بهزیستی ذهنی در اسکیزوفرنی:مقایسه مطالعه تصادفی غیر حقارت کنترل شده برچسب باز 12 ماه با کویتاپین XR ریسپریدون (بازیابی)

Abstract This randomised 12-month open study analysed the effectiveness of quetiapine XR (400–800 mg) versus risperidone (2–6 mg) on subjective well-being in schizophrenia (NCT00600756). Primary objective was to demonstrate non-inferiority of quetiapine XR to risperidone in 6-month responder rate using the Subjective Well-Being under Neuroleptics scale (SWN-K) (per-protocol at Month 6 [PP 6] population). Non-inferiority was defined as the lower limit of the 95% confidence interval (CI) greater than −9.7% for the adjusted difference between quetiapine XR and risperidone. Secondary objectives included non-inferiority of quetiapine XR versus risperidone (lower limit of 95% CI greater than −7.5 points) for SWN-K change from baseline to Month 12 (PP 12). 798 patients were randomised (quetiapine XR, n=395; risperidone, n=403); at Month 12, 212 (54%) and 227 (56%) patients, respectively, completed the study. At Month 6, SWN-K responder rate in the PP 6 population was 65% (136/210) with quetiapine XR and 68% (158/232) with risperidone (adjusted treatment difference: −5.7%; 95% CI: −15.1, 3.7); thus, non-inferiority could not be established. SWN-K change from baseline to Month 12 was 23.2 points for quetiapine XR and 21.1 points for the risperidone group; treatment difference was 2.1 (95% CI: −0.8; 5.0); non-inferiority was established (PP 12). Conclusion: SWN-K response at 6 months was comparable between the two antipsychotics. However, with a lower than expected responder rate and a lower than expected number of evaluable patients in the PP 6 population, non-inferiority was not demonstrated. A secondary objective (SWN-K total score) established non-inferiority of quetiapine XR to risperidone at Month 12.

ntroduction It is widely accepted that multidimensional outcome in schizophrenia comprises three different aspects: symptomatic remission (Andreasen et al., 2005 and Lambert et al., 2010), functional improvement (Emsley et al., 2011) and adequate subjective well-being/quality of life (Lambert et al., 2008). Concurrent and sustained long-term achievement of each of these objectives, so-called recovery, is a desirable property of present and future antipsychotic agents (Correll, 2011). The concept of subjective well-being was originally designed to assess subjective experiences under neuroleptic treatment (Naber, 1995 and Naber et al., 2001). However, further research has expanded the subjective well-being concept to an overall subjective outcome dimension in schizophrenia (Vothknecht and Schoevers, 2011), which is highly associated with quality of life (Vothknecht et al., 2011, Wehmeier et al., 2007 and Wehmeier et al., 2008). Poor subjective well-being is linked to a diagnosis of schizophrenia more than to other psychotic disorders (Lambert et al., 2009a) and is caused by high levels of psychopathology and, similar to quality of life (Tomotake, 2011), specifically by depression, negative symptoms and anxiety (Karow et al., 2005 and Lambert et al., 2009a). Furthermore, adverse subjective well-being is related to high antipsychotic D2 receptor occupancy above 60–70% (de Haan et al., 2000, de Haan et al., 2003 and Mizrahi et al., 2007) and antipsychotic side effects, particularly if they cause a high level of subjective distress (Schimmelmann et al., 2005). Subjective well-being at baseline, and especially its early improvement, are predictors of medication adherence (Karow et al., 2007) and of short- (Schennach-Wolff et al., 2011) and long-term (de Haan et al., 2008) symptomatic remission, as well as short- (Lambert et al., 2007 and Lambert et al., 2009b) and long-term (Lambert et al., 2009c) recovery outcomes. Given the importance of subjective well-being and quality of life from a patient′s point of view and its central position in the patient's recovery definition (Bellack, 2006 and Resnick et al., 2005), it is surprising that we know of only three randomised controlled trials with antipsychotics to date that compare different antipsychotics with regard to subjective well-being or quality of life as the primary endpoint. Two of these trials compare the differential effectiveness of first- and second-generation antipsychotics (Jones et al., 2006 and Silva de Lima et al., 2005) and one trial compares two second-generation antipsychotics (Naber et al., 2005; olanzapine versus clozapine in mostly treatment-resistant patients). To date, we know of no head-to-head randomised controlled trial comparing quetiapine with risperidone for subjective well-being. As such, the present study was designed to assess the long-term subjective well-being of outpatients with schizophrenia, treated with either quetiapine XR or oral risperidone at a flexible dose in a naturalistic setting for a period of one year. Risperidone was chosen as the comparator in this study as it is a well-established oral antipsychotic, and it was hypothesised that comparable results in terms of subjective well-being, quality of life, and effects on depressive symptoms would be observed for quetiapine XR versus risperidone.

Results 3.1. Patients and treatment In total, 798 patients were randomised to receive quetiapine XR (n=395) or risperidone (n=403) ( Figure 1). The first patient enroled on 31 January, 2008 and the last patient completed the study on 22 October, 2009. Baseline characteristics were comparable between the treatment groups, with the possible exception of hospitalisations due to schizophrenia in the previous 6 months (20.7% compared with 15.6% for quetiapine XR and risperidone, respectively, Table 1). At Month 6, 131 patients (33.2%) in the quetiapine XR group and 118 patients (29.3%) in the risperidone group had withdrawn from study treatment ( Figure 1). In total, 185 (46.8%) patients in the quetiapine XR group and 171 (42.4%) patients in the risperidone group were excluded from the PP 6 analysis set. The most common reasons were the SWN-K total score not being available at Month 6 (117 and 107 patients in the quetiapine XR and risperidone groups, respectively) or at least one major protocol deviation (92 and 86 patients in the quetiapine XR and risperidone groups, respectively). In total, 183 patients in the quetiapine XR group and 176 patients in the risperidone group withdrew prematurely from study treatment, with 212 (53.7%) and 227 (56.3%) patients, respectively, completing the study to Month 12. The most common reasons for withdrawal are described in Figure 1. Table 1. Patient demographics and key baseline patient characteristics (ITT). Demographic Quetiapine XRn=379 Risperidonen=392 Sex, n (%) Male 224 (59.1) 225 (57.4) Female 155 (40.9) 167 (42.6) Mean (SD) age, years 39.3 (11.7) 40.0 (11.7) No alcohol or drug abuse or dependence, n (%) 329 (86.8) 338 (86.2) Alcohol use, n (%) 45 (11.9) 48 (12.2) Cannabis use, n (%) 5 (1.3) 10 (2.6) Other, n (%) 4 (1.1) 5 (1.3) DSM-IV schizophrenia subtype diagnosis at baseline, n (%) Paranoid type 239 (63.2) 231 (58.9) Schizoaffective disorder of bipolar type 31 (8.2) 33 (8.4) Schizoaffective disorder of depressive type 26 (6.9) 34 (8.7) Undifferentiated type 28 (7.4) 26 (6.6) Disorganised type 23 (6.1) 24 (6.1) Residual type 19 (5.0) 21 (5.4) Schizophreniform disorder with good prognositic features 10 (2.6) 14 (3.6) Schizophreniform disorder without good prognositic features 2 (0.5) 9 (2.3) Catatonic type 1 (0.3) 0 Median duration of present episode, months (range) 2.0 (0.1–312) 3.0 (0.1–300) Mean (SD) years since first known schizophrenia diagnosis 11.4 (10.4) 11.3 (10.2) Any prior schizophrenia episode, n (%) Yes 333 (87.9) 352 (89.8) No 46 (12.1) 40 (10.2) Mean (SD) no of prior schizophrenia episodes Any 5.3 (5.2) 4.8 (4.4) With hospitalisation 3.6 (5.1) 3.2 (5.4) Mean (SD) no of months since last schizophrenia episode, months 23.6 (48.7) 22.5 (36.0) Hospitalisations due to schizophrenia in previous 6 months, n (%) patients 69 (20.7) 55 (15.6) SWN-K total score, mean (SD) 64.2 (11.4) 64.5 (11.3) SWN-K subscale score, mean (SD) Physical functioning 13.1 (3.4) 13.1 (3.4) Social integration 12.8 (3.4) 12.9 (3.4) Mental functioning 11.8 (3.4) 11.9 (3.2) Self control 12.9 (2.9) 13.1 (3.0) Emotional regulation 13.6 (3.5) 13.5 (3.2) ITT—intent-to-treat; SD—standard deviation. Table options The mean (standard deviation [SD]) duration of exposure to quetiapine XR was 249.2 (138.0) days and the mean duration of exposure to risperidone was 257.8 (135.6) days. The mean (SD) prescribed daily dose over the entire study was 566.5 (140.6) mg/day in the quetiapine XR group and 3.9 (1.3) mg/day in the risperidone group. In the quetiapine XR group, 38.5% of patients were prescribed a mean daily dose <500 mg, 39.8% were prescribed ≥500 mg to <700 mg and 21.6% were prescribed ≥700 mg. In the risperidone group, 23.5% were prescribed <3 mg, 53.1% were prescribed ≥3 mg to <5 mg and 23.5% were prescribed ≥5 mg/day over the course of the study. Mean adherence to study treatment was 97.0% for patients in the quetiapine XR group and 97.3% in the risperidone group. 3.2. Efficacy 3.2.1. Primary objective The SWN-K responder rate at Month 6 in the PP 6 analysis set was 64.8% (136/210) in the quetiapine XR group and 68.1% (158/232) in the risperidone group (Figure 2A). The adjusted difference (quetiapine XR − risperidone, stratified by country) in responder rate between the groups was −5.7% (95% CI: −15.1, 3.7); the lower 95% limit was below the predefined non-inferiority limit of −9.7% (Figure 2B). Non-inferiority for quetiapine XR versus risperidone could not, therefore, be established in terms of responder rate at Month 6. In the ITT analysis set, the SWN-K responder rate at Month 6 was 62.6% (164/262) in the quetiapine XR group and 64.6% (184/285) in the risperidone group. The adjusted difference in responder rate between the groups was −3.4% (95% CI: −11.8, 5.0). Treatment difference in SWN-K responder rates at Month 6 (A) and non-inferiority ... Figure 2. Treatment difference in SWN-K responder rates at Month 6 (A) and non-inferiority analysis at Month 6 (B) (PP 6 analysis set). Difference (Quetiapine XR-risperidone) in response rate (using a Mantel-Haenszel type statistic). Responder, SWN-K total score change from baseline ≥10 points or ≥20% improvement in SWN-K total score; SWN-K, subjective well-being under neuroleptics scale, short form; PP, per protocol; CI, confidence interval. Figure options 3.2.2. Secondary objectives 3.2.2.1. Subjective well-being The least squares mean change in SWN-K total score from baseline to Month 12 was 23.2 points in the quetiapine XR group (n=173) and 21.1 points in the risperidone group (n=191) (difference, 2.1; 95% CI: −0.8, 5.0; Figure 3). The lower 95% limit was above the predefined non-inferiority limit of −7.5 points, thereby indicating non-inferiority for quetiapine XR versus risperidone in terms of change from baseline in SWN-K total score at Month 12. In the ITT analysis set, the least squares mean change in SWN-K total score from baseline to Month 12 was 22.7 points in the quetiapine XR group and 19.4 points in the risperidone group (difference, 3.3; 95% CI: 0.6, 5.9; Figure 3). Treatment difference in least squares mean SWN-K total score at Month 12 ... Figure 3. Treatment difference in least squares mean SWN-K total score at Month 12 (non-inferiority analysis) (PP 12 and ITT analysis sets). Difference (quetiapine XR-risperidone) in LSM change from baseline in SWN-K total score. LSM, least squares means; SWN-K, subjective well-being under neuroleptics scale, short from; ITT, intent-to-treat; PP, per protocol; CI, confidence interval. Figure options There were no significant differences between the groups in terms of mean SWN-K subscale scores (physical functioning [quetiapine XR, 18.0; risperidone, 17.1], social integration [quetiapine XR, 17.5; risperidone, 16.9], mental functioning [quetiapine XR, 16.8; risperidone, 15.9], self-control [quetiapine XR, 17.3; risperidone, 17.0], or emotional regulation [quetiapine XR, 18.2; risperidone, 17.7]) at Month 12 (quetiapine XR, n=210; risperidone, n=227). At Month 6, the SWN-K remission rate was 54.2% (142/262) in the quetiapine XR group compared with 48.1% (137/285) in the risperidone group, with no significant difference between the treatment groups (difference in SWN-K remission rate, 2.9%; 95% CI: −5.7, 11.5). At Month 12, the SWN-K remission rate was 66.2% (139/210) in the quetiapine XR group, compared with 56.4% (138/227) in the risperidone group (difference in SWN-K remission rate, 6.3%; 95% CI: −3.6, 16.2). 3.2.2.2. Schizophrenia symptoms and relapse The mean (SD) change in CGI–SCH overall severity score from baseline to Month 12 was similar in both treatment groups: −1.5 (1.1) in the quetiapine XR group and −1.3 (1.2) in the risperidone group. In total, 83.4% of patients (176/211) were classed as improved for CGI–SCH overall severity in the quetiapine XR group, compared with 78.4% of patients (178/227) in the risperidone group. At Month 12, mean (SD) change from baseline in CGI–SCH severity score for depressive symptoms was −1.3 (1.2) in the quetiapine XR group and −0.8 (1.3) in the risperidone group. The percentage of patients classed as improved for CGI-SCH depressive symptoms was higher in the quetiapine XR group (144/211; 68.2%) than in the risperidone group (131/227; 57.7%: odds ratio for treatment effect, 1.65; 95% CI: 1.01, 2.70). There were no differences between the treatment groups for mean change from baseline to Month 12 in CGI–SCH positive symptom scores (quetiapine XR, −1.3; risperidone, −1.4), negative symptom scores (quetiapine XR, −1.4; risperidone, −1.3) and cognitive symptom scores (quetiapine XR, −1.2; risperidone, −1.1) (see Table 2). Table 2. Change in CGI–SCH scale scores and CDSS total score from baseline to Month 12 (ITT). Scale Quetiapine XRn=379 Risperidonen=392 CGI–SCH overall severity Baseline mean (SD) 3.9 (0.89) 3.8 (0.96) Month 12 mean (SD) 2.3 (0.93) 2.5 (1.11) Change from baseline to Month 12 N 211 227 Mean (SD) −1.5 (1.07) −1.3 (1.15) CGI change score improved, n (%) 176 (83.4%) 178 (78.4%) Treatment effect for improved (95% CI) 1.46 (0.87, 2.43) CGI–SCH positive symptoms Baseline mean (SD) 3.2 (1.27) 3.2 (1.36) Month 12 mean (SD) 1.8 (0.89) 1.8 (1.04) Change from baseline to Month 12 N 211 227 Mean −1.3 (1.22) −1.4 (1.28) (SD) CGI change score improved, n (%) 156 (73.9%) 173 (76.2%) Treatment effect for improved (95% CI) 0.98 (0.57, 1.67) CGI–SCH negative symptoms Baseline mean (SD) 3.8 (1.08) 3.7 (1.14) Month 12 mean (SD) 2.3 (1.00) 2.5 (1.06) Change from baseline to Month 12 N 211 227 Mean (SD) −1.4 (1.18) −1.3 (1.09) CGI change score improved, n (%) 156 (73.9%) 170 (74.9%) Treatment effect for improved (95% CI) 0.94 (0.57, 1.55) CGI–SCH depressive symptoms Baseline mean (SD) 2.9 (1.30) 2.7 (1.38) Month 12 mean (SD) 1.6 (0.75) 1.8 (1.02) Change from baseline to Month 12 N Mean (SD) CGI change score improved, n (%) 211−1.3 (1.22) 144 (68.2%) 227−0.8 (1.27) 131 (57.7%) Treatment effect for improved (95% CI) 1.65 (1.01, 2.70) CGI–SCH cognitive symptoms Baseline mean (SD) 3.5 (1.10) 3.5 (1.12) Month 12 mean (SD) 2.3 (0.94) 2.4 (1.10) Change from baseline to Month 12 N 211 227 Mean (SD) −1.2 (1.14) −1.1 (1.14) CGI change score improved, n (%) 150 (71.1%) 145 (63.9%) Treatment effect for improved (95% CI) 1.52 (0.97, 2.36) CDSS total score Baseline Mean (SD) 6.9 (5.1) 6.6 (5.2) Month 12 Mean (SD) 1.7 (2.4) 2.6 (3.6) Change from baseline to Month 12 N 211 227 Mean (SD) −5.3 (5.10) −3.8 (4.6) Treatment difference (95% CI) −1.0 (−1.6, −0.4) CGI-SCH—Clinical Global Impression-Schizophrenia scale; CDSS—Calgary Depression Scale for Schizophrenia; ITT—intent-to-treat; SD—standard deviation. Table options The mean change in CDSS total score was improved in both treatment groups from baseline to Month 6 and from baseline to Month 12. At Month 6, the mean change from baseline (6.9 in the quetiapine XR group [n=379] and 6.6 in the risperidone group [n=392]) was greater in the quetiapine XR group (−4.3) than in the risperidone group (−2.8; treatment difference, −1.2; 95% CI: −1.8, −0.7). At Month 12, the mean change from baseline was also greater in the quetiapine XR group (−5.3) than in the risperidone group (−3.8; treatment difference, −1.0; 95% CI: −1.6, −0.4) ( Table 2). At Month 12, the percentage of patients classed as having either no depression or mild depression based on CDSS total score was 91.9 in the quetiapine XR group and 81.1 in the risperidone group. CGI–SCH relapse rates were comparable between the treatment groups at Month 12; 11.3% (43/379) in the quetiapine XR group compared with 7.9% (31/392) in the risperidone group (treatment difference, 0.6%; 95% CI: −3.0, 4.2). At Month 6, CGI–SCH relapse rates were 9.2% (35/379) in the quetiapine XR group and 5.6% (22/392) in the risperidone group (treatment difference, 1.3%; 95% CI: −2.1, 4.6). 3.2.2.3. Quality of life and functional outcomes Patient quality of life, measured by the EQ-5D health profile, was similar for both treatment groups at Month 6 and Month 12. The mean (SD) change from baseline to Month 12 in EQ-5D index score was 0.21 (0.25) in the quetiapine XR group and 0.16 (0.24) in the risperidone group. In terms of functional improvement at Month 12, 8/211 patients (3.8%) in the quetiapine XR group and 7/227 patients (3.1%) in the risperidone group reported a real improvement in both occupational and residential status from baseline; 160/211 patients (75.5%) in the quetiapine XR group and 171/227 patients (75.3%) in the risperidone group reported being in stable state for occupational and residential status as recorded at baseline. 3.3. Safety and tolerability The overall incidence of TEAEs was 60.9% in the quetiapine XR group and 64.2% in the risperidone group. The most common TEAEs are shown in Table 3. The incidence of drug-related TEAEs was 50.6% in the quetiapine XR group and 50.7% in the risperidone group. In total, 57 patients (14.6%) in the quetiapine XR group and 48 patients (11.9%) in the risperidone group withdrew from the study as a result of a TEAE in the safety analysis set. Forty-five patients (11.5%) in the quetiapine XR group and 26 patients (6.5%) in the risperidone group experienced serious TEAEs. The most common serious TEAEs (n≥3) were schizophrenia (3.1%), psychotic disorder (2.3%), agitation (0.8%) and suicide attempt (0.8%) in the quetiapine XR group, and schizophrenia (2.1%) and psychotic disorder (1.8%) in the risperidone group. Drug related serious TEAEs were experienced by 13 patients (3.3%) in the quetiapine XR group (schizophrenia [1.5%], psychotic disorder [1.3%], confusional state [0.3%], psychomotor hyperactivity [0.3%] and schizoaffective disorder [0.3%]) and 4 patients (1.0%) in the risperidone group (schizophrenia [0.5%], extrapyramidal disorder [0.2%] and neuroleptic malignant syndrome [0.2%]). One death occurred as a result of a serious TEAE (brainstem stroke) in the risperidone treatment group, which was not determined to be drug related. At least one extrapyramidal TEAE was observed in 38 patients (9.7%) in the quetiapine XR group and 83 patients (20.6%) in the risperidone group. The most common extrapyramidal TEAEs in the quetiapine XR group were hyperkinesia (1.8%), hypokinesia (1.8%), restlessness (1.8%) and tremor (1.8%). The most common extrapyramidal TEAEs in the risperidone group were tremor (7.2%), hypokinesia (4.5%), muscle rigidity (4.0%) and extrapyramidal disorder (3.7%). Table 3. Overview of treatment-emergent adverse events at 12 months and most common treatment-emergent adverse events by preferred term (≥5% in one treatment group) (safety analysis set). Quetiapine XRN=391 RisperidoneN=402 n(%) Number of events n(%) Number of events TEAE 238 (60.9) 791 258 (64.2) 834 Drug-related TEAE 198 (50.6) 493 204 (50.7) 522 TEAE leading to permanent discontinuation of the study 57 (14.6) 72 48 (11.9) 80 Serious TEAEs 45 (11.5) 49 26 (6.5) 31 Drug-related serious TEAEs 13 (3.3) 14 4 (1.0) 4 Serious TEAEs leading to death 0 0 1 (0.2) 1 TEAE Somnolence 71 (18.2) 76 47 (11.7) 48 Insomnia 30 (7.7) 34 52 (12.9) 55 Anxiety 33 (8.4) 34 38 (9.5) 39 Constipation 40 (10.2) 44 23 (5.7) 24 Headache 23 (5.9) 24 39 (9.7) 44 Aptyalism 33 (8.4) 36 51 (5.2) 23 Asthenia 22 (5.6) 22 25 (6.2) 29 Nausea 18 (4.6) 19 26 (6.5) 26 Weight increased 18 (4.6) 18 25 (6.2) 25 Dizziness postural 23 (5.9) 24 18 (4.5) 19 Sedation 25 (6.4) 26 15 (3.7) 15 Tension 18 (4.6) 23 20 (5.0) 20 Tremor 7 (1.8) 8 29 (7.2) 29 N—number of assessable patients, n (%)—patients with individual TEAE. TEAE—treatment-emergent adverse event. Table options Individual symptom scores, measured by a modified UKU side-effect rating scale, are described in Table 4. Table 4. Analysis of individual symptom scores, by modified UKU (safety analysis set). Solicited AE,n/N(%) Treatment group Month 6 Month 12 Autonomic Quetiapine XR 40/263 (15.2) 24/210 (11.4) Risperidone 34/284 (12.0) 19/224 (8.5) P value 0.269 0.305 Psychic Quetiapine XR 29/262 (11.1) 14/211 (6.6) Risperidone 37/285 (13.0) 13/225 (5.8) P value 0.492 0.711 Neurologic Quetiapine XR 10/262 (3.8) 4/209 (1.9) Risperidone 30/284 (10.6)⁎⁎ 20/224 (8.9)⁎⁎⁎ P value 0.003 0.001 Sexual dysfunction (men) Quetiapine XR 9/151 (6.0) 9/120 (7.5) Risperidone 20/154 (13.0)⁎ 13/125 (10.4) P value 0.036 0.427 Hyperprolactinaemia (men) Quetiapine XR 1/151 (0.7) 1/120 (0.8) Risperidone 2/156 (1.3) 0/125 (0) P value 1.000 0.490 Sexual dysfunction (women) Quetiapine XR 5/103 (4.8) 4/85 (4.7) Risperidone 3/124 (2.4) 4/96 (4.2) P value 0.473 1.000 Hyperprolactinaemia (women) Quetiapine XR 0/106 (0) 0/87 (0) Risperidone 16/126 (12.7)⁎⁎⁎ 10/97 (10.3)⁎⁎ P value <0.001 0.002 N—number of assessable patients, n (%)—patients with individual AE, AE—adverse event; UKU—Udvalg for Kliniske Undersøgelser, side-effect rating scale. ⁎ p<0.05. ⁎⁎ p<0.01. ⁎⁎⁎ p≤0.001 quetiapine XR versus risperidone Table options Mean (SD) weight change from baseline to Month 12 was 1.0 (5.6) kg in the quetiapine XR group and 1.1 (5.3) kg in the risperidone treatment group. In total, 63.7% of patients in the risperidone group shifted from a normal prolactin level (normal prolactin range: male, 0–14 ng/ml; female, 0–24 ng/ml) at baseline to a high prolactin level (above the normal prolactin range) at Month 12 compared with 15.1% in the quetiapine XR group. Risperidone was associated with a mean (SD) increase in serum prolactin levels between baseline and Month 12 (16.0 [46.3] ng/ml). Quetiapine XR was associated with a mean (SD) decrease in serum prolactin levels between baseline and Month 12 (−7.7 [32.9] ng/ml). With the exception of serum prolactin, there were no meaningful differences in clinical laboratory parameters (glucose, liver enzymes and absolute neutrophil counts) and vital signs between the treatment groups. The mean (SD) change from baseline to Month 12 in plasma glucose was 0.14 (1.06) mmol/l in the quetiapine XR group and 0.22 (1.71) mmol/l in the risperidone group.