<b>The causal effect of the relative macronutrient intake on diabetic nephropathy: Mendelian randomization, LDSC, and mediator analysis</b>

LDSC analysis showed that the relative intake of protein had genetic correction with DN (r_g=0.341,<i> p</i>=9.07E-05). The MR results showed that relatively higher protein intake could decrease the risk of DN (OR=0.233, 95%CI=0.082~0.666), and this causation was confirmed through MVMR after adjusting for model 1 (OR=0.150, 95%CI=0.061~0.367) and model 2 (OR=0.248, 95%CI=0.091~0.676). A mediated proportion of 93.770% was found between relative protein intake and DN via kidney function. Our results did not find evidence that the relative intake of fat and carbohydrate could affect the progression of DN.