Diabetes Care 1 Angela Galler,1 Esther Bollow,2 Comparison of Glycemic and CLIN CARE/EDUCATION/NUTRITION/PSYCHOSOCIAL Michael Meusers,3 Bela Bartus,4 Metabolic Control in Youth Andrea Nake,¨ 5 Holger Haberland,6 Edith Schober,7 and Reinhard W. Holl,2 for With Type 1 Diabetes With the German Federal Ministry for Education and Research (BMBF) Competence and Without Antipsychotic Network of Diabetes Mellitus Medication: Analysis From the Nationwide German/Austrian Diabetes Survey (DPV) DOI: 10.2337/dc14-2538 OBJECTIVE The objective of this study was to explore metabolic risk factors and glycemic control in youth with type 1 diabetes treated with typical or atypical antipsychotics. RESEARCH DESIGN AND METHODS Data for 60,162 subjects with type 1 diabetes up to the age of 25 years registered in the nationwide German/Austrian Diabetes Survey were included in the analy- sis. BMI; HbA1c; treatment strategy; prevalence of hypertension, dyslipidemia, microalbuminuria, and retinopathy; frequency of hypoglycemia and diabetic ketoacidosis (DKA); and immigrant status among subjects treated with typical 1Charite´dUniversitatsmedizin¨ Berlin, Paediatric or atypical antipsychotics were compared with those without antipsychotic med- Endocrinology and Diabetology, University Hospital for Children and Adolescents, Campus ication and analyzed by regression analysis. Virchow Klinikum, Berlin, Germany 2Institute of Epidemiology and Medical Biometry, RESULTS University of Ulm, Ulm, Germany A total of 291 subjects with type 1 diabetes (median diabetes duration 7.2 years) 3Gemeinschaftskrankenhaus Herdecke, Child and Adolescent Psychiatry, Herdecke, Germany received antipsychotic medications (most commonly risperidone). Subjects trea- 4 P Filderklinik Paediatrics, Filderstadt, Germany ted with antipsychotics had a higher BMI ( = 0.004) and dyslipidemia was more 5University Carl Gustav Carus Dresden, University frequent (P = 0.045) compared with subjects not receiving antipsychotic medica- Hospital for Children and Adolescents, Dresden, tion. Frequencies of severe hypoglycemia and DKA were significantly higher in Germany 6 subjects receiving antipsychotics (P < 0.001). The prevalences of hypertension, Sana Hospital Berlin Lichtenberg, Hospital for Children and Adolescents, Berlin, Germany microalbuminuria, and retinopathy were not different. In subjects treated with 7Medical University Vienna, University Hospital typical antipsychotics, glycemic control did not differ compared with those who for Children and Adolescents, Vienna, Austria did not receive antipsychotic medications. By contrast, subjects treated with Corresponding author: Angela Galler, angela atypical antipsychotics had higher HbA1c levels (P = 0.022). [email protected]. Received 25 October 2014 and accepted 25 Conclusions February 2015. This analysis from a real-life survey demonstrated that subjects with antipsychotic This article contains Supplementary Data online medication had worse glycemic control and a higher rate of acute complications at http://care.diabetesjournals.org/lookup/ compared with those without antipsychotic medication. Health care teams caring suppl/doi:10.2337/dc14-2538/-/DC1. for youth with type 1 diabetes taking antipsychotic medication need to know © 2015 by the American Diabetes Association. Readers may use this article as long as the work about these findings. We suggest monitoring metabolic risk factors as well as is properly cited, the use is educational and not providing diabetes education about prevention of acute complications. for profit, and the work is not altered. Diabetes Care Publish Ahead of Print, published online March 17, 2015 2 Type 1 Diabetes and Antipsychotic Medication Diabetes Care Over the past two decades the use of diabetes treated with atypical or typical ISPAD and American Diabetes Association antipsychotic drugs in medical practice antipsychotics in a real-life setting. guidelines (21). Albuminuria was screened has increased in the U.S. and in Europe according to ISPAD and the German Dia- (1–7). In the pediatric, adolescent, and RESEARCH DESIGN AND METHODS betes Association guidelines (21,22). Albu- young adult populations, antipsychotics The German and Austrian Diabetes Sur- min and creatinine were measured by are primarily used for treatment of psy- vey (DPV) is a prospective, nationwide center-specific laboratory methods that chiatric disorders, mainly schizophrenia, documentation of subjects with type 1 had to meet German internal and external and for concomitant treatment of bipo- diabetes. Demographic, anthropometric, quality requirements for laboratory anal- lar and affective disorders. In clinical and diabetes-related data of subjects ysis, according to the guidelines of the practice antipsychotics also are used for with type 1 diabetes are recorded in German Medical Association (23). Micro- the treatment of symptoms related to 391 diabetes care centers in Germany albuminuria was defined as an albumin behavior and of nonpsychotic disorders, and Austria. All data were derived from excretion rate $20mg/min or urine for example, in obsessive-compulsive hospital charts or were specifically asked albumin-to-creatinine ratio $2.5 mg/mmol. disorders, personality disorders, Tourette for by health professionals. No self-report Macroalbuminuria was defined as syndrome, and autism (1–7). questionnaires were used. The ethics albumin excretion rate $ 200 mg/min Treatment with antipsychotics may committee of Ulm University approved or a urine albumin-to-creatinine ratio result in serious adverse effects. In ad- data collection and anonymous analysis $35 mg/mmol (21,24). Serum concen- dition to the extrapyramidal symptoms for study purposes. Overall, data from trations of cholesterol, HDL, and LDL caused by typical antipsychotics, the ad- 83,748 subjects with type 1 diabetes were measured. Dyslipidemia was diag- verse metabolic effects of atypical anti- were registered in the survey from 1995 nosed if at least one lipid parameter was psychotics are of concern (8–14). The until March 2013. A total of 60,162 chil- increased (cutoff levels of .200 mg/dL use of atypical antipsychotics may result dren, adolescents, and adults (up to the for total cholesterol, .130 mg/dL for in weight gain, changes in lipid concen- age of 25 years) with type 1 diabetes and LDL, and ,35 mg/dL for HDL) and if trations, increased glucose concentra- with a diabetes duration of more than at least two of three consecutive mea- tions, and insulin resistance in healthy 6 months were included in this analysis. surements were above the cutoff subjects (8–14). Weight gain occurs in Clinical characteristics and laboratory levels. Immigration status was defined up to 50% of patients receiving long- parameters were assessed; BMI and as the place of birth of one or both term treatment for schizophrenia (8). blood pressure were recorded, and parents in a country other than Germany Several studies indicate that weight age-specific normal values were ob- or Austria. gain with antipsychotic medication is tained from current guidelines (16–18). Subjects taking antipsychotic medica- more severe in children and adolescents BMI SD score (SDS) was calculated using tions were identified. Antipsychotic compared with adults (8). One study of national reference data in Germany agents (neuroleptics) were classified as children encountered a 10% increase in (16). Hypertension was defined accord- either atypical antipsychotics or typical body weight after 3 months of treat- ing to the Second Task Force on Blood antipsychotics (Table 1). Duration of an- ment with olanzapine (15). Further- Pressure Control in Children and Adoles- tipsychotic medication use was assessed. more, adults and youths taking atypical cents (17,18). Furthermore, treatment antipsychotics have an increased risk of strategy (multiple daily injections or in- Statistical Analysis developing type 2 diabetes (6,13,14). sulin pump treatment); frequency of We used SAS statistic software version Weight gain and changes in metabolic severe hypoglycemia (hypoglycemia 9.3 for data evaluation and statistical parameters also have been described with a loss of consciousness or seizure); analysis. Data are presented as medians for treatment with some typical antipsy- frequency of episodes of diabetic ketoa- and interquartile ranges or as means chotic agents. Interpretation of these cidosis (DKA; pH ,7.30, defined accord- and 95% upper and lower confidence effects is more difficult because of the ing to International Society for Pediatric limits, as appropriate. Prevalences and inadequate evaluation of confounding and Adolescent Diabetes [ISPAD] guide- frequencies of complications are pre- factors in many early studies of typical lines) (19); and frequency of hospital ad- sented as either percentages or as event antipsychotics (8). missions were assessed. Rates of severe rates per 1 patient-year. The Kruskal- Because of the adverse metabolic ef- hypoglycemia, rates of DKA, and fre- Wallis test was performed to compare fects of antipsychotics, patients with quency of hospital admissions were as- age, diabetes duration, BMI SDS, and type 1 diabetes, who already are at risk sessed during the previous year and median HbA1c in subjects with or with- for an unfavorable metabolic profile, calculated and expressed per 1 patient- out treatment with typical or atypical may experience further deterioration year. Glycemic control was assessed as antipsychotics. The x2 test was used to of their metabolic risk factors with median HbA1c during the previous year.