Glycated Hemoglobin A1c Time in Range and Dementia in Older Adults With Diabetes

• Published in: JAMA network open 2024-08, Vol.7 (8), p.e2425354
• Main Authors: Underwood, Patricia C, Zhang, Libin, Mohr, David C, Prentice, Julia C, Nelson, Richard E, Budson, Andrew E, Conlin, Paul R
• Format: Article
• Language: English
• Subjects: Aged; Aged, 80 and over; Cohort Studies; Dementia - blood; Dementia - epidemiology; Diabetes; Diabetes and Endocrinology; Diabetes Mellitus - blood; Diabetes Mellitus - epidemiology; Female; Glycated Hemoglobin - analysis; Hemoglobin; Humans; Hypoglycemia; Incidence; Male; Older people; Online Only; Original Investigation; United States - epidemiology; Veterans - statistics & numerical data
• ISSN: 2574-3805; 2574-3805
• DOI: 10.1001/jamanetworkopen.2024.25354

Individuals with diabetes commonly experience Alzheimer disease and related dementias (ADRD). Factors such as hypoglycemia, hyperglycemia, and glycemic variability have been associated with increased risk of ADRD. Traditional glycemic measures, such as mean glycated hemoglobin A1c (HbA1c), may not identify the dynamic and complex pathophysiologic factors in the association between diabetes and ADRD. The HbA1c time in range (TIR) is a previously developed measure of glycemic control that expresses HbA1c stability over time within specific ranges. This measure may inform the current understanding of the association between glucose levels over time and ADRD incidence. To examine the association between HbA1c TIR and incidence of ADRD in older veterans with diabetes. The study sample for this cohort study was obtained from administrative and health care utilization data from the Veterans Health Administration and Medicare from January 1, 2004, to December 31, 2018. Veterans 65 years or older with diabetes were assessed. Participants were required to have at least 4 HbA1c tests during the 3-year baseline period, which could start between January 1, 2005, and December 31, 2014. Data analysis was conducted between July and December 2023. Hemoglobin A1c TIR was calculated as the percentage of days during baseline in which HbA1c was in individualized target ranges based on clinical characteristics and life expectancy, with higher HbA1c TIR viewed as more favorable. The association between HbA1c TIR and ADRD incidence was estimated. Additional models considered ADRD incidence in participants who were above or below HbA1c target ranges most of the time. The study included 374 021 veterans with diabetes (mean [SD] age, 73.2 [5.8] years; 369 059 [99%] male). During follow-up of up to 10 years, 41 424 (11%) developed ADRD. Adjusted Cox proportional hazards regression models showed that lower HbA1c TIR was associated with increased risk of incident ADRD (HbA1c TIR of 0 to