Objective  To evaluate the correlation of apolipoprotein E (ApoE) gene polymorphisms and Type 2 diabetes mellitus (T2DM) in Chinese population by the method of evidence based medicine.  Methods  Relative literatures on Chinese population about T2DM vs normal glucose tolerance (NGT) were fully searched and screened through computed databases

such Wanfang and VIP

and Medline. The data were extracted for heterogeneity

analysis. The Meta-analysis was employed for summarizing all the studies

calculating the pooled

OR

and its 95% confidence interval (95%

CI

)

and testing the overall effects. Both publication bias analysis by Egger’s publication bias analysis and the fail-safe number for

P

=0.05 (Nfs0.05)

and sensitivity analysis by different effect model and Hardy-Weinberg equilibrium (HWE) genetic principle were performed for evaluating the reliability and stability of the Meta-analysis.  Results  Twenty-two literatures

total 4 452 cases

were obtained. Remarkable heterogeneities were discovered in the allele E3 and genotype E3/E3 of ApoE between T2DM and NGT subjects in these studies (

P

<

0.10). The allele E2

E4 and genotype E2/E3

E3/E4 of ApoE were more frequently in T2DM than NGT in Chinese population. Their

ORs

and 95%

CIs

were 1.48 (1.25

1.75)

1.26 (1.08

1.47)

1.53 (1.26

1.87) and 1.32 (1.10

1.58)

respectively (

P

<

0.01). The allele E3 and genotype E3/E3 of ApoE were found less frequently in T2DM than NGT in these populations. Their

ORs

and 95%

CIs

were 0.72 (0.62

0.83) and 0.66 (0.58

0.79)

respectively (

P

<

0.01). However

E2/E2

E2/E4 and E4/E4 of ApoE gene were no difference between T2DM and NGT subjects. The results of sensitivity analysis and publication bias analysis accounted for the reliability and stability of this Meta-analysis.  Conclusion  ApoE gene polymorphisms are associated with T2DM in Chinese populations. The allele E2 and E4 of ApoE are risk factors for T2DM

and the genotype E3/E3 is protective for T2DM.