Abstract: Objective To explore the changes and combined predictive value of serum apolipoprotein A1/apolipoprotein B(ApoA1/B), homocysteine(Hcy), and lipoprotein(a)［Lp(a)］ in type 2 diabetes mellitus(T2DM) patients with microvascular complications(MVC). 
Methods A total of 134 T2DM patients were selected. According to presence of combined MVC, they were divided into MVC group(n=56) and non-MVC group(n=78). The clinical data, serum ApoA1/B, Hcy, and Lp(a) levels were compared between two groups, and the correlation between serum ApoA1/B, Hcy, and Lp(a) and laboratory indicators was analyzed by Pearson correlation analysis. Logistic regression was used to analyze the influencing factors of T2DM with MVC, and receiver operating characteristic(ROC) curve was used to evaluate the single and combined predictive value of serum ApoA1/B, Hcy, and Lp(a) on T2DM with MVC. 
Results Serum ApoA1/B in MVC group was lower than that in non-MVC group, and Hcy and Lp(a) were higher than those in non-MVC group(P＜0.05). Serum ApoA1/B was negatively correlated with glycosylated hemoglobin(HbAlc), fasting blood-glucose(FBG), 2 h postprandial plasma glucose(2 hPG), serum totalcholesterol(TC), and triglyceride(TG), and Hcy and Lp(a) were positively correlated with FBG, 2 h PG, HbA1c, TC, and TG(P＜0.05). Course of disease, body mass index, combined hypertension, serum ApoA1/B, Hcy, and Lp(a) were influencing factors of T2DM with MVC(P＜0.05). The AUC of combined serum ApoA1/B, Hcy, and Lp(a) in predicting T2DM with MVC was 0.878, the sensitivity was 73.21%, and the specificity was 88.46%, which was better than the individual prediction of each index(P＜0.05). 
Conclusion The expression level of serum ApoA1/B in T2DM patients with MVC is significantly reduced, and the expression levels of Hcy and Lp(a) are significantly increased. Detecting changes in their level can help predict the occurrence of MVC in patients with T2DM in clinical practice, and provide an effective reference for patient condition assessment, prevention and treatment. 

Key words: diabetes mellitus, type 2, microvascular complications, predictive value