Risk Factors for Coronary Heart Disease in Type 1 Diabetic Patients in Europe: The EURODIAB Prospective Complications Study

Full details of the design, methods, and recruitment in the EURODIAB Prospective Complications Study (PCS) have been published elsewhere (10–12). The baseline cross-sectional clinic-based study examined 3,250 type 1 diabetic patients between 1989–1991. Participants were aged between 15 and 60 years and recruited from 31 centers in 16 European countries. The sampling frame was all type 1 diabetic patients attending at least once in the last year for each center. Patients were stratified by age (three categories), diabetes duration (three categories), and sex. Ten patients were then randomly selected from each stratum (10). Type 1 diabetes was defined as diabetes diagnosed before the age of 36 years with a continuous need for insulin within 1 year of diagnosis. Of those invited, 85% participated. Those with a duration of diabetes <1 year and pregnant women were excluded. Ethics committee approval was obtained at each center, and all subjects provided written informed consent.

Seven years after baseline examinations, study participants were invited for re-examination. Of the 3,250 subjects at baseline, 681 individuals could not be assessed or were excluded from follow-up because of the following reasons. Eight patients did not meet the inclusion criteria at baseline, 222 patients had CHD at baseline, and in 14 patients, CHD was not measured. In addition, four centers (n = 437) did not participate in the follow-up examination. Of the remaining 2,569 subjects, 2,329 could be assessed for CHD, of which 151 had CHD at follow-up and 2,178 did not.

All risk factors and microvascular complications were measured at baseline according to a standardized protocol (11). BP was recorded in a sitting position with a random zero sphygmomanometer (Hawskley, Lancing, U.K.) as the mean of two measurements. Hypertension was defined as a systolic BP ≥140 mmHg or diastolic BP ≥90 mmHg and/or the current use of BP-lowering drugs.

Retinopathy was assessed by retinal photographs taken according to the EURODIAB protocol (13). Grading was performed by the Retinopathy Grading Centre at the Hammersmith Hospital, Imperial College London. Retinopathy was classified as none (level 0), nonproliferative (levels 1–3), and proliferative retinopathy (levels 4 and 5).

Peripheral neuropathy was assessed on the basis of neuropathic symptoms and signs, including measurement of vibration perception threshold. Autonomic neuropathy was defined as an R-R ratio of <1.04 and/or a fall in BP from resting to standing of >20 mmHg (14).

A single 24-h urine collection was performed to calculate albumin excretion rate (AER) after excluding proteinuria due to urinary tract infection using a Nephur dipstick test for bacteria. Urinary albumin was measured in a single laboratory by an immunoturbidimetric method (Sanofi Diagnostics Pasteur, Minneapolis, MN) (15). AER was categorized as normoalbuminuria at ≤20 μg/min, microalbuminuria between 20 and 200 μg/min, and macroalbuminuria at ≥200 μg/min. Albuminuria was defined as micro- and macroalbuminuria. A blood sample was taken for the measurement of plasma lipids (fasting triglycerides [FTG], cholesterol, HDL cholesterol) and GHb. Triglyceride (16) and cholesterol (17) concentration of plasma and the cholesterol concentration of HDL (18) were assayed by standard enzymatic methods (Boehringer Mannheim, East Sussex, U.K.) using a cobas-bio centrifugal analyzer (Roche, Welwyn Garden City, Herts, U.K.). For HDL, samples with triglyceride concentrations >268 mg/dl were diluted with 0.15 mol/l sodium chloride solution before chemical precipitation. All analyses were performed centrally. LDL cholesterol was calculated from Friedewald’s formula if triglycerides were below 400 mg/dl (19). Fifteen patients did not have their LDL cholesterol calculated. Non-HDL cholesterol was calculated as HDL cholesterol subtracted from total cholesterol.

The value of GHb was measured centrally by an enzyme immunoassay (Dako, Ely, U.K.) using a monoclonal antibody raised against GHb with a reference range of 2.9–4.8% (20). The intra-assay and interassay coefficients of variation were 2.3–2.4% and 2.6–5.0%, respectively.

CHD (follow-up) was defined as a previous physician diagnosis of CHD or presence of electrocardiogram (ECG) abnormalities. The former included: myocardial infarction (MI), angina pectoris, or coronary artery bypass graft surgery. A conventional 12-lead resting ECG was recorded on each subject. ECG abnormalities were classified by two observers according to the Minnesota Code (21). Any discrepancies between the two observers were adjudicated by a third. ECG abnormalities suggestive of probable ischemia consist of codes 1.1 and 1.2 (major Q/QS waves) and code 7.1 (complete left bundle branch block). Possible ischemia consists of code 1.3 for minor Q waves, codes 4.1, 4.2, and 4.3 for ST segment abnormalities, and codes 5.1, 5.2, and 5.3 for T wave abnormalities. Death due to CHD causes was also included and coded according to ICD-9 (410–414) classification.

All CHD events were captured by questionnaire with additional supporting information from hospital records, death certificates, and other health care documents. In addition, a comparison of allocation of cause of death was performed separately by two observers (J.H.F. and N.C.) with 100% agreement.

The statistical packages SAS (version 8.0; SAS, Cary, NC) and STATA 7 were used to perform all analyses. Descriptive statistics were performed to estimate differences between subjects; Student’s t test and χ² test were used as appropriate. Spearman’s rank correlations were used to test the cross-sectional relationship between risk factors. Non-normally distributed variables were log transformed.

Incidence rates (per 1,000 person-years) for CHD were calculated by sex and age-group.

Sex differences in risk factors were tested by fitting interaction terms between specific risk factors with sex using data on the whole study population. Interactions were only tested for risk factors that were thought to be different between men and women, such as waist-to-hip ratio (WHR) (P = 0.03 in a model with WHR, sex, and an interaction term of WHR and sex), smoking (P = 0.14), and AER (P = 0.19). Furthermore, sex-specific variables were analyzed simultaneously to estimate which risk factors were more important in men and in women. Sex-specific variables were defined by nesting each variable for men and women separately (for example WHR × [sex = 1] for men and WHR × [sex = 0] for women). As interactions and sex-specific variables were observed suggestive of a difference between men and women, findings are presented stratified by sex.

Univariate and multivariate Cox proportional hazards modeling were used to estimate hazard ratios for CHD associated with risk factors. A simultaneous and stepwise approach was used to determine the most important risk factors. Likelihood ratio tests were used to estimate the importance of each risk factor. A full model with a set of risk factors was compared with an incomplete model omitting the risk factor of interest from the previous model. The P value for the effect of removing a given risk factor from the model was obtained by comparing the log likelihoods from the two models. Standardized hazard ratios were estimated from these models by exponentiating the β-coefficient multiplied by the standard deviation (Exp[β × SD]).

Of the 151 CHD events, 34% (n = 51) were classified on the basis of physician-diagnosed angina pectoris, coronary artery bypass grafting, and MI; 11% (n = 17) were due to fatal CHD; and the rest of the events were assessed on the basis of Minnesota coded ischemic ECGs (possible MI [3%] and probable MI [52%]). CHD events were equally distributed between men and women. CHD incidence rates were higher in women than in men, 10.2 (95% CI 8.2–12.7) compared with 8.0 (6.3–10.1) per 1,000 person-years, but this was not statistically significant (P = 0.14). Men and women >40 years of age at baseline had three times higher CHD incidence rates than those <40 years (24 vs. 8 per 1,000 person-years in women and 20 vs. 5 per 1,000 person-years in men) (P = 0.0001).

Baseline risk factors were compared between those included (n = 2,329) and those lost to follow-up (n = 921) for men and women separately (data not shown). A more atherogenic risk factor profile was found in those who dropped out, even when adjusted for age, such as older age and duration of diabetes, worse glycemic control, abnormal lipid levels, higher BP, and more microvascular complications.

Apart from WHR and smoking, there were no major differences in risk factors for CHD between men and women at baseline (Table 1).

After adjustment for diabetes duration, age was a significant risk factor for CHD, whereas diabetes duration, adjusted for age, was not significantly related to CHD (Table 2). Age and diabetes duration were highly correlated (Spearman’s correlation coefficient = 0.51).

In men and women, baseline age and increased AER levels were significant risk factors for CHD. In addition in men, current smoking, WHR, and raised levels of GHb predicted CHD, whereas in women, raised systolic BP and increased levels of FTG were significantly and positively related to CHD. Of the other complications, autonomic neuropathy and total neuropathy were significantly related to CHD in men, whereas retinopathy was related to CHD in women. Further adjustment for center did not alter these results.

The variables that persisted in models with age and diabetes duration adjustment were simultaneously added in multivariate models (Table 3).

In men, baseline age, WHR, and AER were independent predictors of CHD. Current smoking was also related to CHD in men, although it was not statistically significant. Of the other complications, autonomic neuropathy as well as peripheral neuropathy were related to CHD in men with hazard ratios of, respectively, 1.71 (95% CI 1.03–2.85, P = 0.037) and 1.68 (1.01–2.77, P = 0.045), adjusted for age, WHR, and smoking. Further adjustment for hypertension (in a model with age, WHR, and smoking) still showed strong and hardly unaltered hazard ratios for the association between autonomic neuropathy and CHD (hazard ratio [HR] 1.66, 95% CI 0.99–2.74, P = 0.051) and between peripheral neuropathy and CHD (1.70, 1.02–2.82, P = 0.04). In all of these models, the significant estimates for age, WHR, and smoking were not affected.

In women, baseline age and AER were the most important predictors of CHD. Systolic BP and FTG increased the risk of CHD, although these associations only reached borderline statistical significance. Retinopathy was related to CHD in women (HR 1.77, 95% CI 0.93–3.37, P = 0.08) and was adjusted for age, systolic BP, and FTG but was not statistically significant.

Other possibly important risk factors with a less stringent P value of <0.20 (from Table 2) were entered in a forward stepwise manner to estimate independent effects of any additional variables to the models. None of the other variables significantly contributed to the prediction of CHD in men. However, in women, HDL cholesterol (HR 0.74, 95% CI 0.56–0.98) did contribute significantly to the prediction of CHD (independently of age, AER, and systolic BP). Adding in HDL cholesterol made the relationship with FTG disappear due to the high negative correlation between HDL cholesterol and FTG (r = −0.34).

Furthermore, Cox proportional hazards models were carried out in the total population (with more statistical power) using sex-specific variables for age, smoking, WHR, systolic BP, FTG, and AER to estimate which risk factors were more important in men and women. In these analyses, WHR was a more important predictor in men than in women (P = 0.01 in men compared with P = 0.2 in women). Similarly, systolic BP (P = 0.92 in men and 0.003 in women) and FTG (P = 0.82 in men and 0.003 in women) were significant predictors in women but not in men.

Restriction of the CHD events to hard events (omitting ECG-related events) only, bearing in mind the lack of statistical power, did not alter the results. Adjustment for center and BP- and lipid-lowering drugs in the multivariate models did not alter the results.

Incidence rates of CHD were substantially higher in this type 1 diabetic population at ∼1% per year compared with general population incidence rates of 0.1–0.5% per year in those aged 25–74 years (22,23). This compares favorably with the World Health Organization Multinational Study of Vascular Disease in Diabetes with incidence rates of 1.6 and 1.2% per year in men and women, respectively (9), and 2% per year in the Epidemiology of Diabetes Complications (EDC) study.

Higher CHD incidence rates were found in women compared with men in the EURODIAB PCS. Our study suggests clear differences in risk factors predicting CHD in men compared with women, with WHR (and smoking) more important in men and FTG (or HDL cholesterol) and systolic BP more important in women. The EDC study demonstrated the importance of overt nephropathy, WHR, diabetes duration, hypertension, triglycerides, and HDL cholesterol in predicting CHD, but a sex-specific analysis was not performed (8).

We confirm the strong independent relationship of albuminuria with CHD when compared with other risk factors (4,6–9,24,25).

Some studies have shown that an adverse lipid profile is already present at the microalbuminuric stage (26,27). However, limited results have been published for lipid and lipoprotein concentrations in relation to CHD (8,9). Our findings of independent relationships between FTG or HDL cholesterol and CHD have been reported previously (8), but in the only other large cohort study in type 1 diabetic patients, HDL cholesterol was not measured, and no significant relationship was found between FTG and CHD (9).

Our study showed an independent relationship between WHR and CHD, which was not measured or not found to be related to CHD in most previous studies (4,6,7,9,28,29). However, our findings do agree with the EDC study (30). Possible reasons for the discrepancy in findings could be due to the small sample size in previous studies or simply due to the fact that type 1 diabetic patients are generally lean. Even though type 1 diabetic patients are usually lean, weight gain is associated with intensive glycemic control, which could lead to varying the use of insulin as a method to lose or gain weight (31,32). As in our study, several prospective studies did not find a significant relationship between BMI and CHD, even in univariate analyses (7,8,28), perhaps because BMI is a more general marker of obesity, whereas WHR measures, more specifically, central obesity.

Clustering of risk factors, such as triglycerides, elevated BP, central obesity, and perhaps albuminuria as part of the insulin resistance syndrome (33) could play an important role in the pathogenesis of CHD in type 1 diabetes. This proposal is supported by findings from the EDC (34) where estimated glucose disposal rate (a marker for insulin resistance) was prospectively related to lower extremity arterial disease (35).

Conflicting results have been reported regarding the relationship between hyperglycemia and CHD in type 1 diabetic cohort studies, with some studies not finding (6,7,36) and other studies finding (28,37) an independent association. The negative findings reported in our study, which might be due to the difficulty of detecting a statistical difference for GHb values, which are already raised and at the upper end of the distribution, do not imply that optimal glycemic control is not important in the prevention of vascular complications in type 1 diabetes, as clearly demonstrated by the Diabetes Control and Complications Trial (38).

Prior prospective studies in type 1 diabetic patients also showed strong independent associations between elevated BP and CHD (6–9). In the largest of these studies, a higher excess mortality was shown in type 1 than in type 2 diabetic patients, especially when both hypertension and proteinuria were present (39). It is not known, however, whether BP is elevated before the onset of albuminuria or as a consequence of it and further research is needed to explore the pathogenesis (40,41).

In addition to albuminuria, other diabetic complications, such as retinopathy, peripheral, and autonomic neuropathy, were related to the development of CHD. To our knowledge, there are no prospective studies in type 1 diabetic patients that have found strong independent relationships between these complications and CHD. Due to inter-relationships with other established risk factors, it was difficult to separate the effects of these other complications and assess independent associations with CHD.

Differences in type 1 and 2 diabetes can be indicated. In type 1 diabetes, important risk factors for CHD are more diabetes-related risk factors, such as albuminuria and other risk factors such as FTG and WHR. In contrast in type 2 diabetes, the more conventional risk factors (smoking, dyslipidemia, and BP) are important (42). CHD and type 2 diabetes are thought to have common antecedents, the so-called “common soil” hypothesis (43), such as hypertension, insulin resistance, and obesity, whereas type 1 diabetes is a state of insulin deficiency.

The EURODIAB PCS, a clinic-based study with a large sample size, provides a useful European-wide summary as the same standardized methods were used in each center. However, there are some limitations. Just over 50% of CHD events were classified by Minnesota- coded ECGs, and a comparison with baseline ECGs was not performed due to the 7-year gap in coding of baseline and follow-up ECGs. Although viewed as a softer end point than, for example, admission to hospital with an acute event, or a bypass graft operation, the former is at least standardized across centers, whereas as access to health care and thresholds for coronary interventions may differ by center, these latter measures may be less useful for a multicenter study of this type than they first appear. Restriction of the CHD events to hard events (omitting ECG-related events) only, bearing in mind the lack of statistical power, did not alter the results. Of the total baseline sample, 28% participants were lost to follow-up. We showed that these individuals were likely to have a more atherogenic profile than participants in the follow-up. This is likely to mean that we have underestimated the incidence of CHD, and that is true for both men and women. However, our key analyses of interest, the relationship between risk factors at baseline and disease at follow-up, are unlikely to be affected, as it is hard to hypothesize a situation where albuminuria is strongly positively associated with events in responders but negatively associated in nonresponders. Multiple testing could have led to more significant results than there actually are; however, careful statistical analysis and interpretations were made. Furthermore, several studies, although smaller, have confirmed similar risk factors as in our study.

In conclusion, this large study supports the evidence for a strong predictive role of baseline albuminuria in the pathogenesis of CHD in type 1 diabetes. Furthermore, sex-specific risk factors such as systolic BP, FTG (or HDL cholesterol), and WHR were found to be important in the development of CHD and need to be explored further.

The EURODIAB PCS was supported by grants from the Wellcome Trust, the European Community, and Diabetes U.K.

The EURODIAB Prospective Complications Study Group includes B. Karamanos, A. Kofinis, K. Petrou (Hippokration Hospital, Athens, Greece); F. Giorgino, G. Picca, A. Angarano, G. De Pergola, L. Laviola, R. Giorgino (Medicina Interna, Endocrinologia e Malattie Metaboliche, DETO, Università Degli Studi di Bari, Bari, Italy); M. Songini*, A. Casu, M. Pedron, S. Pintus, M. Fossarello* (Department of Internal Medicine, *Ospedale San Michele, Cagliari, Italy); J.B. Ferriss, G. Grealy, D. O’Keefe (Cork Regional Hospital, Cork, Ireland); M. Toeller, C. Arden (Diabetes Research Institute, Heinrich-Heine University, Dusseldorf, Germany); R. Rottiers, C. Tuyttens, H. Priem (University Hospital of Gent, Belgium); P. Ebeling, M. Kylliäinen, V.A. Koivisto (University Hospital of Helsinki, Finland); B. Idzior-Walus, J. Sieradzki, K. Cyganek (Department of Metabolic Diseases, Jagiellonian University, Krakow, Poland); H.H.P.J. Lemkes, M. Krans (University Hospital of Leiden, The Netherlands); J. Nunes-Correa, M.C. Rogado, L. Gardete-Correia, M.C. Cardoso, A. Silva, J. Boavida, M. Machado Sa Marques (Portuguese Diabetic Association, Lisbon, Portugal); G. Michel, R. Wirion, S. Cardillo (Centre Hospitalier, Luxembourg); G. Pozza, R. Mangili, V. Asnaghi (Ospedale San Raffaele, Milan, Italy); E. Standl, B. Schaffler, H. Brand, A. Harms (City Hospital Schwabing, Munich, Germany); Ben Soussan, O. Verier-Mine, P. Fallas, M.C. Fallas (Centre Hospitalier de Valenciennes, France); J.H. Fuller, J. Holloway, L. Asbury, D.J. Betteridge (University College, London); G. Cathelineau, A. Bouallouche, B. Villatte Cathelineau (Hospital Saint-Louis, Paris, France); F. Santeusanio, G. Rosi, V. D’Alessandro, C. Cagini, P. Bottini, P. Reboldi (Instituto di Patologia Medica, Policlinico, Perugia, Italy); R. Navalesi, G. Penno, S. Bandinelli, R. Miccoli, M. Nannipieri (Dipartimento di Endocrinologia e Metabolismo, Pisa, Italy); G. Ghirlanda, C. Saponara, P. Cotroneo, A. Manto, A. Minnella (Universita Cattolica del Sacro Cuore, Rome, Italy); J.D. Ward, S. Tesfaye, S. Eaton, C. Mody (Royal Hallamshire Hospital, Sheffield, U.K.); M. Borra, P. Cavallo Perin, S. Giunti, G. Grassi, G.F. Pagano, M. Porta, R. Sivieri, F. Vitelli, D. Ferrari (Dipartimento di Medicina Interna, Università di Torino and ASO CTO/CRF/Maria Adelaide, Turin, Italy); N. Papazoglou, G. Manes (General Hospital of Thessaloniki, Greece); M. Muggeo, M. Iagulli (V. Cattedra di Malattie del Metabolismo, Verona, Italy); K. Irsigler, H. Abrahamian (Hospital Vienna Lainz, Austria); S. Walford, J. Sinclair, S. Hughes, V. McLelland, J. Ward (New Cross Hospital, Wolverhampton, U.K.); G. Roglic, Z. Metelko, Z.R. Pepeonik, Z. Babic (Vuk Vrhovac Institute for Diabetes, Zagreb, Croatia); C. Ionescu-Tirgoviste, A. Coszma, C. Guja (Clinic of Diabetes, Nutrition and Metabolic Diseases, Bucharest, Romania).

The Steering committee members are J.H. Fuller (London), B. Karamanos, Chairman (Athens), A.-K. Sjolie (Aarhus), N. Chaturvedi (London), M. Toeller (Dusseldorf), G. Pozza, Co-Chairman (Milan), B. Ferriss (Cork), M. Porta (Turin), R. Rottiers (Gent), and G. Michel (Luxembourg).

The coordinating center includes J.H. Fuller, N. Chaturvedi, J. Holloway, D. Webb, L. Asbury, M. Shipley, and S.J. Livingstone (University College, London).

The central laboratories include G.-C. Viberti, R. Swaminathan, P. Lumb, A. Collins, and S. Sankaralingham (Guy’s and St. Thomas Hospital, London, U.K.).