66 REUNION CIENTIFICA DE LA ASOCIACION AMERICANA DE DIABETES.
Washington, USA, Junio 2006.

Resúmenes de comunicaciones.

Glucosa en ayunas alterada y enfermedad cardiovascular: El Estudio del Corazón de Framingham.
Levitsky, J y col. Framingham, Boston, MA.

Pre-diabetic hyperglycemia confers an elevated cardiovascular disease (CVD) risk. Whether the 2003 American Diabetes Association (ADA) definition of impaired fasting glucose (IFG) confers an increased risk of CVD in a community-based sample is unknown.
Participants of the Framingham Offspring cohort free of CVD (n=4058, 53% women, mean age 49) were followed from 1983-2002 and were categorized according to the 1997 ADA IFG definition (110-125 mg/dl) and the 2003 ADA IFG definition (100-125 mg/dl). Endpoints were coronary heart disease (CHD; MI/angina/CHD death) and CVD (CHD/stroke/TIA/intermittent claudication/CHF/CVD death). Covariates included traditional CVD risk factors. Logistic regression was used to calculate event risk in each glycemic category.
Among women, 78 CHD/128 CVD events occurred; 213 CHD/295 CVD events occurred among men. Multivariable-adjusted odds ratios (OR) with 95% CI for CHD and CVD events by category are presented (Table).
Neither the 1997 nor the 2003 definition of IFG identified men at increased risk of CHD or CVD. The 2003 IFG definition was associated with increased risk for CHD in women. However, risk was driven by events in the 110-125 mg/dl group, which approached the risk associated with diabetes. Confirmation in other samples is warranted.

6-OR] Metabolic Syndrome and Its Single Traits as Risk Factors for Atherosclerotic Vascular Disease in Type 2 Diabetes: The Diabetes in Germany (DIG) Study
Koehler, C. y col. Dresden, Alemania.

There is still controversy on whether the overall Metabolic Syndrome (MS) carries more predictive power for atherosclerotic vascular disease (AVD) then its single traits in type 2 diabetes (T2DM). With two additional traits to meet the definition of the MS, six combinations are possible in patients with T2DM that may carry different risk for AVD. Our aim was to investigate the association between MS, its single traits and their combinations with AVD. DIG is a population based observational study in patients with T2DM.
Inclusion criteria were T2DM and age 35-80 years. Exclusions criteria: major cardiovascular event <3 months before entry, heart failure NYHA IV, macroproteinuria. The MS was defined by NCEPIII. AVD was defined as myocardial infarction and/or coronary revascularization and/or stroke.
We included 4,020 T2DM patients with an average duration of T2DM of 8.4 years and HbA1C of 7.0%. Overall 74.4% of the DIG population had the MS without a significant sex difference and 16% had AVD: 20.4% males and 11.1% females. The highest odds ratio (OR) for AVD for a single trait was for hypertension: 4.2(2.1-8.4) in men and 7.7(1.9-31.4) in females. All combinations with hypertension had ORs of more than 5.6 and were significantly higher in women. We used stepwise regression analysis to identify the independent risk factors of AVD in T2DM in two models. In both we included established risk factors: age, sex, smoking habits and LDL-cholesterol (LDLC) along with the single traits of MS in the first and overall MS in the second model. Besides age, male sex, smoking habit and LDLC, hypertension was an independent risk factor for AVD in T2DM in model 1 and overall MS in model 2.
Our data suggest that the MS comprises a very heterogeneous cluster of traits whose association with AVD shows a great variation (ORs 0.87-10.9). The combination of T2DM and hypertension is by far the most hazardous.

[8-OR] Visceral Adiposity and Triglyceride-Rich Lipoprotein Cholesterol Predict Coronary Atherosclerosis in Type 2 Diabetes
Mazzone, T. y col. Chigago, Boston, San Antonio.

CHICAGO is a multicenter, randomized 18-mo study comparing the effects of pioglitazone vs glimepiride on measures of atherosclerotic disease (Carotid Intima-Media Thickness and Electron Beam Tomography of the coronary arteries, EBT) and biochemical indices of cardiovascular risk. We now report the relationship between baseline characteristics and coronary artery calcium (CAC) in 439 ethnically diverse subjects with type 2 diabetes (T2D). A multivariate analysis including traditional cardiac risk factors confirmed that age, SBP, gender, race/ethnicity, calculated LDL and triglyceride levels were significant predictors of CAC. In a multivariate analysis with over 30 traditional and emerging cardiac risk factors (including apolipoproteins, inflammatory and coagulation markers, and visceral and total abdominal fat), age, SBP, gender, and race/ethnicity remained significant predictors of CAC. Only the ratio of visceral adipose tissue to total abdominal adipose tissue (p<0.0001), and the level of cholesterol in triglyceride-rich lipoproteins (VLDL and IDL, p=0.0008) emerged as additional significant predictors of coronary atherosclerosis. The impact of race/ethnicity on CAC was highly significant (p=0.006). Compared to Caucasians, Asians had higher (+260%), and African-Americans and Hispanics had lower (-63 and -25%, respectively) CAC scores. Further, examination of males and females separately showed that inclusion of the intra-abdominal adiposity measure eliminated the influence of race/ethnicity on CAC in females. These results identify the amount of visceral abdominal fat and the cholesterol content of triglyceride-rich lipoproteins as important pathophysiologic and clinical markers for advanced coronary atherosclerosis in subjects with T2D. The impact of race/ethnicity on coronary atherosclerosis in diabetic females is completely accounted for by differences in visceral adiposity.

[24-OR] Renal Function Decline (RFD) in Type 2 Diabetes Mellitus (T2DM) in the Absence of Proteinuria
Nieczas, M y col. Boston, MA.

Our aim was to determine the frequency of RFD (GFR<60 ml/min) in the absence of proteinuria in a large population attending a diabetes clinic and to identify urinary markers associated with RFD.
In 2003-05, 5,627 patients with T2DM aged 35 to 69 attended the Joslin Clinic and had multiple determinations of the urinary albumin to creatinine ratio (ACR) and serum creatinine. These were used to classify patients according to level of albumin excretion and estimate the frequency of RFD and the rate of GFR loss/yr. RFD was present in 11% of 3623 with normoalbuminuria, 21% of 1542 with microalbuminuria and 58% of 462 with proteinuria. Its frequency is only 3.6% in the U.S. population without diabetes in this age range. To identify urinary markers associated with RFD, we compared T2DM patients with RFD selected from each level of albumin excretion with T2DM controls with normal and stable GFR selected from patients with normoalbuminuria (see table). Results were similar after adjustment for urinary creatinine concentrations. Mean HbA1c (in %) during the 5 years preceding the study was 7.5±1.2, 7.8±1.3, 8.0±1.3 and 8.8±1.5, respectively, in the four study groups (p<0.001 for the last).
In conclusion, renal function decline occurs in a significant proportion of T2DM patients without evidence of an impaired glomerular filtration barrier (no proteinuria). These patients do have elevated urinary excretion of TGFbeta - inducible proteins: IL-8 and MCP-1.

[25-OR] Identifying Patients with Type 2 Diabetes at High Risk of Microalbuminuria: The DEMAND Study
Rossi, M y col. Italia.

Aim:To evaluate to what extent risk factors and their interaction increased the likelihood of microalbuminuria (MA) in individuals with DM2.
Methods: Fifty-five Italian diabetes outpatient clinics selected 1874 patients aged 18-80 years, using a random sampling procedure. Albumine/creatinine ratio (A/C) was measured on a morning spot urine sample. A tree-based regression technique (RECPAM) was used to investigate interaction between correlates of MA.
Results: Overall,1677 patients were included in the analyses, while 197 were excluded for urinary infections and 57 for macroalbuminuria. Overall, the prevalence of MA (A/C=30-299 mg/g) was 19.4%. RECPAM led to the identification of 7 classes with a marked difference in the risk of MA. Patients with HbA1c <7%, a waist circumference (WC) ≤102 cm and non-smokers had the lowest prevalence of MA (7.5%), and represented the reference class. The likelihood of MA was 13 times higher in patients with retinopathy, a WC>98 cm, and HbA1c >8% and 5 times higher in patients with the same characteristics but HbA1c levels between 7% and 8%. Individuals without retinopathy with HbA1c levels ≥7.0% also showed a three-fold risk of MA. Among individuals with HbA1c <7.0%, a WC >102 cm or smoking were also associated with a higher risk of MA. A final multivariate analysis with RECPAM classes forced in retained age, systolic blood pressure, and HDL cholesterol as independent correlates of MA.
Conclusion: The likelihood of MA is strongly related to the interaction between metabolic control, retinopathy, abdominal obesity, and smoking. Elevated blood pressure and low HDL-cholesterol levels further increase the risk of MA.

[32-OR] A Randomized, Controlled Trial of a Low-Glycemic Index vs. a Low-Carbohydrate, Ketogenic Diet for Type 2 Diabetes
Westeman, E y col. Durham, NC

While many current guidelines recommend a high-carbohydrate, low-fat, low-calorie diet for type 2 diabetes, recent studies suggest that low carbohydrate diets also lead to improvement in glycemic control. The purpose of this study was to compare a low-calorie, low-glycemic index diet (LGID) to a low-carbohydrate ketogenic diet (LCKD) for overweight and obese type 2 diabetics.
Community volunteers were recruited for this outpatient, randomized, clinical trial. Inclusion criteria included HgbA1c > 6.0, BMI from 27-50, age 18-65 years, diagnosis of type 2 diabetes mellitus for at least one year, onset of diabetes after age 15 years, no history of diabetic ketoacidosis, and motivation to lose weight. Individuals with an unstable or serious medical condition, liver or kidney disease were excluded. Subjects were randomized to instruction in either an LGID or an LCKD. LGID subjects were given an explicit calorie restriction; LCKD subjects were told to limit carbohydrate intake to < 20 grams/day. Both groups received a multivitamin, and a vanadyl sulfate/chromium supplement. Subjects returned every week for 3 months, then every 2 weeks to an outpatient research clinic.
The mean (SD) age of participants was 52.0 (7.5) years, 80% were female, 51% were Caucasian. After 3 months, the mean HgbA1c for the LCKD group (n=27) reduced from 8.7% (1.8) to 7.1% (1.2), and for the LGID group (n=30) from 8.0% (1.8) to 7.5% (1.7), p=0.01 between group comparison. Weight loss was 8.3 kg for the LCKD group, and 5.6 kg for the LGID group, p=0.12 between group comparison. Using an algorithm to manage medications, 79% of the LCKD group had their medications reduced or eliminated, compared to 66% of the LGID group.
Both lifestyle interventions led to medication reduction, weight loss, and improved glycemic control, but the LCKD led to a greater reduction in HgbA1c after 3 months. Because medication reduction is required to avoid hypoglycemia, glucose self-monitoring and medical supervision are recommended if these approaches are used to treat diabetes mellitus.

37-OR] Effect of Counseling with Continuous Glucose Monitoring on Physical Activity Behaviors
Allen, N y col. Worcerster, Amherst, MA.

Despite the known benefits, up to 60% of people with diabetes do not participate in regular physical activity. The purpose of this study was to test the feasibility and implementation of an intervention using Continuous Glucose Monitoring System® (CGMS) technology to change physical activity behavior in individuals with type 2 diabetes.
Individuals were randomized to CGMS counseling intervention (IG) or to standard counseling (CG). Both groups received 90 minutes of diabetes education including meal and walking plans with a follow up phone call at 4 weeks. IG counseling included feedback on a CGMS graph from the participant and from a role model depicting physical activity related reductions in glucose levels. Also, a discussion of physical activity benefits/barriers and goal setting was included. Outcomes were recorded at week one and eight (see table).
Fifty-two subjects (mean age 57 + 14 years, 52% female, 90% white and 8 + 7 years of diabetes) participated. At 8 weeks, IG participants had a higher self-efficacy score indicating more confidence in their ability to stick to a physical activity program. Systolic BP was significantly decreased in the IG. Physical activity counts as measured by Actigraph® accelerometer were not significantly different from baseline in either group. However, counts for IG increased while counts for CG decreased. A1c improved significantly in the IG only.
These data suggest that using CGMS feedback in a physical activity counseling intervention targeting individuals with type 2 diabetes is feasible and may improve risk factors for diabetes-related complications.

• indicates p< 0.05

[38-OR] Program To Enhance Diabetes Education Resources in Primary Care Setting Improves Measures of Care
Harris, S, C. y col. Boston, MA.

Primary care providers often lack resources for diabetes patient education. With a goal of improving clinical and behavioral outcomes, we developed a program to deliver diabetes staff training and patient education materials to PCP offices in Detroit and NYC.
PCPs and office staff had access to CDEs who provided diabetes training and patient education flip charts, booklets and handouts in English and Spanish. Modifications in year 2 included streamlined enrollment, more cost-effective materials and no call center. 800 providers enrolled in year 1 and 820 in year 2. 1,167 office staff received the core diabetes training.
Clinical measures were collected from patient records. Self-reported data were collected from patients. A representative sample of providers and office staff were surveyed, as well as a control provider group.
Mean A1C improved from 7.61% to 7.14% over six-months (N=171, p=0.0001). Mean BP decreased from 133.7 to 128.0 systolic and 78.8 to 76.4 diastolic (p=0.001). BP <140/90 increased from 53% to 65% (p=0.001). A1C change significantly correlated with diastolic BP change (p=0.01). Mean LDL cholesterol decreased from 107.4 mg/dl to 94mg/dl (p=0.001).
Mean A1C for insulin patients declined from 8.98% to 7.49% (self-reported, N=724). Increased months of practice participation was correlated with this decline. PCP confidence in their office providing better educational support significantly increased compared to non-participants and the pre-assessment of participants. PCP comfort level with prescribing insulin for patients not reaching target glucose control on oral medications was significantly higher among participating providers compared to those who did not participate (p=0.006). After diabetes training, office staff knowledge about diabetes significantly increased (p=<0.01).
Conclusion: We demonstrated that a multifaceted program focused on enhancing the management of diabetes in community-based primary care setting is highly effective in improving patient outcomes and clinician confidence.

78-OR] Socioeconomic Deprivation and Trends in Diabetes Prevalence and Detection among U.S. Adults
Becles, G y col. Atlanta, GA.

Diabetes has become increasingly prevalent in the U.S. and is more common among disadvantaged than among more affluent persons. We examined secular trends in diabetes prevalence and detection between 1976 and 2002 according to socioeconomic status among the U.S. population aged 20-74. Using data from three consecutive NHANES surveys (1976-80, 1988-94, and 1999-2002), we examined the trends in diagnosed, undiagnosed, and total diabetes, and the proportion of total cases that were diagnosed, according to levels of poverty income ratio (PIR). All analyses controlled for age, sex, and race/ethnicity distributions of the U.S. census 2000 population. Between 1976 and 2004, prevalence of total and diagnosed diabetes was highest and tended to increase the most among the lowest PIR tertile. Prevalence of total diabetes increased from 7.1% to 11.4% among the lowest PIR tertile, while diagnosed diabetes increased from 4.3% to 7.9%. By comparison, there was little change in total diabetes (5.6% to 6.0%) or diagnosed diabetes (3.4% to 4.1%) among the highest PIR tertile. There was no change in prevalence of undiagnosed diabetes among any of the PIR tertiles. The proportion of total cases that were diagnosed increased among all groups, from 60-61% in 1976-80 to 69-71% in 1999-2002 but there was no difference in this proportion across PIR tertiles. Increasing prevalence of total diabetes was primarily due to increases in diagnosed diabetes; with the most disadvantaged groups disproportionately affected. These findings indicate a need for strengthened diabetes prevention efforts among persons of socioeconomic disadvantage.

[81-OR] The Menstrual Cycle as a Risk Factor for Diabetes in Women
Young Oh, J y col.  Seoul, Corea.

Menstrual irregularity is thought to be related to insulin resistance, type 2 diabetes, or cardiovascular disease in women.
To evaluate whether the menstrual irregularity is related to the risk for type 2 diabetes in women. Questionnaires on menstrual history (cycle length at age 20, 30, 40, and now, age of menarche, and menstrual duration), lifestyle factors, and family history of diabetes were completed in 117 women with type 2 diabetes and 209 age–matched control women. Women with diabetes were more obese, had higher levels of systolic blood pressure and fasting plasma glucose, and less favorable lipid profiles (all Ps <0.05). Menstrual cycle length at present, at age 30 and 40 were significantly longer in women with diabetes than control (33 vs. 30 d; 32 vs. 30 d; 31 vs. 29 d, P < 0.05, respectively). When we stratified all subjects to four groups according to their menstrual cycle length (≤ 25, 26~31, 32~39, and ≥ 40 days), the frequency of diabetes was significantly higher in women with over 40 days of their cycle length compared to those with normal cycle length (63 vs. 31 %, P < 0.01). In multiple logistic regression analysis, long menstrual cycle length was significantly associated with diabetes after adjusting for age, BMI, family history of diabetes, and various lifestyle factors.
In conclusion, oligomenorrhea might be an important risk factor for type 2 diabetes in women, but prospective study will be required to assess the causality.

• Age, BMI, family history of diabetes (first degree relatives), and lifestyle factors (exercise, alcohol, and smoking) were adjusted in this analysis.
  

[84-OR] Family History of Coronary Heart Disease Is Associated with Risk of Type 2 Diabetes in Those with a Parental History of Diabetes
Yeung, E y col. Balitimore, MD, Minneapolis, MN

Type 2 diabetes and cardiovascular disease share common risk factors, including those that may be inherited from one generation to the next. Thus, the purpose of the study was to test whether familial risk of coronary heart disease (CHD) as measured by the family CHD risk score (CHD-FRS) is predictive of diabetes development.
Methods: The study analyzed data for 11,297 participants from the Atherosclerosis Risk in Communities Study. Middle-aged white and black participants free of diabetes at baseline (1987-1989) were followed prospectively with exams at 3 year intervals to determine diabetes status. CHD-FRS, which compares the expected versus observed CHD occurrence in each participant's family, was categorized into 3 classes (low, medium, and high). Cox regression was used to determine the risk of incident diabetes by these 3 levels and adjusting for confounders.
Results: The median follow up time was 9 years (with 1,302 incident cases). Familial CHD risk was significantly associated with increased diabetes risk among those with a positive parental diabetes history (n=2,661). Compared to their counterparts with low risk, those with the highest familial CHD risk had a hazard ratio (HR), adjusted for age, race, and sex, of 1.76 (95%CI: 1.28-2.42). Additionally adjusting for the traditional diabetes lifestyle risk factors did not eliminate this excess risk (HR=1.67; 95%CI:1.22-2.30). Compared with the low CHD-FRS group, the medium risk group also had increased risk (HR=1.28; 95%CI: 1.02-1.59). On the other hand, among those with a negative parental diabetes history, no significant increase in risk was found even with the highest level of CHD-FRS (adjusted HR=1.15; 95%CI: 0.88-1.51).
Conclusions: Familial CHD risk confers significant additional risk for type 2 diabetes among those that already have a family history of diabetes. The double assault of both CHD and diabetes family histories highlights the need for vigorous, early action to prevent type 2 diabetes in individuals with family histories of both conditions.

[99-OR] MRSA Isolation from Diabetic Foot Ulcers -Does It Affect Healing and Is the Problem Getting Worse?
Game, F y col. Nottingham. Gran Bretaña.

The isolation of methicillin resistant Staphylococcus aureus (MRSA) has increased in the last 25 years, and many believe that it is more virulent than its methicillin sensitive counterpart. In order to determine both the changing prevalence of MRSA and any relationship to healing of diabetic foot ulcers, we surveyed cohorts of consecutive patients and compared outcomes in those who were and were not MRSA positive. The first cohort (98 patients with 150 ulcers) comprised all those attending in the calendar months, November 2001 and February 2002. Each person was sampled only once. The second cohort (85 patients with 110 ulcers) was from four weeks in July-August 2004. Vigorous surface sampling for MRSA was performed with a saline-moistened swab. Outcomes at 6 months were determined by a blinded clinician using the detailed database maintained in the clinic, supplemented when necessary by case notes review. MRSA was isolated from 15 ulcers (10.0%) in 10 patients (10.2%) in the 2001/2002 cohort, and from 19 ulcers (17.3%) in 17 patients (20.0%) in 2004. When those positive for MRSA were compared with those which were negative, there were no differences in the proportion of ulcers healed at 6 months in either the 2001/2 (Pearson Chi-Square 0.181, p= 0.774) or the 2004 survey (0.217, p=0.309). There were no differences in time to healing in either cohort (Mann-Whitney U 80.5 p=0.142, and U 76.5, p=0.233, respectively). Similarly, there were no differences in the numbers of major amputations (Chi-Square 0.14, p=0.579, and 4.996, p=0.025) or deaths (Chi-Square 3.6, p=0.191, and 0.536, p=0.464). When the 2 cohorts were combined, no differences were observed between those from whom MRSA was and was not isolated, with respect to healing (Chi-Square 0.845, p=0.428), time to healing (U 474 p=0.615), amputation (Chi-Square 3.34, p=0.087) or death (Chi-Square 0.314, p=0.483). It is concluded that although colonisation of diabetic foot ulcers with MRSA is increasing, it is not associated with worse outcome.

[101-OR] Persistent Tumor Necrosis Factor (TNF) α Synthesis during Cutaneous Inflammation in Persons with Diabetes
Evans, B y col. Bridgetown, Barbados. Gran Bretaña.

Impaired wound healing makes a significant contribution to morbidity, mortality and costs associated with diabetes. However, our understanding of the mechanisms behind diabetic wound healing is limited. Here we have used cantharidin-induced skin blisters as a novel way to probe the evolution of the human cutaneous inflammatory response in persons with uncomplicated type 2 diabetes (n=16) versus healthy controls (n=31). Blister fluid from duplicate forearm skin blisters was analyzed at 16h and 40h, in the acute and resolving phases respectively, for proinflammatory TNFα and wound-healing transforming growth factor (TGF)β cytokines. In healthy persons, skin blisters at 16h contained high levels of TNFα compared to those with diabetes (771.4±141.6 pg/ml vs. 199.6±95.8; p=0.01). In contrast, at 40 hours TNFα was now substantially reduced in healthy persons, yet in diabetes substantially increased (378.0±153.5 vs. 806.3±201.8; p=0.09). The difference in TNFα over time by diabetes status was significant when tested as an interaction (p=0.003). TGFβ, although undetectable in both groups at 16h, was strongly induced at 40h in normals (10.3±3.0 ng/ml) but remained undetectable in diabetes. The balance of regulatory cytokines has thus been reversed in diabetes, with persistence of TNFα but loss of TGFβ in the resolving phase. By examining the phenotype of leukocytes in blister fluid, we were also able to pinpoint a deficiency at 40h in diabetes (p=0.002) in the generation of a wound-healing sub-population of CD163+ macrophages, that is known to be suppressed in the presence of TNFα. In conclusion, this pilot study has identified molecular deficiencies in the cytokine balance and regulatory macrophage populations controlling inflammatory wound healing in diabetes. Our results suggest that anti-TNF therapy may be a treatment option in diabetic foot.

[102-OR] Optimal Cut-Off Point of Ankle-Brachial Index Is 1.0 in Type 2 Diabetic Patients
Jinnouchi, H y col. Kumamoto, Japón.

To clarify optimal cut-off point of Ankle-Brachial Index (ABI) in Type2 diabetes, we surveyed 2491 Type 2 diabetic patients using FormPWV/ABI (Colin Medical Technology Corporation). In these, 206 diabetic patients were suspected to have arteriosclerosis in their lower extremities. Severity of arteriosclerosis was determined by magnetic resonance angiography (Siemens Magnetom Symphony SY Maestro Class 1.5 T). Severity of arteriosclerosis was classified to five grade (none, mild, mediate, severe and complete). Histogram showed that the ABI cut-off point of 0.9 overlook the half of patients with mediate arteriosclerosis. Cumulative logistic probability plot indicated that the ABI cut-off point of 0.9 means severe arteriosclerosis. To prevent the progression of arteriosclerosis from early stage in primary care before macrovascular event, optimal cut-off point of ABI would be 1.0.

[104-OR] Comparison of Three Classification Systems for Diabetic Foot Ulcers in a Tanzanian Population
Abbas, Z. y col. Tanzania, Gran Bretaña.

We have examined the spectrum of foot ulcers in a consecutive series presenting at Muhimbili National Hospital, Dar es Salaam, Tanzania, and have used three different classification systems to seek associations between baseline status and outcome. 326 people (mean age 54.5±11.3 SD; 67.2%M), with 479 ulcers, presented between 1st January 2003 and 30th September 2005. 74 (22.7%) people with 104 (21.7%) ulcers were lost to follow up and were excluded from analysis. Mean age in the remaining 252 patients (67.1%M) was 54.7±11.5 years. Median duration of follow-up was 36 (range 0-973) days. 230 of 375 ulcers (61.3%) had healed by 30th September 2005, while 69 (18.4%) were unhealed, 58 (15.4%) had been treated by amputation and 18 (4.8%) had persisted until the time of patient death. With healing as the endpoint, there was a strong linear by linear association with ulcer depth graded by the criteria common to the University of Texas (UT), S(AD)SAD and PEDIS classification systems (70.558, p<0.001). There was also a strong association with infection at baseline, and this was stronger using the S(AD)SAD classification (61.774, p<0.001) – which distinguishes cellulitis and osteomyelitis, than with PEDIS (37.927, p<0.001) and UT stage – which combines infection and ischemia (32.929, p<0.001). There was a weaker association with ischemia alone (10.799, p<0.001), as well as with neuropathy (12.475, p<0.001), whether graded by either the S(AD)SAD or the PEDIS systems. There was a very weak correlation between healing and cross-sectional area (4.387, p=0.036) in the S(AD)SAD system. Classification by S(AD)SAD or PEDIS may predict outcome better than the UT system in an African population. The associations found between baseline status and healing are very different from those previously reported from UK and US, and emphasise the importance of neuropathy and infection, as opposed to ischaemia and ulcer area in this relatively young population.

[105-OR] A Prospective, Randomized, Controlled, Blinded, Multicenter Pivotal Trial of a Platelet Rich Plasma Gel* Versus Control When Added to the Standard of Care in the Treatment of Non-Healing Diabetic Foot Ulcers
Driver, V. R y col. North Chicago, IL.

A prospective, randomized, controlled, blinded, multicenter trial of Platelet Rich Plasma (PRP) Gel* versus Control for the treatment of non-healing diabetic foot ulcers was conducted at 14 clinical trial sites. Platelet rich plasma gel is an autologous, point-of-care process where a small volume of blood is drawn from the subject; the blood is separated in a centrifuge; the resulting PRP is extracted and activated. This activation causes the platelets to release autologous multiple growth factors; the fibrinogen in the plasma to convert to a fibrin matrix scaffold; the liquid develops a gel consistency. This PRP gel is applied topically to the prepared wound bed.
Subjects were randomized per inclusion/exclusion criteria and were treated twice a week for 12 weeks, or until healing, e.g. 100% epithelialization. Healed wounds were followed for 12 additional weeks.
Seventy two (72) subjects/wounds were enrolled in the study. An independent audit of the sites was conducted to ensure compliance with Good Clinical Practices. Based on the audit, the final per protocol analysis was conducted on 40 subjects from 7 sites.
The majority of the wounds (35/40) 88% were in the size range ≤ 7 cm2 area and ≤ 2 cm3 volume. Results indicated that 81.3% of the PRP gel treated wounds healed compared to 42.1% of the control treated wounds (p=0.036). When the larger wounds were included, 68.4% of the PRP gel healed versus 42.9% of the control treated wounds (p=0.125). There were no reported serious adverse events. One PRP gel healed wounds broke down during the 12 week followup period.
This first prospective, randomized, controlled trial in the US using autologous PRP gel for treating diabetic foot ulcers demonstrates PRP gel healed the majority of the diabetic foot ulcers, without serious adverse events, and minimal recidivism.
*AutoloGel™ System, Cytomedix, Inc, Rockville, MD

[106-OR] The Use of Medical Hyperspectral Imaging (MHSI) To Evaluate Microcirculatory Changes in Diabetic Foot Ulcers and Predict Clinical Outcomes
Dinh, T y col. Boston, MA.

Foot ulceration is a serious complication of diabetes and new techniques that can predict wound healing may prove very helpful in the management of this condition. Medical Hyperspectral Imaging (MHSI) is a novel camera-based diagnostic which can quantify tissue oxy- and deoxyhemoglobin on a pixel by pixel basis and reflect systemic microcirculatory changes associated with diabetes. We tested the feasibility of MHSI to provide spatially localized data that can predict healing vs nonhealing. Ten T1DM patients with 17 foot ulcers, 12 T1DM without ulcers and 11 non-diabetic controls were seen four times over a 6-months period. MHSI measurements, including oxyhemoglobin, deoxyhemoglobin, total-hemoglobin and oxygen-saturation (percentage of oxyhemoglobin to total hemoglobin assessed as the sum of oxy- and deoxyhemoglobin SHSIO2) were performed at the ulcer area, palm, volar forearm, foot dorsum and sole. Foot ulcers images were divided into 25 concentric circles and 8 pie segments forming 200 sectors per ulcer and the above measurements were performed in each sector. MHSI algorithms were developed to identify changes associated with ulcer healing and nonhealing. During the 6-month period, 8 ulcers healed and 9 did not heal. The preliminary analysis was carried on SHSIO2 and OXYHSI and it showed a significant reduction in OXYHSI was observed among ulcers that healed (-1.1 ± 1.2) and did not heal (-9.8 ± 1.2) (p<0.001). TCPO2 and laser Doppler data were not significantly correlated with ulcer healing or progression. Patients who did not heal also demonstrated increased heterogeneity in distant foot and in arm measurements. We conclude that MHSI has the potential to identify microvascular abnormalities in the diabetic foot and assist in managing foot ulceration.

[115-OR] Liraglutide Significantly Improves Glycemic Control, and Lowers Body Weight without Risk of Either Major or Minor Hypoglycemic Episodes in Subjects with Type 2 Diabetes
Vilsboll, T y col. Francia, Holanda. República Eslovaca.

The objective of this 14-week study was to assess the efficacy and safety of liraglutide, a once-daily GLP-1 analogue, in 165 subjects with type 2 diabetes (HbA1c at randomization: 8.1-8.5%). Subjects treated with diet (19 %) or single oral antidiabetic agents (after 4 weeks washout) were randomized to one of three once-daily doses of liraglutide (0.65 mg, 1.25 mg or 1.9 mg) or placebo (1:1:1:1). Significant improvement in HbA1c was achieved in all actively treated groups vs. placebo (p<0.0001), with an estimated difference vs. placebo of -1.74% at the highest dose level (last observation carried forward - LOCF). The fraction of subjects achieving an HbA1c level ≤ 7.0 % was between 0.43 and 0.50 in the three liraglutide arms versus 0.08 in the placebo group. Improvement in glycemic control was achieved without any major or minor hypoglycemic episodes. Dose dependent weight reduction was achieved. In the high-dose liraglutide treatment group, estimated change in body weight was -2.99 kg from baseline and -1.21 kg compared to placebo (p=0.039) (LOCF). Liraglutide was well tolerated; four subjects withdrew due to adverse events in the three liraglutide treated groups vs. three in the placebo treated group. Main adverse events were from the gastrointestinal (GI) system; diarrhea was the most frequent, with an incidence of 19.5% and 12.5% in the high-dose liraglutide group and in the placebo group, respectively. In the high-dose group 10% of the subjects experienced nausea. The frequency of GI events decreased over time. In conclusion, liraglutide used as monotherapy significantly improved glycemic control without risk of major or minor hypoglycemia, was well tolerated, and lowered body weight in subjects with type 2 diabetes.

[120-OR] Sustained Reduction in HbA1c during One-Year Treatment with Vildagliptin in Patients with Type 2 Diabetes (T2DM)
Dejager, S y col. Suiza.

Vildagliptin is a potent and selective DPP-4 inhibitor that improves islet function by increasing α- and β-cell responsiveness to glucose. This 52-wk, multicenter, randomized, double-blind study compared vildagliptin 100 mg daily (50 mg bid, n=526) to metformin 2000 mg daily (Met, 1000 mg bid, n=254) in drug-naive patients (pts) with T2DM. The groups were well-balanced at baseline (BL). Mean age, BMI, HbA1c and disease duration were 53.1 y, 32.4 kg/m2, 8.7% and 2.4 y, respectively. As shown below, both vildagliptin and Met rapidly decreased HbA1c and this improvement was sustained throughout 52-wk treatment. The adjusted mean change (AMΔ) from BL to endpoint was -1.0±0.1% in pts receiving vildagliptin and -1.4±0.1% in those receiving Met. The between-group difference in AMΔ HbA1c did not establish non-inferiority of vildagliptin 100 mg daily to Met 2000 mg daily in these drug-naive pts.

Vildagliptin did not induce weight gain (0.3±0.2 kg) and Met-treated pts had a modest weight loss (-1.9±0.3 kg). The incidence of side effects was 70.1% in patients receiving vildagliptin vs 75.4% with Met. The incidence of GI side effects was significantly less in pts receiving vildagliptin (21.8%) vs Met (43.7%, P<0.001). Diarrhea was reported by 6.0 vs 26.2% of pts receiving vildagliptin vs Met; nausea, by 3.3 vs 10.3%; abdominal pain, by 2.3 vs 7.1%; dyspepsia, by 1.2 vs 4.8%; and vomiting, by 2.1 vs 4.4%. A mild hypoglycemic episode was reported by 3 vildagliptin-treated patients (0.6%) and by one Met-treated patient (0.4%). Conclusions: 1) a clinically meaningful reduction in HbA1c is sustained throughout one-year treatment with vildagliptin 50 mg bid, 2) vildagliptin is well-tolerated and associated with fewer GI side effects than metformin.

[791-P] Effect of Ruboxistaurin (RBX) on Quantitative Measures of Diabetic Peripheral Neuropathy (DPN)
Casellini, N y col. USA. CASELLINI, N y col. USA.

Hyperglycemia in diabetes leads to PKC β overactivation and microvascular dysfunction, possibly resulting in sensory symptoms, neurological impairment, and electrophysiologic changes seen in patients who develop DPN. We tested the effect of 32 mg/d RBX, a specific PKC β inhibitor, on quantitative measures of DPN. We report data for 20 placebo [PBO] and 20 RBX-treated patients (≥18 years old, type 1 or 2 diabetes, and HbA1c ≤11%) enrolled in a randomized, double-blind, single-site trial in the United States. RBX and PBO groups were comparable at baseline with regard to age and duration of diabetes and DPN. At endpoint (6 months), NTSS-6 improved by 39.3% in the RBX group vs 2.3% in the PBO group (p<0.037). There were no significant between-group differences for change from baseline in electrophysiology, skin nerve fiber density, quantitative sensory testing of vibration, cold, or warm sensation thresholds, cardiac autonomic reflex testing, skin microvascular blood flow (SBF), or Neurologic Impairment Score. In the RBX group, sensory symptom scores decreased from baseline (NTSS-6: -39.3%, p<0.003; Neurologic Symptom Score [NSS]: -17.8%, p<0.03), as did sensory signs (SS; -20.5%, p<0.02) and Total Neuropathy Score (TNS; -19.8%, p<0.02). SS (-4.1%, p<0.04) and TNS (-19.6%, p<0.003) also decreased from baseline in the PBO group. Endothelium-dependent SBF measured by Laser Doppler at the distal calf increased from baseline after local warming to 40oC in the RBX (by 92%, p<0.005) but not the PBO group. C fiber-mediated SBF increased from baseline after local warming to 44oC in both groups (RBX: 56%, p<0.008; PL: 100%, p<0.002). RBX decreased NTSS-6, NSS, SS, and TNS, and increased SBF, from baseline after 6 months. Data from this small cohort suggest that enhancing nerve blood flow with RBX may reduce sensory symptoms and signs when affecting nerve or endothelial function.

[796-P] Duloxetine's Long-Term Effects on Glycemic Control in Patients with Diabetic Peripheral Neuropathic Pain
Hardy, T y col. Indianapolis, USA.
 
Duloxetine has been shown to be safe and effective in the management of diabetic peripheral neuropathic pain (DPNP). We report the impact of long-term treatment with duloxetine on fasting blood glucose (FBG), glycosylated hemoglobin (HbA1C), lipids and weight.
In three long-term studies, 867 patients were randomized at 2:1 ratio to duloxetine 60 mg BID (N=580) or routine care (N=287) for 52 weeks after completing 12-week placebo-controlled, acute phase studies. Pooled data from the study extensions were analyzed.
After long-term treatment with duloxetine of up to 52 weeks, there was an increase in HbA1C in both the duloxetine and routine care groups, but the mean increase was greater in the duloxetine-treated group (0.5% vs 0.2%, respectively, p<.001). There was also a modest increase in FBG and a small increase in total cholesterol in duloxetine-treated patients while the routine care group showed a slight decrease (mean change FBG: 0.67 vs -0.64 mmol/L, p<.001; cholesterol: 0.06 vs -0.16 mmol/L, p=.005, respectively). A small decease in HDL cholesterol was seen in both groups, although the decline was significantly smaller with duloxetine than with routine care (mean change: -0.014 vs -0.078 mmol/L, p=.002 respectively). In these trials, after 52 weeks of therapy, the weight increase was smaller in the duloxetine-treated patients, compared to routine care group, but the difference was not statistically significant. The frequency of diabetes-related adverse events was similar between groups. Long-term duloxetine treatment appears to be associated with a slight worsening of metabolic control in patients with DPNP. Patients managed with duloxetine for DPNP should continue to have their glycemic control closely monitored.

[799-P] Dramatic Impact of Surgically-Induced Weight Loss on Cardiovascular Autonomic Function in Severely Obese Diabetic Individuals
Maser, R y col. Newark. USA.

Reduced heart rate variation (HRV) is the earliest indicator of cardiovascular autonomic neuropathy (CAN). CAN is an independent predictor of mortality. Weight loss (diet and surgically-induced) has been shown to improve HRV in non-diabetic subjects. The purpose of this study was to determine if surgically-induced weigh loss improved HRV in diabetic persons. Participants (n=28) (aged 44±9 (mean±SD) years; 86% women, 32% with a history of diabetes) were examined before and 6 months after bariatric surgery (i.e., Roux-en-Y gastric bypass (n=23) or adjustable gastric banding system (n=5)). Autonomic function was assessed via RR-variation during deep breathing (i.e., mean circular resultant [MCR]) and expiration/inspiration [E/I] ratio) and during the Valsalva maneuver. Table 1 shows that autonomic function improved with weight loss.

Table 2 indicates that improved autonomic function was similar for those with and without diabetes.

These preliminary results indicate a greater degree of improvement in HRV than that reported by others using medications (e.g., ACE inhibitors) or diet induced weight loss. An improved autonomic profile in response to a reduction in body mass may help to reduce the risk of cardiovascular morbidity and perhaps mortality for individuals with and without diabetes.

[617-P] Factors That Identify T1DM Drivers at Risk for Future Hypoglycemia-Related Driving Mishaps
Cox, D y col.  USA.

Driving mishaps are more common among individuals with T1DM, accounted for by a subgroup of T1DM drivers. This multi-center study investigated what factors differentiate T1DM drivers at risk for future hypoglycemia-related driving mishaps. All 515 T1DM drivers completed a series of questionnaires and then employed a memory meter and a hand held computer with a driving diary for one month to test a theoretical risk model. Then subjects were queried monthly for a year about driving mishaps. All 493 subjects completed all aspects of data collection.
Zero hypoglycemia-related driving mishaps were reported by 67%, 1 mishap was reported by 15%, and >2 mishaps by 18%. There were 38 episodes of severe hypoglycemia while driving, 15 collisions, and 7 citations, 14 episodes of loosing control of the vehicle, 205 episodes of someone having to take control of the vehicle and 214 episodes of automatic driving, all due to hypoglycemia. Subjects reporting >2 hypoglycemia driving mishaps were more likely to live alone (p=.03), drink more alcohol (p=.02), have a MiniMental Status Exam score <24 (p=.003), take more risks while driving (p=.001), have a history of collisions in the previous two years (p=.001), and experience more severe hypoglycemia in the past year (p=.005), report more hypoglycemia driving mishaps in the previous year (p=.0001), and experience more mild symptomatic hypoglycemia in the previous six months (p=.0005). These subjects did not differ in terms of age, gender, miles driven, visual acuity, sleep difficulties, duration of disease, neuropathy or retinopathy, hypoglycemia unawareness, insulin regimens or glucose in the car
Conclusion: 1) the majority of T1DM drivers do not report hypoglycemia-related driving mishaps, 2) those who will have driving mishaps may be predictable, and 3) risk factors for future driving mishaps may be minimized with behavioral interventions.

[629-P] Hypoglycemia in the Hospital: A Method for Measuring Frequency and Severity
Korytkowsi, M y col.  USA.

As many hospitals explore methods for achieving recommended blood glucose ranges for hospitalized patients with general medical and surgical conditions, it is important that the safety of these interventions be closely monitored. Hypoglycemia stands out as the most important measure of safety for treatment of hyperglycemia. Currently, there is no standard definition or reporting mechanism for hypoglycemia in hospitalized patients.
To address this, we queried a computerized archival database for all bedside blood glucose (BG) values < 70 mg/dl occurring outside a critical care unit on all admissions to our institution in May 2004. Severity was defined as follows: mild (BG 50-69), moderate (40-49), severe (BG<40 mg/dl). All BG values < 70 occurring within a 4 hour interval were counted as a single event (i.e. maximum 6 events per patient per day) with the nadir BG defining the severity of the event.
Of 2562 total admissions, 182 were complicated by the occurrence of hypoglycemia. 239 mild events occurred in 151 admissions (range 1-15/adm); 45 moderate events in 37 admissions (range 1-3/adm), and 40 severe events in 29 admissions (range 1-5/adm). The frequency of hypoglycemia was 15.6, 2.9 and 2.6 events per 1000 patient days, respectively.
Patients with hypoglycemia had higher Charlson Severity of Illness Scores (1.4 vs 2.2, P<0.001), and longer mean hospital length of stay (5.6 + 7.9 vs 11.5 + 17.9 days, p<0.001) than those without. 50% of events occurred within the first 2 days and 73% within the first 5 days of hospitalization. Diagnosis codes observed to have the greatest frequency of hypoglycemia were DM (26.5%), renal disease (23.8%), CHF (14.4%), and MI (8.6%).
In summary, we present an approach to determining the frequency, severity, predictors and impact of hypoglycemia in a tertiary care hospital that can be applied to other institutions wishing to monitor the safety of protocols that promote intensification of inpatient glycemic management.

643-P] Comparison of Serum High-Molecular-Weight (HMW) Adiponectin with Total Adiponectin Concentrations in Type 2 Diabetic Patients with Coronary Artery Disease Using a Novel Enzyme-Linked Immunosorbent Assay To Detect HMW Adiponectin
Aso, Y y col.  Japón.  

Adiponectin (Acrp30), an adipocyte-derived protein, exists in serum as a trimer, a hexamer, and a high-molecular-weight (HMW) form including 12 to18 subunits. As HMW adiponectin may be biologically active, we measured it in serum using a novel enzyme-linked immunosorbent assay (ELISA), comparing HMW with total adiponectin concentration in patients with type 2 diabetes, and confirming by gel-filtration chromatography that the ELISA detected mainly adiponectin with 12 to 18. We next investigated the relationship between serum HMW and coronary artery disease (CAD) in 280 consecutive type 2 diabetic patients including 59 patients with angiographically confirmed CAD. Total adiponectin was measured in serum by a commercially available ELISA. Like serum total adiponectin, HMW adiponectin correlated positively with high-density lipoprotein cholesterol, and negatively with triglyceride, insulin sensitivity, creatinine clearance, and circulating inflammatory markers. Total and HMW adiponectin were significantly higher in women than in men, as was the HMW/total adiponectin ratio. Serum HMW and the HMW/total adiponectin ratio were significantly lower in men with than without CAD (P<0.05, respectively). In women, the ratio, but neither total nor HMW adiponectin, tended to be lower when CAD was present. In conclusion, determination of HMW adiponectin, especially relative to total serum adiponectin, is useful for evaluating CAD in type 2 diabetic patients.

[649-P] The Effect of the Angiotensin II Receptor Blocker, Valsartan on Oxidative Stress and Inflammatory Mediators in Obese Subjects with Type 2 Diabetes
Viswanathan, P y col.  Buffalo. USA.

We have previously shown that the angiotensin II recptor blocker (ARB), valsartan exerts antioxidative stress effect in normal subjects. In the current study, we hypothesized that valsartan exerts an anti-inflammatory effect in diabetic (DM) hypertensive patients Ten DM hypertensive patients (age= 50.1±4.7 years , BMI=40.4±2.5 kg/m2) were recruited and blood samples were collected at baseline and at 4, 12, and 24 weeks following 160 mg per day valsartan treatment. Following valsartan treatment plasma C-reactive protein (CRP) concentrations were lower by 20±7% (P<0.05) by week 24. Monocyte chemoattractant protein-1 (MCP-1) plasma concentration was also suppressed by 15±10% (P<0.05), E-selectin 6±9% (P<0.05) and matrix metalloproteinase-9 (MMP-9) by 8±5% (P<0.05) following 24 weeks of 160mg per day valsartan. These anti-inflammatory effects were independent from changes in blood pressure. In conclusion, Valsartan at a dose of 160mg per day exerts a potent anti-inflammatory effect on diabetic patients with hypertension characterized by a significant reduction of plasma CRP, MCP-1, E-selectin and MMP-9 which is independent of changes in blood pressure.

653-P] Microvascular Myocardial Disease in Type 1 Diabetes Mellitus of Latin American Origin
Drori, R y col.  Los Angeles. USA.

Background: The mechanism of myocardial dysfunction in the absence of coronary artery disease (CAD) in Diabetes Mellitus (DM) is poorly understood. Different pathophysiological mechanisms have been suggested, including autonomic neuropathy, increased collagen cross-linking, oxidative stress and microvascular dysfunction (MD). Diabetes is known to be associated with MD. 31P magnetic resonance spectroscopy (MRS) is a technique used to detect myocardial MD in patients without evidence of CAD. This technique was utilized to evaluate type 1 DM (T1-DM) patients, and correlate with findings in other organ systems.
Methods: 70 T1-DM patients with disease duration >10 years and no evidence of CAD were studied. The patients were recruited from our ongoing registry of T1-DM Latinos at USC Keck School of Medicine. History, physical examination, metabolic panel, ECG, urinary and ophthalmologic evaluation were performed on all subjects. 23 subjects underwent myocardial 31P MRS before and during handgrip stress to assess phosphocreatine/ATP (PCr/ATP) ratio. This study analyzed the 16 patients who had technically adequate 31P MRS.
Results: Mean age 29.1±7.7 years, duration of disease 14.8±4 years, body mass index 29.7± 6.1, hemoglobin A1C 9.7±2.1, cholesterol 181±37 mg/dl, LDL-chol 111±36 mg/dl, HDL-chol 47.5±9.4 mg/dl, triglyceride 109.9±61.5 mg/dl, microalbuminuria was present in 12.5% of subjects, and retinopathy in 31% of subjects. Fifty percents of the subjects showed a significant fall in PCr/ATP (mean -32.7±5.1, P<0.001) during handgrip stress, of these only one had retinopathy and two had microalbuminuria.
Conclusion: Type 1 DM patients demonstrate a significant stress-induced reduction in PCr/ATP. These data suggest myocardial microvascular abnormalities as a mechanism in diabetic cardiomyopathy. Furthermore, myocardial microvascular disease may present earlier than retinal or renal involvement.

[659-P] Tight Intraoperative Glycemic Control and Outcomes in Cardiac Surgery Patients: A Randomized Controlled Trial
Gandhi, G y col. Rochester. USA.

There is no consensus on optimal management of hyperglycemia during cardiac surgery. Safe and effective insulin infusion algorithms that enable maintenance of rigorous glycemic control are lacking. We conducted a randomized trial to determine whether normalization of intraoperative glucose levels with insulin improves outcomes. Patients undergoing cardiac surgery were randomly assigned to receive intensive insulin therapy (IT) intraoperatively to maintain glucose levels between 80 to 100 mg/dL (185 patients) or conventional treatment (CT) (186 patients). Primary outcome was a composite of mortality, sternal infections, prolonged ventilation, cardiac arrhythmias, stroke and renal failure within 30 days after surgery. Mean glucose levels were similar at induction of anesthesia (111±22 IT vs 111±31 mg/dL CT; P=.98). Glucose concentrations were lower in the intensively than conventionally treated group, after cardiopulmonary bypass (123±24 vs 148±35 mg/dL; P<.01) and on arrival to ICU (114±29 vs 157±42 mg/dL; P<.01). Postoperatively, both groups were treated with an insulin infusion to maintain normoglycemia and glucose levels after 24 hours were no different (103±17 IT vs 104±22 mg/dL CT; P=.79). The frequency of hypoglycemia was low (1 episode in each group intraoperatively). Overall, there was no benefit of intensive intraoperative insulin therapy on the primary composite endpoint (44% IT vs 46% CT event rate, OR 0.93; 95% CI .62-1.39; P=.36). There was no difference in length of ICU (2±2 IT vs 2±3 days CT, P=.18) and hospital stays (8±4 IT vs 8±5 days CT, P=.33) between the groups. Intensive insulin therapy to maintain intraoperative normoglycemia in cardiac surgery patients does not reduce mortality or morbidity. When given in a controlled setting using standardized protocols, intensive insulin therapy was safe and able to maintain glucose concentrations close to normal during surgery.

[663-P] Discontinuation of Antihyperglycemic Therapy and Mortality after Acute Myocardial Infarction (AMI)
Inzucchi, S y col.  Denver. USA.

AMI patients with diabetes (DM) have increased mortality; hyperglycemia may contribute to this risk. Often, outpatient antihyperglycemic regimens are stopped and not resumed at hospital discharge, a practice with unknown implications for outcomes. Using data from the National Heart Care Project, we studied Medicare beneficiaries with DM hospitalized for AMI. The independent association between the discontinuation of DM therapy and outcomes was assessed in multivariable Cox models, adjusting for clinical variables, including admission glucose, complications, ventricular function, and concurrent medications. The primary outcome was time to death within 1 year.
The study cohort consisted of 8,751 patients treated with antihyperglycemics prior to admission. Of these, 1170 (13.3%) were discharged off DM therapy (Group A), 37.6% of whom died within 1 year, compared to 28.2% of those continued on DM therapy (Group B; n=7581) (p<0.0001). After multivariable adjustment, increased mortality in Group A persisted (HR 1.42 [95% CI 1.27-1.59]). Slightly fewer readmissions in Group A (HR 0.92 [0.85-1.00]) suggested more out-of-hospital deaths. During the first 30 days, Group A experienced 36.1% of its overall mortality, compared to 22.9% early deaths in Group B. Because of concerns that failure to restart DM therapy occurred more commonly in those needing long-term care, we stratified patients by discharge disposition. The association with mortality remained significant among those discharged to home (HR 1.24 [1.06-1.44]).
In older DM patients with AMI, the failure to continue therapy for DM is associated with excess mortality, especially in the first 30 days. It is unclear whether this is the result of hyperglycemia post-discharge but suggests that maintaining DM therapy after AMI may be important in preventing adverse outcomes.

[667-P] Effect of Increased Dosages of Atorvastatin on LDL Subclass Distribution in Patients with Dyslipidemia
Karalis, D y col. USA.

A common finding in diabetes and the metabolic syndrome is the presence of small dense LDL particles which are thought to increase cardiovascular risk in these patients. Evidence from a small number of studies suggests that statins improve LDL subclass distribution (LDL-SD); however, the effect of increased statin dosages on LDL-SD is not known.
In a retrospective subgroup analysis of 190 patients with hypertriglyceridemia (TG > 200 mg/dL) from the NASDAC study, the effects of atorvastatin 10, 20, 40 and 80 mg on LDL-SD and peak particle diameter in angstroms (Ao) were determined using gradient gel electrophoresis.
All doses of atorvastatin significantly lowered the concentration (mg/dL) of small dense LDL (subclasses IIIa + IIIb) and significantly increased LDL peak particle diameter (range 2.0-3.8 Ao) from baseline. Both low (10/20 mg) and high dose (40/80 mg) atorvastatin significantly reduced the percent distribution of small dense LDL (-8.4%, p<0.004 & -7.345, p<0.006 respectively) from baseline. High dose atorvastatin lowered small dense LDL more than did low dose atorvastatin ( -40.7 % vs. –28.8%; p<0.003).
Atorvastatin altered the atherogenic lipid profile in patients with dyslipidemia resulting in a shift from small dense LDL to more buoyant and less atherogenic particles. Higher doses of atorvastatin lowered small dense LDL more than lower doses and these effects may in part explain the benefits of high dose atorvastatin seen in clinical trials.

Preguntas y Asesoramiento sobre Pie Diabético por los Pacientes

¿Cada cuánto es conveniente hacerse revisar los pies?

Si no hay lesiones o factores de riesgo para úlceras de pie, una vez por año.

Si hay riesgo de desarrollar úlceras de pie, cada tres meses. Hay riesgo cuando hay vasculopatías, neuropatía, alteraciones del apoyo de los pies, antecedentes de úlceras, o antecedentes de amputación. Uno de esos elementos.

Mi padre tiene varios dedos negros (necróticos) con mucho dolor, que le impide dormir de noche. ¿Qué me sugiere?.