“…The meta-regression of confounding factors showed that age (p < 0.01, Adj R-squared = 92.41%; Figure 3A) and CVD (p = 0.01, Adj R-squared = 85.21%; Supplementary Figure 1) exerted an effect on the association of LDL-c with AD; whereas other parameters including BMI (p = 0.063, Adj R-squared = −6.53%; Figure 3B), education (p = 0.50, Adj R-squared = −11.58%; Figure 3C), smoking (p = 0.10, Adj R-squared = 43.90%; Figure 3D), hypertension (p = 0.98, Adj R-squared = −22.11%; Figure 3E) and diabetes mellitus (p = 0.57, Adj R-squared = −13.04%; Figure 3F) had no impact on the outcomes. As shown in Table 2, we found no statistic differences of the pooled weighted characteristics on male gender (OR = 0.86, 95% CI 0.71∼1.04, p = 0.11), education (SMD = 0.26, 95% CI−0.78∼1.30, p < 0.63), smoking (OR = 1.33, 95% CI 0.71∼2.47, p = 0.38), hypertension (OR = 0.91, 95% CI 0.62∼1.35, p = 0.64), diabetes mellitus (OR = 1.02, 95% CI (Caramelli et al, 1999;Paragh et al, 2002;Reitz et al, 2004;Kouzuki et al, 2018;Chen et al, 2019) 340/375 Quartile1 60-70 0.80 (0.23, 1.37) < 0.01 7 (Solfrizzi et al, 2002;Cacabelos et al, 2003;Panza et al, 2003;Wehr et al, 2006;Yavuz et al, 2008;Macesic et al, 2017;Shafagoj et al, 2018) Sample size (n) 11 (Lehtonen and Luutonen, 1986;Caramelli et al, 1999;Lesser et al, 2001;Paragh et al, 2002;Ryglewicz et al, 2002;Solfrizzi et al, 2002;Panza et al, 2003;Wolf et al, 2004;Mamo et al, 2008;Kouzuki et al, 2018;Shafagoj et al, 2018) 0.82∼1.26, p = 0.88) and CVD (OR = 1.28, 95% CI 0.61∼2.70, p = 0.51) between AD and non-dementia controls; whilst there was a positive correlation of age (SMD = 0.62, 95% CI 0.28∼0.95, p < 0.01) and a i...…”