| Objective:To explore the systemic factors influencing retinal vessel diameter and its correlation with visceral fat area(VFA)in patients with type 2 diabetes mellitus(T2DM).Methods:The study population consisted of 690 patients diagnosed with T2DM who met the inclusion and exclusion criteria.They were admitted to Chengde Central Hospital between May 2021 and March 2022.General data of the patients were collected,including gender,age,hypertension history,hyperlipidemia history,and diabetes duration.Peripheral venous blood was collected for measurement of total cholesterol(TC),fasting plasma glucose(FPG),triglycerides(TG),high-density lipoprotein cholesterol(HDL-C),glycosylated hemoglobin A1c(Hb A1c),low-density lipoprotein cholesterol(LDL-C)and other indicators.Blood pressure,height and weight were measured.Body mass index(BMI)was calculated.The ocular examination included slit-lamp examination,intraocular pressure measurement,indirect ophthalmoscopy,and fundus photography.Retinal blood vessel diameters are measured through fundus photography using Automated Retinal Image Analyzer(ARIA)software.The central retinal artery equivalent(CRAE)and central retinal venous equivalent(CRVE)were calculated using the modified Parr-Hubbard formula.Arteriole-to-venule ratio(AVR)was calculated.VFA and subcutaneous fat area(SFA)were measured.Visceral to subcutaneous fat area ratio(VSR),visceral fat area/height~2(VFA/height~2),visceral fat area/weight(VFA/weight),and visceral fat area/body mass index(VFA/BMI)were calculated.The distribution of retinal vessel diameter was described in T2DM population.Spearman correlation was used to analyze the correlation between CRVE,CRAE and systemic influencing factors.Multiple linear regression was used to explore the influencing factors of CRVE and CRAE.According to VFA,the patients were divided into two groups:visceral obesity(VO)group(VFA≥100 cm~2,333 cases)and non-VO group(VFA<100 cm~2,357 cases).The differences between the two groups were compared.The Spearman correlation was used to analyze the correlation between visceral fat and retinal vessel diameter.Three multiple linear regression models were used to adjust for different confounding factors to analyze the robustness of the association between visceral fat and retinal vessel diameter.Results:1.Characteristics of the T2DM study population690 patients were included in the analysis.The mean age was 53.5±11.8years.This study included 286 women(41.4%),404 men(58.6%),and diabetic retinopathy(DR)215(31.2%).The average CRAE,CRVE and AVR were 153.76±12.65μm,231.11±19.49μm and 0.67±0.06,respectively.The average VFA,SFA and VSR were 99.89±36.10 cm~2,196.86±61.97 cm~2and0.52±0.15,respectively.The average VFA/height~2,VFA/weight and VFA/BMI were 35.27±12.26 cm~2/m~2,1.35±0.37 cm~2/kg and 3.81±1.10,respectively.The average BMI was 25.80±3.54 kg/m~2.The frequency distribution histograms show that CRAE and CRVE are approximately normal distribution.2.Factors influencing the retinal blood vessel diameter in patients with T2DM2.1 Population distribution of retinal vessel diameter:The CRVE of males was higher than that of females(232.8±18.6μm vs.228.7±20.4μm,P=0.002).The CRVE was higher in smoking patients compared to non-smoking patients(234.1±18.8μm vs.229.2±19.7μm,P=0.002).The CRVE in patients with DR was higher than that in patients without DR(241.0±19.3μm vs.226.6±17.9μm,P<0.001).The CRVE of patients with diabetes course>8 years was higher than that of patients with diabetes course≤8 years(233.2±20.1μm vs.229.1±18.7μm,P=0.007).CRVE increased with the increase of Hb A1c(P=0.046).CRAE stenosis was greater in patients with hypertension than in those without(150.1±11.6μm vs.157.3±12.6μm,P<0.001).CRAE narrowed as systolic blood pressure(P<0.001),diastolic blood pressure(P=0.011),and BMI(P=0.009)increased.2.2 Spearman correlation analysis between retinal vessel diameter and systemic influencing factors:CRVE was associated with female(r=-0.121,P=0.002),smoking(r=0.119,P=0.002),diabetes duration(r=0.111,P=0.004),Hb A1c(r=0.091,P=0.017),and diastolic blood pressure(r=-0.093,P=0.015).CRAE was associated with TG(r=-0.090,P=0.019),systolic blood pressure(r=-0.206,P<0.001),diastolic blood pressure(r=-0.235,P<0.001),and BMI(r=-0.137,P<0.001).2.3 Multiple linear regression analysis of influencing factors of retinal vessel diameter:Systolic blood pressure[per 10 mm Hg increaseβ-0.864,95%confidence interval(CI)-1.476~-0.252,P=0.006],diastolic blood pressure(per 10 mm Hg increaseβ-1.339,95%CI-2.343~-0.335,P=0.009),BMI(per 1 kg/m~2increaseβ-0.273,95%CI-0.545~-0.002,P=0.048)were independent influencing factors for CRAE.Diabetes duration(per 1 year increaseβ0.350,95%CI 0.137~0.564,P=0.001),Hb A1c(per 1%increaseβ0.986,95%CI 0.264~1.707,P=0.008),and diastolic blood pressure(per 10mm Hg increaseβ-1.323,95%CI-2.550~-0.096,P=0.035)were independent factors of CRVE.3.Correlation analysis between retinal vessel diameter and visceral fat in T2DM patients3.1 Baseline characteristics of patients in visceral obesity group and non-visceral obesity group:Compared with the non-VO group,there are more men(70.0%vs.47.9%,P<0.001),hypertension(56.8%vs.41.5%,P<0.001),hyperlipidemia(70.0%vs.54.3%,P<0.001),smoking(46.5%vs.32.2%,P<0.001),alcohol consumption(52.9%vs.35.9%,P<0.001)patients in the VO group.BMI(27.7±3.4 kg/m~2vs.24.0±2.6 kg/m~2,P<0.001)was higher and CRAE was lower(151.9±12.0μm vs.155.5±13.0μm,P<0.001)in the VO group.There were no significant differences in age,diabetes course,Hb A1c,DR And CRVE between the two groups.3.2 Spearman correlation analysis between retinal vessel diameter and visceral fat:In T2DM patients,VFA(r=-0.165,P<0.001),SFA(r=-0.115,P=0.003),VSR(r=-0.092,P=0.015),VFA/height~2(r=-0.162,P<0.001),VFA/weight(r=-0.135,P<0.001)and VFA/BMI(r=-0.141,P<0.001)were negatively correlated with CRAE,but not with CRVE.3.3 Multiple linear regression analysis of retinal vascular diameter and visceral fat:After adjusting for age and gender by Mode1,CRAE narrowed gradually with the increase of BMI[per 1 standard deviation(SD)increaseβ-1.908,95%CI-2.870~-0.946,P<0.001],VFA(per 1 SD increaseβ-2.086,95%CI-3.055~-1.117,P<0.001),SFA(per 1 SD increaseβ-1.486,95%CI-2.443~-0.529,P=0.002),VSR(per 1 SD increaseβ-1.177,95%CI-2.216~-0.138,P=0.027),VFA/height~2(per 1 SD increaseβ-2.000,95%CI-2.938~-1.062,P<0.001),VFA/weight(per 1 SD increaseβ-1.687,95%CI-2.637~-0.737,P=0.001)and VFA/BMI(per 1 SD increaseβ-1.790,95%CI-2.775~-0.806,P<0.001).After adjusting for age,sex,smoking,alcohol consumption,diabetes duration,Hb A1c,systolic blood pressure,and diastolic blood pressure by Mode2,CRAE still narrowed gradually with the increase of BMI(per 1 SD increaseβ-1.197,95%CI-2.174~-0.220,P=0.016),VFA(per 1 SD increaseβ-1.301,95%CI-2.291~-0.311,P=0.010),VFA/height~2(per 1 SD increaseβ-1.275,95%CI-2.226~-0.323,P=0.009),VFA/weight(per 1 SD increaseβ-1.044,95%CI-1.997~-0.092,P=0.032),and VFA/BMI(per 1 SD increaseβ-1.066,95%CI-2.062~-0.070,P=0.036);however,the association between SFA,VSR and CRAE was not statistically significant.The results was similar to Mode2 after adjusting for age,sex,smoking,alcohol consumption,diabetes course,Hb A1c,hypertension,hyperlipidemia and DR by Mode3.This indicates the association between visceral fat and retinal artery diameter was robust.The correlation between VFA and CRAE was stronger than BMI in all three models.However,even after multivariate adjustment,the association between CRVE and BMI,VFA,SFA,VSR,VFA/height~2,VFA/weight and VFA/BMI remained statistically insignificant.Conclusions:1.Systolic blood pressure,diastolic blood pressure and BMI were negatively correlated with CRAE in type 2 diabetic patients.Diabetes course and Hb A1c were positively correlated with CRVE in type 2 diabetic patients.Diastolic blood pressure was negatively correlated with CRVE in type 2diabetic patients.2.CRAE narrowed with the increase of VFA in type 2 diabetic patients,and VFA had a stronger correlation with CRAE than BMI. |