Skin Autofluorescence is an Independent Predictor of Post Stroke Infection in Diabetes.

Post stroke infection occurs in 15-20% of acute stroke patients and is associated with a poor longterm outcome. In a prospective study on 113 acute ischemic stroke patients with diabetes mellitus 15.9% suffered nosocomial infection. We found chronic hyperglycemia measured by skin autofluorescence in arbitrary units to be an independent predictor of a nosocomial infection post stroke (OR = 3.24 [CI 95%: 1.13; 9.26], p = 0.029). Skin autofluorescence represents the glycemic memory beyond HbA1c. Potential mechanisms leading from increased skin autofluorescence to vulnerability for infectious complications include more severe strokes due to preexisting vasculopathy and exacerbated post stroke immunosuppression.

Measuring of Advanced Glycation End Products in Acute Stroke Care: Skin Autofluorescence as a Predictor of Ischemic Stroke Outcome in Patients with Diabetes Mellitus.

Background: Patients with diabetes mellitus (DM) are known to show poor recovery after stroke. This specific burden might be due to acute and chronic hyperglycemic effects. Meanwhile, the underlying mechanisms are a cause of discussion, and the best measure to predict the outcome is unclear. Skin autofluorescence (SAF) reflects the in-patient load of so-called advanced glycation end products (AGEs) beyond HbA1c and represents a valid and quickly accessible marker of chronic hyperglycemia. We investigated the predictive potential of SAF in comparison to HbA1c and acute hyperglycemia on the functional outcome at 90 days after ischemic stroke in a cohort of patients with DM. Methods: We prospectively included 113 patients with DM type 2 hospitalized for acute ischemic stroke. SAF was measured on each patient's forearm by a mobile AGE-Reader mu© in arbitrary units. HbA1c and the area under the curve (AUC) of the blood sugar profile after admission were assessed. Functional outcome was assessed via phone interview after 90 days. A poor outcome was defined as a deterioration to a modified Rankin Scale score ≥ 3. A good outcome was defined as a modified Rankin Scale score < 3 or as no deterioration from premorbid level. Results: Patients with a poor outcome presented with higher values of SAF (mean 3.38 (SD 0.55)) than patients with a good outcome (mean 3.13 (SD 0.61), p = 0.023), but did not differ in HbA1c and acute glycemia. In logistic regression analysis, age (p = 0.021, OR 1.24 [1.12−1.37]) and SAF (p = 0.021, OR 2.74 [1.16−6.46]) significantly predicted a poor outcome, whereas HbA1c and acute glycemia did not. Patients with a poor 90-day outcome and higher SAF experienced more infections (4.2% vs. 33.3% (p < 0.01)) and other various in-hospital complications (21.0% vs. 66.7% (p < 0.01)) than patients with a good outcome and lower SAF levels. Conclusions: SAF offers an insight into glycemic memory and appears to be a significant predictor of poor stroke outcomes in patients with DM exceeding HbA1c and acute glycemia. Measuring SAF could be useful to identify specifically vulnerable patients at high risk of complications and poor outcomes.