Multiple Factors Analysis On Diabetic Macular Edema

ObjectDiabetic macular edema (DME), which involves retinal thickening in the macular area, is the leading cause of blindness and vision impairment. DME occurs after breakdown of the blood-retinal barrier because of leakage of dilated hyperpermeable capillaries and microaneurysms. Impairment in the visual acuity of diabetic patients with macular edema can undeniably be decreased with systemic and ocular therapeutic intervention at early stages , as shown by numerous controlled studies, but will require a renewed emphasis on early detection. As an implementation of effective technique, OCT is extremely sensitive for detecting early diabetic macular edema, comparing previous methods of examining which fail to identify a sizable number of the macular thickness. Early detection and intervention of DME is critical in preventing visual loss, so we evaluated macular thickness among individuals with diabetes in Beixinjing District by the use of Optical Coherence Tomography (OCT) in diagnosis of macular edema, we investigated the multiple factors of DME to access the control of diabetes-associated metabolic abnormalities which may slow the progression of DME. Outcomes in our understanding of systemic risk factorsof DME will lead to effective treatments that offer the potential for further reduction of visual loss caused by diabetic macular edema. Furthermore, Our study may assist the physician and ophthalmologist in providing patients with the optimal diabetic care available.MethodsThe community based study was done in Beixinjing District, Shanghai. A representative data of 795 diabetic patients, underwent an epidemiologic study of a routine ocular examination between 2003-2004, was studied as observational case series, 128 patients at study entry were chosen to take further ophthalmic examinations. Exclusion criteria were cataract and glaucoma . Participants 128 patients were performed by OCT, diagnosed as DME and graded according to the macular thickness records(normal^ 200nm;mild DME 201~300um;moderate DME301~400um;severe DME>400 inn). Retinal photographs were obtained with a digital non-mydriasis fundus camera, the presence and severity of diabetic retinopathy diagnosis were graded at 2003 International Clinical Classification of DRP. All 128 participants underwent a standardized interview, further examinations and laboratory investigations, such as episode age, body mass index (BMI), WHR, blood pressure, HbAlc, fasting plasma glucose(FPG), postprandial 2 hours plasma glucose(2hPG), serum total protein(TP), albumin(A), A/G, serum creatinine (Scr), blood urea nitrogen (BUN), urea acid(UA),24h urine protein excretion(24hUPE), 24h urine albumin excretion(24hUAE), 24h urine NAG, 24h urine glucose, total cholesterol(TC), high-density lipoprotein cholesterol (HDL), low-density lipoprotein cholesterol (LDL), apolipoprotein Al(ApoAl), apolipoprotein B(ApoB), apolipoprotein E(ApoE),triglyceride(TG). Inputting all records, SPSS11.0 is used to analyze variables statistically.ResultsIn 128 patients 248 eyes , macular thickness of the foveal was examed with OCT . In this population with diabetes, the median thickness of the foveal was 164inn (53.50urn) .1. Multiple factors analysis on progression of DME by bivariate correlation and Partial correlation: No Controlling Variables, we found the macular thickness has significant correlation with age (Pearson coefficient r=0. 237), the duration of DM(0. 369), TG(0. 521), lipoprotein a(0.534) , BUN(0. 266), Scr(0.284), UA(0.280), 24h urine albumin excretion(0.230), high-density lipoprotein cholesterol (-0.247), P <0 .01, 2-tailed Significance;and related with total cholesterol(0. 174), 24h urine protein excretion (0. 176), P <0 .05, 2-tailed Significance. In model controlling for DR, the macular thickness was related with lipoprotein a (Pearson Coefficient r=0. 537), TG(0.452), UA(0.337), Scr (0.332), BUN(0.252), the duration (0.213), age(0. 246), high-density lipoprotein cholesterol (-0.188), P <0 .01, 2-tailed Significance.2. Multiple stepwise regression of macular thickness: ApoE, 24h urine albumin excretion entered the multiple stepwise regression model of M= 4. 562+0. 001[LP(a)]+0.106(TG)+0. 003(Scr)+0. 006(duration). The standardized discriminant function coefficients of them four are 0.349, 0.329, 0.167, 0.146(P <0 .01, 2-tailed Significance ).3. Discriminant analysis of DME: Three variables, LP(a),TG and duration of DM .entered discriminant analysis. The Standardized discriminant function coefficients of them are 0. 635 , 0. 520, 0. 378. Discriminant model is D = 0. 006[LP(a)] + 0. 630(TG) +0. 056(Duration)-3. 374. Canonical discriminant functions evaluated at group means ,group centroids of DME is -0. 646, not DME is 1. 152. Percent of grouped cases correctly classified 84.4%4. Multiple factors analysis on progression of diabetic retinopathy by Spearman' s Test: In different progress stages, seven factors had statistically significant relationship with DR. The factors included the duration (Pearson Coefficient r=0. 549), TG(0.271), ApoE(0.231), 24h urine protein excretion (0.437), 24h urine albumin excretion (0. 674), HDL(-0.197), P<0 . 01, 2-tailed Significance;and HbAlc (0. 218), P <0. 05, 2-tailed Significance.5. Multiple comparison of variables : we created five categories of each variables based on DR grade, The mean between groups is statistically significant different in 24h urine albumin excretion,24h urine protein excretion, TG, duration of DM, (P<0. 001);LP(a)>ApoE(P<0. 01);HbAlc, 24h urine NAG (P<0.05) .6. In the logistic analysis on DR: By Stepwise, ApoE, 24h urine albumin excretion entered the logistic regression model of Y= — 3. 993+0. 056(24h urine albumin excretion)+0. 068(ApoE), P <0 .001, (2-tailed Significance).7. In the logistic analysis on PDR: By Stepwise, four variables, including the duration , 24h urine albumin excretion ,TG, entered the logistic regression model of Y' =-8. 194+0. 152(duration) + 0. 010(24h urine albumin excretion) + 0. 983(TG), (P <0 .01, 2-tailed Significance ).8. The correlation between DME and DR: The incidence of diabetic macular edema increased, and the severity of DME aggravated, with the aggravation of diabetic retinopathy. The multiple liner regression model and curve estimation: M=5. 061 (±0.024)+ 0.077 (±0. 013)DR, (P <0 .001 ),but the R-square of the curve is low. The development of Diabetic macular edema and its classification cannot be brought into definite correspondence or unification with the classification of diabetic retinopathy.Conclusion1. Significant independent risk factors of diabetic macular edema were lipoprotein(a), triglyceride , longer duration of diabetes and serum creatinine. Both the multiple regression model and discriminant model could predict development and progression of DME. LP(a)> TG, were confirmed as independent predictors associated with DME progression. With the elevation of LP(aK TGn duration and Scr, the incidence of DME increased, and the severity of DME aggravated.2. ApoE, 24h urine albumin excretion were confirmed as independent risk factors associated with DR progression. Longer duration, 24h urine albumin excretion and TG were confirmed as independent risk factors associated with PDR progression. Both ApoE and 24h urine albumin excretion are predictors of development and progression of DR. Higher 24h urine albumin excretion is associated with development of diabetic microvascular complications.3. Despite intensive study, current understanding of the pathogenesis of DME and DR remains incomplete, we inferred that DME and DR share several metabolic mechanisms in common, as well as develop in their respective biochemical mechanisms.4. Dyslipidemia has an effect on both DME and DR. Higher serum lipids are associated with increased risk of DME or DR. Our studies have suggested that elevated TG levels maybe an additional common risk factor for DME and DR,. LP(a) is associated with the development mechanisms of DME , but not with DR. ApoE is confirmed to be associated with the development of DR , but not with DME. More monitoring for serum lipid levels should be done for DME and DR patients. More positive treatment and control of the metabolic abnormalities of diabetes is necessary for diabetic patients combined with DME or not.5. 24h urine albumin excretion is the independent risk factor of DR,however, relationships of 24h urine albumin excretion with progression and development of DME were weaker and not significant after adjustment for DR. We inferred that urine albumin excretion made an indirect effect on DME secondarily to DR.6. Relationships between DME and HbAlC were not significant;the progression and development of DR is associated with higher HbAlC.7. At any time during the progression of DR, even with no DR, patients with diabetes can develop DME. The incidence of diabetic macular edema increased, and the severity of DME aggravated, with the aggravation of diabetic retinopathy. Nevertheless, whereas the grade of diabetic macular edema and its classification cannot be brought into definite correspondence or unification with the classification of diabetic retinopathy, hence the typing of DME in another individual classification in DR is necessary.