Effects of smoking during antipsychotic withdrawal in patients with chronic schizophrenia

Schizophrenia Research 46 (2000) 119–127 www.elsevier.com/locate/schres

Effects of smoking during antipsychotic withdrawal in patients with chronic schizophrenia Jose A. Apud *, Michael F. Egan, Richard J. Wyatt Neuroscience Center at St. Elizabeth’s, Neuropsychiatry Branch, National Institute of Mental Health, 2700 Martin Luther King, Jr. Ave., SE, Washington, DC 20032, USA Received 8 September 1999; accepted 15 November 1999

Abstract A number of studies have shown that patients with schizophrenia smoke more than other psychiatric patients and more than the general population. Also, medicated schizophrenics who smoke present more positive symptoms of schizophrenia than non-smokers. The objective of the present study was to assess the effect of smoking on ratings of psychopathology for 30 days following discontinuation of antipsychotic medication. The subjects were 101 treatmentresistant patients with schizophrenia who had been admitted to the inpatient service of Neuroscience Research Hospital (NRH ), National Institute of Mental Health, between 1982 and 1994 to undergo studies involving discontinuation of antipsychotic medication. Patients were rated independently on a daily basis on the 22-item Psychiatric Symptom Assessment Scale (PSAS ), an extended version of the Brief Psychiatric Rating Scale (BPRS). At baseline, ratings for Verbal Positive, Paranoia and Loss of Function were higher in smokers (n=65) than nonsmokers (n=36), but a statistically significant difference was observed only for the Verbal Positive cluster. Analysis by gender revealed that male non-smokers had the lowest psychopathology ratings at baseline. There were no differences in Anxiety/depression, Behavior Positive, Deficit Symptoms or Mannerisms (a measure for abnormal involuntary movements). Following medication discontinuation, repeated-measure analysis demonstrated a ‘time’ effect for all the variables studied and a ‘group’ (smokers vs. non-smokers) effect for Verbal Positive, Paranoia, and Loss of Function. Post-hoc comparisons at individual time points showed significantly higher ratings for smokers at week 1 for Paranoia. No differences were observed at later time points. In conclusion, at baseline, smokers had more positive symptoms and were apparently more functionally impaired than non-smokers. This difference was no longer evident after a 30 day medication discontinuation period. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Antipsychotic; Drug withdrawal; Neuroleptic; Schizophrenia; Smoking

1. Introduction Over the last few years, a growing number of studies have focused on the relationship between * Corresponding author. Present address: Department of Psychiatry — Barton Hall, Psychopharmacology Division, St. Elizabeths Hospital, 2700 Martin Luther King, Jr. Ave., SE, Washington, DC 20032, USA. Tel.: +1-202-645-5733/5722; fax: +1-202-373-5413. E-mail address: [email protected] (J.A. Apud)

cigarette smoking and psychiatric disorders ( Hughes et al., 1986; Goff et al., 1992; Glassman, 1993; de Leon et al., 1995). For the most part, these studies indicate that patients with schizophrenia smoke more than other psychiatric patients and more than the general population (Masterson and O’Shea, 1984; Gopalaswamy and Morgan, 1986; Hughes et al., 1986; Goff et al., 1992; deLeon et al., 1995). It is not well understood whether increased cigarette smoking in schizo-

0920-9964/99/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S0 9 2 0 -9 9 6 4 ( 9 9 ) 0 0 23 0 - 3

120

J.A. Apud et al. / Schizophrenia Research 46 (2000) 119–127

phrenics is related to the effect of smoking on alleviating depressive and/or anxiety symptoms, combating boredom, decreasing side effects associated with antipsychotic medications, or correcting a neuronal deficit involved in the cognitive dysfunction present in some schizophrenics (Hughes et al., 1986; Goff et al., 1992; Glassman, 1993; Goldberg and Weinberger, 1995). Recent findings suggest that cigarette smoking, through its effect on a cholinergic receptor, may play a role in normalizing some of the psychophysiological deficits found in schizophrenia such as auditory evoked potential studies and smooth-pursuit eye movements (Holzman et al., 1988; Klein and Andresen, 1991; Adler et al., 1992, 1993; Freedman et al., 1997). Previous studies have found an association between the symptoms of schizophrenia and smoking. In fact, in a cohort of medicated patients with schizophrenia, smokers were found to present with more positive symptoms of schizophrenia than non-smokers. Heavy smokers had higher positive symptom ratings when compared with light smokers or non-smokers (Goff et al., 1992; Ziedonis et al., 1994), and fewer negative symptoms than light smokers. There were, however, no differences in negative symptoms between light and nonsmokers (Ziedonis et al., 1994). Whether these differences between smokers and non-smokers are related to an effect of smoking on neuroleptic levels, on the underlying neurobiology or on other still unknown factor is not yet clearly understood. Nicotine stimulates catecholaminergic activity both by increasing its synthesis (Hiremagalur and Sabban, 1995) and release (Imperato et al., 1986) and possibly by decreasing its degradation (Hiremagalur and Sabban, 1995). This suggests that nicotine by itself can induce psychotic symptoms. To test this hypothesis, we looked at changes in the severity of psychotic symptoms in schizophrenic patients undergoing discontinuation of antipsychotic medication. Because smokers might be expected to have heightened dopaminergic neurotransmission when antipsychotic medications are discontinued and there is no longer a postsynaptic block, they might have a more rapid increase in their psychotic symptoms. To address this issue, we performed an analysis of the scores on the

22-item Psychiatric Symptom Assessment Scale (PSAS ) (Bigelow and Berthot, 1989) in 101 patients with schizophrenia during a 30 day medication free period.

2. Patients and methods The patients were 29 female and 72 male treatment-resistant inpatient schizophrenics admitted to the research inpatient units at the NIMH Neuropsychiatric Research Hospital at St. Elizabeths between 1982 and 1994. The mean age (±SD) for females was 31.64±7.35 years and for males 31.45±7.93 years. All patients and usually a family member provided informed consent for participation in the study, which had been approved by the NIH Institutional Review Board. The program typically enrolled patients who had already had adequate trials of standard psychotherapeutic and pharmacotherapeutic interventions but were only partially effective in decreasing their symptoms. Upon admission, all patients underwent medical, neurological, and psychiatric evaluation usually lasting 5 weeks. If, from those evaluations, it was concluded that the patient had significant medical problems, a history of violence or suicidal behavior not identified during the prehospitalization screening, or if the patient decided to withdraw his consent, he/she could leave the hospital or participate in other protocols not involving placebo studies. All patients participating in the study were diagnosed with chronic schizophrenia using either DSM-III (American Psychiatry Association, 1980) or DSM-III-R criteria (American Psychiatry Association, 1987). Patients entering medication discontinuation studies did so in order to participate in protocols such as imaging or biochemical studies as well as for diagnostic purposes. The 101 patients had a mean length of illness of 11.53±7.36 years, and had been exposed to typical neuroleptics for most of their illness. Of the 101 patients, 65 (64%) were smokers. Twenty-six (40%) of the smokers and 10 (28%) of the non-smokers had tardive dyskinesia ( TD). All patients included in this study were on typical antipsychotic medications prior to their

J.A. Apud et al. / Schizophrenia Research 46 (2000) 119–127

admission. After the initial evaluation period, patients were maintained on a standard dose of a typical neuroleptic (usually haloperidol 0.4 mg/ kg/24 h) and benztropine if indicated. The patients remained on that medication in an open label fashion for at least several more weeks until going on coded medication. All other medications were discontinued. During the initial phase of the coded medication period, patients were placed either in coded inactive or coded active medication. For the coded inactive medication phase, patients were tapered from the coded active medication over a period of 4 days. Patients were kept on coded inactive medication for at least 4 weeks if their condition did not seriously deteriorate. The mean days of coded inactive medication were 69.67± 49.75 (median: 52.0, range 30–303). Patients who were on coded inactive medication for less than 30 days were excluded from the analysis. At the end of the coded medication period, the patients were restarted on an active medication (usually haloperidol 0.4 mg/kg) for 6 or more weeks before starting another protocol, or until they were well enough to be discharged. Following admission, each patient was rated twice daily on the 22-item Psychiatric Symptom Assessment Scale (PSAS) (Bigelow and Berthot, 1989), which is an extended version of the Brief Psychiatric Rating Scale (BPRS ) (Overall and Gorham, 1962). Mean ratings from the day shift, performed independently by two nurses, were used in the present study. The PSAS is a seven-point rating scale (0–6), with zero indicating the absence of a symptom or behavior. The PSAS differs from the BPRS in several aspects: (1) the information for the ratings comes from the patient’s own statements and/or behavior; (2) new items were added; (3) other items were renamed; and (4) the PSAS has a set of three key words accompanying each item to remind the rater of the multiple elements involved in each rating. The items comprise five different factors (Bigelow and Berthot, 1989): Anxiety/Depression consisting of ratings of anxiety statements, helplessness/hopelessness, guilt feelings, and somatic concerns; Behavior Positive consisting of tension, distractibility, motor hyperactivity, disorientation, hallucinatory statements, and hallucinatory behavior; Deficit Symptoms consisting of

121

depressed mood, social withdrawal, blunted affect, and motor retardation; Paranoia consisting of hostility, suspiciousness, and uncooperativeness; and Verbal Positive consisting of elated mood, grandiose statements, disorganized speech, and unusual ideas. A single item, the Loss of Function rating was also used. In addition, Mannerism, which is normally part of the Behavior Positive subscale, was examined separately. The use of Mannerism independently of the Behavior Positive factor provides a rough measure of involuntary movements (tardive dyskinesia, withdrawal dyskinesia, parkinsonism, increase of mannerisms, and posturing). Not surprisingly, in this cohort, patients with tardive dyskinesia ( TD) present a significantly higher score for Mannerisms than patients with no TD (manuscript in preparation). Thus, in our study, the Mannerisms cluster seems to represent a reliable index of choreoathetosis. Hence, in the final analysis seven items were examined. To evaluate the overall clinical effect of medication withdrawal in smokers and non-smokers, the mean PSAS scores for the different syndromes were pooled in groups of three days. The baseline measure was obtained by pooling ratings from days −5, −6 and −7 before the placebo phase. The Total Score Intraclass Correlation Coefficient (ICC ) (Bartko, 1966) for the Neuroscience Research Hospital nursing staff on the PSAS was previously shown to be 0.93, with most PSAS items ranging between 0.63 to 0.87. For PSAS factors, the ICC ranges between 0.80 and 0.90. The analysis was performed based on PSAS ratings obtained by a group of trained nurses who rotated in the rating of the patients and met once a month to control reliability. To assess the effect of smoking on positive symptoms (Behavior Positive, Verbal Positive and Paranoia), negative symptoms (Deficit Symptoms), mood and anxiety (AnxietyDepression), abnormal movements (Mannerisms) and function (Loss of Function), ANOVAs with time as repeated measure and smoking as a main effect were used. Post-hoc analysis comparing groups was performed using unpaired t-tests with Bonferroni correction. The demographic data were analyzed using unpaired t-tests for continuous data and by x2 for binary data. All analyses were

122

J.A. Apud et al. / Schizophrenia Research 46 (2000) 119–127

performed using Statistica, version Windows (Statsoft, Tulsa, OK ).

5.1

for

3. Results Of 123 patients that participated in the coded medication protocol, 101 are included in this analysis. Seventeen were excluded because their condition significantly deteriorated before completing 30 days on placebo, while five patients were excluded because they had been antipsychotic-free for a significant amount of time prior to starting the protocol. Seventy three per cent of the 22 patients excluded from the protocol were males; 50% were non-smokers and 77% did not have TD. Of the 17 excluded patients who were on neuroleptics prior to the protocol, 94% had been on high-potency antipsychotics. There were no significant differences in psychopathology ratings, duration of illness, age at admission, age at onset (n=22) or dose of neuroleptic (n=17) between smokers and non-smokers. Moreover, there were no significant differences in baseline psychopathology ratings or in demographic characteristics between the excluded patients and those included in the analysis (data not presented) Demographics for the 101 schizophrenic patients participating in this study are presented in Table 1a. The overall frequency of smoking was 64% (n=65). Among smokers, 46/65 (71%) were males, and 19/65 (29%) were females (x2=0.02, P=0.87). Comparisons of age at onset, age at admission, duration of illness, or dose of the antipsychotic medication at baseline expressed in CPZ equivalents (Jeste and Wyatt, 1982) revealed no significant differences between smokers and

nonsmokers ( Table 1). No differences were observed in WAIS-R Verbal Intelligence Quotient, Performance Intelligence Quotient or Full Scale Intelligence Quotient (data not shown). Repeat-measure ANOVAs revealed a significant effect of ‘time’ for Anxiety/depression (F=2.96, P<0.01), Behavior Positive (F=8.75, P<0.01), Verbal Positive (F=5.74, P<0.01), Paranoia (F= 5.99, P<0.01), Loss of Function (F=2.75, P<0.01), Mannerisms (F=6.24, P<0.01) and Deficit Symptoms (F=2.68, P<0.05). There was also a ‘group’ effect for Paranoia (F=10.70, P<0.02), Verbal Positive (F=9.74, P<0.02) and Loss of Function (F=8.26, P<0.04). No significant group×time interactions was found for any of the items analyzed. Unpaired t-tests with Bonferroni correction revealed that, at baseline, both positive symptoms of schizophrenia and functional impairment were higher in smokers than in non-smokers, even though values were statistically significant only for the Verbal Positive cluster (Paranoia, P=0.09; Verbal Positive P<0.05; Loss of Function P=0.07) (Table 2 analysis of variance with Bonferroni correction. Values represent the Mean ± S.D. of Total Rating values for each item. The ‘‘Behaviour Positive’’ scale excludes the ‘‘Mannerisms’’ item.). One week after antipsychotic medications were withdrawn, a significant difference between groups was observed only for the Paranoia factor (P<0.004). Two, 3 and 4 weeks after discontinuation of antipsychotic medications, no significant differences were evident between smokers and non-smokers (Table 2). No significant changes were found for Deficit Symptoms, Anxiety/depression, or Mannerisms either at baseline or at any of the times during the coded inactive medication period (Table 2).

Table 1 Demographic data of schizophrenic patients who completed a 4-week drug-free perioda Smoking

Gender

Age at onset (years)

Age at admission (years)

Duration of illness (years)

Neuroleptic dose (CPZ equivalents)

Yes No

M=46; F=19 M=26; F=10 x2=0.02 0.87

19.68±4.21 20.69±5.03 t=1.08 0.28

31.12±7.1 32.34±8.9 t=0.75 0.45

11.44±6.9 11.65±8.3 t=0.13 0.89

1595.3±1497.6 1236.9±1352.5 t=−1.19 0.24

P

a Values represent the Mean ± S.D. (two-tailed ‘‘t’’-test for independent data).

123

J.A. Apud et al. / Schizophrenia Research 46 (2000) 119–127 Table 2 Neuroleptic withdrawal in schizophrenic patients: effects of smoking on psychopathology ratingsa Syndrome

Anxiety/depression

Behavior positive

Deficit symptoms

Paranoia

Verbal positive

Loss function

Mannerisms

Period

Baseline 1 week 2 weeks 3 weeks 4 weeks Baseline 1 week 2 weeks 3 weeks 4 weeks Baseline 1 week 2 weeks 3 weeks 4 weeks Baseline 1 week 2 weeks 3 weeks 4 weeks Baseline 1 week 2 weeks 3 weeks 4 weeks Baseline 1 week 2 weeks 3 weeks 4 weeks Baseline 1 week 2 weeks 3 weeks 4 weeks

Smoker

Non-smoker

P-value (corrected)

Mean

S.D.

Mean

S.D.

2.60 2.64 3.10 2.98 3.52 6.79 7.37 8.31 8.54 10.13 5.36 5.85 6.33 6.78 6.95 3.22 3.87 3.91 3.87 5.03 3.91 4.13 4.51 4.28 5.60 3.10 3.11 3.23 3.18 3.40 2.25 2.45 2.51 2.62 2.85

2.62 2.89 2.72 3.53 3.80 5.22 5.76 5.65 6.02 6.88 3.05 3.31 3.07 3.22 3.30 1.95 2.93 2.53 2.62 3.15 3.55 3.94 3.58 4.03 4.75 0.83 1.14 0.92 1.00 1.14 1.15 1.33 1.34 1.38 1.48

2.75 2.87 3.94 3.55 3.41 5.32 5.69 6.98 7.09 8.01 6.64 5.82 6.65 6.44 7.16 2.28 2.00 3.03 3.11 3.54 1.89 2.62 2.39 2.93 3.93 2.67 2.77 2.90 2.81 2.93 1.84 1.99 2.38 2.27 2.58

2.56 2.63 3.79 3.74 3.08 4.19 4.67 6.13 7.01 6.72 2.97 2.66 2.68 3.10 3.91 1.77 1.56 3.18 2.99 2.63 2.47 2.91 3.56 3.60 4.34 0.79 0.83 1.09 1.04 1.06 1.14 1.42 1.41 1.30 1.67

n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s n.s. n.s. n.s. n.s. n.s n.s. =0.09 <0.004 n.s. n.s. n.s. <0.05 n.s. n.s. n.s. n.s =0.07 n.s. n.s. n.s n.s n.s n.s. n.s. n.s n.s.

a Data were analyzed by repeated measure analysis of variance with Bonferroni correction. Values represent the Mean±S.D. of Total Rating values for each item. The ‘‘Behavior Positive’’ scale excludes the ‘‘Mannerisms’’ item.

4. Discussion The aim of this retrospective study was to assess the effect of smoking on psychopathology ratings in patients with schizophrenia during withdrawal from antipsychotic medications. Sixty-four per cent of the 101 treatment refractory schizophrenics were smokers. This is consistent with previous studies showing that schizophrenics are more likely

to smoke than other psychiatric patients and than the general population (Masterson and O’Shea, 1984; Gopalaswamy and Morgan, 1986; Hughes et al., 1986; Goff et al., 1992; DeLeon et al., 1995). Furthermore, in the Goff et al. (1992) study, patients with schizophrenia who smoked were found to receive twice the daily dose of antipsychotic medication with respect to non-smokers. In the present study, however, smokers and non-

124

J.A. Apud et al. / Schizophrenia Research 46 (2000) 119–127

smokers were receiving similar doses of neuroleptic medication (in CPZ equivalents) on admission to our research center ( Table 1). One explanation for the different doses of neuroleptic medication in the present study and that of Goff et al. between smokers and non-smokers is that the patients in the present analysis were a relatively homogenous population of chronic schizophrenic inpatients. Also, because the patients in our study were treatment-refractory, prior to admission, their treating psychiatrists used doses of antipsychotic medication at least two- to threefold higher than those reported for patients included in other withdrawal studies (Gilbert et al., 1995). There were no sociodemographic differences between smokers and non-smokers for gender, age, age at onset, duration of illness ( Table 1) or IQ (data not shown). As was found in previous investigations (Goff et al., 1992; Ziedonis et al., 1994) at baseline, smokers presented more positive symptoms of schizophrenia than non-smokers ( Verbal Positive, P<0.05; Paranoia, P=0.09) ( Table 2). A similar trend was observed for Behavior Positive, even though the difference did not reach statistical significance after the Bonferroni correction (P= 0.14). Furthermore, patients in the smoking group were more functionally impaired (Loss of Function, P=0.07). Interestingly, with the exception of the Paranoia factor, the differences observed at baseline for Verbal Positive, and Loss of Function were no longer evident 1 week following withdrawal from antipsychotic medications. At weeks 2–4 of the coded inactive medication period, no significant difference was observed between smokers and non-smokers on any of the rated items. These findings suggest that smoking worsens psychosis. However, post-hoc analysis indicates that this is true mostly at baseline when patients are still on neuroleptics or, in some cases, during the first week on coded inactive medication. This difference disappears after 2– 4 weeks on placebo. When the data were analyzed by gender, a significant effect of smoking on positive symptoms was found. In fact, male smokers were more impaired at baseline and up to 2 weeks following antipsychotic withdrawal than nonsmokers, while no differences were observed in the female group (manuscript in preparation).

The statistically significant effect of ‘time’ indicates that both smokers and non-smokers unsurprisingly undergo a clinical deterioration during a 30 day coded inactive medication period. The loss of a statistically significant difference in the psychopathology rating between non-smokers and smokers for positive symptoms and function after 1– 2 weeks on coded inactive medication may suggest that there might be different patterns of deterioration between the patient groups. This contention, however, is not supported by the repeated-measure analysis since no significant ‘group×time’ interaction was noted. To test the effect of smoking exclusively in patients on coded inactive medication and without the residual effect of neuroleptics, a ‘group×time’ interaction analysis was performed during weeks 2–4. Even with these less stringent conditions, no additional significant findings were observed (results not shown). The finding that withdrawal from antipsychotic medications in patients with schizophrenia failed to lead to a more rapid clinical deterioration in smokers than in non-smokers was unexpected in view of the pharmacological effects of nicotine on the brain. Nicotine stimulates catecholaminergic activity both by increasing its synthesis ( Hiremagalur and Sabban, 1995) and release ( Imperato et al., 1986) and possibly by decreasing dopamine degradation (Hiremagalur and Sabban, 1995). Thus, we hypothesized a more rapid clinical deterioration in smokers than in non-smokers in the absence of antipsychotic medication. Possible explanations for this unexpected finding include; (1) a pharmacokinetic interaction between smoking and antipsychotic medications; (2) other pharmacological effects of smoking, (3) a positive effect of smoking on cognitive function, (4) compensatory actions of smoking on particular neurophysiological deficits present in schizophrenics. Regarding the possibility of a pharmacokinetic interaction between smoking and neuroleptics, smoking has been shown to decrease neuroleptic blood levels by increasing hepatic metabolism and renal excretion (Salokangas et al., 1997). This may explain the higher psychopathology rating for positive symptoms observed in smokers than non-smokers being treated with antipsychotic medications. Unfortunately, we do not have any neuroleptic blood levels for our patients. However, if

J.A. Apud et al. / Schizophrenia Research 46 (2000) 119–127

blood levels of the antipsychotic medication were lower in our smoking population, then the initial impact of the withdrawal from antipsychotic medications on psychotic decompensation may not be as dramatic in smokers. This could, in turn, lead to a loss of difference in psychotic symptoms between the two groups. Concerning other pharmacological effects of smoking, there is certainly no reason to assume that the usefulness of smoking is limited only to the single acute pharmacological action of nicotine. In fact, many central nicotinic receptors, including those that mediate the effects of nicotine on dopamine release, are chronically desensitized by longterm exposure to nicotinic agonists ( Wonnacott, 1990; Balfour, 1991). Thus, caution is warranted when assuming that chronic nicotine inhalation from tobacco smoke invariably results in the release of dopamine with activation of the dopamine-mediated reward system. In fact, under chronic stimulation conditions, the opposite may occur, that is, the desensitization of the nicotinic cholinergic receptors by nicotine or by any of the active minor tobacco alkaloids may reduce the fraction of released dopamine at the level of the synaptic cleft (Dwoskin et al., 1995). This adaptive mechanism may prove useful in view of the longterm effects of antipsychotic medications on the dopaminergic system (Racagni et al., 1980). Both acute and chronic administration of antipsychotic medication increase the release of dopamine in different brain areas (Egan et al., 1996). This increase may continue to be observed days or weeks after the medication is withdrawn. Thus, discontinuation of antipsychotic medication would make an unblocked receptor available to an increased amount of dopamine. The decreased release of dopamine secondary to the desensitization of the nicotinic cholinergic receptors may help explain the observation that smoking had no effect on the rate of deterioration in chronic schizophrenic patients after withdrawal from antipsychotic medications. A third possibility for the present findings is the alleged positive effect of smoking on cognitive function in chronic schizophrenics. Nicotine has been found to reverse scopolamine-induced impairment of performance in the passive avoidance task in rats (Nitta et al., 1994), to improve concen-

125

tration and decrease distractibility in normal subjects ( Wesnes and Warburton, 1983). Nicotine also reverses the haloperidol-induced impairment in memory performance and complex reaction time in schizophrenic patients (Levin et al., 1996). Schizophrenics have been found to be particularly impaired in some cognitive tests (Goldberg and Weinberger, 1995), raising the possibility that they may be using smoking as a tool to compensate for some of those cognitive deficits. Finally, another explanation for our findings may be that smoking has a compensatory effect on the P50 and smooth pursuit eye movement deficits present in schizophrenics. Auditory evoked potential studies indicate that the P50 wave deficit is inherited in schizophrenic pedigrees (Freedman et al., 1997). Similarly, schizophrenics display abnormalities in smooth-pursuit eye movements ( Holzman et al., 1988). Cigarette smoking has been found to normalize the P50 auditory-evoked potential gating deficit (Adler et al., 1992, 1993), and to partially reverse abnormalities in smoothpursuit eye movements ( Klein and Andresen, 1991). Whether the normalizing effect is due to nicotine’s ability to alter cholinergic transmission in schizophrenic patients or whether this effect is mediated through other neurotransmitter systems modulated by acetylcholine is not yet understood (Clarke and Pert, 1985; Egan and North, 1986). This effect, however, does not seem to be related to other confounding factors such as concurrent treatment with antipsychotic medications or anticholinergics ( Waldo et al., 1995). There are some limitations to this study. For instance, a number of patients had to be excluded from the study due to a severe deterioration of their mental status and could not remain off antipsychotic medications for 30 days. Because these patients had more rapid or severe deterioration, if they had remained in the study, the outcome could have been different. However, the number of smokers and non-smokers that were excluded from the study was similar, so it is unlikely that the exclusion of those patients posed a major impact on the final analysis. A second limitation is that the daily number of cigarettes smoked by the patients was not recorded, so it is unclear whether there was an increase, a decrease or no changes in cigarette consumption

126

J.A. Apud et al. / Schizophrenia Research 46 (2000) 119–127

for the smokers during the placebo period. Also, given the naturalistic design of this study, no biochemical or functional correlates such as nicotine, aconitine, carbon monoxide or bombesin levels were followed. Any of these indices could have provided an indication of how much patients were smoking. A third limitation to this study is that we do not have any information on the caffeine consumption in these patients. Heavy smoking is directly correlated with caffeine consumption, and thus increased psychopathology in smokers may also be related to caffeine consumption. Previous studies, however, failed to demonstrate any consistent psychotogenic effect of caffeine when administered over long periods ( Koczapsi et al., 1989; Hyde, 1990). A fourth important point relates to the relationship between medication dose and neuroleptic levels. In this study, we do not have the information necessary to establish a correlation between dose and level. A fifth limitation of the study is that we have used a global tool (Mannerism) to assess involuntary movements, whereas previous studies have concluded that dyskinetic movements of the fingers show the most significant increase during discontinuation of antipsychotic medications (Dixon et al., 1993). The lack of sensitivity of the Mannerism ratings in terms of abnormal involuntary movements may have contributed to our inability to observe any significant differences between smokers and non-smokers after withdrawal of antipsychotic medications. A sixth limitation is the fact that the ratings were not done by the same nurses throughout the study, which may have increased the variability in the ratings. In fact, this research spanned a period of about 12 years. Nevertheless, a group of highly trained research nurses who met regularly to test for the reliability of their ratings performed the ratings. In conclusion, the high incidence of smoking in schizophrenics was confirmed using a selected population of treatment resistant chronic schizophrenics admitted to the NIMH Neuropsychiatry Research Hospital at St. Elizabeths, between 1982 and 1994. At baseline, smokers presented with more positive symptoms and were more function-

ally impaired than non-smokers. After the introduction of coded inactive medication, the difference between groups remained significant for 1 week only for the Paranoia factor. When stratified by gender, however, the difference between smokers and non-smokers for the Paranoia as well as the Verbal Positive factors remained significant up to 2 weeks following neuroleptic withdrawal in the male group. No differences were observed in the female group (manuscript in preparation). The failure of smoking to accelerate the psychotic decompensation in patients on placebo could be explained through a pharmacokinetic interaction between antipsychotic and smoking, a unique pharmacological mechanism as a result of chronic exposure to nicotine, or to a compensatory effect of nicotine on cognitive functions or on neurophysiological deficits in schizophrenics. Further information on the effects of smoking on psychopathology may be obtained by analyzing the recovery rate in smokers and non-smokers after reinstitution of antipsychotic medications. If the pharmacokinetic interaction hypothesis is true, we should observe a slower trend for recovery in smokers with respect to non-smokers because of the difficulty in achieving appropriate neuroleptic levels due to a smoking-induced increase in antipsychotic metabolism.

Acknowledgements The authors would like to thank the patients, their families and the staff of the Neuropsychiatry Research Hospital, NIMH, for their participation and collaboration. We also thank Dr Karen Pettigrew for helping us with the statistical evaluation and Mrs Ioline Henter for her help in retrieving information and editorial assistance.

References Adler, L.E., Hoffer, L.D., Griffith, J., Waldo, M.C., Freedman, R., 1992. Normalization by nicotine of deficient auditory sensory gating in the relatives of schizophrenics. Biol. Psychiatry 32, 607–616. Adler, L.E., Hoffer, L.D., Wiser, A., Freedman, R., 1993. Normalization of auditory physiology by cigarette smoking in schizophrenic patients. Am. J. Psychiatry 150, 1856–1861. American Psychiatric Association, 1987. DSM-III: Diagnostic

J.A. Apud et al. / Schizophrenia Research 46 (2000) 119–127 and Statistical Manual of Mental Disorders. American Psychiatric Association, Washington, DC. American Psychiatric Association, 1987. DSM-III-R: Diagnostic and Statistical Manual of Mental Disorders. third edition, revised, American Psychiatric Association, Washington, DC. Balfour, D.J.K., 1991. The neurochemical mechanisms underlying nicotine tolerance and dependence. In: Pratt, J.A. ( Ed.), The Biological Basis of Drug Tolerance and Dependence. Academic Press, London, pp. 121–151. Bartko, J.J., 1966. The intra-class correlation coefficient as a measure of reliability. Psychol. Rep 19, 3–11. Bigelow, L.B., Berthot, B.D., 1989. The psychiatric symptom assessment scale (PSAS). Psychopharmacol. Bull. 25, 168–173. Clarke, P., Pert, A., 1985. Autoradiographic evidence for nicotine receptors on nigrostriatal and mesolimbic dopaminergic neurons. Brain Res. 348, 355–358. de Leon, J., Dadvand, M., Canuso, C., White, A.O., Stanilla, J.K., Simpson, G.M., 1995. Schizophrenia and smoking: an epidemiological survey in a state hospital. Am. J. Psychiatry 152, 453–455. Dixon, L., Thakker, G., Conley, R., Ross, D., Cascella, N., Tamminga, C., 1993. Changes in psychopathology and dyskinesia after neuroleptic withdrawal in a double-blind design. Schizophr. Res. 10, 267–271. Dwoskin, L.P., Teng, L., Buxton, S.T., Ravard, A., Dei, N., Crooks, P.A., 1995. Minor alkaloids of tobacco release [3H ]-dopamine from superfused rat striatal slices. Eur. J. Pharmacol. 276, 195–199. Egan, T.M., North, R.A., 1986. Actions of Acetylcholine and nicotine on rat locus coeruleus in vitro. Neuroscience 19, 565–571. Egan, M.F., Chrapusta, S., Karoum, F., Lipska, B., Wyatt, R.J., 1996. Effects of chronic neuroleptic treatment on dopamine release: insights from studies using 3-methoxytyramine. J. Neural Transm. 103, 777–805. Freedman, R., Coon, H., Myles-Worsley, M., Orr-Utreger, A., Olincy, A., Davis, A., Polymeropoulos, M., Holik, K., Adler, L.E., Leonard, S., Byerley, W., 1997. Linkage of a neurophysiological deficit in schizophrenia to a chromosome 15 locus. Proc. Natl. Acad. Sci. USA 94, 587–592. Gilbert, P.L., Harris, M.J., McAdams, L.A., Jeste, D.V., 1995. Neuroleptic withdrawal in schizophrenic patients. Arch. Gen. Psychiatry 52, 173–188. Glassman, A.H., 1993. Cigarette smoking implications for psychiatric illness. Am. J. Psychiatry 150, 546–553. Goff, D.C., Henderson, D.C., Amico, E., 1992. Cigarette and smoking. Am. J. Psychiatry 149, 1189–1194. Goldberg, T.E., Weinberger, D.R., 1995. Thought disorder, working memory and attention: interrelationship and the effects of neuroleptic medications. Int. Clin. Psychopharmacol. 10, Suppl. 3, 99–104. Gopalaswamy, A.K., Morgan, R., 1986. Smoking in chronic schizophrenia. Br. J. Psychiatry 149, 523–528. Hiremagalur, B., Sabban, E.L., 1995. Nicotine elicits changes in expression of adrenal catecholamine biosynthetic enzymes, neuropeptide Y and immediate early genes by

127

injection but not continuous administration. Brain Res. Mol. Brain Res. 32, 109–115. Holzman, P.S., Kringlen, E., Matthysee, S., Flanagan, S.D., Lipton, D., Cramer, R.B., Levin, G., Lange, S., Levey, D.L., 1988. A single dominant gene can account for eye tracking dysfunctions and schizophrenia in offspring of discordant twins. Arch. Gen. Psychiatry 45, 641–647. Hughes, J.R., Hatsukami, D.K., Mitchell, J.E., Dahlgren, L.A., 1986. Prevalence of smoking among psychiatric outpatients. Am. J. Psychiatry 143, 993–997. Hyde, A.P., 1990. Response to ‘Effects of caffeine on behavior of schizophrenic inpatients’. Schizophr. Bull. 16, 371–375. Imperato, A., Mulas, A., DiChiara, G., 1986. Nicotine preferentially stimulates dopamine release in the limbic system of freely moving rats. Eur. J. Pharmacol. 132, 337–338. Jeste, D.V., Wyatt, R.J., 1982. Understanding and treating tardive dyskinesia. In: Kupfer, D.J., Wyatt, R.J. ( Eds.), Foundations of Modern Psychiatry. The Guilford Press, New York, pp. 81–87. Klein, C., Andresen, B., 1991. On the influence of smoking upon smooth pursuit eye movements of schizophrenics and normal controls. J. Psychophysiol. 5, 361–369. Koczapsi, A., Paredes, J., Kogan, C., Ledwidge, B., Higenbottam, J., 1989. Effects of caffeine on behavior of schizophrenic inpatients. Schizophr. Bull. 15, 339–344. Levin, E.D., Wilson, W., Rose, J.E., McEvoy, J., 1996. Nicotine–haloperidol interactions and cognitive performance in schizophrenics. Neuropsychopharmacology 15, 429–436. Masterson, E., O’Shea, B., 1984. Smoking and malignancy in schizophrenia. Br. J. Psychiatry 145, 429–432. Nitta, A., Katono, Y., Itoh, A., Hasegawa, T., Nabeshima, T., 1994. Nicotine reverses scopolamine-induced impairment of performance in passive avoidance task in rats through its action on the dopaminergic neuronal system. Pharmacol. Biochem. Behav. 49, 807–812. Overall, J.E., Gorham, D.R., 1962. The Brief Psychiatry Rating Scale. Psychol. Rep. 10, 799–812. Racagni, G., Bruno, F., Bugatti, A., Parenti, M., Apud, J.A., Santini, V., Carenzi, A., Gropetti, F., Cattabeni, F., 1980. Behavioral and biochemical correlates after haloperidol and clozapine long-term treatment. In: Racagni, G., Spano, P.F. ( Eds.), Long-term Effects of Neuroleptics. Raven Press, New York. Salokangas, R.K., Saarijarvi, S., Taiminen, T., Lehto, H., Niemi, H., Ahola, V., Syvalathi, E., 1997. Effect of smoking on neuroleptics in schizophrenia. Schizophr. Res. 17, 55–60. Waldo, M., Myles-Worsley, M., Madison, A., Byerley, W., Freedman, R., 1995. Sensory gating in parents of schizophrenics. Am. J. Med. Genet. 60, 506–511. Wesnes, K., Warburton, D.M., 1983. Smoking, nicotine and human performance. Pharmacol. Ther. 21, 189–208. Wonnacott, S., 1990. Characterization of nicotinic receptor sites in brain. In: Wonnacott, S., Russel, M.A.H., Strollerman, I.P. ( Eds.), Nicotine Psychopharmacology: Molecular, Cellular and Behavioral Aspects. Oxford University Press, Oxford, pp. 227–277. Ziedonis, D.M., Kosten, T.R., Glazer, W.M., Frances, R.J., 1994. Nicotine dependence and schizophrenia. Hospital and Community Psychiatry 45, 204–206.