Background
Continuous hyperglycaemia has been related with dementia. However, it remains unclear whether prediabetes poses a higher risk of dementia. A meta‐analysis was therefore conducted to comprehensively investigate the possible role of prediabetes as a risk factor of dementia.
Methods
Prospective cohort studies reporting the association of prediabetes and dementia were identified from PubMed, Web of Science, and Embase databases. A random‐effects model was applied to combine the results by incorporating the influence of heterogeneity. Subgroup analyses were also conducted to explore the influences of study features on the relationship. Sensitivity analysis re‐estimated the combined effect size after excluding single studies separately to explore the robustness of the results.
Results
Nine studies involving 29 986 adults from the general population, 6265 (20.9%) of whom had prediabetes, were included. It was shown that prediabetes was not independently associated with a higher incidence of dementia compared with normoglycaemia (adjusted risk ratio (RR): 1.01, 95% confidence interval (CI): 0.85–1.21, P = 0.89, I2 = 39%). Subgroup analyses according to the definitions of prediabetes, follow‐up duration, method for diagnosis of dementia, and quality score produced similar findings (P for all subgroup differences >0.05). In addition, prediabetes was not independently associated with the incidence of Alzheimer's disease (RR: 1.24, 95% CI: 0.98–1.56, P = 0.07, I2 = 0%) or vascular dementia (RR: 1.16, 95% CI: 0.70–1.92, P = 0.56, I2 = 0%). Different definitions of prediabetes have the potential to influence the results, as reflected in the subgroup analysis for Alzheimer's disease (RR: 1.30, 95% CI: 1.06–1.60, P = 0.01, I2 = 0%).
Conclusions
Prediabetes may not be an independent risk factor of all‐cause dementia or vascular dementia in the general adult population. However, changing the definition of prediabetes may have an impact on the outcome for Alzheimer's disease.