Effect of Intensive Glycemic Control on Levels of Markers of Inflammation in Type 1 Diabetes Mellitus in the Diabetes Control and Complicati
http://www.100md.com
《循环学杂志》
the Division of Preventive Medicine (D.A.S., R.J.G., J.E.M., P.M.R.) and Center for Cardiovascular Disease Prevention (P.M.R.), Brigham and Women’s Hospital, the Department of Ophthalmology (D.A.S.), Schepens Eye Research Institute, the Department of Laboratory Medicine (N.R.), Children’s Hospital, and the Diabetes Center (D.M.N.), Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, Mass
Departments of Epidemiology (D.A.S., J.E.M.) and Biostatistics (R.J.G.), Harvard School of Public Health, Boston, Mass
Division of Endocrinology (A.J.J., T.J.L.), Medical University of South Carolina
Section of Endocrinology (A.J.J., T.J.L.), University of Oklahoma Health Sciences Center.
Abstract
Background— Type 1 diabetes mellitus is associated with an increased risk of cardiovascular disease (CVD) that is not fully explained by conventional risk factors. The Diabetes Control and Complications Trial (DCCT) showed that intensive diabetes therapy reduced levels of LDL cholesterol and triglycerides but increased the risk of major weight gain, which might adversely affect CVD risk. The present study examined the effect of intensive therapy on levels of several markers of inflammation that have been linked to risk of CVD.
Methods and Results— We measured levels of inflammatory biomarkers in stored baseline and 3-year follow-up serum specimens from a random sample of 385 participants in the DCCT, a multicenter trial in which 1441 subjects aged 13 to 39 years with type 1 diabetes mellitus were randomized to intensive or conventional diabetes treatment. The markers included high-sensitivity C-reactive protein (hsCRP), soluble intercellular adhesion molecule type 1 (sICAM-1), soluble vascular cell adhesion molecule type 1 (sVCAM-1), and the 55-kDa soluble tumor necrosis factor- receptor 1 (sTNF-R1). We examined the effect of intensive therapy on the change in levels of the inflammatory markers. In unadjusted analyses, levels of hsCRP and sTNF-R1 increased in both treatment groups after 3 years of follow-up, with no significant difference between groups for hsCRP (P=0.53) but with a greater increase of sTNF-R1 in the intensive therapy group (P=0.002). In contrast, mean levels of sICAM-1 and sVCAM-1 decreased among participants assigned to intensive therapy, whereas they did not change among those in the conventional treatment group (P=0.03 for sICAM-1; P=0.03 for sVCAM-1). After adjustment for baseline levels and other factors, intensive therapy remained associated with a significant decrease in sICAM-1 (P=0.02) and an increase in sTNF-R1 (P=0.03). For hsCRP, there was a significant interaction between the top third of weight gain and treatment assignment (P=0.03). In subgroup analyses among subjects undergoing intensive therapy, hsCRP levels increased among those who gained the most weight, whereas it decreased among those in the bottom third of weight gain (P=0.0004).
Conclusions— Intensive therapy in patients with type 1 diabetes mellitus reduced levels of sICAM-1 and increased levels of sTNF-R1 and of hsCRP among those who gained weight. These data demonstrate that the effect of intensive therapy on inflammation is complex and, to the extent that hsCRP is a risk factor, suggest that the risk of atherosclerosis among diabetic patients may be influenced by the degree of weight gain while undergoing intensive therapy.
Key Words: inflammation ; epidemiology ; cardiovascular diseases ; diabetes mellitus ; obesity
Introduction
Cardiovascular disease (CVD) is a major cause of morbidity and the leading cause of mortality among people with type 1 diabetes mellitus. People with diabetes mellitus have a higher coincidence of a number of traditional CVD risk factors, including hyperlipidemia and hypertension.1–3 Moreover, chronic hyperglycemia is a risk factor for atherosclerosis in type 1 diabetes mellitus.4 However, neither
See p 2414
traditional CVD risk factors nor hyperglycemia fully explains the risk for CVD. An additional factor that may play a role in the increased risk of CVD among diabetics is an increase in the overall level of systemic inflammatory activity.
Compelling evidence suggests that CVD is at least in part an inflammatory process.5 Moreover, prior research has shown increased levels of several markers of inflammation among people with diabetes,6 and these levels generally correlate with measures of glycemic control, such as glycosylated hemoglobin A1c (HbA1c). For example, HbA1c significantly correlates with levels of C-reactive protein (CRP),7,8 an acute-phase reactant that is associated with early atherosclerotic lesions9 and predicts future risk of CVD,10–12 including among subjects with type 1 diabetes mellitus.7 CRP also predicts onset of type 2 diabetes mellitus among those with low HbA1c13 and adds prognostic information on vascular risk among those with metabolic syndrome.14 Other markers of inflammation, including soluble intercellular adhesion molecule 1 (sICAM-1) and soluble tumor necrosis factor alpha- receptors (sTNF-R), predict risk of CVD in nondiabetic persons15,16 and are increased in type 2 diabetes mellitus.17,18
In the Diabetes Control and Complications Trial (DCCT), intensive diabetes control was associated with a clear and sustained reduction in levels of HbA1c and of microvascular complications in type 1 diabetes mellitus; however, there was an increased risk of weight gain and obesity, accompanied by other cardiovascular risk factors, among those assigned to intensive diabetes therapy.19 Recently, higher body mass index has been linked with increased levels of inflammatory markers, particularly high-sensitivity CRP (hsCRP).7,20,21 In both the DCCT22 and a large trial among subjects with type 2 diabetes mellitus,23 it was not proved conclusively whether intensive diabetes therapy reduces the risk of CVD. Nevertheless, long-term follow-up of the DCCT cohort has revealed a beneficial effect of previous intensive therapy on atherosclerosis as measured by carotid intima-media thickness.4 The present study was undertaken to examine the effect of intensive metabolic control on circulating levels of inflammatory molecules, including hsCRP, sICAM-1, soluble vascular cell adhesion molecule-1 (sVCAM-1), and the 55-kDa sTNF-R1, putative early markers of cardiovascular risk. The potential contribution of weight gain to the inflammatory molecules was also explored.
Methods
The DCCT was a multicenter trial of intensive versus conventional treatment of type 1 diabetes mellitus, designed to examine development and progression of microvascular complications.24 Subjects were 1441 individuals with type 1 diabetes mellitus between the ages of 13 and 39 years, recruited at 29 centers from 1983 through 1989. All DCCT participants had baseline HbA1c levels >3 SDs above the mean for nondiabetics. Subjects were recruited into 2 strata: (1) a primary prevention cohort with no retinopathy, diabetes duration of 1 to 5 years, and no evidence of microalbuminuria at baseline (albuminuria <28 μg/min [<40 mg/24 h]) or (2) a secondary intervention cohort with minimal to moderate nonproliferative retinopathy, diabetes duration between 1 and 15 years, and albuminuria 140 μg/min (200 mg/24 h). Subjects with baseline total cholesterol levels >3 SDs above the mean for sex and age, calculated LDL cholesterol levels >4.9 mmol/L (190 mg/dL), body weight >30% above ideal, major ECG abnormalities, a history of coronary heart disease, or symptoms of peripheral vascular disease were excluded. For the present study, we randomly selected a subgroup of 400 DCCT participants stratified by baseline retinopathy subgroup and randomized treatment assignment. This sample size was based on a calculation to achieve power of 85% to detect a difference of 0.15 SDs of the change in biomarker levels between treatment groups with a type 1 error rate of 0.05. None of the participants in the present study had hypertension or were taking hormone replacement therapy.
Intensive insulin therapy consisted of the administration of insulin 3 or more times per day by injection or external insulin pump therapy, guided by results of self-monitoring of blood glucose performed at least 4 times daily. Conventional therapy was with 1 to 2 daily injections of insulin, with daily self-monitoring of urine or blood glucose. Glycosylated hemoglobin levels fell rapidly in the intensive therapy group, reaching a nadir within 6 months, and average levels remained approximately 2 percentage points lower than in the conventional therapy group for the duration of the trial. After a mean follow-up of 6.5 years (range 3 to 9 years), the DCCT was stopped, having shown clinically and statistically highly significant reductions (range of 35% to >70%) in retinopathy, nephropathy, and neuropathy with intensive compared with conventional therapy.24
DCCT Blood and Urine Collection and Storage
Fasting plasma and serum samples were obtained from DCCT participants at baseline and each annual visit. For stored plasma samples, blood was drawn into EDTA tubes with 500 U of Trasylol (aprotinin) per milliliter of whole blood, immediately placed on ice, and then centrifuged at 2000 rpm for 10 minutes. The plasma was then transferred to 1.8-mL cryotubes and immediately placed on dry ice. For stored serum samples, blood was drawn into a red-topped tube, allowed to clot for at least 20 minutes, and then spun in a centrifuge at room temperature for 10 minutes at 3000 rpm. Serum was then divided into 1.8-mL cryotubes and promptly frozen. Samples were maintained at –70°C25 at the DCCT Central Biochemistry Laboratory, Department of Laboratory Medicine and Pathology, University of Minnesota until preparation for the present study.
Laboratory Analysis
Baseline and 3-year serum samples were thawed and assayed for hsCRP with a latex-enhanced immunonephelometric assay on a BN II analyzer (Dade Behring). Serum levels of sICAM-1, sVCAM-1, and sTNF-R1 were measured by ELISAs (R&D Systems). All samples were handled in an identical and masked fashion, with baseline and 3-year specimens interspersed for analysis.
Statistical Analysis
We examined the change in levels of each biomarker from baseline to the 3-year follow-up visit and tested whether this was related to the randomization assignment using the Mann-Whitney U test. These relationships were also estimated with linear regression, with the natural logarithm of the 3-year follow-up levels of the biomarker as the dependent variable and treatment assignment, the natural log of baseline levels of the biomarker, and other factors as the independent variables. To enhance comparability of results among markers, we retained a common set of covariates for each marker of inflammation. These covariates included baseline levels of the marker, baseline HbA1c, age, sex, duration of diabetes, body mass index, and smoking status. For continuous variables, we subtracted the mean value from the measured value to obtain a distribution centered on the mean, to aid in interpretation of the models (particularly the intercepts). For each marker, we then fit an additional model in which we adjusted for the average HbA1c level over follow-up, to examine the relationships with glucose control per se. In further models, we considered interactions between baseline levels of the markers and randomized treatment assignment by including product terms in the models. Finally, for both hsCRP and sTNF-R1, molecules strongly correlated with body weight, we explored whether there was a relationship with weight gain or an interaction between weight gain and treatment assignment. For these analyses, we divided weight gain into approximate thirds based on the distribution of weight gain in the total DCCT population. We included terms for thirds of weight gain and product terms that represented their interaction with treatment assignment in the regression models.
We performed regression diagnostics to identify unduly influential observations and repeated models that excluded these observations. Because exclusion of the most influential observations did not change interpretation of any findings, we present only the models for all participants. In a secondary analysis, we used the baseline to 3-year change in levels of the biomarkers as the independent variable for our models. Because the distributions of the changes in inflammatory markers were skewed, and log transformations were not possible because of negative values (because a participant’s marker level could have either decreased or increased), we obtained the normalized rank transformations of each of the markers and regressed covariates on these ranks to obtain probability values for the regression coefficients. There were no substantial differences in this analysis compared with the findings presented.
Results
Of the 400 DCCT participants randomly selected for the present study, 385 had sufficient baseline and 3-year follow-up serum samples to measure at least 1 of the selected biomarkers; thus, any potential effect of dropouts on results is minimal and not of concern. In this subset of DCCT participants, there were no significant differences in levels of the inflammatory markers hsCRP, sICAM-1, sVCAM-1, or sTNF-R1 at baseline between subjects assigned to intensive versus conventional therapy (Table 1).
Median levels of hsCRP increased over time by 0.19 mg/L (23.9%) in the conventional treatment group and by 0.08 mg/L (9.3%) among those assigned to intensive therapy (Table 2). This difference was not statistically significant (P=0.53). In contrast, median levels of both sICAM-1 and sVCAM-1 increased from baseline to 3 years of follow-up by 4.9 ng/mL (1.7%) and 4.6 ng/mL (1.0%), respectively, among subjects assigned to conventional treatment but decreased by 2.5 mg/L (0.9%) and 9.5 mg/L (2.1%), respectively, among subjects assigned to intensive therapy. The differences in change in levels for intensive versus conventional therapy were statistically significant (each P=0.03). Levels of sTNF-R1 increased among subjects in both treatment arms of the DCCT but by a greater amount among subjects assigned to intensive therapy (6.0 pg/mL [0.6%] for conventional treatment versus 108.1 pg/mL [10.9%] for intensive therapy; P=0.002). Because baseline levels had a substantial impact on changes over time, as would be expected as more extreme levels regressed toward the mean, we reexamined associations with treatment assignment in models that adjusted for baseline levels of the markers, as well as for other potential predictors of change.
In analyses of hsCRP, after adjustment for baseline levels of hsCRP (P<0.0001), female sex (P<0.0001), and other factors, 3-year follow-up hsCRP levels were not significantly different among those assigned to intensive versus conventional control (P=0.98; Table 3). In models for sICAM-1, both baseline sICAM-1 levels (P<0.0001) and current cigarette smoking at baseline (P<0.0001) were significant predictors of 3-year follow-up levels. After adjustment for these and other variables, intensive therapy was still associated with a significant decrease in sICAM-1 levels after 3 years of follow-up (P=0.02; Table 3). For sVCAM-1, after adjustment for baseline levels (P<0.0001), age (P=0.0005), and other factors, the association of intensive therapy with change in sVCAM-1 was no longer significant (P=0.19; Table 3). Lower baseline levels of sTNF-R1 (P<0.0001) and older age (P=0.01) were associated with an increase in sTNF-R1 at 3 years. After adjustment for these and other covariates, intensive diabetes control was associated with a significantly greater increase in sTNF-R1 levels than conventional therapy (P=0.03; Table 3).
In further analyses, we included the average level of HbA1c over follow-up in models to investigate the extent to which any observed effect of intensive control might be mediated through the degree of glucose control achieved. For hsCRP, adjustment for average levels of HbA1c revealed a significant increase in hsCRP among those assigned to intensive control (P=0.04). In this model, the level of hsCRP after 3 years of follow-up was inversely associated with baseline HbA1c levels (P=0.003), but there was a direct association with average HbA1c during follow-up (P=0.005; Table 4). In contrast, for sICAM-1, sVCAM-1, and sTNF-R1, adjustment for average HbA1c attenuated the effect of intensive control, which was not significant for any of these markers (Table 4). Additional adjustment for changes in cigarette smoking during follow-up did not change any of these findings (data not shown).
Finally, given the known tendency for subjects to gain weight after initiation of intensive therapy, the strong relationship between body mass index and hsCRP levels, and the observation that intensive control increased hsCRP levels after adjustment for average HbA1c levels, we examined weight change and whether there might be an interaction between weight change and randomized treatment assignment that might provide some explanation for the lack of an overall effect of intensive therapy on change in levels of hsCRP. In this analysis, we identified a significant interaction between weight change and treatment assignment (P=0.03; Table 5). In separate analyses in the 2 treatment groups, among subjects assigned to intensive therapy, there was a median increase of 0.54 mg/L (75%) in hsCRP levels at 3 years of follow-up among those in the top third of weight change (mean body mass index at 3 years of follow-up 27.0 kg/m2) compared with a decrease of 0.24 mg/L (20%) among subjects in the lowest third of weight change (mean body mass index at 3 years of follow-up 22.8 kg/m2; P=0.0004; Figure). In contrast, among those assigned to conventional control, there was no significant difference in the change in hsCRP according to weight change category (P=0.88). We also investigated potential effects of weight gain with regard to sTNF-R1 and found a significantly greater increase in sTNF-R1 levels among subjects who gained the most weight (P=0.007; Table 5). In contrast to hsCRP, for sTNFR-1, there was no significant interaction of weight gain with intensive therapy (data not shown).
Interaction of weight gain and intensive diabetes control on change in levels of hsCRP in DCCT. Median change in hsCRP (mg/L) from baseline to 3 years, by randomization assignment and third of weight gain over same period. White bars are for conventional treatment group, gray bars for those assigned to intensive diabetes therapy. Probability values (Wilcoxon rank sum test) for change in hsCRP were 0.88 and 0.0004, respectively, for subgroups undergoing conventional and intensive therapy.
Discussion
Randomized clinical trials have not conclusively shown that intensive glycemic control reduces the risk of CVD among people with either type 126 or type 223 diabetes mellitus. Some epidemiological data suggest a stronger relationship with insulin resistance-related factors than with HbA1c.27 In the DCCT, intensive therapy was associated with a higher risk of overweight and obesity,28 which may adversely affect CVD risk. Several lines of evidence have come to light that suggest that inflammation is a key player in the development of CVD.5 A number of studies have found that several circulating molecules involved in inflammation, such as sICAM-1, hsCRP, and TNF-R1, predict risk of CVD, and some of these molecules, notably hsCRP and sTNF-R1, are highly influenced by body weight. The present data from the DCCT indicate that among subjects with type 1 diabetes mellitus, intensive glycemic control is associated with a significant reduction in levels of sICAM-1 but not with change in levels of hsCRP, the marker of inflammation most strongly and consistently predictive of CVD in epidemiological studies. Furthermore, the present analyses indicated that there was a significant rise in levels of hsCRP among intensively treated subjects who gained the most weight. There was also a significant rise in sTNF-R1 among intensively treated subjects, which appears to be explained by greater weight gain in the intensively treated group.
Although this is the largest study to date of these relationships among individuals with type 1 diabetes mellitus, several limitations deserve consideration. Given the uniqueness of the DCCT population, generalizability of these findings to individuals with type 2 diabetes mellitus and to ethnic minority groups is uncertain. The present analyses of weight change as a potential predictor of change may be controversial, because they include in the model a factor (weight gain) that is directly influenced by treatment assignment and is thus an intermediate variable. Multiple testing may have also resulted in an inflation of the type 1 error rate. However, we think that our analyses are informative because they provide a potentially relevant explanation for the overall lack of an effect of intensive therapy on hsCRP levels, in contrast to a significant increase after adjustment for average HbA1c during follow-up, and for the greater increase in sTNF-R1 among subjects undergoing intensive therapy. We were also limited to a single measure of the inflammatory markers at each time point, which may not accurately reflect the long-term or average levels of exposure that are likely to be most relevant to CVD. However, single measures have shown significant predictive value for CVD in prior prospective studies.10–12,15,16 Long-term follow-up of the DCCT cohort in the Epidemiology of Diabetes Interventions and Complications (EDIC) study presents the opportunity for future studies of the persistence of the observed changes in inflammatory biomarkers and their potential associations with CVD risk.
hsCRP has been strongly related to risk of CVD in a number of prospective studies and with early carotid atherosclerosis in young subjects with type 1 diabetes mellitus.9,29 Most studies have shown a positive correlation between hsCRP and HbA1c.6,30 In the present study, however, we found no overall difference in the change in levels of hsCRP comparing intensive and conventional therapy. There are few other data available, and none that we know of for type 1 diabetes mellitus or with comparable follow-up. A significant reduction in CRP was noted with an intensive insulin regimen in critically ill subjects.31 Moreover, CRP fell after 16 weeks of insulin treatment of 22 subjects with poorly controlled type 2 diabetes mellitus.32 In the present study of type 1 diabetes mellitus, we measured change in hsCRP after 3 years of intensive or conventional treatment and found no overall change in hsCRP. Interestingly, however, any potential benefit of improved glycemic control may have been limited by the weight gain in subjects assigned to intensive therapy.
By the end of the DCCT, the prevalence of obesity reached 33% among subjects assigned to intensive therapy versus 19% in the conventionally treated group.28 Subjects who gained the most weight with intensive therapy tended to experience metabolic changes similar to those seen in the insulin resistance syndrome.19 The present study adds to these findings by demonstrating an increase in hsCRP among intensively treated subjects who gained the most weight. This finding is consistent with the known direct relationship between body weight and hsCRP33 and with a large number of studies that show a relationship of CRP with the development of insulin resistance.13,34–38 Particularly in light of evidence that suggests a stronger relationship of insulin resistance-related factors than glycemia with CVD,27 if the increase in hsCRP levels associated with weight gain persists over time, this in conjunction with the other adverse metabolic changes that occur with excessive weight gain during intensive therapy may result in an increased risk of CVD.
Raised circulating levels of adhesion molecules are thought to indicate a state of endothelial activation with consequent induction of immunological activity.39,40 Levels of sICAM-115 but not sVCAM-141 predict future cardiovascular events. Although adhesion molecules have also been implicated in the pathogenesis of type 1 diabetes mellitus,42,43 there is little information about their relationship with metabolic control or their potential role in the pathogenesis of complications. Subjects with diabetes mellitus appear to have elevated levels of sICAM-1 and sVCAM-1.44 In a small study, sICAM-1 was significantly higher among diabetic subjects who later developed CVD than among those who did not.45 In the present study, levels of sICAM-1 fell significantly after the introduction of intensive therapy. Our analyses suggested that this decrease was mediated through reduction of glucose. Given the previously described association of sICAM-1 with future CVD, the fall in sICAM-1 with intensive therapy might exert a beneficial influence on future risk of CVD. However, this finding must be interpreted cautiously in light of the inconsistent effects of intensive therapy on other predictive markers, in particular hsCRP.
The presence of advanced glycosylation end products and glycated albumin in the diabetic milieu can enhance expression of tumor necrosis factor (TNF)-.46,47 However, TNF- has a short half-life, and high levels of its receptors can interfere with its detection,48 which limits its usefulness as a biomarker. Although less widely studied than other markers of inflammation, levels of TNF receptors are elevated in a variety of settings.49,50 In a cross-sectional study, higher levels of TNF receptors were associated with carotid atherosclerosis in subjects less than 70 years of age.16 In the present study, we observed a significantly greater increase in sTNF-R1 among subjects assigned to intensive therapy; however, after adjustment for weight gain during follow-up, the effect of intensive therapy was not significant. Rather, mean levels of sTNF-R1 increased to a greater extent among subjects who gained weight, regardless of type of therapy.
In summary, we examined the effect of intensive versus conventional therapy on 4 selected markers of inflammation in the DCCT. The effect of intensive control on levels of inflammatory markers was not uniform, which suggests a complex relationship between intensive diabetes therapy and inflammation. Levels of sICAM-1 were reduced, but there was an increase in sTNF-R1 among intensively treated subjects and of hsCRP among intensively treated subjects who gained the most weight. To the extent that hsCRP is a risk factor, the risk of atherosclerosis among subjects with type 1 diabetes mellitus may be influenced by whether or not they gain weight while undergoing intensive therapy.
Acknowledgments
This study was supported by Juvenile Diabetes Research Foundation grant No. 1-2000-646 (Dr Schaumberg), the Earle P. Charlton, Jr. Charitable Foundation (Dr Nathan), and a contract with the Division of Diabetes, Endocrinology, and Metabolic Diseases of the National Institute of Diabetes and Digestive and Kidney Diseases.
References
Laing SP, Swerdlow AJ, Slater SD, Burden AC, Morris A, Waugh NR, Gatling W, Bingley PJ, Patterson CC. Mortality from heart disease in a cohort of 23,000 patients with insulin-treated diabetes. Diabetologia. 2003; 46: 760–765.
Green A, Hougaard P. Epidemiological studies of diabetes mellitus in Denmark: 5: mortality and causes of death among insulin-treated diabetic patients. Diabetologia. 1984; 26: 190–194.
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. 1984; 33: 271–276.
Nathan DM, Lachin J, Cleary P, Orchard T, Brillon DJ, Backlund JY, O’Leary DH, Genuth S. Intensive diabetes therapy and carotid intima-media thickness in type 1 diabetes mellitus. N Engl J Med. 2003; 348: 2294–2303.
Libby P, Ridker PM, Maseri A. Inflammation and atherosclerosis. Circulation. 2002; 105: 1135–1143.
Gomes MB, Piccirillo LJ, Nogueira VG, Matos HJ. Acute-phase proteins among patients with type 1 diabetes. Diabetes Metab. 2003; 29: 405–411.
Kilpatrick ES, Keevil BG, Jagger C, Spooner RJ, Small M. Determinants of raised C-reactive protein concentration in type 1 diabetes. QJM. 2000; 93: 231–236.
King DE, Mainous AG III, Buchanan TA, Pearson WS. C-reactive protein and glycemic control in adults with diabetes. Diabetes Care. 2003; 26: 1535–1539.
Hayaishi-Okano R, Yamasaki Y, Katakami N, Ohtoshi K, Gorogawa S, Kuroda A, Matsuhisa M, Kosugi K, Nishikawa N, Kajimoto Y, Hori M. Elevated C-reactive protein associates with early-stage carotid atherosclerosis in young subjects with type 1 diabetes. Diabetes Care. 2002; 25: 1432–1438.
Ridker PM, Rifai N, Rose L, Buring JE, Cook NR. Comparison of C-reactive protein and low-density lipoprotein cholesterol levels in the prediction of first cardiovascular events. N Engl J Med. 2002; 347: 1557–1565.
Ridker PM, Hennekens CH, Buring JE, Rifai N. C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med. 2000; 342: 836–843.
Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men [published erratum appears in N Engl J Med. 1997;337:356]. N Engl J Med. 1997; 336: 973–979.
Pradhan AD, Manson JE, Rifai N, Buring JE, Ridker PM. C-reactive protein, interleukin 6, and risk of developing type 2 diabetes mellitus. JAMA. 2001; 286: 327–334.
Ridker PM, Buring JE, Cook NR, Rifai N. C-reactive protein, the metabolic syndrome, and risk of incident cardiovascular events: an 8-year follow-up of 14 719 initially healthy American women. Circulation. 2003; 107: 391–397.
Ridker PM, Hennekens CH, Roitman-Johnson B, Stampfer MJ, Allen J. Plasma concentration of soluble intercellular adhesion molecule 1 and risks of future myocardial infarction in apparently healthy men. Lancet. 1998; 351: 88–92.
Elkind MS, Cheng J, Boden-Albala B, Rundek T, Thomas J, Chen H, Rabbani LE, Sacco RL. Tumor necrosis factor receptor levels are associated with carotid atherosclerosis. Stroke. 2002; 33: 31–37.
Meigs JB, Hu FB, Rifai N, Manson JE. Biomarkers of endothelial dysfunction and risk of type 2 diabetes mellitus. JAMA. 2004; 291: 1978–1986.
Hu FB, Meigs JB, Li TY, Rifai N, Manson JE. Inflammatory markers and risk of developing type 2 diabetes in women. Diabetes. 2004; 53: 693–700.
Purnell JQ, Hokanson JE, Marcovina SM, Steffes MW, Cleary PA, Brunzell JD. Effect of excessive weight gain with intensive therapy of type 1 diabetes on lipid levels and blood pressure: results from the DCCT: Diabetes Control and Complications Trial [published erratum appears in JAMA. 1998;280:1484]. JAMA. 1998; 280: 140–146.
Bermudez EA, Rifai N, Buring J, Manson JE, Ridker PM. Interrelationships among circulating interleukin-6, C-reactive protein, and traditional cardiovascular risk factors in women. Arterioscler Thromb Vasc Biol. 2002; 22: 1668–1673.
Vikram NK, Misra A, Dwivedi M, Sharma R, Pandey RM, Luthra K, Chatterjee A, Dhingra V, Jailkhani BL, Talwar KK, Guleria R. Correlations of C-reactive protein levels with anthropometric profile, percentage of body fat and lipids in healthy adolescents and young adults in urban North India. Atherosclerosis. 2003; 168: 305–313.
Effect of intensive diabetes management on macrovascular events and risk factors in the Diabetes Control and Complications Trial. Am J Cardiol. 1995; 75: 894–903.
UK Prospective Diabetes Study (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.
The 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. 1993; 329: 977–986.
The Diabetes Control and Complications Trial Research Group. DCCT Manual of Operations. Springfield, Va: Department of Commerce, National Technical Information Service; 1993.
The Diabetes Control and Complications Trial Research Group. Effect of intensive diabetes management on macrovascular events and risk factors in the Diabetes Control and Complications Trial. Am J Cardiol. 1995; 75: 894–903.
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. 2003; 26: 1374–1379.
Influence of intensive diabetes treatment on body weight and composition of adults with type 1 diabetes in the Diabetes Control and Complications Trial. Diabetes Care. 2001; 24: 1711–1721.
Colhoun HM, Schalkwijk C, Rubens MB, Stehouwer CD. C-reactive protein in type 1 diabetes and its relationship to coronary artery calcification. Diabetes Care. 2002; 25: 1813–1817.
Hansen TK, Thiel S, Knudsen ST, Gravholt CH, Christiansen JS, Mogensen CE, Poulsen PL. Elevated levels of mannan-binding lectin in patients with type 1 diabetes. J Clin Endocrinol Metab. 2003; 88: 4857–4861.
Hansen TK, Thiel S, Wouters PJ, Christiansen JS, Van den Berghe G. Intensive insulin therapy exerts antiinflammatory effects in critically ill patients and counteracts the adverse effect of low mannose-binding lectin levels. J Clin Endocrinol Metab. 2003; 88: 1082–1088.
Yudkin JS, Panahloo A, Stehouwer C, Emeis JJ, Bulmer K, Mohamed-Ali V, Denver AE. The influence of improved glycaemic control with insulin and sulphonylureas on acute phase and endothelial markers in type II diabetic subjects. Diabetologia. 2000; 43: 1099–1106.
Ford ES. Body mass index, diabetes, and C-reactive protein among U.S. adults. Diabetes Care. 1999; 22: 1971–1977.
Barzilay JI, Abraham L, Heckbert SR, Cushman M, Kuller LH, Resnick HE, Tracy RP. The relation of markers of inflammation to the development of glucose disorders in the elderly: the Cardiovascular Health Study. Diabetes. 2001; 50: 2384–2389.
Festa A, D’Agostino R Jr, Tracy RP, Haffner SM. Elevated levels of acute-phase proteins and plasminogen activator inhibitor-1 predict the development of type 2 diabetes: the Insulin Resistance Atherosclerosis Study. Diabetes. 2002; 51: 1131–1137.
Festa A, D’Agostino R Jr, Howard G, Mykkanen L, Tracy RP, Haffner SM. Chronic subclinical inflammation as part of the insulin resistance syndrome: the Insulin Resistance Atherosclerosis Study (IRAS). Circulation. 2000; 102: 42–47.
Schmidt MI, Duncan BB, Sharrett AR, Lindberg G, Savage PJ, Offenbacher S, Azambuja MI, Tracy RP, Heiss G. Markers of inflammation and prediction of diabetes mellitus in adults (Atherosclerosis Risk in Communities study): a cohort study. Lancet. 1999; 353: 1649–1652.
Wu T, Dorn JP, Donahue RP, Sempos CT, Trevisan M. Associations of serum C-reactive protein with fasting insulin, glucose, and glycosylated hemoglobin: the Third National Health and Nutrition Examination Survey, 1988–1994. Am J Epidemiol. 2002; 155: 65–71.
Gearing AJ, Newman W. Circulating adhesion molecules in disease. Immunol Today. 1993; 14: 506–512.
Roep BO, Heidenthal E, de Vries RR, Kolb H, Martin S. Soluble forms of intercellular adhesion molecule-1 in insulin-dependent diabetes mellitus. Lancet. 1994; 343: 1590–1593.
de Lemos JA, Hennekens CH, Ridker PM. Plasma concentration of soluble vascular cell adhesion molecule-1 and subsequent cardiovascular risk. J Am Coll Cardiol. 2000; 36: 423–426.
Lampeter ER, Kishimoto TK, Rothlein R, Mainolfi EA, Bertrams J, Kolb H, Martin S. Elevated levels of circulating adhesion molecules in IDDM patients and in subjects at risk for IDDM. Diabetes. 1992; 41: 1668–1671.
Yang XD, Michie SA, Mebius RE, Tisch R, Weissman I, McDevitt HO. The role of cell adhesion molecules in the development of IDDM: implications for pathogenesis and therapy. Diabetes. 1996; 45: 705–710.
Fasching P, Veitl M, Rohac M, Streli C, Schneider B, Waldhausl W, Wagner OF. Elevated concentrations of circulating adhesion molecules and their association with microvascular complications in insulin-dependent diabetes mellitus. J Clin Endocrinol Metab. 1996; 81: 4313–4317.
Jude EB, Douglas JT, Anderson SG, Young MJ, Boulton AJ. Circulating cellular adhesion molecules ICAM-1, VCAM-1, P- and E-selectin in the prediction of cardiovascular disease in diabetes mellitus. Eur J Intern Med. 2002; 13: 185–189.
Neumann A, Schinzel R, Palm D, Riederer P, Munch G. High molecular weight hyaluronic acid inhibits advanced glycation endproduct-induced NF-kappaB activation and cytokine expression. FEBS Lett. 1999; 453: 283–287.
Abordo EA, Thornalley PJ. Synthesis and secretion of tumour necrosis factor-alpha by human monocytic THP-1 cells and chemotaxis induced by human serum albumin derivatives modified with methylglyoxal and glucose-derived advanced glycation endproducts. Immunol Lett. 1997; 58: 139–147.
Moreau E, Philippe J, Couvent S, Leroux-Roels G. Interference of soluble TNF-alpha receptors in immunological detection of tumor necrosis factor-alpha. Clin Chem. 1996; 42: 1450–1453.
Steiner G, Studnicka-Benke A, Witzmann G, Hofler E, Smolen J. Soluble receptors for tumor necrosis factor and interleukin-2 in serum and synovial fluid of patients with rheumatoid arthritis, reactive arthritis and osteoarthritis. J Rheumatol. 1995; 22: 406–412.
Studnicka-Benke A, Steiner G, Petera P, Smolen JS. Tumour necrosis factor alpha and its soluble receptors parallel clinical disease and autoimmune activity in systemic lupus erythematosus. Br J Rheumatol. 1996; 35: 1067–1074.(Effect of Intensive Glyce)
Departments of Epidemiology (D.A.S., J.E.M.) and Biostatistics (R.J.G.), Harvard School of Public Health, Boston, Mass
Division of Endocrinology (A.J.J., T.J.L.), Medical University of South Carolina
Section of Endocrinology (A.J.J., T.J.L.), University of Oklahoma Health Sciences Center.
Abstract
Background— Type 1 diabetes mellitus is associated with an increased risk of cardiovascular disease (CVD) that is not fully explained by conventional risk factors. The Diabetes Control and Complications Trial (DCCT) showed that intensive diabetes therapy reduced levels of LDL cholesterol and triglycerides but increased the risk of major weight gain, which might adversely affect CVD risk. The present study examined the effect of intensive therapy on levels of several markers of inflammation that have been linked to risk of CVD.
Methods and Results— We measured levels of inflammatory biomarkers in stored baseline and 3-year follow-up serum specimens from a random sample of 385 participants in the DCCT, a multicenter trial in which 1441 subjects aged 13 to 39 years with type 1 diabetes mellitus were randomized to intensive or conventional diabetes treatment. The markers included high-sensitivity C-reactive protein (hsCRP), soluble intercellular adhesion molecule type 1 (sICAM-1), soluble vascular cell adhesion molecule type 1 (sVCAM-1), and the 55-kDa soluble tumor necrosis factor- receptor 1 (sTNF-R1). We examined the effect of intensive therapy on the change in levels of the inflammatory markers. In unadjusted analyses, levels of hsCRP and sTNF-R1 increased in both treatment groups after 3 years of follow-up, with no significant difference between groups for hsCRP (P=0.53) but with a greater increase of sTNF-R1 in the intensive therapy group (P=0.002). In contrast, mean levels of sICAM-1 and sVCAM-1 decreased among participants assigned to intensive therapy, whereas they did not change among those in the conventional treatment group (P=0.03 for sICAM-1; P=0.03 for sVCAM-1). After adjustment for baseline levels and other factors, intensive therapy remained associated with a significant decrease in sICAM-1 (P=0.02) and an increase in sTNF-R1 (P=0.03). For hsCRP, there was a significant interaction between the top third of weight gain and treatment assignment (P=0.03). In subgroup analyses among subjects undergoing intensive therapy, hsCRP levels increased among those who gained the most weight, whereas it decreased among those in the bottom third of weight gain (P=0.0004).
Conclusions— Intensive therapy in patients with type 1 diabetes mellitus reduced levels of sICAM-1 and increased levels of sTNF-R1 and of hsCRP among those who gained weight. These data demonstrate that the effect of intensive therapy on inflammation is complex and, to the extent that hsCRP is a risk factor, suggest that the risk of atherosclerosis among diabetic patients may be influenced by the degree of weight gain while undergoing intensive therapy.
Key Words: inflammation ; epidemiology ; cardiovascular diseases ; diabetes mellitus ; obesity
Introduction
Cardiovascular disease (CVD) is a major cause of morbidity and the leading cause of mortality among people with type 1 diabetes mellitus. People with diabetes mellitus have a higher coincidence of a number of traditional CVD risk factors, including hyperlipidemia and hypertension.1–3 Moreover, chronic hyperglycemia is a risk factor for atherosclerosis in type 1 diabetes mellitus.4 However, neither
See p 2414
traditional CVD risk factors nor hyperglycemia fully explains the risk for CVD. An additional factor that may play a role in the increased risk of CVD among diabetics is an increase in the overall level of systemic inflammatory activity.
Compelling evidence suggests that CVD is at least in part an inflammatory process.5 Moreover, prior research has shown increased levels of several markers of inflammation among people with diabetes,6 and these levels generally correlate with measures of glycemic control, such as glycosylated hemoglobin A1c (HbA1c). For example, HbA1c significantly correlates with levels of C-reactive protein (CRP),7,8 an acute-phase reactant that is associated with early atherosclerotic lesions9 and predicts future risk of CVD,10–12 including among subjects with type 1 diabetes mellitus.7 CRP also predicts onset of type 2 diabetes mellitus among those with low HbA1c13 and adds prognostic information on vascular risk among those with metabolic syndrome.14 Other markers of inflammation, including soluble intercellular adhesion molecule 1 (sICAM-1) and soluble tumor necrosis factor alpha- receptors (sTNF-R), predict risk of CVD in nondiabetic persons15,16 and are increased in type 2 diabetes mellitus.17,18
In the Diabetes Control and Complications Trial (DCCT), intensive diabetes control was associated with a clear and sustained reduction in levels of HbA1c and of microvascular complications in type 1 diabetes mellitus; however, there was an increased risk of weight gain and obesity, accompanied by other cardiovascular risk factors, among those assigned to intensive diabetes therapy.19 Recently, higher body mass index has been linked with increased levels of inflammatory markers, particularly high-sensitivity CRP (hsCRP).7,20,21 In both the DCCT22 and a large trial among subjects with type 2 diabetes mellitus,23 it was not proved conclusively whether intensive diabetes therapy reduces the risk of CVD. Nevertheless, long-term follow-up of the DCCT cohort has revealed a beneficial effect of previous intensive therapy on atherosclerosis as measured by carotid intima-media thickness.4 The present study was undertaken to examine the effect of intensive metabolic control on circulating levels of inflammatory molecules, including hsCRP, sICAM-1, soluble vascular cell adhesion molecule-1 (sVCAM-1), and the 55-kDa sTNF-R1, putative early markers of cardiovascular risk. The potential contribution of weight gain to the inflammatory molecules was also explored.
Methods
The DCCT was a multicenter trial of intensive versus conventional treatment of type 1 diabetes mellitus, designed to examine development and progression of microvascular complications.24 Subjects were 1441 individuals with type 1 diabetes mellitus between the ages of 13 and 39 years, recruited at 29 centers from 1983 through 1989. All DCCT participants had baseline HbA1c levels >3 SDs above the mean for nondiabetics. Subjects were recruited into 2 strata: (1) a primary prevention cohort with no retinopathy, diabetes duration of 1 to 5 years, and no evidence of microalbuminuria at baseline (albuminuria <28 μg/min [<40 mg/24 h]) or (2) a secondary intervention cohort with minimal to moderate nonproliferative retinopathy, diabetes duration between 1 and 15 years, and albuminuria 140 μg/min (200 mg/24 h). Subjects with baseline total cholesterol levels >3 SDs above the mean for sex and age, calculated LDL cholesterol levels >4.9 mmol/L (190 mg/dL), body weight >30% above ideal, major ECG abnormalities, a history of coronary heart disease, or symptoms of peripheral vascular disease were excluded. For the present study, we randomly selected a subgroup of 400 DCCT participants stratified by baseline retinopathy subgroup and randomized treatment assignment. This sample size was based on a calculation to achieve power of 85% to detect a difference of 0.15 SDs of the change in biomarker levels between treatment groups with a type 1 error rate of 0.05. None of the participants in the present study had hypertension or were taking hormone replacement therapy.
Intensive insulin therapy consisted of the administration of insulin 3 or more times per day by injection or external insulin pump therapy, guided by results of self-monitoring of blood glucose performed at least 4 times daily. Conventional therapy was with 1 to 2 daily injections of insulin, with daily self-monitoring of urine or blood glucose. Glycosylated hemoglobin levels fell rapidly in the intensive therapy group, reaching a nadir within 6 months, and average levels remained approximately 2 percentage points lower than in the conventional therapy group for the duration of the trial. After a mean follow-up of 6.5 years (range 3 to 9 years), the DCCT was stopped, having shown clinically and statistically highly significant reductions (range of 35% to >70%) in retinopathy, nephropathy, and neuropathy with intensive compared with conventional therapy.24
DCCT Blood and Urine Collection and Storage
Fasting plasma and serum samples were obtained from DCCT participants at baseline and each annual visit. For stored plasma samples, blood was drawn into EDTA tubes with 500 U of Trasylol (aprotinin) per milliliter of whole blood, immediately placed on ice, and then centrifuged at 2000 rpm for 10 minutes. The plasma was then transferred to 1.8-mL cryotubes and immediately placed on dry ice. For stored serum samples, blood was drawn into a red-topped tube, allowed to clot for at least 20 minutes, and then spun in a centrifuge at room temperature for 10 minutes at 3000 rpm. Serum was then divided into 1.8-mL cryotubes and promptly frozen. Samples were maintained at –70°C25 at the DCCT Central Biochemistry Laboratory, Department of Laboratory Medicine and Pathology, University of Minnesota until preparation for the present study.
Laboratory Analysis
Baseline and 3-year serum samples were thawed and assayed for hsCRP with a latex-enhanced immunonephelometric assay on a BN II analyzer (Dade Behring). Serum levels of sICAM-1, sVCAM-1, and sTNF-R1 were measured by ELISAs (R&D Systems). All samples were handled in an identical and masked fashion, with baseline and 3-year specimens interspersed for analysis.
Statistical Analysis
We examined the change in levels of each biomarker from baseline to the 3-year follow-up visit and tested whether this was related to the randomization assignment using the Mann-Whitney U test. These relationships were also estimated with linear regression, with the natural logarithm of the 3-year follow-up levels of the biomarker as the dependent variable and treatment assignment, the natural log of baseline levels of the biomarker, and other factors as the independent variables. To enhance comparability of results among markers, we retained a common set of covariates for each marker of inflammation. These covariates included baseline levels of the marker, baseline HbA1c, age, sex, duration of diabetes, body mass index, and smoking status. For continuous variables, we subtracted the mean value from the measured value to obtain a distribution centered on the mean, to aid in interpretation of the models (particularly the intercepts). For each marker, we then fit an additional model in which we adjusted for the average HbA1c level over follow-up, to examine the relationships with glucose control per se. In further models, we considered interactions between baseline levels of the markers and randomized treatment assignment by including product terms in the models. Finally, for both hsCRP and sTNF-R1, molecules strongly correlated with body weight, we explored whether there was a relationship with weight gain or an interaction between weight gain and treatment assignment. For these analyses, we divided weight gain into approximate thirds based on the distribution of weight gain in the total DCCT population. We included terms for thirds of weight gain and product terms that represented their interaction with treatment assignment in the regression models.
We performed regression diagnostics to identify unduly influential observations and repeated models that excluded these observations. Because exclusion of the most influential observations did not change interpretation of any findings, we present only the models for all participants. In a secondary analysis, we used the baseline to 3-year change in levels of the biomarkers as the independent variable for our models. Because the distributions of the changes in inflammatory markers were skewed, and log transformations were not possible because of negative values (because a participant’s marker level could have either decreased or increased), we obtained the normalized rank transformations of each of the markers and regressed covariates on these ranks to obtain probability values for the regression coefficients. There were no substantial differences in this analysis compared with the findings presented.
Results
Of the 400 DCCT participants randomly selected for the present study, 385 had sufficient baseline and 3-year follow-up serum samples to measure at least 1 of the selected biomarkers; thus, any potential effect of dropouts on results is minimal and not of concern. In this subset of DCCT participants, there were no significant differences in levels of the inflammatory markers hsCRP, sICAM-1, sVCAM-1, or sTNF-R1 at baseline between subjects assigned to intensive versus conventional therapy (Table 1).
Median levels of hsCRP increased over time by 0.19 mg/L (23.9%) in the conventional treatment group and by 0.08 mg/L (9.3%) among those assigned to intensive therapy (Table 2). This difference was not statistically significant (P=0.53). In contrast, median levels of both sICAM-1 and sVCAM-1 increased from baseline to 3 years of follow-up by 4.9 ng/mL (1.7%) and 4.6 ng/mL (1.0%), respectively, among subjects assigned to conventional treatment but decreased by 2.5 mg/L (0.9%) and 9.5 mg/L (2.1%), respectively, among subjects assigned to intensive therapy. The differences in change in levels for intensive versus conventional therapy were statistically significant (each P=0.03). Levels of sTNF-R1 increased among subjects in both treatment arms of the DCCT but by a greater amount among subjects assigned to intensive therapy (6.0 pg/mL [0.6%] for conventional treatment versus 108.1 pg/mL [10.9%] for intensive therapy; P=0.002). Because baseline levels had a substantial impact on changes over time, as would be expected as more extreme levels regressed toward the mean, we reexamined associations with treatment assignment in models that adjusted for baseline levels of the markers, as well as for other potential predictors of change.
In analyses of hsCRP, after adjustment for baseline levels of hsCRP (P<0.0001), female sex (P<0.0001), and other factors, 3-year follow-up hsCRP levels were not significantly different among those assigned to intensive versus conventional control (P=0.98; Table 3). In models for sICAM-1, both baseline sICAM-1 levels (P<0.0001) and current cigarette smoking at baseline (P<0.0001) were significant predictors of 3-year follow-up levels. After adjustment for these and other variables, intensive therapy was still associated with a significant decrease in sICAM-1 levels after 3 years of follow-up (P=0.02; Table 3). For sVCAM-1, after adjustment for baseline levels (P<0.0001), age (P=0.0005), and other factors, the association of intensive therapy with change in sVCAM-1 was no longer significant (P=0.19; Table 3). Lower baseline levels of sTNF-R1 (P<0.0001) and older age (P=0.01) were associated with an increase in sTNF-R1 at 3 years. After adjustment for these and other covariates, intensive diabetes control was associated with a significantly greater increase in sTNF-R1 levels than conventional therapy (P=0.03; Table 3).
In further analyses, we included the average level of HbA1c over follow-up in models to investigate the extent to which any observed effect of intensive control might be mediated through the degree of glucose control achieved. For hsCRP, adjustment for average levels of HbA1c revealed a significant increase in hsCRP among those assigned to intensive control (P=0.04). In this model, the level of hsCRP after 3 years of follow-up was inversely associated with baseline HbA1c levels (P=0.003), but there was a direct association with average HbA1c during follow-up (P=0.005; Table 4). In contrast, for sICAM-1, sVCAM-1, and sTNF-R1, adjustment for average HbA1c attenuated the effect of intensive control, which was not significant for any of these markers (Table 4). Additional adjustment for changes in cigarette smoking during follow-up did not change any of these findings (data not shown).
Finally, given the known tendency for subjects to gain weight after initiation of intensive therapy, the strong relationship between body mass index and hsCRP levels, and the observation that intensive control increased hsCRP levels after adjustment for average HbA1c levels, we examined weight change and whether there might be an interaction between weight change and randomized treatment assignment that might provide some explanation for the lack of an overall effect of intensive therapy on change in levels of hsCRP. In this analysis, we identified a significant interaction between weight change and treatment assignment (P=0.03; Table 5). In separate analyses in the 2 treatment groups, among subjects assigned to intensive therapy, there was a median increase of 0.54 mg/L (75%) in hsCRP levels at 3 years of follow-up among those in the top third of weight change (mean body mass index at 3 years of follow-up 27.0 kg/m2) compared with a decrease of 0.24 mg/L (20%) among subjects in the lowest third of weight change (mean body mass index at 3 years of follow-up 22.8 kg/m2; P=0.0004; Figure). In contrast, among those assigned to conventional control, there was no significant difference in the change in hsCRP according to weight change category (P=0.88). We also investigated potential effects of weight gain with regard to sTNF-R1 and found a significantly greater increase in sTNF-R1 levels among subjects who gained the most weight (P=0.007; Table 5). In contrast to hsCRP, for sTNFR-1, there was no significant interaction of weight gain with intensive therapy (data not shown).
Interaction of weight gain and intensive diabetes control on change in levels of hsCRP in DCCT. Median change in hsCRP (mg/L) from baseline to 3 years, by randomization assignment and third of weight gain over same period. White bars are for conventional treatment group, gray bars for those assigned to intensive diabetes therapy. Probability values (Wilcoxon rank sum test) for change in hsCRP were 0.88 and 0.0004, respectively, for subgroups undergoing conventional and intensive therapy.
Discussion
Randomized clinical trials have not conclusively shown that intensive glycemic control reduces the risk of CVD among people with either type 126 or type 223 diabetes mellitus. Some epidemiological data suggest a stronger relationship with insulin resistance-related factors than with HbA1c.27 In the DCCT, intensive therapy was associated with a higher risk of overweight and obesity,28 which may adversely affect CVD risk. Several lines of evidence have come to light that suggest that inflammation is a key player in the development of CVD.5 A number of studies have found that several circulating molecules involved in inflammation, such as sICAM-1, hsCRP, and TNF-R1, predict risk of CVD, and some of these molecules, notably hsCRP and sTNF-R1, are highly influenced by body weight. The present data from the DCCT indicate that among subjects with type 1 diabetes mellitus, intensive glycemic control is associated with a significant reduction in levels of sICAM-1 but not with change in levels of hsCRP, the marker of inflammation most strongly and consistently predictive of CVD in epidemiological studies. Furthermore, the present analyses indicated that there was a significant rise in levels of hsCRP among intensively treated subjects who gained the most weight. There was also a significant rise in sTNF-R1 among intensively treated subjects, which appears to be explained by greater weight gain in the intensively treated group.
Although this is the largest study to date of these relationships among individuals with type 1 diabetes mellitus, several limitations deserve consideration. Given the uniqueness of the DCCT population, generalizability of these findings to individuals with type 2 diabetes mellitus and to ethnic minority groups is uncertain. The present analyses of weight change as a potential predictor of change may be controversial, because they include in the model a factor (weight gain) that is directly influenced by treatment assignment and is thus an intermediate variable. Multiple testing may have also resulted in an inflation of the type 1 error rate. However, we think that our analyses are informative because they provide a potentially relevant explanation for the overall lack of an effect of intensive therapy on hsCRP levels, in contrast to a significant increase after adjustment for average HbA1c during follow-up, and for the greater increase in sTNF-R1 among subjects undergoing intensive therapy. We were also limited to a single measure of the inflammatory markers at each time point, which may not accurately reflect the long-term or average levels of exposure that are likely to be most relevant to CVD. However, single measures have shown significant predictive value for CVD in prior prospective studies.10–12,15,16 Long-term follow-up of the DCCT cohort in the Epidemiology of Diabetes Interventions and Complications (EDIC) study presents the opportunity for future studies of the persistence of the observed changes in inflammatory biomarkers and their potential associations with CVD risk.
hsCRP has been strongly related to risk of CVD in a number of prospective studies and with early carotid atherosclerosis in young subjects with type 1 diabetes mellitus.9,29 Most studies have shown a positive correlation between hsCRP and HbA1c.6,30 In the present study, however, we found no overall difference in the change in levels of hsCRP comparing intensive and conventional therapy. There are few other data available, and none that we know of for type 1 diabetes mellitus or with comparable follow-up. A significant reduction in CRP was noted with an intensive insulin regimen in critically ill subjects.31 Moreover, CRP fell after 16 weeks of insulin treatment of 22 subjects with poorly controlled type 2 diabetes mellitus.32 In the present study of type 1 diabetes mellitus, we measured change in hsCRP after 3 years of intensive or conventional treatment and found no overall change in hsCRP. Interestingly, however, any potential benefit of improved glycemic control may have been limited by the weight gain in subjects assigned to intensive therapy.
By the end of the DCCT, the prevalence of obesity reached 33% among subjects assigned to intensive therapy versus 19% in the conventionally treated group.28 Subjects who gained the most weight with intensive therapy tended to experience metabolic changes similar to those seen in the insulin resistance syndrome.19 The present study adds to these findings by demonstrating an increase in hsCRP among intensively treated subjects who gained the most weight. This finding is consistent with the known direct relationship between body weight and hsCRP33 and with a large number of studies that show a relationship of CRP with the development of insulin resistance.13,34–38 Particularly in light of evidence that suggests a stronger relationship of insulin resistance-related factors than glycemia with CVD,27 if the increase in hsCRP levels associated with weight gain persists over time, this in conjunction with the other adverse metabolic changes that occur with excessive weight gain during intensive therapy may result in an increased risk of CVD.
Raised circulating levels of adhesion molecules are thought to indicate a state of endothelial activation with consequent induction of immunological activity.39,40 Levels of sICAM-115 but not sVCAM-141 predict future cardiovascular events. Although adhesion molecules have also been implicated in the pathogenesis of type 1 diabetes mellitus,42,43 there is little information about their relationship with metabolic control or their potential role in the pathogenesis of complications. Subjects with diabetes mellitus appear to have elevated levels of sICAM-1 and sVCAM-1.44 In a small study, sICAM-1 was significantly higher among diabetic subjects who later developed CVD than among those who did not.45 In the present study, levels of sICAM-1 fell significantly after the introduction of intensive therapy. Our analyses suggested that this decrease was mediated through reduction of glucose. Given the previously described association of sICAM-1 with future CVD, the fall in sICAM-1 with intensive therapy might exert a beneficial influence on future risk of CVD. However, this finding must be interpreted cautiously in light of the inconsistent effects of intensive therapy on other predictive markers, in particular hsCRP.
The presence of advanced glycosylation end products and glycated albumin in the diabetic milieu can enhance expression of tumor necrosis factor (TNF)-.46,47 However, TNF- has a short half-life, and high levels of its receptors can interfere with its detection,48 which limits its usefulness as a biomarker. Although less widely studied than other markers of inflammation, levels of TNF receptors are elevated in a variety of settings.49,50 In a cross-sectional study, higher levels of TNF receptors were associated with carotid atherosclerosis in subjects less than 70 years of age.16 In the present study, we observed a significantly greater increase in sTNF-R1 among subjects assigned to intensive therapy; however, after adjustment for weight gain during follow-up, the effect of intensive therapy was not significant. Rather, mean levels of sTNF-R1 increased to a greater extent among subjects who gained weight, regardless of type of therapy.
In summary, we examined the effect of intensive versus conventional therapy on 4 selected markers of inflammation in the DCCT. The effect of intensive control on levels of inflammatory markers was not uniform, which suggests a complex relationship between intensive diabetes therapy and inflammation. Levels of sICAM-1 were reduced, but there was an increase in sTNF-R1 among intensively treated subjects and of hsCRP among intensively treated subjects who gained the most weight. To the extent that hsCRP is a risk factor, the risk of atherosclerosis among subjects with type 1 diabetes mellitus may be influenced by whether or not they gain weight while undergoing intensive therapy.
Acknowledgments
This study was supported by Juvenile Diabetes Research Foundation grant No. 1-2000-646 (Dr Schaumberg), the Earle P. Charlton, Jr. Charitable Foundation (Dr Nathan), and a contract with the Division of Diabetes, Endocrinology, and Metabolic Diseases of the National Institute of Diabetes and Digestive and Kidney Diseases.
References
Laing SP, Swerdlow AJ, Slater SD, Burden AC, Morris A, Waugh NR, Gatling W, Bingley PJ, Patterson CC. Mortality from heart disease in a cohort of 23,000 patients with insulin-treated diabetes. Diabetologia. 2003; 46: 760–765.
Green A, Hougaard P. Epidemiological studies of diabetes mellitus in Denmark: 5: mortality and causes of death among insulin-treated diabetic patients. Diabetologia. 1984; 26: 190–194.
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. 1984; 33: 271–276.
Nathan DM, Lachin J, Cleary P, Orchard T, Brillon DJ, Backlund JY, O’Leary DH, Genuth S. Intensive diabetes therapy and carotid intima-media thickness in type 1 diabetes mellitus. N Engl J Med. 2003; 348: 2294–2303.
Libby P, Ridker PM, Maseri A. Inflammation and atherosclerosis. Circulation. 2002; 105: 1135–1143.
Gomes MB, Piccirillo LJ, Nogueira VG, Matos HJ. Acute-phase proteins among patients with type 1 diabetes. Diabetes Metab. 2003; 29: 405–411.
Kilpatrick ES, Keevil BG, Jagger C, Spooner RJ, Small M. Determinants of raised C-reactive protein concentration in type 1 diabetes. QJM. 2000; 93: 231–236.
King DE, Mainous AG III, Buchanan TA, Pearson WS. C-reactive protein and glycemic control in adults with diabetes. Diabetes Care. 2003; 26: 1535–1539.
Hayaishi-Okano R, Yamasaki Y, Katakami N, Ohtoshi K, Gorogawa S, Kuroda A, Matsuhisa M, Kosugi K, Nishikawa N, Kajimoto Y, Hori M. Elevated C-reactive protein associates with early-stage carotid atherosclerosis in young subjects with type 1 diabetes. Diabetes Care. 2002; 25: 1432–1438.
Ridker PM, Rifai N, Rose L, Buring JE, Cook NR. Comparison of C-reactive protein and low-density lipoprotein cholesterol levels in the prediction of first cardiovascular events. N Engl J Med. 2002; 347: 1557–1565.
Ridker PM, Hennekens CH, Buring JE, Rifai N. C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med. 2000; 342: 836–843.
Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men [published erratum appears in N Engl J Med. 1997;337:356]. N Engl J Med. 1997; 336: 973–979.
Pradhan AD, Manson JE, Rifai N, Buring JE, Ridker PM. C-reactive protein, interleukin 6, and risk of developing type 2 diabetes mellitus. JAMA. 2001; 286: 327–334.
Ridker PM, Buring JE, Cook NR, Rifai N. C-reactive protein, the metabolic syndrome, and risk of incident cardiovascular events: an 8-year follow-up of 14 719 initially healthy American women. Circulation. 2003; 107: 391–397.
Ridker PM, Hennekens CH, Roitman-Johnson B, Stampfer MJ, Allen J. Plasma concentration of soluble intercellular adhesion molecule 1 and risks of future myocardial infarction in apparently healthy men. Lancet. 1998; 351: 88–92.
Elkind MS, Cheng J, Boden-Albala B, Rundek T, Thomas J, Chen H, Rabbani LE, Sacco RL. Tumor necrosis factor receptor levels are associated with carotid atherosclerosis. Stroke. 2002; 33: 31–37.
Meigs JB, Hu FB, Rifai N, Manson JE. Biomarkers of endothelial dysfunction and risk of type 2 diabetes mellitus. JAMA. 2004; 291: 1978–1986.
Hu FB, Meigs JB, Li TY, Rifai N, Manson JE. Inflammatory markers and risk of developing type 2 diabetes in women. Diabetes. 2004; 53: 693–700.
Purnell JQ, Hokanson JE, Marcovina SM, Steffes MW, Cleary PA, Brunzell JD. Effect of excessive weight gain with intensive therapy of type 1 diabetes on lipid levels and blood pressure: results from the DCCT: Diabetes Control and Complications Trial [published erratum appears in JAMA. 1998;280:1484]. JAMA. 1998; 280: 140–146.
Bermudez EA, Rifai N, Buring J, Manson JE, Ridker PM. Interrelationships among circulating interleukin-6, C-reactive protein, and traditional cardiovascular risk factors in women. Arterioscler Thromb Vasc Biol. 2002; 22: 1668–1673.
Vikram NK, Misra A, Dwivedi M, Sharma R, Pandey RM, Luthra K, Chatterjee A, Dhingra V, Jailkhani BL, Talwar KK, Guleria R. Correlations of C-reactive protein levels with anthropometric profile, percentage of body fat and lipids in healthy adolescents and young adults in urban North India. Atherosclerosis. 2003; 168: 305–313.
Effect of intensive diabetes management on macrovascular events and risk factors in the Diabetes Control and Complications Trial. Am J Cardiol. 1995; 75: 894–903.
UK Prospective Diabetes Study (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.
The 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. 1993; 329: 977–986.
The Diabetes Control and Complications Trial Research Group. DCCT Manual of Operations. Springfield, Va: Department of Commerce, National Technical Information Service; 1993.
The Diabetes Control and Complications Trial Research Group. Effect of intensive diabetes management on macrovascular events and risk factors in the Diabetes Control and Complications Trial. Am J Cardiol. 1995; 75: 894–903.
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. 2003; 26: 1374–1379.
Influence of intensive diabetes treatment on body weight and composition of adults with type 1 diabetes in the Diabetes Control and Complications Trial. Diabetes Care. 2001; 24: 1711–1721.
Colhoun HM, Schalkwijk C, Rubens MB, Stehouwer CD. C-reactive protein in type 1 diabetes and its relationship to coronary artery calcification. Diabetes Care. 2002; 25: 1813–1817.
Hansen TK, Thiel S, Knudsen ST, Gravholt CH, Christiansen JS, Mogensen CE, Poulsen PL. Elevated levels of mannan-binding lectin in patients with type 1 diabetes. J Clin Endocrinol Metab. 2003; 88: 4857–4861.
Hansen TK, Thiel S, Wouters PJ, Christiansen JS, Van den Berghe G. Intensive insulin therapy exerts antiinflammatory effects in critically ill patients and counteracts the adverse effect of low mannose-binding lectin levels. J Clin Endocrinol Metab. 2003; 88: 1082–1088.
Yudkin JS, Panahloo A, Stehouwer C, Emeis JJ, Bulmer K, Mohamed-Ali V, Denver AE. The influence of improved glycaemic control with insulin and sulphonylureas on acute phase and endothelial markers in type II diabetic subjects. Diabetologia. 2000; 43: 1099–1106.
Ford ES. Body mass index, diabetes, and C-reactive protein among U.S. adults. Diabetes Care. 1999; 22: 1971–1977.
Barzilay JI, Abraham L, Heckbert SR, Cushman M, Kuller LH, Resnick HE, Tracy RP. The relation of markers of inflammation to the development of glucose disorders in the elderly: the Cardiovascular Health Study. Diabetes. 2001; 50: 2384–2389.
Festa A, D’Agostino R Jr, Tracy RP, Haffner SM. Elevated levels of acute-phase proteins and plasminogen activator inhibitor-1 predict the development of type 2 diabetes: the Insulin Resistance Atherosclerosis Study. Diabetes. 2002; 51: 1131–1137.
Festa A, D’Agostino R Jr, Howard G, Mykkanen L, Tracy RP, Haffner SM. Chronic subclinical inflammation as part of the insulin resistance syndrome: the Insulin Resistance Atherosclerosis Study (IRAS). Circulation. 2000; 102: 42–47.
Schmidt MI, Duncan BB, Sharrett AR, Lindberg G, Savage PJ, Offenbacher S, Azambuja MI, Tracy RP, Heiss G. Markers of inflammation and prediction of diabetes mellitus in adults (Atherosclerosis Risk in Communities study): a cohort study. Lancet. 1999; 353: 1649–1652.
Wu T, Dorn JP, Donahue RP, Sempos CT, Trevisan M. Associations of serum C-reactive protein with fasting insulin, glucose, and glycosylated hemoglobin: the Third National Health and Nutrition Examination Survey, 1988–1994. Am J Epidemiol. 2002; 155: 65–71.
Gearing AJ, Newman W. Circulating adhesion molecules in disease. Immunol Today. 1993; 14: 506–512.
Roep BO, Heidenthal E, de Vries RR, Kolb H, Martin S. Soluble forms of intercellular adhesion molecule-1 in insulin-dependent diabetes mellitus. Lancet. 1994; 343: 1590–1593.
de Lemos JA, Hennekens CH, Ridker PM. Plasma concentration of soluble vascular cell adhesion molecule-1 and subsequent cardiovascular risk. J Am Coll Cardiol. 2000; 36: 423–426.
Lampeter ER, Kishimoto TK, Rothlein R, Mainolfi EA, Bertrams J, Kolb H, Martin S. Elevated levels of circulating adhesion molecules in IDDM patients and in subjects at risk for IDDM. Diabetes. 1992; 41: 1668–1671.
Yang XD, Michie SA, Mebius RE, Tisch R, Weissman I, McDevitt HO. The role of cell adhesion molecules in the development of IDDM: implications for pathogenesis and therapy. Diabetes. 1996; 45: 705–710.
Fasching P, Veitl M, Rohac M, Streli C, Schneider B, Waldhausl W, Wagner OF. Elevated concentrations of circulating adhesion molecules and their association with microvascular complications in insulin-dependent diabetes mellitus. J Clin Endocrinol Metab. 1996; 81: 4313–4317.
Jude EB, Douglas JT, Anderson SG, Young MJ, Boulton AJ. Circulating cellular adhesion molecules ICAM-1, VCAM-1, P- and E-selectin in the prediction of cardiovascular disease in diabetes mellitus. Eur J Intern Med. 2002; 13: 185–189.
Neumann A, Schinzel R, Palm D, Riederer P, Munch G. High molecular weight hyaluronic acid inhibits advanced glycation endproduct-induced NF-kappaB activation and cytokine expression. FEBS Lett. 1999; 453: 283–287.
Abordo EA, Thornalley PJ. Synthesis and secretion of tumour necrosis factor-alpha by human monocytic THP-1 cells and chemotaxis induced by human serum albumin derivatives modified with methylglyoxal and glucose-derived advanced glycation endproducts. Immunol Lett. 1997; 58: 139–147.
Moreau E, Philippe J, Couvent S, Leroux-Roels G. Interference of soluble TNF-alpha receptors in immunological detection of tumor necrosis factor-alpha. Clin Chem. 1996; 42: 1450–1453.
Steiner G, Studnicka-Benke A, Witzmann G, Hofler E, Smolen J. Soluble receptors for tumor necrosis factor and interleukin-2 in serum and synovial fluid of patients with rheumatoid arthritis, reactive arthritis and osteoarthritis. J Rheumatol. 1995; 22: 406–412.
Studnicka-Benke A, Steiner G, Petera P, Smolen JS. Tumour necrosis factor alpha and its soluble receptors parallel clinical disease and autoimmune activity in systemic lupus erythematosus. Br J Rheumatol. 1996; 35: 1067–1074.(Effect of Intensive Glyce)