Cerebral microbleeds are associated with slowed processing speed in middle-aged adults with type 1 diabetes

Introduction

Type 1 diabetes (T1D) is an important risk factor for cerebral small vessel disease (SVD). Cerebral microbleeds (CMB) and white matter hyperintensities (WMH) can already be observed in early midlife in individuals with T1D, even though generally, this pathology is seen decades later in those without diabetes. Whether these changes affect cognitive functions, remains unclear. We investigated the associations of CMB and WMH with processing speed and attention in adults with T1D without major neurological symptoms.

Methods

As part of an ongoing FinnDiane sub-study, brain magnetic resonance imaging, and extensive clinical and neuropsychological examinations were done for 142 participants with T1D (age 47.2±7.6 years, diabetes duration 31.6±11.0 years). CMB and WMH were evaluated visually by an experienced neuroradiologist and categorized according to number and severity (CMB: 0 vs 1-2 vs ≥3; WMH: Fazekas score 0 vs ≥1). The assessment of processing speed and attention included the Wechsler Adult Intelligence Scale-IV (WAIS-IV) Coding and Symbol search subtests, the Stroop test, and the computerized Flexible Attention Test (FAT) Simple Visuomotor Speed, Numbers and Number-Letter subtasks.

Results

CMB (≥3) and mild WMH were associated with poorer performance in WAIS-IV Coding and Symbol Search, and Stroop and FAT subtasks (p<0.05) in univariate linear regression analyses. After controlling for age and years of education, the associations of CMB with Coding (stand. β=-0.17, p=0.038), FAT Simple Visuomotor Speed (stand. β=0.16, p=0.048) and Stroop color- incongruent part (stand. β=0.18, p=0.038) remained significant, whereas WMH were no longer related to cognitive performance. When CMB and WMH were entered together in multiple linear regression models adjusted for age and education, CMB had independent negative contributions to Coding (stand. β=-0.17, p=0.045). The effect sizes of the significant associations were small (Cohen's f2=0.04) (Fig. 1).

Discussion

CMB were related to a subtle decline in processing speed and attention in middle-aged adults with T1D. The associations were only partly explained by age. Effect sizes on a group level were small indicating minor clinical significance. However, our results provide insight into the development of SVD-related cognitive changes already in midlife and suggest an increased risk of cognitive decline in individuals with T1D.