Investigation of Concomitant Use of Alzheimer’s Disease Drugs with Sulfonylureas and Serious Hypoglycemia
Year: 2020
Type: article
Study design: Observational Study
Abstract: To the Editor: Serious hypoglycemia from the use of insulin secretagogues is one of the most common and clinically significant adverse drug events.1 A screening study by Han et al2 identified signals of drugs that might be associated with serious hypoglycemia when used with insulin secretagogues, and Alzheimer’s disease (AD) drugs were among those signals. That study used the self-controlled case-series design, which inherently controls for confounding by time-invariant patient factors but did not control for time-varying confounders. We sought to confirm these signals in an independent database (Medicaid), using the self-controlled case-series design with adjustment for time-varying factors. We used Medicaid claims data from five states (California, Florida, New York, Ohio, and Pennsylvania) from 1999 to 2012, supplemented with Medicare claims for dual enrollees. The exposure of interest was use of an AD drug—donepezil, galantamine, memantine, or rivastigmine (precipitant drugs3)—during use of an insulin secretagogue—glimepiride, glipizide, glyburide, nateglinide, or repaglinide—or metformin (object drugs3). The outcome of interest was hospital presentation with serious hypoglycemia, ascertained by discharge diagnosis codes appearing in the principal position of inpatient claims or any position of emergency department claims (positive predictive value, 78%4–89%5; eTable 1; https://links.lww.com/EDE/B726). We used the self-controlled case-series design and additionally controlled for time-varying factors: drugs that may cause hypo- or hyperglycemia; drugs that may interact with insulin secretagogues; and acute infection (eTable 2; https://links.lww.com/EDE/B726). Object drug-specific observation time was censored by the first occurrence of end of supply of the object drug; Medicaid disenrollment; dataset end. We used conditional Poisson regression to calculate confounder-adjusted rate ratios (RRs) with 95% confidence intervals (CIs) for serious hypoglycemia during use of an object drug with versus without an AD drug. Adults (age ≥18 years) continuously enrolled in Medicaid for at least 180 days free of a dispensing for an object drug before the first observed episode of that object drug, who experienced at least one outcome event during the object drug observation time, were eligible. Analyses were conducted using SAS version 9.4. The study was approved by the institutional review board of the University of Pennsylvania. The number of persons experiencing the outcome, the number of outcome occurrences during observation time, person-days of observation time, and the median observation time per individual by object drug are reported in the Table. The median age at start of observation was 61.2–71.1 years, and the proportion of females ranged from 62.6% to 66.6% (eTable 3; https://links.lww.com/EDE/B726). The Figure shows adjusted RRs (sensitivity analyses results were similar; eTables 4; https://links.lww.com/EDE/B726 and 5; https://links.lww.com/EDE/B726). AD drugs examined were not associated with an elevated risk of serious hypoglycemia when used with sulfonylureas. Galantamine and donepezil showed elevated RRs in users of metformin (RR = 1.84; 95% CI = 1.03, 3.29) and nateglinide (RR = 3.19; 95% CI = 1.34, 7.59), respectively. It is unclear whether these associations represent drug interactions, since neither drug pair was previously identified as a potential signal, and we included metformin as a negative control. Further, we know of no pharmacokinetic or pharmacodynamic mechanisms to explain drug interactions of these agents. Nateglinide is metabolized primarily through cytochrome P450 (CYP) 2C9, with a minor portion through CYP3A4,6 while donepezil does not affect CYP2C9 and is unlikely to affect the metabolism of CYP3A4 substrates at therapeutic concentrations.7 Metformin is excreted renally,8 and we are aware of no mechanism by which an interaction with AD drugs could cause hypoglycemia in users of metformin. Also, chance findings are possible given the large number of associations examined. TABLE. - Number of Persons Experiencing the Outcome, Number of Outcome Occurrences During Observation Time, Person-days of Observation Time, and the Median Observation Time per Individual by Object Drug Antidiabetes Drugs (Object Drugs) Glimepiride Glipizide Glyburide Nateglinide Repaglinide Metformin Number of persons experiencing the outcomea 7,828 13,509 14,046 982 1,925 24,931 Number of outcome occurrences during observation time 9,708 17,188 17,332 1,297 2,535 32,098 Person-days of observation time, total 2,234,976 3,755,050 3,667,018 188,109 363,141 8,990,281 Median observation time in days per individual (Q1, Q3) 156 (66, 358) 153 (68, 342) 138 (64. 310) 96 (38, 212) 97 (38, 221) 219 (102, 450) Q1 indicates the first quartile; Q3, the third quartile.aOutcome: serious hypoglycemia. Figure.: Serious hypoglycemia occurrence rate ratios from the concomitant use of insulin secretagogues or metformin (object drugs) and Alzheimer’s disease drugs (precipitant drugs) vs the use of the object drugs without the precipitant drugs. aRate ratio: [(outcome occurrence rate during precipitant − exposed time)/(outcome occurrence rate during precipitant − unexposed time)], for each object–precipitant drug pair; confounder-adjusted. bNo estimates due to a small number of outcome occurrences (n < 5).Thus, our findings did not confirm previous signals of an increased rate in serious hypoglycemia with concomitant use of AD drugs and sulfonylureas. Acknowledgments The authors thank Ms. Qing Liu and Ms. Min Du of the Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania, for their assistance with biostatistics computer programming. The study was approved by the institutional review board of the University of Pennsylvania. Medicaid and Medicare claims data are available via a data use agreement with the Centers for Medicare and Medicaid Services (CMS) (https://www.cms.gov/). The authors did not have any special access privileges that others would not have. The procedures to obtain access to these data are described in the CMS website (https://www.cms.gov/Research-Statistics-Data-and-Systems/Files-for-Order/Data-Disclosures-Data-Agreements/Researchers) and the Research Data Assistance Center (ResDAC) website (https://www.resdac.org/research-identifiable-files-rif-requests). Young Hee NamDepartment of Population MedicineHarvard Medical School and Harvard Pilgrim Health Care InstituteBoston, MA Colleen M. BrensingerWarren B. BilkerCharles E. LeonardSean HennessyCenter for Pharmacoepidemiology Research and TrainingCenter for Clinical Epidemiology and BiostatisticsDepartment of BiostatisticsEpidemiology, and InformaticsPerelman School of MedicineUniversity of PennsylvaniaPhiladelphia, PA[email protected]
Source: Epidemiology
Language: English
Subfield: Toxicology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W3090759901