Meta-analysis of Alzheimer's disease risk with obesity, diabetes, and related disorders.

Late-onset Alzheimer's disease (AD) is a multifactorial and heterogeneous disorder with major risk factors including advanced age, presence of an apolipoprotein E epsilon4 (APOE4) allele, and family history of AD. Other risk factors may be obesity and diabetes and related disorders, which are highly prevalent. We reviewed longitudinal epidemiological studies of body mass, diabetes, metabolic syndrome, and glucose and insulin levels on risk for AD. We conducted meta-analyses of the results from these studies. For obesity assessed by body mass index, the pooled effect size for AD was 1.59 (95% confidence interval [CI] 1.02-2.5; z = 2.0; p = .042), and for diabetes, the pooled effect size for AD was 1.54 (95% CI 1.33-1.79; z = 5.7; p < .001). Egger's test did not find significant evidence for publication bias in the meta-analysis for obesity (t = -1.4, p = .21) or for diabetes (t = -.86, p = .42). Since these disorders are highly comorbid, we conducted a meta-analysis combining all studies of obesity, diabetes, and abnormal glucose or insulin levels, which yielded a highly significant pooled effect size for AD of 1.63 (95% CI 1.39-1.92; z = 5.9; p < .001). Obesity and diabetes significantly and independently increase risk for AD. Though the level of risk is less than that with the APOE4 allele, the high prevalence of these disorders may result in substantial increases in future incidence of AD. Physiological changes common to obesity and diabetes plausibly promote AD.