Incidence of Acute Myocardial Infarction in Patients with Diabetes and Its Association with Mortalit...

Romina Altamirano, Michael Caponigro, Gabriela Carrion, Juan Carlos Zevallos, Grettel Castro, Juan A. González Sánchez, Noël C. Barengo

American Journal of Public Health Research

Incidence of Acute Myocardial Infarction in Patients with Diabetes and Its Association with Mortality and Cardiopulmonary Complications in Puerto Rico

Romina Altamirano1,#, Michael Caponigro2,#, Gabriela Carrion1,#, Juan Carlos Zevallos3, Grettel Castro3, Juan A. González Sánchez4, Noël C. Barengo3,

1Universidad de Las Americas, Quito, Ecuador

2American University of Antigua College of Medicine, Coolidge, Antigua and Barbuda

3Department of Medical and Health Sciences Research, Herbert Wertheim College of Medicine, Florida International University, Miami, United States

4University of Puerto Rico Medical Science Campus, San Juan, Puerto Rico

#Shared first co-authorship

Abstract

Diabetes mellitus (DM) patients are at an increased risk of acute myocardial infarction (AMI). Adequate DM control may reduce in-hospital mortality and cardio-pulmonary complications after an AMI. The objective of this study was to determine whether uncontrolled DM in patients with an incidental AMI was associated with an increased risk of in-hospital mortality and selected cardio-pulmonary complications. A secondary data analysis of the Puerto Rican Cardiovascular Surveillance System during 2007, 2009, and 2011 was conducted. The study included men and women living in Puerto Rico who were hospitalized due to an incidental AMI and had information on HbA1c measurement (n=220). Patients were divided according to their HbA1c levels into two groups (i) <7% (controlled); and (ii) >7% (uncontrolled). Mortality and complications were defined according to ICD-9 codes. Univariate and multivariate logistic regression models were used to test for associations between HbA1c and mortality and cardio-pulmonary outcomes. The model was adjusted for gender, obesity (BMI >30kg/m2), hyperlipidemia, hypertension, chronic obstructive pulmonary disease, smoking, insurance and age. The results revealed no statistically significant association between the controlled and the uncontrolled DM patients and mortality, respectively cardiopulmonary complications (unadjusted OR 2.1; 95% CI: 0.7-6.4; adjusted OR 2.4; 95% CI: 0.5-10.5). The uncontrolled DM group was statistically significantly younger than those with controlled DM (mean age 65.2 vs. 71.1 years; p-value 0.002). There was no statistically significant difference in the prevalence of uncontrolled DM between patients with complications (80%) and those without complication (66%; p=value 0.203). AMI patients with uncontrolled DM did not have an increased risk of mortality and cardiopulmonary complications compared to patients with controlled DM. This needs further evaluation in a larger study population and DM patients with an AMI should have HbA1c measured to estimate their risk of complications.

Cite this article:

  • Romina Altamirano, Michael Caponigro, Gabriela Carrion, Juan Carlos Zevallos, Grettel Castro, Juan A. González Sánchez, Noël C. Barengo. Incidence of Acute Myocardial Infarction in Patients with Diabetes and Its Association with Mortality and Cardiopulmonary Complications in Puerto Rico. American Journal of Public Health Research. Vol. 4, No. 6, 2016, pp 196-201. http://pubs.sciepub.com/ajphr/4/6/1
  • Altamirano, Romina, et al. "Incidence of Acute Myocardial Infarction in Patients with Diabetes and Its Association with Mortality and Cardiopulmonary Complications in Puerto Rico." American Journal of Public Health Research 4.6 (2016): 196-201.
  • Altamirano, R. , Caponigro, M. , Carrion, G. , Zevallos, J. C. , Castro, G. , Sánchez, J. A. G. , & Barengo, N. C. (2016). Incidence of Acute Myocardial Infarction in Patients with Diabetes and Its Association with Mortality and Cardiopulmonary Complications in Puerto Rico. American Journal of Public Health Research, 4(6), 196-201.
  • Altamirano, Romina, Michael Caponigro, Gabriela Carrion, Juan Carlos Zevallos, Grettel Castro, Juan A. González Sánchez, and Noël C. Barengo. "Incidence of Acute Myocardial Infarction in Patients with Diabetes and Its Association with Mortality and Cardiopulmonary Complications in Puerto Rico." American Journal of Public Health Research 4, no. 6 (2016): 196-201.

Import into BibTeX Import into EndNote Import into RefMan Import into RefWorks

1. Introduction

Diabetes mellitus (DM) is a growing epidemic that has affected people from all regions and has become a rising global burden [1]. It has affected the health condition of millions as well as developed into a financial liability internationally. In 2015 there was an estimated 415 million people with DM and over 5 millions deaths due to the disease [1, 2]. This chronic condition seems to affect all systems of the body, but the most common cause of death and disability was due to cardiovascular disease [1]. People with DM have a higher prevalence of coronary artery disease (CAD) leading to an increased risk of myocardial infarction (MI), mortality, and other cardiopulmonary complications. The National Cholesterol Education Program report from the United States and European guidelines consider DM to be a coronary heart disease (CHD) equivalent, thereby elevating it to the highest risk category [1]. This classification was based on coronary mortality for patients with DM without a prior MI and patients without DM who had a prior MI [2]. According to the American Diabetes Association (ADA) [4], the assessment of HbA1c levels in diabetes patients are categorized as good control (<7%), suboptimal control (HbA1c 7% to 9%), or poor control (HbA1c >9%)4. Poor long-term glycemic control is an important risk factor for adverse outcomes in patients with DM [4]. Epidemiologic studies suggest that an increase in HbA1c of 1% is associated with a 20% increased risk in subsequent mortality for patients with coronary disease and concurrent DM [4]. Information regarding associations between DM control and 30 days in-hospital survival are contradictory [5, 8, 9]. Previous studies had controversial results with several limitations that included selection bias [6], small sample size causing low power to detect differences, and focus on special subgroups of the population [12].

The aim of our study was to investigate whether diabetes control was associated with mortality and cardiopulmonary complications such as atrial fibrillation, ventricular tachycardia, ventricular fibrillation, cardiac arrest, pulmonary infarction, pulmonary edema, respiratory failure, and stroke in DM patient with an incident AMI in Puerto Rico.

2. Materials and Methods

2.1. Study Design

We conducted a cross-sectional study of the Puerto Rican Cardiovascular Surveillance database during 2007, 2009 and 2011 [10].

2.2. Study Population

All previous health records were reviewed before including a patient. The definition used in this study as inclusion criteria for T2DM was previous history of DM found in the medical records according to the ADA. After a patient was found to have the diagnosis of T2DM, the record was reviewed to see if HbA1c levels were recorded to be included in the study. Only those with HbA1c measurements taken within the first 24 hours were included. The inclusion criteria for an AMI was that each case was validated using the widely accepted diagnostic definition developed by the World Health Organization (WHO) [4] This definition requires that at least two out of three criteria be present for the confirmation of AMI. This uses information from the patient’s clinical history, serum enzyme elevations, and serial ECG findings. This secondary data analysis was conducted in patients presenting with an incidental (first time) AMI who had HbA1c measures taken within the first 24 hours of hospital admission and reported in the medical chart. We included all patients with ICD-9 codes: 410-414; specifically: New MI, 411: Other acute/sub-acute forms ischemic heart disease, 412: Old MI, 413 Angina pectoris, and 414: Other forms of chronic ischemic heart disease, and excluded 412: patients with previous MI. The records of any previous hospitalizations for cardiovascular events were reviewed when available and if past records indicated similar presentation they were excluded from the study. We excluded patients who did not meet the diagnosis criteria for DM (n=1,138) as stated above, missing key variables of the study, had ECG changes consistent with previous MI, documented history of MI, patients who developed AMI from any procedural events during their hospital stay, and any cardiopulmonary complication the arises after 24 hours of initial hospital arrival. The cardiopulomanary complications that were included in our study was atrial fibrillation, ventricular tachycardia, ventricular fibrillation, cardiac arrest, pulmonary infarction, pulmonary edema, respiratory failure, and stroke. All of these complications were documented in the patient’s medical chart and had to be within 24 hours of admission The study population consisted of Puerto Ricans (a mostly Hispanic population) men and women with T2DM living on the Island who were hospitalized for an incidental AMI. In the database, there were 2,204 patients with an AMI during 2007, 2009, and 2011. A total of 1,470 patients were excluded for not having an incidental AMI or not being diagnosed with T2DM. Out of the remaining 734 patients, 532 were excluded, as they did not have information on HbA1c measurement. Thus, 202 patients comprised our study population [10].

The study population was divided into two groups according to their HbA1C level: controlled DM (HbA1c <7) and uncontrolled DM (HbA1c >7). The covariables that were analyzed in this study were age, gender, hypertension, obesity (defined as BMI>30kg/m2), hyperlipidemia, chronic obstructive pulmonary disease (COPD), current cigarette smokers, and insurance status. These variables were selected because they were related to both HbA1c levels as well as to AMI complications.

All data were obtained from medical documentation of each individual patient that was entered into the database system. Participants have been appropriately de-identified to maintain minimal risk exposure of sensitive information, which was our main concern when obtaining the data for our study.

2.3. Statistical Analysis

STATA version 14.0 was used for the statistical analysis. The medical records of patients discharged with a diagnosis consistent with AMI were reviewed utilizing a standardized electronic data abstraction form, which has been adapted to use in Puerto Rico from the Worcester Heart Failure Study. Data were captured electronically utilizing the REDCapTM web application exclusively designed for research studies that allows secure online management of surveys and databases. For the purpose of present analysis, patients were divided into controlled (HbA1c ≤7%) and uncontrolled DM (, HbA1c >7%) based on admission HbA1c. Categorical variables are presented as percentage and number of patients and the continuous variable are presented as mean ± standard deviation. Baseline characteristics of the two groups were compared using the χ2 test, t student test, and the Fisher exact test as appropriate. Logistic regression models (univariate and multivariate analysis) were used to test the association between the independent variable and the outcome variable, which is in-hospital mortality and cardiopulmonary complications. The Hosmer-Lemeshow test was used to test for the goodness-of-fit of the logistic regression models. The multivariate logistic regression model was adjusted for sex, age, body mass index (BMI), hypertension, hyperlipidemia, chronic obstructive pulmonary disease (COPD), current smokers, insurance status, and age. A p-value <0.05 was considered as statistically significant.

3. Results

Table 1 presents the characteristics of the study participants according to DM control (HbA1C<7% versus HbA1C≥7%). The only characteristic that was statistically significantly different was that the patients in the uncontrolled DM group were younger than those in the group with controlled DM (mean age 65.2 vs. 71.1 years; p-value 0.002). No statistically significant differences were found in the distribution of gender, obesity (BMI>30kg/m2), hyperlipidemia, hypertension, COPD, current smokers, or insurance between the two groups.

The association between HbA1c, mortality and cardio-pulmonary complications among DM patients are presented in Table 2. In the univariate analysis uncontrolled DM patients were twice as likely to die or to have complications than those that had DM controlled (OR 2.1; 95% CI: 0.7-6.4). After adjustment for gender, BMI>30 kg/m2, hyperlipidemia, hypertension, COPD, current smokers, insurance status, and age the OR only marginally changed and remained statistically non-significant (OR 2.4; 95% CI: 0.5-10.3). The OR for an association between obesity and mortality was not statistically significant after multivariate adjustment (OR 0.8; 95% CI 0.2-2.8).

Table 1. Characteristics of diabetes mellitus (DM) patients with first time acute myocardial infarction according to DM control in Puerto Rico during 2007, 2009, and 2011

Table 2. Unadjusted and adjusted odds ratios for the association between HbA1c, mortality and complications among diabetes mellitus patients with first time acute myocardial infarction in Puerto Rico during 2007, 2009, and 2011

Out of 734 DM patients with an AMI, only 202 had their HbA1c levels measured (Table 3). DM patients without HbA1C measurement seemed to have a higher prevalence of mortality and complications (12.8%) compared to measured patients (9.9%). However, these results were not statistically significant (p=0.283). There was a higher prevalence of COPD in the group where HbA1c was measured (9.0%) compared to those with no HbA1c measured (4.7%; p-value=0.031). Furthermore, the prevalence of obesity (BMI>30 kg/m2) was higher in the group who had HbA1c measured (71.9%) compared to those who were not measured (62.8%; p-value=0.037). There was a higher distribution of patients with insurance in the group where HbA1c was measured (76.2%) compared to those who did not have measurement (66.4%; p-value=0.010). There was no statistical significant difference in the rest of the characteristics studied.

Table 3. Characteristics of all previously diabetes mellitus patients with first time acute myocardial infarction according to whether HbA1C was measured or not in Puerto Rico during 2007, 2009, and 2011

4. Discussion

Even though our study found that AMI patients with uncontrolled DM had a two-fold increased risk of mortality and cardiopulmonary complications compared to controlled DM, this risk increase was not statistically significant.

A possible explanation may be a power issue and future studies may adjust for this with a larger study population. This may reveal whether poor DM control is associated with in-hospital mortality and cardio-pulmonary complications. The assessment of chronic glycemic control is highly variable among patients with AMIs and DM. Because much of this variability occurs at the hospital level, the evaluation of DM control could represent an additional quality indicator and an opportunity to advance patient-centered AMI care [4]. Every year over 735,000 Americans citizens have an AMI and people with known DM have an increased risk [6, 7]. In addition, it might be desirable to achieve HbA1c as close to the normal glycemic range as possible [8].

Poor DM control has shown to increase the risk of mortality through several mechanisms [11]. When glucose levels are chronically elevated they damage blood vessels and make them more susceptible to atherosclerosis build up. This causes people to develop atherosclerosis at an early age, leading to hypertension and coronary artery disease [4]. DM patients are also more prone to develop dyslipidemia. Their LDL particles are smaller then non-diabetics making it easier to penetrate vessels. This makes the particles more susceptible to oxidation and damage [4]. Higher HbA1c was associated with poor glycol-metabolic control, older patients, obesity, hypertension, Killip's class>1, tachycardia, initial bundle branch block, atrial fibrillation, and higher mortality during follow-up [5].  These patients with disturbed glucose metabolism had worse early outcomes, characterized by progressive increased rates of in-hospital mortality according to HbA1c. In hospital mortality was 14% higher in patients with HbA1c >6.4% [7]. Diabetes patients with low or high HbA1c are at higher risk of all cause mortality compared wit patients with relatively moderate Hba1c levels [8]. This compared to our study by looking for similar outcomes in terms of mortality and cardiopulmonary complications. Our study also tried to looks for an increased risk of mortality and cardiopulmonary complications based on Hba1c levels. But our odds ratio was not statistically significant and we had a wide confidence interval showing a lack of precision in our study.

Previous studies have shown that elevated glucose levels increase the risk of mortality after an AMI in both diabetic and non-diabetic patients [17]. Both DM and non-DM patients had significantly higher plasma glucose levels, while HbA1c-adjusted levels were not associated with post-AMI prognosis [19]. Our study did not analyze admission glucose but rather HbA1c levels within 24 hours of admission. We did not find any significant evidence to support the claim that DM patients with AMI had increased mortality and complications. We also did not evaluate medical records past the initial 24 hours of hospital stay to evaluate long-term prognosis for patients with controlled and uncontrolled HbA1c levels. A secondary data analysis of The National Cardiovascular Data Registry (NCDR) observed no association between HbA1C and mortality in patients with diabetes presenting with AMI [18]. In a multivariable regression, they observed no association between low or high HbA1c [18]. Similar to their study we did not find any statistically significant association between HbA1c and mortality and complications. As their study, our study had similar limitations DM patients with AMI did not have information on HbA1c measurement (59%). In another study Hba1c and blood glucose was measured within 3 hours of admission of an AMI to evaluate for myocardial perfusion abnormalities [15]. Patients with the highest HbA1c (>8.5%) showed the highest morality, double-vessel, and triple-vessel disease compared to lower HbA1c levels. We had similar outcomes in terms of mortality and trying to relate it to chronic glycemic control [19]. But this study had a follow up at four weeks and did not accurately measure the immediate outcome for elevated HbA1c levels as our study did. We also did not find any statistically significant association between HbA1c and mortality as this study did at the 4 week follow up. To detect myocardial perfusion defects and relate it to mortality a coronary angiography performed within 24 hours would have been a better predictor to relate it to plasma glucose, HbA1c, and mortality. Some studies used a different definition of DM control [14, 15] as our study used HbA1c above 7% as uncontrolled DM, while their study used 8.5%.

The limitations of our study were small sample size and limited number of outcomes. Out of a potential 2,204 patients with an AMI, only 202 had DM and HbA1c measured within 24 hours of hospital admission. This affected the power of our study to detect an association. Since there was a low study population we completed a non-responder analysis based on whether HbA1c was measured (Table three). The results indeed showed that the non-measured HbA1c patients had different profile than those with measured HbA1c. The patients with HbA1c measured were statistically significantly different in terms of BMI>30 kg/m2 (p-value 0.037), have COPD (p-value 0.031), and have insurance (p-value 0.010) In addition to small sample size we also had a limited number of outcomes. Out of the 202 patients in our study population, only 20 cases had complications. If a greater number of complications were found in a study population it could possibly change future management strategies for DM patients with AMI. Despite not measuring glucose levels, hyperglycemia could also affect mortality both acutely and chronically. Follow up HbA1c levels should have been checked at monthly or yearly intervals to look for a possible association with mortality, as it might be a predictor of future mortality rather than short-term prognosis. Finally, we were not able to assess whether hypertension treatment or adecuate hypertension control was associated with mortality due to a lack of this particular information. Even though we did not find a statistically significantly increased risk of being classified as hypertensive, we are aware that due to the above mentioned limitation, no conclusions can be made in regard to a possible association between uncontrolled hypertension and mortality

5. Conclusion

In conclusion, our findings suggest that AMI patients with uncontrolled DM did not have an increased risk of mortality and cardiopulmonary complications compared to controlled DM. However, this needs further evaluation on a larger study population in order to give more accurate results evaluate DM patients. We recommend that all DM patients with an AMI should have HbA1c measured to estimate their risk, manage patient’s glycemic control better, and possibly prevent future complications from happening. In addition, other factors such as tobacco consumption, hyperlipidemia, obesity, and hypertension are important factors to be assessed and controlled for in order to improve survival in AIM.”

Acknowledgements

This project was supported with Grant 5U01CK000274 from the Centers for Disease Control and Prevention and Grants 5S21MD000242 and 5S21MD000138, from National Center for Minority Health and Health Disparities, National Institutes of Health.

Statement of Competing Interests

The authors declare no competing interests.

List of Abbreviations

DM Diabetes mellitus, CAD coronary artery disease, MI myocardial infarction, CHD coronary heart disease, ADA American Diabetes Association, AMI acute myocardial infarction , T2DM diabetes mellitus type 2, WHO World Health Organization, ECG electrocardiogram, ICD international classification of diseases, BMI body mass index, COPD chronic obstructive pulmonary disease, REDCapTM (), OR odds ratio, HbA1c Glycosylated hemoglobin, A1C Glycosylated hemoglobin, IRB Institutional Review Board, FBG fasting blood sugar, GV variability glycemic, NSTEMI Non ST- segment elevation myocardial infarction, MACE Major adverse cardiovascular events, MAGE mean amplitude of glycemic excursions, IC confidence interval, MeSH medical subject headings, IDF International Diabetes Federation, ACS acute coronary syndrome, APG admission plasma glucose.

References

[1]  International Diabetes Federation, IDF Diabetes Atlas 7th Edition, International Diabetes Federation, 2015. [E-book] Available: www.diabetesatlas.org.
In article      
 
[2]  Adler, A.I., Neil, H.A., Manley, S.E., Holman, R.R., Turner, R.C, “Hyperglycemia and hyperinsulinemia at diagnosis of diabetes and their association with subsequent cardiovascular disease in the United Kingdom prospective diabetes study (UKPDS 47),” Am Heart J, 138 (5). S353-S359. November 1999.
In article      View Article
 
[3]  Al-Nozha, O., Mojadadi, M., Mosaad, M., El-Bab, M.F, “Assessment of coronary heart diseases in diabetics in al-Madinah al-Munawarah,” International Journal of General Medicine. 2012 (5). 143-149. February 2012.
In article      
 
[4]  ADA. Diabetes Management Guidelines. ADA, 2016. http://www.ndei.org/ADA-diabetes-management-guidelines-diagnosis-A1C-testing.aspx.html.
In article      
 
[5]  Stolker, J., Sun, D., Conaway, D., Jones, P., Masoudi F., Peterson, P, “Importance of Measuring Glycosylated Hemoglobin in Patients With Myocardial Infarction and Known Diabetes Mellitus,” The American Journal of Cardiology. 105(8). 1090-1094. April 2010.
In article      View Article  PubMed
 
[6]  Lee, S., Cho, S., Jeong, M., Kim, Y., Kim. C., Cho, M, “Hypoglycemia at Admission in Patients With Acute Myocardial Infarction Predicts a Higher 30-Day Mortality in Patients With Poorly Controlled Type 2 Diabetes Than in Well-Controlled Patients,” Diabetes Care. 37(8). 2366-2373. August 2014.
In article      View Article  PubMed
 
[7]  CDC. Heart Disease Facts & Statistic. CDC. 2016 http://www.cdc.gov/HeartDisease/facts.htm.
In article      
 
[8]  Blasco, M., Sanjuan, R., Palacios, L., Huerta, R., Carratala, A., Nunez, J, “Prognostic value of admission glycated haemoglobin in unknown diabetic patients with acute myocardial infarction,” European Heart Journal: Acute Cardiovascular Care. 3(4). 347-353. December 2014.
In article      View Article  PubMed
 
[9]  Arnold, L., Wang, Z, “The HbA1c and All-Cause Mortality Relationship in Patients with Type 2 Diabetes is J-Shaped: A Meta-Analysis of Observational Studies,” Journal of the society for biomedical diabetes research: The Review of Diabetic Studies. 11(2). 138-152. August 2014.
In article      View Article  PubMed
 
[10]  Puerto Rican Cardiovascular Surveillance System.
In article      
 
[11]  Hadjadj, S., Coisne, D., Mauco, G., Ragot. S., Duengler, F., Sosner, P, “Prognostic value of admission plasma glucose and HbA1c in acute myocardial infarction,” Diabetic Medicine. 21(4). 305-310. March 2004.
In article      View Article  PubMed
 
[12]  Chan, C., Li, R., Chan, J., Zhang, Q., Chan, C., Dong, M, “ The Value of Admission HbA1c Level in Diabetic Patients With Acute Coronary Syndrome,” Clinical Cardiology. 34(8). 507-512. June 2011.
In article      View Article  PubMed
 
[13]  Britton, K., Aggarwal, V., Chen, A., Alexander, K., Amsterdam, E., Fraulo, E, “No association between hemoglobin A1c and in-hospital mortality in patients with diabetes and acute myocardial infarction,” American Heart Journal. 161(4). 657-663. April 2011.
In article      View Article  PubMed
 
[14]  Cakmak, M., Cakmak, N., Cetemen, S., Tanriverdi, H., Enc, Y., Teskin, O, “The value of admission glycosylated hemoglobin level in patients with acute myocardial infarction,” Canadian Journal of Cardiology. 24(5). 375-378. May 2008.
In article      View Article
 
[15]  Timmer, J., Ottervanger J.P., Bilo, H.J., Dambrink, J.H, “Prognostic value of admission glucose and glycosylated haemoglobin levels in acute coronary syndromes,” QJM. 99(4). 237-243. April 2006.
In article      View Article  PubMed
 
[16]  Strojek, K., Raz, I., Jermendy, G., Gitt, A., Liu, R., Zhang, Q, “Factors Associated With Cardiovascular Events in Patients With Type 2 Diabetes and Acute Myocardial Infarction,” The Journal of Clinical Endocrinology & Metabolism. 101(1). 243-253. January 2016.
In article      View Article  PubMed
 
[17]  Ahn, J., Hong, T., Park, J., Lee, H., Oh, J., Choi, J, “Clinical influence of early follow-up glycosylated hemoglobin levels on cardiovascular outcomes in diabetic patients with ST-segment elevation myocardial infarction after coronary reperfusion,” Coronary Artery Disease. 26(7). 555-561. November 2015.
In article      View Article  PubMed
 
[18]  Britton, K., Aggarwal, V., Chen, A., Alexander, K., Amsterdam, E., Fraulo, E, “No association between hemoglobin A1c and in-hospital mortality in patients with diabetes and acute myocardial infarction,” American Heart Journal. 161(4). 657-663. April 2011.
In article      View Article  PubMed
 
[19]  Vujosevic , S., Radojevic, N., Belada, N, “Influence of admission glucose profile and hemoglobin A1c on complications of acute myocardial infarction in diabetic patients,” European review for medical and pharmacological sciences. 9(17). 1252-1257. May 2013.
In article      
 
  • CiteULikeCiteULike
  • MendeleyMendeley
  • StumbleUponStumbleUpon
  • Add to DeliciousDelicious
  • FacebookFacebook
  • TwitterTwitter
  • LinkedInLinkedIn