Second-Generation Antipsychotics and the Risk of Insulin Resistance, Dyslipidemia in Children

July 1, 2005

Comparative data are emerging on the metabolic effects of the second-generation antipsychotics on adolescents and children. Data presented at national meetings is discussed. When using these medications, the risks and benefits need to be carefully weighed until more definitive information is available.

Psychiatric Times

July 2005

Vol. XXII

Issue 8

Comparative data are emerging on the metabolic effects of second-generation antipsychotics (SGAs) used in treating children and adolescents suffering from psychotic, mood and disruptive behavior disorders.

Christoph Correll, M.D., a child and adolescent psychiatrist and research scientist affiliated with Zucker Hillside Hospital and Schneider Children's Hospital in New York, and Anil Malhotra, M.D., director of psychiatry research at Zucker Hillside Hospital, are principal investigators in a longitudinal study examining biological and genetic risk factors for weight gain and metabolic abnormalities in young people taking atypical antipsychotics.

"We are looking at three different treatment conditions: antipsychotic-naive, pretreated but currently drug free and switchers, and we are also looking at six different medications," Correll told Psychiatric Times. The medications are aripiprazole (Abilify), clozapine (Clozaril), olanzapine (Zyprexa), quetiapine (Seroquel), risperidone (Risperdal) and ziprasidone (Geodon).

In order to really say something definitively, Correll said they wanted to have enough antipsychotic-naive patients for the first pass, explaining that combining such patients with those who had been pretreated and those were being switched from one SGA to another would not lead to clear, clean findings. In studies of adults, he pointed out that what may obscure some of the findings about the real risk "is that many of those patients have been chronically pretreated or were switchers."

"We are now at 421 patients. Two months ago, we locked the database, and we are at the tail end of the first set of analyses on just antipsychotic-naive children," Correll said. Currently he and his colleagues are preparing a journal article that will focus on the first 159 antipsychotic-naive youngsters who have had at least one post-baseline assessment. The article will compare olanzapine, quetiapine and risperidone with regard to weight gain, body composition, insulin resistance and dyslipidemia.

Although the study is not randomized, it is the largest study ever done in antipsychotic-naive individuals taking novel antipsychotics, Correll explained, and the only one to date on children that includes prospective, systematic measurements of glucose and lipid metabolism.

"Most of the studies, even in adults, are cross-sectional or based upon pharmacoepidemiologic data; there is very little prospective data available on these very important [safety] issues," he said. "And there is not a single study in kids that is prospective. So we can inform the field."

In April at the XX International Congress on Schizophrenia Research in Savannah, Ga., Correll presented a subanalysis of 93 antipsychotic-naive children taking olanzapine, risperidone or quetiapine, reporting the effects of those SGAs on glucose metabolism and insulin resistance (Correll et al., 2005a). Being antipsychotic naive is important, Correll said, because there is the debate as to whether it is the underlying illness, the patient's lifestyle/behavior or the drugs that cause the insulin resistance, which is a precursor for diabetes and metabolic syndrome.

"We have people who are young and start out often not being obese, and then we follow them for several months, so we will be able more to determine the effect of the medications," he said.

In the subanalysis presented in April, the researchers reported on 93 participants who ranged in age from 5 to 18 (mean age=14.1±3.4) and who had psychotic, mood and/or disruptive behavior spectrum disorders (Correll et al., 2005a). More than half (54.8%) were male and 46.2% were white.

Participants' height, weight, fat mass and percentage (via impedantiometry), waist circumference, fasting glucose, insulin, prolactin, leptin, and SGA levels (ensuring compliance) had been measured at baseline and at four and 12 weeks. The homeostatic model (HOMA-IR=insulin umol x glucose mmol/22.5) was used to calculated insulin resistance.

In the antipsychotic-naive youngsters treated with risperidone (n=51), olanzapine (n=30) or quetiapine (n=12) for eight to 13 weeks, fasting glucose, insulin and insulin resistance increased significantly. One premorbidly obese youngster (1.1%) developed diabetes on quetiapine. Medications did not differ in their effect on glucose and insulin, or on absolute and relative HOMA-IR changes (olanzapine=0.81±1.6, 62.2%; quetiapine=0.95±-2.4, 19.1%; risperidone=0.30±1.5, 32.0%).

When individual SGAs were considered, increases in glucose reached significance only for risperidone, while changes in insulin and HOMA-IR were only significant for olanzapine. Glucose increase was correlated with low baseline glucose levels and male gender (R2=0.48, pp

Based on their subanalysis, the researchers concluded, "Increased insulin resistance in young people after three months of treatment with olanzapine, risperidone or quetiapine ... is of considerable concern." They recommended that clinicians institute routine monitoring of weight and glucose metabolism in this population.

In a related subanalysis, data were presented on dyslipidemia in a larger set of 258 youngsters (mean age=13.8 ±3.5, 62.0% male, 47.9% white) who completed eight to 13 weeks of treatment with aripiprazole (n=48), olanzapine (n=66), quetiapine (n=51), risperidone (n=81) or ziprasidone (n=12) (Olshanskiy et al., 2005). Dyslipidemia was defined as fasting cholesterol 200 mg/dL and/or triglycerides 150 mg/dL. Participants' height, weight, body mass index (BMI), fat mass (bioimpedantiometry), waist circumference, lipid profile, leptin and SGA levels (assessing compliance) were measured at baseline and monthly thereafter. With regard to the entire group, none of the individual lipid parameters increased significantly from baseline to end point, but 19.2% of youths developed new-onset dyslipidemia, without statistical differences between the medication groups.

Correll presented data from a larger cohort of that same study at the Society of Biological Psychiatry meeting and at the American Psychiatric Association meeting, both held in Atlanta. In an abstract prepared for the biological psychiatry meeting, the investigators discussed dyslipidemia in 131 antipsychotic-naive youngsters (mean age=14.1±3.5, 64.9% male, 49.6% white) treated with aripiprazole (n=15), olanzapine (n=37), quetiapine (n=15) or risperidone (n=64) for eight to 14 weeks (Correll et al., 2005b). They concluded, "New-onset dyslipidemia is a relevant side effect of SGA treatment in antipsychotic-naive youths." In the total group, fasting total cholesterol and triglycerides increased significantly compared to baseline.

Analyzing changes for each SGA, lipid increases in total cholesterol low-density lipoprotein (LDL) cholesterol and triglycerides were significant only for olanzapine. Differences between medications reached significance regarding total cholesterol and LDL cholesterol, with olanzapine causing significantly greater increases than risperidone. However, as reported in the earlier group of patients with mixed treatment histories (Olshanskiy et al., 2005), the rates of new-onset dyslipidemia in antipsychotic-naive youths were not different between groups.

Using logistic regression, cholesterol increase was correlated with low baseline cholesterol levels and weight gain. Increases in triglycerides were correlated with weight gain, low baseline triglycerides, divalproex (Depakote) co-treatment, and Asian or African-American race. Also, new-onset dyslipidemia was associated with weight gain and co-treatment with divalproex. Overall, the investigators recommended more frequent lipid monitoring.

At that same meeting, Correll presented data showing that SGAs can adversely affect body composition in children and adolescents (Correll et al., 2005c). Correll and colleagues analyzed prospective data from 301 youngsters, ranging in age from 4 to 19 (mean age=13.7±3.5) who had a DSM-IV diagnosis of schizophrenia spectrum (31.0%), mood (41.1%) or disruptive behavior (27.9%) disorders requiring initiation of SGA treatment. Of the participants, 63.3% were postpubertal, 64.7% were male, and 47.5% were white. They had completed eight to 14 weeks of SGA treatment.

According to Correll, in the entire sample--which included switchers and pretreated individuals as well as antipsychotic-naive participants--body weight, fat mass and waist circumference increased significantly for all SGAs, except aripiprazole, while BMI percentiles increased significantly only for patients on olanzapine and risperidone.

When switchers and pretreated youngsters were removed from the analyses, weight as well as age -and sex-adjusted BMI percentiles increased with all atypical antipsychotic drugs for which data in a sufficient number of never-treated youths were available (i.e., aripiprazole, olanzapine, quetiapine and risperidone). Individuals found to be at high-risk for adverse changes in body composition included those who were antipsychotic-naive, were older with schizophrenia or mood disorders, were treated with olanzapine, were relatively leptin-resistant, and experienced significant early weight gain.

Asked what clinicians should be aware of in terms of treating children and adolescents with second-generation antipsychotics, Correll responded that while novel antipsychotics have been shown to be effective for a number of psychiatric conditions, including nonpsychotic disorders, clinicians have to carefully weigh the risks and benefits of these treatments.

"There seems to be a class effect for some adverse effects in body weight, body composition and related metabolic indices," he said. Also, "there is a differential effect, across agents, but all seem to be able to induce adverse effects at least in the first three months of treatment, so one has to carefully weigh the options. We are collecting data on six-, nine- and 12-month time points to really see where some of these changes level off. It seems some agents that have adverse changes early on level off much earlier than some other medications, which is the reason that in some switch studies medications come off nearly weight neutral, while others still induce some of these adverse changes."

References

Correll CU, Olshanskiy V, Mughal T et al. (2005a), Prospective study of second-generation antipsychotic-induced insulin resistant in antipsychotic-naive children and adolescents. Presented at the XX International Congress on Schizophrenia Research. Savannah, Ga.; April 2-6.

Correll CU, Parikh UH, Mughal T et al. (2005b), Dyslipidemia in antipsychotic-naive youngsters treated with atypical antipsychotics. Presented at the 60th Annual Meeting of the Society of Biological Psychiatry. Atlanta; May 19-21.

Correll CU, Parikh UH, Mughal T et al. (2005c), Effect of past treatment exposure on risk estimation of atypical antipsychotic-induced body composition changes in children and adolescents. Presented at the 60th Annual Meeting of the Society of Biological Psychiatry. Atlanta; May 19-21.

Olshanskiy V, Malhotra AK, Parikh UH et al. (2005), New-onset dyslipidemia in children and adolescents treated with second-generation antipsychotics. Presented at the XX International Congress on Schizophrenia Research. Savannah, Ga.; April 2-6.