Epidemiology Of Diabetes Interventions And Complications Research Articles

Ultra-Widefield and Early Treatment Diabetic Retinopathy Study 7-Field Grading of Diabetic Retinopathy

High concordance in diabetic retinopathy (DR) outcomes between 7-field (7F) and ultra-widefield (UWF) images would allow for combining longitudinal assessments based on the 2 modalities both in clinical studies and clinical care. To compare 7F and UWF imaging with regard to DR severity and the associations of DR severity with risk factors, such as hemoglobin A1c, age, diabetes duration, and sex. This cross-sectional study describes the outcomes of the randomized clinical Diabetes Control and Complications Trial (DCCT) and its subsequent observational study, the Epidemiology of Diabetes Interventions and Complications (EDIC) study. Of the 1441 participants with type 1 diabetes in the DCCT, 1375 were enrolled in the EDIC study. Of the 1171 participants who were active between March 2019 and December 2021, 200° UWF color imaging and 7F fundus photographs were obtained for 785 participants once at the same visit. Central graders assessed 7F-UWF with a 7F template masking the retinal periphery and the full UWF image (UWF-global). Data were analyzed from January 2022 to March 2023. Hemoglobin A1c was assessed quarterly during the DCCT and annually during the EDIC study using high-performance liquid chromatography. Retinopathy was determined independently for all imaging as mild, moderate, or severe nonproliferative DR (SNPDR) using the Early Treatment Diabetic Retinopathy Study (ETDRS) grading scale for the 7F images and the global ETDRS grading scale for the UWF images. Panretinal and focal photocoagulation were self-reported or based on scarring location and pattern observed during grading. Proliferative DR (PDR) was defined by observed neovascularization or evidence of panretinal photocoagulation. Among the 785 participants included in this study, 420 (53%) were male and 365 (47%) were female. The mean (SD) age was 61 (7) years. DR grading between UWF-7F and 7F imaging was correlated for all outcomes, including for severe outcomes, such as SNPDR (κ, 0.73; concordance, 96%), PDR (κ, 0.74; concordance, 97%), scatter photocoagulation (κ, 0.97; concordance, 99%), and focal photocoagulation (κ, 0.71; concordance, 98%). Most DR severity scores were within 1 step (1410 of 1529 [92%]), and 3% (51 of 1529) were more than 2 steps apart (κ, 0.45; 95% CI, 0.42-0.49; weighted κ, 0.63; 95% CI, 0.60-0.67) on the ETDRS severity scale. DR severity assessed within the UWF-global area was higher compared to 7F (median [IQR] UWF-global score, 3 [2-3] vs median 7F level score, 2.0 [1-3]; P < .001), although the 2 modalities were correlated (1225 of 1508 [81%] 1-step agreement; weighted κ, 0.41). Standard ETDRS 7F and UWF evaluations of DR were comparable for ETDRS severity levels as previously reported by Diabetic Retinopathy Clinical Research Retina Network reports. In addition, these evaluations of DR were comparable for DCCT/EDIT study outcomes and major study conclusions, suggesting that use of UWF imaging is not likely to introduce relevant measurement biases in future longitudinal studies. ClinicalTrials.gov Identifiers: NCT00360815.

Intensive Glycemic Management Is Associated With Reduced Retinal Structure Abnormalities on Ocular Coherence Tomography in the DCCT/EDIC Study.

To investigate quantitative and qualitative changes in retinal structure using optical coherence tomography (OCT) and their associations with systemic or other risk factors in individuals with type 1 diabetes (T1D). In the Epidemiology of Diabetes Interventions and Complications (EDIC) study, OCT images were obtained during study years 25-28 (2019-2022) in 937 participants; 54% and 46% were from the original intensive (INT) and conventional (CONV) glycemic management treatment groups, respectively. Average age for participants was 61 years old, diabetes duration 39 years, and HbA1c 7.6%. Participants originally in the CONV group were more likely to have disorganization of retinal inner layers (DRIL) (CONV 27.3% vs. INT 18.7%; P = 0.0003), intraretinal fluid (CONV 24.4% vs. INT 19.2%; P = 0.0222), and intraretinal cysts (CONV 20.8% vs. INT 16.6%; P = 0.0471). In multivariable models, sex, age, smoking, mean updated systolic blood pressure, and history of "clinically significant" macular edema (CSME) and of anti-VEGF treatment were independently associated with changes in central subfield thickness, while HbA1c, BMI, and history of CSME and of ocular surgery were associated with DRIL. Visual acuity (VA) decline was associated with significant thinning of all retinal subfields except for the central and inner nasal subfields. Early intensive glycemic management in T1D is associated with a decreased risk of DRIL. This important morphological abnormality was associated with a history of macular edema, a history of ocular surgery, and worse VA. This study reveals benefits of intensive glycemic management on the retina beyond features detected by fundus photographs and ophthalmoscopy.

Epicardial and intra-thoracic adipose tissue and cardiovascular calcifications in type 1 diabetes (T1D) in epidemiology of diabetes Interventions and Complications (EDIC): A pilot study

ObjectiveCoronary artery, aortic valve, and descending aorta calcification (CAC, AVC, DAC) are manifestations of atherosclerosis, and cardiac epicardial adipose tissue (EAT) indicates heart adiposity. This study explored the association between cardiac adipose tissue and cardiovascular calcification in participants with long-standing T1D. MethodsEAT and intra-thoracic adipose tissue (IAT) were measured in 100 T1D subjects with cardiac computed tomography (CT) scans in the EDIC study. Volume analysis software was used to measure fat volumes. Spearman correlations were calculated between CAC, AVC, DAC with EAT, and IAT. Associations were evaluated using multiple linear and logistic regression models. ResultsParticipants ranged in age from 32 to 57. Mean EAT, and IAT were 38.5 and 50.8 mm3, respectively, and the prevalence of CAC, AVC, and DAC was 43.6 %, 4.7 %, and 26.8 %, respectively. CAC was positively correlated with age (p-value = 0.0001) and EAT (p-value = 0.0149) but not with AVC and DAC; IAT was not associated with calcified lesions. In models adjusted for age and sex, higher levels of EAT and IAT were associated with higher CAC (p-value < 0.0001 for both) and higher AVC (p-values of 0.0111 and 0.0053, respectively), but not with DAC. The associations with CAC remained significant (p-value < 0.0001) after further adjustment for smoking, systolic blood pressure, BMI, and LDL, while the associations with AVC did not remain significant. ConclusionIn participants with T1D, higher EAT and IAT levels are correlated with higher CAC scores. EAT and IAT were not independently correlated with DAC or AVC.

Associations of Kidney Tubular Biomarkers With Incident Macroalbuminuria and Sustained Low eGFR in DCCT/EDIC.

Tubulointerstitial injury contributes to diabetic kidney disease (DKD) progression. We tested tubular biomarker associations with DKD development in type 1 diabetes (T1D). We performed a case-cohort study examining associations of tubular biomarkers, measured across seven time points spanning ∼30 years, with incident macroalbuminuria ("severely elevated albuminuria," urinary albumin excretion rate [AER] ≥300 mg/day) and sustained low estimated glomerular filtration rate (eGFR) (persistent eGFR <60 mL/min/1.73 m2) in the Diabetes Control and Complications Trial (DCCT)/Epidemiology of Diabetes Interventions and Complications (EDIC) study. Biomarkers included KIM-1 and sTNFR1 in serum/plasma, MCP-1 and EGF in urine, and a composite tubular secretion score reflecting secreted solute clearance. We assessed biomarkers using single values, as mean values from consecutive time points, and as change over consecutive time points, each as time-updated exposures. At baseline, mean diabetes duration was 5.9 years, with mean HbA1c 8.9%, eGFR 125 mL/min/1.73 m2, and AER 16 mg/day. There were 4.8 and 3.5 cases per 1,000 person-years of macroalbuminuria and low eGFR, respectively. Assessed according to single biomarker values, KIM-1 was associated with risk of subsequent macroalbuminuria and low eGFR (hazard ratio [HR] per 20% higher biomarker 1.11 [95% CI 1.06, 1.16] and 1.12 [1.04, 1.21], respectively) and sTNFR1 was associated with subsequent macroalbuminuria (1.14 [1.03, 1.25]). Mean KIM-1 and EGF-to-MCP-1 ratio were associated with subsequent low eGFR. In slope analyses, increases in KIM-1 and sTNFR1 were associated with subsequent macroalbuminuria (per 20% biomarker increase, HR 1.81 [1.40, 2.34] and 1.95 [1.18, 3.21]) and low eGFR (2.26 [1.65, 3.09] and 2.94 [1.39, 6.23]). Serial KIM-1 and sTNFR1 are associated with incident macroalbuminuria and sustained low eGFR in T1D.

Neuropathic Pain With and Without Diabetic Peripheral Neuropathy in Type 1 Diabetes.

Diabetic peripheral neuropathy (DPN) is common; however, the features and burden of neuropathic pain (NP) in type 1 diabetes (T1D) are poorly understood. We evaluated the incidence of first occurrence, annual prevalence, remission, and risk factors for NP during long-term follow-up of participants with T1D. The Michigan Neuropathy Screening Instrument (MNSI) was administered annually (1994-2020) for 1,324 participants in the Epidemiology of Diabetes Interventions and Complications (EDIC) study. NP with clinical signs of DPN (NP DPN+) was defined according to self-reported NP plus an examination score >2, while NP without clinical signs of DPN (NP DPN-) was defined according to self-reported NP and an examination score ≤2. At EDIC year 1, median age for participants was 36 years (interquartile range 30, 41), diabetes duration 13 years (10, 18), and HbA1c 7.9% (7.2, 8.9). At year 26 (median diabetes duration 39 years), cumulative incidence of NP was 57%, regardless of concomitant clinical signs of DPN (36% NP DPN+ vs. 46% NP DPN-). NP prevalence was 20% at 26 years (11% NP DPN+ and 9% NP DPN-), suggesting frequent remission. Annualized remission rates were similar regardless of pain medication use. In addition to HbA1c, female sex was associated with NP DPN-. NP incidence in T1D was high and frequently occurred in the absence of clinical signs of neuropathy, as assessed with the MNSI. Pain remission was not explained by pain medication use. Effective clinical strategies for identification and management are needed.

Methylglyoxal Adducts Are Prognostic Biomarkers for Diabetic Kidney Disease in Patients With Type 1 Diabetes.

Over 30% of patients with type 1 diabetes develop diabetic kidney disease (DKD), significantly increasing mortality risk. The Diabetes Control and Complications Trial (DCCT) and follow-up study Epidemiology of Diabetes Interventions and Complications (EDIC) established that glycemic control measured by HbA1c predicts DKD risk. However, the continued high incidence of DKD reinforces the urgent need for additional biomarkers to supplement HbA1c. Here, we assessed biomarkers induced by methylglyoxal (MG), a metabolic by-product that forms covalent adducts on DNA, RNA, and protein called MG-adducts. Urinary MG-adducts were measured from patients with type 1 diabetes enrolled in DCCT/EDIC who did (cases, n=90) or did not (controls, n=117) develop DKD. Univariate and multivariable analysis revealed that measurements of MG-adducts independently predict DKD, even before established DKD biomarkers such as glomerular filtration rate and albumin excretion rate. Elevated levels of MG-adducts bestowed the greatest risk of developing DKD in a multivariable model that included HbA1c and other clinical covariates. Our work establishes a novel class of biomarkers to predict DKD risk and suggests that inclusion of MG-adducts may be a valuable tool to improve existing predictors of complications like DKD prior to overt disease to aid in identifying at-risk individuals and personalized risk management.

Utility of the NIH Toolbox Cognition Battery in middle to older aged adults with longstanding type 1 diabetes: The DCCT/EDIC study

Objective: Adults with type 1 diabetes (T1D) face an increased risk for cognitive decline and dementia. Diabetes-related and vascular risk factors have been linked to cognitive decline using detailed neuropsychological testing; however, it is unclear if cognitive screening batteries can detect cognitive changes associated with aging in T1D. Method: 1,049 participants with T1D (median age 59 years; range 43–74) from the Diabetes Control and Complications Trial (DCCT), and the follow-up Epidemiology of Diabetes Interventions and Complications (EDIC) study, completed the NIH Toolbox Cognition Battery (NIHTB-C) and Montreal Cognitive Assessment (MoCA). Neuropsychological assessments, depression, glycated hemoglobin levels (HbA1c), severe hypoglycemia, T1D complications, and vascular risk factors were assessed repeatedly over 32 years to determine associations with current NIHTB-C performance. Available cognitive data was clinically adjudicated to determine cognitive impairment status. Results: NIHTB-C scores had moderate associations (r = 0.36–0.53) with concurrently administered neuropsychological tests. In multivariate models, prior severe hypoglycemic episodes, depression symptoms, nephropathy, lower BMI, and higher HbA1c and LDL cholesterol were associated with poorer NIHTB-C Fluid Cognition Composite scores. The NIHTB-C adequately detected adjudicated cognitive impairment (Area Under the Curve = 0.86; optimal cut score ≤90). The MoCA performed similarly (Area Under the Curve = 0.83; optimal cut score ≤25). Conclusions: The NIHTB-C is sensitive to the cognitive effects of diabetes-related and vascular risk factors, correlated with neuropsychological testing, and accurately detects adjudicated cognitive impairment. These data support its use as a screening test in middle to older aged adults with T1D to determine if referral for detailed neuropsychological assessment is needed.

Patterns of Regional Brain Atrophy and Brain Aging in Middle- and Older-Aged Adults With Type 1 Diabetes

Little is known about structural brain changes in type 1 diabetes (T1D) and whether there are early manifestations of a neurodegenerative condition like Alzheimer disease (AD) or evidence of premature brain aging. To evaluate neuroimaging markers of brain age and AD-like atrophy in participants with T1D in the Diabetes Control and Complications Trial (DCCT)/Epidemiology of Diabetes Interventions and Complications (EDIC) study, identify which brain regions are associated with the greatest changes in patients with T1D, and assess the association between cognition and brain aging indices. This cohort study leveraged data collected during the combined DCCT (randomized clinical trial, 1983-1993) and EDIC (observational study, 1994 to present) studies at 27 clinical centers in the US and Canada. A total of 416 eligible EDIC participants and 99 demographically similar adults without diabetes were enrolled in the magnetic resonance imaging (MRI) ancillary study, which reports cross-sectional data collected in 2018 to 2019 and relates it to factors measured longitudinally in DCCT/EDIC. Data analyses were performed between July 2020 and April 2022. T1D diagnosis. Psychomotor and mental efficiency were evaluated using verbal fluency, digit symbol substitution test, trail making part B, and the grooved pegboard. Immediate memory scores were derived from the logical memory subtest of the Wechsler memory scale and the Wechsler digit symbol substitution test. MRI and machine learning indices were calculated to predict brain age and quantify AD-like atrophy. This study included 416 EDIC participants with a median (range) age of 60 (44-74) years (87 of 416 [21%] were older than 65 years) and a median (range) diabetes duration of 37 (30-51) years. EDIC participants had consistently higher brain age values compared with controls without diabetes, indicative of approximately 6 additional years of brain aging (EDIC participants: β, 6.16; SE, 0.71; control participants: β, 1.04; SE, 0.04; P < .001). In contrast, AD regional atrophy was comparable between the 2 groups. Regions with atrophy in EDIC participants vs controls were observed mainly in the bilateral thalamus and putamen. Greater brain age was associated with lower psychomotor and mental efficiency among EDIC participants (β, -0.04; SE, 0.01; P < .001), but not among controls. The findings of this study suggest an increase in brain aging among individuals with T1D without any early signs of AD-related neurodegeneration. These increases were associated with reduced cognitive performance, but overall, the abnormal patterns seen in this sample were modest, even after a mean of 38 years with T1D.

A Machine Learning-Based Severity Prediction Tool for the Michigan Neuropathy Screening Instrument.

Diabetic sensorimotor polyneuropathy (DSPN) is a serious long-term complication of diabetes, which may lead to foot ulceration and amputation. Among the screening tools for DSPN, the Michigan neuropathy screening instrument (MNSI) is frequently deployed, but it lacks a straightforward rating of severity. A DSPN severity grading system has been built and simulated for the MNSI, utilizing longitudinal data captured over 19 years from the Epidemiology of Diabetes Interventions and Complications (EDIC) trial. Machine learning algorithms were used to establish the MNSI factors and patient outcomes to characterise the features with the best ability to detect DSPN severity. A nomogram based on multivariable logistic regression was designed, developed and validated. The extra tree model was applied to identify the top seven ranked MNSI features that identified DSPN, namely vibration perception (R), 10-gm filament, previous diabetic neuropathy, vibration perception (L), presence of callus, deformities and fissure. The nomogram's area under the curve (AUC) was 0.9421 and 0.946 for the internal and external datasets, respectively. The probability of DSPN was predicted from the nomogram and a DSPN severity grading system for MNSI was created using the probability score. An independent dataset was used to validate the model's performance. The patients were divided into four different severity levels, i.e., absent, mild, moderate, and severe, with cut-off values of 10.50, 12.70 and 15.00 for a DSPN probability of less than 50, 75 and 100%, respectively. We provide an easy-to-use, straightforward and reproducible approach to determine prognosis in patients with DSPN.

Nomogram-Based Chronic Kidney Disease Prediction Model for Type 1 Diabetes Mellitus Patients Using Routine Pathological Data.

Type 1 diabetes mellitus (T1DM) patients are a significant threat to chronic kidney disease (CKD) development during their life. However, there is always a high chance of delay in CKD detection because CKD can be asymptomatic, and T1DM patients bypass traditional CKD tests during their routine checkups. This study aims to develop and validate a prediction model and nomogram of CKD in T1DM patients using readily available routine checkup data for early CKD detection. This research utilized 1375 T1DM patients’ sixteen years of longitudinal data from multi-center Epidemiology of Diabetes Interventions and Complications (EDIC) clinical trials conducted at 28 sites in the USA and Canada and considered 17 routinely available features. Three feature ranking algorithms, extreme gradient boosting (XGB), random forest (RF), and extremely randomized trees classifier (ERT), were applied to create three feature ranking lists, and logistic regression analyses were performed to develop CKD prediction models using these ranked feature lists to identify the best performing top-ranked features combination. Finally, the most significant features were selected to develop a multivariate logistic regression-based CKD prediction model for T1DM patients. This model was evaluated using sensitivity, specificity, accuracy, precision, and F1 score on train and test data. A nomogram of the final model was further generated for easy application in clinical practices. Hypertension, duration of diabetes, drinking habit, triglycerides, ACE inhibitors, low-density lipoprotein (LDL) cholesterol, age, and smoking habit were the top-8 features ranked by the XGB model and identified as the most important features for predicting CKD in T1DM patients. These eight features were selected to develop the final prediction model using multivariate logistic regression, which showed 90.04% and 88.59% accuracy in internal and test data validation. The proposed model showed excellent performance and can be used for CKD identification in T1DM patients during routine checkups.

Plasma apoM Levels and Progression to Kidney Dysfunction in Patients With Type 1 Diabetes.

Apolipoprotein M (apoM), primarily carried by HDL, has been associated with several conditions, including cardiovascular disease and diabetic nephropathy. This study proposes to examine whether plasma apoM levels are associated with the development of diabetic kidney disease, assessed as progression to macroalbuminuria (MA) and chronic kidney disease (CKD). Plasma apoM was measured using an enzyme immunoassay in 386 subjects from the Diabetes Control and Complications Trial (DCCT)/Epidemiology of Diabetes Interventions and Complications (EDIC) cohort at DCCT entry and closeout and the concentrations used to determine the association with risk of progression to kidney dysfunction from the time of measurement through 18 years of EDIC follow-up. apoM levels, at DCCT baseline, were higher in patients who developed CKD than in those who retained normal renal function. At DCCT closeout, participants who progressed to MA, CKD, or both MA and CKD also had significantly higher apoM levels than those who remained normal, and increased levels of apoM were associated with increased risk of progression to both MA (risk ratio [RR] 1.30 [95% CI 1.01, 1.66]) and CKD (RR 1.69 [95% CI 1.18, 2.44]). Our results strongly suggest that alterations in apoM and therefore in the composition and function of HDL in type 1 diabetes are present early in the disease process and are associated with the development of nephropathy.

Brain Structure Among Middle-aged and Older Adults With Long-standing Type 1 Diabetes in the DCCT/EDIC Study.

Individuals with type 1 diabetes mellitus (T1DM) are living to ages when neuropathological changes are increasingly evident. We hypothesized that middle-aged and older adults with long-standing T1DM will show abnormal brain structure in comparison with control subjects without diabetes. MRI was used to compare brain structure among 416 T1DM participants in the Epidemiology of Diabetes Interventions and Complications (EDIC) study with that of 99 demographically similar control subjects without diabetes at 26 U.S. and Canadian sites. Assessments included total brain (TBV) (primary outcome), gray matter (GMV), white matter (WMV), ventricle, and white matter hyperintensity (WMH) volumes and total white matter mean fractional anisotropy (FA). Biomedical assessments included HbA1c and lipid levels, blood pressure, and cognitive assessments of memory and psychomotor and mental efficiency (PME). Among EDIC participants, HbA1c, severe hypoglycemia history, and vascular complications were measured longitudinally. Mean age of EDIC participants and control subjects was 60 years. T1DM participants showed significantly smaller TBV (least squares mean ± SE 1,206 ± 1.7 vs. 1,229 ± 3.5 cm3, P < 0.0001), GMV, and WMV and greater ventricle and WMH volumes but no differences in total white matter mean FA versus control subjects. Structural MRI measures in T1DM were equivalent to those of control subjects who were 4-9 years older. Lower PME scores were associated with altered brain structure on all MRI measures in T1DM participants. Middle-aged and older adults with T1DM showed brain volume loss and increased vascular injury in comparison with control subjects without diabetes, equivalent to 4-9 years of brain aging.

Physical Function in Middle-aged and Older Adults With Type 1 Diabetes: Long-term Follow-up of the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications Study.

To describe the prevalence and clinical correlates of functional limitations in middle-aged and older adults with long-standing type 1 diabetes. Functional limitations were assessed for 1,094 participants in the Epidemiology of Diabetes Interventions and Complications (EDIC) study, a multicenter, longitudinal, observational follow-up of participants with type 1 diabetes randomly assigned to intensive or conventional diabetes therapy during the Diabetes Control and Complications Trial (DCCT). The primary outcome measure was a score <10 on the Short Physical Performance Battery (SPPB). The secondary outcome, self-reported functional limitation, was assessed by written questionnaire. Logistic regression models were used to assess associations of both outcomes with demographic and clinical factors (glycemic and nonglycemic factors, micro- and macrovascular complications, DCCT cohort, and treatment assignment). Participants were 53% male, with mean ± SD age 59.5 ± 6.8 years and diabetes duration 37.9 ± 4.9 years. The prevalence of SPPB score <10 was 21%. The prevalence of self-reported functional limitations was 48%. While DCCT treatment assignment was not associated with physical function outcomes measured ∼25 years after the end of the DCCT, the time-weighted mean DCCT/EDIC HbA1c was associated with both outcomes. Other clinical factors associated with both outcomes in multivariable analyses were BMI, general psychological distress, and cardiac autonomic neuropathy. Almost half of the middle-aged and older adults with long-standing type 1 diabetes reported functional limitations, which were associated with higher HbA1c and BMI, general psychological distress, and cardiac autonomic neuropathy. Future research is needed to determine whether these findings are generalizable.

Performance Analysis of Conventional Machine Learning Algorithms for Diabetic Sensorimotor Polyneuropathy Severity Classification Using Nerve Conduction Studies.

Background Diabetic sensorimotor polyneuropathy (DSPN) is a major form of complication that arises in long-term diabetic patients. Even though the application of machine learning (ML) in disease diagnosis is very common and well-established in the field of research, its application in DSPN diagnosis using nerve conduction studies (NCS), is very limited in the existing literature. Method In this study, the NCS data were collected from the Diabetes Control and Complications Trial (DCCT) and its follow-up Epidemiology of Diabetes Interventions and Complications (EDIC) clinical trials. The NCS variables are median motor velocity (m/sec), median motor amplitude (mV), median motor F-wave (msec), median sensory velocity (m/sec), median sensory amplitude (μV), Peroneal Motor Velocity (m/sec), peroneal motor amplitude (mv), peroneal motor F-wave (msec), sural sensory velocity (m/sec), and sural sensory amplitude (μV). Three different feature ranking techniques were used to analyze the performance of eight different conventional classifiers. Results The ensemble classifier outperformed other classifiers for the NCS data ranked when all the NCS features were used and provided an accuracy of 93.40%, sensitivity of 91.77%, and specificity of 98.44%. The random forest model exhibited the second-best performance using all the ten features with an accuracy of 93.26%, sensitivity of 91.95%, and specificity of 98.95%. Both ensemble and random forest showed the kappa value 0.82, which indicates that the models are in good agreement with the data and the variables used and are accurate to identify DSPN using these ML models. Conclusion This study suggests that the ensemble classifier using all the ten NCS variables can predict the DSPN severity which can enhance the management of DSPN patients.

Type 1 diabetes and oral health: Findings from the Epidemiology of Diabetes Interventions and Complications (EDIC) study

ObjectiveTo describe long-term oral health outcomes and examine associations between sociodemographic factors, clinical characteristics, and markers of diabetes control on tooth loss in participants with type 1 diabetes enrolled in the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) study. Research design and methodsOral health outcomes related to tooth loss were reported at annual visits during EDIC years 22–26 (2015–2019). Generalized estimating equation models were used to assess the association of individual risk factors and tooth loss, over repeated time points. ResultsA total of 165 (17%) participants with type 1 diabetes reported 221 oral health outcomes related to tooth loss over a five-year period. After controlling for age and current tobacco use, the presence of diabetic peripheral neuropathy was significantly associated with an increased odds of tooth loss (OR = 1.88, 95% CI 1.24, 2.87) while higher mean HDL/LDL cholesterol ratio was significantly associated with a decreased odds of tooth loss (OR = 0.87, 95% CI = 0.79, 0.97). ConclusionsThese findings suggest that diabetes-related complications, either resulting from or independent of poor glycemia, may be directly associated with oral health conditions, and support the need for individuals with type 1 diabetes and providers to implement lifestyle and medical interventions to reduce oral health risks.

Coronary Artery Disease Events and Carotid Intima-Media Thickness in Type 1 Diabetes in the DCCT/EDIC Cohort.

BackgroundCarotid artery intima‐media thickness (IMT) is associated with the risk of subsequent cardiovascular events in the general population. This association has not been established in type 1 diabetes.Methods and ResultsWe studied if carotid IMT is associated with the risk of a first coronary artery disease event in participants with type 1 diabetes in the EDIC (Epidemiology of Diabetes Interventions and Complications) study, the long‐term observational follow‐up of the DCCT (Diabetes Control and Complications Trial). Between 1994 and 1996, common carotid artery and internal carotid artery IMT were measured with high‐resolution ultrasound in 1309 study participants with a mean age of 35 years and diabetes duration of 13.8 years; 52% were men. Cox proportional hazards models evaluated the association of standardized common carotid artery IMT and internal carotid artery IMT with subsequent cardiovascular events over the next 17 years. Models were adjusted for age, sex, mean hemoglobin A1c levels, and traditional cardiovascular risk factors. Associations of common carotid artery IMT with subsequent CAD were significant after adjustment for imaging device, sex, and age (hazard ratio [HR], 1.23 per 0.09 mm [95% CI, [1.04–1.45]; P=0.0141), but did not remain significant after further adjustment for traditional risk factors and hemoglobin A1c (HR, 1.14 per 0.09 mm [95% CI, 0.97–1.33]; P=0.1206). No significant associations with subsequent coronary artery disease events were seen for internal carotid artery IMT.ConclusionsIn the DCCT/EDIC cohort with type 1 diabetes, common carotid artery IMT, but not internal carotid artery IMT, is weakly associated with subsequent coronary artery events, an association eliminated after adjusting for coexistent traditional cardiovascular risk factors.RegistrationURL: https://www.clinicaltrials.gov; Unique identifiers: NCT00360815 and NCT00360893.

Performance Analysis of Conventional Machine Learning Algorithms for Identification of Chronic Kidney Disease in Type 1 Diabetes Mellitus Patients.

Chronic kidney disease (CKD) is one of the severe side effects of type 1 diabetes mellitus (T1DM). However, the detection and diagnosis of CKD are often delayed because of its asymptomatic nature. In addition, patients often tend to bypass the traditional urine protein (urinary albumin)-based CKD detection test. Even though disease detection using machine learning (ML) is a well-established field of study, it is rarely used to diagnose CKD in T1DM patients. This research aimed to employ and evaluate several ML algorithms to develop models to quickly predict CKD in patients with T1DM using easily available routine checkup data. This study analyzed 16 years of data of 1375 T1DM patients, obtained from the Epidemiology of Diabetes Interventions and Complications (EDIC) clinical trials directed by the National Institute of Diabetes, Digestive, and Kidney Diseases, USA. Three data imputation techniques (RF, KNN, and MICE) and the SMOTETomek resampling technique were used to preprocess the primary dataset. Ten ML algorithms including logistic regression (LR), k-nearest neighbor (KNN), Gaussian naïve Bayes (GNB), support vector machine (SVM), stochastic gradient descent (SGD), decision tree (DT), gradient boosting (GB), random forest (RF), extreme gradient boosting (XGB), and light gradient-boosted machine (LightGBM) were applied to developed prediction models. Each model included 19 demographic, medical history, behavioral, and biochemical features, and every feature’s effect was ranked using three feature ranking techniques (XGB, RF, and Extra Tree). Lastly, each model’s ROC, sensitivity (recall), specificity, accuracy, precision, and F-1 score were estimated to find the best-performing model. The RF classifier model exhibited the best performance with 0.96 (±0.01) accuracy, 0.98 (±0.01) sensitivity, and 0.93 (±0.02) specificity. LightGBM performed second best and was quite close to RF with 0.95 (±0.06) accuracy. In addition to these two models, KNN, SVM, DT, GB, and XGB models also achieved more than 90% accuracy.

Comparison of central laboratory HbA1c measurements obtained from a capillary collection versus a standard venous whole blood collection in the GRADE and EDIC studies.

BackgroundWe compared HbA1c values obtained from capillary blood collection kits versus venous whole blood collections in study participants with type 1 or type 2 diabetes.MethodsA total of 122 subjects, 64 with type 2 diabetes participating in the Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness (GRADE) Study and 58 with type 1 diabetes from the Epidemiology of Diabetes Interventions and Complications (EDIC) Study, participated in the validation study. Capillary tubes were filled by fingerstick by the participants on the same day as the collection of venous whole blood samples in EDTA-containing test tubes and were mailed to the central laboratory. HbA1c in all samples was measured with the same high-performance liquid chromatography. GRADE participants also completed a questionnaire on the ease of performing capillary collections.ResultsParticipants from 22 clinical centers (GRADE n = 5, EDIC n = 17) were between 35 and 86 years of age, with 52% male and diverse race/ethnicities. Venous HbA1c results ranged between 5.4–11.9% (35.5–106.6 mmol/mol) with corresponding capillary results ranging between 4.2–11.9% (22.4–106.6 mmol/mol). The venous and capillary results were highly correlated (R2 = 0.993) and 96.7% differed by ≤0.2% (2.2 mmol/mol). Of participants surveyed, 69% indicated that the instructions and collection were easy to follow and 97% felt the collection method would be easy to do at home.ConclusionsThe capillary blood HbA1c results compared well with the conventional venous whole blood results. The capillary kits can be employed in other studies to reduce interruption of critical data collection and potentially to augment clinical care when in-person visits are not possible.

A nomogram-based diabetic sensorimotor polyneuropathy severity prediction using Michigan neuropathy screening instrumentations

BackgroundDiabetic Sensorimotor polyneuropathy (DSPN) is one of the major indelible complications in diabetic patients. Michigan neuropathy screening instrumentation (MNSI) is one of the most common screening techniques used for DSPN, however, it does not provide any direct severity grading system. MethodFor designing and modeling the DSPN severity grading systems for MNSI, 19 years of data from Epidemiology of Diabetes Interventions and Complications (EDIC) clinical trials were used. Different Machine learning-based feature ranking techniques were investigated to identify the important MNSI features associated with DSPN diagnosis. A multivariable logistic regression-based nomogram was generated and validated for DSPN severity grading using the best performing top-ranked MNSI features. ResultsTop-10 ranked features from MNSI features: Appearance of Feet (R), Ankle Reflexes (R), Vibration perception (L), Vibration perception (R), Appearance of Feet (L), 10-gm filament (L), Ankle Reflexes (L), 10-gm filament (R), Bed Cover Touch, and Ulceration (R) were identified as important features for identifying DSPN by Multi-Tree Extreme Gradient Boost model. The nomogram-based prediction model exhibited an accuracy of 97.95% and 98.84% for the EDIC test set and an independent test set, respectively. A DSPN severity score technique was generated for MNSI from the DSPN severity prediction model. DSPN patients were stratified into four severity levels: absent, mild, moderate, and severe using the cut-off values of 17.6, 19.1, 20.5 for the DSPN probability less than 50%, 75%–90%, and above 90%, respectively. ConclusionsThe findings of this work provide a machine learning-based MNSI severity grading system which has the potential to be used as a secondary decision support system by health professionals in clinical applications and large clinical trials to identify high-risk DSPN patients.

Longitudinal patterns of urinary incontinence and associated predictors in women with type 1 diabetes.

Urinary incontinence (UI) in women is a dynamic condition with numerous risk factors yet most studies have focused on examining its prevalence at a single time. The objective of this study was to describe the long-term time course of UI in women with type 1 diabetes (T1D). Longitudinal data in women with T1D were collected from 568 women in the Epidemiology of Diabetes Interventions and Complications (EDIC) study, the observational follow-up of the Diabetes Control and Complications Trial (DCCT) cohort. Over a 12-year period, participants annually responded to whether they had experienced UI in the past year. We identified four categories of UI in this population over time: 205 (36.1%) women never reported UI (no UI), 70 (12.3%) reported it one or two consecutive years only (isolated UI), 247 (43.5%) periodically changed status between UI and no UI (intermittent UI), and 46 (8.1%) reported UI continuously after the first report (persistent UI). Compared to women reporting no/isolated UI, women displaying the intermittent phenotype were significantly more likely to be obese (OR:1.86, 95% CI 1.15, 3.00) and report prior hysterectomy (OR: 2.57, 95% CI: 1.39, 4.77); whereas women with persistent UI were significantly more likely to have abnormal autonomic function (OR: 2.36, 95% CI: 1.16-4.80). UI is a dynamic condition in women with T1D. Varying risk factors observed for the different phenotypes of UI suggest distinctive pathophysiological mechanisms. These findings have the potential to be used to guide individualized interventions for UI in women with diabetes.