The Evaluation of Screening and Early Detection Strategies for Type 2 Diabetes and Impaired Glucose Tolerance (DETECT-2) update of the Finnish diabetes risk score for prediction of incident type 2 diabetes

External validation of three diabetes prediction scores in a Spanish cohort: does adding high risk for depression improve the validation of the FINDRISC score (FINDRISC-MOOD)?

OBJECTIVES

To evaluate the external validity of the FINDRISC, DESIR and ADA risk scores for the prediction of diabetes in a Spanish population aged >45 years and to test the possible improvement of FINDRISC by adding a new variable of high risk of depression when Patient Health Questionnaire-9 (PHQ-9) questionnaire score ≥10 (FINDRISC-MOOD).

DESIGN

Prospective population-based cohort study.

SETTING

10 primary healthcare centres in the north of the city of Madrid (Spain).

PARTICIPANTS

A total of 1242 participants without a history of diabetes and with 2-hour oral glucose tolerance test (OGTT) plasma glucose <200 mg/dL (<11.1 mmol/L) were followed up for 7.3 years (median) using their electronic health records (EHRs) and telephone contact.

PRIMARY AND SECONDARY OUTCOME MEASURES

Diabetes risk scores (FINDRISC, DESIR, ADA), PHQ-9 questionnaire and 2-hour-OGTT were measured at baseline. Incident diabetes was defined as treatment for diabetes, fasting plasma glucose ≥126 mg/dL (≥7.0 mmol/L), new EHR diagnosis or self-reported diagnosis. External validation was performed according to optimal cut-off, sensitivity, specificity and Youden Index. Comparison between diabetes risk scores, including FINDRISC-MOOD (original FINDRISC score plus five points if PHQ-9 ≥10), was measured by area under the receiver operating characteristic curve (AUROC).

RESULTS

During follow-up, 104 (8.4%; 95% CI, 6.8 to 9.9) participants developed diabetes and 185 had a PHQ-9 score ≥10. The AUROC values were 0.70 (95% CI, 0.67 to 0.72) for FINDRISC-MOOD and 0.68 (95% CI, 0.65 to 0.71) for the original FINDRISC. The AUROCs for DESIR and ADA were 0.66 (95% CI, 0.63 to 0.68) and 0.66 (95% CI, 0.63 to 0.69), respectively. There were no significant differences in AUROC between FINDRISC-MOOD and the other scores.

CONCLUSIONS

The results of FINDRISC-MOOD were like those of the other risk scores and do not allow it to be recommended for clinical use.

A non-invasive risk score including skin autofluorescence predicts diabetes risk in the general population

Increased skin autofluorescence (SAF) predicts the development of diabetes-related complications and cardiovascular disease. We assessed the performance of a simple model which includes SAF to identify individuals at high risk for undiagnosed and incident type 2 diabetes, in 58,377 participants in the Lifelines Cohort Study without known diabetes. Newly-diagnosed diabetes was defined as fasting blood glucose ≥ 7.0 mmol/l and/or HbA_1c ≥ 6.5% (≥ 48 mmol/mol) or self-reported diabetes at follow-up. We constructed predictive models based on age, body mass index (BMI), SAF, and parental history of diabetes, and compared to results with the concise FINDRISC model. At 2nd visit to Lifelines, 1113 (1.9%) participants were identified with undiagnosed diabetes and 1033 (1.8%) participants developed diabetes during follow-up. A model comprising age, BMI and SAF yielded an AUC of 0.783 and was non-inferior to the concise FINDRISC model, which had an AUC of 0.797 to predict new diabetes. At a score of 5.8, sensitivity was 78% and specificity of 66%. Model 2 which also incorporated parental diabetes history, had an AUC of 0.792, and a sensitivity of 74% and specificity of 70% at a score of 6.5. Net reclassification index (NRI) did not improve significantly (NRI 1.43% (- 0.50-3.37 p = 0.15). The combination of an easy to perform SAF measurement with age and BMI is a good alternative screening tool suitable for medical and non-medical settings. Parental history of diabetes did not significantly improve model performance in this homogeneous cohort.

Accuracy of Machine Learning Classification Models for the Prediction of Type 2 Diabetes Mellitus: A Systematic Survey and Meta-Analysis Approach

Soft-computing and statistical learning models have gained substantial momentum in predicting type 2 diabetes mellitus (T2DM) disease. This paper reviews recent soft-computing and statistical learning models in T2DM using a meta-analysis approach. We searched for papers using soft-computing and statistical learning models focused on T2DM published between 2010 and 2021 on three different search engines. Of 1215 studies identified, 34 with 136952 patients met our inclusion criteria. The pooled algorithm's performance was able to predict T2DM with an overall accuracy of 0.86 (95% confidence interval [CI] of [0.82, 0.89]). The classification of diabetes prediction was significantly greater in models with a screening and diagnosis (pooled proportion [95% CI] = 0.91 [0.74, 0.97]) when compared to models with nephropathy (pooled proportion = 0.48 [0.76, 0.89] to 0.88 [0.83, 0.91]). For the prediction of T2DM, the decision trees (DT) models had a pooled accuracy of 0.88 [95% CI: 0.82, 0.92], and the neural network (NN) models had a pooled accuracy of 0.85 [95% CI: 0.79, 0.89]. Meta-regression did not provide any statistically significant findings for the heterogeneous accuracy in studies with different diabetes predictions, sample sizes, and impact factors. Additionally, ML models showed high accuracy for the prediction of T2DM. The predictive accuracy of ML algorithms in T2DM is promising, mainly through DT and NN models. However, there is heterogeneity among ML models. We compared the results and models and concluded that this evidence might help clinicians interpret data and implement optimum models for their dataset for T2DM prediction.

The 6-Month Efficacy of an Intensive Lifestyle Modification Program on Type 2 Diabetes Risk Among Rural Women with Prior Gestational Diabetes Mellitus: a Cluster Randomized Controlled Trial

This study aimed to evaluate the efficacy of an intensive lifestyle modification program tailored to rural Chinese women with prior gestational diabetes mellitus compared with usual care. In a cluster randomized controlled trial, 16 towns (clusters) in two distinct rural areas in China were randomly selected (8 towns per district); and 320 women with prior gestational diabetes mellitus were recruited from these towns. With stratification for the two study districts, eight towns (160 women) were randomly assigned to the intervention group of a tailored intensive lifestyle modification program and 8 towns (160 women) to the control group. Process measures were collected on attendance, engagement, fidelity, and satisfaction. Primary efficacy outcomes included glycemic and weight-related outcomes, while secondary efficacy outcomes were behavioral outcomes and type 2 diabetes risk score, which were collected at baseline, 3-month, and 6-month follow-up. Generalized estimation equations were used to analyze the data. High attendance (72% of sessions), engagement (67% of interactive activities and group discussions), fidelity (98%), and satisfaction (92%) with the tailored intensive lifestyle modification program were achieved. There were significant reductions in fasting blood glucose, oral glucose tolerance test 2 h, waist circumference, and type 2 diabetes risk score of participants in the intervention group compared to the control group (p < .05). There was no significant intervention effect on body mass index or behavioral outcomes (p > .05). In this study, we demonstrate the successful efficacy of an Intensive Lifestyle Modification Program in reducing type 2 diabetes risk among younger women with prior gestational diabetes mellitus. This tailored program delivered by local healthcare providers is a promising approach for diabetes prevention in rural China, reducing health disparities in rural communities about diabetes prevention. Registered in the Chinese Clinical Trial Registry (ChiCTR2000037956) on 3rd Jan 2018.

Development of Various Diabetes Prediction Models Using Machine Learning Techniques

BACKGROUND

There are many models for predicting diabetes mellitus (DM), but their clinical implication remains vague. Therefore, we aimed to create various DM prediction models using easily accessible health screening test parameters.

METHODS

Two sets of variables were used to develop eight DM prediction models. One set comprised 62 easily accessible examination results of commonly used variables from a tertiary university hospital. The second set comprised 27 of the 62 variables included in the national routine health checkups. Gradient boosting and random forest algorithms were used to develop the models. Internal validation was performed using the stratified 10-fold cross-validation method.

RESULTS

The area under the receiver operating characteristic curve (ROC-AUC) for the 62-variable DM model making 12-month predictions for subjects without diabetes was the largest (0.928) among those of the eight DM prediction models. The ROC-AUC dropped by more than 0.04 when training with the simplified 27-variable set but still showed fairly good performance with ROC-AUCs between 0.842 and 0.880. The accuracy was up to 11.5% higher (from 0.807 to 0.714) when fasting glucose was included.

CONCLUSION

We created easily applicable diabetes prediction models that deliver good performance using parameters commonly assessed during tertiary university hospital and national routine health checkups. We plan to perform prospective external validation, hoping that the developed DM prediction models will be widely used in clinical practice.

Development and validation of the type 2 diabetes mellitus 10-year risk score prediction models from survey data

AIMS

In this paper, we demonstrate the development and validation of the 10-years type 2 diabetes mellitus (T2DM) risk prediction models based on large survey data.

METHODS

The Survey of Health, Ageing and Retirement in Europe (SHARE) data collected in 12 European countries using 53 variables representing behavioural as well as physical and mental health characteristics of the participants aged 50 or older was used to build and validate prediction models. To account for strongly unbalanced outcome variables, each instance was assigned a weight according to the inverse proportion of the outcome label when the regularized logistic regression model was built.

RESULTS

A pooled sample of 16,363 individuals was used to build and validate a global regularized logistic regression model that achieved an area under the receiver operating characteristic curve of 0.702 (95% CI: 0.698-0.706). Additionally, we measured performance of local country-specific models where AUROC ranged from 0.578 (0.565-0.592) to 0.768 (0.749-0.787).

CONCLUSIONS

We have developed and validated a survey-based 10-year T2DM risk prediction model for use across 12 European countries. Our results demonstrate the importance of re-calibration of the models as well as strengths of pooling the data from multiple countries to reduce the variance and consequently increase the precision of the results.

Non-Alcoholic Fatty Liver Disease Markers Associated with Fasting Serum Insulin and Urinary Albumin Excretion Independent of Fasting Plasma Glucose

Objective: We examined the association between non-alcoholic fatty liver disease (NAFLD) markers and fasting serum immunoreactive insulin (FIRI) and urinary albumin excretion (UAE). Subjects and methods: This study comprised Periods I and II from January 2007 to May 2009, and from June 2009 to December 2011, respectively. After excluding people with ethanol intake ≥210 g/week in men and ≥140 g/week in women, 961 people (613 men, 348 women; mean age: 44 years) were included. We evaluated the fatty liver using ultrasonography score (FLUS) and measured liver enzymes. Results: The mean observation period was 25 ± 9 months. We stratified people into two groups by fasting plasma glucose (FPG) in Period I. The cutoff point between the lower FPG and higher FPG was 100 mg/dL. In regression analysis, serum alanine aminotransferase (ALT) (p < 0.001), FLUS (p < 0.001) and γ-glutamyl transpeptidase (GGTP) (p = 0.022) in Period I were independently associated with FIRI in Period II, whereas in all participants FPG was not. ALT (p < 0.001) and GGTP (p = 0.001) were also independently associated with UAE in people with FPG < 100 mg/dL in Period II. Conclusions: Some NAFLD markers were associated with FIRI and UAE independently of fasting plasma glucose.

Cardiometabolic risk prediction algorithms for young people with psychosis: a systematic review and exploratory analysis

OBJECTIVE

Cardiometabolic risk prediction algorithms are common in clinical practice. Young people with psychosis are at high risk for developing cardiometabolic disorders. We aimed to examine whether existing cardiometabolic risk prediction algorithms are suitable for young people with psychosis.

METHODS

We conducted a systematic review and narrative synthesis of studies reporting the development and validation of cardiometabolic risk prediction algorithms for general or psychiatric populations. Furthermore, we used data from 505 participants with or at risk of psychosis at age 18 years in the ALSPAC birth cohort, to explore the performance of three algorithms (QDiabetes, QRISK3 and PRIMROSE) highlighted as potentially suitable. We repeated analyses after artificially increasing participant age to the mean age of the original algorithm studies to examine the impact of age on predictive performance.

RESULTS

We screened 7820 results, including 110 studies. All algorithms were developed in relatively older participants, and most were at high risk of bias. Three studies (QDiabetes, QRISK3 and PRIMROSE) featured psychiatric predictors. Age was more strongly weighted than other risk factors in each algorithm. In our exploratory analysis, calibration plots for all three algorithms implied a consistent systematic underprediction of cardiometabolic risk in the younger sample. After increasing participant age, calibration plots were markedly improved.

CONCLUSION

Existing cardiometabolic risk prediction algorithms cannot be recommended for young people with or at risk of psychosis. Existing algorithms may underpredict risk in young people, even in the face of other high-risk features. Recalibration of existing algorithms or a new tailored algorithm for the population is required.

Agreement between Type 2 Diabetes Risk Scales in a Caucasian Population: A Systematic Review and Report

Early detection of people with undiagnosed type 2 diabetes (T2D) is an important public health concern. Several predictive equations for T2D have been proposed but most of them have not been externally validated and their performance could be compromised when clinical data is used. Clinical practice guidelines increasingly incorporate T2D risk prediction models as they support clinical decision making. The aims of this study were to systematically review prediction scores for T2D and to analyze the agreement between these risk scores in a large cross-sectional study of white western European workers. A systematic review of the PubMed, CINAHL, and EMBASE databases and a cross-sectional study in 59,042 Spanish workers was performed. Agreement between scores classifying participants as high risk was evaluated using the kappa statistic. The systematic review of 26 predictive models highlights a great heterogeneity in the risk predictors; there is a poor level of reporting, and most of them have not been externally validated. Regarding the agreement between risk scores, the DETECT-2 risk score scale classified 14.1% of subjects as high-risk, FINDRISC score 20.8%, Cambridge score 19.8%, the AUSDRISK score 26.4%, the EGAD study 30.3%, the Hisayama study 30.9%, the ARIC score 6.3%, and the ITD score 3.1%. The lowest agreement was observed between the ITD and the NUDS study derived score (κ = 0.067). Differences in diabetes incidence, prevalence, and weight of risk factors seem to account for the agreement differences between scores. A better agreement between the multi-ethnic derivate score (DETECT-2) and European derivate scores was observed. Risk models should be designed using more easily identifiable and reproducible health data in clinical practice.

Type 2 Diabetes Risk Among University Students in Malaysia

BACKGROUND

Type 2 diabetes (T2D) is a preventable condition. Targeting those who are at risk of getting this disease is essential.

OBJECTIVE

To examine T2D risk among university students in Malaysia and determine its relationship with socio-demographic characteristics and physical activity.

METHODS

The study was conducted cross-sectionally on 390 students selected using quota sampling method from 13 faculties in Universiti Kebangsaan Malaysia. A short form of the International Physical Activity Questionnaire and Finnish Diabetes Risk Score were used to measure the physical activity and T2D risk.

RESULTS

The T2D risk was found to be low (M = 5.23, SD = 3.32) with more than two-third of the student population at the low risk level while a significant proportion of 23.8%, 5.6% and 0.3% having slightly elevated, moderate and high risk respectively. The T2D risk was significantly related to their age (rho = 0.197, p < 0.000), gender (U = 12641, p = 0.011), ethnic group (Χ2 = 18.86, p < 0.000), marital status (Χ2 = 6.597, p = 0.037), residence (U = 10345, p = 0.008), academic year (Χ2 = 14.24, p = 0.007) and physical activity (rho = -0.205, p < 0.000 and Χ2 = 13.515, p = 0.001). Of these, only age (β=0.130) and physical activity (β=-0.159) remained significant in the regression analysis.

CONCLUSION

The findings call for a radical change in the nursing practice to target the amendable factors that are significant in order to prevent the progression of the risk towards type 2 diabetes.

Driving Type 2 Diabetes Risk Scores into Clinical Practice: Performance Analysis in Hospital Settings

Electronic health records and computational modelling have paved the way for the development of Type 2 Diabetes risk scores to identify subjects at high risk. Unfortunately, few risk scores have been externally validated, and their performance can be compromised when routine clinical data is used. The aim of this study was to assess the performance of well-established risk scores for Type 2 Diabetes using routinely collected clinical data and to quantify their impact on the decision making process of endocrinologists. We tested six risk models that have been validated in external cohorts, as opposed to model development, on electronic health records collected from 2008-2015 from a population of 10,730 subjects. Unavailable or missing data in electronic health records was imputed using an existing validated Bayesian Network. Risk scores were assessed on the basis of statistical performance to differentiate between subjects who developed diabetes and those who did not. Eight endocrinologists provided clinical recommendations based on the risk score output. Due to inaccuracies and discrepancies regarding the exact date of Type 2 Diabetes onset, 76 subjects from the initial population were eligible for the study. Risk scores were useful for identifying subjects who developed diabetes (Framingham risk score yielded a c-statistic of 85%), however, our findings suggest that electronic health records are not prepared to massively use this type of risk scores. Use of a Bayesian Network was key for completion of the risk estimation and did not affect the risk score calculation (p > 0.05). Risk score estimation did not have a significant effect on the clinical recommendation except for starting pharmacological treatment (p = 0.004) and dietary counselling (p = 0.039). Despite their potential use, electronic health records should be carefully analyzed before the massive use of Type 2 Diabetes risk scores for the identification of high-risk subjects, and subsequent targeting of preventive actions.

Use of Non-invasive Parameters and Machine-Learning Algorithms for Predicting Future Risk of Type 2 Diabetes: A Retrospective Cohort Study of Health Data From Kuwait

Objective: In recent decades, the Arab population has experienced an increase in the prevalence of type 2 diabetes (T2DM), particularly within the Gulf Cooperation Council. In this context, early intervention programmes rely on an ability to identify individuals at risk of T2DM. We aimed to build prognostic models for the risk of T2DM in the Arab population using machine-learning algorithms vs. conventional logistic regression (LR) and simple non-invasive clinical markers over three different time scales (3, 5, and 7 years from the baseline). Design: This retrospective cohort study used three models based on LR, k-nearest neighbours (k-NN), and support vector machines (SVM) with five-fold cross-validation. The models included the following baseline non-invasive parameters: age, sex, body mass index (BMI), pre-existing hypertension, family history of hypertension, and T2DM. Setting: This study was based on data from the Kuwait Health Network (KHN), which integrated primary health and hospital laboratory data into a single system. Participants: The study included 1,837 native Kuwaiti Arab individuals (equal proportion of men and women) with mean age as 59.5 ± 11.4 years. Among them, 647 developed T2DM within 7 years of the baseline non-invasive measurements. Analytical methods: The discriminatory power of each model for classifying people at risk of T2DM within 3, 5, or 7 years and the area under the receiver operating characteristic curve (AUC) were determined. Outcome measures: Onset of T2DM at 3, 5, and 7 years. Results: The k-NN machine-learning technique, which yielded AUC values of 0.83, 0.82, and 0.79 for 3-, 5-, and 7-year prediction horizons, respectively, outperformed the most commonly used LR method and other previously reported methods. Comparable results were achieved using the SVM and LR models with corresponding AUC values of (SVM: 0.73, LR: 0.74), (SVM: 0.68, LR: 0.72), and (SVM: 0.71, LR: 0.70) for 3-, 5-, and 7-year prediction horizons, respectively. For all models, the discriminatory power decreased as the prediction horizon increased from 3 to 7 years. Conclusions: Machine-learning techniques represent a useful addition to the commonly reported LR technique. Our prognostic models for the future risk of T2DM could be used to plan and implement early prevention programmes for at risk groups in the Arab population.

Validity of the FINDRISC as a prediction tool for diabetes in a contemporary Norwegian population: a 10-year follow-up of the HUNT study

OBJECTIVE

The Finnish Diabetes Risk Score (FINDRISC) is a recommended tool for type 2 diabetes prediction. There is a lack of studies examining the performance of the current 0-26 point FINDRISC scale. We examined the validity of FINDRISC in a contemporary Norwegian risk environment.

RESEARCH DESIGN AND METHODS

We followed 47 804 participants without known diabetes and aged ≥20 years in the HUNT3 survey (2006-2008) by linkage to information on glucose-lowering drug dispensing in the Norwegian Prescription Database (2004-2016). We estimated the C-statistic, sensitivity and specificity of FINDRISC as predictor of incident diabetes, as indicated by incident use of glucose-lowering drugs. We estimated the 10-year cumulative diabetes incidence by categories of FINDRISC.

RESULTS

The C-statistic (95% CI) of FINDRISC in predicting future diabetes was 0.77 (0.76 to 0.78). FINDRISC ≥15 (the conventional cut-off value) had a sensitivity of 38% and a specificity of 90%. The 10-year cumulative diabetes incidence (95% CI) was 4.0% (3.8% to 4.2%) in the entire study population, 13.5% (12.5% to 14.5%) for people with FINDRISC ≥15 and 2.8% (2.6% to 3.0%) for people with FINDRISC <15. Thus, FINDRISC ≥15 had a positive predictive value of 13.5% and a negative predictive value of 97.2% for diabetes within the next 10 years. To approach a similar sensitivity as in the study in which FINDRISC was developed, we would have to lower the cut-off value for elevated FINDRISC to ≥11. This would yield a sensitivity of 73%, specificity of 67%, positive predictive value of 7.7% and negative predictive value of 98.5%.

CONCLUSIONS

The validity of FINDRISC and the risk of diabetes among people with FINDRISC ≥15 is substantially lower in the contemporary Norwegian population than assumed in official guidelines. To identify ~3/4 of those developing diabetes within the next 10 years, we would have to lower the threshold for elevated FINDRISC to ≥11, which would label ~1/3 of the entire adult population as having an elevated FINDRISC necessitating a glycemia assessment.

Development of a new scoring system to predict 5-year incident diabetes risk in middle-aged and older Chinese

AIMS

The aim of this study was to develop a new risk score system to predict 5-year incident diabetes risk among middle-aged and older Chinese population.

METHODS

This prospective study included 17,690 individuals derived from the Dongfeng-Tongji cohort. Participants were recruited in 2008 and were followed until October 2013. Incident diabetes was defined as self-reported clinician diagnosed diabetes, fasting glucose ≥7.0 mmol/l, or the use of insulin or oral hypoglycemic agent. A total of 1390 incident diabetic cases were diagnosed during the follow-up period. β-Coefficients were derived from Cox proportional hazard regression model and were used to calculate the risk score.

RESULTS

The diabetes risk score includes BMI, fasting glucose, hypertension, hyperlipidemia, current smoking status, and family history of diabetes. The β-coefficients of these variables ranged from 0.139 to 1.914, and the optimal cutoff value was 1.5. The diabetes risk score was calculated by multiplying the β-coefficients of the significant variables by 10 and rounding to the nearest integer. The score ranges from 0 to 36. The area under the receiver operating curve of the score was 0.751. At the optimal cutoff value of 15, the sensitivity and specificity were 65.6 and 72.9%, respectively. Based upon these risk factors, this model had the highest discrimination compared with several commonly used diabetes prediction models.

CONCLUSIONS

The newly established diabetes risk score with six parameters appears to be a reliable screening tool to predict 5-year risk of incident diabetes in a middle-aged and older Chinese population.

Integration of Distributed Services and Hybrid Models Based on Process Choreography to Predict and Detect Type 2 Diabetes

Life expectancy is increasing and, so, the years that patients have to live with chronic diseases and co-morbidities. Type 2 diabetes is one of the most prevalent chronic diseases, specifically linked to being overweight and ages over sixty. Recent studies have demonstrated the effectiveness of new strategies to delay and even prevent the onset of type 2 diabetes by a combination of active and healthy lifestyle on cohorts of mid to high risk subjects. Prospective research has been driven on large groups of the population to build risk scores that aim to obtain a rule for the classification of patients according to the odds for developing the disease. Currently, there are more than two hundred models and risk scores for doing this, but a few have been properly evaluated in external groups and integrated into a clinical application for decision support. In this paper, we present a novel system architecture based on service choreography and hybrid modeling, which enables a distributed integration of clinical databases, statistical and mathematical engines and web interfaces to be deployed in a clinical setting. The system was assessed during an eight-week continuous period with eight endocrinologists of a hospital who evaluated up to 8080 patients with seven different type 2 diabetes risk models implemented in two mathematical engines. Throughput was assessed as a matter of technical key performance indicators, confirming the reliability and efficiency of the proposed architecture to integrate hybrid artificial intelligence tools into daily clinical routine to identify high risk subjects.

Comparison of traditional diabetes risk scores and HbA1c to predict type 2 diabetes mellitus in a population based cohort study

AIMS

Compare performances of diabetes risk scores and glycated hemoglobin (HbA1c) to estimate the risk of incident type 2 diabetes mellitus (T2DM) in Northeast Germany.

METHODS

We studied 2916 subjects (20 to 81years) from the Study of Health in Pomerania (SHIP) in a 5-year follow-up period. Diabetes risk scores included the Cooperative Health Research in the Region of Augsburg (KORA) base model, the Danish diabetes risk score and the Data from the Epidemiological Study on the Insulin Resistance syndrome (D.E.S.I.R) clinical risk score. We assessed the performance of each of the diabetes risk scores and the HbA1c for 5-year risk of T2DM by the area under the receiver-operating characteristic curve (AUC) and calibration plots.

RESULTS

In SHIP, the incidence of T2DM was 5.4% (n=157) in the 5-year follow-up period. Diabetes risk scores and HbA1c achieved AUCs ranging from 0.76 for the D.E.S.I.R. clinical risk score to 0.82 for the KORA base model. For diabetes risk scores, the discriminative ability was lower for the age group 55 to 74years. For HbA1c, the discriminative ability also decreased for the group 55 to 74years while it was stable in the age group 30 to 64years old.

CONCLUSIONS

All diabetes risk scores and the HbA1c showed a good prediction for the risk of T2DM in SHIP. Which model or biomarker should be used is driven by its context of use, e.g. the practicability, implementation of interventions and availability of measurement.

Skin autofluorescence improves the Finnish Diabetes Risk Score in the detection of diabetes in a large population-based cohort: The LifeLines Cohort Study

AIM

The aim of the present study was to investigate whether skin autofluorescence would improve the Finnish Diabetes Risk Score (FINDRISC) in detecting undiagnosed diabetes in a large population-based cohort.

METHODS

Included were participants from the Dutch LifeLines Cohort Study. Skin autofluorescence was assessed in an unselected subset of participants using the AGE Reader. After the exclusion of participants with previously diagnosed diabetes (n=1635), pregnant women (n=58) and those using corticosteroids (n=345), 79,248 subjects were eligible for analysis. Diabetes was defined as fasting plasma glucose ≥7.0mmol/L, non-fasting plasma glucose ≥11.1mmol/L or HbA1c ≥6.5% (48mmol/mol).

RESULTS

Diabetes was detected in 1042 participants (aged 55±12 years; 54% male). Skin autofluorescence improved the area under the receiver operating characteristic (AUROC) curve of the FINDRISC model from 0.802 to 0.811 (P<0.001). Furthermore, the addition of skin autofluorescence to FINDRISC reclassified 8-15% of all participants into more accurate risk categories (NRI: 0.080, 95% CI: 0.052-0.110). The proportion of reclassified participants was especially high (>30%) in the intermediate (1% to <5% and 5% to<10%) risk categories. When skin autofluorescence was added to a simplified model (age+body mass index), its discriminatory performance was similar to the full model+skin autofluorescence (AUROC: 0.806, P=0.062).

CONCLUSION

Skin autofluorescence is a non-invasive tool that can be used to further improve the FINDRISC for diabetes detection. The new resultant model is especially useful for reclassifying people in the intermediate-risk categories, where additional blood glucose testing is needed to confirm the presence of diabetes.

Updated risk factors should be used to predict development of diabetes

AIMS

Predicting incident diabetes could inform treatment strategies for diabetes prevention, but the incremental benefit of recalculating risk using updated risk factors is unknown. We used baseline and 1-year data from the Nateglinide and Valsartan in Impaired Glucose Tolerance Outcomes Research (NAVIGATOR) Trial to compare diabetes risk prediction using historical or updated clinical information.

METHODS

Among non-diabetic participants reaching 1year of follow-up in NAVIGATOR, we compared the performance of the published baseline diabetes risk model with a "landmark" model incorporating risk factors updated at the 1-year time point. The C-statistic was used to compare model discrimination and reclassification analyses to demonstrate the relative accuracy of diabetes prediction.

RESULTS

A total of 7527 participants remained non-diabetic at 1year, and 2375 developed diabetes during a median of 4years of follow-up. The C-statistic for the landmark model was higher (0.73 [95% CI 0.72-0.74]) than for the baseline model (0.67 [95% CI 0.66-0.68]). The landmark model improved classification to modest (<20%), moderate (20%-40%), and high (>40%) 4-year risk, with a net reclassification index of 0.14 (95% CI 0.10-0.16) and an integrated discrimination index of 0.01 (95% CI 0.003-0.013).

CONCLUSIONS

Using historical clinical values to calculate diabetes risk reduces the accuracy of prediction. Diabetes risk calculations should be routinely updated to inform discussions about diabetes prevention at both the patient and population health levels.

Predicting glycated hemoglobin levels in the non-diabetic general population: Development and validation of the DIRECT-DETECT prediction model - a DIRECT study

Aims/hypothesis

To develop a prediction model that can predict HbA1c levels after six years in the non-diabetic general population, including previously used readily available predictors.

Methods

Data from 5,762 initially non-diabetic subjects from three population-based cohorts (Hoorn Study, Inter99, KORA S4/F4) were combined to predict HbA1c levels at six year follow-up. Using backward selection, age, BMI, waist circumference, use of anti-hypertensive medication, current smoking and parental history of diabetes remained in sex-specific linear regression models. To minimize overfitting of coefficients, we performed internal validation using bootstrapping techniques. Explained variance, discrimination and calibration were assessed using R², classification tables (comparing highest/lowest 50% HbA1c levels) and calibration graphs. The model was externally validated in 2,765 non-diabetic subjects of the population-based cohort METSIM.

Results

At baseline, mean HbA1c level was 5.6% (38 mmol/mol). After a mean follow-up of six years, mean HbA1c level was 5.7% (39 mmol/mol). Calibration graphs showed that predicted HbA1c levels were somewhat underestimated in the Inter99 cohort and overestimated in the Hoorn and KORA cohorts, indicating that the model’s intercept should be adjusted for each cohort to improve predictions. Sensitivity and specificity (95% CI) were 55.7% (53.9, 57.5) and 56.9% (55.1, 58.7) respectively, for women, and 54.6% (52.7, 56.5) and 54.3% (52.4, 56.2) for men. External validation showed similar performance in the METSIM cohort.

Conclusions/interpretation

In the non-diabetic population, our DIRECT-DETECT prediction model, including readily available predictors, has a relatively low explained variance and moderate discriminative performance, but can help to distinguish between future highest and lowest HbA1c levels. Absolute HbA1c values are cohort-dependent.

Update and Next Steps for Real-World Translation of Interventions for Type 2 Diabetes Prevention: Reflections From a Diabetes Care Editors' Expert Forum

The International Diabetes Federation estimates that 415 million adults worldwide now have diabetes and 318 million have impaired glucose tolerance. These numbers are expected to increase to 642 million and 482 million, respectively, by 2040. This burgeoning pandemic places an enormous burden on countries worldwide, particularly resource-poor regions. Numerous landmark trials evaluating both intensive lifestyle modification and pharmacological interventions have persuasively demonstrated that type 2 diabetes can be prevented or its onset can be delayed in high-risk individuals with impaired glucose tolerance. However, key challenges remain, including how to scale up such approaches for widespread translation and implementation, how to select appropriately from various interventions and tailor them for different populations and settings, and how to ensure that preventive interventions yield clinically meaningful, cost-effective outcomes. In June 2015, a Diabetes Care Editors' Expert Forum convened to discuss these issues. This article, an outgrowth of the forum, begins with a summary of seminal prevention trials, followed by a discussion of considerations for selecting appropriate populations for intervention and the clinical implications of the various diagnostic criteria for prediabetes. The authors outline knowledge gaps in need of elucidation and explore a possible new avenue for securing regulatory approval of a prevention-related indication for metformin, as well as specific considerations for future pharmacological interventions to delay the onset of type 2 diabetes. They conclude with descriptions of some innovative, pragmatic translational initiatives already under way around the world.

Evaluation of Non-Laboratory and Laboratory Prediction Models for Current and Future Diabetes Mellitus: A Cross-Sectional and Retrospective Cohort Study

BACKGROUND

Various diabetes risk scores composed of non-laboratory parameters have been developed, but only a few studies performed cross-validation of these scores and a comparison with laboratory parameters. We evaluated the performance of diabetes risk scores composed of non-laboratory parameters, including a recently published Korean risk score (KRS), and compared them with laboratory parameters.

METHODS

The data of 26,675 individuals who visited the Seoul National University Hospital Healthcare System Gangnam Center for a health screening program were reviewed for cross-sectional validation. The data of 3,029 individuals with a mean of 6.2 years of follow-up were reviewed for longitudinal validation. The KRS and 16 other risk scores were evaluated and compared with a laboratory prediction model developed by logistic regression analysis.

RESULTS

For the screening of undiagnosed diabetes, the KRS exhibited a sensitivity of 81%, a specificity of 58%, and an area under the receiver operating characteristic curve (AROC) of 0.754. Other scores showed AROCs that ranged from 0.697 to 0.782. For the prediction of future diabetes, the KRS exhibited a sensitivity of 74%, a specificity of 54%, and an AROC of 0.696. Other scores had AROCs ranging from 0.630 to 0.721. The laboratory prediction model composed of fasting plasma glucose and hemoglobin A1c levels showed a significantly higher AROC (0.838, P < 0.001) than the KRS. The addition of the KRS to the laboratory prediction model increased the AROC (0.849, P = 0.016) without a significant improvement in the risk classification (net reclassification index: 4.6%, P = 0.264).

CONCLUSIONS

The non-laboratory risk scores, including KRS, are useful to estimate the risk of undiagnosed diabetes but are inferior to the laboratory parameters for predicting future diabetes.

Abdominal aortic aneurysms--glycaemic status and mortality

AIMS

The prevalence of diabetes mellitus (DM) and mortality with respect to glycaemic status in patients with abdominal aortic aneurysms (AAA) was evaluated. Glycaemic status was assessed by an oral glucose tolerance test (OGTT) and by HbA1c.

METHODS

Sixty-six patients with AAA admitted to the vascular surgery unit for elective surgery between October 2006 and September 2007 were included. Seven patients had previously known DM. OGTT and HbA1c results were available from 58 patients. The patients were categorized as having DM, prediabetes and normoglycaemia according to the WHO's and American Diabetes Association's criteria.

RESULTS

The prevalence of newly diagnosed DM according to the OGTT and HbA1c results were 12% and 14% respectively. Mean follow-up time was 68 months and all-cause mortality 43%. HbA1c was an independent predictor for mortality in the DM category. Hazard ratio of all-cause mortality in the DM category defined by the HbA1c values was 6.35, 95% [CI 1.49-27.1]; p = 0.01.

CONCLUSIONS

DM defined by HbA1c ≥ 6.5% is an important determinant of mortality following surgical treatment for AAA. Half the patients with AAA and DM were unaware of their DM diagnosis. All patients with AAA should be tested for DM using HbA1c. The results should be confirmed in a larger prospective study.

Analysis of the Japanese Diabetes Risk Score and fatty liver markers for incident diabetes in a Japanese cohort

AIMS

We examined the effectiveness of the Japanese Diabetes Risk Score (JPDRISC) and fatty liver markers for predicting incident diabetes.

METHODS

We created the JPDRISC. The study periods I and II were January 2007 to May 2009 and June 2009 to December 2011, respectively. A total of 2084 people (1389 men, 695 women; mean age: 46 years) were included. People with diabetes in the Period I and those with ethanol intake >140 g/week were excluded. A total of 1515 people were included. Fatty liver using ultrasonography scores (FLUS) were assigned.

RESULTS

The mean observation period was 26.3 months, and 24 people had developed diabetes between the Periods I and II. In logistic regression analysis, the JPDRISC (OR=1.197, 95% C.I.: 1.062-1.350, p=0.003) and FLUS (OR=2.591, 95% C.I.: 1.411-4.758, p=0.002) in the Period I were independent determinants of incident diabetes. In receiver operating characteristic analysis, sensitivity and specificity for incident diabetes were 0.885 and 0.536, respectively, in people with both FLUS≥1 and the total JPDRISC≥6 in the Period I. The sensitivity was better than the JPDRISC alone (sensitivity 0.696) and FLUS alone (sensitivity 0.750).

CONCLUSIONS

JPDRISC and FLUS were independently associated with incident diabetes and their combination is useful.

Validating prediction scales of type 2 diabetes mellitus in Spain: the SPREDIA-2 population-based prospective cohort study protocol

INTRODUCTION

The incidence of type 2 diabetes mellitus (T2DM) is increasing worldwide. When diagnosed, many patients already have organ damage or advance subclinical atherosclerosis. An early diagnosis could allow the implementation of lifestyle changes and treatment options aimed at delaying the progression of the disease and to avoid cardiovascular complications. Different scores for identifying undiagnosed diabetes have been reported, however, their performance in populations of southern Europe has not been sufficiently evaluated. The main objectives of our study are: to evaluate the screening performance and cut-off points of the main scores that identify the risk of undiagnosed T2DM and prediabetes in a Spanish population, and to develop and validate our own predictive models of undiagnosed T2DM (screening model), and future T2DM (prediction risk model) after 5-year follow-up. As a secondary objective, we will evaluate the atherosclerotic burden of the population with undiagnosed T2DM.

METHODS AND ANALYSIS

Population-based prospective cohort study with baseline screening, to evaluate the performance of the FINDRISC, DANISH, DESIR, ARIC and QDScore, against the gold standard tests: Fasting plasma glucose, oral glucose tolerance and/or HbA1c. The sample size will include 1352 participants between the ages of 45 and 74 years.

ANALYSIS

sensitivity, specificity, positive predictive value, negative predictive value, likelihood ratio positive, likelihood ratio negative and receiver operating characteristic curves and area under curve. Binary logistic regression for the first 700 individuals (derivation) and last 652 (validation) will be performed. All analyses will be calculated with their 95% CI; statistical significance will be p<0.05.

ETHICS AND DISSEMINATION

The study protocol has been approved by the Research Ethics Committee of the Carlos III Hospital (Madrid). The score performance and predictive model will be presented in medical conferences, workshops, seminars and round table discussions. Furthermore, the predictive model will be published in a peer-reviewed medical journal to further increase the exposure of the scores.

Predictive modeling for incident and prevalent diabetes risk evaluation

With half of individuals with diabetes undiagnosed worldwide and a projected 55% increase of the population with diabetes by 2035, the identification of undiagnosed and high-risk individuals is imperative. Multivariable diabetes risk prediction models have gained popularity during the past two decades. These have been shown to predict incident or prevalent diabetes through a simple and affordable risk scoring system accurately. Their development requires cohort or cross-sectional type studies with a variable combination, number and definition of included risk factors, with their performance chiefly measured by discrimination and calibration. Models can be used in clinical and public health settings. However, the impact of their use on outcomes in real-world settings needs to be evaluated before widespread implementation.

Factors affecting the decline in incidence of diabetes in the Diabetes Prevention Program Outcomes Study (DPPOS)

During the first 7 years of the Diabetes Prevention Program Outcomes Study (DPPOS), diabetes incidence rates, when compared with the Diabetes Prevention Program (DPP), decreased in the placebo (-42%) and metformin (-25%), groups compared with the rates in the intensive lifestyle intervention (+31%) group. Participants in the placebo and metformin groups were offered group intensive lifestyle intervention prior to entering the DPPOS. The following two hypotheses were explored to explain the rate differences: "effective intervention" (changes in weight and other factors due to intensive lifestyle intervention) and "exhaustion of susceptible" (changes in mean genetic and diabetes risk scores). No combination of behavioral risk factors (weight, physical activity, diet, smoking, and antidepressant or statin use) explained the lower DPPOS rates of diabetes progression in the placebo and metformin groups, whereas weight gain was the factor associated with higher rates of progression in the intensive lifestyle intervention group. Different patterns in the average genetic risk score over time were consistent with exhaustion of susceptibles. Results were consistent with exhaustion of susceptibles for the change in incidence rates, but not the availability of intensive lifestyle intervention to all persons before the beginning of the DPPOS. Thus, effective intervention did not explain the lower diabetes rates in the DPPOS among subjects in the placebo and metformin groups compared with those in the DPP.

Challenges of diabetes prevention in the real world: results and lessons from the Melbourne Diabetes Prevention Study

OBJECTIVE

To assess effectiveness and implementability of the public health programme Life! Taking action on diabetes in Australian people at risk of developing type 2 diabetes.

RESEARCH DESIGN AND METHODS

Melbourne Diabetes Prevention Study (MDPS) was a unique study assessing effectiveness of Life! that used a randomized controlled trial design. Intervention participants with AUSDRISK score ≥15 received 1 individual and 5 structured 90 min group sessions. Controls received usual care. Outcome measures were obtained for all participants at baseline and 12 months and, additionally, for intervention participants at 3 months. Per protocol set (PPS) and intention to treat (ITT) analyses were performed.

RESULTS

PPS analyses were considered more informative from our study. In PPS analyses, intervention participants significantly improved in weight (-1.13 kg, p=0.016), waist circumference (-1.35 cm, p=0.044), systolic (-5.2 mm Hg, p=0.028) and diastolic blood pressure (-3.2 mm Hg, p=0.030) compared with controls. Based on observed weight change, estimated risk of developing diabetes reduced by 9.6% in the intervention and increased by 3.3% in control participants. Absolute 5-year cardiovascular disease (CVD) risk reduced significantly for intervention participants by 0.97 percentage points from 9.35% (10.4% relative risk reduction). In control participants, the risk increased by 0.11 percentage points (1.3% relative risk increase). The net effect for the change in CVD risk was -1.08 percentage points of absolute risk (p=0.013).

CONCLUSIONS

MDPS effectively reduced the risk of diabetes and CVD, but the intervention effect on weight and waist reduction was modest due to the challenges in recruiting high-risk individuals and the abbreviated intervention.

Metabolic syndrome and lifestyle modification

A clustering of metabolic abnormalities such as dyslipidemia, hypertension, and diabetes mellitus, all of which are major risk factors for cardiovascular disease (CVD), occurs more often than by chance. Numerous epidemiological studies, as well as basic researches, have revealed that visceral fat accumulation is closely involved in this risk clustering. This morbid condition is now well recognized as the metabolic syndrome. The concept of the metabolic syndrome, i.e., the involvement of visceral adiposity in the clustering of CVD risk factors, implies that an effective CVD risk reduction will be accomplished by an intervention to reduce visceral fat deposits. The primary strategy of the intervention is lifestyle modification, which can be put into practice in healthcare fields, without necessity of medical treatment. Now that CVD is a leading global health burden, the metabolic syndrome attracts increasing attention in the world. To take global action against the syndrome, several working groups developed its internationally unified diagnostic criteria. Most recently, the International Diabetes Federation (IDF) and the American Heart Association/National Heart, Lung, and Blood Institute (AHA/NHLBI) jointly proposed the criteria, although some cautions will be needed in their practical use. In this review, we mainly focus on the findings observed in clinical and epidemiological studies, to discuss a practical strategy of the management of the metabolic syndrome in healthcare fields.

Update of the German Diabetes Risk Score and external validation in the German MONICA/KORA study

AIMS

Several published diabetes prediction models include information about family history of diabetes. The aim of this study was to extend the previously developed German Diabetes Risk Score (GDRS) with family history of diabetes and to validate the updated GDRS in the Multinational MONItoring of trends and determinants in CArdiovascular Diseases (MONICA)/German Cooperative Health Research in the Region of Augsburg (KORA) study.

METHODS

We used data from the European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam study for extending the GDRS, including 21,846 participants. Within 5 years of follow-up 492 participants developed diabetes. The definition of family history included information about the father, the mother and/or sibling/s. Model extension was evaluated by discrimination and reclassification. We updated the calculation of the score and absolute risks. External validation was performed in the MONICA/KORA study comprising 11,940 participants with 315 incident cases after 5 years of follow-up.

RESULTS

The basic ROC-AUC of 0.856 (95%-CI: 0.842-0.870) was improved by 0.007 (0.003-0.011) when parent and sibling history was included in the GDRS. The net reclassification improvement was 0.110 (0.072-0.149), respectively. For the updated score we demonstrated good calibration across all tenths of risk. In MONICA/KORA, the ROC-AUC was 0.837 (0.819-0.855); regarding calibration we saw slight overestimation of absolute risks.

CONCLUSIONS

Inclusion of the number of diabetes-affected parents and sibling history improved the prediction of type 2 diabetes. Therefore, we updated the GDRS algorithm accordingly. Validation in another German cohort study showed good discrimination and acceptable calibration for the vast majority of individuals.

Discovery of biomarkers for glycaemic deterioration before and after the onset of type 2 diabetes: rationale and design of the epidemiological studies within the IMI DIRECT Consortium

AIMS/HYPOTHESIS

The DIRECT (Diabetes Research on Patient Stratification) Study is part of a European Union Framework 7 Innovative Medicines Initiative project, a joint undertaking between four industry and 21 academic partners throughout Europe. The Consortium aims to discover and validate biomarkers that: (1) predict the rate of glycaemic deterioration before and after type 2 diabetes onset; (2) predict the response to diabetes therapies; and (3) help stratify type 2 diabetes into clearly definable disease subclasses that can be treated more effectively than without stratification. This paper describes two new prospective cohort studies conducted as part of DIRECT.

METHODS

Prediabetic participants (target sample size 2,200-2,700) and patients with newly diagnosed type 2 diabetes (target sample size ~1,000) are undergoing detailed metabolic phenotyping at baseline and 18 months and 36 months later. Abdominal, pancreatic and liver fat is assessed using MRI. Insulin secretion and action are assessed using frequently sampled OGTTs in non-diabetic participants, and frequently sampled mixed-meal tolerance tests in patients with type 2 diabetes. Biosamples include venous blood, faeces, urine and nail clippings, which, among other biochemical analyses, will be characterised at genetic, transcriptomic, metabolomic, proteomic and metagenomic levels. Lifestyle is assessed using high-resolution triaxial accelerometry, 24 h diet record, and food habit questionnaires.

CONCLUSIONS/INTERPRETATION

DIRECT will yield an unprecedented array of biomaterials and data. This resource, available through managed access to scientists within and outside the Consortium, will facilitate the development of new treatments and therapeutic strategies for the prevention and management of type 2 diabetes.

Evaluation of Finnish Diabetes Risk Score in screening undiagnosed diabetes and prediabetes among U.S. adults by gender and race: NHANES 1999-2010

Objective

To evaluate the performance of Finnish Diabetes Risk Score (FINDRISC) in detecting undiagnosed diabetes and prediabetes among U.S. adults by gender and race.

Methods

This cross-sectional analysis included participants (aged ≥20 years) from the National Health and Nutrition Examination Survey (NHANES) 1999–2010. Sensitivity, specificity, area under the receiver operating characteristic (ROC) curve and the optimal cutoff points for identifying undiagnosed diabetes and prediabetes were calculated for FINDRISC by gender and race/ethnicity.

Results

Among the 20,633 adults (≥20 years), 49.8% were women and 53.0% were non-Hispanic White. The prevalence of undiagnosed diabetes and prediabetes was 4.1% and 35.6%, respectively. FINDRISC was positively associated with the prevalence of diabetes (OR = 1.48 for 1 unit increase, p<0.001) and prediabetes (OR = 1.15 for 1 unit increase, p<0.001). The area under ROC for detecting undiagnosed diabetes was 0.75 for total population, 0.74 for men and 0.78 for women (p = 0.04); 0.76 for White, 0.76 for Black and 0.72 for Hispanics (p = 0.03 for White vs. Hispanics). The area under ROC for detecting prediabetes was 0.67 for total population, 0.66 for men and 0.70 for women (p<0.001); 0.68 for White, 0.67 for Black and 0.65 for Hispanics (p<0.001 for White vs. Hispanics). The optimal cutoff point was 10 (sensitivity = 0.75) for men and 12 (sensitivity = 0.72) for women for detecting undiagnosed diabetes; 9 (sensitivity = 0.61) for men and 10 (sensitivity = 0.69) for women for detecting prediabetes.

Conclusions

FINDRISC is a simple and non-invasive screening tool to identify individuals at high risk for diabetes in the U.S. adults.

Evaluation of the Finnish Diabetes Risk Score (FINDRISC) as a screening tool for the metabolic syndrome

OBJECTIVES

Traditionally, the Finnish Diabetes Risk Score (FINDRISC) questionnaire is a screening tool to estimate risk of type 2 diabetes. In this study, we evaluated the ability of FINDRISC to predict the development of the metabolic syndrome (MetS) in an Iranian population without diabetes and MetS.

METHODS

A total of 1,010 first-degree relatives of consecutive patients with type 2 diabetes, 30-70 years old (274 men and 736 women), without diabetes and MetS, were examined and followed up over 8.0 ± 1.6 years (mean ± SD) for MetS incidence. The incidence of MetS was examined across quartiles of FINDRISC, and a receiver operating characteristic (ROC) curve was plotted to assess the discrimination. At baseline and through follow-ups, participants underwent a standard 75 g 2-hour oral glucose tolerance test (OGTT). Data for determining FINDRISC were available from each participant.

RESULTS

During 8,089 person-years of follow-up, 69 men and 209 women without MetS and diabetes at baseline subsequently developed MetS. The incidence of MetS was 31.4 per 1000 person-years in men and 35.5 in women. The FINDRSC at baseline was significantly associated with MetS evolution. Participants in the top quartile of FINDRISC were 4.4 times more likely to develop MetS than those in the bottom quartile (rate ratio 4.4; 95% CI 2.7-7.0). The area under the ROC curve was 65.0% (95% CI 61.3-68.7).

CONCLUSION

The results of this study suggest that FINDRISC can be applied to detect MetS in a high-risk population.

Light to moderate alcohol consumption is protective for type 2 diabetes mellitus in normal weight and overweight individuals but not the obese

OBJECTIVE

To examine the association between alcohol consumption and risk of type 2 diabetes mellitus (T2DM) overall and by body mass index.

METHODS

Cross-sectional study of employed individuals. Daily alcohol intakes were calculated from a self-administered food frequency questionnaire by 5,512 Maori, Pacific Island, and European workers (3,992 men, 1520 women) aged 40 years and above.

RESULTS

There were 170 new cases of T2DM. Compared to the group with no alcohol consumption and adjusting for age, sex, and ethnicity, the group consuming alcohol had relative risks of T2DM of 0.23 (95% CI: 0.08, 0.65) in normal weight individuals, 0.38 (0.18, 0.81) in overweight individuals, and 0.99 (0.59, 1.67) in obese individuals. After further adjusting for total cholesterol, HDL-cholesterol, triglycerides, smoking habit, physical activity, socioeconomic status, body mass index, and hypertension, the relative risks of T2DM were 0.16 (0.05, 0.50) in normal weight individuals, 0.43 (0.19, 0.97) in overweight individuals, and 0.92 (0.52, 1.60) in overweight individuals. Across the categories of alcohol consumption, there was an approximate U-shaped relationship for new cases of T2DM. There was no significant association between alcohol consumption and IGT.

CONCLUSIONS

Alcohol consumption was protective against diagnosis of T2DM in normal and overweight individuals but not in the obese.

The potential of novel biomarkers to improve risk prediction of type 2 diabetes

The incidence of type 2 diabetes can be reduced substantially by implementing preventive measures in high-risk individuals, but this requires prior knowledge of disease risk in the individual. Various diabetes risk models have been designed, and these have all included a similar combination of factors, such as age, sex, obesity, hypertension, lifestyle factors, family history of diabetes and metabolic traits. The accuracy of prediction models is often assessed by the area under the receiver operating characteristic curve (AROC) as a measure of discrimination, but AROCs should be complemented by measures of calibration and reclassification to estimate the incremental value of novel biomarkers. This review discusses the potential of novel biomarkers to improve model accuracy. The range of molecules that serve as potential predictors of type 2 diabetes includes genetic variants, RNA transcripts, peptides and proteins, lipids and small metabolites. Some of these biomarkers lead to a statistically significant increase of model accuracy, but their incremental value currently seems too small for routine clinical use. However, only a fraction of potentially relevant biomarkers have been assessed with regard to their predictive value. Moreover, serial measurements of biomarkers may help determine individual risk. In conclusion, current risk models provide valuable tools of risk estimation, but perform suboptimally in the prediction of individual diabetes risk. Novel biomarkers still fail to have a clinically applicable impact. However, more efficient use of biomarker data and technological advances in their measurement in clinical settings may allow the development of more accurate predictive models in the future.

Non-invasive risk scores for prediction of type 2 diabetes (EPIC-InterAct): a validation of existing models

BACKGROUND

The comparative performance of existing models for prediction of type 2 diabetes across populations has not been investigated. We validated existing non-laboratory-based models and assessed variability in predictive performance in European populations.

METHODS

We selected non-invasive prediction models for incident diabetes developed in populations of European ancestry and validated them using data from the EPIC-InterAct case-cohort sample (27,779 individuals from eight European countries, of whom 12,403 had incident diabetes). We assessed model discrimination and calibration for the first 10 years of follow-up. The models were first adjusted to the country-specific diabetes incidence. We did the main analyses for each country and for subgroups defined by sex, age (<60 years vs ≥60 years), BMI (<25 kg/m(2)vs ≥25 kg/m(2)), and waist circumference (men <102 cm vs ≥102 cm; women <88 cm vs ≥88 cm).

FINDINGS

We validated 12 prediction models. Discrimination was acceptable to good: C statistics ranged from 0·76 (95% CI 0·72-0·80) to 0·81 (0·77-0·84) overall, from 0·73 (0·70-0·76) to 0·79 (0·74-0·83) in men, and from 0·78 (0·74-0·82) to 0·81 (0·80-0·82) in women. We noted significant heterogeneity in discrimination (pheterogeneity<0·0001) in all but one model. Calibration was good for most models, and consistent across countries (pheterogeneity>0·05) except for three models. However, two models overestimated risk, DPoRT by 34% (95% CI 29-39%) and Cambridge by 40% (28-52%). Discrimination was always better in individuals younger than 60 years or with a low waist circumference than in those aged at least 60 years or with a large waist circumference. Patterns were inconsistent for BMI. All models overestimated risks for individuals with a BMI of <25 kg/m(2). Calibration patterns were inconsistent for age and waist-circumference subgroups.

INTERPRETATION

Existing diabetes prediction models can be used to identify individuals at high risk of type 2 diabetes in the general population. However, the performance of each model varies with country, age, sex, and adiposity.

FUNDING

The European Union.

Finnish Diabetes Risk Score to predict type 2 diabetes in the Isfahan diabetes prevention study

AIM

The strong association between the Finnish Diabetes Risk Score (FINDRISC) and risk of diabetes reported in European populations cannot necessarily be generalized to other populations. The aim of this study was to evaluate the ability of FINDRISC to predict progression to diabetes in an Iranian population without diabetes.

METHODS

A total of 1537 first-degree relatives (FDR) without diabetes of consecutive people with type 2 diabetes 30-70 years old (376 men and 1161 women) were examined and followed for a mean (SD) of 7.8 (1.7) years for diabetes incidence. We examined the incidence of diabetes across quartiles of FINDRISC and plotted a receiver operating characteristic (ROC) curve to assess discrimination. At baseline and through follow-up, participants underwent a standard 75-g 2-h oral glucose tolerance test. Data for the FINDRISC were available from each participant.

RESULTS

During 12,046 person-years of follow-up, 41 men and 154 women developed diabetes. The incidence of type 2 diabetes was 14.0 per 1000 person-years in men and 16.9 in women. Those in the top quartile of FINDRISC were 21.7 times more likely to develop diabetes than those in the bottom quartile (relative risk 21.7; 95% CI 9.90, 47.39). The area under the ROC was 75.1% (95% CI 71.3, 78.8).

CONCLUSIONS

The results of this study show that FINDRISC is a robust predictor of type 2 diabetes in high-risk individuals in Iran.

Family history of diabetes is associated with higher risk for prediabetes: a multicentre analysis from the German Center for Diabetes Research

AIMS/HYPOTHESIS

Prediabetes is a collective term for different subphenotypes (impaired glucose tolerance [IGT] and/or impaired fasting glucose [IFG]) with different pathophysiologies. A positive family history for type 2 diabetes (FHD) is associated with increased risk for type 2 diabetes. We assumed that it would also associate with prediabetes, but wondered whether all subphenotypes are related to a positive family history.

METHODS

In a study population of 8,106 non-diabetic individuals of European origin collected from four study centres (normal glucose tolerance, NGT n = 5,482, IFG and/or IGT n = 2,624), we analysed whether having at least one first degree relative with diabetes is associated with prediabetes. The analyses were performed using the same models in each population separately. Afterwards, a meta-analysis was performed.

RESULTS

FHD was significantly associated with the risk for prediabetes (IFG and/or IGT, OR 1.40; 95% CI 1.27, 1.54). This association remained significant in multivariable logistic regression models including sex, age and BMI (OR 1.26; 95% CI 1.14, 1.40). When different prediabetic outcomes were considered separately, the association was found for isolated IFG (OR 1.37; 95% CI 1.20, 1.57), isolated IGT (OR 1.25; 95% CI 1.07, 1.46) as well as for the combination IFG+IGT (OR 1.64; 95% CI 1.40, 1.93). After stratification on BMI, association between FHD and prediabetes was seen only in non-obese individuals (BMI < 30 kg/m(2)).

CONCLUSIONS/INTERPRETATION

We found that FHD is an important risk factor for prediabetes, especially for combined IGT and IFG. Its relevance seems to be more evident in the non-obese.

Type 2 Diabetes Risk among Asian Indians in the US: A Pilot Study

The purpose of this pilot study was to investigate type 2 diabetes risk among Asian Indians of Kerala ethnicity living in a West Texas County of the USA. The study used a descriptive correlational design with thirty-seven adult nondiabetic Asian Indian subjects between 20 and 70 years of age. The measurement included nonbiochemical indices of obesity, family history of type 2 diabetes, length of immigration in the US, history of hypertension, physical activity pattern, and fruit and vegetable intake. The majority of the subjects showed an increased nonbiochemical indices corresponding with overweight and obesity, placing them at risk for type 2 diabetes and associated cardiovascular complications. The physical activity pattern indicated a sedentary lifestyle. The decreased physical activity was associated with a higher Body Mass Index (BMI) and body fat percentage; length of residence in the US greater than 10 years was associated with increased body fat percentage and BMI; family history of type 2 diabetes was associated with an increase in body fat percentage. Fruit and vegetable intake pattern was not associated with a risk for type 2 diabetes. Further studies are recommended for risk surveillance among Asian Indian population living in the US.

Risk scores for diabetes and impaired glycaemia in the Middle East and North Africa

AIMS

To develop risk scores for diabetes and diabetes or impaired glycaemia for individuals living in the Middle East and North Africa region. In addition, to derive national risk scores for Algeria, Saudi Arabia and the United Arab Emirates and to compare the performance of the regional risk scores with the national risk scores.

METHODS

An opportunistic sample of 6588 individuals aged 30-75 years was screened. Screening consisted of a questionnaire and a clinical examination including measurement of HbA(1c). Two regional risk scores and national risk scores for each of the three countries were derived separately by stepwise backwards multiple logistic regression with diabetes [HbA(1c) ≥ 48 mmol/mol (≥ 6.5%)] and diabetes or impaired glycaemia [HbA(1c) ≥ 42 mmol/mol (≥ 6.0%)] as outcome. The performance of the regional and national risk scores was compared in data from each country by receiver operating characteristic analysis.

RESULTS

The eight risk scores all included age and BMI, while additional variables differed between the scores. The areas under the receiver operating characteristic curves were between 0.67 and 0.70, and for sensitivities approximately 75%; specificities varied between 50% and 57%. The regional and the national risk scores performed equally well in the three national samples.

CONCLUSIONS

Two regional risk scores for diabetes and diabetes or impaired glycaemia applicable to the Middle East and North Africa region were identified. The regional risk scores performed as well as the national risk scores derived in the same manner.

Prevalence of undiagnosed abnormal glucose tolerance in adult patients cared for by general practitioners in Hungary. Results of a risk-stratified screening based on FINDRISC questionnaire

BACKGROUND

The prevalence of type 2 diabetes mellitus is rapidly increasing, worldwide and also in Hungary. Timely diagnosis and early treatment could be aided by targeted screening. Recognizing this, the Hungarian Diabetes Association initiated a risk-stratified screening with the involvement of primary care physicians.

MATERIAL/METHODS

In the first phase of screening, the FINDRISC questionnaire was completed, followed by an oral glucose tolerance test (OGTT) for those with a score of ≥12. Between September 1, 2010 and March 31, 2011, 70,432 non-diabetic adults, who visited their general practitioners for any reason, were involved in the screening. Of these, 68,476 questionnaires proved to be suitable for processing.

RESULTS

From the questionnaires, 28,077 (41.0%) had a score of ≥12. A valid OGTT was performed in 22,846 cases; of this group 3,217 subjects (14.1%) had elevated fasting glucose levels, 5,663 (24.8%) had impaired glucose tolerance, and 1,750 (7.6%) had manifest, previously undiagnosed, diabetes mellitus. Overall, from the valid OGTT group, 46.5% subjects had some degree of glucose intolerance.

CONCLUSIONS

Based on the FINDRISC questionnaire, the risk-stratified screening for diabetes mellitus proved to be simple and cost-effective method for the early detection of carbohydrate metabolism disorders. Using this method, the prevalence rate of previously undiagnosed abnormal glucose tolerance was high in adult patients cared for by general practitioners in Hungary.

Development of a new scoring system for predicting the 5 year incidence of type 2 diabetes in Japan: the Toranomon Hospital Health Management Center Study 6 (TOPICS 6)

AIMS/HYPOTHESIS

The aims of this study were to assess the clinical significance of introducing HbA(1c) into a risk score for diabetes and to develop a scoring system to predict the 5 year incidence of diabetes in Japanese individuals.

METHODS

The study included 7,654 non-diabetic individuals aged 40-75 years. Incident diabetes was defined as fasting plasma glucose (FPG) ≥7.0 mmol/l, HbA(1c) ≥6.5% (48 mmol/mol) or self-reported clinician-diagnosed diabetes. We constructed a risk score using non-laboratory assessments (NLA) and evaluated improvements in risk prediction by adding elevated FPG, elevated HbA(1c) or both to NLA.

RESULTS

The discriminative ability of the NLA score (age, sex, family history of diabetes, current smoking and BMI) was 0.708. The difference in discrimination between the NLA + FPG and NLA + HbA(1c) scores was non-significant (0.836 vs 0.837; p = 0.898). A risk score including family history of diabetes, smoking, obesity and both FPG and HbA(1c) had the highest discrimination (0.887, 95% CI 0.871, 0.903). At an optimal cut-off point, sensitivity and specificity were high at 83.7% and 79.0%, respectively. After initial screening using NLA scores, subsequent information on either FPG or HbA(1c) resulted in a net reclassification improvement of 42.7% or 52.3%, respectively (p < 0.0001). When both were available, net reclassification improvement and integrated discrimination improvement were further improved at 56.7% (95% CI 47.3%, 66.1%) and 10.9% (9.7%, 12.1%), respectively.

CONCLUSIONS/INTERPRETATION

Information on HbA(1c) or FPG levels after initial screening by NLA can precisely refine diabetes risk reclassification.

Prediction models for risk of developing type 2 diabetes: systematic literature search and independent external validation study

OBJECTIVE

To identify existing prediction models for the risk of development of type 2 diabetes and to externally validate them in a large independent cohort.

DATA SOURCES

Systematic search of English, German, and Dutch literature in PubMed until February 2011 to identify prediction models for diabetes.

DESIGN

Performance of the models was assessed in terms of discrimination (C statistic) and calibration (calibration plots and Hosmer-Lemeshow test).The validation study was a prospective cohort study, with a case cohort study in a random subcohort.

SETTING

Models were applied to the Dutch cohort of the European Prospective Investigation into Cancer and Nutrition cohort study (EPIC-NL).

PARTICIPANTS

38,379 people aged 20-70 with no diabetes at baseline, 2506 of whom made up the random subcohort.

OUTCOME MEASURE

Incident type 2 diabetes.

RESULTS

The review identified 16 studies containing 25 prediction models. We considered 12 models as basic because they were based on variables that can be assessed non-invasively and 13 models as extended because they additionally included conventional biomarkers such as glucose concentration. During a median follow-up of 10.2 years there were 924 cases in the full EPIC-NL cohort and 79 in the random subcohort. The C statistic for the basic models ranged from 0.74 (95% confidence interval 0.73 to 0.75) to 0.84 (0.82 to 0.85) for risk at 7.5 years. For prediction models including biomarkers the C statistic ranged from 0.81 (0.80 to 0.83) to 0.93 (0.92 to 0.94). Most prediction models overestimated the observed risk of diabetes, particularly at higher observed risks. After adjustment for differences in incidence of diabetes, calibration improved considerably.

CONCLUSIONS

Most basic prediction models can identify people at high risk of developing diabetes in a time frame of five to 10 years. Models including biomarkers classified cases slightly better than basic ones. Most models overestimated the actual risk of diabetes. Existing prediction models therefore perform well to identify those at high risk, but cannot sufficiently quantify actual risk of future diabetes.

One risk assessment tool for cardiovascular disease, type 2 diabetes, and chronic kidney disease

OBJECTIVE

Individuals at high risk for chronic cardiometabolic disease (cardiovascular disease [CVD], type 2 diabetes, and chronic kidney disease [CKD]) share many risk factors and would benefit from early intervention. We developed a nonlaboratory-based risk-assessment tool for identification of people at high cardiometabolic disease risk.

RESEARCH DESIGN AND METHODS

Data of three population-based cohorts from different regions of the Netherlands were merged. Participants were 2,840 men and 3,940 women, white, aged 28-85 years, free from CVD, type 2 diabetes, and CKD diagnosis at baseline. The outcome was developing cardiometabolic disease during 7 years follow-up.

RESULTS

Age, BMI, waist circumference, antihypertensive treatment, smoking, family history of myocardial infarction or stroke, and family history of diabetes were significant predictors, whereas former smoking, history of gestational diabetes, and use of lipid-lowering medication were not. The models showed acceptable calibration (Hosmer and Lemeshow statistics, P > 0.05) and discrimination (area under the receiver operating characteristic [ROC] curve 0.82 [95% CI 0.81-0.83] for women and 0.80 [0.78-0.82] for men). Discrimination of individual outcomes was lowest for diabetes (area under the ROC curve 0.70 for men and 0.73 for women) and highest for CVD mortality (0.83 for men and 0.85 for women).

CONCLUSIONS

We demonstrate that a single risk stratification tool can identify people at high risk for future CVD, type 2 diabetes, and/or CKD. The present risk-assessment tool can be used for referring the highest risk individuals to health care for further (multivariable) risk assessment and may as such serve as an important part of prevention programs targeting chronic cardiometabolic disease.

Risk models and scores for type 2 diabetes: systematic review

OBJECTIVE

To evaluate current risk models and scores for type 2 diabetes and inform selection and implementation of these in practice.

DESIGN

Systematic review using standard (quantitative) and realist (mainly qualitative) methodology. Inclusion criteria Papers in any language describing the development or external validation, or both, of models and scores to predict the risk of an adult developing type 2 diabetes.

DATA SOURCES

Medline, PreMedline, Embase, and Cochrane databases were searched. Included studies were citation tracked in Google Scholar to identify follow-on studies of usability or impact.

DATA EXTRACTION

Data were extracted on statistical properties of models, details of internal or external validation, and use of risk scores beyond the studies that developed them. Quantitative data were tabulated to compare model components and statistical properties. Qualitative data were analysed thematically to identify mechanisms by which use of the risk model or score might improve patient outcomes.

RESULTS

8864 titles were scanned, 115 full text papers considered, and 43 papers included in the final sample. These described the prospective development or validation, or both, of 145 risk prediction models and scores, 94 of which were studied in detail here. They had been tested on 6.88 million participants followed for up to 28 years. Heterogeneity of primary studies precluded meta-analysis. Some but not all risk models or scores had robust statistical properties (for example, good discrimination and calibration) and had been externally validated on a different population. Genetic markers added nothing to models over clinical and sociodemographic factors. Most authors described their score as "simple" or "easily implemented," although few were specific about the intended users and under what circumstances. Ten mechanisms were identified by which measuring diabetes risk might improve outcomes. Follow-on studies that applied a risk score as part of an intervention aimed at reducing actual risk in people were sparse.

CONCLUSION

Much work has been done to develop diabetes risk models and scores, but most are rarely used because they require tests not routinely available or they were developed without a specific user or clear use in mind. Encouragingly, recent research has begun to tackle usability and the impact of diabetes risk scores. Two promising areas for further research are interventions that prompt lay people to check their own diabetes risk and use of risk scores on population datasets to identify high risk "hotspots" for targeted public health interventions.

The performance of risk scores and hemoglobin A1c to find undiagnosed diabetes with isolated postload hyperglycemia

The aim of this study is to develop strategies to screen diabetic subjects with isolated postload hyperglycemia (IPH) in Chinese population. We included 1175 adult subjects who did not report diabetes were included. Diabetes was diagnosed by oral glucose tolerance tests. IPH was defined as fasting plasma glucose (FPG)<7 mmol/l and 2-hour post-load plasma glucose (2hPG) greater than 11.1 mmol/l. Using FPG criteria, only 59.8% of diabetic subjects were not identified, showing a poor agreement between FPG and 2hPG criteria (kappa 0.294). Age, FPG, total cholesterol, triglycerides, blood pressure, body mass index, HbA1c and medication for hypertension were associated factors for IPH. Four scores were constructed using all these factors, age and blood test results, age and HbA1c, and data from non-invasive examinations, respectively. The area under the ROC curve were 0.9296(95%CI 0.8948-0.9643), 0.9111(95%CI 0.8713-0.9508), 0.8902(95%CI 0.8341-0.9646), 0.8924(95%CI 0.7835-0.8753), and 0.8654(95%CI 0.7963-0.9345) for score 1, 2, 3, 4, and HbA1c, respectively. The sensitivity of all four risk scores to detect IPH was better than that of impaired fasting glucose (IFG). The sensitivity and specificity of HbA1c at cutoff 6.2% for detecting IPH was also better than that of IFG. In conclusion, the risk scores and HbA1c are useful to identify subjects with undiagnosed IPH, with better performance than IFG.