Higher glucose parameters at baseline are related to hippocampal volume decline after a 2‐year multidomain lifestyle intervention to prevent cognitive decline in people at risk of developing Alzheimer’s disease: Exploratory results from the FINGER study

Abstract

AbstractBackgroundLifestyle interventions for preventing cognitive decline have provided promising results in older at‐risk individuals. Two modifiable risk factors are insulin resistance and diabetes mellitus, since they’ve been associated with higher risk of cognitive decline. However, it remains unclear if healthy individuals with higher normal glucose levels are free from suffering the negative effect of pathological glucose values on brain and cognition. The Oral Glucose Tolerance Test (OGTT) is a sensitive measurement for non‐detected glucose metabolism disorders and might be related with changes in brain structure and cognition. In this study we hypothesised that changes in glucose parameters will be associated with brain structural changes.Method1.259 participants of the Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability (FINGER) were included [Table 1]. A sample of 132 participants had brain MRI scans, and 47 also underwent PiB‐PET and FDG‐PET scans. Global GMv, hippocampal volumes, WMHv and cortical thickness measurements were analysed to assess the association between brain structural changes and glucose homeostasis. AD‐signature composites were also calculated from FDG‐PET and PiB‐PET scans data for the analysis. SPSS was used for linear regression models.ResultsBaseline AUC‐OGTT, 120min‐OGTT and fasting glucose were negatively associated (p<0.05) with 2‐year change in hippocampal volume [Table 2]. A negative association was also found between baseline AUC and change in FDG‐PET‐AD‐composite (p = 0.028). Three negative associations were additionally found at baseline between AUC‐OGTT and cortical thickness (p = 0.015), fasting glucose and total GMv (p = 0.015), and HbA1c and FDG‐PET‐AD‐composite (p = 0.033) [Table 3]. Interestingly, a positive association was found in a smaller sample (n = 38) between the TyG index and change in PiB‐PET‐AD‐composite (0.014). No associations were found between glucose parameters and the remaining imaging data.ConclusionIn a sample of participants at risk for dementia, baseline higher measures of peripheral glucose metabolism such as AUC‐OGTT and 120min‐OGTT glucose were related to greater decline in hippocampal volume, irrespective of the randomisation group and after considering abnormal glucose metabolism. These results suggest that glucose homeostasis might have an impact on brain structure in participants with higher normal glucose parameters measurements.