisted of two groups of boys with low and high DBH activities. There were 24 boys in the low group and 26 boys in the high group on whom 24-hour urinary catecholamine data were available. Three subjects, two in the low and one in the high group, did not have norepinephrine (NE), epinephrine (EP), or dopamine (DA) measured. Subjects had the range of diagnoses found in an inpatient population, and most subjects had multiple diagnoses. Diagnostic data and 24-hour urine values in the high and low DBH groups are reported elsewhere (Rogeness et al., 1988). The mean age of the subjects was I I. I (SD 2.7) and the mean IQ was 99 (SD 16). There was no significant difference in age or IQ between the two groups. Subjects were medication free for 2 or more weeks before urines were collected. There were 18 subjects in the CDU group, 20 in the CDS group, and 12 in the Neither group. The number of low DBH subjects was not equally distributed among groups. The number of low DBH subjects in each of these groups was 14, 6, and 4, respectively. Their respective mean ages were 10.0 (SD 2.3) I I .7 (SD 2.7) and I I .6 (SD 2.9). The CDU group was nonsignificantly younger (F = 2.4; df= 2, 47; p = 0. IO). Clinical Measures. There were two measures of ADD symptoms. The attending child psychiatrist completed a DSM-III checklist on the subject (Rogeness et al., 1986) rating a symptom present or absent. If there was a significant past history of the symptom, it was rated as present even though it may not have been currently prominent. The items for concentration and hyperactivity were taken verbatim from DSM-IIf. The number of items rated present was summed. After a subject was entered into the study, the child-care workers on the day and evening shift were asked to complete Conners’ Behavioral Problem Checklist for Parents (CBPC) ‘(Conners, 1973) based on the subject’s behavior on the unit during his first l-2 weeks of

244 hospitalization. A single score for each item was obtained by averaging the scores from the day and evening shift. The impulsive/ hyperactive factor was one of eight factors derived from the checklist (Conners, 1973). Correlations (Pearson’s) with the biochemical data were controlled for age by treating age as a covariate. Biochemical Measures. Blood for DBH measurement was drawn from patients within l-4 days of admission when other routine blood work was being done. Plasma DBH was measured photometrically by the method described by Nagatsu and Udenfriend (1972). Twenty-four hour urines were collected on 3 consecutive days on subjects who were medication free for at least 2 weeks. The urine was collected in opaque plastic jugs containing 700 mg sodium metabisulfite. The urine was kept on ice during the collection. Urines were considered complete on the basis of the following criteria: (I) The staff collecting the urines considered them complete. (2) If two urines had similar creatinine, and the 24-hour creatinine fell within two standard deviations of the mean expected creatinine, the urine collection was considered complete. (3) If all three creatinines were dissimilar but one of the 24-hour urines fell within one standard deviation of the mean, it was considered complete. The formula used to determine expected creatinine was from Barratt and Chantler (1975). Our creatinine methodology initially resulted in lower than expected creatinine values, and urines that did not meet the low criterion, but had a high level of staff confidence that they were complete, were included as complete. There was no significant difference in creatinine between the high and the low DBH groups, 0.69 (SD 0.40) g and 0.70 (SD 0.39) g, respectively. Fifty subjects, 24 in the low group and 26 in the high group, were considered to have complete 24-hour urines. Twenty of these consisted of one complete urine, and 30 were the average of two or three complete urines. Diet was not controlled. Catecholamines and their metabolites were measured using high performance liquid chromatography with electrochemical detection. NE, EP, and DA were measured by the method described by Smedes et al. (1982). Vanillyl-

mandelic acid (VMA), MHPG, and HVA and 5hydroxyindoleacetic measured by the method described by Joseph et al. (198 I).

acid (SHIAA) were

Results Table 1 shows the correlation of the 24-hour urinary amines and metabolites with age. Positive correlations are present in the combined group for NE, EP, MHPG, VMA, and HVA. The correlations in the low DBH group are stronger than in the high DBH group, with only MHPG and VMA positively correlated with age in the high group. In the comparisons in the rest of this article, the 24-hour urine values are expressed in 24 hour/m*. This decreases the influence of age and physical size. When the values are expressed per m*, there are no significant correlations with age in the combined group except for SHIAA (-0.26). In the low group, only the correlation

Table 1. Correlations of 24-hour urinary amine measures (pg/24 hours) with age Combined DBH I 6 DBH I 15

NE

EP

0.52’ 0.70’

0.20’ 0.64’

0.29

-0.17

DA 0.19

0.66’ -0.13

MHPG

VMA

HVA

SHIAA

0.46’ 0.59

0.56’ 0.67’

0.35’ 0.50

0.22 0.25

0.36’

0.39’

0.18

0.15

Note. NE = norepinephrine. EP = epinephrine. DA = dopamine. MHPG = 3-methoxy-4-hydroxyphenylglycol. VMA = vanillylmandelic acid. HVA = homovanillic acid. 5HIAA = Shydroxyindoleacetic acid. DBH = dopamineP-hydroxylase. 1.p
245 with NE is significant (0.49). In the high group, there are three significant correlations: EP, -0.44; DA, -0.39; and SHIAA, -0.37. Table 2 shows the correlations between the urine values and ADD symptoms broken down by DBH. There are several positive correlations between ADD measures and the urine values in the combined group. As with the correlations with age, the number and strength of the correlations with catecholamine measures are greater in the low DBH group-eight in the low group and one in the high group. SHIAA is also correlated with one ADD measure in each group.

Table 2. Correlations (r) between ADD symptoms and amine measures in 24-hour urines h/24 hours/m2)

DA

HVA

NE

VMA

MHPG

Inattention (DSM-I// checklist)

Hyperactivity (DSM-//I checklist)

Combined

0.29’

0.17

0.392

DBH 5 6

0.32

0.26

0.28

DBH 1 15

0.34

0.15

0.45’

Combined

0.21

0.29’

0.392

DBH 5 6

0.22

0.593

0.522 0.25

DBH 1 15

0.21

0.05

Combined

0.23

0.16

0.473

DBH 5 6

0.39’

0.40’

0.733

DBH Z 15

0.18

0.11

0.34

Combined

0.11

0.26’

0.352

DBH 5 6

0.22

0.482

0.33

DBH 2 15

0.13

0.10

0.25

Combined

0.31’

0.10

0.13

DBH I

0.532

0.37’

0.24

0

6

DBH 2 15 EP

0

0

Combined

-0.06

0.07

0.12

DBH 5 6

0.02

0.37

0.38

-0.17

0.07

0.01

Combined

0.23

0.18

-0.09

DBH 5 6

0.29

0.582

0.33

DBH 2 15

0.563

0.29

0

DBH 2 15 SHIAA

Impulsive/hyperactive (Conners checklist)

Note. ADD = attentiondeficitdisorder.See note to Table 1 for other abbreviations.In the combinedgroup, n = 43 for VMA, MHPG, and HVA; for DA, NE, and EP, n = 41. In the low DBH group, n = 20 and 19; in the high DBH group, n = 23 and 22. 1.p
Table 3 compares three clinical groups of patients: CDU, CDS, and Neither on the 24-hour urine measures. The CDS group has significantly higher output of NE than the CDU or Neither group and higher VMA output than the Neither group. In Table 4, the groups are divided into CDU, CDS, and Neither to determine what effect the breakdown by these diagnoses may have on the correlations with the ADD measures. The correlations in the CDU group are more frequent and stronger than in the other groups. This may be due to the predominance of low DBH subjects in this

246

Table 3. Comparison of urinary amine measures @g/24 hours/m2) between conduct disorder subtypes CDU (n = 18) DA

211

HVA

5174

NE’

25

f

Neither (n = 12)

CDS (n =20)

57

262

+2103

5042 7

190

+ 1442

4045

P

2.6

0.09

+ 1239 7

4.9

0.01

772

2506

f

557

1822

+

508

4.2

0.02

MHPG

1075

f

369

1002

+

387

798

+

207

2.7

0.08

5.8 +

2.0

-t 1586

15

61

f

4057

+

+

2306

5HIAA

33

123

‘/MA2

EP

f

t-

F

6.1 + 4035

21

3.3

+ 1462

+

6.3 + 2918

+

2.9 936

Note. CDU = conduct disorder, undersocialized. CDS = conduct drsorder. socialized. conduct disorder. See note to Table 1 for other abbreviations.

Nerther = subfects without

1. CDU significantly different from CDU and Neither, p < 0.05, Duncan procedure. 2. CDS significantly different from Neither, p < 0.05, Duncan procedure.

group, but diagnostic effects are also a possibility. For example, 5H IAA is positively correlated with ADD symptoms in the CDU group but not in the CDS group. The correlations in the Neither group should be interpreted cautiously since the number

Table 4. Correlations (r) between ADD symptoms and urinary amine measures @g/24 hours/m2) Inattention checklist)

IDSM-III DA

HVA

NE

MHPG

EP

5HIAA

Impulsive/hyperactive (Canners checklist1

CDU

0.39

0.64s

0.22

CDS

0.36

0.32

0.43’

Neither

0.11

-0.28

CDU

0.11

0.39

CDS

0.20

0.35

0.23

Neither

0.11

0.02

-0.45

CDU

0.36

0.833

0.39

CDS

0.05

-0.14

-0.16

-0.25

CDU

0.03

0.46’

CDS

Neither VMA

Hyperactivity checklist)

(DSM-III

-0.47 0.562

0.42’ 0.26 0.552

-0.13

0.13

-0.02

Neither

0.01

-0.07

-0.66’ -0.19

CDU

0.34

0.16

CDS

0.11

-0.05

0.32

Neither

0.27

0.03

-0.60

0.803

0.14

CDU

0.08

CDS

-0.18

-0.30

0.14

Neither

-0.57’

-0.60’

0.12

CDU

0.47’

0.602

CDS

0.14

0.04

-0.24

Neither

0.51

0.59’

-0.26

Note. ADD = attention deficit disorder. See note to Table 1 for other abbreviations. 1. p < 0.05. 2. p < 0.01. 3. p < 0.005.

0.F18~

247 of subjects is small and the number of ADD symptoms in this group would be less than in the other two groups. The stronger correlations between the biochemical measures and age and of the biochemical measures and ADD symptoms in the low DBH group suggest either that the relationships are more visible because the group is more biochemically homogeneous in its catecholamine metabolism or that there are qualitative differences between the two groups. Correlations of the urinary measures with each other may provide clues as to whether there are qualitative differences in amine metabolism between the two DBH groups. Similarly, a correlation matrix may clarify whether there are qualitative differences in amine metabolism among the diagnostic groups. Tables 5 and 6 show the correlations of the monoamines and their metabolites. We hypothesized earlier that biochemical-behavioral relationships might be stronger or more visible in the low DBH group since it is more homogeneous in relationship to catecholamine metabolism. If there is decreased release of NE or a decreased efficiency in its production in the low DBH group, one could hypothesize differences in correlations between NE and its metabolites in the low group as compared to the high group. Since others have found high correlations in a sample that would be in strength of heterogeneous for plasma DBH (Roy et al., 1988), the reduction correlations would more likely be in the low DBH group. Similarly, if the “normal” metabolism of NE is being interfered with in the low DBH group, the relationships between NE function and DA and serotonin function may be different in the low DBH group compared to the high DBH group. In the low DBH group, MHPG and VMA are not correlated in contrast to a 0.62 correlation in the high DBH group. Correlations between other measures also show a difference in the low group with DA not correlating with HVA, VMA, MHPG, or SHIAA. DA has a 0.76 correlation with HVA in the high DBH group in contrast to a 0. I5 correlation in the low group, suggesting a shift away from the normal metabolic relationship between DA and HVA in the low DBH group. Interestingly, there was a positive correlation between 24-hour urinary HVA and plasma HVA in the high group (r = 0.47) but not in the low group (r = 0.19). There are several differences in the correlations when the CDU and CDS diagnostic groups are compared. Differences occur between MHPG and NE, VMA, Table 5. Correlation

coefficients

for urinary catecholamine

and metabolites

NE-E

VMA-MHPG

VMA-E

All subjects

0.472

0.452

0.442

0.27’

0.18

0.33’

DBH

5

6

0.613

0.39’

0.502

0.12

0.26

0.46’

DBH

2

15

0.442

0.522

0.42’

0.623

0.12

0.32

0.19

0.50’

0.40’

0.07

0.11

-0.08

0.66’

0.35

0.42

0.23

0.09

-0.02

0.19

0.642

0.53’

0.16

NE-YHPG

CDU CDS Neither

CDU

NE-WA

MHPG-E

0.67’ -0.04

Note. CDU = conduct disorder, undersocialized. CDS = conduct disorder, socialized. Sea note to Table 1 for other abbreviations. 1.p
0.21

CDU

CDS

0.05

0.38

0.32

0.40’

0.32

0.352

NEHVA

0.16

0.49’

0.23

0.633

0.06

0.392

VMADA

0.62’

0.773

0.843

0.783

0.793

0.773

VMAHVA

0.35

0.34

0.64*

0.542

0.29

0.382

IWHPGDA

0.61’

0.36

0.28

0.542

0.20

0.342

MHPGHVA

0.09

-0.16

0.592

0.20

0.09

0.15

NE5HIAA

0.32

0.31

0.39’

0.39’

0.42’

0.413

VMA5HlAA

0.51’

-0.29

0.23

0.16

0.11

0.11

MHPG5HIAA

l.p
Note. CDU = conduct disorder, undersocialized. CDS = conduct disorder, socialized. See note to Table 1 for other abbreviations.

nor CDS

0.29

0.542

DBHZ15

CDU

0.40’

DBHs6

Neither

0.402

0.43’

All subjects

NEDA

Table 6. Correlation coefficients for urinary monoamines and metabolites

0.15 0.793

0.56* 0.36

0.57’

0.763

0.432

0.57’

0.523 0.15

0.51 s 0.593

DAHVA

HVASHIAA

0.40

0.45

0.19

0.512

0.04

0.332

DASHIAA

249 and EP, with correlations greater in the CDS group. Differences also occur in the correlations between DA and other measures and between SHIAA and other measures.

Discussion Urinary catecholamines showed stronger correlations with age and with ADD symptoms in the low DBH group than in the high DBH group. Since the low DBH group may be more genetically homogeneous with regard to catecholamine function (Ciaranello and Boehme, 1981) catecholamines may be more directly associated with changes in behavior in this homogeneous group or the more homogeneous group may make the relationships more visible. In a more heterogeneous group, the same relationships may occur but be obscured by differing baseline values or other factors. There was only one significant correlation with catecholamine measures in the high DBH group, and given the number of correlations, this could have occurred by chance. Studies on ADD subjects reported by others would have included subjects who were heterogeneous for DBH, so our high DBH group is consistent with those studies in showing no relationship between the biochemical measures and ADD symptoms. The comparison of the correlation coefficients of the monoamines and their metabolites between the high and low DBH groups raises some possibilities regarding differences in metabolism in the low DBH group. We previously reported that the MHPG/VMA ratio was higher in the low DBH group (Rogeness et al., 1988). This finding, based on a review by Maas et al. (1987) is consistent with decreased extracellular release of NE. We hypothesized less efficient noradrenergic functioning in the low DBH group. The lack of correlation between MHPG and VMA in the low DBH group may reflect this less efficient system with a breakdown in the normal or usual relationship between MHPG and VMA. One could also speculate that the absence of a correlation between DA and HVA in the low group might result from DA’s being produced in unusual amounts in sites such as NE neurons, thereby resulting in a dissociation of the normal relationship between DA and HVA. The CDS group had higher levels of 24-hour NE and VMA production, raising the possibility of two types of conduct disorder-one with lower NE function and one with higher NE function. It is possible that biological-behavioral relationships may be different in different diagnostic groups-analogous to what we showed between the DBH groups since the biology may be different in each diagnostic group. The CDU group showed stronger correlations between the catecholamine measures and ADD symptoms. This may be due to the majority of the subjects in the CDU group having low DBH. However, the correlations between ADD symptoms and SHIAA in the CDU group suggest that biochemical differences may be present between the two groups just as the increased 24-hour urinary NE output suggests biochemical differences between the two groups. An examination of the correlation matrix also suggests differences. Differences are suggested within catecholamine metabolism as shown by MHPG correlations and between NE and serotonin metabolism as shown by the SHIAA correlations.

250

A comparison of the correlations of the ratios NE/SHIAA, MHPG/SHIAA, and VMA/SHIAA and the impulsive/hyperactive scale of the CBPC also suggests possible biochemical-behavioral differences between the two types of conduct disorder (Table 7). The symptoms are correlated positively with the ratios in the CDS group and negatively in the CDU group. The data suggest that there may be two biological subtypes of conduct disorder differentiated by noradrenergic function and different functional interaction between noradrenergic and serotonergic systems.

Table 7. Correlation of ratios with the hyperactive/ impulsive factor (Conners checklist) CDU

CDS

NE/SHIAA

-0.25

0.58’

MHPG/5HIAA

-0.57’

0.44’

VMAISHIAA

-0.08

0.37

Note. CDU = conduct disorder, undersocialized. CDS = conduct disorder, socialized. See note to Table 1 for other abbreviations. l.p
In summary, the data suggest that classifying groups by a biochemical variable may make biochemical-behavioral relationships visible that are obscured in more heterogeneous groups. Also, studying relationships between monoamines and their metabolites among diagnostic groups may provide clues to biochemical-behavioral relationships that are not evident when individual amines or metabolites are examined in isolation. Acknowledgment. This research was supported by grants M H-38679 and MH-40935 from the National Institute of Mental Health, the Meadows Foundation, the San Antonio Area Foundation, and the Abell-Hanger Foundation. The authors thank Ms. Lisa Boutin and Ms. Linda Crawford for their technical assistance.

References American Psychiatric Association. DSM-Ill: Diagnostic and Statistical Manual of’ Mental Disorders. 3rd ed. Washington, DC: APA, 1980. Barratt, T.M., and Chantler, C. Clinical assessment of renal function. In: Rubin, M.I., and Barratt, T.M., eds. Pediatric Nephrology. Baltimore: Williams & Wilkins Company, 1975. p. 56. Buchsbaum, M.S., and Rieder, R.O. Biologic heterogeneity and psychiatric research. Archives of’ General Psychiatry, 36: I l63- I 169, 1979. Ciaranello, R.D., and Boehme, R.E. Biochemical genetics of neurotransmitter enzymes and receptors: Relationships to schizophrenia and other major psychiatric disorders. Clinical Genetics. 19:358-372, 198 I. Canners, C.K. Rating scales for use in drug studies with children. Psychopharmacology Bulletin, (Special Issue): 24, 1973. Joseph, M.H.; Kadam, B.V.; and Risby, D. Simple high-performance liquid chromatographic method for the concurrent determination of the amine metabolites vanillylmandelic acid, 3-methoxy-4-hydroxyphenylglycol, 5-hydroxyindoleacetic acid, dihydroxyphenylacetic acid and homovanillic acid in urine using electrochemical detection. Journal of Chromatography. 226:36 I-368, I98 I. Kahn, A.U., and Dekirmenjian, H. Urinary excretion of catecholamine metabolites in hyperkinetic child syndrome. American Journal of’ Psychiatry, 138:108-l IO, 1981.

251 Koslow, S.H.; Maas, J.W.; Bowden, C.L.; Davis, J.W.; Hanin, 1.; and Javaid, J. CSF and urinary biogenic amines and metabolites in depression and mania. Archives of General Psychiatry, 40:999-IO IO, 1983. Maas, J.W.; Koslow, S.H.; Davis, J.M.; Katz, M.; Frazer, A.; Bowden, C.L.; Berman, N.; Gibbons, R.; Stokes, P.; and Landis, D.H. Catecholamine metabolism and disposition in healthy and depressed subjects. Archives of General Psychiatry, 44:337-344, 1987. Nagatsu, T., and Udenfriend, S. Photometric assay of dopamine-D-hydroxylase activity in human blood. Clinical Chemistry, 18:980-983, 1972. Quay, H.C. Attention deficit disorder and the behavioral inhibition system: The relevance of the neuropsychological theory of Jeffrey A. Gray. Journal of Child Psychology and Psychiatry, in press. Rapoport, J.L.; Lott, I.; Alexander, D.; and Abramson, J. Urinary noradrenaline and playroom behavior in hyperactive boys. Lancet, II: I 141, 1970. Rapoport, J.L.; Mikkelsen, E.J.; Ebert, M.H.; Brown, G.L.; Weise, V.K.; and Kopin, I.J. Urinary catecholamines and amphetamine excretion in hyperactive and normal boys. Journal of Nervous and Mental Disease, I66:73 I-737, 1978. Rogeness, G.A.; Hernandez, J.M.; Macedo, C.A.; Amrung, S.A.; and Hoppe, S.K. Nearzero plasma dopamine-P-hydroxylase and conduct disorder in emotionally disturbed boys. Journal of the American Academy of Child and Adolescent Psychiatry, 25521-527, 1986. Rogeness, G.A.; Hernandez, J.M.; Macedo, C.A.; Mitchell, E.L.; Amrung, S.A.; and Harris, W.R. Clinical characteristics of emotionally disturbed boys with very low activities of dopamine-P-hydroxylase. Journal of the American Academy of Child and Adolescent Psychiatry. 23:203-208, 1984. Rogeness, G.A.; Javors, M.A.; Macedo, C.A.; and Fischer, C. Plasma dopamine# hydroxylase, HVA, MHPG and conduct disorder in emotionally disturbed boys. Biological Psychiatry, 22: I 158-l 162, 1987. Rogeness, G.A.; Maas, J.W.; Javors, M.A.; Macedo, C.A.; Harris, W.R.; and Hoppe, S.K. Diagnoses, catecholamine metabolism and plasma dopamine-/3-hydroxylase. Journal of the American Academy of Child and Adolescent Psychiatry, 27: I2 I - 125, 1988. Roy, A.; Guthrie, S.; Karoum, F.; Pickar, D.; and Linnoila, M. High intercorrelations among urinary outputs of norepinephrine and its major metabolites. Archives of General Psychiatry. 45:158-161, 1988. Shekim, W.O.; Dekirmenjian, H.; and Chapel, J.L. Urinary MHPG excretion in minimal brain dysfunction and its modification by d-amphetamine. American Journal of Psychiatry, 136:667-671, 1979. Shekim, W.O.; Javaid, J.; Davis, J.M.; and Bylund, D.B. Urinary MHPG and HVA excretion in boys with attention deficit disorder and hyperactivity treated with damphetamine. Biological Psychiatry, I8:707-7 14, 1983. Smedes, F.; Kraak, J.C.; and Poppe, H. Simple and fast solvent extraction system for selective and quantitative isolation of adrenaline, noradrenaline and dopamine from plasma and urine. Journal of Chromatography. 23 I :25-39, 1982. Weinshilboum, R.M. Biochemical genetics of catecholamines in man. Mayo Clinic Proceedings, 58:3 19-330, 1983. Weinshilboum, R.M.; Raymond, F.A.; Elveback, L.R.; and Weidman, W.H. Serum dopamine-/3-hydroxylase activity: Sibling-sibling correlation. Science, 181:943-945, 1973. Wender, P.H.; Epstein, R.S.; Kopin, I.J.; and Gordon, E.K. Urinary monoamine metabolites in children with minimal brain dysfunction. American Journal of Psychiatry. 127:141 l-1415, 1971.