This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kothari, V. Articles by Holman, R. R. Search for Related Content PubMed PubMed Citation Articles by Kothari, V. Articles by Holman, R. R. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Hazardous Substances DB CHOLESTEROL Medline Plus Health Information Diabetes Stroke Related Collections Lipids Primary and Secondary Stroke Prevention Risk Factors for Stroke

(Stroke. 2002;33:1776.)
© 2002 American Heart Association, Inc.

Original Contributions

UKPDS 60

Risk of Stroke in Type 2 Diabetes Estimated by the UK Prospective Diabetes Study Risk Engine

Viti Kothari, MSc; Richard J. Stevens, PhD; Amanda I. Adler, MD; Irene M. Stratton, MSc; Susan E. Manley, PhD; H. Andrew Neil, FRCP Rudy R. Holman, FRCP for the UK Prospective Diabetes Study Group

From the Diabetes Trials Unit (V.K., R.J.S., A.I.A., I.M.S., S.E.M., R.R.H.), Oxford Centre for Diabetes, Endocrinology and Metabolism, University of Oxford, Radcliffe Infirmary, and Department of Primary Health Care (H.A.N.), Institute of Health Sciences, Oxford, United Kingdom.

Correspondence to Dr Richard J. Stevens, Diabetes Trials Unit, Oxford Centre for Diabetes, Endocrinology and Metabolism, University of Oxford, Radcliffe Infirmary, Woodstock Rd, Oxford OX2 6HE, UK. E-mail richard.stevens@ dtu.ox.ac.uk

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion Appendix 1 References

 
Background and Purpose— People with type 2 diabetes are at elevated risk of stroke compared with those without diabetes. Relative risks have been examined in earlier work, but there is no readily available method for predicting the absolute risk of stroke in a diabetic individual. We developed mathematical models to estimate the risk of a first stroke using data from 4549 newly diagnosed type 2 diabetic patients enrolled in the UK Prospective Diabetes Study.

Methods— During 30 700 person-years of follow-up, 188 first strokes (52 fatal) occurred. Model fitting was carried out by maximum likelihood estimation using the Newton-Raphson method. Diagnostic plots were used to compare survival probabilities calculated by the model with those calculated using nonparametric methods.

Results— Variables included in the final model were duration of diabetes, age, sex, smoking, systolic blood pressure, total cholesterol to high-density lipoprotein cholesterol ratio and presence of atrial fibrillation. Not included in the model were body mass index, hemoglobin A1c, ethnicity, and ex-smoking status. The use of the model is illustrated with a hypothetical study power calculation.

Conclusions— This model forecasts the absolute risk of a first stroke in people with type 2 diabetes using variables readily available in routine clinical practice.

Key Words: blood pressure • cholesterol • diabetes mellitus • statistics • stroke

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion Appendix 1 References

 
In people with type 2 diabetes, there is a 2- to 5-fold increased risk for stroke compared with those without diabetes.^1,2 Risk factors for increased risk of stroke in the type 2 diabetic population include elevated blood pressure, smoking, age, male sex, atrial fibrillation (AF), and hyperglycemia.^1–9 In the general population, the relationship between lipid levels and stroke is complicated by an inverse association between total cholesterol and hemorrhagic stroke, ¹⁰ but depressed high-density lipoprotein (HDL) cholesterol, elevated ratio of total to HDL cholesterol (T:H), and elevated triglycerides have all been reported to increase overall risk of stroke.^11–13 In the diabetic population, elevated total cholesterol, decreased HDL cholesterol, and elevated triglycerides have all been identified as risk factors for stroke.¹⁴

Tools for calculation of cardiovascular risk are now frequently used to assist decisions about clinical management for both individuals and populations.¹⁵ Charts and computer programs have been developed to make calculations in the course of a patient-doctor visit, but typically use coronary heart disease risk as a surrogate for total cardiovascular risk.^16–19 The majority of the models currently in use for predicting vascular risk in individuals with diabetes use equations from the Framingham study, ^12,20–23 which included only 337 individuals with diabetes out of 5573.²⁴

Data from the 5102 patients with type 2 diabetes in the UK Prospective Diabetes Study (UKPDS) provide an opportunity to derive a diabetes-specific model for risk of stroke.^25,26 The model presented here is specifically designed for a type 2 diabetic population using variables available in routine clinical practice, including duration of diabetes and age at diagnosis.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion Appendix 1 References

 
Study Subjects
The UKPDS has been described elsewhere.²⁵ Briefly, between 1977 and 1991, general practitioners in the catchment areas of 23 participating UKPDS hospitals were asked to refer all patients age 25 to 65 years presenting with newly diagnosed diabetes. There were 5102 patients recruited to the study, who met inclusion criteria listed previously.²⁵ Patients with a myocardial infarction within the last year, or with more than 1 vascular episode, were excluded.²⁵

For this analysis, data from 4549 patients of white, Afro-Caribbean, or Asian-Indian ethnic groups were included. Excluded were 39 patients of other ethnic groups; 37 with stroke before diagnosis of diabetes; 215 with missing data for blood pressure, lipids, or electrocardiography; and 262 with follow-up times too short (<4 years) for the model-fitting process. Characteristics of those included in the model are shown in Table 1.

View this table: [in this window] [in a new window]   Table 1. Characteristics of Subjects: Mean±sd, or n (Percentage)

Patients had blood pressure and biochemical measurements, including hemoglobin (Hb) A_1c, lipid, and lipoprotein fractions recorded at diagnosis of diabetes, at the end of a 3- to 4-month dietary run-in period, and then at annual intervals.²⁵ HbA_1c was measured by high-performance liquid chromatography (Biorad Diamat Automated Glycosylated Hemoglobin Analyzer), nondiabetic range 4.5% to 6.2%.^26,27 Cholesterol measurements were within the limits for accuracy of the lipid standardization program of the Centers for Disease Control and Prevention (CDC) in Atlanta, Ga, but when compared with the CDC reference laboratory in the United Kingdom were 2.5% lower and 1.2% lower (total cholesterol and HDL cholesterol, respectively).^27–29 Systolic blood pressure (SBP) was recorded each year as the mean of three measurements, with a coefficient of variation <15%, taken at the same visit.⁴ AF was identified clinically or by electrocardiography.

End Points
Stroke in the UKPDS is defined as a neurological deficit with symptoms or signs lasting 1 month or more.²⁵ No distinction was made between ischemic, embolic, and hemorrhagic strokes. In patients with multiple strokes, only the first stroke is considered here. The model treats death from causes other than stroke as censored.

Statistical Model and Methods
The mathematical form of the model and its derivation have been described elsewhere.³⁰ Table 2 shows the variables considered for inclusion in the model. Model building then proceeded stepwise using likelihood ratio tests for significance. The following well-documented risk factors for stroke were tested for inclusion at a significance level of 0.05: sex, age at diagnosis of diabetes, SBP, atrial fibrillation, smoking, and ratio of total:HDL cholesterol.^7,10,31 The following additional possible risk factors were tested at a significance level of 0.005, to reduce the danger of a type I error (erroneous inclusion) occurring: HbA_1c, Afro-Caribbean ethnicity, Asian-Indian ethnicity, body mass index (BMI), and duration of diagnosed diabetes. Proportional-hazards assumptions were verified with log-cumulative hazard plots³² (data not shown). For AF, there were too few cases for a reliable log-cumulative hazard plot.

View this table: [in this window] [in a new window]   Table 2. Variables Considered for Inclusion in the Model

To improve model stability, SBP, total cholesterol, HDL cholesterol, and HbA_1c were each taken to be the mean of values taken 1 and 2 years after diagnosis of diabetes, as in our previous article.³⁰ Model fitting was carried out by maximum likelihood estimation, using the Newton-Raphson method as implemented in the Numerical Algorithms Group C Library.³³

In a supplementary analysis, we fitted alternative models in which the ratio T:H was replaced with each of the following: total cholesterol, HDL cholesterol, low-density lipoprotein (LDL) cholesterol, triglycerides, and the ratio LDL:HDL cholesterol.

During the early years of the study, all-cause mortality was not significantly different from that in the general population, possibly because people with life-threatening illnesses were excluded from the study.³⁴ To avoid selection bias, years 0 to 4 of data were not used in model fitting; 262 patients with follow-up times <4 years were excluded as described above, and the fitting process for the remaining data used likelihoods conditional on surviving to 4 years without a stroke.

Model adequacy was checked with diagnostic plots, comparing modeled survival probabilities, calculated as the average across the 4549 patients, to survival probabilities for the study population calculated by nonparametric (life-table) methods.³²

External validity of the model was examined by comparison with the rates of fatal stroke reported in 1370 patients of the Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR) older-onset cohort, who were followed for an average of 8.3 years.³⁵ The risk equation of this article was used to calculate 8.3-year risk of stroke in smoking and nonsmoking men and women with type 2 diabetes and characteristics as reported by WESDR.^36–38 Because the blood pressure and cholesterol values taken are the mean of a large number, correction for regression dilution was incorporated. These risk estimates were then combined in a weighted average, reflecting the male-to-female mix and the prevalence of smoking in the WESDR cohort, to give an estimate of the overall risk of stroke, for comparison with the rates of fatal stroke reported by WESDR. Rates of nonfatal stroke were not reported by WESDR.

	Results

Top Abstract Introduction Materials and Methods Results Discussion Appendix 1 References

 
Median follow-up time was 10.5 years. There were 30 700 person-years of follow-up available for model fitting, during which 52 fatal and 136 nonfatal first strokes occurred.

Table 3 shows the parameters included in the final model. Not significant were HbA_1c (P=0.086), Afro-Caribbean (P>0.5) and Asian-Indian (P>0.5) ethnicity, and ex-smoking status (P>0.25). BMI achieved conventional significance (P=0.0085) but did not meet our criteria for inclusion in the model.

View this table: [in this window] [in a new window]   Table 3. Parameters Included in Model Equation

In a patient who has had no stroke in the first T years of diagnosed diabetes, the probability of a stroke in the next t years is R_T(t) = exp[-qx1.145^Tx(1- 1.145^t)/(1-1.145)], where q=0.00186 x 1.092^age-55x 0.700^sex x 1.547^smokingx8.554^AFx 1.122^(SBP-135.5)/10x 1.138^T:H-5.11.

When the input values are of greater or lesser stability than those used to develop the model—for example, blood pressure on a single occasion, rather than averaged over 2 occasions—the effect of regression dilution must be considered, as indicated in the Appendix.

This model uses the ratio T:H for compatibility with our existing model software.³⁰ However, it was found that models of similarly good fit to the data could be built using either total cholesterol (P=0.0044), LDL cholesterol (P=0.0037), or the ratio LDL:HDL cholesterol (P=0.0013) as the only measure of dyslipidemia. Nonnested models cannot be compared by likelihood ratio tests, but graphical comparisons found these models to be very similar to the T:H ratio model (data not shown). Triglycerides were significant (P=0.037) as the sole lipid measure, but not in models already containing LDL cholesterol, LDL:HDL cholesterol ratio, or T:H ratio. HDL cholesterol was not significant in any model (P>0.1).

Example
Consider a power calculation for a hypothetical trial of lipid-lowering therapy in primary prevention of stroke, among white men with type 2 diabetes. Based on data from WESDR, assume a cohort with mean SBP 147 mm Hg, total cholesterol 5.65 mmol/L, HDL cholesterol 1.11 mmol/L, age 55 years at diagnosis of diabetes, and with 12 years of diagnosed diabetes.^35,36 For a nonsmoking man, with characteristics as above and assuming no AF, calculations are as follows: q= 0.00186x1.092^55-55x 1x1x1x1.122^{(147- 135.5)/10}x 1.138^{(5.65/1.11)- 5.11}=0.00212, and hence R₁₂(5)= 1-exp[-0.00212x1.145¹² x(1-1.145⁵)/(1- 1.145)]=0.069, so that such a patient has a 6.9% probability of a stroke within the next 5 years, conditional on not dying from causes other than stroke. For a patient who smokes, the probability is 10.5%. Assuming a 30% smoking prevalence, the average 5-year risk in the population is (0.3x10.5%)+(0.7x6.9%)=8.0%. To have 80% power to detect an effect of a therapy that lowers risk by 15%, a 5-year trial would need 15 000 patients. These event rates are likely to be overestimates for the whole population, because not all patients will remain in the study for 5 years, as a result of deaths from cardiac and other causes as well as loss to follow-up. A more thorough analysis would correct also for regression dilution and account for the variation in risk factors across the population rather than use mean values.

Validation
Figure 1 shows that survival rates predicted by the model lie close to the rates observed in the UKPDS. Observed and modeled survival rates are shown in Figure 2a and 2b, split by SBP and by the ratio T:H, confirming the ability of the model to adjust for these risk factors.

View larger version (12K): [in this window] [in a new window]   Figure 1. Observed survival rates (), with 95% CIs (dashed line) and modeled survival rates (solid line), from year 4.

View larger version (22K): [in this window] [in a new window]   Figure 2. Modeled and observed survival rates from year 4 by SBP (a) and lipid ratio (b). a, indicates observed (SBP 135 mm Hg); dashed line, modeled (SBP 135 mm Hg); , observed (SBP > 135 mm Hg); and dotted line, modeled (SBP >135 mm Hg). b, indicates observed (lipid ratio 5.0); dashed line, modeled (lipid ratio 5.0); , observed (lipid ratio >5.0); and dotted line, modeled (lipid ratio > 5.0).

For the WESDR cohort, in which 59 deaths from stroke were observed, the equation derived here predicts a 13.2% risk of fatal and nonfatal stroke in 8.3 years, corresponding to 197 strokes in 1370 patients. Previous studies have reported stroke fatality rates between 23% and 30% in diabetes.^7,39,40 Applying a 26.5% fatality rate to our prediction of 197 strokes gives 52 fatal strokes. For comparison, the Framingham equations for mortality predict 33 fatal strokes in the same cohort.⁴¹

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion Appendix 1 References

 
To our knowledge, this is the first time a model has been developed to predict the risk of stroke specifically in a population with type 2 diabetes. Because our equations are based on a cohort identified at diagnosis of diabetes and without previous stroke, the model predicts risk of stroke as a primary event. The inputs required by the model are all readily measured in clinical practice, and routinely reported in research, for maximum applicability. The model has been incorporated in the UKPDS Risk Engine software, which is available to noncommercial users without charge from our website.⁴²

Several limitations arise from the use of clinical trial rather than epidemiologic data, although we have attempted to correct for the impact of trial selection criteria, as described above. For time periods <4 years or >20 years after diagnosis of diabetes, or for ages >65 at diagnosis of diabetes, predictions from this model are extrapolations. Our definition of nonfatal stroke is more stringent than is sometimes used, requiring signs or symptoms to persist for more than a month. Our model does not distinguish between fatal and nonfatal strokes or between ischemic and hemorrhagic strokes. It would be ideal to test for any effect of antihypertensive and other therapies, but this is not practical because many patients took many different therapies over the years of UKPDS follow-up.^4,43 Certain limitations are generic to all modeling exercises. Users should be aware that the risk ratio between individuals at 2 levels of a risk factor is not necessarily equal to the risk reduction obtainable for a single individual by changing levels of the risk factor. For tight blood pressure control, the risk reduction obtained in the UKPDS was greater than the risk ratio in an observational model would suggest.³ The current version of the model does not provide for estimation when data are missing. HbA_1c was not found to be significant in this analysis, as it was in a previous UKPDS analysis that used more person-years of data and a more informative variable, updated mean HbA_1c.⁶ BMI achieved conventional significance (P<0.05) but did not meet our inclusion criteria, which were designed to protect the generalizability of the risk equation. A risk equation including BMI might not generalize well, because previous studies have found that BMI is not a risk factor for stroke in type 2 diabetes or have even reported a possible protective effect of obesity.^{44– 47}

The comparison with stroke death rates in WESDR suggests that the model may generalize well to other populations. Although our derived figure, 52 fatal strokes, is lower than the 59 fatal strokes, this is to be expected because we model first stroke only. There are few other published models for the risk of stroke, and none specific to diabetes. The best-known previously published risk equations for stroke are derived from the Framingham cohort,^20,41 and have been used in guidelines for clinical practice⁴⁸ and in health economic modeling in the general population⁴⁹ and in diabetes.^21–23 The 5734 individuals in Framingham were older than in the UKPDS, (mean age 66 years) and experienced a higher event rate, 472 strokes in 10 years, but fewer than 10% had diabetes. The Framingham profile adjusts for left ventricular hypertrophy. Exclusion of the latter from the present paper gives our equations greater applicability, at the cost of some discriminatory power. The British Regional Heart Study model for risk of stroke does not adjust for presence of diabetes,⁵⁰ and a recent cardiovascular risk score provides only relative risk in the case of stroke.⁵¹

Although guidelines for primary prevention often use risk of coronary heart disease as a surrogate for risk of cardiovascular disease,^48,52 accurate estimation of stroke as well as coronary risk is preferable, for example, in hypertension management¹⁷ and for patients of African origin.¹⁵ The model here is suitable for use in a clinical setting.

Risk estimators also have a role to play in research and in policy formulation. The example in the Results section has demonstrated the use of the model in a power calculation, indicating that in primary prevention the large numbers needed to establish a lipid-lowering effect on stroke may require large-scale meta-analyses.^53,54 The literature applying the Framingham equation to cost-effectiveness analyses in diabetes demonstrates a role for a diabetes-specific risk estimator.^22,23,55 We hope that the present article will be of use to researchers, planners, and clinicians in the management of cerebrovascular risk in diabetes.

	Acknowledgments

 
The major grants for this study were from the United Kingdom Medical Research Council; British Diabetic Association; UK Department of Health; National Eye Institute and National Institute of Digestive, Diabetes, and Kidney Disease of the National Institutes of Health; British Heart Foundation; Novo Nordisk; Bayer; Bristol-Myers Squibb; Hoechst; Lilly; Lipha; and Farmalita Carlo Erba. R.J.S. was supported by Wellcome Trust Fellowship 054470/Z/98/Z/DG/NOS/sh.

The cooperation of the patients and staff of the UKPDS is much appreciated. The participating centers in the UKPDS were Radcliffe Infirmary, Oxford; Royal Infirmary, Aberdeen; University Hospital, Birmingham; St. George’s Hospital and Hammersmith Hospital, London; City Hospital, Belfast; North Staffordshire Royal Infirmary, Stoke-on-Trent; Royal Victoria Hospital, Belfast; St. Helier Hospital, Carshalton; Whittington Hospital, London; Norfolk & Norwich Hospital; Lister Hospital, Stevenage; Ipswich Hospital; Ninewells Hospital, Dundee; Northampton Hospital; Torbay Hospital; Peterborough General Hospital; Scarborough Hospital; Derbyshire Royal Infirmary; Manchester Royal Infirmary; Hope Hospital, Salford; Leicester General Hospital; Royal Devon & Exeter Hospital. We are grateful to Prof Peter Bennet and to Dr Andrew Farmer for helpful comments.

	Appendix 1

Top Abstract Introduction Materials and Methods Results Discussion Appendix 1 References

 
Appendix: Regression Dilution
The term regression dilution describes the behavior of parameter estimates in regression and in survival analysis when a predictor variable is subject to biovariability or assay variability.⁵⁶ The risk equation reported in this article is recommended for use with estimates of blood pressure, and of total and HDL cholesterol, of similar precision to those with which the model was built, that is, the mean of 6 measurements of blood pressure, the mean of 2 measurements of total cholesterol, and the mean of 2 measurements of HDL cholesterol.

Adjustment for regression dilution is appropriate if the available data are not subject to biovariability, for example, if the blood pressure and cholesterol levels are the means over very large populations. Based on a maximum likelihood method,⁵⁷ the following corrections are appropriate. For blood pressure, replace the parameter estimate 1.122 with 1.122^1.25=1.155. For lipid ratio, replace the parameter estimate 1.138 with 1.138^1.28=1.181. Because variables are centered around their means in the model equation (for example, blood pressure is used as SBP-135.5), no adjustment is necessary to the intercept parameter.

We use a method reported previously³⁰ to estimate the adjustment necessary if the input data are of greater biovariability than ours. If the available blood pressure is the mean of just 3 measurements, replace 1.122 with 1.101. If blood pressure is a single measurement, replace 1.122 with 1.060. If the lipid ratio is based on single measurements of HDL and total cholesterol, replace the parameter 1.138 with 1.111.

Received November 19, 2001; revision received February 18, 2002; accepted April 3, 2002.

	References

Top Abstract Introduction Materials and Methods Results Discussion Appendix 1 References

 
1. Manson JAE, Colditz GA, Stampfer MJ, Willett WC, Krolewski AS, Rosner B, Arky RA, Speizer FE, Hennekens CHA. A prospective study of maturity-onset diabetes mellitus and risk of coronary heart disease and stroke in women. Arch Intern Med. 1991; 151: 1141–1147.[Abstract/Free Full Text]

2. Stamler J, Vaccaro O, Neaton JD, Wentworth D. Diabetes, other risk factors, and 12 year cardiovascular mortality for men screened in the Multiple Risk Factor Intervention Trial. Diabetes Care. 1993; 16: 434–444.[Abstract]

3. UKPDS Group. Association of systolic blood pressure with macrovascular and microvascular complications of type 2 diabetes (UKPDS 36). BMJ. 2000; 321: 412–419.[Abstract/Free Full Text]

4. UKPDS Group. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes (UKPDS 38). BMJ. 1998; 317: 703–713.[Abstract/Free Full Text]

5. Kuller L. Stroke and diabetes.In: Harris MI, Cowie CC, Stern MP, Boyko EJ, Reiber GE, Bennett PH, eds. Diabetes in America. Washington, DC: US Government Printing Office; 1995.

6. UKPDS Group. Risk factors for stroke in type 2 diabetes mellitus: United Kingdom Prospective Diabetes Study (UKPDS) 29. Arch Intern Med. 1999; 159: 1033–1034.[Free Full Text]

7. Stegmayr B, Asplund K. Diabetes as a risk factor for stroke: a population perspective. Diabetologia. 1995; 38: 1061–1068.[Medline] [Order article via Infotrieve]

8. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation: a major contributor to stroke in the elderly: the Framingham Study. Arch Intern Med. 1987; 147: 1561–1564.[Abstract/Free Full Text]

9. UKPDS Group. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35). BMJ. 2000; 321: 405–412.[Abstract/Free Full Text]

10. Goldstein LB, Adams R, Becker K, Freberg CD, Gorelick PB, Hademenos G, Hill M, Howard G, Howard VJ, Jacobs B, Levine SR, Mosca L, Sacco RL, Sherman DG, Wolf PA, del Zoppo GJ. AHA Scientific Statement: Primary Prevention of Stroke: a Statement for Healthcare Professionals from the Stroke Council of the American Heart Association. Circulation. 2001; 103: 163–182.[Free Full Text]

11. Wannamethee SG, Shaper AG, Ebrahim S. HDL-cholesterol, total cholesterol, and the risk of stroke in middle-aged British men. Stroke. 2000; 31: 1882–1888.[Abstract/Free Full Text]

12. Anderson KM, Odell PM, Wilson PWF, Kannel WB. Cardiovascular disease risk profiles. Am Heart J. 1990; 121: 293–298.[CrossRef]

13. Salonen JT, Puska P. Relation of serum cholesterol and triglycerides to the risk of acute myocardial infarction, cerebral stroke and death in eastern Finnish male population. Int J Epidemiol. 1983; 12: 26–31.[Abstract/Free Full Text]

14. Lehto S, Ronnemaa T, Pyörälä K, Laakso M. Predictors of stroke in middle-aged patients with non-insulin-dependent diabetes. Stroke. 1996; 27: 63–68.[Abstract/Free Full Text]

15. Robson J, Feder G. Predicting and reducing cardiovascular risk. Heart. 2001; 85: 487–488.[Free Full Text]

16. Isles CG, Ritchie LD, Murchie P, Norrie J. Risk assessment in primary prevention of coronary heart disease: randomised comparison of three scoring methods. BMJ. 2000; 320: 690–691.[Free Full Text]

17. Ramsay LE, Williams B, Johnston GD, MacGregor GA, Poston L, Potter JF, Poulter NR, Russell G. British Hypertension Society guidelines for hypertension management 1999: summary. BMJ. 1999; 319: 630–635.[Free Full Text]

18. British Cardiac Society, British Hyperlipidaemia Association, British Hypertension Society, British Diabetic Association. Joint British recommendations on prevention of coronary heart disease in clinical practice: summary. BMJ. 2000; 320: 705–708.[Free Full Text]

19. Wood D, Backer GD, Faergeman O, Graham I, Mancia G, Pyorala K. Prevention of coronary heart disease in clinical practice: recommendations of the Second Joint Task Force of European and Other Societies on Coronary Prevention. Atherosclerosis. 1998; 140: 199–270.[CrossRef][Medline] [Order article via Infotrieve]

20. Wolf PA, D’Agostino RB, Belanger AJ, Kannel WB. Probability of stroke: a risk profile from the Framingham study. Stroke. 1991; 22: 312–318.[Abstract/Free Full Text]

21. Eastman RC, Javitt JC, Herman WH, Dasbach EJ, Zbrozek AS, Dong F, Manninen D, Garfield SA, Copley-Merriman C, Maier W, Eastman JF, Kotsanos J, Cowie CC, Harris M. Model of complications of NIDDM, I: model construction and assumptions. Diabetes Care. 1997; 20: 725–734.[Abstract]

22. Brown JB, Russell A, Chan W, Pedula K, Aickin M. The global diabetes model: user friendly version 3.0. Diabetes Res Clin Pract. 2000; 50 (suppl 3): S15–S46.[Medline] [Order article via Infotrieve]

23. Palmer AJ, Brandt A, Gozzoli V, Weiss C, Stock H, Wenzel H. Outline of a diabetes management model: principles and applications. Diabetes Res Clin Pract. 2000; 50 (suppl 3): S47–S56.[Medline] [Order article via Infotrieve]

24. Anderson KM, Wilson PW, Odell PM, Kannel WB. An updated coronary risk profile: a statement for health professionals. Circulation. 1991; 83: 356–362.[Free Full Text]

25. UKPDS Group. UK Prospective Diabetes Study VIII: study design, progress and performance. Diabetologia. 1991; 34: 877–890.[CrossRef][Medline] [Order article via Infotrieve]

26. UKPDS Group. Intensive blood glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet. 1998; 352: 837–853.[CrossRef][Medline] [Order article via Infotrieve]

27. UKPDS Group. Biochemical risk factors in type 2 diabetic patients at diagnosis compared with age-matched normal subjects: UK Prospective Diabetes Study XI. Diabet Med. 1994; 11: 534–544.[Medline] [Order article via Infotrieve]

28. Laboratory Standardization Panel. Current status of blood cholesterol measurement in clinical chemistry laboratories in the United States. Clin Chem. 1988; 34: 193–201.[Free Full Text]

29. Packard CJ, Bell MA, Eaton RH, Dagen MM, Cassidy M, Shepherd JA. A pilot scheme for improving the accuracy of serum cholesterol measurement in Scotland and Northern Ireland. Ann Clin Biochem. 1993; 30(pt 4): 387–393.[Medline] [Order article via Infotrieve]

30. UKPDS Group. The UKPDS Risk Engine: a model for the risk of coronary heart disease in type 2 diabetes (UKPDS 56). Clin Sci. 2001; 101: 671–679.[Medline] [Order article via Infotrieve]

31. Bell DSH. Stroke in the diabetic patient. Diabetologia. 1994; 17: 213–219.[CrossRef]

32. Collett D. Modelling Survival Data in Medical Research. London, UK: Chapman and Hall; 1994.

33. Numerical Algorithms Group C. Numerical Algorithms Group C Library.Version 5. Oxford, UK: Numerical Algorithms Group; 1998.

34. UKPDS Group. Risk factors for coronary artery disease in non-insulin dependent diabetes (UKPDS 23). BMJ. 1998; 316: 823–828.[Abstract/Free Full Text]

35. Klein R. Hyperglycemia and microvascular and macrovascular disease in diabetes. Diabetes Care. 1995; 18: 258–268.[Abstract]

36. Klein R, Klein BE, Moss SE, Davis MD, DeMets DL. The Wisconsin Epidemiologic Study of Diabetic Retinopathy, X: four-year incidence and progression of diabetic retinopathy when age at diagnosis is 30 years or more. Arch Ophthalmol. 1989; 107: 244–249.[Abstract/Free Full Text]

37. Klein BEK, Moss SE, Klein R, Surawicz T. Serum cholesterol in the Wisconsin Epidemiologic Study of Diabetic Retinopathy. Diabetes Care. 1992; 15: 282–287.[Abstract]

38. Moss SE, Klein R, K. KBE. Association of cigarette smoking with diabetic retinopathy. Diabetes Care. 1991; 14: 119–126.[Abstract]

39. Stevens R, Adler A, Gray A, Briggs A, Holman R. Life-expectancy projection by modelling and computer simulation (UKPDS 46). Diabetes Res Clin Pract. 2000; 50 (suppl 3): S5–S13.[Medline] [Order article via Infotrieve]

40. Sprafka JM, Virnig BA, Shahar E, McGovern PG. Trends in diabetes prevalence among stroke patients and the effect of diabetes on stroke survival: the Minnesota Survey. Diabet Med. 1994; 11: 678–684.[Medline] [Order article via Infotrieve]

41. Anderson KM, Odell PM, Wilson PW, Kannel WB. Cardiovascular disease risk profiles. Am Heart J. 1991; 121: 293–298.[CrossRef][Medline] [Order article via Infotrieve]

42. Diabetes Trials Unit, Oxford Centre for Diabetes, Endocrinology and Metabolism, University of Oxford. UKPDS Risk Engine.Available at: http://www.dtu.ox.ac.uk/riskengine/. Accessed May 10, 2002.

43. UKPDS Group. Glycaemic control with diet, sulphonylurea, metformin and insulin therapy in patients with type 2 diabetes: progressive requirement for multiple therapies (UKPDS 49). JAMA. 1999; 281: 2005–2012.[Abstract/Free Full Text]

44. Folsom AR, Rasmussen ML, Chambless LE, Howard G, Cooper LS, Schmidt MI, Heiss G. Prospective associations of fasting insulin fat distribution, and diabetes with risk of ischemic stroke: the Atherosclerosis Risk in Communities (ARIC) Study Investigators. Diabetes Care. 1999; 22: 1077–1083.[Abstract/Free Full Text]

45. Tuomilehto J, Rastenyte D, Jousilahti P, Sarti C, Vartiainen E. Diabetes mellitus as a risk factor for death from stroke: prospective study of the middle-aged Finnish population. Stroke. 1996; 27: 210–215.[Abstract/Free Full Text]

46. Guerrero-Romero F, Rodriguez-Moran M. Proteinuria is an independent risk factor for ischemic stroke in non-insulin-dependent diabetes mellitus. Stroke. 1999; 30: 1787–1791.[Abstract/Free Full Text]

47. Klein R, Klein BE, Moss SE. Is obesity related to microvascular and macrovascular complications in diabetes? The Wisconsin Epidemiologic Study of Diabetes Retinopathy. Arch Intern Med. 1997; 157: 650–656.[Abstract/Free Full Text]

48. British Cardiac Society, British Hyperlipidaemia Association, British Hypertension Society, British Diabetic Association. Joint British recommendations on prevention of coronary heart disease in clinical practice. Heart. 1998; 80 (suppl 2): S-1–S-29.[Free Full Text]

49. Matchar DB, Samsa GP, Matthews JR, Ancukiewicz M, Parmigiani G, Hasselblad V, Wolf PA, D’Agostino RB, Lipscomb J. The Stroke Prevention Policy Model: linking evidence and clinical decisions. Ann Intern Med. 1997; 127: 704–711.[Abstract/Free Full Text]

50. Coppola WGT, Whincup PH, Papacosta O, Walker M, Ebrahim S. Scoring system to identify men in high risk of stroke: a strategy for general practice. Br J Gen Pract. 1995; 45: 185–189.[Medline] [Order article via Infotrieve]

51. Pocock SJ, McCormack V, Gueyffier F, Boutitie F, Fagard RH, Boissel J-P. A score for predicting risk of death from cardiovascular disease in adults with raised blood pressure, based on individual patient data from randomised controlled trials. BMJ. 2001; 323: 75–81.[Abstract/Free Full Text]

52. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA. 2001; 285: 2486–2497.[Free Full Text]

53. Austin MA. Plasma triglyceride as a risk factor for coronary heart disease. Am J Epidemiol. 1989; 129: 249–259.[Free Full Text]

54. Cholesterol Treatment Trialist (CTT) Collaboration. Protocol for a prospective collaborative overview of all current and planned randomized trials of cholesterol treatment regimens. Am J Cardiol. 1995; 75: 1130–1134.[CrossRef][Medline] [Order article via Infotrieve]

55. Eastman RC, Javitt JC, Herman WH, Dasbach EJ, Copley-Merriman C, Maier W, Dong F, Manninen D, Zbrozek AS, Kotsanos J, Garfield SA, Harris M. Model of complications of NIDDM, II: analysis of the health benefits and cost-effectiveness of treating NIDDM with the goal of normoglycemia. Diabetes Care. 1997; 20: 735–744.[Abstract]

56. MacMahon S, Peto R, Cutler J, Collins R, Sorlie P, Neaton J, Abbott R, Godwin J, Dyer A, Stamler J. Blood pressure, stroke, and coronary heart disease, 1: prolonged differences in blood pressure: prospective observational studies corrected for the regression dilution bias. Lancet. 1990; 335: 765–774.[CrossRef][Medline] [Order article via Infotrieve]

57. Frost C, Thompson S. Correcting for regression dilution bias: comparison of methods for a single predictor variable. J R Stat Soc Ser A. 2000; 163: 173–190.[CrossRef]

This article has been cited by other articles:

				R. H. Fagard Smoking Amplifies Cardiovascular Risk in Patients With Hypertension and Diabetes Diabetes Care, November 1, 2009; 32(suppl_2): S429 - S431. [Full Text] [PDF]

				V. A. Fonseca, H. Smith, N. Kuhadiya, S. M. Leger, C. L. Yau, K. Reynolds, L. Shi, R. H. McDuffie, T. Thethi, and J. John-Kalarickal Impact of a Natural Disaster on Diabetes: Exacerbation of disparities and long-term consequences Diabetes Care, September 1, 2009; 32(9): 1632 - 1638. [Abstract] [Full Text] [PDF]

				B. T. Denton, M. Kurt, N. D. Shah, S. C. Bryant, and S. A. Smith Optimizing the Start Time of Statin Therapy for Patients with Diabetes Med Decis Making, May 1, 2009; 29(3): 351 - 367. [Abstract] [PDF]

				J. L. Rosenzweig, E. Ferrannini, S. M. Grundy, S. M. Haffner, R. J. Heine, E. S. Horton, and R. Kawamori Primary Prevention of Cardiovascular Disease and Type 2 Diabetes in Patients at Metabolic Risk: An Endocrine Society Clinical Practice Guideline J. Clin. Endocrinol. Metab., October 1, 2008; 93(10): 3671 - 3689. [Abstract] [Full Text] [PDF]

				J. Cederholm, K. Eeg-Olofsson, B. Eliasson, B. Zethelius, P. M. Nilsson, S. Gudbjornsdottir, and on behalf of the Swedish National Diabetes Registe Risk Prediction of Cardiovascular Disease in Type 2 Diabetes: A risk equation from the Swedish National Diabetes Register Diabetes Care, October 1, 2008; 31(10): 2038 - 2043. [Abstract] [Full Text] [PDF]

				P. Gaede, W. J. Valentine, A. J. Palmer, D. M.D. Tucker, M. Lammert, H.-H. Parving, and O. Pedersen Cost-Effectiveness of Intensified Versus Conventional Multifactorial Intervention in Type 2 Diabetes: Results and projections from the Steno-2 study Diabetes Care, August 1, 2008; 31(8): 1510 - 1515. [Abstract] [Full Text] [PDF]

				E. S. Huang, Q. Zhang, N. Gandra, M. H. Chin, and D. O. Meltzer The Effect of Comorbid Illness and Functional Status on the Expected Benefits of Intensive Glucose Control in Older Patients with Type 2 Diabetes: A Decision Analysis Ann Intern Med, July 1, 2008; 149(1): 11 - 19. [Abstract] [Full Text] [PDF]

				A. J. Hayes, P. M. Clarke, P. G. Glasziou, R. J. Simes, P. L. Drury, and A. C. Keech Can Self-Rated Health Scores Be Used for Risk Prediction in Patients With Type 2 Diabetes? Diabetes Care, April 1, 2008; 31(4): 795 - 797. [Abstract] [Full Text] [PDF]

				A. H Barnett The importance of treating cardiometabolic risk factors in patients with type 2 diabetes Diabetes and Vascular Disease Research, March 1, 2008; 5(1): 9 - 14. [Abstract] [PDF]

				E. L. Air and B. M. Kissela Diabetes, the Metabolic Syndrome, and Ischemic Stroke: Epidemiology and possible mechanisms Diabetes Care, December 1, 2007; 30(12): 3131 - 3140. [Full Text] [PDF]

				C. A. Cull, C. C. Jensen, R. Retnakaran, and R. R. Holman Impact of the Metabolic Syndrome on Macrovascular and Microvascular Outcomes in Type 2 Diabetes Mellitus: United Kingdom Prospective Diabetes Study 78 Circulation, November 6, 2007; 116(19): 2119 - 2126. [Abstract] [Full Text] [PDF]

				T. J. Hoerger, K. A. Hicks, S. W. Sorensen, W. H. Herman, R. E. Ratner, R. T. Ackermann, P. Zhang, and M. M. Engelgau Cost-Effectiveness of Screening for Pre-Diabetes Among Overweight and Obese U.S. Adults Diabetes Care, November 1, 2007; 30(11): 2874 - 2879. [Abstract] [Full Text] [PDF]

				The Mount Hood 4 Modeling Group Computer Modeling of Diabetes and Its Complications: A report on the Fourth Mount Hood Challenge Meeting Diabetes Care, June 1, 2007; 30(6): 1638 - 1646. [Abstract] [Full Text] [PDF]

				C. B. Giorda, A. Avogaro, M. Maggini, F. Lombardo, E. Mannucci, S. Turco, S. S. Alegiani, R. Raschetti, M. Velussi, E. Ferrannini, et al. Incidence and Risk Factors for Stroke in Type 2 Diabetic Patients: The DAI Study Stroke, April 1, 2007; 38(4): 1154 - 1160. [Abstract] [Full Text] [PDF]

				T. Wada, M. Naito, H. Kenmochi, H. Tsuneki, and T. Sasaoka Chronic Nicotine Exposure Enhances Insulin-Induced Mitogenic Signaling via Up-Regulation of {alpha}7 Nicotinic Receptors in Isolated Rat Aortic Smooth Muscle Cells Endocrinology, February 1, 2007; 148(2): 790 - 799. [Abstract] [Full Text] [PDF]

				X. Yang, W.-Y. So, A. P.S. Kong, C.-S. Ho, C. W.K. Lam, R. J. Stevens, R. R. Lyu, D. D. Yin, C. S. Cockram, P. C.Y. Tong, et al. Development and Validation of Stroke Risk Equation for Hong Kong Chinese Patients With Type 2 Diabetes: The Hong Kong Diabetes Registry Diabetes Care, January 1, 2007; 30(1): 65 - 70. [Abstract] [Full Text] [PDF]

				A. Balamurugan, R. Ohsfeldt, T. Hughes, and M. Phillips Diabetes Self-management Education Program for Medicaid Recipients: A Continuous Quality Improvement Process The Diabetes Educator, November 1, 2006; 32(6): 893 - 900. [Abstract] [Full Text] [PDF]

				N. Asfandiyarova, N. Kolcheva, I. Ryazantsev, and V. Ryazantsev Risk factors for stroke in type 2 diabetes mellitus Diabetes and Vascular Disease Research, May 1, 2006; 3(1): 57 - 60. [Abstract] [PDF]

				H. Zhou, D. J.M. Isaman, S. Messinger, M. B. Brown, R. Klein, M. Brandle, and W. H. Herman A Computer Simulation Model of Diabetes Progression, Quality of Life, and Cost Diabetes Care, December 1, 2005; 28(12): 2856 - 2863. [Abstract] [Full Text] [PDF]

				E. Selvin, J. Coresh, S. H. Golden, F. L. Brancati, A. R. Folsom, and M. W. Steffes Glycemic Control and Coronary Heart Disease Risk in Persons With and Without Diabetes: The Atherosclerosis Risk in Communities Study Arch Intern Med, September 12, 2005; 165(16): 1910 - 1916. [Abstract] [Full Text] [PDF]

				R. M. Clifford, W. A. Davis, K. T. Batty, and T. M.E. Davis Effect of a Pharmaceutical Care Program on Vascular Risk Factors in Type 2 Diabetes: The Fremantle Diabetes Study Diabetes Care, April 1, 2005; 28(4): 771 - 776. [Abstract] [Full Text] [PDF]

				W. H. Herman, T. J. Hoerger, M. Brandle, K. Hicks, S. Sorensen, P. Zhang, R. F. Hamman, R. T. Ackermann, M. M. Engelgau, R. E. Ratner, et al. The Cost-Effectiveness of Lifestyle Modification or Metformin in Preventing Type 2 Diabetes in Adults with Impaired Glucose Tolerance Ann Intern Med, March 1, 2005; 142(5): 323 - 332. [Abstract] [Full Text] [PDF]

				L. Rodriguez-Manas, M. Castro, and J. Manzarbeitia Review: Cardiovascular risk and diabetes in old people The British Journal of Diabetes & Vascular Disease, January 1, 2005; 5(1): 7 - 11. [Abstract] [PDF]

				E. Selvin, S. Marinopoulos, G. Berkenblit, T. Rami, F. L. Brancati, N. R. Powe, and S. H. Golden Meta-Analysis: Glycosylated Hemoglobin and Cardiovascular Disease in Diabetes Mellitus Ann Intern Med, September 21, 2004; 141(6): 421 - 431. [Abstract] [Full Text] [PDF]

				American Diabetes Association Consensus Panel Guidelines for Computer Modeling of Diabetes and Its Complications Diabetes Care, September 1, 2004; 27(9): 2262 - 2265. [Full Text] [PDF]

				R. J. Stevens, R. L. Coleman, A. I. Adler, I. M. Stratton, D. R. Matthews, and R. R. Holman Risk Factors for Myocardial Infarction Case Fatality and Stroke Case Fatality in Type 2 Diabetes: UKPDS 66 Diabetes Care, January 1, 2004; 27(1): 201 - 207. [Abstract] [Full Text] [PDF]

				P. Khairy, C. P. O'Donnell, and M. J. Landzberg Transcatheter Closure versus Medical Therapy of Patent Foramen Ovale and Presumed Paradoxical Thromboemboli: A Systematic Review Ann Intern Med, November 4, 2003; 139(9): 753 - 760. [Abstract] [Full Text] [PDF]

				W. H. Herman Diabetes Modeling Diabetes Care, November 1, 2003; 26(11): 3182 - 3183. [Full Text] [PDF]

				K. Eguchi, K. Kario, and K. Shimada Greater Impact of Coexistence of Hypertension and Diabetes on Silent Cerebral Infarcts Stroke, October 1, 2003; 34(10): 2471 - 2474. [Abstract] [Full Text] [PDF]

				R.M. Conroy, K. Pyorala, A.P. Fitzgerald, S. Sans, A. Menotti, G. De Backer, D. De Bacquer, P. Ducimetiere, P. Jousilahti, U. Keil, et al. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project Eur. Heart J., June 1, 2003; 24(11): 987 - 1003. [Abstract] [Full Text] [PDF]

				OTHER ARTICLES NOTED (Nov 01 to 18 Oct 02) Evid. Based Nurs., January 1, 2003; 6(1): e1 - 1. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kothari, V. Articles by Holman, R. R. Search for Related Content PubMed PubMed Citation Articles by Kothari, V. Articles by Holman, R. R. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Hazardous Substances DB CHOLESTEROL Medline Plus Health Information Diabetes Stroke Related Collections Lipids Primary and Secondary Stroke Prevention Risk Factors for Stroke