Poorer Cognitive Performance Post-Stroke Observed in Patients With Diabetes


Compared to those with normal fasting glucose levels, those with Type 2 diabetes had poorer cognitive performance at 3 and 6 months after suffering a stroke, based on data from the STROKOG consortium.

Jessica Lo

Jessica W. Lo, MSc, research associate, Centre for Healthy Brain Ageing, School of Psychiatry, UNSW Syndey

Jessica W. Lo, MSc

A newly published analysis from the Stroke and Cognition (STROKOG) consortium studies suggests that the presence of diabetes mellitus is associated with poorer cognitive performance for patients 3 to 6 months after stroke. Prediabetes, however, was not associated with any such impact on cognitive performance.1,2

All told, the analysis included 7 international observational studies and 1601 patients. The patients with Type 2 diabetes (T2D) had significantly poorer performance on global cognition assessments (standard deviation [SD], −0.59; 95% CI, −0.82 to −0.36; P <.001) and in all domains compared with patients with normal fasting glucose levels. Stroke location was not found to be significant (P >.1) for any cognitive outcome except the memory domain (P <.001).

Additionally, no significant differences were observed between those with impaired fasting glucose and those with normal fasting glucose in global cognition (SD, −0.10; 95% CI, −0.45 to 0.24; P = .55) or any cognitive domain. The data were analyzed by Jessica W. Lo, MSc, research associate, Centre for Healthy Brain Ageing, School of Psychiatry, UNSW Syndey, and colleagues, including Perminder Sachdev, MD, PhD, senior author, Scientia professor, Centre for Healthy Brain Ageing, UNSW Sydney.

“The deficits we found in all areas of cognitive function highlight the importance of assessing the capacity for self-care in patients with Type 2 diabetes following a stroke,” Lo said in a statement. “We need to ensure that stroke survivors have the mental competency to manage the complex and intertwined tasks to effectively treat Type 2 diabetes, which can include measuring glucose levels multiple times a day, managing glucose monitoring devices, adjusting medication doses, self-administering insulin or other medications, and understanding food labels and portion sizes to adjust what is eaten at each meal or snack.”

READ MORE: Recommendations for Prehospital Stroke Care During COVID-19 Published

The mean age of the cohort was 66.0 years (range, 22—96), with 63% of the cohort being men. As for racial demographics, 70% of the patients were Asian, 26% were white, and 2.6% were African American. In total 99% of the patients had had an ischemic stroke, and 36% (n = 578) were categorized as having T2D, 12% (n = 193) had impaired fasting glucose, and 52% (n = 830) had normal fasting glucose. Patients were assigned according to the World Health Organization criteria, with T2D defined as fasting glucose levels ≥7 mmol/L, having a previous diagnosis of diabetes mellitus, or using a diabetes mellitus medication; impaired was defined as fasting glucose levels 6.1 to 6.9 mmol/L; and normal was defined as fasting glucose levels <6.1 mmol/L.

As well, Lo and colleagues conducted an analysis with a second definition of impaired fasting glucose, the 2003 American Diabetes Association criteria. These criteria defined a broader range for impaired fasting glucose (5.6—6.9 mmol/L). Additionally, there was a subgroup analysis of 4 studies conducted based on hemoglobin A1c (HbA1c) ≥6.5% to define diabetes mellitus and HbA1c between 5.7% and 6.4% to define prediabetes.

That analysis produced similar results, with effect sizes changing between ±2% and 13% for T2D versus normal, or remaining near 0 for impaired versus normal. The subanalysis from 4 studies (n = 1176) to define diabetes mellitus (33%) and prediabetes (31%) produced similar results as well.

A history of depression was not significantly related to global cognition (P = .91) or any cognitive domain, but the sample size was reduced to 454 because of a lack of data for 72% of the sample. An analysis performed in the subgroup of patients without diabetes mellitus (n = 1023), fasting glucose levels as a continuous measure was not significantly related to global cognition or performance in any cognitive domain, and the effect sizes were close to 0.

“T2D increases the risk of stroke and has been associated with cognitive impairment and may increase dementia risk. That’s why T2D is another important target in the prevention of dementia, and the focus should be on early treatment for prediabetes to delay or prevent the progression to T2D,” Sachdev said in a statement. “In this study, we wanted to know if stroke patients with prediabetes also have worse cognitive function compared to stroke patients without prediabetes or diabetes.”

“This is important because prediabetes is very common, and individuals can have prediabetes for several years before progressing to T2D. Early and aggressive treatment of prediabetes can delay or prevent T2D. If we target the treatment of prediabetes, could this prevent the development of dementia in some individuals?” Sachdev added.

The study did not identify any significant interactions between impaired fasting glucose and sex for global cognition nor any cognitive domain. In women, the link between T2D and poorer cognitive function were significant for the memory (P = .025), perceptual motor (P = .040), and executive function domains (P = .044), but these associations were not so in men. Similarly, no significant interactions were observed between impaired fasting glucose and age for global cognition or any domain. T2D and age did have a single significant association for the memory domain (P = .03), with a stronger association between T2D and poorer memory in younger stroke patients.

Finally, when Lo and colleagues assumed a 20% elevation in fasting glucose levels because of stress hyperglycemia, similar results—with small changes in effect sizes&mdash;were observed despite the new glucose group sizes (T2D: 30.8%; IFG: 4.4%; normal: 64.8%).

“While our study is focused on cognition after a stroke, there is strong evidence that T2D is associated with cognitive impairment,” Lo said. “This is an important message for the general public. Since our study shows no evidence that prediabetes is associated with worse cognitive performance, this emphasizes the importance of the early diagnosis and treatment of prediabetes (which is often under-diagnosed) in order to delay or prevent the progression to T2D.”


1. Lo JW, Crawford JD, Samaras K, et al. Association of Prediabetes and T2DWith Cognitive Function After Stroke: A STROKOG Collaboration Study. Stroke. 2020;51:00-00. DOI: 10.1161/STROKEAHA.119.028428.

2. Type 2 diabetes linked to worse cognitive performance after a stroke; prediabetes not linked, but prevention needed [press release]. Dallas, TX: American Heart Association; Published May 14, 2020. Accessed May 19, 2020. apnews.com/5d0a96355a99ffd2437a55ba5e7155ec

Related Videos
Video 2 - 4 KOLs are feature in, "Changes in Presentation of Spasticity Over Time"
Video 1 - 4 KOLs are feature in, "Definition and Pathophysiology of Spasticity"
Dolores D. Santamaria, MD
 Bruce Cree, MD, PhD, MAS, FAAN
Video 3 - 5 KOLs are featured in "Transitions of Care: Moving Spinal Muscular Atrophy Patients from Pediatric/Adolescent Care to Adult Clinics"
Video 3 - 5 KOLs are featured in "Presentation of Adult Spinal Muscular Atrophy in Clinics"
Fawad Khan, MD, FACNS
Valerie J. Block, PT, DPTSc
© 2024 MJH Life Sciences

All rights reserved.