| Background:With a rapid development of economic growth,an increase in life expectancy,and changes in lifestyle,the prevalence of type 2 diabetes mellitus(T2DM)is increasing year by year.The previous International Diabetes Federation(IDF)estimates showed that there were 366 million people with diabetes in 2011,and there were more than 280 million people with prediabetes.The prevalence of diabetes,which is expected to rise to 552 million by 2030,is increasing at an averaged annual growth of 2.7%.Moreover,most people with diabetes live in low-and middle-income countries.The prevalence of DM has increased from 0.67%in 1980 to 11.6%in 2010 among Chinese population.There were about 113.9 million Chinese adults with diabetes and 493.4 million with prediabetes in 2010.T2DM,which do serious harm to the human health,has become the third chronic disease after malignant tumor and cardiovascular disease.Overweight and obesity as another global disease,are closely related to social economy development.There are approximately 1 billion adults with overweight in the whole world,and there are approximately 0.3 billion adults with obesity around the world.However,developing countries are more and more suffering from the dangers of obesity.The Chinese survey data between 2007 and 2008 showed that the prevalence of overweight and obesity among adults was 25.1%and 5%,respectively.The total prevalence of overweight and obesity was 30.1%,the prevalence was much higher than that in 2002.Previous studies proved that overweight and obesity,which was the independent risk factors associated with T2DM,was the important risk factor related to chronic noncommunicable diseases.The risk for presence of TIDM is positively associated with BMI.The prevalence of T2DM in people with obesity was 3-7 times of the prevalence in people with normal body weight.The prevalence of T2DM in people with BMI?35 kg/m2 was about 20 times of the prevalence in people with 18.5 ?BMI ?24.9 kg/m2.There is no doubt that overweight and obesity plays an important role in the increasing prevalence of T2DM.Type 2 diabetes mellitus(T2DM)as a group of metabolic diseases with chronic hyperglycemia as the main clinical features,is caused by insulin secretion disorders and(or)insulin action defects.T2DM is associated with chronic metabolic noncommunicable diseases such as hypertension,dyslipidemia,metabolic syndrome,hyperuricemia,and fatty liver disease etc..They interact with each other,affect the quality of human life seriously,but also bring heavy psychological and economic burdens to the family and society.The new diabetes map of the IDF showed that there were about 4.6 million patients died of diabetes in 2011 around the world,and 1 people died of diabetes on average every 7 seconds.The diabetes medical costs,accounted for about 11%of the world's total health expenses,was up to 46.5 billion dollars in 2011 all over the world.In China,diabetes direct medical costs,which was about 57.469 billion yuan,accounted for 7.57%of the total health expenses,and the ratio had been close to or more than some developed countries(the European Union countries,7.20%;America,10.61%).With the increasing prevalence of diabetes,the economic burden caused by diabetes will be more and more serious.Diabetic retinopathy(DR)is a common chronic microvascular diabetic complication,and it is the leading cause of visual impairment.The American Wisconsin Epidemiologic Study of Diabetic Retinopathy(WESDR)showed that almost all patients with type 1 diabetes mellitus and approximately 60%of patients with T2DM suffered from DR after had diabetes for 20 or more years.Apart from visual morbidity,the presence of DR may indicate microcirculatory dysfunction in other organ systems.Therefore,screening patients with high risk for DR,diagnosing and treating DR timely play important roles in preventing or delaying the development the DR and other diabetic complications.Mydriatic fundus examination,mydriatic fundus digital radiography,and Fundus fluorescence angiography are all effective tools for the screening of DR,and they all require skilled ophthalmologist to carry out the work.Moreover,above methods,which are costly,time-consuming,invasive,not easy to be accepted,are not suitable for large-scale population census.Last but not the least,although the economy of China has made rapid progress in the past few decades,the knowledge of health care remains deficient,and not enough adequate emphasis is placed on regular physical examinations.Therefore,it is valuable to produce a simple,convenient,fast,economic,noninvasive,and receptive DR risk score.Recently,risk score had been widely used in screening chronic Non-communicable diseases.The principle of the risk score is based on the important risk factors of the disease,and the risk score can be completed by the patients themselves.At present,the diabetes risk score(FINDRISC),which can be completed by subjects themselves,contains the Finland diabetes risk score,Dutch diabetes risk score,Danish diabetes risk score,india diabetes risk score,Thailand diabetes risk score and Qingdao diabetes risk score.In addition,studies revealed that the FINDRISC was associated with insulin resistance and the development of T2DM,and it could be used in screening chronic heart disease,stroke,mortality,hepatic steatosis,and metabolic syndrome.However,there is no DR risk score used in screening patients with high risk for DR in patients T2DM.The present study was based on the survey data of subjects with high risk for diabetes in Guangzhou and patients with overweight/obese patients with T2DM in Guangdong province.The main subjective of this study was to investigate the prevalence and risk factors of DR.Moreover,another objective of this study was to analyze the performance of the modified-FINDRISC for screening for DR in subjects with high risk for diabetes.Last but not the least,the important objective of this study was to produce a simple,convenient,fast,economic,noninvasive,and receptive DR risk score,in order to prevent or delay the presence of development of DR.Part ? A cross-sectional survey and a risk score for screening for diabeticretinopathy in a high-risk populationObjective:The objectives of this study were to investigate the prevalence and risk factors of the DR in subjects with high risk for diabetes,and investigate the performance of the risk score for screening for DR.Subjects:A total of 650 subjects,aged over 45 years without known diabetes,were randomly recruited from five communities in Guangzhou between July 2009 and May 2010.Exclusion criteria were known DM,cancer,hepatic failure,renal failure,severe psychiatric disturbance,and any other systemic medical condition.Methods:This study used the Finnish Diabetes Risk Score(FINDRISC)as the basic screening tool.However,this study revised two questions according to the diagnostic cut-off value of the overweight/obese and central obesity regarding Chinese population.In addition,this study added family history of diabetes in the questionnaire.The revised questionnaire was named as modified-FINDRISC.Therefore,the modified-FINDRISC questionnaire included age,body mass index(BMI)and waist circumference,family history of diabetes,history of antihypertensive drug treatment and impaired glucose regulation(IGR),physical activity,and daily consumption of fruits or vegetables.Subjects with modified-FINDRISC?9 were included in this study.A total of 208 included subjects received a screening for DR,a standard oral glucose tolerance test(OGTT),a physical examination(height,weight,waist circumference,hip circumference,and blood pressure),a collection of blood samples to determine the level of fasting plasma glucose(FPG),two-hour postprandial plasma glucose(2hPG)levels,glycated hemoglobin(Hb A1c)levels,high-density lipoprotein cholesterol(HDL-C)concentration,low-density lipoprotein cholesterol(LDL-C)levels,total cholesterol(TC),triglycerides(TG),and a collection of first morning urine to determine the level of urinary albumin.Performance of the diabetic retinopathy risk score was tested using receiver operating characteristic(ROC)curve analysis.Statistical analysis:All analyses were performed using SPSS software version 13.0.Normally distributed and continuous variables were presented as mean ± standard deviation,and non-normally distributed variables were presented as medians(quartiles 25%and 75%).The Independent-Samples t-test was used to examine differences in normally distributed and continuous variables,and the Mann-Whitney U was used to examine differences in non-normally distributed variables.Categorical variables were expressed as percentages,and the ?2 test was used for comparisons of proportions.Binary logistic regression was used to assess the associations between DR and associated risk factors.Odds ratios(OR)and 95%confidence intervals(95%CI)were calculated.Performance of the diabetic retinopathy risk score was tested using receiver operating characteristic(ROC)curve analysis.A value of P<0.05 was considered statistically significant(two-tailed).Results:1.Basic information of subjects with high risk for diabetesAll 619 subjects completed the modified-FINDRISC questionnaire,208 subjects(86 males)with FINDRISC score>9 were included in the study,and the mean age was 69.2±8.5 years.Among 208 subjects with a FINDRISC score ? 9,55 subjects(26.4%)had DM,84 subjects(40.4%)had IGR,and 69 subjects(33.2%)had normal glucose tolerance(NGT).2.Clinical characteristics of subjects with and without DRCompared with subjects without DR,subjects with DR had higher HbA1c,FPG,and 2hPG levels,higher FINDRISC scores,higher urinary albumin levels,higher ratios of microalbuminuria,and history of IGR(P<0.05).The mean age,SBP,DBP,BMI,waist circumference,hip circumference,TC,TG,HDL-cholesterol and LDL-cholesterol were similar among subjects without and with DR(P>0.05).The ratio of gender,the percentage of history of antihypertensive drug treatment,family history of DM,physical activity?4h/week,daily consumption of fruits or vegetables,were also similar among subjects without and with DR(P>0.05).3.Prevalence of DRAmong 208 included subjects with a FINDRISC score ? 9,the total prevalence of DR was 14.9%.The prevalence of DR in males was similar in females(P=0.472).The prevalence of DR was comparable in different age groups(P=0.641).The prevalence of DR increased with increasing modified-FINDRISC(p=0.041).The prevalence of patients with FINDRISC scores?10,? 12,and>12 was 8.2%,20.3%,and 21.2%,respectively.The prevalence of DR was comparable in patients with normal weight,overweight and obesity(P=0.702).The prevalence of DR in subjects with DM was dramatically higher than in subjects with IGR and normal glucose tolerance(P<0.001).Compared with subjects without history of IGR,subjects with history of IGR had higher prevalence of DR(47.1%vs.12.0%,p=0.001).The prevalence of DR was similar in subjects with and without family history of diabetes(18.2%vs.14.7%,p=0.780).The prevalence of DR was similar in subjects with and without history of antihypertensive drug treatment(15.5%vs.14.1%,p=0.780).The prevalence of DR was comparable in patients with and without physical activity? 4h/week(14.5%vs.15.4%,P=0.864).The prevalence of DR was similar in subjects with and without daily consumption of fruits or vegetables(16.2%vs.11.1%,P=0.363).4.Risk factors for DRIn the present study,we carried out the binary logistic analysis to evaluate the risk factors for DR.The presence of DR was a dependent parameter,and age,gender,family history of DM,history of IGR,SBP,DBP,BMI,waist circumference,hip circumference,HbA1c,FPG,2hPG,TC,TG,HDL-cholesterol,LDL-cholesterol,modified-FINDRISC,physical activity,daily consumption of fruits or vegetables,history of antihypertensive drug treatment,and microalbuminuria were independent parameters.In the concise model,the results of the binary logistic analysis showed that history of impaired glucose tolerance(OR,6.493;95%CI:2.278-18.505),modified-FINDRISC(OR,1.188;95%CI:1.019-1.384),HbA1c(OR,3.083;95%CI:1.858-5.118),FPG(OR,1.274;95%CI:1.160-1.398),OGTT 2hPG(OR,1.823;95%CI:1.342-2.477)and microalbuminuria(OR,4.792;95%CI:1.926-11.923)were risk factors associated with DR.In the full model model,the results of the binary logistic analysis showed that history of impaired glucose tolerance(OR,7.194;95%CI:1.083-47.810),HbA1c(OR,2.912;95%CI:1.009-8.402),OGTT 2hPG(OR,1.014;95%CI:1.003-1.025)and microalbuminuria(OR,5.387;95%CI:1.255-23.129)were independent risk factors associated with DR.5.Performance of the modified-FINDRSC for screening for DRThe ROC curve represented the diagnostic accuracy of the modified-FINDRSC for DR.The area under the ROC curve for DR in training samples was 0.626(95%CI 0.524-0.729).And the result partly presented that the modified-FINDRSC may be a reliable tool for screening for DR in patients with high risk for diabetes.Conclusions:In summary,the prevalence of DR was high in a Chinese population at high risk for diabetes.The history of IGR and the presence of microalbuminuria were strong risk factors for DR in the Chinese high-risk population.Risk factors also included higher HbAlc and 2hPG levels.The modified-FINDRISC questionnaire may also be a reliable tool to identify individuals with high risk for diabetes in China,further studies need are needed to clarify these findings.Part ? A cross-sectional survey and a risk score for screening for diabeticretinopathy in overweight/obese patients with type 2 diabetes mellitusObjectives:The objectives of this study were to investigate the prevalence and risk factors of the DR in Chinese overweight/obese patients with T2DM.Furthermore,the important objectives of this study were to develop a simple risk calculator that could be conveniently used in primary care and also by patients themselves,in order to provide preliminary theoretical and clinical basis in developing DR risk score.Subjects:This was a hospital-based multicentric cross-sectional study in Guangdong Province.In each hospital,patients were enrolled according to all of the following criteria between August 2011 and March 2012:1)Chinese T2DM patients aged over 20 years had lived in Guangdong Province for>1 year,2)body mass index(BMI)?25 Kg/m2,3)waist circumference was measured,4)history of DR,history of antihypertensive drug treatment,physical activity,diet control,and date of diabetes diagnosis were recorded.Subjects with cancer,pregnancy,severe psychiatric disturbance,hepatic failure and end-stage renal failure were excluded.Methods:If the patient had a previous ophthalmologist diagnosis of DR,the grading of DR was already recorded by the ophthalmologist.Doctors in each hospital checked the medical record about DR and recorded the grading of DR.Retinopathy grading was based on the results of the worst eye,and the retinopathy severity score was assigned according to the International Clinical DR Disease Severity Scale.Because the aims of the present study were to produce a simple risk calculator that could be conveniently used in primary care and also by patients themselves,only parameters that are easy to assess without any clinical measurements requiring special skills or any laboratory tests were considered into the model.Therefore,age,duration of DM,hypertensive drug treatment,BMI,waist circumference,physical inactivity,and diet control were considered as candidate risk factors.Candidate risk factors which were statistically significant(P<0.05)in the concise model were further fitted into the full logistic regression model using the enter method.1869 patients(approximately 70%)were selected by using systematic random sampling method to form training samples,and the remaining 830 patients(approximately 30%)were used as testing samples.In the training samples,beta coefficients derived from the full logistic regression were used to calculate the diabetic retinopathy risk score.A score for each variable in the full model was calculated by multiplying the ?-coefficient by 10.A sum score was calculated for every patient by adding the score for each variable.Cross-validation was used to validate the method for establishment of the diabetic retinopathy risk score.Performance of the diabetic retinopathy risk score was tested using receiver operating characteristic(ROC)curve analysis.Based on the ROC analysis of the training samples,the best cutoff value of the diabetic retinopathy risk score was determined from the highest Youden's Index,which is defined as follows:(sensitivity+specificity-1).Using the diabetic retinopathy risk score cut-off value of 20 points to identify DR,if the sensitivity and specificity in testing samples was similar as that in training samples,the method for establishment of the DR risk score was correct.Statistical analysis:All analyses were performed using spss software version 13.0.Normally distributed and continuous variables were presented as mean ± standard deviation,and non-normally distributed variables were presented as medians(quartiles 25%and 75%).The Independent-Samples t-test was used to examine differences in normally distributed and continuous variables,and the Mann-Whitney U was used to examine differences in non-normally distributed variables.Categorical variables were expressed as percentages,and the ?2 test was used for comparisons of proportions.The ranked data was analyzed by K Independent-Samples t-test.A value of P<0.05 was considered statistically significant(two-tailed).Binary logistic regression was used to assess the associations between DR and associated risk factors.Odds ratios(OR)and 95%confidence intervals(95%CI)were calculated.Performance of the diabetic retinopathy risk score was tested using receiver operating characteristic(ROC)curve analysis.Based on the ROC analysis,the best cutoff value of the diabetic retinopathy risk score was determined from the highest Youden's Index,which is defined as follows:(sensitivity+specificity-1).Results:1.Clinical characteristics of the overweight or obese patient with T2DMThe evaluated 2699 patients included 1263 males and 1436 females,with an average age of 59.4±13.0 years(20-90 years).Compared with patients who did not develop DR,patients with DR had older age,higher waist circumference,higher HbAlc,longer duration of DM,higher systolic blood pressure,and higher TC(P<0.05).The frequency of females,hypertension,hypertensive drug treatment,and family history of diabetes were significantly higher in patients with DR than patients without DR(P<0.05).BMI,TG,HDL-cholesterol,LDL-cholesterol,percentage of regular physical exercise and diet control,however,were similar among patients without and with DR(P>0.05).2.Development of the DR risk scoreBinary logistic regression analysis was performed to estimate the strength of the association of candidate factors to the presence of DR.Age,waist circumference,history of antihypertensive drug treatment,and duration of DM,which were significantly(P<0.05)associated with the presence of DR in the concise model,were further fitted into the full logistic regression model using the enter method.Based on the beta coefficient of the full model,the risk score was constructed by multiplying the beta coefficients by 10 and rounding to the nearest integer.The risk score for patients with 20 ? age ? 44,45 ? age ? 64,age? 65 was 0,5,4,respectively.The risk score for patients without central obesity and with,central obesity was 0 and 2,respectively.The risk score for patients with diabetic duration?1,1<diabetic duration?5,5<diabetic duration?10,10<diabetic duration?15,15<diabetic duration was 0,9,10,14,and 20,respectively.The risk score for patients without history of antihypertensive drug treatment and with history of antihypertensive drug treatment was 0 and 7,respectively.A sum score was calculated for every patient by adding the score for each variable.3.Performance of DR risk score and the risk score cut-off value of DRThe ROC curve represented the diagnostic accuracy of the diabetic retinopathy risk score for DR.The area under the ROC curve for DR in training samples was 0.700(95%CI 0.671-0.729).The area under the ROC curve for DR intesting samples was 0.697(95%CI 0.654-0.741).Comparing the Youden's Index of different values,the optimal cutoff point to predict DR in training samples was 20.Using the diabetic retinopathy risk score cut-off value of 20 points to identify DR,the sensitivity and specificity in testing samples(73.1%and 52.2%)was similar as that in training samples(77.2%and 55.1%).And among 167 patients with DR deriving from testing samples,74.9%(125)of patients with diabetic retinopathy risk score>20 were diagnosed as DR.4.Prevalence of DR19.7%(531)of the 2699 patients were diagnosed with DR.The prevalence of DR significantly increased with the increasing of the diabetic duration(P<0.001).And the prevalence of the mild non-proliferative DR,moderate non-proliferative DR,severe non-proliferative DR,and proliferative DR all significantly increased with the diabetic retinopathy risk score(P<0.001).The prevalence was significantly higher in females than in males(22.8%vs.16.2%,P<0.001).The prevalence of DR significantly increased with increasing age(P<0.001).And the prevalence of DR increased with increasing duration of DM(P<0.001).The prevalence of DR in patients with history of antihypertensive drug treatment was dramatically higher than in patients without history of antihypertensive drug treatment(25.5%vs.11.9%,P<0.001).The prevalence of DR in patients with HbAlc<7%was dramatically higher than in patients with HbAlc>7%(21.9%vs.16.3%,P=0.003).However,the prevalence of DR was comparable in overweight patients and obese patients(19.4%vs.21.1%,P=0.392),the prevalence of DR was similar in patients with and without central obesity(20.0%vs.16.8%,P=0.187).5.Risk factors for DRIn the present study,we carried out the binary logistic analysis to evaluate the risk factors for DR.The presence of DR was a dependent parameter,and HbAlc,FPG,TC,TG,HDL-cholesterol,LDL-cholesterol and DR risk score were independent parameters.In the concise model,the results of the binary logistic analysis showed that HbA1c(OR=1.046;95%CI:1.002,1.092),TC(OR=1.062;95%CI:1.001,1.126)and DR risk score(OR=1.104;95%CI:1.089,1.120)were risk factors associated with DR.In the full model,after adjusted for other variables,DR risk score(OR=1.102;95%CI:1.084,1.119)and HbAlc(OR=1.078;95%CI:1.019,1.141)were independent risk factors associated with DR.Conclusions:In conclusion,the prevalence of DR in overweight/obese patients with T2DM was 19.7%.Age,duration of DM,hyperglycemia,history of antihypertensive drug treatment and central obesity were risk factors associated with the presence of DR.The diabetic retinopathy risk score,which was constructed by age,duration of DM,history of antihypertensive drug treatment,and waist circumference,had reasonable performance for screening for DR in Chinese overweight/obese patients with T2DM. |
PDF Full Text Request