- Email: [email protected]
Early prediction of olanzapine-induced weight gain for schizophrenia patients
Psychiatry Research 263 (2018) 207–211
Contents lists available at ScienceDirect
Psychiatry Research journal homepage: www.elsevier.com/locate/psychres
Early prediction of olanzapine-induced weight gain for schizophrenia patients Ching-Hua Lina,b, Shih-Chi Lina, Yu-Hui Huanga, Fu-Chiang Wanga, Chun-Jen Huangb,c,
T
⁎
a
Kaohsiung Municipal Kai-Syuan Psychiatric Hospital, Kaohsiung, Taiwan Department of Psychiatry, School of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan c Department of Psychiatry, Kaohsiung Medical University Hospital, Taiwan b
A R T I C LE I N FO
A B S T R A C T
Keywords: Schizophrenia Olanzapine Weight gainers Early prediction Receiver operating characteristic curve
The aim of this study was to determine whether weight changes at week 2 or other factors predicted weight gain at week 6 for schizophrenia patients receiving olanzapine. This study was the secondary analysis of a six-week trial for 94 patients receiving olanzapine (5 mg/d) plus trifluoperazine (5 mg/d), or olanzapine (10 mg/d) alone. Patients were included in analysis only if they had completed the 6-week trial (per protocol analysis). Weight gain was defined as a 7% or greater increase of the patient's baseline weight. The receiver operating characteristic curve was employed to determine the optimal cutoff points of statistically significant predictors. Eleven of the 67 patients completing the 6-week trial were classified as weight gainers. Weight change at week 2 was the statistically significant predictor for ultimate weight gain at week 6. A weight change of 1.0 kg at week 2 appeared to be the optimal cutoff point, with a sensitivity of 0.92, a specificity of 0.75, and an AUC of 0.85. Using weight change at week 2 to predict weight gain at week 6 is favorable in terms of both specificity and sensitivity. Weight change of 1.0 kg or more at 2 weeks is a reliable predictor.
1. Introduction Obesity is prevalent among schizophrenia patients (Dickerson et al., 2006), and many antipsychotics used to treat schizophrenia may also result in weight gain (Allison et al., 1999b). For example, weight gain associated with olanzapine is commonly reported (Allison et al., 1999b; Duggan et al., 2005). Weight gain can lead to various medical complications such as diabetes mellitus, dyslipidemia, cardiovascular disease, and osteoarthritis (Bellanger and Bray, 2005; Huxley et al., 2010; Kawachi, 1999; Newcomer and Haupt, 2006). In addition, weight gain induced by antipsychotics may contribute to low self-esteem, medication nonadherence and subsequent relapse (Allison et al., 1999a). Psychiatrists need to balance the risk of weight gain against the benefit of symptomatic control. Therefore, reliable information about the risk for weight gain is required for patients receiving antipsychotics. The mechanism of weight gain during treatment with olanzapine is not fully understood, and is likely to be multifactorial (Albaugh et al., 2006). In studies on the prediction of weight gain for patients taking antipsychotics, risk factors already identified include male gender, nonwhite ethnicity, first-episode psychosis, increased appetite, better clinical response, lower baseline body mass index (BMI), younger age, and gene polymorphism (Basson et al., 2001; Kinon et al., 2005; Ou
⁎
et al., 2013; Ujike et al., 2008). Increased appetite is most likely a result of the antipsychotic medication, probably because antipsychotic medication is an antagonist of serotonergic 5-HT2c receptors, histaminergic H1-receptors, and dopamine receptors (Bak et al., 2014; Starrenburg and Bogers, 2009). However, to date, few of these risk factors have led to the development of a clinically useful decision-making tool. Three industry-sponsored studies have reported that early weight increase during the course of olanzapine treatment may predict substantial weight gain following long-term use (Kinon et al., 2005; Lipkovich et al., 2006, 2008). A recent meta-analysis study (Musil et al., 2015) has concluded that weight increase is common during the first weeks of treatment. These findings indicate that early prediction of weight gain and, consequently, the timely implementation of weight management strategies is of great benefit. Consensus guidelines emphasize the importance of appropriate baseline screening and measuring of body weight, beginning in the initial weeks of treatment for patients taking olanzapine (American Diabetes Association, 2004; De Hert et al., 2009). We recently conducted a six-week, randomized, double-blind study to compare the efficacy and safety of low-dose olanzapine (5 mg/d) plus low-dose trifluoperazine (5 mg/d) versus full-dose olanzapine (10 mg) in the acute treatment of schizophrenia (Lin et al., 2017). One of the
Corresponding author at: Department of Psychiatry, Kaohsiung Medical University Hospital, 100, Shih-Chuan 1st Road, San Ming District, Kaohsiung 807, Taiwan. E-mail address: [email protected] (C.-J. Huang).
https://doi.org/10.1016/j.psychres.2018.02.058 Received 19 July 2017; Received in revised form 24 December 2017; Accepted 28 February 2018 0165-1781/ © 2018 Elsevier B.V. All rights reserved.
Psychiatry Research 263 (2018) 207–211
C.-H. Lin et al.
findings was that no significantly different changes in body weight occurred at weeks 1, 2, 3, 4, and 6 between the two groups after treatment. The explanation for this is that weight change due to olanzapine is not dose-related within a range of 5–20 mg/day (Kinon et al., 2001), whereas weight increase due to trifluoperazine is unusual (Stahl, 2014). Therefore, possible weight change induced by trifluoperazine may be minimal if patients receive 5 mg daily of trifluoperazine plus 5 mg daily of olanzapine for short-term treatment. The aim of the study was to determine whether weight change, percentage of weight change, and/or BMI change at weeks 1 or 2 predicted weight gain at week 6 for schizophrenia patients receiving olanzapine treatment, and therefore to determine which predictor was most powerful and user-friendly.
7% or greater increase of the baseline weight (kg) to endpoint. Weight gainers and non-weight gainers were compared in terms of sex, age, age at onset, baseline PANSS, baseline weight (kg), baseline BMI (kg/m2), weight change at week 1 (kg), percentage of weight change at week 1, BMI change at week 1, weight change at week 2 (kg), percentage of weight change at week 2, and BMI change at week 2. Age at onset was regarded as the age at which the first psychotic symptoms occurred. Each statistically significant predictor at weeks 1 and 2 from the univariate analysis was separately entered into a receiver operating characteristic (ROC) analysis for determining which one was a powerful and user-friendly predictor. The ROC curve was used to determine the optimal cutoff point of predictor between the weight gainers and nonweight gainers, maximizing both the sensitivity and specificity of the predictor(s) so that false positive and false negative rates could be minimized. The area under the ROC curve (AUC) is a measure of the overall discriminative power. In practice, an AUC ≥ 0.8 indicates a good discriminative capacity (Weinstein and Fineberg, 1980). All data were processed by SPSS version 17.0 for Windows (SPSS Inc., Chicago, IL, USA). Pearson χ2 test was used to compare categorical variables; independent t-test for continuous variables. All tests were two-tailed, and significance was defined as an alpha of less than 0.05.
2. Methods 2.1. Ethics The study was the secondary analysis of a clinical trial, conducted from January 2012 to February 2016, as documented elsewhere (Lin et al., 2017). This study was approved by the Kai-Syuan Psychiatric Hospital's institutional review board (IRB) and conducted in accordance with Good Clinical Practice procedures, the Declaration of Helsinki, and national legal requirements (Taiwan's Human Subjects Research Act). This study was registered on Clinicaltrials.gov (Identifier number: NCT02704962).
3. Results 3.1. Patient characteristics
2.2. Patients, study design, procedures, and body weight assessment
A total of 94 newly hospitalized schizophrenia patients with acute exacerbation were randomly assigned for treatment with olanzapine plus trifluoperazine (N = 47) or olanzapine alone (N = 47). No antipsychotic-naive or first episode schizophrenia patients were enrolled. Twenty-seven of the 94 patients dropped out early from the study. One patient withdrew from the trial after 2 weeks of treatment due to a weight increase of 5.5 kg. Sixty-seven patients who completed the 6week trial were enrolled in the analysis. The mean age of the patients was 39.4 (SD = 8.5) years, and 47.8% (N = 32) were males. The mean PANSS score of 95.0 (SD = 16.5) at baseline reflected a fairly severely symptomatic population (Table 1). Eleven of the 67 patients (16.4%) were classified as weight gainers. There were no between-group differences in terms of sex, age, age at onset, baseline PANSS, baseline weight, and baseline BMI (Table 1). Table 2 lists the mean body weight, mean BMI, mean weight change, mean percentage of weight change, and mean BMI change of the two groups at each assessment.
As previously described in detail (Lin et al., 2017), all newly hospitalized schizophrenia patients for acute treatment were evaluated by trained and board-certified psychiatrists. Inpatients were enrolled into this study if they: 1) were physically healthy and had all laboratory parameters within normal limits, 2) were aged between 18 and 55 years, 3) satisfied DSM-IV (APA, 2000) criteria for schizophrenia, 4) had a baseline Clinical Global Impression-Severity of Illness Scale (CGIS) (Guy, 1976) ≥4, 5) had no DSM-IV diagnosis of substance abuse or dependence (including alcohol) in the past 6 months, 6) had not received depot antipsychotics during the preceding three months, and 7) had not taken olanzapine for at least 4 weeks before participating in the trial. Patients excluded from this study were: 1) those with histories of serious adverse reactions to olanzapine or trifluoperazine, 2) those with histories of tardive dyskinesia or neuroleptic malignant syndrome, 3) female subjects who were pregnant or at risk of pregnancy or lactation, and 4) those diagnosed with treatment-resistant schizophrenia (Kane et al., 1988), or having previously received clozapine or electroconvulsive therapy. Written informed consent was obtained from every patient and his or her legal guardian. After a washout period of at least 3 days, patients were randomly assigned to a 1:1 ratio of 5 mg/day of olanzapine plus 5 mg/day of trifluoperazine, or 10 mg/day of olanzapine alone for 6 weeks. Patient adherence and safety were closely monitored by the research staff. Benzodiazepine was allowed as needed for insomnia or agitation. Biperiden, up to 6 mg/day, was also allowed to treat motor side effects, but prophylactic use of biperiden was prohibited. No other psychotropic agents were permitted during the study. Symptom severity was rated using the Positive and Negative Syndrome Scale (PANSS) (Kay et al., 1987). At weeks 0, 1, 2, 3, 4, and 6 (or upon early termination), body weight (kg) was measured in the morning after emptying the bladder and before breakfast, using a highquality, calibrated digital scale, with the patient wearing indoor clothing but no shoes.
3.2. Early prediction of weight gain Weight change at week 1, percentage of weight change at week 1, and BMI change at week 1 were significant predictors of weight gainers at week 1. Similarly, weight change at week 2, percentage of weight change at week 2, and BMI change at week 2 were also significant predictors of weight gainers at week 2 (Table 1). ROC analysis was employed to determine the optimal cutoff point and AUC of each significant predictor. Table 3 shows the cutoff point, sensitivity, specificity, predictive power, and AUC of each predictor at weeks 1 and 2. At week 1, all the AUCs provided by weight change (0.75), percentage of weight change (0.75), and BMI change (0.75) were less than 0.8. At week 2, a weight change of 1.0 kg appeared to be the optimal cutoff point for predicting weight gainers, with a sensitivity of 0.92, a specificity of 0.75, a predictive power of 0.88, and an AUC of 0.85. A weight change of 1.23% appeared to be the optimal cutoff point for predicting weight gain, with a sensitivity of 0.91, a specificity of 0.70, a predictive power of 0.81, and an AUC of 0.84. A BMI change of 0.36 appeared to be the optimal cutoff point for predicting weight gain, with a sensitivity of 0.91, a specificity of 0.75, a predictive power of 0.83, and an AUC of 0.85. ROC curves of weight change at weeks 1 and 2 are illustrated in Fig. 1.
2.3. Statistical analyses Patients were included in the analysis only if they had completed the 6-week trial (per protocol analysis). Weight gain was defined as a 208
Psychiatry Research 263 (2018) 207–211
C.-H. Lin et al.
Table 1 Comparison of clinical characteristics between weight gainersa and non-weight gainers. Variable
Total (N = 67)
Weight gainersa (n = 11)
Non-weight gainers (n = 56)
Sex Male, n (%) Female, n (%) Age, mean (SD), years Age at onset, mean (SD), years Baseline PANSSb, mean (SD) Baseline weight, mean (SD), kg Baseline BMIc, mean (SD), kg/m2 Weight change at week 1, mean (SD), kg Percentage of weight change at week 1, mean (SD), kg BMI change at week 1 Weight change at week 2, mean (SD), kg Percentage of weight change at week 2, mean (SD), kg BMI change at week 2
32 (47.8) 35 (52.2) 39.4 (8.5) 25.4 (8.1) 95.0 (16.5) 63.6 (13.7) 24.0 (4.5) 0.2 (1.1) 0.5 (1.9) 0.1 (0.4) 0.8 (1.8) 1.4 (3.0) 0.3 (0.7)
5 (45.5) 6 (54.5) 37.2 (8.6) 21.7 (5.7) 92.4 (14.2) 58.7 (9.2) 22.4 (3.6) 1.1 (1.5) 2.0 (2.6) 0.4 (0.5) 2.7 (1.8) 5.0 (3.7) 0.0 (0.4)
27 (48.2) 29 (51.8) 39.9 (8.5) 26.1 (8.4) 95.5 (16.9) 64.6 (14.3) 24.3 (4.6) 0.1 (1.0) 0.2 (1.5) 1.0 (0.7) 0.4 (1.5) 0.7 (2.2) 0.1 (0.6)
p 0.884d
0.336e 0.100e 0.570e 0.198e 0.187e 0.037e 0.049e 0.032e < 0.000 0.003e < 0.000
Bold, statistically significant. a weight gainers = patients who experienced at least 7% weight increase at end point. b PANSS = Positive and Negative Syndrome Scale. c BMI = body mass index. d Pearson's χ2 test. e independent t-test.
4. Discussion
the other hand, 75% of patients (i.e., specificity) with a weight change < 1 kg at week 2 were correctly identified as non-weight gainers. Compared to a study conducted in the U.S. by Simpson et al. (2004) for schizophrenia inpatients (N = 133) treated with 5–15 mg daily of olanzapine for 6 weeks, weight gain was defined as a 7% increase in weight from baseline. About 36% of patients were classified as weight gainers (Parsons et al., 2009). There were only 4 Asian patients (3.0%) enrolled in this study. However, in a Taiwanese study (Chiu et al., 2010), schizophrenia inpatients received 5–20 mg daily of olanzapine for 8 weeks. Six of 33 completers (18.2%) met the weight gain criterion (i.e., ≥7% weight gain). This result is comparable to our findings, and indicates that ethnic factors should be considered (Basson et al., 2001; Citrome et al., 2011). Study by Basson et al. (2001) found that males gained significantly more weight than females if patients with schizophrenia received olanzapine. In the present study, males and females had comparable rates of weight gain after the 6-week treatment (Table 1). To avoid loss of power from dichotomizing a continuous variable (i.e., weight gainers vs. non-weight gainers) (Streiner, 2002; Taylor et al., 2006), weight change at week 6 (i.e., a continuous variable) was regarded as a dependent variable. Analysis of sex difference was performed by analysis of covariance, with sex as a fixed factor and baseline weight as covariate. There was no significant between-sex difference (p = 0.845) in body weight change at week 6 (data not shown in the table). A per protocol analysis for a trial may be biased because
The first main finding of this study was that as early as week 1, weight gainers had significantly greater weight change, percentage of weight change, and BMI change than non-weight gainers (see Table 1). However, these three predictors at week 1 did not provide adequate discriminative capacity (i.e., AUC < 0.8) (see Table 3). The second main finding was that weight change at week 2, percentage of weight change at week 2, or BMI change at week 2 provided comparable AUCs (i.e., 0.84 vs. 0.85 vs. 0.84). The explanation of such comparable AUCs among three predictors was that weight change at week 2, percentage of weight change at week 2, and BMI change at week 2 were highly correlated each other (r2 = 0.94, p < 0.001 for weight change at week 2 vs. percentage of weight change at week 2; r2 = 0.99, p < 0.001 for weight change at week 2 vs. BMI change at week 2; r2 = 0.95, p < 0.001 for percentage of weight change at week 2 vs. BMI change at week 2). All AUCs provided by the 3 predictors at week 2 were greater than 0.80. Therefore, weight change at week 2, percentage of weight change at week 2, or BMI change at week 2 had good discriminative capacity for predicting an ultimate 7% or more weight increase (Table 3). Compared with using a percentage of weight change of 1.23% at week 2 or a BMI change of 0.36 at week 2 for predicting ultimate weight gainers, a weight change of 1.0 kg at week 2 was more user-friendly in clinical practice. A weight change ≥ 1.0 kg at week 2 of initiating olanzapine treatment for patients with schizophrenia correctly identified 92% of weight gainers at endpoint (i.e., sensitivity). On
Table 2 Mean body weight, mean BMI, mean weight change, mean percentage of weight change, and mean BMI change of the two groups at each assessment (N = 67).
Weight gainers Body weight, mean (SD), kg BMI, mean (SD), kg/m2 Weight change Percentage of weight change BMI change Non-weight gainers Body weight, mean (SD), kg BMI, mean (SD), kg/m2 Weight change Percentage of weight change BMI change
Week 0 Mean (SD)
Week 1 Mean (SD)
Week 2 Mean (SD)
Week 3 Mean (SD)
Week 4 Mean (SD)
Week 6 Mean (SD)
58.7 (9.2) 22.4 (3.6)
59.9 (9.5) 22.8 (3.7) 1.1 (1.5) 2.0 (2.6) 0.4 (0.5)
61.5 (8.4) 23.4 (3.3) 2.7 (1.8) 5.0 (3.7) 1.0 (0.7)
62.0 (8.6) 23.6 (3.4) 3.3 (2.0) 5.9 (4.1) 1.2 (0.7)
63.0 (8.4) 24.0 (3.2) 4.2 (2.7) 7.7 (5.6) 1.6 (1)
64.3 (9.7) 24.5 (3.6) 5.5 (1.8) 9.6 (3.7) 2.1 (0.6)
64.6 (14.3) 24.3 (4.6)
64.7 (14.0) 24.4 (4.5) 0.1 (1.0) 0.2 (1.5) 0.02 (0.4)
65.0 (14.2) 24.5 (4.5) 0.4 (1.5) 0.7 (2.2) 0.1 (0.6)
65.1 (14.4) 24.5 (4.6) 0.5 (1.6) 0.9 (2.5) 0.2 (0.6)
65.3 (14.2) 24.6 (4.6) 0.7 (1.8) 1.2 (2.5) 0.3 (0.7)
65.6 (13.9) 24.7 (4.5) 1.0 (2.0) 1.7 (2.9) 0.4 (0.7)
209
Psychiatry Research 263 (2018) 207–211
C.-H. Lin et al.
Table 3 Cutoff point, sensitivity, specificity, predictive power, and AUC of each predictor at weeks 1 and 2: ROC Predictors Week 1 Weight change Percentage of weight change BMI change Week 2 Weight change Percentage of weight change BMI change b
a
analysis.
Cutoff point
Sensitivity (True positive)
Specificity (True negative)
Predictive power
AUCa (95% CI)
0.50 0.93 0.15
0.64 0.73 0.72
0.80 0.79 0.71
0.75 0.76 0.72
0.75 (0.63–0.85) 0.75 (0.63–0.85) 0.75 (0.63–0.84)
1.00 1.23 0.36
0.92 0.91 0.91
0.75 0.70 0.75
0.88 0.81 0.83
0.85 (0.74–0.93) 0.84 (0.74–0.92) 0.85 (0.75–0.93)
AUC = area under the ROC curve. a ROC = receiver operating characteristic.
after one year of treatment. The second study, by Lipkovich et al. (2006), reported that a weight increase of 2 − 3 kg during the first 3 weeks of olanzapine treatment for patients with bipolar disorder predicted substantial weight gain by 30 weeks. Weight change at 3 weeks plus baseline characteristics for predicting weight gain provided an AUC of 0.80. The third study (Lipkovich et al., 2008), indicates that weight change from baseline to weeks 1, 2, 3, or 4 alone could be used to predict substantial weight gain defined as at least a 7% weight increase. However, these studies (Kinon et al., 2005; Lipkovich et al., 2008) did not provide the AUC values for potential predictors. Two of the three studies (Kinon et al., 2005; Lipkovich et al., 2006) did not reveal the potential predictors at 2 weeks. Several strengths of this study should be addressed. First, our study provided data on Asian patients for comparison. Second, the subjects were inpatients for acute treatment. In inpatient settings, patients are carefully monitored, and confounding factors (Strassnig et al., 2003) related to weight gain, such as low levels of physical activity and inappropriate food intake, could be minimized. Third, some proposed cutoff points for predicting substantial weight gain may be chosen arbitrarily. Sensitivity and specificity analyses were performed in the present study to determine the optimal cut-off for predicting weight gain. The optimal cut-off point, a weight change of 1.0 kg at week 2, to assess the individual's risk of experiencing substantial weight gain at week 6, is simple for clinicians to remember and apply. In addition, because this was a racially homogeneous group for study, the confounding variable of ethnicity could be decreased. There were also some limitations to this study. First, this was a trial with small sample sizes, a 6-week design, and was conducted in only one psychiatric center. The results may not therefore be applicable to weight change which occurs during long-term treatment. However, the initial weight gain within the first 6 weeks of pharmacotherapy is most important, as patients will not lose weight thereafter (Bak et al., 2014). Second, although there were no significant differences in weight gain between the two groups (i.e., 5 mg/d of olanzapine plus 5 mg/d of trifluoperazine versus 10 mg/d of olanzapine) in our recent study (Lin et al., 2017), a meta-analysis study indicates that some patients taking trifluoperazine may gain weight (Koch et al., 2014). The possibility that receiving one of these treatments predicts weight gain at follow-up should therefore be explored. Third, generalization may not extend to those patients aged more than 55 years old, to patients treated with other atypical antipsychotics, to antipsychotic-naive patients (Bak et al., 2014; Tarricone et al., 2010), to first episode schizophrenia patients (Álvarez-Jiménez et al., 2008), or to patients with other mental disorders (Allison et al., 1999b; Basson et al., 2001; Moteshafi et al., 2012). Fourth, the original study was not designed for estimating the predictive value of a weight gain in the population. Finally, use of an all-Taiwanese sample may limit generalizability to other countries. In conclusion, a weight change of 1 kg or more at week 2 could be used to predict a 7% or more weight increase for schizophrenia patients after 6 weeks of treatment with olanzapine. This finding provided a sensitivity of 0.92, a specificity of 0.75, and an AUC of 0.85. Further
Fig. 1. The ROC curves for weight gainers versus non-weight gainers at weeks 1 and 2.
participants who fail to complete the trial will be excluded (Sedgwick, 2015). If analysis was on an intent-to-treat basis, missing data would be replaced according to the principle of last observation carried forward. Twelve of the 94 patients (12.8%) were classified as weight gainers. There were no between-group differences in terms of sex, age, age at onset, baseline PANSS, baseline weight, and baseline BMI. Weight change at week 2 was still a significant predictor of weight gainers, and was therefore entered into a ROC analysis. A weight change of 1.0 kg at week 2 still appeared to be the optimal cutoff point, with a sensitivity of 0.92, a specificity of 0.74, and an AUC of 0.88 (data not shown in the table). Symptomatic change at 2 weeks has also been reported to be a powerful predictor of clinical response or remission in antipsychotic treatment (Ascher-Svanum et al., 2008; Samara et al., 2015). Our major finding, that weight change of 1.0 kg or more at 2 weeks predicts substantial weight gain at week 6, can provide reliable information for treating psychiatrists to make risk/benefit-based decisions about continuing or changing treatment. If the treating psychiatrist decides to continue olanzapine treatment for an individual patient at risk of weight gain, remedies for preventing or reducing further weight increase should be applied early in the course of treatment (Lipkovich et al., 2008; Musil et al., 2015). These remedies should include dietary advice, exercise, behavioral interventions, or pharmacological treatment (Daumit et al., 2013; Ganguli, 2007; Mizuno et al., 2014). The three studies (Kinon et al., 2005; Lipkovich et al., 2006, 2008) mentioned above found that early increase in weight is a clinically useful predictor for gaining weight after long-term treatment with olanzapine. The first study (Kinon et al., 2005), revealed that schizophrenic patients with at least a 7% weight increase during the first 6 weeks of treatment were at greatest risk for gaining the most weight 210
Psychiatry Research 263 (2018) 207–211
C.-H. Lin et al.
studies, preferably involving larger inpatient or outpatient groups from multi-centers, and treatment durations of longer than 6 weeks, are needed to better confirm the predictive power of weight change at week 2 in experiencing substantial weight gains after olanzapine treatment.
schizophrenic. A double-blind comparison with chlorpromazine. Arch. Gen. Psychiatry 45, 789–796. Kawachi, I., 1999. Physical and psychological consequences of weight gain. J. Clin. Psychiatry 60 (Suppl. 21), S5–S9. Kay, S.R., Fiszbein, A., Opler, L.A., 1987. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr. Bull. 13, 261–276. Kinon, B.J., Basson, B.R., Gilmore, J.A., Tollefson, G.D., 2001. Long-term olanzapine treatment: weight change and weight-related health factors in schizophrenia. J. Clin. Psychiatry 62, 92–100. Kinon, B.J., Kaiser, C.J., Ahmed, S., Rotelli, M.D., Kollack-Walker, S., 2005. Association between early and rapid weight gain and change in weight over one year of olanzapine therapy in patients with schizophrenia and related disorders. J. Clin. Psychopharmacol. 25, 255–258. Koch, K., Mansi, K., Haynes, E., Adams, C.E., Sampson, S., Furtado, V.A., 2014. Trifluoperazine versus placebo for schizophrenia. Cochrane Database Syst. Rev (CD010226). Lin, C.H., Wang, F.C., Lin, S.C., Huang, Y.H., Chen, C.C., 2017. A randomized, doubleblind, comparison of the efficacy and safety of low-dose olanzapine plus low-dose trifluoperazine versus full-dose olanzapine in the acute treatment of schizophrenia. Schizophr. Res. 185, 80–87. Lipkovich, I., Citrome, L., Perlis, R., Deberdt, W., Houston, J.P., Ahl, J., et al., 2006. Early predictors of substantial weight gain in bipolar patients treated with olanzapine. J. Clin. Psychopharmacol. 26, 316–320. Lipkovich, I., Jacobson, J.G., Hardy, T.A., Hoffmann, V.P., 2008. Early evaluation of patient risk for substantial weight gain during olanzapine treatment for schizophrenia, schizophreniform, or schizoaffective disorder. BMC Psychiatry 8, 78. Mizuno, Y., Suzuki, T., Nakagawa, A., Yoshida, K., Mimura, M., Fleischhacker, W.W., et al., 2014. Pharmacological strategies to counteract antipsychotic-induced weight gain and metabolic adverse effects in schizophrenia: a systematic review and metaanalysis. Schizophr. Bull. 40, 1385–1403. Moteshafi, H., Zhornitsky, S., Brunelle, S., Stip, E., 2012. Comparing tolerability of olanzapine in schizophrenia and affective disorders: a meta-analysis. Drug Saf. 35, 819–836. Musil, R., Obermeier, M., Russ, P., Hamerle, M., 2015. Weight gain and antipsychotics: a drug safety review. Expert Opin. Drug Saf. 14, 73–96. Newcomer, J.W., Haupt, D.W., 2006. The metabolic effects of antipsychotic medications. Can. J. Psychiatry 51, 480–491. Ou, J.J., Xu, Y., Chen, H.H., Fan, X., Gao, K., Wang, J., et al., 2013. Comparison of metabolic effects of ziprasidone versus olanzapine treatment in patients with first-episode schizophrenia. Psychopharmacology 225, 627–635. Parsons, B., Allison, D.B., Loebel, A., Williams, K., Giller, E., Romano, S., et al., 2009. Weight effects associated with antipsychotics: a comprehensive database analysis. Schizophr. Res. 110, 103–110. Samara, M.T., Leucht, C., Leeflang, M.M., Anghelescu, I.G., Chung, Y.C., Crespo-Facorro, B., et al., 2015. Early improvement as a predictor of later response to antipsychotics in schizophrenia: a diagnostic test review. Am. J. Psychiatry 172, 617–629. Sedgwick, P., 2015. Intention to treat analysis versus per protocol analysis of trial data. BMJ 350, 681. Simpson, G.M., Glick, I.D., Weiden, P.J., Romano, S.J., Siu, C.O., 2004. Randomized, controlled, double-blind multicenter comparison of the efficacy and tolerability of ziprasidone and olanzapine in acutely ill inpatients with schizophrenia or schizoaffective disorder. Am. J. Psychiatry 161, 1837–1847. Stahl, S.M., 2014. Stahl’s Essential Psychopharmacology: Prescriber’s Guide, Fifth edition. ed. Starrenburg, F., Bogers, J., 2009. How can antipsychotics cause diabetes mellitus? Insights based on receptor-binding profiles, humoral factors and transporter proteins. Eur. Psychiatry 24, 164–170. Strassnig, M., Brar, J.S., Ganguli, R., 2003. Nutritional assessment of patients with schizophrenia: a preliminary study. Schizophr. Bull. 29, 393–397. Streiner, D.L., 2002. Breaking up is hard to do: the heartbreak of dichotomizing continuous data. Can. J. Psychiatry 47, 262–266. Tarricone, I., Ferrari Gozzi, B., Serretti, A., Grieco, D., Berardi, D., 2010. Weight gain in antipsychotic-naive patients: a review and meta-analysis. Psychol. Med. 40, 187–200. Taylor, A.B., West, S.G., Aiken, L.S., 2006. Loss of power in logistic, ordinal logistic, and probit regression when an outcome variable is coarsely categorized. Educ. Psychol. Meas. 66, 228–239. Ujike, H., Nomura, A., Morita, Y., Morio, A., Okahisa, Y., Kotaka, T., et al., 2008. Multiple genetic factors in olanzapine-induced weight gain in schizophrenia patients: a cohort study. J. Clin. Psychiatry 69, 1416–1422. Weinstein, M.C., Fineberg, H.V., 1980. Clinical Decision Analysis. Saunders, Philadelphia; London.
Acknowledgments This study was funded by the Kaohsiung Municipal Kai-Syuan Psychiatric Hospital (KSPH-2011-22) and the Ministry of Science and Technology, Taiwan (MOST-103–2314-B-280-001-MY3). The authors report no conflicts of interest. References American Diabetes Association, 2004. Consensus development conference on antipsychotic drugs and obesity and diabetes. Diabetes Care 27, 596–601. Albaugh, V.L., Henry, C.R., Bello, N.T., Hajnal, A., Lynch, S.L., Halle, B., et al., 2006. Hormonal and metabolic effects of olanzapine and clozapine related to body weight in rodents. Obesity 14, 36–51. Allison, D.B., Fontaine, K.R., Heo, M., Mentore, J.L., Cappelleri, J.C., Chandler, L.P., et al., 1999a. The distribution of body mass index among individuals with and without schizophrenia. J. Clin. Psychiatry 60, 215–220. Allison, D.B., Mentore, J.L., Heo, M., Chandler, L.P., Cappelleri, J.C., Infante, M.C., et al., 1999b. Antipsychotic-induced weight gain: a comprehensive research synthesis. Am. J. Psychiatry 156, 1686–1696. Álvarez-Jiménez, M., González-Blanch, C., Crespo-Facorro, B., Hetrick, S., RodríguezSánchez, J.M., Pérez-Iglesias, R., et al., 2008. Antipsychotic-induced weight gain in chronic and first-episode psychotic disorders. CNS Drugs 22, 547–562. APA, 2000. Diagnostic and Statistical Manual of Mental Disorders: DSM-IV-TR, 4th ed., text revision. American Psychiatric Association, Washington, DC. Ascher-Svanum, H., Nyhuis, A.W., Faries, D.E., Kinon, B.J., Baker, R.W., Shekhar, A., 2008. Clinical, functional, and economic ramifications of early nonresponse to antipsychotics in the naturalistic treatment of schizophrenia. Schizophr. Bull. 34, 1163–1171. Bak, M., Fransen, A., Janssen, J., van Os, J., Drukker, M., 2014. Almost all antipsychotics result in weight gain: a meta-analysis. PLoS One 9, e94112. Basson, B.R., Kinon, B.J., Taylor, C.C., Szymanski, K.A., Gilmore, J.A., Tollefson, G.D., 2001. Factors influencing acute weight change in patients with schizophrenia treated with olanzapine, haloperidol, or risperidone. J. Clin. Psychiatry 62, 231–238. Bellanger, T.M., Bray, G.A., 2005. Obesity related morbidity and mortality. J. La State Med. Soc. 157 (Spec 1), 42–49. Chiu, C.C., Chen, C.H., Chen, B.Y., Yu, S.H., Lu, M.L., 2010. The time-dependent change of insulin secretion in schizophrenic patients treated with olanzapine. Prog. Neuropsychopharmacol. Biol. Psychiatry 34, 866–870. Citrome, L., Holt, R.I., Walker, D.J., Hoffmann, V.P., 2011. Weight gain and changes in metabolic variables following olanzapine treatment in schizophrenia and bipolar disorder. Clin. Drug Investig. 31, 455–482. Daumit, G.L., Dickerson, F.B., Wang, N.Y., Dalcin, A., Jerome, G.J., Anderson, C.A., et al., 2013. A behavioral weight-loss intervention in persons with serious mental illness. N. Engl. J. Med. 368, 1594–1602. De Hert, M., Dekker, J.M., Wood, D., Kahl, K.G., Holt, R.I., Moller, H.J., 2009. Cardiovascular disease and diabetes in people with severe mental illness position statement from the European Psychiatric Association (EPA), supported by the European Association for the Study of Diabetes (EASD) and the European Society of Cardiology (ESC). Eur. Psychiatry 24. pp. 412–424. Dickerson, F.B., Brown, C.H., Kreyenbuhl, J.A., Fang, L., Goldberg, R.W., Wohlheiter, K., et al., 2006. Obesity among individuals with serious mental illness. Acta Psychiatr. Scand. 113, 306–313. Duggan, L., Fenton, M., Rathbone, J., Dardennes, R., El-Dosoky, A., Indran, S., 2005. Olanzapine for schizophrenia. Cochrane Database Syst. Rev (CD001359). Ganguli, R., 2007. Behavioral therapy for weight loss in patients with schizophrenia. J. Clin. Psychiatry 68 (Suppl. 4), S19–S25. Guy, W., 1976. ECDEU Assessment Manual for Psychopharmacology. DHEW Publication 76338, Washington DC. Huxley, R., Mendis, S., Zheleznyakov, E., Reddy, S., Chan, J., 2010. Body mass index, waist circumference and waist: hip ratio as predictors of cardiovascular risk–a review of the literature. Eur. J. Clin. Nutr. 64, 16–22. Kane, J., Honigfeld, G., Singer, J., Meltzer, H., 1988. Clozapine for the treatment-resistant
211
Contents lists available at ScienceDirect
Psychiatry Research journal homepage: www.elsevier.com/locate/psychres
Early prediction of olanzapine-induced weight gain for schizophrenia patients Ching-Hua Lina,b, Shih-Chi Lina, Yu-Hui Huanga, Fu-Chiang Wanga, Chun-Jen Huangb,c,
T
⁎
a
Kaohsiung Municipal Kai-Syuan Psychiatric Hospital, Kaohsiung, Taiwan Department of Psychiatry, School of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan c Department of Psychiatry, Kaohsiung Medical University Hospital, Taiwan b
A R T I C LE I N FO
A B S T R A C T
Keywords: Schizophrenia Olanzapine Weight gainers Early prediction Receiver operating characteristic curve
The aim of this study was to determine whether weight changes at week 2 or other factors predicted weight gain at week 6 for schizophrenia patients receiving olanzapine. This study was the secondary analysis of a six-week trial for 94 patients receiving olanzapine (5 mg/d) plus trifluoperazine (5 mg/d), or olanzapine (10 mg/d) alone. Patients were included in analysis only if they had completed the 6-week trial (per protocol analysis). Weight gain was defined as a 7% or greater increase of the patient's baseline weight. The receiver operating characteristic curve was employed to determine the optimal cutoff points of statistically significant predictors. Eleven of the 67 patients completing the 6-week trial were classified as weight gainers. Weight change at week 2 was the statistically significant predictor for ultimate weight gain at week 6. A weight change of 1.0 kg at week 2 appeared to be the optimal cutoff point, with a sensitivity of 0.92, a specificity of 0.75, and an AUC of 0.85. Using weight change at week 2 to predict weight gain at week 6 is favorable in terms of both specificity and sensitivity. Weight change of 1.0 kg or more at 2 weeks is a reliable predictor.
1. Introduction Obesity is prevalent among schizophrenia patients (Dickerson et al., 2006), and many antipsychotics used to treat schizophrenia may also result in weight gain (Allison et al., 1999b). For example, weight gain associated with olanzapine is commonly reported (Allison et al., 1999b; Duggan et al., 2005). Weight gain can lead to various medical complications such as diabetes mellitus, dyslipidemia, cardiovascular disease, and osteoarthritis (Bellanger and Bray, 2005; Huxley et al., 2010; Kawachi, 1999; Newcomer and Haupt, 2006). In addition, weight gain induced by antipsychotics may contribute to low self-esteem, medication nonadherence and subsequent relapse (Allison et al., 1999a). Psychiatrists need to balance the risk of weight gain against the benefit of symptomatic control. Therefore, reliable information about the risk for weight gain is required for patients receiving antipsychotics. The mechanism of weight gain during treatment with olanzapine is not fully understood, and is likely to be multifactorial (Albaugh et al., 2006). In studies on the prediction of weight gain for patients taking antipsychotics, risk factors already identified include male gender, nonwhite ethnicity, first-episode psychosis, increased appetite, better clinical response, lower baseline body mass index (BMI), younger age, and gene polymorphism (Basson et al., 2001; Kinon et al., 2005; Ou
⁎
et al., 2013; Ujike et al., 2008). Increased appetite is most likely a result of the antipsychotic medication, probably because antipsychotic medication is an antagonist of serotonergic 5-HT2c receptors, histaminergic H1-receptors, and dopamine receptors (Bak et al., 2014; Starrenburg and Bogers, 2009). However, to date, few of these risk factors have led to the development of a clinically useful decision-making tool. Three industry-sponsored studies have reported that early weight increase during the course of olanzapine treatment may predict substantial weight gain following long-term use (Kinon et al., 2005; Lipkovich et al., 2006, 2008). A recent meta-analysis study (Musil et al., 2015) has concluded that weight increase is common during the first weeks of treatment. These findings indicate that early prediction of weight gain and, consequently, the timely implementation of weight management strategies is of great benefit. Consensus guidelines emphasize the importance of appropriate baseline screening and measuring of body weight, beginning in the initial weeks of treatment for patients taking olanzapine (American Diabetes Association, 2004; De Hert et al., 2009). We recently conducted a six-week, randomized, double-blind study to compare the efficacy and safety of low-dose olanzapine (5 mg/d) plus low-dose trifluoperazine (5 mg/d) versus full-dose olanzapine (10 mg) in the acute treatment of schizophrenia (Lin et al., 2017). One of the
Corresponding author at: Department of Psychiatry, Kaohsiung Medical University Hospital, 100, Shih-Chuan 1st Road, San Ming District, Kaohsiung 807, Taiwan. E-mail address: [email protected] (C.-J. Huang).
https://doi.org/10.1016/j.psychres.2018.02.058 Received 19 July 2017; Received in revised form 24 December 2017; Accepted 28 February 2018 0165-1781/ © 2018 Elsevier B.V. All rights reserved.
Psychiatry Research 263 (2018) 207–211
C.-H. Lin et al.
findings was that no significantly different changes in body weight occurred at weeks 1, 2, 3, 4, and 6 between the two groups after treatment. The explanation for this is that weight change due to olanzapine is not dose-related within a range of 5–20 mg/day (Kinon et al., 2001), whereas weight increase due to trifluoperazine is unusual (Stahl, 2014). Therefore, possible weight change induced by trifluoperazine may be minimal if patients receive 5 mg daily of trifluoperazine plus 5 mg daily of olanzapine for short-term treatment. The aim of the study was to determine whether weight change, percentage of weight change, and/or BMI change at weeks 1 or 2 predicted weight gain at week 6 for schizophrenia patients receiving olanzapine treatment, and therefore to determine which predictor was most powerful and user-friendly.
7% or greater increase of the baseline weight (kg) to endpoint. Weight gainers and non-weight gainers were compared in terms of sex, age, age at onset, baseline PANSS, baseline weight (kg), baseline BMI (kg/m2), weight change at week 1 (kg), percentage of weight change at week 1, BMI change at week 1, weight change at week 2 (kg), percentage of weight change at week 2, and BMI change at week 2. Age at onset was regarded as the age at which the first psychotic symptoms occurred. Each statistically significant predictor at weeks 1 and 2 from the univariate analysis was separately entered into a receiver operating characteristic (ROC) analysis for determining which one was a powerful and user-friendly predictor. The ROC curve was used to determine the optimal cutoff point of predictor between the weight gainers and nonweight gainers, maximizing both the sensitivity and specificity of the predictor(s) so that false positive and false negative rates could be minimized. The area under the ROC curve (AUC) is a measure of the overall discriminative power. In practice, an AUC ≥ 0.8 indicates a good discriminative capacity (Weinstein and Fineberg, 1980). All data were processed by SPSS version 17.0 for Windows (SPSS Inc., Chicago, IL, USA). Pearson χ2 test was used to compare categorical variables; independent t-test for continuous variables. All tests were two-tailed, and significance was defined as an alpha of less than 0.05.
2. Methods 2.1. Ethics The study was the secondary analysis of a clinical trial, conducted from January 2012 to February 2016, as documented elsewhere (Lin et al., 2017). This study was approved by the Kai-Syuan Psychiatric Hospital's institutional review board (IRB) and conducted in accordance with Good Clinical Practice procedures, the Declaration of Helsinki, and national legal requirements (Taiwan's Human Subjects Research Act). This study was registered on Clinicaltrials.gov (Identifier number: NCT02704962).
3. Results 3.1. Patient characteristics
2.2. Patients, study design, procedures, and body weight assessment
A total of 94 newly hospitalized schizophrenia patients with acute exacerbation were randomly assigned for treatment with olanzapine plus trifluoperazine (N = 47) or olanzapine alone (N = 47). No antipsychotic-naive or first episode schizophrenia patients were enrolled. Twenty-seven of the 94 patients dropped out early from the study. One patient withdrew from the trial after 2 weeks of treatment due to a weight increase of 5.5 kg. Sixty-seven patients who completed the 6week trial were enrolled in the analysis. The mean age of the patients was 39.4 (SD = 8.5) years, and 47.8% (N = 32) were males. The mean PANSS score of 95.0 (SD = 16.5) at baseline reflected a fairly severely symptomatic population (Table 1). Eleven of the 67 patients (16.4%) were classified as weight gainers. There were no between-group differences in terms of sex, age, age at onset, baseline PANSS, baseline weight, and baseline BMI (Table 1). Table 2 lists the mean body weight, mean BMI, mean weight change, mean percentage of weight change, and mean BMI change of the two groups at each assessment.
As previously described in detail (Lin et al., 2017), all newly hospitalized schizophrenia patients for acute treatment were evaluated by trained and board-certified psychiatrists. Inpatients were enrolled into this study if they: 1) were physically healthy and had all laboratory parameters within normal limits, 2) were aged between 18 and 55 years, 3) satisfied DSM-IV (APA, 2000) criteria for schizophrenia, 4) had a baseline Clinical Global Impression-Severity of Illness Scale (CGIS) (Guy, 1976) ≥4, 5) had no DSM-IV diagnosis of substance abuse or dependence (including alcohol) in the past 6 months, 6) had not received depot antipsychotics during the preceding three months, and 7) had not taken olanzapine for at least 4 weeks before participating in the trial. Patients excluded from this study were: 1) those with histories of serious adverse reactions to olanzapine or trifluoperazine, 2) those with histories of tardive dyskinesia or neuroleptic malignant syndrome, 3) female subjects who were pregnant or at risk of pregnancy or lactation, and 4) those diagnosed with treatment-resistant schizophrenia (Kane et al., 1988), or having previously received clozapine or electroconvulsive therapy. Written informed consent was obtained from every patient and his or her legal guardian. After a washout period of at least 3 days, patients were randomly assigned to a 1:1 ratio of 5 mg/day of olanzapine plus 5 mg/day of trifluoperazine, or 10 mg/day of olanzapine alone for 6 weeks. Patient adherence and safety were closely monitored by the research staff. Benzodiazepine was allowed as needed for insomnia or agitation. Biperiden, up to 6 mg/day, was also allowed to treat motor side effects, but prophylactic use of biperiden was prohibited. No other psychotropic agents were permitted during the study. Symptom severity was rated using the Positive and Negative Syndrome Scale (PANSS) (Kay et al., 1987). At weeks 0, 1, 2, 3, 4, and 6 (or upon early termination), body weight (kg) was measured in the morning after emptying the bladder and before breakfast, using a highquality, calibrated digital scale, with the patient wearing indoor clothing but no shoes.
3.2. Early prediction of weight gain Weight change at week 1, percentage of weight change at week 1, and BMI change at week 1 were significant predictors of weight gainers at week 1. Similarly, weight change at week 2, percentage of weight change at week 2, and BMI change at week 2 were also significant predictors of weight gainers at week 2 (Table 1). ROC analysis was employed to determine the optimal cutoff point and AUC of each significant predictor. Table 3 shows the cutoff point, sensitivity, specificity, predictive power, and AUC of each predictor at weeks 1 and 2. At week 1, all the AUCs provided by weight change (0.75), percentage of weight change (0.75), and BMI change (0.75) were less than 0.8. At week 2, a weight change of 1.0 kg appeared to be the optimal cutoff point for predicting weight gainers, with a sensitivity of 0.92, a specificity of 0.75, a predictive power of 0.88, and an AUC of 0.85. A weight change of 1.23% appeared to be the optimal cutoff point for predicting weight gain, with a sensitivity of 0.91, a specificity of 0.70, a predictive power of 0.81, and an AUC of 0.84. A BMI change of 0.36 appeared to be the optimal cutoff point for predicting weight gain, with a sensitivity of 0.91, a specificity of 0.75, a predictive power of 0.83, and an AUC of 0.85. ROC curves of weight change at weeks 1 and 2 are illustrated in Fig. 1.
2.3. Statistical analyses Patients were included in the analysis only if they had completed the 6-week trial (per protocol analysis). Weight gain was defined as a 208
Psychiatry Research 263 (2018) 207–211
C.-H. Lin et al.
Table 1 Comparison of clinical characteristics between weight gainersa and non-weight gainers. Variable
Total (N = 67)
Weight gainersa (n = 11)
Non-weight gainers (n = 56)
Sex Male, n (%) Female, n (%) Age, mean (SD), years Age at onset, mean (SD), years Baseline PANSSb, mean (SD) Baseline weight, mean (SD), kg Baseline BMIc, mean (SD), kg/m2 Weight change at week 1, mean (SD), kg Percentage of weight change at week 1, mean (SD), kg BMI change at week 1 Weight change at week 2, mean (SD), kg Percentage of weight change at week 2, mean (SD), kg BMI change at week 2
32 (47.8) 35 (52.2) 39.4 (8.5) 25.4 (8.1) 95.0 (16.5) 63.6 (13.7) 24.0 (4.5) 0.2 (1.1) 0.5 (1.9) 0.1 (0.4) 0.8 (1.8) 1.4 (3.0) 0.3 (0.7)
5 (45.5) 6 (54.5) 37.2 (8.6) 21.7 (5.7) 92.4 (14.2) 58.7 (9.2) 22.4 (3.6) 1.1 (1.5) 2.0 (2.6) 0.4 (0.5) 2.7 (1.8) 5.0 (3.7) 0.0 (0.4)
27 (48.2) 29 (51.8) 39.9 (8.5) 26.1 (8.4) 95.5 (16.9) 64.6 (14.3) 24.3 (4.6) 0.1 (1.0) 0.2 (1.5) 1.0 (0.7) 0.4 (1.5) 0.7 (2.2) 0.1 (0.6)
p 0.884d
0.336e 0.100e 0.570e 0.198e 0.187e 0.037e 0.049e 0.032e < 0.000 0.003e < 0.000
Bold, statistically significant. a weight gainers = patients who experienced at least 7% weight increase at end point. b PANSS = Positive and Negative Syndrome Scale. c BMI = body mass index. d Pearson's χ2 test. e independent t-test.
4. Discussion
the other hand, 75% of patients (i.e., specificity) with a weight change < 1 kg at week 2 were correctly identified as non-weight gainers. Compared to a study conducted in the U.S. by Simpson et al. (2004) for schizophrenia inpatients (N = 133) treated with 5–15 mg daily of olanzapine for 6 weeks, weight gain was defined as a 7% increase in weight from baseline. About 36% of patients were classified as weight gainers (Parsons et al., 2009). There were only 4 Asian patients (3.0%) enrolled in this study. However, in a Taiwanese study (Chiu et al., 2010), schizophrenia inpatients received 5–20 mg daily of olanzapine for 8 weeks. Six of 33 completers (18.2%) met the weight gain criterion (i.e., ≥7% weight gain). This result is comparable to our findings, and indicates that ethnic factors should be considered (Basson et al., 2001; Citrome et al., 2011). Study by Basson et al. (2001) found that males gained significantly more weight than females if patients with schizophrenia received olanzapine. In the present study, males and females had comparable rates of weight gain after the 6-week treatment (Table 1). To avoid loss of power from dichotomizing a continuous variable (i.e., weight gainers vs. non-weight gainers) (Streiner, 2002; Taylor et al., 2006), weight change at week 6 (i.e., a continuous variable) was regarded as a dependent variable. Analysis of sex difference was performed by analysis of covariance, with sex as a fixed factor and baseline weight as covariate. There was no significant between-sex difference (p = 0.845) in body weight change at week 6 (data not shown in the table). A per protocol analysis for a trial may be biased because
The first main finding of this study was that as early as week 1, weight gainers had significantly greater weight change, percentage of weight change, and BMI change than non-weight gainers (see Table 1). However, these three predictors at week 1 did not provide adequate discriminative capacity (i.e., AUC < 0.8) (see Table 3). The second main finding was that weight change at week 2, percentage of weight change at week 2, or BMI change at week 2 provided comparable AUCs (i.e., 0.84 vs. 0.85 vs. 0.84). The explanation of such comparable AUCs among three predictors was that weight change at week 2, percentage of weight change at week 2, and BMI change at week 2 were highly correlated each other (r2 = 0.94, p < 0.001 for weight change at week 2 vs. percentage of weight change at week 2; r2 = 0.99, p < 0.001 for weight change at week 2 vs. BMI change at week 2; r2 = 0.95, p < 0.001 for percentage of weight change at week 2 vs. BMI change at week 2). All AUCs provided by the 3 predictors at week 2 were greater than 0.80. Therefore, weight change at week 2, percentage of weight change at week 2, or BMI change at week 2 had good discriminative capacity for predicting an ultimate 7% or more weight increase (Table 3). Compared with using a percentage of weight change of 1.23% at week 2 or a BMI change of 0.36 at week 2 for predicting ultimate weight gainers, a weight change of 1.0 kg at week 2 was more user-friendly in clinical practice. A weight change ≥ 1.0 kg at week 2 of initiating olanzapine treatment for patients with schizophrenia correctly identified 92% of weight gainers at endpoint (i.e., sensitivity). On
Table 2 Mean body weight, mean BMI, mean weight change, mean percentage of weight change, and mean BMI change of the two groups at each assessment (N = 67).
Weight gainers Body weight, mean (SD), kg BMI, mean (SD), kg/m2 Weight change Percentage of weight change BMI change Non-weight gainers Body weight, mean (SD), kg BMI, mean (SD), kg/m2 Weight change Percentage of weight change BMI change
Week 0 Mean (SD)
Week 1 Mean (SD)
Week 2 Mean (SD)
Week 3 Mean (SD)
Week 4 Mean (SD)
Week 6 Mean (SD)
58.7 (9.2) 22.4 (3.6)
59.9 (9.5) 22.8 (3.7) 1.1 (1.5) 2.0 (2.6) 0.4 (0.5)
61.5 (8.4) 23.4 (3.3) 2.7 (1.8) 5.0 (3.7) 1.0 (0.7)
62.0 (8.6) 23.6 (3.4) 3.3 (2.0) 5.9 (4.1) 1.2 (0.7)
63.0 (8.4) 24.0 (3.2) 4.2 (2.7) 7.7 (5.6) 1.6 (1)
64.3 (9.7) 24.5 (3.6) 5.5 (1.8) 9.6 (3.7) 2.1 (0.6)
64.6 (14.3) 24.3 (4.6)
64.7 (14.0) 24.4 (4.5) 0.1 (1.0) 0.2 (1.5) 0.02 (0.4)
65.0 (14.2) 24.5 (4.5) 0.4 (1.5) 0.7 (2.2) 0.1 (0.6)
65.1 (14.4) 24.5 (4.6) 0.5 (1.6) 0.9 (2.5) 0.2 (0.6)
65.3 (14.2) 24.6 (4.6) 0.7 (1.8) 1.2 (2.5) 0.3 (0.7)
65.6 (13.9) 24.7 (4.5) 1.0 (2.0) 1.7 (2.9) 0.4 (0.7)
209
Psychiatry Research 263 (2018) 207–211
C.-H. Lin et al.
Table 3 Cutoff point, sensitivity, specificity, predictive power, and AUC of each predictor at weeks 1 and 2: ROC Predictors Week 1 Weight change Percentage of weight change BMI change Week 2 Weight change Percentage of weight change BMI change b
a
analysis.
Cutoff point
Sensitivity (True positive)
Specificity (True negative)
Predictive power
AUCa (95% CI)
0.50 0.93 0.15
0.64 0.73 0.72
0.80 0.79 0.71
0.75 0.76 0.72
0.75 (0.63–0.85) 0.75 (0.63–0.85) 0.75 (0.63–0.84)
1.00 1.23 0.36
0.92 0.91 0.91
0.75 0.70 0.75
0.88 0.81 0.83
0.85 (0.74–0.93) 0.84 (0.74–0.92) 0.85 (0.75–0.93)
AUC = area under the ROC curve. a ROC = receiver operating characteristic.
after one year of treatment. The second study, by Lipkovich et al. (2006), reported that a weight increase of 2 − 3 kg during the first 3 weeks of olanzapine treatment for patients with bipolar disorder predicted substantial weight gain by 30 weeks. Weight change at 3 weeks plus baseline characteristics for predicting weight gain provided an AUC of 0.80. The third study (Lipkovich et al., 2008), indicates that weight change from baseline to weeks 1, 2, 3, or 4 alone could be used to predict substantial weight gain defined as at least a 7% weight increase. However, these studies (Kinon et al., 2005; Lipkovich et al., 2008) did not provide the AUC values for potential predictors. Two of the three studies (Kinon et al., 2005; Lipkovich et al., 2006) did not reveal the potential predictors at 2 weeks. Several strengths of this study should be addressed. First, our study provided data on Asian patients for comparison. Second, the subjects were inpatients for acute treatment. In inpatient settings, patients are carefully monitored, and confounding factors (Strassnig et al., 2003) related to weight gain, such as low levels of physical activity and inappropriate food intake, could be minimized. Third, some proposed cutoff points for predicting substantial weight gain may be chosen arbitrarily. Sensitivity and specificity analyses were performed in the present study to determine the optimal cut-off for predicting weight gain. The optimal cut-off point, a weight change of 1.0 kg at week 2, to assess the individual's risk of experiencing substantial weight gain at week 6, is simple for clinicians to remember and apply. In addition, because this was a racially homogeneous group for study, the confounding variable of ethnicity could be decreased. There were also some limitations to this study. First, this was a trial with small sample sizes, a 6-week design, and was conducted in only one psychiatric center. The results may not therefore be applicable to weight change which occurs during long-term treatment. However, the initial weight gain within the first 6 weeks of pharmacotherapy is most important, as patients will not lose weight thereafter (Bak et al., 2014). Second, although there were no significant differences in weight gain between the two groups (i.e., 5 mg/d of olanzapine plus 5 mg/d of trifluoperazine versus 10 mg/d of olanzapine) in our recent study (Lin et al., 2017), a meta-analysis study indicates that some patients taking trifluoperazine may gain weight (Koch et al., 2014). The possibility that receiving one of these treatments predicts weight gain at follow-up should therefore be explored. Third, generalization may not extend to those patients aged more than 55 years old, to patients treated with other atypical antipsychotics, to antipsychotic-naive patients (Bak et al., 2014; Tarricone et al., 2010), to first episode schizophrenia patients (Álvarez-Jiménez et al., 2008), or to patients with other mental disorders (Allison et al., 1999b; Basson et al., 2001; Moteshafi et al., 2012). Fourth, the original study was not designed for estimating the predictive value of a weight gain in the population. Finally, use of an all-Taiwanese sample may limit generalizability to other countries. In conclusion, a weight change of 1 kg or more at week 2 could be used to predict a 7% or more weight increase for schizophrenia patients after 6 weeks of treatment with olanzapine. This finding provided a sensitivity of 0.92, a specificity of 0.75, and an AUC of 0.85. Further
Fig. 1. The ROC curves for weight gainers versus non-weight gainers at weeks 1 and 2.
participants who fail to complete the trial will be excluded (Sedgwick, 2015). If analysis was on an intent-to-treat basis, missing data would be replaced according to the principle of last observation carried forward. Twelve of the 94 patients (12.8%) were classified as weight gainers. There were no between-group differences in terms of sex, age, age at onset, baseline PANSS, baseline weight, and baseline BMI. Weight change at week 2 was still a significant predictor of weight gainers, and was therefore entered into a ROC analysis. A weight change of 1.0 kg at week 2 still appeared to be the optimal cutoff point, with a sensitivity of 0.92, a specificity of 0.74, and an AUC of 0.88 (data not shown in the table). Symptomatic change at 2 weeks has also been reported to be a powerful predictor of clinical response or remission in antipsychotic treatment (Ascher-Svanum et al., 2008; Samara et al., 2015). Our major finding, that weight change of 1.0 kg or more at 2 weeks predicts substantial weight gain at week 6, can provide reliable information for treating psychiatrists to make risk/benefit-based decisions about continuing or changing treatment. If the treating psychiatrist decides to continue olanzapine treatment for an individual patient at risk of weight gain, remedies for preventing or reducing further weight increase should be applied early in the course of treatment (Lipkovich et al., 2008; Musil et al., 2015). These remedies should include dietary advice, exercise, behavioral interventions, or pharmacological treatment (Daumit et al., 2013; Ganguli, 2007; Mizuno et al., 2014). The three studies (Kinon et al., 2005; Lipkovich et al., 2006, 2008) mentioned above found that early increase in weight is a clinically useful predictor for gaining weight after long-term treatment with olanzapine. The first study (Kinon et al., 2005), revealed that schizophrenic patients with at least a 7% weight increase during the first 6 weeks of treatment were at greatest risk for gaining the most weight 210
Psychiatry Research 263 (2018) 207–211
C.-H. Lin et al.
studies, preferably involving larger inpatient or outpatient groups from multi-centers, and treatment durations of longer than 6 weeks, are needed to better confirm the predictive power of weight change at week 2 in experiencing substantial weight gains after olanzapine treatment.
schizophrenic. A double-blind comparison with chlorpromazine. Arch. Gen. Psychiatry 45, 789–796. Kawachi, I., 1999. Physical and psychological consequences of weight gain. J. Clin. Psychiatry 60 (Suppl. 21), S5–S9. Kay, S.R., Fiszbein, A., Opler, L.A., 1987. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr. Bull. 13, 261–276. Kinon, B.J., Basson, B.R., Gilmore, J.A., Tollefson, G.D., 2001. Long-term olanzapine treatment: weight change and weight-related health factors in schizophrenia. J. Clin. Psychiatry 62, 92–100. Kinon, B.J., Kaiser, C.J., Ahmed, S., Rotelli, M.D., Kollack-Walker, S., 2005. Association between early and rapid weight gain and change in weight over one year of olanzapine therapy in patients with schizophrenia and related disorders. J. Clin. Psychopharmacol. 25, 255–258. Koch, K., Mansi, K., Haynes, E., Adams, C.E., Sampson, S., Furtado, V.A., 2014. Trifluoperazine versus placebo for schizophrenia. Cochrane Database Syst. Rev (CD010226). Lin, C.H., Wang, F.C., Lin, S.C., Huang, Y.H., Chen, C.C., 2017. A randomized, doubleblind, comparison of the efficacy and safety of low-dose olanzapine plus low-dose trifluoperazine versus full-dose olanzapine in the acute treatment of schizophrenia. Schizophr. Res. 185, 80–87. Lipkovich, I., Citrome, L., Perlis, R., Deberdt, W., Houston, J.P., Ahl, J., et al., 2006. Early predictors of substantial weight gain in bipolar patients treated with olanzapine. J. Clin. Psychopharmacol. 26, 316–320. Lipkovich, I., Jacobson, J.G., Hardy, T.A., Hoffmann, V.P., 2008. Early evaluation of patient risk for substantial weight gain during olanzapine treatment for schizophrenia, schizophreniform, or schizoaffective disorder. BMC Psychiatry 8, 78. Mizuno, Y., Suzuki, T., Nakagawa, A., Yoshida, K., Mimura, M., Fleischhacker, W.W., et al., 2014. Pharmacological strategies to counteract antipsychotic-induced weight gain and metabolic adverse effects in schizophrenia: a systematic review and metaanalysis. Schizophr. Bull. 40, 1385–1403. Moteshafi, H., Zhornitsky, S., Brunelle, S., Stip, E., 2012. Comparing tolerability of olanzapine in schizophrenia and affective disorders: a meta-analysis. Drug Saf. 35, 819–836. Musil, R., Obermeier, M., Russ, P., Hamerle, M., 2015. Weight gain and antipsychotics: a drug safety review. Expert Opin. Drug Saf. 14, 73–96. Newcomer, J.W., Haupt, D.W., 2006. The metabolic effects of antipsychotic medications. Can. J. Psychiatry 51, 480–491. Ou, J.J., Xu, Y., Chen, H.H., Fan, X., Gao, K., Wang, J., et al., 2013. Comparison of metabolic effects of ziprasidone versus olanzapine treatment in patients with first-episode schizophrenia. Psychopharmacology 225, 627–635. Parsons, B., Allison, D.B., Loebel, A., Williams, K., Giller, E., Romano, S., et al., 2009. Weight effects associated with antipsychotics: a comprehensive database analysis. Schizophr. Res. 110, 103–110. Samara, M.T., Leucht, C., Leeflang, M.M., Anghelescu, I.G., Chung, Y.C., Crespo-Facorro, B., et al., 2015. Early improvement as a predictor of later response to antipsychotics in schizophrenia: a diagnostic test review. Am. J. Psychiatry 172, 617–629. Sedgwick, P., 2015. Intention to treat analysis versus per protocol analysis of trial data. BMJ 350, 681. Simpson, G.M., Glick, I.D., Weiden, P.J., Romano, S.J., Siu, C.O., 2004. Randomized, controlled, double-blind multicenter comparison of the efficacy and tolerability of ziprasidone and olanzapine in acutely ill inpatients with schizophrenia or schizoaffective disorder. Am. J. Psychiatry 161, 1837–1847. Stahl, S.M., 2014. Stahl’s Essential Psychopharmacology: Prescriber’s Guide, Fifth edition. ed. Starrenburg, F., Bogers, J., 2009. How can antipsychotics cause diabetes mellitus? Insights based on receptor-binding profiles, humoral factors and transporter proteins. Eur. Psychiatry 24, 164–170. Strassnig, M., Brar, J.S., Ganguli, R., 2003. Nutritional assessment of patients with schizophrenia: a preliminary study. Schizophr. Bull. 29, 393–397. Streiner, D.L., 2002. Breaking up is hard to do: the heartbreak of dichotomizing continuous data. Can. J. Psychiatry 47, 262–266. Tarricone, I., Ferrari Gozzi, B., Serretti, A., Grieco, D., Berardi, D., 2010. Weight gain in antipsychotic-naive patients: a review and meta-analysis. Psychol. Med. 40, 187–200. Taylor, A.B., West, S.G., Aiken, L.S., 2006. Loss of power in logistic, ordinal logistic, and probit regression when an outcome variable is coarsely categorized. Educ. Psychol. Meas. 66, 228–239. Ujike, H., Nomura, A., Morita, Y., Morio, A., Okahisa, Y., Kotaka, T., et al., 2008. Multiple genetic factors in olanzapine-induced weight gain in schizophrenia patients: a cohort study. J. Clin. Psychiatry 69, 1416–1422. Weinstein, M.C., Fineberg, H.V., 1980. Clinical Decision Analysis. Saunders, Philadelphia; London.
Acknowledgments This study was funded by the Kaohsiung Municipal Kai-Syuan Psychiatric Hospital (KSPH-2011-22) and the Ministry of Science and Technology, Taiwan (MOST-103–2314-B-280-001-MY3). The authors report no conflicts of interest. References American Diabetes Association, 2004. Consensus development conference on antipsychotic drugs and obesity and diabetes. Diabetes Care 27, 596–601. Albaugh, V.L., Henry, C.R., Bello, N.T., Hajnal, A., Lynch, S.L., Halle, B., et al., 2006. Hormonal and metabolic effects of olanzapine and clozapine related to body weight in rodents. Obesity 14, 36–51. Allison, D.B., Fontaine, K.R., Heo, M., Mentore, J.L., Cappelleri, J.C., Chandler, L.P., et al., 1999a. The distribution of body mass index among individuals with and without schizophrenia. J. Clin. Psychiatry 60, 215–220. Allison, D.B., Mentore, J.L., Heo, M., Chandler, L.P., Cappelleri, J.C., Infante, M.C., et al., 1999b. Antipsychotic-induced weight gain: a comprehensive research synthesis. Am. J. Psychiatry 156, 1686–1696. Álvarez-Jiménez, M., González-Blanch, C., Crespo-Facorro, B., Hetrick, S., RodríguezSánchez, J.M., Pérez-Iglesias, R., et al., 2008. Antipsychotic-induced weight gain in chronic and first-episode psychotic disorders. CNS Drugs 22, 547–562. APA, 2000. Diagnostic and Statistical Manual of Mental Disorders: DSM-IV-TR, 4th ed., text revision. American Psychiatric Association, Washington, DC. Ascher-Svanum, H., Nyhuis, A.W., Faries, D.E., Kinon, B.J., Baker, R.W., Shekhar, A., 2008. Clinical, functional, and economic ramifications of early nonresponse to antipsychotics in the naturalistic treatment of schizophrenia. Schizophr. Bull. 34, 1163–1171. Bak, M., Fransen, A., Janssen, J., van Os, J., Drukker, M., 2014. Almost all antipsychotics result in weight gain: a meta-analysis. PLoS One 9, e94112. Basson, B.R., Kinon, B.J., Taylor, C.C., Szymanski, K.A., Gilmore, J.A., Tollefson, G.D., 2001. Factors influencing acute weight change in patients with schizophrenia treated with olanzapine, haloperidol, or risperidone. J. Clin. Psychiatry 62, 231–238. Bellanger, T.M., Bray, G.A., 2005. Obesity related morbidity and mortality. J. La State Med. Soc. 157 (Spec 1), 42–49. Chiu, C.C., Chen, C.H., Chen, B.Y., Yu, S.H., Lu, M.L., 2010. The time-dependent change of insulin secretion in schizophrenic patients treated with olanzapine. Prog. Neuropsychopharmacol. Biol. Psychiatry 34, 866–870. Citrome, L., Holt, R.I., Walker, D.J., Hoffmann, V.P., 2011. Weight gain and changes in metabolic variables following olanzapine treatment in schizophrenia and bipolar disorder. Clin. Drug Investig. 31, 455–482. Daumit, G.L., Dickerson, F.B., Wang, N.Y., Dalcin, A., Jerome, G.J., Anderson, C.A., et al., 2013. A behavioral weight-loss intervention in persons with serious mental illness. N. Engl. J. Med. 368, 1594–1602. De Hert, M., Dekker, J.M., Wood, D., Kahl, K.G., Holt, R.I., Moller, H.J., 2009. Cardiovascular disease and diabetes in people with severe mental illness position statement from the European Psychiatric Association (EPA), supported by the European Association for the Study of Diabetes (EASD) and the European Society of Cardiology (ESC). Eur. Psychiatry 24. pp. 412–424. Dickerson, F.B., Brown, C.H., Kreyenbuhl, J.A., Fang, L., Goldberg, R.W., Wohlheiter, K., et al., 2006. Obesity among individuals with serious mental illness. Acta Psychiatr. Scand. 113, 306–313. Duggan, L., Fenton, M., Rathbone, J., Dardennes, R., El-Dosoky, A., Indran, S., 2005. Olanzapine for schizophrenia. Cochrane Database Syst. Rev (CD001359). Ganguli, R., 2007. Behavioral therapy for weight loss in patients with schizophrenia. J. Clin. Psychiatry 68 (Suppl. 4), S19–S25. Guy, W., 1976. ECDEU Assessment Manual for Psychopharmacology. DHEW Publication 76338, Washington DC. Huxley, R., Mendis, S., Zheleznyakov, E., Reddy, S., Chan, J., 2010. Body mass index, waist circumference and waist: hip ratio as predictors of cardiovascular risk–a review of the literature. Eur. J. Clin. Nutr. 64, 16–22. Kane, J., Honigfeld, G., Singer, J., Meltzer, H., 1988. Clozapine for the treatment-resistant
211