当前位置: 首页 > 期刊 > 《糖尿病学杂志》 > 2006年第12期 > 正文
编号:11410957
The Effect of Intensive Glycemic Treatment on Coronary Artery Calcific
http://www.100md.com 《糖尿病学杂志》
     1 Biostatistics Center, George Washington University, Rockville, Maryland

    2 University of Pittsburgh, Pittsburgh, Pennsylvania

    3 Case Western Reserve University, Cleveland, Ohio

    4 University of California, Irvine, California

    5 Harbor UCLA (University of California, Los Angeles) Medical Center, Torrance, California

    6 University of Toronto, Toronto, Canada

    7 Harvard Medical School, Boston, Massachusetts

    CAC, coronary artery calcification; CT, computed tomography; CVD, cardiovascular disease; DCCT, Diabetes Control and Complications Trial; EDIC, Epidemiology of Diabetes Interventions and Complications; IMT, intima-media thickness; ROC, receiver operating characteristics

    ABSTRACT

    The Epidemiology of Diabetes Interventions and Complications (EDIC) study, an observational follow-up of the Diabetes Control and Complications Trial (DCCT) type 1 diabetes cohort, measured coronary artery calcification (CAC), an index of atherosclerosis, with computed tomography (CT) in 1,205 EDIC patients at 7–9 years after the end of the DCCT. We examined the influence of the 6.5 years of prior conventional versus intensive diabetes treatment during the DCCT, as well as the effects of cardiovascular disease risk factors, on CAC. The prevalences of CAC >0 and >200 Agatston units were 31.0 and 8.5%, respectively. Compared with the conventional treatment group, the intensive group had significantly lower geometric mean CAC scores and a lower prevalence of CAC >0 in the primary retinopathy prevention cohort, but not in the secondary intervention cohort, and a lower prevalence of CAC >200 in the combined cohorts. Waist-to-hip ratio, smoking, hypertension, and hypercholesterolemia, before or at the time of CT, were significantly associated with CAC in univariate and multivariate analyses. CAC was associated with mean HbA1c (A1C) levels before enrollment, during the DCCT, and during the EDIC study. Prior intensive diabetes treatment during the DCCT was associated with less atherosclerosis, largely because of reduced levels of A1C during the DCCT.

    Patients with type 1 diabetes have a high risk of developing cardiovascular disease (CVD), which in young adulthood can be 10-fold higher than in the general population (1,2). The reasons for this increased risk have not been fully elucidated and can only be partly explained by standard cardiovascular risk factors. Surprisingly, cumulative hyperglycemia has not been shown consistently to be a risk factor for cardiovascular events in type 1 diabetes (3–8).

    The Diabetes Control and Complications Trial (DCCT)/Epidemiology of Diabetes Interventions and Complications (EDIC) study provides an opportunity to explore the complex relationships among traditional CVD risk factors, glycemia, and CVD outcomes (9). The DCCT demonstrated the importance of glycemic control in preventing or delaying microvascular complications (10), but it did not reach a clear conclusion with respect to macrovascular complications, owing to the low prevalence of macrovascular disease in the relatively young cohort (11). The long-term EDIC follow-up included assessments of subclinical CVD with measurement of carotid intima-media thickness (IMT) (12,13) and with computerized tomography (CT) of the heart to detect and quantitate calcification in the coronary arteries, a marker of atherosclerosis (14). The EDIC study demonstrated a significant protective effect of prior intensive diabetes therapy, compared with conventional therapy, on the progression of IMT over a 6-year period that was associated with the level of HbA1c (A1C) (13). Progression of IMT was associated with the level of glycemia (A1C) over time, consistent with some (15), but not all (16,17), reports in type 1 diabetes.

    The association of glycemia with coronary artery calcification (CAC) in type 1 diabetes is unclear. Two studies (18,19) failed to demonstrate a relationship between glycemia and CAC in type 1 diabetes, whereas one did (20). These studies showed relationships with traditional CVD risk factors, but a smaller sex difference in CAC than is usual for the nondiabetic population, consistent with the reduction in the difference between sexes for CAD in type 1 diabetes.

    The DCCT/EDIC study has recently reported (21) that DCCT intensive therapy significantly reduced the long-term risk of clinical CVD by 42%; however, the cumulative incidence of such events remains low, precluding multivariate analyses at this time. Measurement of CAC provides an opportunity to assess the effects of putative and established CVD risk factors on the progression of atherosclerosis in type 1 diabetes that may ultimately translate into effects on risk of clinical events.

    Herein we assess the long-term effects of original DCCT (conventional versus intensive) treatment assignment on the degree of CAC measured 8 years after the completion of the DCCT. We also examine the association of CAC with history of glycemia, with other risk factors and markers of CVD, and with clinically prevalent CVD.

    RESEARCH DESIGN AND METHODS

    Between 1983 and 1989, the DCCT enrolled 1,441 subjects with type 1 diabetes who, at baseline, were 13–39 years of age, had type 1 diabetes for 1–15 years, and were in generally good health (10). The DCCT consisted of two cohorts: the primary prevention cohort had type 1 diabetes for 1–5 years, no retinopathy, and urinary albumin excretion <40 mg per 24 h at baseline; the secondary intervention cohort had type 1 diabetes for 1–15 years, very mild to moderate nonproliferative retinopathy, and urinary albumin excretion 200 mg per 24 h at baseline. At the end of the DCCT in 1993, after 6.5 years of mean follow-up, intensive therapy was recommended for all subjects, and they returned to their own health care providers for diabetes care. In 1994, 1,375 (96%) of the 1,425 surviving members volunteered to participate in the EDIC observational follow-up study (9).

    Computed tomography of coronary calcification.

    Computed tomography (CT) was performed between November 2000 and March 2003 (11–20 years after enrolment into the DCCT, 7–9 years after its end) in 1,205 (86%) of the surviving 1,404 participants, with specific patient consent. CT was performed in 19 scanning sites (see APPENDIX) using a C-150 cardiac-gated electron beam CT scanner (n = 9; Imatron, San Francisco, CA), a Lightspeed (n = 7; General Electric Medical Systems, Waukesha, WI) or a Volume Zoom (Siemens, Erlanger, Germany) multidetector CT system, a Lightspeed Marconi MX-8000 (GE), or a Somatom 4+ (Siemens) (n = 3). All participants were scanned twice over calibration phantoms of known physical calcium concentration.

    Scans were read centrally at the Harbor-UCLA (University of California, Los Angeles) Research and Education Institute (Torrance, CA) to identify and quantify CAC, calibrated according to the readings of the phantom using the method of Agatston et al. (22). The average score from the two scans was used in the analysis. Readers were masked to subject identity and prior treatment assignment.

    Scans were evaluated by the staff at the reading center on seven criteria: motion artifact, streak artifact, phantom placement, slice registration, lack of noise, axis coverage, and xy axis coverage. The 19 scanning centers were monitored monthly on these criteria. The intra- and interreader precision was evaluated with the use of a set of standard scans that were reread by the same reader and another reader at the reading center. The kappa measure of intrareader agreement beyond chance for the presence or absence of calcification was 0.81, and the interreader kappa was 0.86. The coefficient of reliability for the calcification scores was 0.99 for both inter- and intrareader.

    Other procedures.

    Each EDIC subject had an annual history, physical examination, electrocardiogram, and laboratory testing, including serum creatinine and A1C, determined as they were during the DCCT (9,10,23). Fasting lipid profiles and 4-h urine collections for measurement of albumin excretion rate and creatinine clearance were obtained in alternate years during the EDIC study (9). Carotid IMT was measured by B-mode ultrasonography in 1994 and again in 2000–2001 (12,13). Combined IMT was defined as the sum of the standardized intima-media measurements of the common and internal carotid arteries. Standardized IMT was defined as: (variable – mean) ÷ SD, as described by O’Leary et al. (24).

    Hypertension was defined as systolic blood pressure 140 mmHg or diastolic blood pressure 90 mmHg, documented hypertension, or use of antihypertensive agents. Hypercholesterolemia was defined as LDL cholesterol 130 mg/dl or use of lipid-lowering agents. Cardiovascular clinical outcomes included fatal and nonfatal myocardial infarctions, stroke, revascularization (bypass surgery or angioplasty), angina confirmed with a positive exercise tolerance test or angiogram, or silent myocardial infarctions identified using criteria applied to the follow-up electrocardiogram. For classification of cardiovascular clinical outcomes, information from death certificates; medical records from hospitalizations; autopsy reports, if available; and interviews with the participants were reviewed by the EDIC mortality and morbidity committee, masked to treatment assignment, and the event was classified according to specified guidelines (25).

    Statistical analyses.

    Analyses were conducted using SAS (26). Clinical characteristics were compared using Wilcoxon’s rank-sum test for continuous quantitative variables and 2 or Fisher’s exact tests for categorical variables. Analyses used the prevalence of CAC scores of >0 and >200 Agatston units; the latter has been a predictor of CVD events in other studies (18,20,27). The Mantel-Haenszel 2 test of nonzero correlation was used to test for a linear trend in proportions (28). The stratified adjusted Mantel-Haenszel analysis adjusted for other qualitative covariate effects on the OR or test of trend (28). Homogeneity of treatment effect over strata was assessed by the Breslow-Day test (28). Logistic regression examined the relationship between covariates and the prevalence of CAC (28,29). The entropy R2 coefficient was used to describe the proportion of variation in risk explained by the model (29).

    Tobit-censored regression models (30) assessed covariate effects on the observed CAC score, which is a mixture of a discrete random variable (any measurable calcification, yes/no) and a quantitative random variable (the amount of CAC, if measurable). The Tobit regression coefficient represents the log ratio of the geometric mean CAC score per unit increase in the covariate, assuming some true measurable calcification for all subjects, including those with undetectable levels. Ordinary multiple regression of the observed measurable values is biased, whereas logistic regression of the prevalence of measurable calcification is inefficient (31). The Tobit model was fit using the LIFEREG procedure in SAS (26), where the natural logarithm of the nonzero CAC scores was reduced by subtracting the natural logarithm of the lowest detectable CAC score (the lower limit of quantification of CAC). Goodness of fit was assessed by applying the Hosmer-Lemeshow test to the estimated probabilities from the Tobit model as a predictor for presence of CAC, and by the Spearman correlation coefficient between the measured score of CAC and that predicted from the Tobit model (32).

    The intent-to-treat effects of prior DCCT conventional versus intensive therapy were assessed in basic logistic and Tobit models, adjusted for baseline age, type 1 duration, sex, scanning site, and DCCT baseline retinopathy cohort (primary versus secondary). Additional multivariate models explored effects of covariates measured up to the time of CAC measurement. The most significant factor among similar variables (e.g., systolic and diastolic blood pressure) was used. Backward elimination was used to select two-way interactions with treatment group (28,29), retaining those nominally significant at P 0.05. In models with an interaction, the overall treatment effect was assessed using a test of both the group and the group by covariate interaction on 2 degrees of freedom (df) (29). Receiver operating characteristics (ROC) plots were used to describe the sensitivity and specificity of CAC for prevalent CVD (21,33–35).

    RESULTS

    Clinical characteristics.

    Table 1 shows the clinical characteristics of the participants in the CT study at DCCT baseline and at the exam immediately before, or at time of, the CT scan, stratified by sex and original DCCT treatment group. Of the 1,205 participants, 95% were Caucasian, and 53% were male. In both male and female subjects, blood pressure and lipids were not significantly different between the former DCCT conventional- and intensive-treatment groups at the time of the CT scans. In men only, prevalence of hypertension and aspirin use were significantly lower in the intensive group. The A1C level before CT was similar in the treatment groups in both sexes, but mean A1C level during the 9 years of EDIC follow-up before the CT scan was slightly higher in the conventional than in the intensive group in men (8.2 vs. 8.0 ± 1.1%, P < 0.01). Albumin excretion rates were significantly lower in the intensive than in the conventional-treatment group (P < 0.01). Measurements of adiposity were not different between the treatment groups. A comparison of the EDIC subjects who participated in the CT study with those who did not participate revealed similar distributions of sex, race, and treatment group. However, the participants who did not have CT scans were younger (25 vs. 27 years at DCCT entry) and had higher mean A1C (8.5 vs. 8.2%, P < 0.003) during the DCCT.

    CAC distributions.

    Figure 1 shows the distribution of CAC scores, including the prevalence of CAC = 0, 1–200, and >200 Agatston units by treatment group stratified by primary prevention and secondary intervention cohorts. These analyses showed differences between treatment groups in the primary cohort (P = 0.03), but not in the secondary cohort (P = 0.41), not adjusted for scanning site or other baseline factors.

    Figure 2A–D shows the prevalences of CAC >0 and >200 Agatston units within each treatment group, stratified by sex and also by decade of age at the time of the scan. The overall prevalences for CAC >0 and >200 Agatston units were 31.0 and 8.5%, respectively. The prevalences of CAC >0 and >200 Agatston units increased linearly by decade of age within each sex (P < 0.01 for each). Within both sexes, and within each treatment group, the prevalence of calcification increased significantly with age, except for CAC >200 Agatston units in the intensive group among women, for which the overall incidence was lower than for men or conventional group women, especially among those 50 years.

    DCCT treatment group differences.

    In the primary cohort, compared with the conventional-treatment group, the intensive-treatment group had a significantly lower prevalence of CAC >0 Agatston units (21.7 vs. 29.8%, respectively), with an adjusted odds ratio (OR) for conventional versus intensive therapy of 1.59 (95% CI 1.06–2.39, P = 0.024). By contrast, there was no treatment effect in the secondary cohort (OR 0.94). The difference in ORs between cohorts (i.e., group by cohort interaction) was barely significant (P = 0.049).

    The adjusted effect of conventional versus intensive therapy on the prevalence of CAC >200 Agatston units within the primary prevention cohort (OR 2.13) was not significantly different from that in the secondary intervention cohort (1.50, P = 0.474 for test of interaction). For the two cohorts combined, the adjusted treatment group effect had an OR of 1.65 (95% CI 1.06–2.58, P = 0.026).

    In a Tobit regression analysis of the log(CAC) score in the primary prevention cohort, allowing for nonmeasurable values, the conventional-treatment group on average had a 3.7-fold higher CAC score than the intensive group (95% CI 1.3–10.5, P = 0.014). However, there was no treatment group difference within the secondary intervention cohort, the geometric mean ratio being 1.00 (0.4–2.5, P = 0.87). The difference between the primary versus secondary cohorts (i.e., the 3.7 vs. 1.0 treatment group effect) approached statistical significance (P = 0.060 for the test of interaction). Further analyses in the secondary intervention cohort revealed a significant group by diabetes duration interaction effect (P = 0.0008) and the overall treatment group effect with 2 df was significant (P = 0.003). The treatment effect increased with longer duration of diabetes at DCCT baseline, with the geometric mean ratio 1.4-fold higher for each additional year of diabetes duration (Fig. 3). This interaction was diminished after adjusting for A1C at DCCT entry, smoking, hypertension, and waist-to-hip ratio at the time of CT scan, but it remained significant (P = 0.0097).

    A further analysis adjusting for the differences in the log mean A1C between treatment groups during the DCCT explained virtually all of the treatment group effect within the primary prevention cohort. For the Tobit model, the treatment group effect (P = 0.01) became nonsignificant (P = 0.69) after adjustment for A1C. In the Tobit model within the primary prevention cohort, and with the above baseline factors, a 10% increase in the DCCT mean A1C was associated with a 1.85-fold increase in the geometric mean CAC score (P < 0.0001); within the secondary intervention cohort, it was associated with a 1.32-fold increase (P = 0.051). The difference between cohorts (test of interaction) was not significant (P = 0.21).

    Correlation between CAC and CAD risk factors.

    Univariate rank correlations, partially adjusted for DCCT baseline age and sex, assessed the association between the prevalence of CAC >0 Agatston units, CAC >200 Agatston units, and the log CAC, with covariates (Table 2). The results were similar across all of the CAC outcomes. Attained age (r = 0.34), male sex (r = 0.21), waist-to-hip ratio (r = 0.14), combined common and internal carotid IMT at year 6 (r = 0.17), hypercholesterolemia (r = 0.14), and hypertension (r = 0.15) had the strongest association with the log CAC. There was no significant correlation with ankle-to-arm blood pressure ratio. Other weaker or less consistent correlates were with smoking, DCCT baseline retinopathy status, microalbuminuria, menopausal status, and a parental history of type 2 diabetes. Neither total cholesterol nor LDL cholesterol levels were significant correlates, whereas triglyceride level was positively and HDL cholesterol level was negatively correlated with CAC. A1C before the DCCT, the mean levels during the DCCT, at DCCT closeout, before the CAC scan, mean during EDIC, and mean during DCCT and EDIC all correlated with CAC (P < 0.01).

    Multivariate regression models.

    Expanded multivariate regression models assessed the association of additional risk factors, measured either at DCCT baseline or close to the time of the CT scan during EDIC, with CAC. In a logistic model, hypercholesterolemia, increased waist-to-hip ratio, and smoking were significantly associated with prevalence of CAC >0 Agatston units. As observed in the intent-to-treat analysis of the DCCT treatment group effect, there was a significant interaction between treatment group and baseline retinopathy strata (primary versus secondary cohort). Overall, the model explained 20% (R2) of the variation in risk. A logistic regression model with the prevalence of CAC >200 Agatston units produced similar results.

    The multivariate Tobit regression model (Table 3) provided results similar to the two logistic regression models. The CAC score increased 1.3-fold per year of age, was 3.4-fold more for men than women, was 7.1-fold greater among current smokers, and increased 1.3-fold per 10 mg per 24 increase in albumin excretion rate, 2.6-fold per 10% increase in waist-to-hip ratio, 2.8-fold among those with hypercholesterolemia, 2.8-fold among those with hypertension, and 1.4-fold per 10% increase in DCCT mean A1C. The Hosmer-Lemeshow test (P = 0.14), comparing the Tobit model predicted probability of having CAC to the detected presence of CAC, and the significant Spearman correlation coefficient (0.4781, P < 0.0001) between the predicted and observed CAC scores both suggest that this model fits the observed data.

    Sensitivity and specificity.

    The sensitivities and specificities of CAC >0 and >200 Agatston units for nonfatal myocardial infarction (n = 10) were 80 and 69.4% and 70 and 92.1%, respectively. The sensitivities and specificities for all CVD events (see RESEARCH DESIGN AND METHODS, n = 44) for CAC >0 and >200 Agatston units were 72.7 and 70.5% and 47.7 and 93.0%, respectively. The ROC curve (Fig. 4) reflects the characteristics of the CAC score with discriminating power for CVD. CAC >0 Agatston units showed the highest (29%) false-positive rate and highest (73%) true-positive rate. The false- and true-positive rates for CAC >200 Agatston units were 7 and 44%, respectively. The area under the ROC curve was 0.78 (95% CI 0.69–0.86).

    DISCUSSION

    The most important and new observation of this study is that the prevalence of CAC, and the CAC scores, in the DCCT/EDIC type 1 diabetes cohort are significantly lower in the former intensive treatment compared with the former conventional-treatment group. The prevalence of a clinically significant CAC score of >200 Agatston units was 7.0% in the former intensive-treatment group and 9.9% in the former conventional-treatment group.

    The beneficial effect of prior intensive therapy during the 6.5 years of the DCCT was greater among those entered into the primary prevention cohort with 1–5 years’ diabetes duration on entry, among other factors, than among those entered into the secondary intervention cohort with 1–15 years’ duration. The beneficial effect of intensive therapy is largely attributable to the differences between groups in the level of A1C during the DCCT.

    The lesser treatment effect in the secondary intervention cohort appeared to be the result of an interaction between baseline diabetes duration and treatment group (Fig. 3). The beneficial treatment effect at longer durations was washed out by the negative treatment effect at shorter durations, mainly 1–5 years. The absence of a treatment effect among those in the secondary intervention cohort with only 1–5 years’ duration could be an artifact of the selection criteria. Prior epidemiological modeling (36) suggests that subjects who have retinopathy present after only 1–5 years’ duration tend to have higher preexisting levels of A1C that could in turn diminish the long-term effectiveness of intensive therapy on other outcomes, such as CAC. In addition, there were some imbalances between treatment groups in the subcohort of the secondary intervention cohort with 1–5 years’ diabetes duration. Most importantly, compared with the conventional-treatment patients, intensive-treatment patients had higher A1C levels at DCCT entry (9.6 vs. 8.8%).

    The 31% prevalence of CAC >0 Agatston units in the DCCT/EDIC cohort is lower than in other reports of type 1 diabetic cohorts (18–20,37). This may reflect a substantially lower CVD risk, based on eligibility criteria and lower mean A1C, than in other studies. In addition, unlike some (18–20) but not all (37) reports, female subjects in the DCCT/EDIC study had a lower prevalence of CAC level in all age-groups. The male-to-female OR of 2.7 for CAC >0 Agatston units in our cohort is similar to the OR of 2.5 reported in a previous study of type 1 diabetes (37).

    Most previous studies in type 1 diabetes have not shown a significant association between glycemic control and CVD or CAD events, either cross-sectionally (3–5) or prospectively (6). Moreover, two previous studies have not shown a relationship between glycemic control and CAC (18,19). However, a recent report (20) showed a greater risk of progression of CAC over 3 years among those with A1C >7.5% than patients with A1C <7.5%. Perhaps owing to the small numbers who progressed (21 of 109), A1C was not a significant univariate risk factor. In contrast, the DCCT/EDIC study, with a much larger sample size and detailed longitudinal assessment of metabolic control, has shown that A1C measured before DCCT enrollment, mean A1C during the DCCT (which had the strongest correlation), and mean level during the EDIC study were each significantly correlated with CAC after adjustment for age and sex, and were independent of waist-to-hip ratio, smoking, hypertension, and hypercholesterolemia before CT scan. The DCCT cohort was selected to exclude patients with hypertension and hypercholesterolemia on entry to the DCCT, and the impact of hyperglycemia may have become more apparent without the influence of these conventional CVD risk factors at baseline.

    In univariate analyses, most of the known risk factors for coronary artery disease, including age, smoking, systolic blood pressure, hypertension, waist-to-hip ratio, hypercholesterolemia, HDL cholesterol and triglyceride levels, and microalbuminuria, were significantly associated with the presence of CAC in the DCCT/EDIC study. Although both microalbuminuria and hypertension were also reduced by prior intensive treatment (38), the effect of glycemic control on coronary artery calcium control was independent of these two risk factors. The LDL cholesterol level was not by itself a significant continuous risk factor for the prevalence of coronary artery calcium in the DCCT/EDIC study.

    In the Tobit multivariate regression model, the mean CAC score was significantly greater in older individuals, men, and those with higher waist-to-hip ratios, retinopathy at DCCT baseline, and higher albumin excretion rates. The association of early signs of microvascular complications with later CAC may reflect some common elements in their development and in the pathogenesis of atherosclerosis, such as hypertension or hyperglycemia.

    We previously reported that the rate of progression of carotid artery IMT from DCCT closeout to 6 years later was reduced by prior intensive treatment (13). This effect was largely explained by the difference in A1C levels that existed between the two prior treatment groups. We have now shown that CAC correlates significantly with carotid IMT that was measured 1–3 years earlier (Table 2). Taken together, these prolonged effects of intensive treatment on both coronary artery calcium and carotid IMT support the interpretation that lowering glycemia with intensive treatment results in a slowing of atherosclerosis.

    The mechanism(s) by which hyperglycemia causes atherosclerosis is incompletely understood but could involve long-lived advanced glycation end products (39,40) in vessel walls and their interactions with advanced glycation end product receptors (41). The demonstration of a delayed benefit from intensive treatment on atherosclerosis is also consistent with the persistent beneficial effect of prior intensive treatment on retinopathy and nephropathy during the EDIC study (38,42). The observation of a treatment effect principally within the primary prevention cohort with a preexisting mean duration of diabetes of 2.5 years on entry adds further weight to the recommendation that intensive treatment be started as early in the course of type 1 diabetes as safely and practically possible.

    Certain limitations of this study should be noted. A baseline assessment of CAC was not obtained either at the beginning of the DCCT or at the beginning of EDIC study. Although such assessments are not necessary to document a treatment effect on CAC levels, without baseline levels it is not possible to describe the magnitude or rate of change in CAC over time. Theoretically, it is possible that these findings simply reflect a chance baseline imbalance in the CAC levels. However, this appears unlikely given the similarity of characteristics relevant to atherosclerosis in the conventional-treatment and intensive-treatment groups at DCCT baseline (9,10). Although CAC scores measured by electron-beam CT and multidetector CT correlate very well (43), the use of several different machines to measure CAC may have added variability to the measurements and interfered with our ability to establish correlations. However, even if we looked separately at the results from the nine electron-beam CT scanners and the results from the other scanners, the trend in differences between intensive- and conventional-treatment groups remained (data not shown).

    The overwhelming majority of the subjects did not have any detectable levels of calcification, thus reducing the sensitivity of these analyses to detect an effect of intensive therapy, or to assess the relative effects of glycemia and known risk factors. The low prevalence of measurable CAC also limits the ability to assess the predictive value of CAC for future macrovascular events. Nevertheless, significant risk factor effects were observed, and an ROC analysis showed significant association between CAC score and prevalence of overt CVD at the time of assessment. A recent DCCT/EDIC analysis showed that intensive therapy reduced the risk of aggregate CVD events by 42% and the risk of major clinical events (fatal or nonfatal myocardial infarction or stroke) by 57% (each P < 0.02) (21). However, the predictive value of a given level of CAC for the risk of a future CVD event cannot be definitively assessed, owing to the small number of such events that were observed after the measurement of CAC.

    Finally, although coronary artery calcium is a quantitative marker of coronary atherosclerosis burden (44,45) and predicts coronary artery disease measured by angiography (46,47), it is not known whether the reduction in the prevalence of CAC with intensive therapy will translate into a reduction in the incidence of coronary artery disease and other CVD events (48,49). The usefulness of screening asymptomatic patients with CT is controversial (50,51). The majority of DCCT/EDIC participants with a mean age of 43 years at the time of CAC measurement displayed no detectable coronary artery calcium. Further follow-up with assessment of clinical events will permit us to assess the predictive power of CAC for incident CVD events.

    APPENDIX

    Study chairmen: S. Genuth, D. Nathan, and O. Crofford (past).

    Albert Einstein College of Medicine: S. Engel, J.B. Friday, H. Martinez (past), H. Shamoon, and H. Engel; Case Western Reserve University: W. Dahms, L. Mayer, S. Pendegras, H. Zegarra, D. Miller, L. Singerman, S. Smith-Brewer, and S. Genuth (past); Cornell University Medical Center: D. Brillion, M. Lackaye, M. Heinemann, V. Reppuci, and T. Lee; Henry Ford Health System: F. Whitehouse, D. Kruger, A. Galpern, and J.D. Carey; International Diabetes Center: R. Bergenstal, M. Johnson, D. Kendall, M. Spencer, D. Noller, K. Morgan, and D. Etzwiler (deceased); Joslin Diabetes Center: A. Jacobson, E. Golden, G. Sharuk, Paul Arrigg, R. Baeser, O. Ganda, J. Rosenzweig, H. Wolpert, P. Economides, O. Handy, and L. Rand (past); Massachusetts General Hospital: D. Nathan, M. Larkin, S. Fritz (past), J. Godine, C. McKitrick, and P. Lou; Mayo Foundation: F.J. Service, G. Ziegler, J. Pach, and J. Lindsey; Medical University of South Carolina: J. Colwell, D. Wood, R. Mayfield, K. Hermayer, M. Szpiech, T. Lyons, J. Parker, A. Farr, S. Elsing, T. Thompson, J. Selby, and M. Bracey; Northwestern University: M. Molitch, B. Schaefer, L. Jampol, D. Weinberg, A. Lyon, Z. Strugula, J. Shankle, and P. Astlesford; University of California, San Diego: O. Kolterman, G. Lorenzi, and M. Goldbaum; University of Iowa: W. Sivitz, M. Bayless, R. Zeither (past), T. Weingeist, E. Stone, H. Culver Boidt, K. Gehres, and S. Russell; University of Maryland School of Medicine: M. Hebdon, D. Counts, S. Johnsonbaugh, A. Kowarski (past), D. Ostrowski (past), T. Donner, S. Steidl, and B. Jones; University of Michigan: W. Herman, D. Greene (past), C. Martin, M.J. Stevens, A.K. Vine, and S. Elner; University of Minnesota: J. Bantle, B. Rogness, T. Olsen, and E. Steuer; University of Missouri: D. Hainsworth, D. Goldstein (past), S. Hitt, and J. Giangiacomo; University of New Mexico: D. Schade, M. Burge, J. Canady, M. Schluter, A. Das, and D. Hornbeck (past); University of Pennsylvania: S. Schwartz, P.A. Bourne, B.J. Maschak-Carey (past), L. Baker (deceased), S. Braunstein, and A. Brucker; University of Pittsburgh: T. Orchard, N. Silvers, T. Songer, B. Doft, S. Olson, R.L. Bergren, L. Lobes, M. Fineman, and A. Drash (past); University of South Florida: J. Malone, E.A. Tanaka, J. Vaccaro-Kish (past), C. Berger, R. Gstalder, P.R. Pavan, and A. Morrison; University of Tennessee: S. Dagogo-Jack, C. Wigley, S. Schussler (past), A. Kitabchi, H. Lambeth (past) M.B. Murphy, S. Moser, D. Meyer, A. Iannacone, M. Bryer-Ash (past), and E. Chaum; University of Texas Southwestern University Medical Center: P. Raskin, S. Strowig, A. Edwards, J. Alappatt (past), C. Wilson (past), S. Park (past), and Y. He; University of Toronto: B. Zinman, A. Barnie, S. MacLean, R. Devenyi, M. Mandelcorn, and M. Brent; University of Washington: J. Palmer, S. Catton, J. Kinyoun, L. Van Ottingham (past), and J. Ginsberg (past); University of Western Ontario: J. Dupre, J. Harth, C. Canny (past), and D. Nicolle; Vanderbilt University: M. May, J. Lipps, R. Lorenz (past), L. Survant, S. Feman (past), K. Tawansy, A. Agarwal, and T. Adkins; Washington University, St. Louis: N. White, L. Levandoski, J. Santiago (deceased), I. Boniuk, G. Grand, M. Thomas, D. Burgess, D. Joseph, K. Blinder, and G. Shah; Yale University School of Medicine: W. Tamborlane, P. Gatcomb, K. Stoessel, and K. Taylor.

    CT scanning sites: University of Pittsburgh: Dan Edmundovitch, Steven Reis, Donna McKinley, and Lori Givens; Cleveland Clinic: Art Stillman, Richard White, Stacie Kumiac, Wadih Karim, and Sandy Halliburton; Johns Hopkins University: Elliot Fishman, Joao Lima, John Phelps, and Susan Nigro; University of Pennsylvania: Warren Gefter, and Richard Poynton; St Francis Dematteis Center, Roslyn, NY: Allen Guerci, Yadon Arad, Bill Schoepfer, Michael Cummings, Stephanie Kampfer, Kathleen Rapelte, and Therese Rhatigan; Beth Israel, Boston, MA: Melvin Clause, Dawn Weeks, Betty Scholtz, Robert Butler, Sue Hansen, and Usama Abraham; Mayo Clinic: Patrick Sheedy, Barbara Kruger, and Colleen Braun; University of Minnesota: Charles Dietz, Robert Rickard, Dan Busian, Gregory Snyder, Jessica England, Sandy Pash, Ron Joslin-Zirngible, and John Borchert; University of Iowa: William Stanford, Scott Heery, and Mary Burr; Lifetest, Nashville, TN: Michael Rosinbaum, Janet Calhoun, and Taige Weldy; LifeScore, San Diego, CA: Michael Wright, Marianne Saulino, and Don Davis; Medical University of South Carolina (MUSC): Bayne Selby, Alden Finlayson, and Karen Parish, Laurie Oberholtzer, Judy Webber, Adnrea Bantz, and Becky Orr; University of Texas, SW Dallas, TX: Ronald Peshock, Brian Fox, Jerri Payne, Tommy Tillary, and Beth Murray; Northwestern University: Eric Hart, George Mahoney, Peter Dubbs, and David Channin; Heart Hospital of New Mexico: Martha England, Alan Cravitz, Chris Reese, and Steve Colson; St Johns, Detroit, MI: Julius Gardin, Terri Reagan, Michele Vigliotti, Tina Markovic; Center for Diagnostic Imaging (CDI Seattle): Beau Bordner, Randy Jorgensen, and Shane Miller; BodyScan Imaging, Tampa, FL: Laura Van Horn, Jane Totten, and Delores Valari.

    Clinical Coordinating Center (Case Western Reserve University): B. Dahms, R. Trail, and J. Quin. Data Coordinating Center (The George Washington University, Biostatistics Center): J. Lachin, P. Cleary, D. Kenny (past), J. Backlund, W. Sun, B. Rutledge, B. Waberski, K. Klump, K. Chan, L. Diminick, B. Petty (past), A. Determan (past), and M. Hawkins; National Institute of Diabetes and Digestive and Kidney Disease Program Office: C. Cowie, J. Fradkin, C. Siebert (past), R. Eastman (past); Central Fundus Photograph Reading Center (University of Wisconsin): M. Davis, R. Danis, L. Hubbard, P. Geithman, L. Kastorff, M. Neider, D. Badal, B. Esser, K. Miner, H. Wabers, K. Glander, J. Joyce, N. Robinson, C. Hurtenbach, and C. Hannon; Central Biochemistry Laboratory (University of Minnesota): M. Steffes, J. Bucksa, and B. Chavers; Central Carotid Ultrasound Unit (New England Medical Center): D. O’Leary, L. Funk, and J. Polak; Central ECG Reading Unit (University of Minnesota): R. Crow, C. O’Donnell (past), B. Gloeb, and S. Thomas; Computed Tomography Reading Center (Harbor UCLA Research and Education Institute): R. Detrano, N. Wong, M. Fox, L. Kim, R. Oudiz; External Advisory Committee: G. Weir (Chairman), C. Clark, R. D’Agostino, M. Espeland, B. Klein, T. Manolio, L. Rand, D. Singer, and M. Stern; Molecular Risk Factors Program Project (Medical University of South Carolina): M. Lopes-Virella, W. T. Garvey, T. J. Lyons, A. Jenkins, R. Klein, G. Virella, A.A. Jaffa, D. Zheng, D. Lackland, D. McGee, R.K. Mayfield, and M. Brabham; Genetic Studies Group (Hospital for Sick Children): A. Boright, A. Paterson, S. Scherer, and B. Zinman; Lipoprotein Distribution/Obesity Group (University of Washington: J. Brunzell, J. Hokanson, S. Marcovina, J. Purnell, S. Sibley, S. Deeb, and K. Edwards.

    Editor, EDIC Publications: D. Nathan.

    ACKNOWLEDGMENTS

    Supported by contracts with the Division of Diabetes, Endocrinology and Metabolic Disease of the National Institute of Diabetes and Digestive and Kidney Diseases and by the General Clinical Research Center Program, National Center for Research Resources, National Institutes of Health.

    FOOTNOTES

    A complete list of individuals and institutions participating in the DCCT/EDIC Research Group appears in the APPENDIX.

    The Writing Group of the DCCT/EDIC Research Group takes responsibility for the contents of this article.

    The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

    REFERENCES

    Krolewski AS, Kosinski EJ, Warram JH, Leland OS, Busick EJ, Asmal AC et al: Magnitude and determinants of coronary artery disease in juvenile-onset, insulin-dependent diabetes mellitus. Am J Cardiol 59:750–755, 1987

    Dorman JS, Laporte RE, Kuller LH, Cruickshanks KJ, Orchard TJ, Wagener DK, Becker DJ, Cavender DE, Drash AL: The Pittsburgh insulin-dependent diabetes mellitus (IDDM) morbidity and mortality study: mortality results. Diabetes 33:271–276, 1984

    Orchard TJ, Olson JC, Erbey JR, Williams K, Forrest KY, Smithline Kinder L, Ellis D, Becker DJ: Insulin resistance-related factors, but not glycemia, predict coronary artery disease in type 1 diabetes: 10-year follow-up data from the Pittsburgh Epidemiology of Diabetes Complications Study. Diabetes Care 26:1374–1379, 2003

    Forrest KY, Becker DJ, Kuller LH, Wolfson SK, Orchard TJ: Are predictors of coronary heart disease and lower-extremity arterial disease in type 1 diabetes the same A prospective study. Atherosclerosis 148:159–169, 2000

    Koivisto VA, Stevens LK, Mattock M, Ebeling P, Muggeo M, Stephenson J, Idzior-Walus B: Cardiovascular disease and its risk factors in IDDM in Europe: EURODIAB IDDM Complications Study Group. Diabetes Care 19:689–697, 1996

    Fuller JH, Stevens LK, Wang SL: Risk factors for cardiovascular mortality and morbidity: the WHO multinational study of vascular disease in diabetes. Diabetologia 44 (Suppl. 2):S54–S64, 2001

    Lehto S, Rnnemaa T, Pyrl K, Laakso M: Poor glycemic control predicts coronary heart disease events in patients with type 1 diabetes without nephropathy. Arterioscler Thromb Vasc Biol 19:1014–1019, 1999

    Moss SE, Klein R, Klein BE, Meuer SM: The association of glycemia and cause-specific mortality in a diabetic population. Arch Int Med 154:2473–2479, 1994

    Epidemiology of Diabetes Interventions and Complications (EDIC) Research Group: Epidemiology of Diabetes Interventions and Complications (EDIC): design, implementation, preliminary results of a long-term follow-up of the Diabetes Control and Complications Trial cohort. Diabetes Care 22:99–111, 1999

    Diabetes Control and Complications Trial Research Group: The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 329:977–986, 1993

    The DCCT Research Group: Effect of intensive diabetes management on macrovascular events and risk factors in the Diabetes Control and Complications Trial. Am J Cardiol 75:894–903, 1995

    Epidemiology of Diabetes Interventions and Complications (EDIC) Research Group: Effect of intensive diabetes treatment on carotid artery wall thickness in the epidemiology of diabetes interventions and complications. Diabetes 48:383–390, 1999

    Nathan DM, Lachin J, Cleary P, Orchard T, Brillon DJ, Backlund JY, O’Leary DH, Genuth S; Diabetes Control and Complications Trial; Epidemiology of Diabetes Interventions and Complications Research Group: Intensive diabetes therapy and carotid intima-media thickness in type 1 diabetes mellitus. N Engl J Med 348:2294–2303, 2003

    Zinman B, Orchard T, Cleary P, Genuth S, Jacobson A, Wong N, Detrano R, Backlund JC: Coronary Calcification in the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Cohort. Diabetologia 46 (Suppl. 2):A23; 59, 2003

    Kanters SD, Algra A, Banga JD: Carotid intima-media thickness in hyperlipidemic type I and type II diabetic patients. Diabetes Care 20:276–280, 1997

    Yamasaki Y, Kawamori R, Matsushima H, Nishizawa H, Kodama M, Kajimoto Y, Morishima T, Kamada T: Atherosclerosis in carotid artery of young IDDM patients monitored by ultrasound high-resolution B-mode imaging. Diabetes 43:634–639, 1994

    Peppa-Patrikiou M, Scordili M, Antoniou A, Giannaki M, Dracopoulou M, Dacou-Voutetakis C: Carotid atherosclerosis in adolescents and young adults with IDDM: relation to urinary endothelin, albumin, free cortisol, and other factors. Diabetes Care 21:1004–1007, 1998

    Olson JC, Edmundowicz D, Becker DJ, Kuller LH, Orchard TJ: Coronary calcium in adults with type 1 diabetes: a stronger correlate of clinical coronary artery disease in men than in women. Diabetes 49:1571–1578, 2000

    Colhoun HM, Rubens MR, Underwood R, Fuller JH: The effect of type 1 diabetes mellitus on the gender difference in coronary artery calcification. J Am Coll Cardiol 36:2160–2167, 2000

    Snell-Bergeon JK, Hokanson JE, Jensen L, MacKenzie T, Kinney G, Dabelea D, Eckel RH, Ehrlich J, Garg S, Rewers M: Progression of coronary artery calcification in type 1 diabetes: the importance of glycemic control. Diabetes Care 26:2923–2928, 2003

    Nathan DM, Cleary PA, Backlund JY, Genuth SM, Lachin JM, Orchard TJ, Raskin P, Zinman B; Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Study Research Group: Intensive diabetes treatment and cardiovascular disease in patients with type 1 diabetes. N Engl J Med 353:2643–2653, 2005

    Agatston AS, Janowitz WH, Hildner FJ, Zusmer NR, Viamonte M Jr, Detrano R: Quantitation of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol 27:827–832, 1996

    Steffes M, Cleary P, Goldstein D, Little R, Wiedmeyer HM, Rohlfing C, England J, Bucksa J, Nowicki M: Hemoglobin A1c measurements over nearly two decades: sustaining comparable values throughout the Diabetes Control and Complications Trial and the Epidemiology of Diabetes Interventions and Complications study. Clin Chem 51:753–578, 2005

    O’Leary DH, Polak JF, Kronmal RA, Savage PJ, Borhani NO, Kittner SJ, Tracy R, Gardin JM, Price TR, Furberg CD: Thickening of the carotid wall: a marker for atherosclerosis in the elderly Cardiovascular Health Study Collaborative Research Group. Stroke 27:224–231, 1996

    Baclund JY, Determan A, Cleary P, for the DCCT/EDIC Research Group: Validation of deaths and reported cardiovascular and cerebrovascular events. Con Clin Trials 23:2S, 2003

    SAS/STAT User’s Guide: Version 8 Edition. Cary, NC, SAS Inst., 1999

    Davis PH, Dawson JD, Mahoney LT, Lauer RM: Increased carotid intimal-medial thickness and coronary calcification are related in young and middle-aged adults. Circulation 100:838–842, 1999

    Stokes ME, Davis CS, Kock GG: Categorical Data Analysis Using the SAS System. Cary, NC, SAS Institute, 1999

    Lachin JM: Biostatistical Methods. The Assessment of Relative Risks. New York, John Wiley & Sons, 2000

    Tobin J: Estimation of relationships for limited dependent variables. Econometrics 26:24–36, 1958

    Breen R: Regression Models. Censored, Sample Selected or Truncated Data. Thousand Oaks, CA, Sage Publications, 1996

    Hosmer DW, Lemeshow S: Applied Logistic Regression. New York, John Wiley & Sons, 1989

    Stead AG, MacDonald KG: Constructing ROC Curves with SAS System. Durham, NC, Organon Tekmika.

    Altman DG, Bland JM: Diagnostic tests. I. Sensitivity and specificity. BMJ 308:1552, 1994

    Zweig MH, Campbell G: Receiver-operating characteristics (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin Chem 39:561–577, 1993

    The DCCT Research Group: The relationship of glycemic exposure (HbA1c) to the risk of development and progression of retinopathy in the Diabetes Control and Complications Trial. Diabetes 44:968–983, 1995

    Dabelea D, Kinney G, Snell-Bergeon JK, Hokanson JE, Eckel RH, Ehrlich J, Garg S, Hamman R, Rewers M: Effect of type 1 diabetes on the gender difference in coronary artery calcification: a role for insulin resistance The Coronary Artery Calcification in Type 1 Diabetes (CACTI) study. Diabetes 52:2833–2839, 2003 [erratum Diabetes 53:2177, 2004]

    Steffes M, Molitch M, Chavers B, Cleary P, Lachin J, Genuth S, Nathan D: Sustained effect of intensive treatment of type 1 diabetes mellitus on development and progression of diabetic nephropathy: the Epidemiology of Diabetes Interventions and Complications (EDIC) study. JAMA 290:2159–2167, 2003

    Vlassara H, Palace MR: Diabetes and advanced glycation endproducts. J Int Med 251:87–101, 2002

    Monnier VM, Bautista O, Kenny D, Sell DR, Fogarty J, Dahms W, Cleary P, Lachin J, Genuth S, the DCCT Skin Collagen Ancillary Study Group: Skin collagen glycation, glycoxidation, and crosslinking are lower in subjects with long-term intensive versus conventional therapy of type 1 diabetes: relevance of glycated collagen products versus HbA1c as markers of diabetic complications. Diabetes 48:870–880, 1999

    Yan SF, Ramasamy R, Naka Y, Schmidt AM: Glycation, inflammation, and RAGE: a scaffold for the macrovascular complications of diabetes and beyond. Circ Res 93:1159–1169, 2003

    Lachin JM, Genuth S, Cleary P, Davis MD, Nathan DM: Retinopathy and nephropathy in patients with type 1 diabetes four years after a trial of intensive therapy. N Engl J Med 342:381–389, 2000

    Becker CR, Kleffel T, Crispin A, Knez A, Young J, Schoepf UJ, Haberl R, Reiser MF: Coronary artery calcium measurement: agreement of multirow detector and electron beam CT. AJR Am J Roentgenol 176:1295–1298, 2001

    Simons DB, Schwartz RS, Edwards WD, Sheedy PF, Breen JF, Rumberger JA: Noninvasive definition of anatomic coronary artery disease by ultrafast computed tomographic scanning: a quantitative pathologic comparison study. J Am Coll Cardiol 20:1118–1126, 1992

    Sangiorgi G, Rumberger JA, Severson A, Edwards WD, Gregorire J, Fitzpatrick L, Schwartz RS: Arterial calcification and not lumen stenosis is highly correlated with atherosclerotic plaque burden in humans: a histologic study of 723 coronary artery segments using nondecalcifying methodology. J Am Coll Cardiol 31:126–133, 1998

    Detrano R, Hsiai T, Wang S, Puentes G, Fallavollita J, Shields P, Stanford W, Wolfkiel C, Georgiou D, Budoff M, Reed J: Prognostic value of coronary calcification and angiographic stenoses in patients undergoing coronary angiography. J Am Coll Cardiol 27:285–290, 1996

    Fiorino AS: Electron-beam computed tomography, coronary artery calcium and evaluation of patients with coronary artery disease. Ann Int Med 128:839–847, 1998

    Wexler L, Brundage B, Crouse J, Detrano R, Fuster V, Maddahi Jamshid, Rumberger J, Stanford W, White R, Taubert K, AHA Staff: Coronary artery calcification: pathophysiology, epidemiology, imaging methods, and clinical implications. Circulation 94:1175–1192, 1996

    Arad Y, Spadaro LA, Goodman K, Newstein D, Guerci AD: Prediction of coronary events with electron beam computed tomography. J Am Coll Cardiol 36:1253–1260, 2000

    Clouse ME: How useful is computed tomography for screening for coronary artery disease Circulation 113:125–135, 2006

    Chen J, Krumholz JM: Screening for coronary artery disease with electron-beam computed tomography is not useful. Circulation 113:135–146, 2006(Patricia A. Cleary, Trevor J. Orchard, S)