Higher visit-to-visit glucose and HbA1c variability are associated with lower gray matter and higher white matter hyperintensity volumes in older adults: The Multi-Ethnic Study of Atherosclerosis (MESA)
Abstract Body (Do not enter title and authors here): Background: Dysglycemia is a well-established risk factor for cognitive decline. The association of long-term glucose control with early neuroimaging markers of dementia remains largely unexplored. Hypothesis: Higher visit-to-visit variability in glucose and HbA1c, indicating worse glucose control, are associated with lower total and regional gray matter volume (GMV) and higher white matter hyperintensity (WMH) volume. Methods: Fasting glucose (FG) was measured in MESA participants at Exam 1 (2000–02) and each of 5 follow-up exams through Exam 6 (2016-18), and HbA1c was measured at Exam 2 (2002–04), Exam 5 (2010–12), and Exam 6. We calculated intraindividual variability in FG and HbA1c as the SD (SDFG; SDA1c), average real variability (ARVFG; ARVA1c), and variability independent of the mean (VIMFG; VIMA1c) through Exam 6. Total and regional GMV and WMH volume were assessed by 3T brain MRI (2017-22). We report the SD difference in GMV and log-unit difference in WMH volume per two-fold increment in glucose or HbA1c variability with 95% CI from multivariable linear regression models adjusted for total intracranial volume, study site, age, sex, race/ethnicity, education, vascular risk factors, and APOE ε4 allele status. Results: A total of 1,424 Exam 6 participants (mean (SD) age 72 (8) yrs; 46% men; 28% with diabetes at Exam 6) had ≥3 FG measurements (mean: 5.9 measurements) and MRI data. Two-fold increments of SDFG, ARVFG, and VIMFG were associated with –0.08 (–0.11, –0.06), –0.09 (–0.12, –0.06), and –0.10 (–0.14, –0.06) SD differences in total GMV and 0.13 (0.06, 0.19), 0.13 (0.06, 0.20), and 0.16 (0.05, 0.26) log-unit differences in WMH volumes, respectively (Table 1). Associations appeared strongest with frontal lobe GMV. Further adjustment for mean FG through Exam 6 only modestly attenuated results. In 839 participants with 3 HbA1c measurements, two-fold increments of SDA1c, ARVA1c, and VIMA1c were associated with –0.20 (–0.30, –0.10), –0.77 (–1.20, –0.38), and –0.26 (–0.39, –0.12) SD differences in total GMV and 0.3 (0.04, 0.55), 1.40 (0.38, 2.3), and 0.36 (0.02, 0.70) log-unit differences in WMH volumes, respectively (Table 2). Neither glucose nor HbA1c variability were associated with temporal lobe or hippocampal volumes. Associations did not differ by diabetes status. Conclusion: Higher variability of glucose and HbA1c over 16–18 years were associated with lower GMV and higher WMH volumes in a diverse, community dwelling sample of older adults.
Krizan, Ivan
( Wake Forest School of Medicine
, Winston Salem
, North Carolina
, United States
)
Hughes, Timothy
( Wake Forest School of Medicine
, Winston Salem
, North Carolina
, United States
)
Mongraw-chaffin, Morgana
( MedStar Health Research Institute
, Winston Salem
, North Carolina
, United States
)
Schaich, Christopher
( Wake Forest University
, Winston-Salem
, North Carolina
, United States
)
Tanley, Jordan
( Wake Forest School of Medicine
, Winston Salem
, North Carolina
, United States
)
Brown, Elili
( Wake Forest School of Medicine
, Winston Salem
, North Carolina
, United States
)
Bancks, Michael
( Wake Forest Univ. School of Med.
, Winston Salem
, North Carolina
, United States
)
Nasrallah, Ilya
( University of Pennsylvania
, Philadelphia
, Pennsylvania
, United States
)
Bryan, Nick
( University of Pennsylvania
, Philadelphia
, Pennsylvania
, United States
)
Heckbert, Susan
( UNIVERSITY OF WASHINGTON
, Seattle
, Washington
, United States
)
Luchsinger, Jose
( Columbia University
, New York
, New York
, United States
)
Hayden, Kathleen
( Wake Forest School of Medicine
, Winston-Salem
, United States Minor Outlying Islands
)
Author Disclosures:
Ivan Krizan:DO NOT have relevant financial relationships
| Timothy Hughes:DO NOT have relevant financial relationships
| Morgana Mongraw-Chaffin:DO NOT have relevant financial relationships
| Christopher Schaich:DO NOT have relevant financial relationships
| Jordan Tanley:DO NOT have relevant financial relationships
| Elili Brown:No Answer
| Michael Bancks:DO NOT have relevant financial relationships
| Ilya Nasrallah:DO have relevant financial relationships
;
Consultant:Subtle Medical:Past (completed)
; Advisor:Eisai:Past (completed)
| Nick Bryan:No Answer
| Susan Heckbert:DO NOT have relevant financial relationships
| Jose Luchsinger:DO have relevant financial relationships
;
Consultant:Novo Nordisk:Active (exists now)
; Other (please indicate in the box next to the company name):Merck KGaA:Active (exists now)
; Other (please indicate in the box next to the company name):Wolters Kluwer:Active (exists now)
; Research Funding (PI or named investigator):NIH:Active (exists now)
; Consultant:Merck KGaA:Past (completed)
| Kathleen Hayden:No Answer