Cerebral Microbleeds Are Associated With Deficits In Cognitive Processing Speed And Executive Functions In Middle-Aged Adults With Type 1 Diabetes
Kylaeheiko, I.; Tarkkonen, A.; Kuusela, L.; Martola, J.; Paajanen, T. I.; Virkkala, J.; Groop, P.-H.; Thorn, L. M.; Tatlisumak, T.; Putaala, J.; Gordin, D.; Jokinen, H.; FinnDiane Study Group,
Show abstract
BackgroundAdults with type 1 diabetes (T1D) have an increased risk of developing cerebral small vessel disease (cSVD)-related brain changes already in midlife, yet their significance for cognitive functions remains poorly understood. We investigated the associations between cerebral microbleeds (CMBs), white matter hyperintensities (WMHs), and cognitive functions, including processing speed, executive functions, and episodic memory, in individuals without detected neurological symptoms. MethodsAdults with T1D (n=167; age 46.4{+/-}7.7 years) underwent clinical and biochemical evaluations, brain magnetic resonance imaging (MRI), and a comprehensive neuropsychological assessment. CMB number and topography (lobar, deep or infratentorial, or mixed location) were rated. WMHs were quantified with volumetric analysis. ResultsCompared to absence of CMBs, higher burden of CMBs ([≥]3) was associated independently of age with worse performance in processing speed (stand. {beta}=0.18-0.23, p<0.05) and executive functions (stand. {beta}=0.18-[-0.25], p<0.05), but not with episodic memory. Mild WMHs had no independent relationships with cognition. Compared to strictly lobar or deep or infratentorial CMBs, mixed location of CMBs was more often negatively related to cognitive performance (stand. {beta}=0.20- 0.32, p<0.05). ConclusionsCMBs were related to a subtle, yet systematic impairment in processing speed and executive functions, whereas no such association was observed for WMHs. The results provide insight into the development of early cSVD-related cognitive changes already in midlife and suggest an increased risk of cognitive decline in T1D.
Matching journals
The top 12 journals account for 50% of the predicted probability mass.