Written by Peter Liu
Spoon Feed
Poor glycemic control was associated with worse cognitive scores in patients with type 2 diabetes. However, none of the investigated treatments showed a superior effect on cognitive scores, despite insulin glargine and liraglutide offering better glycemic control.
Low cognition, high blood sugar – a risky combination in diabetes
Control of hyperglycemia seems to benefit patients largely due to prevention of three microvascular disease complications: retinopathy, nephropathy, and peripheral neuropathy. Macrovascular complications, such as stroke and heart attack, show less benefit from good glycemic control. It appears that cognitive outcomes show a similar result. There is a clear association between diabetes and worse cognitive outcomes, but no clear therapy that wins out as a brain-saving treatment.
In the GRADE RCT, 3721 adults with type 2 diabetes of <10 years’ duration were assigned to receive glargine, glimepiride, liraglutide, or sitagliptin added to metformin (four treatment arms total). Cognitive tests were measured at baseline and on 4-year follow-up. No significant between-group differences were found in 4-year cognitive outcomes. However, a 1% increase in time-weighted HbA1c was associated with lower Digit Symbol Substitution Test scores (−0.94; 95%CI −1.30 to −0.57, p<0.001) and Spanish-English Verbal Learning Test scores (−0.27; 95% CI, −0.49 to −0.06, p=0.01). These results clearly link lower cognition with worse glycemic outcomes in diabetes.
While these results are interesting, and supported by other observational trials, the findings do not clearly imply how clinicians should respond. For example, it is not clear whether lower cognition causes the poor glycemic outcomes or vice versa. Probably, the relationship is bidirectional. Also, despite a modest increase in the efficacy of insulin and GLP-1 agonists over sulfonylureas and DPP4 inhibitors, these benefits do not seem to translate into a clear difference in cognitive outcomes.
How does this change my practice?
I have a long list of standard items that I consider during my evaluation of diabetic patients, which includes adequacy of glycemic control, risk of hypoglycemia, presence of irreversible end-organ damage, and socioeconomic factors. It seems wise to add current cognitive status to this. Low baseline cognition, or declining cognition over time, could be a red flag for poorer prognosis, even if its interaction with diabetes is multifactorial and slightly mysterious. Addressing cognitive barriers to medication and lifestyle adherence could also be a useful focus during clinic visits.
Source
Glucose-Lowering Medications, Glycemia, and Cognitive Outcomes: The GRADE Randomized Clinical Trial. JAMA Intern Med. 2025 May 19:e251189. doi: 10.1001/jamainternmed.2025.1189. Epub ahead of print. PMID: 40388190
