Risk prediction models for the development of diabetes in Mauritian Indians
Department of Public Health, University of Helsinki, FIN-00014 Helsinki, Finland. Diabetic Medicine
(Impact Factor: 3.12).
10/2009; 26(10):996-1002. DOI: 10.1111/j.1464-5491.2009.02810.x
To develop risk prediction models of future diabetes in Mauritian Indians.
Three thousand and ninety-four Mauritian Indians (1141 men, aged 20-65 years) without diabetes in 1987 or 1992 were followed up to 1992 or 1998. Subjects underwent repeated oral glucose tolerance tests and diabetes was diagnosed according to 2006 World Health Organization/International Diabetes Federation criteria. Cox regression models for interval censored data were performed using data from 1544 randomly selected participants. Predicted probabilities for diabetes were calculated and validated in the remaining 1550 subjects.
Over 11 years of follow-up, there were 511 cases of diabetes. Among variables tested, family history of diabetes, obesity (body mass index, waist circumference) and glucose were significant predictors of diabetes. Predicted probabilities derived from a simple model fitted with sex, family history of diabetes and obesity ranged from 0.05 to 0.64 in men and 0.03 to 0.49 in women. To predict the onset of diabetes, area under the receiver operating characteristic (ROC) curve (AROC) of predicted probabilities was 0.62 (95% confidence interval, 0.56-0.68) in men and 0.64 (0.59-0.69) in women. At a cut-off point of 0.12, the sensitivity and specificity were 0.72 (0.71-0.74) and 0.47 (0.45-0.49) in men and 0.77 (0.75-0.78) and 0.50 (0.48-0.52) in women, respectively. Addition of fasting plasma glucose (FPG) to the model improved the prediction slightly [AROC curve 0.70 (0.65-0.76) in men, 0.71 (0.67-0.76) in women].
A diabetes prediction model based on obesity and family history yielded moderate discrimination in Mauritian Indians, which was slightly inferior to the model with the FPG but may be useful in low-income countries to promote identification of people at high risk of diabetes.
Figures in this publication
Available from: Gary S Collins
- "In terms of geography, all but two risk prediction models were developed using patient data from single countries [38,40]. Eight articles (21%) were from the USA [31,34,36,39,43,56,57,59], thirteen articles (33%) were from Europe [23-25,32,33,35,40,42,46,52,54,55], thirteen articles (33%) were from Asia [26,27,29,37,41,44,45,47-49,51,60], two were from Africa [30,53], one was from Australia  and one was from Brazil . "
[Show abstract] [Hide abstract]
ABSTRACT: The World Health Organisation estimates that by 2030 there will be approximately 350 million people with type 2 diabetes. Associated with renal complications, heart disease, stroke and peripheral vascular disease, early identification of patients with undiagnosed type 2 diabetes or those at an increased risk of developing type 2 diabetes is an important challenge. We sought to systematically review and critically assess the conduct and reporting of methods used to develop risk prediction models for predicting the risk of having undiagnosed (prevalent) or future risk of developing (incident) type 2 diabetes in adults.
We conducted a systematic search of PubMed and EMBASE databases to identify studies published before May 2011 that describe the development of models combining two or more variables to predict the risk of prevalent or incident type 2 diabetes. We extracted key information that describes aspects of developing a prediction model including study design, sample size and number of events, outcome definition, risk predictor selection and coding, missing data, model-building strategies and aspects of performance.
Thirty-nine studies comprising 43 risk prediction models were included. Seventeen studies (44%) reported the development of models to predict incident type 2 diabetes, whilst 15 studies (38%) described the derivation of models to predict prevalent type 2 diabetes. In nine studies (23%), the number of events per variable was less than ten, whilst in fourteen studies there was insufficient information reported for this measure to be calculated. The number of candidate risk predictors ranged from four to sixty-four, and in seven studies it was unclear how many risk predictors were considered. A method, not recommended to select risk predictors for inclusion in the multivariate model, using statistical significance from univariate screening was carried out in eight studies (21%), whilst the selection procedure was unclear in ten studies (26%). Twenty-one risk prediction models (49%) were developed by categorising all continuous risk predictors. The treatment and handling of missing data were not reported in 16 studies (41%).
We found widespread use of poor methods that could jeopardise model development, including univariate pre-screening of variables, categorisation of continuous risk predictors and poor handling of missing data. The use of poor methods affects the reliability of the prediction model and ultimately compromises the accuracy of the probability estimates of having undiagnosed type 2 diabetes or the predicted risk of developing type 2 diabetes. In addition, many studies were characterised by a generally poor level of reporting, with many key details to objectively judge the usefulness of the models often omitted.
Available from: Inger Njølstad
[Show abstract] [Hide abstract]
ABSTRACT: Aims: To determine the gender-specific incidence and risk factors of type-2 diabetes mellitus (T2DM) in a general population. Methods: The study is based on 12,431 men and 13,737 women aged 25-98 years, attending the Tromsø Study in 1994 and followed through 2005, who did not have diabetes when entering the study. Sex-specific hazard ratios were estimated from Cox proportional hazard models. Results: A total of 522 cases of T2DM were registered, 308 among men and 214 among women. The age-standardised incidence rate was higher in men than in women, 2.6 (95% CI 2.32-2.90) and 1.6(95% CI 1.40-1.83) per 1000 person-years, respectively. In multivariate survival analysis, age, body mass index (BMI),triglycerides, high-density lipoprotein (HDL) cholesterol, hypertension, family history of diabetes, low education and smoking were independent predictors of T2DM in both genders (p<0.05). Total cholesterol and lack of leisure-time physical activity were independent predictors in men only. We found an interaction between HDL cholesterol and triglyceride levels(p<0.001) and between triglyceride levels and a positive family history of diabetes (p=.04). These interactions were independent of BMI. A positive family history combined with triglycerides in the highest tertile and BMI >25 kg/m(2) conveyed a 10-year risk of T2DM of 10% (95% CI 8-12%) vs. 0.2% (95% CI 0.08-0.31%) for the lowest risk group.Conclusions: A family history of diabetes, elevated BMI, and high triglyceride levels identifies independent of cardiovascular risk factors, a group with especially high risk of T2DM. [corrected]
Available from: Emilie Agardh
[Show abstract] [Hide abstract]
ABSTRACT: We conducted a systematic review and meta-analysis, the first to our knowledge, summarizing and quantifying the published evidence on associations between type 2 diabetes incidence and socio-economic position (SEP) (measured by educational level, occupation and income) worldwide and when sub-divided into high-, middle- and low-income countries.
Relevant case-control and cohort studies published between 1966 and January 2010 were searched in PubMed and EMBASE using the keywords: diabetes vs educational level, occupation or income. All identified citations were screened by one author, and two authors independently evaluated and extracted data from relevant publications. Risk estimates from individual studies were pooled using random-effects models quantifying the associations.
Out of 5120 citations, 23 studies, including 41 measures of association, were found to be relevant. Compared with high educational level, occupation and income, low levels of these determinants were associated with an overall increased risk of type 2 diabetes; [relative risk (RR) = 1.41, 95% confidence interval (CI): 1.28-1.51], (RR = 1.31, 95% CI: 1.09-1.57) and (RR = 1.40, 95% CI: 1.04-1.88), respectively. The increased risks were independent of the income levels of countries, although based on limited data in middle- and low-income countries.
The risk of getting type 2 diabetes was associated with low SEP in high-, middle- and low-income countries and overall. The strength of the associations was consistent in high-income countries, whereas there is a strong need for further investigation in middle- and low-income countries.
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.