199
Psychiatry Research. 43: 199-2 I3 Elsevier
Noradrenergic Responses to Clonidine in Acute and Remitted Depressed Male Patients Robert L. Trestman, Timothy L. Lawrence, Emil F. Coccaro, Philip Harvey, David Bernstein, Edith K. Lawrence, Virginia Condello, Theresa Mahon, Ren-Kui Yang, Peter Knott, Thomas B. Horvath, and Larry J. Siever Received
May 22, 1991; revised version received June 15, 1992; accepted August 2, 1992.
Abstract. To investigate noradrenergic function in depression, plasma 3methoxy-4-hydroxyphenylglycol (MHPG), plasma norepinephrine (NE), mean arterial pressure (MAP), and heart rate responses to intravenous clonidine (2 pg/ kg), an a,-adrenergic agonist, were measured in 27 acutely depressed patients, I8 remitted depressed patients, and 27 normal control subjects; a placebo infusion was administered to a subgroup. Clonidine compared with placebo, over a 150minute time course, decreased plasma NE, MAP, and heart rate, but not plasma MHPG, in the control subjects. Plasma MHPG, plasma NE, MAP, and heart rate at baseline or in response to clonidine and placebo over 150 minutes did not indicate any group differences. The only significant plasma MHPG response to clonidine in the normal control subjects occurred 60 minutes after the infusion. A significantly diminished plasma MHPG response to clonidine at 60 minutes was found in the acutely depressed group compared with the normal control subjects. These results suggest that peripheral inhibitory noradrenergic responses to clonidine are normal in depressed patients, while plasma MHPG responses to clonidine, which have a limited central contribution, appear to be a weak reflection of central noradrenergic function and appear insufficiently robust for a meaningful evaluation of hypothetical group differences in central inhibitory a,-adrenergic activity in this population. Key Words. Affective pressure, heart rate. A reduced
responsiveness
disorder,
3-methoxy-4-hydroxyphenylglycol,
of a,-adrenergic
receptors
mean arterial
has been hypothesized
to be
Robert L. Trestman, Ph.D., M.D., is Research Director, Outpatient Psychiatry, Bronx Veterans Affairs Medical Center, and Assistant Professor of Psychiatry, Mt. Sinai School of Medicine, New York, NY. Timothy L. Lawrence, M.D., is Staff Psychiatrist, Stony Lodge, Ossining, NY. Emil F. Coccaro, M.D., is Associate Professor of Psychiatry, Medical College of Pennsylvania/ Eastern Pennsylvania Psychiatric Institute, Philadelphia, PA. Philip Harvey, Ph.D., is Associate Professor of Psychiatry, Mount Sinai School of Medicine, New York, NY. David Bernstein, Ph.D., is Instructor of Psychiatry, Mt. Sinai School of Medicine, New York, NY. Edith K. Lawrence, Ph.D., and Virginia Condello, M.D., are both in private practice in New York State. Theresa Mahon, B.A., is a medical student at the Albert Einstein School of Medicine. Bronx. NY. Ren-Kui Yang. M.D., is Research Scientist. Bronx Veterans Affairs Medical Center, Bronx, NY: Peter Knott, Ph.D,is Research Scientist, Bronx Veterans Affairs Medical Center, and Assistant Professor of Psychiatry, Mt. Sinai School of Medicine, New York, NY. Thomas B. Horvath, M.D., is Professor of Psychiatry, Downstate Medical Center, Brooklyn, NY. Larry J. Siever, M.D., is Professor of Psychiatry, Mt. Sinai School of Medicine, New York, NY. (Reprint requests to Dr. R.L. Trestman, Psychiatry Service, 116A, Bronx VA Medical Center, 130 W. Kingsbridge Road, Bronx, NY 10468, USA.) 01651781/92/$05.00
@ 1992 Elsevier Scientific
Publishers
Ireland
Ltd.
200 associated with major depressive disorder (Bunney and Murphy 1975; Cohen et al., 1980; Matussek et al., 1980; Siever et al., 1981, 1984; Siever and Uhde, 1984; Uhde et al., 1984). Consistent with this hypothesis, the concentration of plasma 3-methoxy-4hydroxyphenylglycol (MHPG) has been reported to show a modest decrease in response to intravenous clonidine, a centrally acting a,-adrenergic agonist, in normal subjects (Leckman et al., 1981; Siever et al., 1984; Nutt and Molyneux 1986) but not in acutely depressed patients (Siever et al., 1984). The magnitude of the reduction in plasma MHPG may partially reflect the responsiveness of central inhibitory a,-adrenergic receptors (Leckman et al., 1980, 1981; Charney et al., 1982~; Siever et al., 1984; Nutt and Molyneux, 1986; Hauger et al., 1988). Clonidine also results in decreases in levels of plasma norepinephrine (NE) in most (Siever et al., 1984; Veith et al., 1984; Featherstone et al., 1987), but not all studies (Hoehe et al., 1988) of normal control subjects, as well as in studies of acutely depressed patients (Matussek et al., 1980; Siever et al., 1984; Veith et al., 1984; Hauger et al., 1988). No differences between the plasma NE responses to clonidine of the acutely depressed patients and the normal control subjects were observed, in contrast with the differential response of plasma MHPG; this may reflect the purely peripheral origins of plasma NE as well as the greater variability in plasma NE responses to clonidine (Siever et al., 1984; Veith et al., 1984; Featherstone et al., 1987; Hoehe et al., 1988; Lechin et al., 1985). Furthermore, while diminished clonidine-induced heart rate reductions have been reported in acute depressed patients as compared with normal control subjects (Siever et al., 1984) normal blood pressure responses to clonidine have been consistently reported in acutely depressed patients (Matussek et al., 1980; Checkley et al., 1981; Charney et al., 1982a; Siever et al., 1984; Horton et al., 1986; Mitchell et al., 1988). The one recent exception (Mitchell et al., 1991) found a diminished systolic, but not diastolic, blood pressure reduction in 10 endogenous depressed patients compared with normal control subjects. The responses of noradrenergic and sympathetic indices to intravenous clonidine have not, however, been assessed in remitted depressed patients. A reduced plasma MHPG response to clonidine in the remitted state of depression might suggest a persistent deficit in the responsiveness of qadrenergic receptors and a vulnerability to recurrent depression. In contrast, normal plasma MHPG responses to clonidine in the remitted, but not the acute state of depression, might suggest a state-related disturbance, perhaps due to excessive noradrenergic activity in conjunction with altered regulation of the adrenergic receptors. The present study attempts to (1) replicate earlier findings of diminished MHPG and autonomic responses to intravenous clonidine in an independent sample of male patients with acute major depressive disorder and (2) examine whether a cohort of remitted depressed patients exhibited blunted noradrenergic responses to clonidine in order to evaluate the state dependence of noradrenergic responses to clonidine. Methods Subjects. Twenty-seven acutely depressed male patients, remission, and 27 male normal control subjects participated
18 depressed male patients in in this study. Other data on a
201
subgroup of these patients have been previously reported (Siever et al., 1992). All acutely depressed patients met Research Diagnostic Criteria (RDC; Spitzer et al., 1978) for definite or probable major depressive disorder at that time as determined by one or two experienced raters who used a structured interview, the Schedule for Affective Disorders and Schizophrenia-Lifetime Version (SADS-L; Endicott and Spitzer, 1978); the interrater reliability for a SADS-L diagnosis of major disorder is high (K = 0.83; Spitzer et al., 1987). Final diagnosis was determined after discrepancies were resolved in a consensus meeting chaired by an experienced diagnostician (T.B.H.) who was not directly involved in hypothesis generation and biological testing. The study group includes eight patients who did not participate in a formal SADS-L interview; their diagnoses were confirmed in a clinical interview that used the RDC. Clinical subgroups (unipolar, bipolar, psychotic, and endogenous patients) were determined by SADS-L criteria. Patients who met RDC for major depressive disorder only on the basis of self-report of symptoms before hospitalization and had a prominent personality disorder that preceded the first onset of depression were considered to have primary personality disorders and were excluded from study. The severity of depression at the time of study was assessed with the 23-item Hamilton Rating Scale for Depression (HRSD; Guy, 1976). Depressed patients in remission were determined to have had an episode of major depressive disorder with significant impairment or a successful response to antidepressant medication as determined by SADS-L interview. They had not been hospitalized for at least 6 months and were rated as having a score < 8 on the 23-item HRSD. Four patients were studied in both the acute and remitted states. Normal control subjects were recruited by advertisement; were determined to be free of RDC or D&V-III-R (American Psychiatric Association, 1987) Axis I disorders or prominent Axis II psychopathology on the basis of the Personality Diagnostic Questionnaire (PDQ), a self-report inventory (Hyler et al., 1982) and an extensive clinical interview with one of us (E.F.C.); and denied any family history of major psychiatric disorder. All subjects were medically screened by a psychiatrist with a medical history interview, physical examination, electrocardiogram, chest x-ray, complete blood count, blood chemistry profile, a flocculation test for syphilis, urinalysis, and thyroid function tests. Subjects with evidence of significant cardiovascular, endocrine, or other systemic illness on the basis of review by an internist were excluded from study. Procedure. Subjects were free of medication for at least 2 weeks before the protocol. They were placed on a low monoamine diet for 3 or more days beforehand and were kept awake and at bedrest throughout the study after an overnight fast. Smoking was not allowed on protocol days. An intravenous catheter was inserted at OS:00 hours and kept patent with a normal saline infusion. Baseline plasma samples for plasma NE and MHPG were drawn at 09:30 hours, 09:45 hours, and 1O:OOhours. Immediately after the blood sample was taken at IO:00 hours, either (1) clonidine at a dose of 2 pg/ kg was administered intravenously over 5 minutes in 10 ml of normal saline (clonidine challenge) or (2) 10 ml of normal saline were administered intravenously over 5 minutes (placebo infusion). Fourteen normal control subjects, 23 acutely depressed patients, and eight remitted depressed patients received the placebo infusion (mean days between studies = 30.1, SD = 57.3). Of these, eight normal control subjects, six acutely depressed patients, and three remitted depressed patients received the clonidine infusion first. No effects due to order of infusion were observed in any measured variable. Subjects were unaware of the experimental condition. Samples for the measurement of NE and MHPG were then drawn 30, 60, 90, 120, and 150 minutes after the infusion. Baseline blood pressure (calculated as mean arterial pressure [MAP]) and heart rate were measured at 09:45 hours and at 1O:OOhours immediately before the clonidine infusion, and at 30-minute intervals thereafter for 3 hours with an automated cuff (Dinamap Model 1846SX, Criticon, Tampa, FL). Plasma Analyses.
MHPG
samples
were collected
in heparinized
tubes,
kept on ice, and
202 centrifuged at 4 “C. The separated plasma was then frozen and stored at -70 “C within I hour of collection. The samples were later assayed by high pressure liquid chromatography with electrochemical detection (HPLC/ECD) techniques including an extraction procedure (Yang et al., 1988). The intra-assay and interassay coefficients of variation were 5.2% and 13.0%, respectively. The NE samples were collected and prepared in a similar manner. but the separated plasma was then frozen and stored at -70 “C within 30 minutes of rollection in tubes that contained 0. I ml of glutathione. The samples were later assayed by H PLC/ ECD, following an alumina extraction technique (Hallman et al., 1978). The interassay and intra-assay coefficients of variation were both 8%. Our earliest MHPG samples (from 8 acute depressed, 6 remitted depressed, and 9 normal control subjects) were assayed by gas chromatography/mass spectrometry (GCiMS). The experimental results were unchanged regardless of whether we included or excluded patients’ samples analyzed by the CC! MS. For consistency, therefore, we are presenting only those MHPG data derived by HPLC techniques: samples from I9 acute depressed, 12 remitted depressed, and 18 control sub,jects. Data Analyses. Between-group differences in clinical and demographic variables were tested by analysis of variance (ANOVA) or two-tailed Student’s f test. Most response variables were not normally distributed: all were log ,,,-transformed for parametric analysis; all data are presented untransformed for interpretability. Primary analysis for each variable (M H PG. NE, MAP, and heart rate) was a repeated measures ANOVA. with and without covariation of baseline age, MAP (for non-MAP analyses), and heart rate (for nonheart rate analyses). Factors were drug (clonidine and placebo infusion), group (normal control, acutely depressed, and remitted depressed), and time (baseline, 30.60, 90, 120, and 150 minutes). The model was a within-group design (i.e., all people of a group who participated in a clonidine challenge and all those who participated in a placebo infusion) instead of being restricted to only those who participated on both study days. This procedure allowed us to maximize available data. Where these analyses yielded significant (p < 0.05) results, restricted analyses of covariance (ANCOVAs) and t tests (with Bonferroni correc!ion) were then performed to Isolate the differences. Baseline MPPG and NE values on each protocol day were calculated as the mean of data collected at 30, IS. and 0 minutes before infusion. Before an earlier study (Siever et al.. 1984) the 60-minute time point had been found to be the time of maximum plasma MHPG response to intravenously administered clonidine (Siever, unpublished data). For that reason. and for comparability with previous studies (Siever et al., 1984, 1987) we also calculated the absolute change in plasma MHPG and NE, MAP and heart rate from baseline to 60 minutes after the clonidine infusion (A 60) for each group of subjects. Given the a priori hypothesis, and previous data suggesting diminished responses in the acutely depressed patient group versus the normal control group (Siever et al., 1984) one-tailed t tests were used. Correlational analyses were performed between the response measures and clinical and demographic variables by diagnostic group; Bonferroni correction for multiple comparisons was made. All values are presented in tables and text as mean + SD.
Results Demographic
and Clinical
The acute depressed patients (n = 27) the I8), and the normal control subjects (n = 27) were and clinical characteristics, except that the group significantly more depressed (as assessed by the hospitalized at the time of study, and included a depressed patients than did the group of remitted
Variables.
remitted depressed patients (n = comparable in both demographic of acute depressed patients were HRSD), were more likely to be larger proportion of endogenously depressed patients (Table 1).
203 Table 1. Demographic and clinical characteristics
Acute
Age (yr)
depressed (h= 27)
Remitted depressed (h= 18)
Normal controls (n= 27)
55.59 f 9.82
52.61 f 11.36
47.44 f 15.77 1.06 zt 0.13 -
Days not receiving
any medication’
57.35 f 35.35
1.15+0.21* 57.07 f 35.27
Days not receiving
antidepressants*
230.75 f 243.1
86.62 + 96.8
-
(15 - 365) 3.64 + 3.08
-
(0 - 8) 2.07 zt 1.73
-
1.12f0.15
Relative weight
(24
(range) HRSD scores
- 730)
26.72 -f (13
6.38
- 38.5)
(range) Number of hospitalizations3
3.18 + 2.30
% (n)’ outpatients
30.00 (6)’
94.44 (17)
-
% (n) bipolar depression
37.0 (10)
22.2 (4)
-
25.9 (7)
22.2 (4)
-
80.8 (21)
17.6 (3)
-
52.0 I1 31 **
22.2 (4)
-
% (#
psychotic
depression
% (rq5.8 endogenous % in) imoaired
depression
4,g
Note. Values are given as mean * SD ‘1” One/two missing case(s). 1. n for acute patients = 23, remitted = 15. 2. n for acute patients = 20, remitted = 16; excludes 7 acute and 2 remitted subjects who never used antidepressants. 3. n for acute patients = 22, remitted = 14. 4. For remitted patients refers to past depressive episode. 5. n for acute patients = 26, remitted = 17. 6. t = 16.17, df = 39.89, p < 0.001; acute > remitted. 7.~*=18.48,df=1,o
Psychiatry Research. 43: 199-2 I3 Elsevier
Noradrenergic Responses to Clonidine in Acute and Remitted Depressed Male Patients Robert L. Trestman, Timothy L. Lawrence, Emil F. Coccaro, Philip Harvey, David Bernstein, Edith K. Lawrence, Virginia Condello, Theresa Mahon, Ren-Kui Yang, Peter Knott, Thomas B. Horvath, and Larry J. Siever Received
May 22, 1991; revised version received June 15, 1992; accepted August 2, 1992.
Abstract. To investigate noradrenergic function in depression, plasma 3methoxy-4-hydroxyphenylglycol (MHPG), plasma norepinephrine (NE), mean arterial pressure (MAP), and heart rate responses to intravenous clonidine (2 pg/ kg), an a,-adrenergic agonist, were measured in 27 acutely depressed patients, I8 remitted depressed patients, and 27 normal control subjects; a placebo infusion was administered to a subgroup. Clonidine compared with placebo, over a 150minute time course, decreased plasma NE, MAP, and heart rate, but not plasma MHPG, in the control subjects. Plasma MHPG, plasma NE, MAP, and heart rate at baseline or in response to clonidine and placebo over 150 minutes did not indicate any group differences. The only significant plasma MHPG response to clonidine in the normal control subjects occurred 60 minutes after the infusion. A significantly diminished plasma MHPG response to clonidine at 60 minutes was found in the acutely depressed group compared with the normal control subjects. These results suggest that peripheral inhibitory noradrenergic responses to clonidine are normal in depressed patients, while plasma MHPG responses to clonidine, which have a limited central contribution, appear to be a weak reflection of central noradrenergic function and appear insufficiently robust for a meaningful evaluation of hypothetical group differences in central inhibitory a,-adrenergic activity in this population. Key Words. Affective pressure, heart rate. A reduced
responsiveness
disorder,
3-methoxy-4-hydroxyphenylglycol,
of a,-adrenergic
receptors
mean arterial
has been hypothesized
to be
Robert L. Trestman, Ph.D., M.D., is Research Director, Outpatient Psychiatry, Bronx Veterans Affairs Medical Center, and Assistant Professor of Psychiatry, Mt. Sinai School of Medicine, New York, NY. Timothy L. Lawrence, M.D., is Staff Psychiatrist, Stony Lodge, Ossining, NY. Emil F. Coccaro, M.D., is Associate Professor of Psychiatry, Medical College of Pennsylvania/ Eastern Pennsylvania Psychiatric Institute, Philadelphia, PA. Philip Harvey, Ph.D., is Associate Professor of Psychiatry, Mount Sinai School of Medicine, New York, NY. David Bernstein, Ph.D., is Instructor of Psychiatry, Mt. Sinai School of Medicine, New York, NY. Edith K. Lawrence, Ph.D., and Virginia Condello, M.D., are both in private practice in New York State. Theresa Mahon, B.A., is a medical student at the Albert Einstein School of Medicine. Bronx. NY. Ren-Kui Yang. M.D., is Research Scientist. Bronx Veterans Affairs Medical Center, Bronx, NY: Peter Knott, Ph.D,is Research Scientist, Bronx Veterans Affairs Medical Center, and Assistant Professor of Psychiatry, Mt. Sinai School of Medicine, New York, NY. Thomas B. Horvath, M.D., is Professor of Psychiatry, Downstate Medical Center, Brooklyn, NY. Larry J. Siever, M.D., is Professor of Psychiatry, Mt. Sinai School of Medicine, New York, NY. (Reprint requests to Dr. R.L. Trestman, Psychiatry Service, 116A, Bronx VA Medical Center, 130 W. Kingsbridge Road, Bronx, NY 10468, USA.) 01651781/92/$05.00
@ 1992 Elsevier Scientific
Publishers
Ireland
Ltd.
200 associated with major depressive disorder (Bunney and Murphy 1975; Cohen et al., 1980; Matussek et al., 1980; Siever et al., 1981, 1984; Siever and Uhde, 1984; Uhde et al., 1984). Consistent with this hypothesis, the concentration of plasma 3-methoxy-4hydroxyphenylglycol (MHPG) has been reported to show a modest decrease in response to intravenous clonidine, a centrally acting a,-adrenergic agonist, in normal subjects (Leckman et al., 1981; Siever et al., 1984; Nutt and Molyneux 1986) but not in acutely depressed patients (Siever et al., 1984). The magnitude of the reduction in plasma MHPG may partially reflect the responsiveness of central inhibitory a,-adrenergic receptors (Leckman et al., 1980, 1981; Charney et al., 1982~; Siever et al., 1984; Nutt and Molyneux, 1986; Hauger et al., 1988). Clonidine also results in decreases in levels of plasma norepinephrine (NE) in most (Siever et al., 1984; Veith et al., 1984; Featherstone et al., 1987), but not all studies (Hoehe et al., 1988) of normal control subjects, as well as in studies of acutely depressed patients (Matussek et al., 1980; Siever et al., 1984; Veith et al., 1984; Hauger et al., 1988). No differences between the plasma NE responses to clonidine of the acutely depressed patients and the normal control subjects were observed, in contrast with the differential response of plasma MHPG; this may reflect the purely peripheral origins of plasma NE as well as the greater variability in plasma NE responses to clonidine (Siever et al., 1984; Veith et al., 1984; Featherstone et al., 1987; Hoehe et al., 1988; Lechin et al., 1985). Furthermore, while diminished clonidine-induced heart rate reductions have been reported in acute depressed patients as compared with normal control subjects (Siever et al., 1984) normal blood pressure responses to clonidine have been consistently reported in acutely depressed patients (Matussek et al., 1980; Checkley et al., 1981; Charney et al., 1982a; Siever et al., 1984; Horton et al., 1986; Mitchell et al., 1988). The one recent exception (Mitchell et al., 1991) found a diminished systolic, but not diastolic, blood pressure reduction in 10 endogenous depressed patients compared with normal control subjects. The responses of noradrenergic and sympathetic indices to intravenous clonidine have not, however, been assessed in remitted depressed patients. A reduced plasma MHPG response to clonidine in the remitted state of depression might suggest a persistent deficit in the responsiveness of qadrenergic receptors and a vulnerability to recurrent depression. In contrast, normal plasma MHPG responses to clonidine in the remitted, but not the acute state of depression, might suggest a state-related disturbance, perhaps due to excessive noradrenergic activity in conjunction with altered regulation of the adrenergic receptors. The present study attempts to (1) replicate earlier findings of diminished MHPG and autonomic responses to intravenous clonidine in an independent sample of male patients with acute major depressive disorder and (2) examine whether a cohort of remitted depressed patients exhibited blunted noradrenergic responses to clonidine in order to evaluate the state dependence of noradrenergic responses to clonidine. Methods Subjects. Twenty-seven acutely depressed male patients, remission, and 27 male normal control subjects participated
18 depressed male patients in in this study. Other data on a
201
subgroup of these patients have been previously reported (Siever et al., 1992). All acutely depressed patients met Research Diagnostic Criteria (RDC; Spitzer et al., 1978) for definite or probable major depressive disorder at that time as determined by one or two experienced raters who used a structured interview, the Schedule for Affective Disorders and Schizophrenia-Lifetime Version (SADS-L; Endicott and Spitzer, 1978); the interrater reliability for a SADS-L diagnosis of major disorder is high (K = 0.83; Spitzer et al., 1987). Final diagnosis was determined after discrepancies were resolved in a consensus meeting chaired by an experienced diagnostician (T.B.H.) who was not directly involved in hypothesis generation and biological testing. The study group includes eight patients who did not participate in a formal SADS-L interview; their diagnoses were confirmed in a clinical interview that used the RDC. Clinical subgroups (unipolar, bipolar, psychotic, and endogenous patients) were determined by SADS-L criteria. Patients who met RDC for major depressive disorder only on the basis of self-report of symptoms before hospitalization and had a prominent personality disorder that preceded the first onset of depression were considered to have primary personality disorders and were excluded from study. The severity of depression at the time of study was assessed with the 23-item Hamilton Rating Scale for Depression (HRSD; Guy, 1976). Depressed patients in remission were determined to have had an episode of major depressive disorder with significant impairment or a successful response to antidepressant medication as determined by SADS-L interview. They had not been hospitalized for at least 6 months and were rated as having a score < 8 on the 23-item HRSD. Four patients were studied in both the acute and remitted states. Normal control subjects were recruited by advertisement; were determined to be free of RDC or D&V-III-R (American Psychiatric Association, 1987) Axis I disorders or prominent Axis II psychopathology on the basis of the Personality Diagnostic Questionnaire (PDQ), a self-report inventory (Hyler et al., 1982) and an extensive clinical interview with one of us (E.F.C.); and denied any family history of major psychiatric disorder. All subjects were medically screened by a psychiatrist with a medical history interview, physical examination, electrocardiogram, chest x-ray, complete blood count, blood chemistry profile, a flocculation test for syphilis, urinalysis, and thyroid function tests. Subjects with evidence of significant cardiovascular, endocrine, or other systemic illness on the basis of review by an internist were excluded from study. Procedure. Subjects were free of medication for at least 2 weeks before the protocol. They were placed on a low monoamine diet for 3 or more days beforehand and were kept awake and at bedrest throughout the study after an overnight fast. Smoking was not allowed on protocol days. An intravenous catheter was inserted at OS:00 hours and kept patent with a normal saline infusion. Baseline plasma samples for plasma NE and MHPG were drawn at 09:30 hours, 09:45 hours, and 1O:OOhours. Immediately after the blood sample was taken at IO:00 hours, either (1) clonidine at a dose of 2 pg/ kg was administered intravenously over 5 minutes in 10 ml of normal saline (clonidine challenge) or (2) 10 ml of normal saline were administered intravenously over 5 minutes (placebo infusion). Fourteen normal control subjects, 23 acutely depressed patients, and eight remitted depressed patients received the placebo infusion (mean days between studies = 30.1, SD = 57.3). Of these, eight normal control subjects, six acutely depressed patients, and three remitted depressed patients received the clonidine infusion first. No effects due to order of infusion were observed in any measured variable. Subjects were unaware of the experimental condition. Samples for the measurement of NE and MHPG were then drawn 30, 60, 90, 120, and 150 minutes after the infusion. Baseline blood pressure (calculated as mean arterial pressure [MAP]) and heart rate were measured at 09:45 hours and at 1O:OOhours immediately before the clonidine infusion, and at 30-minute intervals thereafter for 3 hours with an automated cuff (Dinamap Model 1846SX, Criticon, Tampa, FL). Plasma Analyses.
MHPG
samples
were collected
in heparinized
tubes,
kept on ice, and
202 centrifuged at 4 “C. The separated plasma was then frozen and stored at -70 “C within I hour of collection. The samples were later assayed by high pressure liquid chromatography with electrochemical detection (HPLC/ECD) techniques including an extraction procedure (Yang et al., 1988). The intra-assay and interassay coefficients of variation were 5.2% and 13.0%, respectively. The NE samples were collected and prepared in a similar manner. but the separated plasma was then frozen and stored at -70 “C within 30 minutes of rollection in tubes that contained 0. I ml of glutathione. The samples were later assayed by H PLC/ ECD, following an alumina extraction technique (Hallman et al., 1978). The interassay and intra-assay coefficients of variation were both 8%. Our earliest MHPG samples (from 8 acute depressed, 6 remitted depressed, and 9 normal control subjects) were assayed by gas chromatography/mass spectrometry (GCiMS). The experimental results were unchanged regardless of whether we included or excluded patients’ samples analyzed by the CC! MS. For consistency, therefore, we are presenting only those MHPG data derived by HPLC techniques: samples from I9 acute depressed, 12 remitted depressed, and 18 control sub,jects. Data Analyses. Between-group differences in clinical and demographic variables were tested by analysis of variance (ANOVA) or two-tailed Student’s f test. Most response variables were not normally distributed: all were log ,,,-transformed for parametric analysis; all data are presented untransformed for interpretability. Primary analysis for each variable (M H PG. NE, MAP, and heart rate) was a repeated measures ANOVA. with and without covariation of baseline age, MAP (for non-MAP analyses), and heart rate (for nonheart rate analyses). Factors were drug (clonidine and placebo infusion), group (normal control, acutely depressed, and remitted depressed), and time (baseline, 30.60, 90, 120, and 150 minutes). The model was a within-group design (i.e., all people of a group who participated in a clonidine challenge and all those who participated in a placebo infusion) instead of being restricted to only those who participated on both study days. This procedure allowed us to maximize available data. Where these analyses yielded significant (p < 0.05) results, restricted analyses of covariance (ANCOVAs) and t tests (with Bonferroni correc!ion) were then performed to Isolate the differences. Baseline MPPG and NE values on each protocol day were calculated as the mean of data collected at 30, IS. and 0 minutes before infusion. Before an earlier study (Siever et al.. 1984) the 60-minute time point had been found to be the time of maximum plasma MHPG response to intravenously administered clonidine (Siever, unpublished data). For that reason. and for comparability with previous studies (Siever et al., 1984, 1987) we also calculated the absolute change in plasma MHPG and NE, MAP and heart rate from baseline to 60 minutes after the clonidine infusion (A 60) for each group of subjects. Given the a priori hypothesis, and previous data suggesting diminished responses in the acutely depressed patient group versus the normal control group (Siever et al., 1984) one-tailed t tests were used. Correlational analyses were performed between the response measures and clinical and demographic variables by diagnostic group; Bonferroni correction for multiple comparisons was made. All values are presented in tables and text as mean + SD.
Results Demographic
and Clinical
The acute depressed patients (n = 27) the I8), and the normal control subjects (n = 27) were and clinical characteristics, except that the group significantly more depressed (as assessed by the hospitalized at the time of study, and included a depressed patients than did the group of remitted
Variables.
remitted depressed patients (n = comparable in both demographic of acute depressed patients were HRSD), were more likely to be larger proportion of endogenously depressed patients (Table 1).
203 Table 1. Demographic and clinical characteristics
Acute
Age (yr)
depressed (h= 27)
Remitted depressed (h= 18)
Normal controls (n= 27)
55.59 f 9.82
52.61 f 11.36
47.44 f 15.77 1.06 zt 0.13 -
Days not receiving
any medication’
57.35 f 35.35
1.15+0.21* 57.07 f 35.27
Days not receiving
antidepressants*
230.75 f 243.1
86.62 + 96.8
-
(15 - 365) 3.64 + 3.08
-
(0 - 8) 2.07 zt 1.73
-
1.12f0.15
Relative weight
(24
(range) HRSD scores
- 730)
26.72 -f (13
6.38
- 38.5)
(range) Number of hospitalizations3
3.18 + 2.30
% (n)’ outpatients
30.00 (6)’
94.44 (17)
-
% (n) bipolar depression
37.0 (10)
22.2 (4)
-
25.9 (7)
22.2 (4)
-
80.8 (21)
17.6 (3)
-
52.0 I1 31 **
22.2 (4)
-
% (#
psychotic
depression
% (rq5.8 endogenous % in) imoaired
depression
4,g
Note. Values are given as mean * SD ‘1” One/two missing case(s). 1. n for acute patients = 23, remitted = 15. 2. n for acute patients = 20, remitted = 16; excludes 7 acute and 2 remitted subjects who never used antidepressants. 3. n for acute patients = 22, remitted = 14. 4. For remitted patients refers to past depressive episode. 5. n for acute patients = 26, remitted = 17. 6. t = 16.17, df = 39.89, p < 0.001; acute > remitted. 7.~*=18.48,df=1,o
