help button home button Endocrine Society JCEM
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Journal of Clinical Endocrinology & Metabolism , doi:10.1210/jc.2007-2405
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Maiolino, G.
Right arrow Articles by Rossi, G. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Maiolino, G.
Right arrow Articles by Rossi, G. P.
Related Collections
Right arrow Cardiovascular Endocrinology
The Journal of Clinical Endocrinology & Metabolism Vol. 93, No. 9 3333-3340
Copyright © 2008 by The Endocrine Society

Plasma Adiponectin for Prediction of Cardiovascular Events and Mortality in High-Risk Patients

Giuseppe Maiolino, Maurizio Cesari, Daniele Sticchi, Mario Zanchetta, Luigi Pedon, Katia Antezza, Achille C. Pessina and Gian Paolo Rossi

Division of Cardiology (M.Z., L.P.), Cittadella Hospital, 35013 Cittadella, Italy; and Department of Clinical and Experimental Medicine (G.M., M.C., D.S., K.A., A.C.P., G.P.R.), University of Padova, 35128 Padova, Italy

Address all correspondence and requests for reprints to: Professor Gian Paolo Rossi, M.D., F.A.C.C., F.A.H.A., Department of Clinical and Experimental Medicine, Clinica Medica 4, University Hospital, via Giustiniani, 2, 35128 Padova, Italy. E-mail: gianpaolo.rossi{at}unipd.it.


   Abstract
Top
 Abstract
Introduction
 Patients and Methods
 Results
 Discussion
 References
 
Context: The prognostic value of plasma levels of adiponectin, an adipocytokine with antiatherogenic, antiinflammatory, and insulin-sensitizing effects, is contentious.

Objective: The objective of the study was to investigate whether plasma adiponectin levels predict cardiovascular (CV) events and mortality in high-risk coronary artery disease (CAD) patients.

Design, Setting, Participants, and Main Outcome Measure: We measured plasma adiponectin and examined its impact on the incidence of CV deaths and events at follow-up in the context of all potentially relevant background covariates in 712 high-risk patients of the Genetic and ENvironmental factors in Coronary Atherosclerosis study who underwent coronary angiography for suspected CAD. Based on the population plasma adiponectin median (6.38 µg/ml, interquartile range 4.2–10.2), we split the patients in a high- and a low-plasma adiponectin subgroup. After a median follow-up of 3.8 years (interquartile range 3.3–4.3 yr), outcome data were obtained in 100% of the patients and 45 CV deaths (6.4%) were recorded. Kaplan-Meier analysis unexpectedly showed a higher CV death rate in high-plasma adiponectin than low-plasma adiponectin patients. By contrast, multivariate Cox regression analysis, in which potential confounders, including ongoing medical treatment, were considered, showed no impact of plasma adiponectin on CV death. Similar negative results were obtained using the propensity score that considered all relevant covariables and medical treatment rate, which differed between the high- and low-plasma adiponectin group.

Conclusions: In high-risk CAD patients, plasma adiponectin above the median (6.38 µg/ml) implies a paradoxical higher risk of CV death. However, when relevant covariates that differ between high- and low-plasma adiponectin groups are considered, this association wanes, indicating that the clustering of plasma adiponectin with other covariates can abolish its impact on CV prognosis.


   Introduction
Top
 Abstract
 Introduction
Patients and Methods
 Results
 Discussion
 References
 
Adiponectin acts as an insulin-sensitizer (antidiabetic) (for review, see Ref. 1) and exerts antiinflammatory activities by specifically binding to collagen I, III, and V in the subendothelial space of acutely injured arterial walls (2), thus preventing neointimal formation in response to injury and promoting angiogenesis (3, 4, 5). Moreover, adiponectin blunts several crucial steps of the atherogenic process, such as adhesin expression in endothelial cells, monocytes adhesion to endothelial cells, nuclear factor-{kappa}B signaling, growth and function of macrophages including their transformation to foam cells, and proliferation and migration of vascular smooth muscle cells (2, 6, 7, 8). These effects might depend, at least in part, on the ability of adiponectin to stimulate nitric oxide production from the endothelium (9) and suppress TNF-{alpha} secretion from macrophages with ensuing reduced expression of adhesins on the endothelium (6). In keeping with these evidences, cross-sectional studies have reported low plasma adiponectin levels in conditions that are characterized by accelerated atherosclerosis, such as insulin resistance and hyperinsulinemia, type 2 diabetes mellitus, overweight/obesity, cigarette smoking, and coronary artery disease (CAD) (10, 11, 12, 13, 14, 15, 16, 17, 18). Of note, we have recently shown that plasma adiponectin levels are strongly heritable (19). Moreover, low plasma adiponectin levels have been associated with endothelial dysfunction, which is partly genetically determined (20) and entails a key early feature of cardiovascular (CV) disease and a predictor of CV events (21).

Notwithstanding the body of evidences linking low plasma adiponectin levels with CV phenotypes, information on their impact on prognosis is limited and conflicting. Some studies suggested an increased CV risk associated with low plasma adiponectin levels (22, 23, 24, 25, 26), but others in patients with CAD (27, 28), chronic heart failure (29), chronic renal failure (30), and in older black Americans (31) thereafter showed the opposite. However, in these studies the impact of ongoing medical therapy, which is the rule in these high-risk patients and might interact with plasma adiponectin levels in determining outcome, was neglected. We therefore hypothesized that controversies on the prognostic impact of plasma adiponectin levels could be resolved by examining prospectively a high CV risk cohort, in which the antiatherogenic properties of adiponectin in the subendothelial space of unstable plaques could be magnified, and by giving proper consideration to all potentially relevant covariates, including treatment rate, with use of multivariate survival analyses and calculation of the propensity score (32). The Genetic and ENvironmental factors in Coronary Atherosclerosis (GENICA) study contains exhaustive information on outcome at follow-up and also on overall CV risk factors and ongoing medical treatment at enrollment. Therefore, it provided a unique opportunity to test the hypothesis that plasma adiponectin levels carried independent prognostic information. The aim of the present study was to investigate whether plasma adiponectin levels represent a risk marker for future CV events and mortality in high-risk patients with established CAD.


   Patients and Methods
Top
 Abstract
 Introduction
 Patients and Methods
Results
 Discussion
 References
 
Study participants

The protocol and enrollment criteria of the GENICA study have already been published (33, 34). Briefly, consecutive Caucasian patients referred for coronary angiography for investigation of chest pain and/or suspected CAD between 1999 and 2001 were enrolled with signature of a consent form to participate in this study. Refusal to participate was the only exclusion criteria. The study protocol was approved by the Medical Ethics Committee. Information on medical history, smoking habits, presence/absence of arterial hypertension, diabetes mellitus, hypercholesterolemia, hypertriglyceridemia, and ongoing medications was gathered with a staff-administered questionnaire. Definitions for body mass index (BMI), smoking status, diabetes mellitus, impaired glucose tolerance, hypercholesterolemia, and hypertriglyceridemia were already reported (33). Blood pressure (BP) was measured by mercury sphygmomanometer using Korotkoff phase V for diastolic, according to the World Health Organization guidelines; arterial hypertension was defined according to the European Society of Cardiology/European Society of Hypertension (ESC/ESH) guidelines criteria, or use of antihypertensive agent(s). Hypertension was defined as systolic pressure 140 mm Hg or greater and/or diastolic pressure 90 mmHg or greater and/or use of antihypertensive drugs.

Coronary angiography

Left ventricular ejection fraction (LVEF) was measured as described (33, 35), and the grading of the coronary atherosclerotic burden was carried out with use of a modified Duke Prognostic Index score, as described (33, 35). This score considers only major epicardial coronary arteries with 50% or greater diameter stenosis, including left main trunk, and goes from 0 (all major coronary arteries with lesions <50% diameter stenosis) to 100 (≥95% left main stenosis) but predicted 5-yr mortality of medically treated patients (36, 37).

Laboratory measurements

Patients were studied between 0830 and 1200 h. Blood samples were taken immediately before coronary angiography, put on ice, and centrifuged at 3000 x g (at 4 C for 10 min). Total cholesterol, high-density lipoprotein (HDL) cholesterol, triglycerides, glucose, sodium, potassium, blood urea nitrogen, and creatinine levels were measured with conventional methods. Plasma adiponectin concentration was evaluated by an ELISA method (Human Adiponectin/Acrp30 Quantikine ELISA kit, no. DRP300; R&D Systems, Minneapolis, MN, info{at}RnDSystems.com) that uses monoclonal antibodies recognizing the globular domain (essentially the C terminus of the protein) and full-length adiponectin. Inter- and intraarray coefficient of variation was 6.9 and 4.7%, respectively.

Follow-up data

The long-term outcome of the patients enrolled in the GENICA study was assessed blindly to baseline patient’s data with a predefined form through review of medical charts for the patients regularly seen at referring hospitals and with telephone interviews of patients and/or first-degree relatives and/or family doctors for those not attending regular follow-up visits. On-site investigation was carried out to ascertain the alive/death status for all the patients lost at follow-up. Predetermined primary end point was CV death, comprising sudden death and death due to congestive heart failure, acute coronary syndromes, or stroke (35). Moreover, a composite CV end point that comprised CV death and nonfatal myocardial infarction, acute coronary syndrome, stroke, and vascular surgery was examined. Events were validated by the adjudication committee blinded to patients’ biochemical profile; their exact date of occurrence was known, and therefore, survival data are shown in days on the time scale.

Statistical analysis

Serum triglycerides, creatinine, CAD Duke Prognostic Index score, LVEF, and plasma adiponectin levels were examined after log or square root transformation was taken when deemed necessary to improve approximation to Gaussian distributions. After data inspection it was decided a priori to identify and exclude univariate and multivariate outliers. The former were identified by calculating standardized z scores and excluded based on a z scores of |3.29| (absolute value) that corresponds to a P < 0.001; the latter were identified by calculating Mahalanobis distance using the procedure described by Tabachnick and Fidell (38): cases with {chi}2 in excess of 34.528 (13 df at {alpha} = 0.001) were considered outliers.

Student’s t test for unpaired groups and {chi}2 analysis was used to compare quantitative variables and the frequencies of categorical CAD risk factors between groups, respectively. Multiple linear regression analysis was used for identifying [among gender, age, LVEF, HDL cholesterol, low-density lipoprotein (LDL) cholesterol, BMI, creatinine, homocysteine, smoking habits, and triglycerides] significant predictors of plasma adiponectin levels. The backward strategy for the selection of variables was preferred to the more common forward strategy because it carries a lower risk of failing to find a relationship when one exists (39). CV death rates were estimated by Kaplan-Meier analysis with the log-rank test.

Cox sequential multiple regression analysis was used to identify the predictors of CV death or events at follow-up. Major known CV risk factors and prognostic determinants, history variables, and treatment data were initially entered in several models in blocks as follows: block 1 comprised gender, age, BMI, and cigarette smoking; block 2 comprised history of CV events (transient ischemic attack, stroke, angina, myocardial infarction, coronary artery bypass graft, percutaneous transluminal coronary angioplasty, peripheral vascular disease, vascular surgery); block 3 comprised ongoing treatment variables (with antiplatelets, calcium channel blockers, β-blockers, angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor blockers (ARBs), nitrates, warfarin, {alpha}-blockers, diuretics, heparin, digoxin, statins, and fibrates); block 4 comprised biochemical and clinical variables (lipids levels, creatinine, glomerular filtration rate, and history of hypertension); block 5 comprised LVEF and Duke Prognostic Index score; and block 6 comprised plasma adiponectin levels. After every block the variables that were not significant independent predictors of events were removed. After exclusion of redundant variables, a model with a maximum of five covariates, as permitted by the observed number of deaths or events, was eventually determined. The modified Duke Prognostic Index score was inserted in the model as either a continuous or categorical (dichotomized or divided into quartiles) variable.

Squared multiple correlations were used to assess absence of multicollinearity: covariates with a communality value in excess of 0.90 were considered to be redundant and therefore removed from further analysis (38).

Because several variables were imbalanced between the high- and low-plasma adiponectin group (Tables 1–3GoGoGo) to simultaneously balance all these covariates and obtain valid inferences about plasma adiponectin levels effects, we also exploited use of the propensity score. Exact adjustments made with this score have been described, on average, to effectively remove all the bias in the background covariates (32) and to be most useful for controlling for the effect of potentially significant covariates, particularly when the relatively small number of events limits the use of potentially relevant covariates in the Cox regression model (32). Therefore, we used logistic regression to create the propensity scores by entering in the model all the variables, including history of CV events and ongoing treatment, that differed significantly between the high- and low-plasma adiponectin levels at baseline (Tables 1–3GoGoGo). The Cox regression analysis was then repeated with only the propensity score and plasma adiponectin levels as covariates. Statistical significance was defined as P < 0.05. SPSS 15.00 for Windows (SPSS Italy Inc., Bologna, Italy) was used for all analyses.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Anthropometric and clinical features of the patients classified into high- and low-plasma adiponectin levels

 

View this table:
[in this window]
[in a new window]

  		TABLE 2. History and medical treatment in the patients classified into high- and low-plasma adiponectin levels

 

View this table:
[in this window]
[in a new window]

  		TABLE 3. Distribution of CV events of patients classified into high- and low-plasma adiponectin levels

 

   Results
Top
 Abstract
 Introduction
 Patients and Methods
 Results
Discussion
 References
 
Clinical characteristics

Based on sample size and power calculations, we randomly selected for this study 712 patients from the 1273 consecutive patients originally enrolled in the GENICA study. We found no univariate outlier and four multivariate outliers, who were therefore excluded from further analysis. The whole cohort median values of plasma adiponectin levels were 6.38 µg/ml (interquartile range 4.2–10.2). At multiple linear regression analysis, the predictors of plasma adiponectin levels (adjusted R2 square = 0.283, P < 0.0001) were: male gender (β = –0.283, P < 0.0001), age (β = 0.207, P < 0.0001), LVEF (β = –0.151, P < 0.0001), triglycerides (β = –0.152, P < 0.0001), HDL cholesterol (β = 0.177, P < 0.0001), BMI (β = –0.140, P < 0.0001), and creatinine (β = 0.105, P = 0.003).

Tables 1Go and 2Go show the anthropometric and clinical features of the patients divided according to the plasma adiponectin levels median value into high- and low-plasma adiponectin group. Age, gender, cigarette smoking, BMI, heart rate, glomerular filtration rate, HDL cholesterol, triglycerides, LVEF, CAD score, and plasma adiponectin levels differed significantly between groups. The low-plasma adiponectin group had significantly less women and more coronary atherosclerosis than the high-plasma adiponectin group, thus confirming our previous findings in a smaller cohort of nondiabetic patients. History of transient ischemic attack was less common, and that of myocardial infarction and angina was more prevalent in the low-plasma adiponectin group; however, there were no differences of history of coronary revascularization between the groups (Table 2Go). With regard to treatment rate, the antiplatelet agents, β-blockers, ACE inhibitors and ARBs, nitrates, warfarin, diuretics, and digoxin differed significantly between the groups (Table 2Go). In particular, treatment rate with warfarin, ACE inhibitors, ARBs, diuretics, and digoxin was higher in the high-plasma adiponectin group. On the other hand, treatment rate with antiplatelet agents, β-blockers, and nitrates was higher in the low-plasma adiponectin group.

By the National Cholesterol Education Program NCEP (Adult Treatment Panel ATP III) criteria, 96.1% of our patients fell into the highest class of risk (which confer a 10 yr risk for CAD >20%), 3.9% in the average (which confer a 10 yr risk for CAD between 10 and 20%), and none in the lowest.

Follow-up data

Complete follow-up data were obtained in all patients. After a median follow-up of 3.8 yr (interquartile range 3.3–4.3 yr), we observed 45 CV deaths (6.4%). This high rate agrees with the aforementioned attribution of the patients to the highest National Cholesterol Education Program class of risk. The incidence of CV events was also high, albeit similar in patients with low- and high-plasma adiponectin levels (Table 3Go). Proportionality of hazards for each covariate was confirmed at Cox regression analysis. Total cholesterol was removed from Cox regression analysis because of its collinearity with LDL cholesterol.

Prognostic impact of plasma adiponectin levels

Kaplan-Meier analysis showed a significantly (P = 0.019) higher CV death rate in the high-plasma, compared with the low-plasma, adiponectin group (Fig. 1Go). At variance, hierarchical multivariate Cox regression analysis with inclusion of potential confounders, whereas confirming an impact of some known predictors of CV death and events (Table 4Go), showed no effect of plasma adiponectin levels. Similarly, with use of the propensity score to adjust for the impact of all the background covariates (Tables 1Go and 2Go) that differed between the high- and low-plasma adiponectin group, we found no significant effect of plasma adiponectin levels on CV death or the composite CV end point.


Figure 1
View larger version (26K):
[in this window]
[in a new window]

  		FIG. 1. Results of the univariate Kaplan-Meier survival analysis showing an association of high-plasma adiponectin levels with CV death.

 

View this table:
[in this window]
[in a new window]

  		TABLE 4. Predictors of event-free survival at Cox regression analysis

 
The variables that remained in the final model predicting the outcome CV death and the composite CV end point are shown in Table 4Go ranked by decreasing Wald coefficient value. For CV mortality the most potent predictor was LVEF followed by treatment with calcium channel blockers. For the composite CV end point, the most potent predictor was the CAD Duke Prognostic Index score followed by LVEF and age. For both outcomes the odds risk (OR) associated with plasma adiponectin levels did not achieve significance (Table 4Go); the findings were similar when plasma adiponectin levels was examined in the context of the other variables summarized in the propensity score: for CV death the OR was 1.659 [95% confidence interval (CI) 0.988–2.788, P = 0.056], for the composite CV endpoint the OR was 1.094 (95% CI 0.815–1.468, P = NS). For both end points, the variable that had a stronger impact on removing the apparent association of plasma adiponectin levels with CV outcome in the univariate analysis was LVEF.


   Discussion
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Discussion
References
 
Because plasma adiponectin levels might exert its antiatherogenic activities, particularly in the subendothelial space of acutely injured arterial walls (2, 3, 4, 5), the choice of a high-risk CAD cohort as that recruited in the GENICA study, which likely carried a good deal of prone to destabilization atherosclerotic plaques, might be the ideal setting for challenging the hypothesis of a protective CV effect of the peptide. Moreover, this study has unique features that allowed gathering solid novel information in the conflicting area of the prognostic role of plasma adiponectin levels. They comprised: 1) the enrollment of a mixed-gender cohort of patients who had an accurate assessment of their global coronary atherosclerotic burden with coronary angiography; 2) the availability of thorough information on CV risk profile and medical treatment at baseline and comprehensive data at follow-up. This allowed investigation of the effect of plasma adiponectin levels in the context of the overall CV risk profile and ongoing medical treatment by state-of-the-art techniques, including assessment of the propensity score.

Prognostic effect of plasma adiponectin levels

Based on findings of decreased plasma adiponectin levels in CAD and diabetic patients (13), it was initially suggested that the presence of CAD and/or diabetes was a major determinant of plasma adiponectin levels and, alternatively, that low plasma adiponectin levels were involved in the development of these conditions (18). Subsequently it was reported that in 225 men who underwent coronary angiography, the lowest (<4.0 µg/ml) compared with the highest quartile (≥7.0 µg/ml) of plasma adiponectin levels implied a 2-fold increased risk of CAD, which remained significant after adjustment for diabetes, dyslipidemia, hypertension, smoking, and BMI (16). Similar results were reported in 34 patients with acute myocardial infarction (40) and in 400 nondiabetic high-risk Caucasian patients (41). In contrast, no association between plasma adiponectin levels and CAD was found in American Indians with a high prevalence of diabetes mellitus and obesity after adjustment for other CV risk factors (42).

These conflicting results might be due to differences of study design, inclusion criteria, and ethnic background but also, most likely in our view, to the cross-sectional nature of these studies that enrolled only survivors of CV deaths and therefore, by definition, were prone to a selection bias and serendipitous findings.

Prospective cohort studies, as the present one, are unlikely to be affected by these potential problems but have also given conflicting results. In end-stage renal disease patients followed up for 31 months, plasma adiponectin levels were significantly (P < 0.05) but weakly inversely related to CV events, albeit they failed to predict mortality (24). In a population-based study of 832 elderly men, 74.4% of whom were hypertensives, 36.3% dyslipidemic, and 10.7% diabetics, elevated serum levels of plasma adiponectin levels were found to be associated with a lower risk for coronary heart disease (CHD) after adjustment for classical risk factors but not for ongoing medical therapy (22). Likewise in 745 men with type 2 diabetes mellitus, increased plasma adiponectin levels were associated with a moderately decreased CAD risk (after adjustment for age, BMI, smoking, alcohol consumption, duration of diabetes). Of interest, the association was strongly reduced after adjustment for HDL cholesterol in both stable CHD patients (17) and men with type 2 diabetes (25), suggesting that the latter might partly mediate the association between adiponectin and CHD risk. Also in healthy men, two nested case-control studies reported a significant association between increased plasma adiponectin levels and lower risk of CHD (23, 26).

At variance, plasma adiponectin levels did not predict development of CHD in women of the British Women’s Heart and Health Study (43) and CV events in American Indians men and women of the Strong Heart Study (42). According to other prospective cohort studies, low plasma adiponectin levels would be protective: in patients with CAD (27, 28), heart failure (29), chronic renal failure patients (30), and in older African Americans (31). To justify these paradoxical findings, it was proposed that plasma adiponectin levels represents a marker of wasting that, by itself, is an important predictor of outcomes in these patient populations. Therefore, a high plasma adiponectin level implied a worse prognosis simply because it reflected a more severe stage of the undergoing condition (29). Importantly, it must be underlined that the effect of ongoing medical therapy was not taken into any consideration in the studies of prognostic impact of plasma adiponectin levels in CAD patients (27, 28, 31).

Prognostic effect of plasma adiponectin levels at univariate analysis

By Kaplan-Meier analysis, we found that plasma adiponectin levels had a significant prognostic impact: a lower rate of CV deaths was seen in the low compared with the high plasma adiponectin levels patients (Fig. 1Go), thus supporting previous results in high-risk patients (27, 28, 29, 30, 31). Whereas of some interest from risk stratification purposes, this finding lends no support to the hypothesis that plasma adiponectin levels plays a causative role in triggering CV events, as discussed below.

Prognostic effect of plasma adiponectin levels at multivariate analyses

Given the unbalanced distribution of background covariates between the two plasma adiponectin groups (Tables 1–3GoGoGo) we hypothesized that the worse prognosis associated with the higher plasma adiponectin levels could originate from the inability of univariate statistics to account for the effects of potentially relevant covariates. To test this hypothesis, we adopted two different strategies. First we used a hierarchical multivariate Cox hazard regression statistics strategy (38), which entailed removing at every step the covariates that were not relevant in the model, before adding those of potential interest at the next step. This approach allows adjustment for the factors that might be associated with the risk of CV death and events, including the previous medical history and therapy. Moreover, it is held to furnish more robust and reliable results than classical multivariate stepwise techniques (38). In these models we inserted as covariates treatment at enrollment, history of previous CV events, and, among common CV risk factors, the modified Duke CAD score of coronary atherosclerotic burden and LVEF, e.g. several well-established risk factors, some of which were never used previously in studies on the prognostic impact of plasma adiponectin levels. Of note, in one of the largest available prospective studies, the rate of treatment with aspirin, statins, ACE inhibitors, and oral estrogens was assessed at baseline, but only aspirin use was then entered in the regression models assessing the impact of plasma adiponectin levels on CV events (31). In another large study showing that high plasma adiponectin levels independently predicted a poorer outcome in CAD patients, the ongoing medical treatment was totally neglected (28).

It is therefore of interest that with utmost care to adjusting for covariates and concurrent treatments with agents that are commonly prescribed in the high-risk patients undergoing coronary angiography, we could not confirm the results of the univariate analysis: at the Cox regression analysis, the independent predictors of CV death were low LVEF and calcium channel blockers treatment but not the CAD Duke index score, plasma adiponectin levels, and age (Table 4Go). When the determinants of the composite end point were examined, the variables identified as predictors were the CAD Duke index score, a low LVEF, and age but not calcium channel blockers treatment and plasma adiponectin levels (Table 4Go).

To further support these results, we next used a more robust approach to regression adjustment that has been recently popularized (32), the calculation of the propensity score, followed by regression (covariance) adjustment. We used an ample set of potentially relevant covariates to estimate the propensity score, which was then entered with plasma adiponectin levels in the Cox regression model. It has been pointed out that this smaller model may allow performing diagnostic checks on the fit of the model more reliably than if many covariates were included in the model. Of note, with this analysis, we reached exactly the same conclusions as in the case in the hierarchical Cox multivariate regression, e.g. that there was no independent plasma adiponectin level effect on either CV death or composite CV events (Table 4Go). The variable that had the strongest impact on removing the apparent association of plasma adiponectin levels with CV outcome in the univariate analysis was LVEF. Thus, whereas higher plasma adiponectin levels are associated with a worse long-term CV outcome at univariate analysis, its effect is not an independent one but rather due to a clustering of plasma adiponectin levels with several other variables, including LVEF, which therefore deserve proper attention in prospective cohort studies. Noteworthy, similar conclusions on the lack of independent prognostic effect of plasma total and high-molecular-weight adiponectin were reached in a study in 1051 European patients with stable CHD at baseline (44) that was published after the submission of this manuscript.

Prognostic impact of plasma adiponectin levels and gender

There were fewer women and more coronary atherosclerosis in the low- than high-plasma adiponectin group (Table 1Go). Because women with CAD entails a selected group, it might be argued that this unbalanced distribution between plasma adiponectin groups could contribute to explaining the higher risk for CV deaths in the high-plasma adiponectin group. However, this explanation seems, in our view, unlikely because a significant impact of plasma adiponectin levels on CV mortality was not found disappeared at multivariate analyses after adjustment for gender and the levels on CHD was previously excluded in women (43).

Limitations of the study

Notwithstanding the robust prospective cohort study design used, some limitations need to be mentioned. The number of cardiovascular deaths observed in this cohort (n = 45 CV deaths) was relatively small, thus lowering the power of the study; therefore, some caution in interpreting results is advised. Moreover, as for all investigations of high-risk patients, who are prone to potentially fatal events disease, those who died before enrollment were excluded from our study. Therefore, our conclusions apply only to the CAD patients who survived long enough to be referred for coronary angiography. The selection of a subgroup of patients less susceptible to fatal complications of CAD might explain the intriguing finding of an unbalanced distribution of some risk factors, such as male gender, between the high- and low-adiponectin group shown in Table 1Go. Moreover, because the GENICA cohort comprised high-risk CV patients, our results are not relevant for evaluating the independent role of plasma adiponectin in healthy subjects for risk of CV disease events. Finally, the possibility that changes of baseline treatment occurring during follow-up, which we could not assess, could also have impacted on outcome should be mentioned.

Conclusions

This study demonstrated that higher-than-average plasma adiponectin levels is a marker of a worse CV prognosis in high-risk CAD patients of both genders. However, with use of multivariate Cox regression and propensity score calculation for covariates regression adjustment, we could not confirm an independent effect of plasma adiponectin levels on CV outcome. Thus, our results indicate that the overall risk profile, LVEF and also medical history, and concurrent medical treatment should be considered when assessing the usefulness of potential biomarkers of CV disease, as plasma adiponectin levels, in high-risk patients.


   Footnotes
 
This work was supported by grants from the University of Padova (to A.C.P.), research grants from Regione Veneto (863/01/98), Unindustria of Treviso and F.O.R.I.C.A. (to G.P.R.), and a grant from the Società Italiana dell’Ipertensione Arteriosa (to A.C.P.).

Disclosure Statement: G.M., M.C., D.S., M.Z., L.P., K.A., A.C.P., and G.P.R. have nothing to declare.

First Published Online August 12, 2008

Abbreviations: ACE, Angiotensin-converting enzyme; ARB, angiotensin receptor blocker; BMI, body mass index; BP, blood pressure; CAD, coronary artery disease; CHD, coronary heart disease; CI, confidence interval; CV, cardiovascular; GENICA, Genetic and ENvironmental factors in Coronary Atherosclerosis; HDL, high-density lipoprotein; LDL, low-density lipoprotein; LVEF, left ventricular ejection fraction; OR, odds risk.

Received October 29, 2007.

Accepted June 2, 2008.


   References
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Discussion
 References
 

  1. Berg AH, Scherer PE 2005 Adipose tissue, inflammation, and cardiovascular disease. Circ Res 96:939–949[Abstract/Free Full Text]
  2. Okamoto Y, Arita Y, Nishida M, Muraguchi M, Ouchi N, Takahashi M, Igura T, Inui Y, Kihara S, Nakamura T, Yamashita S, Miyagawa J, Funahashi T, Matsuzawa Y 2000 An adipocyte-derived plasma protein, adiponectin, adheres to injured vascular walls. Horm Metab Res 32:47–50[Medline]
  3. Yamauchi T, Kamon J, Minokoshi Y, Ito Y, Waki H, Uchida S, Yamashita S, Noda M, Kita S, Ueki K, Eto K, Akanuma Y, Froguel P, Foufelle F, Ferre P, Carling D, Kimura S, Nagai R, Kahn BB, Kadowaki T 2002 Adiponectin stimulates glucose utilization and fatty-acid oxidation by activating AMP-activated protein kinase. Nat Med 8:1288–1295[CrossRef][Medline]
  4. Ouchi N, Kobayashi H, Kihara S, Kumada M, Sato K, Inoue T, Funahashi T, Walsh K 2004 Adiponectin stimulates angiogenesis by promoting cross-talk between AMP-activated protein kinase and Akt signaling in endothelial cells. J Biol Chem 279:1304–1309[Abstract/Free Full Text]
  5. Kubota N, Terauchi Y, Yamauchi T, Kubota T, Moroi M, Matsui J, Eto K, Yamashita T, Kamon J, Satoh H, Yano W, Froguel P, Nagai R, Kimura S, Kadowaki T, Noda T 2002 Disruption of adiponectin causes insulin resistance and neointimal formation. J Biol Chem 277:25863–25866[Abstract/Free Full Text]
  6. Ouchi N, Kihara S, Arita Y, Okamoto Y, Maeda K, Kuriyama H, Hotta K, Nishida M, Takahashi M, Muraguchi M, Ohmoto Y, Nakamura T, Yamashita S, Funahashi T, Matsuzawa Y 2000 Adiponectin, an adipocyte-derived plasma protein, inhibits endothelial NF-{kappa}B signaling through a cAMP-dependent pathway. Circulation 102:1296–1301[Abstract/Free Full Text]
  7. Ouchi N, Kihara S, Arita Y, Nishida M, Matsuyama A, Okamoto Y, Ishigami M, Kuriyama H, Kishida K, Nishizawa H, Hotta K, Muraguchi M, Ohmoto Y, Yamashita S, Funahashi T, Matsuzawa Y 2001 Adipocyte-derived plasma protein, adiponectin, suppresses lipid accumulation and class A scavenger receptor expression in human monocyte-derived macrophages. Circulation 103:1057–1063[Abstract/Free Full Text]
  8. Arita Y, Kihara S, Ouchi N, Maeda K, Kuriyama H, Okamoto Y, Kumada M, Hotta K, Nishida M, Takahashi M, Nakamura T, Shimomura I, Muraguchi M, Ohmoto Y, Funahashi T, Matsuzawa Y 2002 Adipocyte-derived plasma protein adiponectin acts as a platelet-derived growth factor-BB-binding protein and regulates growth factor-induced common postreceptor signal in vascular smooth muscle cell. Circulation 105:2893–2898[Abstract/Free Full Text]
  9. Chen H, Montagnani M, Funahashi T, Shimomura I, Quon MJ 2003 Adiponectin stimulates production of nitric oxide in vascular endothelial cells. J Biol Chem 278:45021–45026[Abstract/Free Full Text]
  10. Weyer C, Funahashi T, Tanaka S, Hotta K, Matsuzawa Y, Pratley RE, Tataranni PA 2001 Hypoadiponectinemia in obesity and type 2 diabetes: close association with insulin resistance and hyperinsulinemia. J Clin Endocrinol Metab 86:1930–1935[Abstract/Free Full Text]
  11. Yamauchi T, Kamon J, Waki H, Terauchi Y, Kubota N, Hara K, Mori Y, Ide T, Murakami K, Tsuboyama-Kasaoka N, Ezaki O, Akanuma Y, Gavrilova O, Vinson C, Reitman ML, Kagechika H, Shudo K, Yoda M, Nakano Y, Tobe K, Nagai R, Kimura S, Tomita M, Froguel P, Kadowaki T 2001 The fat-derived hormone adiponectin reverses insulin resistance associated with both lipoatrophy and obesity. Nat Med 7:941–946[CrossRef][Medline]
  12. Miyazaki T, Shimada K, Mokuno H, Daida H 2003 Adipocyte derived plasma protein, adiponectin, is associated with smoking status in patients with coronary artery disease. Heart 89:663
  13. Ouchi N, Kihara S, Arita Y, Maeda K, Kuriyama H, Okamoto Y, Hotta K, Nishida M, Takahashi M, Nakamura T, Yamashita S, Funahashi T, Matsuzawa Y 1999 Novel modulator for endothelial adhesion molecules: adipocyte-derived plasma protein adiponectin. Circulation 100:2473–2476[Abstract/Free Full Text]
  14. Dzielinska Z, Januszewicz A, Wiecek A, Demkow M, Makowiecka-Ciesla M, Prejbisz A, Kadziela J, Mielniczuk R, Florczak E, Janas J, Januszewicz M, Ruzyllo W 2003 Decreased plasma concentration of a novel anti-inflammatory protein—adiponectin—in hypertensive men with coronary artery disease. Thromb Res 110:365–369[CrossRef][Medline]
  15. Ohashi K, Ouchi N, Kihara S, Funahashi T, Nakamura T, Sumitsuji S, Kawamoto T, Matsumoto S, Nagaretani H, Kumada M, Okamoto Y, Nishizawa H, Kishida K, Maeda N, Hiraoka H, Iwashima Y, Ishikawa K, Ohishi M, Katsuya T, Rakugi H, Ogihara T, Matsuzawa Y 2004 Adiponectin I164T mutation is associated with the metabolic syndrome and coronary artery disease. J Am Coll Cardiol 43:1195–1200[Abstract/Free Full Text]
  16. Kumada M, Kihara S, Sumitsuji S, Kawamoto T, Matsumoto S, Ouchi N, Arita Y, Okamoto Y, Shimomura I, Hiraoka H, Nakamura T, Funahashi T, Matsuzawa Y 2003 Association of hypoadiponectinemia with coronary artery disease in men. Arterioscler Thromb Vasc Biol 23:85–89[Abstract/Free Full Text]
  17. Rothenbacher D, Brenner H, Marz W, Koenig W 2005 Adiponectin, risk of coronary heart disease and correlations with cardiovascular risk markers. Eur Heart J 26:1640–1646[Abstract/Free Full Text]
  18. Hotta K, Funahashi T, Arita Y, Takahashi M, Matsuda M, Okamoto Y, Iwahashi H, Kuriyama H, Ouchi N, Maeda K, Nishida M, Kihara S, Sakai N, Nakajima T, Hasegawa K, Muraguchi M, Ohmoto Y, Nakamura T, Yamashita S, Hanafusa T, Matsuzawa Y 2000 Plasma concentrations of a novel, adipose-specific protein, adiponectin, in type 2 diabetic patients. Arterioscler Thromb Vasc Biol 20:1595–1599[Abstract/Free Full Text]
  19. Cesari M, Narkiewicz K, De Toni R, Aldighieri E, Williams CJ, Rossi GP 2007 Heritability of plasma adiponectin levels and body mass index in twins. J Clin Endocrinol Metab 92:3082–3088[Abstract/Free Full Text]
  20. Rossi GP, Taddei S, Virdis A, Cavallin M, Ghiadoni L, Favilla S, Versari D, Sudano I, Pessina AC, Salvetti A 2003 The T-786C and Glu298Asp polymorphisms of the endothelial nitric oxide gene affect the forearm blood flow responses of Caucasian hypertensive patients. J Am Coll Cardiol 41:938–945[Abstract/Free Full Text]
  21. Suwaidi JA, Hamasaki S, Higano ST, Nishimura RA, Holmes Jr DR, Lerman A 2000 Long-term follow-up of patients with mild coronary artery disease and endothelial dysfunction. Circulation 101:948–954[Abstract/Free Full Text]
  22. Frystyk J, Berne C, Berglund L, Jensevik K, Flyvbjerg A, Zethelius B 2007 Serum adiponectin is a predictor of coronary heart disease: a population-based 10-year follow-up study in elderly men. J Clin Endocrinol Metab 92:571–576[Abstract/Free Full Text]
  23. Pischon T, Girman CJ, Hotamisligil GS, Rifai N, Hu FB, Rimm EB 2004 Plasma adiponectin levels and risk of myocardial infarction in men. JAMA 291:1730–1737[Abstract/Free Full Text]
  24. Zoccali C, Mallamaci F, Tripepi G, Benedetto FA, Cutrupi S, Parlongo S, Malatino LS, Bonanno G, Seminara G, Rapisarda F, Fatuzzo P, Buemi M, Nicocia G, Tanaka S, Ouchi N, Kihara S, Funahashi T, Matsuzawa Y 2002 Adiponectin, metabolic risk factors, and cardiovascular events among patients with end-stage renal disease. J Am Soc Nephrol 13:134–141[Abstract/Free Full Text]
  25. Schulze MB, Shai I, Rimm EB, Li T, Rifai N, Hu FB 2005 Adiponectin and future coronary heart disease events among men with type 2 diabetes. Diabetes 54:534–539[Abstract/Free Full Text]
  26. Koenig W, Khuseyinova N, Baumert J, Meisinger C, Lowel H 2006 Serum concentrations of adiponectin and risk of type 2 diabetes mellitus and coronary heart disease in apparently healthy middle-aged men: results from the 18-year follow-up of a large cohort from southern Germany. J Am Coll Cardiol 48:1369–1377[Abstract/Free Full Text]
  27. Cavusoglu E, Ruwende C, Chopra V, Yanamadala S, Eng C, Clark LT, Pinsky DJ, Marmur JD 2006 Adiponectin is an independent predictor of all-cause mortality, cardiac mortality, and myocardial infarction in patients presenting with chest pain. Eur Heart J 27:2300–2309[Abstract/Free Full Text]
  28. Pilz S, Mangge H, Wellnitz B, Seelhorst U, Winkelmann BR, Tiran B, Boehm BO, Marz W 2006 Adiponectin and mortality in patients undergoing coronary angiography. J Clin Endocrinol Metab 91:4277–4286[Abstract/Free Full Text]
  29. Kistorp C, Faber J, Galatius S, Gustafsson F, Frystyk J, Flyvbjerg A, Hildebrandt P 2005 Plasma adiponectin, body mass index, and mortality in patients with chronic heart failure. Circulation 112:1756–1762[Abstract/Free Full Text]
  30. Menon V, Li L, Wang X, Greene T, Balakrishnan V, Madero M, Pereira AA, Beck GJ, Kusek JW, Collins AJ, Levey AS, Sarnak MJ 2006 Adiponectin and mortality in patients with chronic kidney disease. J Am Soc Nephrol 17:2599–2606[Abstract/Free Full Text]
  31. Kanaya AM, Wassel FC, Vittinghoff E, Havel PJ, Cesari M, Nicklas B, Harris T, Newman AB, Satterfield S, Cummings SR 2006 Serum adiponectin and coronary heart disease risk in older Black and White Americans. J Clin Endocrinol Metab 91:5044–5050[Abstract/Free Full Text]
  32. D'Agostino Jr RB 2007 Propensity scores in cardiovascular research. Circulation 115:2340–2343[Free Full Text]
  33. Rossi GP, Cesari M, De Toni R, Zanchetta M, Maiolino G, Pedon L, Ganzaroli C, Maiolino P, Pessina AC 2003 Antibodies to oxidized low-density lipoproteins and angiographically assessed coronary artery disease in white patients. Circulation 108:2467–2472[Abstract/Free Full Text]
  34. Rossi GP, Cesari M, Zanchetta M, Colonna S, Maiolino G, Pedon L, Cavallin M, Maiolino P, Pessina AC 2003 The T-786C endothelial nitric oxide synthase genotype is a novel risk factor for coronary artery disease in Caucasian patients of the GENICA study. J Am Coll Cardiol 41:930–937[Abstract/Free Full Text]
  35. Cesari M, Zanchetta M, Burlina A, Pedon L, Maiolino G, Sticchi D, Pessina AC, Rossi GP 2005 Hyperhomocysteinemia is inversely related with left ventricular ejection fraction and predicts cardiovascular mortality in high-risk coronary artery disease hypertensives. Arterioscler Thromb Vasc Biol 25:115–121[Abstract/Free Full Text]
  36. Mark DB, Nelson CL, Califf RM, Harrell Jr FE, Lee KL, Jones RH, Fortin DF, Stack RS, Glower DD, Smith LR 1994 Continuing evolution of therapy for coronary artery disease. Initial results from the era of coronary angioplasty. Circulation 89:2015–2025[Abstract/Free Full Text]
  37. Califf RM, Armstrong PW, Carver JR, D'Agostino RB, Strauss WE 1996 27th Bethesda Conference: matching the intensity of risk factor management with the hazard for coronary disease events. Task Force 5. Stratification of patients into high, medium and low risk subgroups for purposes of risk factor management. J Am Coll Cardiol 27:1007–1019[CrossRef][Medline]
  38. Tabachnick BJ, Fidell LS 2001 Using multivariate statistics. Boston: Pearson Education Inc.; 548–558
  39. Scott Menard 2001 Applied logistic regression analysis. 2nd ed. London: Sage Publications
  40. Kojima S, Funahashi T, Sakamoto T, Miyamoto S, Soejima H, Hokamaki J, Kajiwara I, Sugiyama S, Yoshimura M, Fujimoto K, Miyao Y, Suefuji H, Kitagawa A, Ouchi N, Kihara S, Matsuzawa Y, Ogawa H 2003 The variation of plasma concentrations of a novel, adipocyte derived protein, adiponectin, in patients with acute myocardial infarction. Heart 89:667
  41. Cesari M, Pessina AC, Zanchetta M, De Toni R, Avogaro A, Pedon L, Dorigatti F, Maiolino G, Rossi GP 2006 Low plasma adiponectin is associated with coronary artery disease but not with hypertension in high-risk nondiabetic patients. J Intern Med 260:474–483[CrossRef][Medline]
  42. Lindsay RS, Resnick HE, Zhu J, Tun ML, Howard BV, Zhang Y, Yeh J, Best LG 2005 Adiponectin and coronary heart disease: the Strong Heart Study. Arterioscler Thromb Vasc Biol 25:e15–e16
  43. Lawlor DA, Davey SG, Ebrahim S, Thompson C, Sattar N 2005 Plasma adiponectin levels are associated with insulin resistance, but do not predict future risk of coronary heart disease in women. J Clin Endocrinol Metab 90:5677–5683[Abstract/Free Full Text]
  44. von Eynatten M, Hamann A, Twardella D, Nawroth PP, Brenner H, Rothenbacher D 2008 Atherogenic dyslipidaemia but not total- and high-molecular weight adiponectin are associated with the prognostic outcome in patients with coronary heart disease. Eur Heart J 29:1307–1315[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 DiabetesHome page
M. von Eynatten, D. Liu, C. Hock, D. Oikonomou, M. Baumann, B. Allolio, G. Korosoglou, M. Morcos, V. Campean, K. Amann, et al.
Urinary Adiponectin Excretion: A Novel Marker for Vascular Damage in Type 2 Diabetes
Diabetes, September 1, 2009; 58(9): 2093 - 2099.
[Abstract] [Full Text] [PDF]

Home page
 British Journal of Diabetes & Vascular DiseaseHome page
K. Karastergiou, V. Mohamed-Ali, M. Jahangiri, and J.-C. Kaski
Review: Adiponectin for prediction of cardiovascular risk?
The British Journal of Diabetes & Vascular Disease, July 1, 2009; 9(4): 150 - 154.
[Abstract] [PDF]

Home page
 Eur Heart JHome page
M. Cesari and G. P. Rossi
Adiponectin and prognostic outcome in patients with coronary artery disease
Eur. Heart J., October 2, 2008; 29(20): 2578 - 2579.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Maiolino, G.
Right arrow Articles by Rossi, G. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Maiolino, G.
Right arrow Articles by Rossi, G. P.
Related Collections
Right arrow Cardiovascular Endocrinology

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals