Research Article

Cross Talk between Lipid Metabolism and Inflammatory Markers in Patients with Diabetic Retinopathy

Table 3

Results of the multivariate multinomial logistic regression analysis. Bold cells denote statistically significant associations.

VariableCategory/incrementPDR/non-CSME versus ref.NPDR/CSME versus ref.PDR/CSME versus ref.
RR (95% CI)p valueRR (95% CI)p valueRR (95% CI)p value

Core model: clinical variables
 Male sexMale versus female2.08 (0.85–5.09) 0.1061.54 (0.72–3.28) 0.2630.30 (0.08–1.11) 0.072
 Duration of diabetesOne-year increase1.10 (1.04–1.16) 0.0011.04 (0.99–1.09) 0.1661.09 (1.01–1.18) 0.037
 HypertensionYes versus no2.91 (1.25–6.80) 0.0133.52 (1.65–7.48) 0.0016.83 (1.70–27.39) 0.007
Serum parameters alternatively introduced to the model
 Apo-B 0.1 g/L increase1.20 (1.06–1.36)0.0031.27 (1.14–1.42)<0.0011.20 (1.01–1.43)0.039
 Apo-B/Apo-A0.1 increase1.18 (1.01–1.38)0.0381.24 (1.08–1.42)0.0021.25 (0.99–1.59)0.059
 VEGF 100 pg/mL increase1.15 (0.99–1.35)0.0721.17 (1.02–1.35)0.0261.24 (1.02–1.50)0.034
 TNF-alpha 10 pg/mL increase1.59 (1.01–2.50)0.0461.68 (1.10–2.56)0.0152.07 (1.20–3.59)0.009

PDR: proliferative diabetic retinopathy; CSME: clinically significant macular edema; RR: relative ratio; CI: confidence interval; VEGF: vascular endothelial growth factor.
ref.: nonproliferative/non-CSME patients, set as reference category; †: adjusted for the parameters included in the core model (male sex, duration of disease, and hypertension).