Severe hypoglycaemia, mild cognitive impairment, dementia and brain volumes in older adults with type 2 diabetes: the Atherosclerosis Risk in Communities (ARIC) cohort study.

AIMS/HYPOTHESIS: We aimed to evaluate the link between severe hypoglycaemia and domain-specific cognitive decline, smaller brain volumes and dementia in adults with type 2 diabetes, which so far has been relatively poorly characterised.

METHODS: We included participants with diagnosed diabetes from the community-based Atherosclerosis Risk in Communities (ARIC) study. At the participants' fifth study visit (2011-2013), we examined the cross-sectional associations of severe hypoglycaemia with cognitive status, brain volumes and prior 15 year cognitive decline. We also conducted a prospective survival analysis of incident dementia from baseline, visit 4 (1996-1998), to 31 December 2013. Severe hypoglycaemia was identified, using ICD-9 codes, from hospitalisations, emergency department visits and ambulance records. Prior cognitive decline was defined as change in neuropsychological test scores from visit 4 (1996-1998) to visit 5 (2011-2013). At visit 5, a subset of participants underwent brain MRIs. Analyses were adjusted for demographics, APOE genotype, use of diabetes medication, duration of diabetes and glycaemic control.

RESULTS: Among 2001 participants with diabetes at visit 5 (mean age 76 years), a history of severe hypoglycaemia (3.1% of participants) was associated with dementia (vs normal cognitive status): OR 2.34 (95% CI 1.04, 5.27). In the subset of participants who had undergone brain MRI (n = 580), hypoglycaemia was associated with smaller total brain volume (-0.308 SD, 95% CI -0.612, -0.004). Hypoglycaemia was nominally associated with a 15 year cognitive change (-0.14 SD, 95% CI -0.34, 0.06). In prospective analysis (n = 1263), hypoglycaemia was strongly associated with incident dementia (HR 2.54, 95% CI 1.78, 3.63).

CONCLUSIONS/INTERPRETATION: Our results demonstrate a strong link between severe hypoglycaemia and poor cognitive outcomes, suggesting a need for discussion of appropriate diabetes treatments for high-risk older adults.