Glucose and lipid metabolism of long-term risperidone monotherapy in patients with schizophrenia


Mari Murashita, MD, PhD, Department of Psychiatry, Hokkaido University Graduate School of Medicine, North-14, West-7, Sapporo 060-8638, Japan. Email:


Abstract  Risperidone has a relatively low risk of causing obesity and diabetes mellitus and is a first-line treatment for schizophrenia. The aim of the present study was to investigate glucose and lipid metabolism, and feeding-control parameters in schizophrenia patients treated with long-term risperidone monotherapy. Fifteen patients with paranoid-type schizophrenia who had been treated with risperidone and had Global Assessment of Function (GAF) scores >70 were selected and compared with healthy volunteers (n = 25). Single assessments of psychotic symptoms, side-effects, Drug-Induced Extrapyramidal Symptoms Scale (DIEPSS) score, bodyweight, body fat percentage and blood sampling were performed. Fasting blood glucose, insulin, hemoglobin A1c, homeostasis model assessment insulin resistance index (HOMA-IR), total cholesterol, triglyceride, high density lipoprotein (HDL)-, low density lipoprotein-cholesterol, adiponectin, prolactin and feeding-control parameters (ghrelin and leptin) were analyzed. The body fat percentage (P = 0.0018), body mass index (BMI) (P = 0.0150), fasting blood glucose (P = 0.0358), triglyceride (P = 0.0377), leptin (P = 0.0243), total ghrelin (P = 0.0067), active ghrelin (P = 0.0241) and prolactin (P < 0.0001) levels of patients treated with risperidone were significantly higher than those of healthy volunteers, while the HDL-cholesterol level (P = 0.0222) was significantly lower. Although the patients had very mild psychiatric symptoms and maintained functionally high levels, the glucose and lipid parameters were significantly impaired compared to healthy volunteers. A high level of plasma ghrelin might increase appetite, leading to exacerbation of metabolic impairment.


Recently, atypical antipsychotics, especially clozapine and olanzapine, have been shown to be associated with weight gain1,2 and diabetes mellitus,3–5 whereas the risks of weight gain and developing diabetes mellitus are significantly lower with risperidone compared to olanzapine and clozapine.1,6 Therefore, risperidone seems to be relatively safe with a low risk of metabolic side-effects in chronic schizophrenia.7 Moreover, a long-term prospective study of schizophrenia confirmed that risperidone was superior to haloperidol with regard to the relapse rate, improving psychotic symptoms and reducing extrapyramidal symptoms (EPS).8 Thus, risperidone is a first-line drug for long-term treatment of schizophrenia.

However, there is no systematic report regarding lipid and glucose metabolic states after long-term treatment with risperidone monotherapy compared to healthy controls. Impaired metabolism is associated with several factors, including oversedation or inactivity, caused by severe psychotic symptoms or side-effects.9 Accordingly, to compare metabolic markers with those of healthy volunteers, we selected patients with Global Assessment of Function (GAF) scores >70. We investigated lipid and glucose metabolic states in the long-term risperidone monotherapy group, who had good global function. Glucose, insulin, HbA1c and homeostasis model assessment insulin resistance index (HOMA-IR) were measured for assessing glucose metabolism, while total cholesterol, triglyceride, high-density lipoprotein (HDL)-cholesterol, low-density lipoprotein (LDL)-cholesterol and adiponectin were measured for assessing lipid metabolism. A few case reports have suggested that clozapine and olanzapine increase appetite and weight gain.10,11 Recently, we found that plasma levels of ghrelin and leptin were increased significantly after 6 months of olanzapine treatment compared to those before treatment.12 Ghrelin is the first hormone found to stimulate appetite and food intake in humans.13 Ghrelin was originally isolated from the rat and human stomach as a cognate endogenous ligand for growth hormone (GH) secretagogue receptor,14 and its composition was determined to be an acylated peptide of 28 amino acids in which the serine-3 residue is n-octanoylated.14,15 This n-octanoylation is indispensable for biological activity of ghrelin. The concentration of circulating ghrelin is increased under conditions of negative energy balance such as starvation and anorexia nervosa, whereas the circulating ghrelin is decreased under those of positive energy balance such as feeding and obesity.16 Because central administration of ghrelin induces food intake via neuropeptide-Y and agouti-related protein,17 ghrelin provides a peripheral signal to the hypothalamus to stimulate food intake and adiposity in rodents.18,19 In contrast, the adipocyte-derived circulating leptin signals the state of fat stores to the hypothalamus, inhibiting food intake and further accumulation of fat.20 Ghrelin and leptin may have opposite actions in the regulation of bodyweight. Therefore, we analyzed both ghrelin and leptin as feeding control parameters in order to clarify the mechanism of increased appetite and weight gain in patients with risperidone monotherapy.



This naturalistic study was managed at the Department of Psychiatry at Hokkaido University Hospital. The study conformed to the guidelines of the Ethical Committee on Human Research of Hokkaido University Graduate School of Medicine, and all subjects gave their written informed consent.

Fifteen patients on risperidone monotherapy were recruited from the 415-patient schizophrenia pool of the outpatient clinic. All patients were diagnosed with paranoid chronic schizophrenia based on the DSM-IV criteria and the GAF scores were >70. No one had severe physical complications or organic or neurological diseases. Subjects were treated with risperidone monotherapy for at least 6 months.

The mean age of the 15 schizophrenia patients (male : female, 5:10) was 40.8 ± 11.8 years (range, 21–62 years). Duration of the disease was 8.9 ± 6.4 years (range, 1–20 years). The duration of treatment with risperidone was 2.5 ± 1.9 years (range, 0.5–8.3 years) and the daily dose was 2.6 ± 1.1 mg/day (range, 1–5 mg/day). The average GAF score was 81.9 ± 8.0. Six patients were initially treated with risperidone from the first episode of schizophrenia, and the remaining nine patients had received other typical antipsychotics (haloperidol 2–3 mg/day) and changed to risperidone. Seven patients had biperiden (1–3 mg/day), two patients had clonazepam (0.5 mg/day). The average education duration was 13.6 ± 1.3 years. Five patients were married. Four patients were housewives, two patients were college or university students, and the other nine patients worked in offices or managed a coffee shop. Five patients had family histories of diabetes mellitus. Three of the 10 female patients were postmenopausal.

Twenty-five healthy volunteers (male : female, 12:13) were used as a control group in order to compare metabolic markers with those of the risperidone group. The average age of control subjects was 38.0 ± 11.3 years (range, 21–58), and all of them worked without a night shift. All control subjects had no history of chronic medication or psychiatric diagnosis. None had diabetes mellitus, hypertension, or hyperlipidemia. Two of the control subjects had family histories of diabetes mellitus, and their BMI was <25 kg/m2.


Single assessments of psychotic symptoms, side-effects, GAF, weight measurement and blood sampling were done on the same day at the outpatient clinic. Symptoms of schizophrenia were assessed using the Positive and Negative Symptoms Scale (PANSS)21 by examining positive symptoms, negative symptoms, general psychopathology and the total symptom score. EPS were assessed using the Drug-Induced Extrapyramidal Symptoms Scale (DIEPSS).22 Other side-effects, including the condition of constipation, appetite, urination, menstruation, drowsiness, palpitation and orthostatic hypotension were reported. Patients were interviewed regarding appetite, and classified into ‘increased’, ‘unchanged’, or ‘decreased’ compared with appetite before risperidone treatment. GAF scores23 were examined to assess the global severity of illness and the quality of life.

Weight and body fat percentage were measured using a Tanita TBF-410 (Tanita, Japan). BMI was expressed in kg/m2. Blood pressure and heart rate were examined. After overnight fasting, blood samples were collected between 08.30 and 09.30 hours. Glucose, insulin and HbA1c were measured for assessing glucose metabolism, while, for assessing insulin resistance (IR) and β-cell function, HOMA-IR was calculated as: HOMA-IR = [fasting insulin (µU/mL) × fasting glucose (mg/dL)]/405. Total cholesterol, triglyceride, HDL-cholesterol, LDL-cholesterol and adiponectin were measured for assessing lipid metabolism. Plasma ghrelin and leptin levels were measured for assessing appetite-stimulating and -inhibiting parameter, respectively. Serum prolactin and GH levels were analyzed, because ghrelin stimulates GH release.

Twenty-five healthy volunteers were measured for bodyweight, body fat percentage and the same parameters of blood sampling as those measured in patients.

Biological assay

Serum leptin and adiponectin levels were determined using radioimmunoassay (RIA) kits (Linco Research, St Charles, MO, USA). Serum GH was analyzed with an immunoradiometric assay (IRMA) kit (Daiichi Radioisotope Laboratories, Tokyo, Japan).

Plasma ghrelin levels were measured with two RIA systems.24 One RIA system measured the acylated active ghrelin concentration and the other the total ghrelin concentration, including active and des-acylated inactive forms.

Statistical analysis

The data are presented as mean ± SD. The significance of categorical variables was analyzed with the χ2 test or Fisher exact test when appropriate. The unpaired Student’s t-test was used for comparison of continuous variables in two groups. Probabilities of <0.05 were considered significant.


The average scores of the four PANSS subscales of 15 patients were positive, 8.2 ± 2.2 (range: 7–15); negative, 10.6 ± 2.7 (range 7–16); general, 20.8 ± 6.1 (range: 16–34); and total, 39.6 ± 10.1 (30–65). The average DIEPPS score was 0.3 ± 0.7 (range: 0–2). Other side-effects, such as constipation 3/15 (20%), abnormality of urination 0/15 (0%), abnormality of menstruation 1/7 (14%), somnolence 5/15 (33%), palpitation 1/20 (6.7%), and orthostatic hypotension 4/15 (26.7%) were reported by interview. No one had tardive dyskinesia, abnormal blood pressure, or abnormal heart rate. Regarding appetite, five of the 15 patients (33%) reported increased appetite and the other patients reported unchanged appetite.

The clinical characteristics of the 15 schizophrenia patients were compared to the 25 healthy volunteers and there was no significant difference in sex (P = 0.3637), age (P = 0.4531) or height (P = 0.4981) between the two groups.

Table 1 shows the metabolic changes of the risperidone group compared with healthy volunteers. Lipid metabolic parameters such as the body fat percentage (P = 0.0018), BMI (P = 0.0150), triglyceride (P = 0.0377) and leptin (P = 0.0243) in the risperidone group were significantly higher than those in the healthy volunteers. HDL-cholesterol (P = 0.0222) was significantly lower in the risperidone group.

Table 1.  Subject data
 Risperidone group (range)Healthy volunteers (range)P
  • P < 0.05.

  • BMI, body mass index; FBS, fasting blood glucose; GH, growth hormone; HbA1c, hemoglobin A1c; HDL, high-density lipoprotein; HOMA-IR, homeostasis model assessment insulin resistance index; LDL, low-density lipoprotein.

Bodyweight (kg)66.3 ± 11.1 (51.2–88.8)62.4 ± 10.6 (48.2–82.9)0.2802
Body fat percentage (%)30.7 ± 8.6 (17.0–54.1)23.2 ± 5.6 (9.1–37.6)0.0018*
BMI (kg/m2)25.2 ± 3.5 (20.6–35.5)22.8 ± 2.1 (18.0–25.0)0.0150*
FBS (mg/dL)98.7 ± 10.2 (87.0–123.0)92.8 ± 7.1 (80.0–109.0)0.0358*
Insulin (µIU/mL)6.7 ± 4.2 (3.0–18.3)5.5 ± 2.2 (2.2–11.4)0.2492
HbA1c (%)5.0 ± 0.5 (4.5–6.1)4.8 ± 0.4 (4.0–5.5)0.2508
HOMA-IR1.7 ± 1.1 (0.71–4.47)1.4 ± 0.9 (0.5–5.2)0.3598
Total cholesterol (mg/dL)195.9 ± 32.1 (152.0–246.0)182.7 ± 29.5 (116.0–216.0)0.1952
Triglyceride (mg/dL)100.9 ± 9.7 (44.0–161.0)77.7 ± 5.9 (24.0–149.0)0.0377*
HDL-cholesterol (mg/dL)58.7 ± 16.9 (28.0–100.0)70.9 ± 14.9 (47.0–101.0)0.0222*
LDL-cholesterol (mg/dL)123.4 ± 31.9 (60.0–168.0)105.1 ± 24.5 (54.0–140.0)0.0513
Adiponectin (mg/dL)10.1 ± 4.2 (4.4–10.8)12.9 ± 6.5 (4.1–30.7)0.1406
Prolactin (ng/mL)46.7 ± 28.7 (19.0–138.3)9.1 ± 3.9 (4.1–19.9)<0.0001*
GH (ng/mL)1.2 ± 1.3 (0.11–4.8)2.3 ± 3.4 (0.1–11.5)0.2164
Leptin (ng/mL)10.4 ± 4.1 (1.2–21.5)6.4 ± 4.2 (2.3–19.9)0.0243*
Total ghrelin (fmol/mL)271.4 ± 182.3 (146.1–636.0)159.2 ± 57.4 (138.2–280.0)0.0067*
Active ghrelin (fmol/mL)24.2 ± 20.8 (3.0–66.2)13.5 ± 7.5 (3.9–30.18)0.0241*

Among the glucose metabolic parameters, only fasting blood glucose level (P = 0.0358) was significantly higher in the risperidone group.

As for the feeding-control parameters, total ghrelin (P = 0.0067) and active ghrelin (P = 0.0241) levels were significantly higher in the risperidone group. The serum leptin level was also significantly higher in the risperidone group.

The prolactin level was significantly higher (P < 0.0001) in the risperidone group.


In the present report we found mild impairment of glucose and lipid metabolism, as well as abnormal feeding control parameters in patients treated with long-term risperidone monotherapy compared to healthy volunteers. The results of GAF and PANSS scores showed that the present patients were well controlled and had good global function and very mild symptoms. In fact, all 15 patients were students, housewives or office workers. Furthermore, the frequencies of occurrence of EPS and other side-effects determined by interview were low and not severe, respectively.

Nevertheless, patients on risperidone monotherapy had significantly impaired body fat percentage, BMI, triglyceride, HDL-cholesterol and leptin compared to healthy volunteers. Although the present study has limitations, such as lack of comparison with other antipsychotics, chronic risperidone treatment may cause such metabolic abnormalities. In addition, we did not directly compare metabolic and feeding control parameters before and after long-term risperidone monotherapy. Consistently, a retrospective chart review reported that risperidone-treated patients had significant increases in weight, BMI and fasting triglyceride at 1 year of treatment compared with the baseline.25 The safety for metabolic impairment in long-term risperidone treatment is a matter of concern, suggesting that this study is worthwhile.

Multiple cases of diabetes mellitus and diabetes ketoacidosis have been reported during treatment with clozapine and olanzapine,3–5 whereas there are only three case reports of new-onset diabetes mellitus associated with risperidone.26,27 In the present study, fasting glucose levels in the risperidone group were significantly higher than those of the controls, but did not reach a pathological range.

The risk of weight gain, diabetes and hyperlipidemia in patients treated with risperidone is lower than in those treated with clozapine and olanzapine.28 However, a possible increase in risk would be predicted to occur in association with any second-generation antipsychotic that increases weight and adiposity. Therefore, regular metabolic monitoring is required during risperidone treatment.

Newly discovered appetite-stimulating hormone, ghrelin, was measured in the present study. In obese humans the circulating ghrelin level is known to be reduced29 and the leptin level is increased.30 The circulating ghrelin level is negatively correlated with the leptin concentration as well as the body fat mass.30 In the present study, although the leptin level, body fat percentage and BMI were significantly increased, the concentrations of both circulating total and active ghrelin were significantly higher in the risperidone group than those of healthy volunteers. This contrary finding suggests that the increased ghrelin level may be directly influenced by risperidone. Thus, a high level of circulating ghrelin increased appetite, and it could lead to weight gain and metabolic impairment in patients treated with risperidone. Certainly, five patients (33%) reported the increased appetite subjectivity; this result was not always parallel with plasma ghrelin level. Therefore, circulating ghrelin level could be an objective marker of appetite stimulation.

Feeding behavior is extremely complicated because not only hormonal regulation but also many other factors, including emotion, memory, learning, and cognitive factors, are involved. Certainly, ghrelin is not the only factor to regulate appetite; but high ghrelin levels in patients on long-term risperidone monotherapy could partially affect metabolic impairment. In conclusion, glucose and lipid metabolism were mildly impaired in schizophrenia patients with clinically mild symptoms and good global function after long-term risperidone monotherapy. Impaired regulation of plasma ghrelin might be associated with the abnormality of lipid and glucose metabolic parameters. Risperidone is highly recommended for the treatment of schizophrenia patients intolerant of clozapine or olanzapine because of weight gain and metabolic impairment, but monitoring of metabolic and feeding parameters during risperidone treatment is needed for preventing metabolic syndrome and maintaining good quality of life. In future, a large prospective follow-up study comparing other antipsychotic drugs will be needed to confirm the influence of long-term risperidone treatment on metabolic and feeding-control parameters.


The authors thank Mrs Seiko Furukawa for analyses by RIA (leptin and adiponectin) and IRMA (GH) assay.