ResearchPad - clinical-studies-in-obesity-diabetes-risk-and-cardiovascular-outcomes Default RSS Feed en-us © 2020 Newgen KnowledgeWorks <![CDATA[SAT-613 Comparison of CV Risk Scores to Evaluate Cardiovascular Risks in Thai Type 2 Diabetes]]> Objective: Various cardiovascular risk scores have been developed and the RAMA-EGAT risk score was developed by Thai database with minority of diabetes. This study aims to compare the predictability of the ADVANCE, UKPDS, SCORE, Framingham Risk Score (FRS) and RAMA-EGAT risk score for carotid atherosclerosis, arterial stiffness and peripheral arterial disease in Thai T2DM patients. Methods: A cross-sectional study was conducted in T2DM patients without established CVD at a tertiary care hospital. Demographic and DM-specific data were collected. Carotid intima-media thickness (CIMT), carotid plaque, cardio-ankle vascular index (CAVI) and ankle-brachial index (ABI) were measured as the markers of atherosclerosis. Risks of CVD were calculated according to the ADVANCE, UKPDS, SCORE, FRS and RAMA-EGAT risk scores. These risk scores were correlated with the atherosclerotic markers by odds ratio using logistic regression and the proper points of the risk scores to predict atherosclerosis were calculated by the areas under the curve (AUC). Results: There were 180 T2DM participants with the mean age of 60-year-old, diabetes duration of 13 years and mean A1C 7.4%. The highest sensitive risk score was FRS, following by UKPDS, SCORE, ADVANCE and RAMA-EGAT risk score, which indicated high-risk patients as 44.8%, 27.6%, 18.9%, 13.8% and 0% accordingly. There were 40.3% of the patients with arterial stiffness detected by CAVI > 9, 24.0% with carotid atherosclerosis defined by CIMT > 0.07 mm or presenting of carotid plague and 8.3% with ABI < 0.9. The odds ratios (OR) of 4 risk scores increased by the quartiles for carotid plaque, CIMT, CAVI and ABI while the OR of RAMA-EGAT scores increased by the quartiles only for carotid plaque. The highest quartile of ADVANCE, UKPDS, SCORE and FRS significantly (P<0.01) increased the risk of abnormal CIMT; OR 2.64-8.75, carotid plaque; OR 1.51-11.21, CAVI; OR 11.38-19.00, and ABI; OR 1.18-12.57. The highest quartile of RAMA-EGAT score significantly increased the risk of carotid plaque; OR 5.35 (1.44-19.91) P<0.01. ROC analysis revealed that ADVANCE > 3.0% in 4-year, UKPDS > 11% in 10-year, fatal-SCORE > 6% in 10-year and FRS > 18% in 10-year were predictive of carotid atherosclerosis with sensitivity of 76-84% and specificity of 61-69% and they were predictive of arterial stiffness with the sensitivity of 71-80% and specificity of 64-68%. Conclusion: There was no significant difference when comparing the predictability of the ADVANCE, UKPDS, FRS and SCORE risk estimation for carotid atherosclerosis, arterial stiffness and peripheral arterial disease and they were more correlative with atherosclerotic markers than RAMA-EGAT score in Thai type 2 diabetic patients.

<![CDATA[SAT-624 Predictive Ability of Lipoprotein Insulin Resistance (LPIR) Score in South Asians: A Comparison of Surrogate Indices of Insulin Sensitivity/Resistance]]> South Asians (SA) are at higher risk for developing insulin resistance (IR) and type 2 diabetes. Consequently, identifying IR in this population is important. Lack of standardization and harmonization of insulin assays limit the clinical use of insulin-based surrogate indexes of insulin resistance. The lipoprotein insulin resistance (LPIR) score, a metabolomic marker, reflects the lipoprotein abnormalities observed in insulin-resistant states. The reliability of the LPIR score to predict IR in South Asians is currently unknown. In this study, we aimed to evaluate the predictive accuracy of LPIR compared to other fasting-based surrogate indices in SA.

In a cross-sectional study of 59 non-diabetic SA subjects (age 36 ± 8 years, BMI 26.5 ± 5.2 kg/m2), we used calibration model analysis to assess the ability of the LPIR score and other simple surrogate indices [homeostasis model assessment (HOMA-IR), quantitative insulin sensitivity check index (QUICKI) and Adipose tissue insulin sensitivity (Adipo-SI)] to predict insulin sensitivity derived from the reference frequently sampled intravenous glucose tolerance test (FSIVGTT) and Minimal Model analysis (SiMM). LPIR scores were calculated using six lipoprotein particle concentrations and sizes measured by nuclear magnetic resonance (NMR) spectroscopy. Further, quantitative predictive accuracy and index comparisons were determined by root mean squared error (RMSE) of prediction and leave-one-out cross-validation-type RMSE of prediction (CVPE). Receiver operating characteristic (ROC) curve analysis was performed to determine how well LPIR distinguished insulin resistant individuals, categorized as an SiMM < 3.

As determined by calibration model analysis, Adipo-SI, HOMA-IR, and QUICKI showed moderate correlations with for SiMM (Adipo-SI: r = 0.66; HOMA-IR: r = 0.60; QUICKI: r = 0.57, p = <0.0001). No significant differences were noted among CVPE or RMSE from any of the routinely used surrogate indices when compared with LPIR. The ROC area under the curve was 0.76 (95% CI 0.64–0.87) suggesting that LPIR performed well in identifying insulin resistant subjects. The optimal cut-off in IR individuals was LPIR >46 (sensitivity: 75.9 %, specificity: 70.0%). We conclude that NMR-derived LPIR may be an appropriate index to assess insulin resistance in South Asians.

<![CDATA[SAT-621 Framingham Cardiovascular Disease 10-Year-Risk Score Is Associated with Myocardial Perfusion in Asymptomatic Diabetic Patients]]> Background: Even without atherosclerosis, diabetes increases the risk of death from coronary heart disease and heart failure. Myocardial perfusion dysfunction may occur in the early stage of diabetic cardiomyopathy, but its examination method is relatively complex. It is very important to carry out targeted cardiac screening to find the factors related to diabetic myocardial perfusion in the early stage.

Methods: We enrolled 77 patients with diabetes and 30 controls, performed anthropometric and laboratory tests such as blood glucose and lipids, and calculated Framingham Cardiovascular Disease 10-year-risk Score (FRS). All participants underwent cardiac magnetic resonance examinations and recorded their cardiac structure, functional indicators (such as ejection fraction (EF), end-diastolic volume (EDV), end-systolic volume (ESV), stroke volume(SV), peak filling rate (PFR),myocardial perfusion index (maximum upslope (Slope), half time to maximum signal intensity (Time50Max (s)), time to maximum signal intensity (TimeMax (s)), the maximum signal intensity (MaxSI),, basic signal intensity (Baseline),the ratio of MaxSI and Baseline ((MaxSI (BL) %), the difference value between MaxSI and Baseline (MaxSI (BL))).

Results: Compared with normal group, no cardiovascular symptoms of left ventricular and right ventricular systolic function in patients with diabetes and end-diastolic and end systolic volume had no obvious difference, left ventricular PFR is lower than normal (279.65 + 57.62 vs. 322.57 + / - 78.29, p = 0.02), in the subgroup analysis we found that the FRS high-risk groups, ventricular septal thickening tend to, and Slope, MaxSI, MaxSI BL (%), MaxSI (BL) were significantly lower than the high risk group, Time50Max and TimeMax were significantly longer than the non-high-risk group, and FRS was negatively correlated with Slope, MaxSI(%BL) and positively correlated with TimeMax(s) and Time50Max(s), with statistical significance.

Conclusion: Systolic function remains and diastolic function decreases in asymptomatic diabetic patients. Moreover, the patients with high risk of FRS had significant decreased perfusion function, and the quantitative indexes of perfusion function were closely related to FRS. It is of great value to pay attention to the changes of FRS score for early screening and diagnosis of diabetic heart disease.

<![CDATA[SAT-622 Heterogeneity of Familial Partial Lipodystrophy Type 2 from a Genotype-Phenotype Perspective]]> Phenotypic heterogeneity is a well-known feature of Familial Partial Lipodystrophy Type 2 (FPLD2) which is caused by pathogenic variants in the LMNA gene. Clinical diagnosis can be challenging in some cases. Likewise, trained physicians can report differences in body composition and clinical manifestation of FPLD2, highlighting the importance of accurate phenotyping. In this study, we aimed to identify phenotype-genotype correlations in a cohort of systematically evaluated patients with FPLD2. We retrospectively evaluated 43 patients diagnosed with FPLD2 (age 50.3±16.1 years, 79.1% women). Per pathogenic variants, patients were divided into two groups; 24 with R482Q (RQ: 55 ± 3.2 years, 70.8% women) and 19 with non-R482Q (Non-RQ: 46 ± 3.2 years, 84.2% women). Non-RQ group consisted of several pathogenic LMNA variants in exons 1, and 5 through 11. Also, DEXA parameters were studied in a subgroup of 19 patients with available assessments (in 11 RQ and 8 non-RQ patients) that were matched for age, sex and BMI. Patients in the RQ group were older when they were first diagnosed with lipodystrophy (48.6 ± 3.2 years and 37.4 ± 3.1 years, p = 0.03). Although the prevalence of diabetes, hepatic steatosis and other co-morbidities associated with metabolic control were similar in both groups at the time of the study, patients with RQ pathogenic variants were diagnosed later in life with diabetes (46.0 ± 4.2 years vs. 35 ± 3.5 years, p = 0.03) and hepatic steatosis (45.3 ± 6.9 years vs. 30.1 ± 3.7 years, p < 0.01. Although more pancreatitis episodes were reported in the RQ group (13 ± 3 vs. 2 ± 1, p = 0.02), the number of patients with a history of pancreatitis was similar across the groups suggesting the occurrence of recurrent pancreatitis episodes in selected patients with RQ pathogenic variant. Pain was a common complaint among the patients, but it was less severe in the RQ group (4.2±2.1 vs 2.3±2.0, p=0.05). In terms of body composition, patients with RQ pathogenic variants had greater bone mass (legs: 879 ± 59.3 g vs. 703.5 ± 33.7 g, p= 0.01; trunk 914.2 ± 65.5 g vs. 674.1 ± 28.0 g, p = 0.005, total body: 2643.7 ± 158.9 g vs. 2140.6 ± 78.4 g, p = 0.005) and higher fat mass in the legs (19 vs. 14%, p = 0.02). Similarly, patients with RQ pathogenic variants had less lean percentage (76 vs. 81%, p = 0.009), and accordingly, less fat-free mass percentage (80 vs. 85%, p = 0.02) in the legs. Total fat-free mass of the RQ group was also lower (66 vs. 76%, p = 0.0009). Genotype-phenotype characterization is important not only for understanding the natural history and clinical manifestation of the disease but also for establishing more accurate and precise diagnostic criteria or therapeutic approaches. Our data suggest more fat preservation in LMNA R482Q carriers, presumably leading to a later diagnosis of lipodystrophy and metabolic abnormalities. More studies are needed to confirm the differences observed in body composition.

<![CDATA[SAT-612 Features of the Severity of Cardiovascular Remodeling and Metabolic Disorders in Hypertensive Patients with Obesity in the Presence of Two Unfavorable Genotypes of the ADIPOQ and IRS-1 Genes]]> The results of a number of studies have shown that in arterial hypertension (AH), G/T and T/T genotypes of the adiponectin gene (ADIPOQ) and Gly/Arg and Arg/Arg genotypes of the insulin receptor substrate 1 gene (IRS-1) are associated with a greater severity of metabolic disorders and hemodynamic parameters compared with G/G and Gly/Gly genotypes of these genes.

The aim of the study: to evaluate the severity of cardiovascular remodeling and metabolic disorders in hypertensive obese patients in the simultaneous presence of two unfavorable genotypes of the ADIPOQ and IRS-1 genes.

Methods: We examined 300 AH patients: 200 patients with AH and obesity, 50 patients with AH and normal body weight, 50 patients with AH and overweight, 40 patients with AH, obesity and type 2 diabetes mellitus (DM2), 30 healthy individuals. The polymorphisms of the ADIPOQ and IRS-1 was assessed by molecular genetic method.

Results: It was found that in all groups of hypertensive patients, regardless of body weight and the presence of DM2, the simultaneous presence of two unfavorable genotypes of the ADIPOQ and IRS-1 genes occurred significantly more often than in healthy individuals: in 41% of AH patients with obesity, 30% of AH patients with normal weight, 40% of AH with overweight, 57.5% of AH with obesity and DM2 vs. 13.3% of healthy individuals. In hypertensive patients, in the presence of overweight and obesity, the frequency of combination of the two unfavorable genotypes of these genes was significantly higher than in AH patients with normal body weight.

Conducting comparative evaluation of AH patients with obesity depending on the presence of two unfavorable genotypes or two protective genotypes of the ADIPOQ and IRS-1 genes showed that carriers of the combination of the G/T + T/T genotype of the ADIPOQ and the Gly/Arg + Arg/Arg genotype of the IRS-1 had a higher body mass index, more pronounced insulin resistance, cardiovascular remodeling, adipokine imbalance, impaired carbohydrate and lipid metabolism.

Conclusions: In AH patients, the frequency of the simultaneous presence of two unfavorable polymorphisms of ADIPOQ and IRS-1 genes was higher than in healthy individuals. In AH patients with overweight and obesity, the frequency of combination of the two unfavorable genotypes of the ADIPOQ and IRS-1 genes was significantly higher than in normal body weight. The presence of a combination of two unfavorable genotypes of the ADIPOQ and IRS-1 genes in patients with AH and obesity was associated with a greater severity of cardiovascular remodeling and metabolic disorders compared with the combination of two protective genotypes of these genes.

<![CDATA[SAT-611 Dietary Reduction of Branched-Chain Amino Acids]]> Background: One of the primary risk factors for the development of diabetes is obesity. Although moderate weight loss can lead to improvements in metabolic health, reduced-calorie diets are difficult to sustain. A number of groups have shown that low protein diets are associated with metabolic health in both rodents and humans. In particular, the branched-chain amino acids (BCAAs) leucine, isoleucine, and valine are associated with insulin resistance and diabetes in humans. Blood levels of the BCAAs decrease in humans fed a low protein diet, and we recently showed that reducing either dietary BCAAs or protein rapidly restored normal body composition and insulin sensitivity to diet-induced obese mice without reducing calorie intake. We are determining the effect of a low BCAA diet in humans with prediabetes and overweight/obesity.

Objectives: The primary outcome is the reduction of dietary BCAA intake by at least 50% in subjects in the low BCAA group while maintaining overall baseline calories. Secondary outcomes are compliance and tolerability of the low BCAA protein powder.

Method: This is a randomized, controlled, single-blind pilot study. The intervention arm uses a low BCAA protein powder to replace two meals per day for 60 days. The control arm uses a control protein powder with standard amounts of amino acids to replace two meals per day for 60 days. We are enrolling 16 males with the following criteria: ages 35 to 65, BMI 28 to 35, and hemoglobin A1c 5.7%-6.4% or fasting glucose 101-125 mg/dL. A registered dietitian reviews a 4-day food diary prior to diet initiation and creates an individualized meal plan based on those values in order to maintain baseline calories during the study diet. Baseline measurements prior to diet initiation include waist circumference, body mass index, fasting insulin and glucose, an oral glucose tolerance test, resting metabolic rate, body composition testing using dual energy x-ray absorptiometry, jumping mechanography to assess muscle function, and a stool sample to assess the microbiome. These tests are repeated after 60 days on the diet. Safety labs are performed while on the diet and 2-3 weeks after the end of the diet. Weekly safety telephone calls occur while on the diet. The food diaries are repeated after 30 and 60 days on the diet.

Results/Conclusion: Ten of sixteen subjects have completed the trial to date. One out of four subjects in the low BCAA group dropped out; the remainder successfully completed the study. BCAA intake was successfully reduced by 50%. Missed beverages were uncommon. No significant safety concerns or side effects have been noted. In conclusion, our early results suggest that replacement of two meals a day with a protein powder lacking BCAA for up to two months is a safe and feasible intervention. Ongoing analysis will determine if this intervention impacts metabolic health.

<![CDATA[SAT-616 Associations Of Body Mass Index And Waist Circumference In Young Adulthood With Later Life Incident Diabetes]]>


Overweight and obesity are known risk factors for incident diabetes, but it remains unclear if exposures during young adulthood (age 18 to 39 years) contribute to mid and late-life (age ≥40 years, collectively labeled here as “later-life”) risk of incident diabetes independent of later-life risk factor exposures.

We sought to assess the independent associations between young adult exposures to overweight and obesity, as assessed by body mass index (BMI) and waist circumference (WC), with later-life incident diabetes, accounting for later-life exposures.

We pooled data from six US cohorts (ARIC, CARDIA, CHS, Framingham Offspring, Health ABC, and MESA), and imputed life-course risk factor trajectories for BMI and WC, as well as for multiple cardiometabolic risk factors, annually from age 18 years to end of follow-up for each participant. Incident diabetes was defined by observed fasting blood glucose ≥126 mg/dL, non-fasting glucose ≥200, or use of diabetes medications. We used Cox proportional hazards models to examine the independent associations between time-weighted average exposures to BMI and WC during young adulthood and incident diabetes. We also performed mediation analyses to assess whether these associations were mediated by young adult exposures to other cardiometabolic risk factors (blood pressure, lipids, insulin resistance).

30,780 participants were included (mean age at first in-person visit 53.1±16.2 years; 56.1% female). Over a 9-year median follow-up, 4,323 participants had incident diabetes. Both young adult BMI and WC were associated with diabetes risk in a dose-dependent manner, independent of later-life BMI and WC. Compared to BMI 18.5–24.9 kg/m

, hazard ratios (HR) for incident diabetes were 1.27 (95%CI: 1.14–1.41) and 1.99 (95%CI: 1.67–2.37) for BMI 25–29 kg/m

and ≥30 kg/m

, respectively. Similarly, compared to normal WC (≤80 cm women; ≤94 cm men), the HRs were 1.42 (95%CI: 1.26–1.59) for WC 81-88cm (women)/95-102cm (men) and 2.13 (95%CI: 1.87–2.43) for WC >88cm (women)/>102cm (men). Young adult homeostatic model of insulin resistance (HOMA-IR) mediated 49% (95%CI: 23–76) and 44% (95%CI: 26–62) of the association between young adult BMI and WC with later-life incident diabetes, respectively.

Elevated BMI or WC during young adulthood were independently associated with later-life incident diabetes, after accounting for later-life BMI and WC, with insulin resistance suggested as a key mediator.

<![CDATA[SAT-617 Potential Utility of the Mixed Meal Test for Differential Diagnosis of Partial Lipodystrophy from Common Type 2 Diabetes and Truncal Obesity]]>


Better clinical tools are needed to improve the differential diagnosis of partial lipodystrophy (PL) from type 2 diabetes (DM) with truncal obesity. Here, we investigated differences in metabolic parameters during a mixed meal test in PL and DM patients to determine if this test may have a role in this regard.

We retrospectively evaluated data collected from 17 PL patients (4M/13F, ages 12-64) and 20 DM patients (13F/7M, ages 24-72) with truncal obesity, who also had nonalcoholic fatty liver disease. All patients underwent a Mixed Meal Test (MMT) with 474 ml of Optifast (320 kcal, 50% carbs, 15% fat, and 35% protein). Blood was collected before and at 30, 60, 90, 120, and 180 minutes post-meal to measure glucose, insulin, free fatty acids (FFA), triglycerides, inflammatory markers, GIP, GLP-1, PYY, and Ghrelin. All samples of the same cohort were run at the same time in duplicates and results were averaged. Mixed linear models were constructed to compare PL and DM cohorts taking into account within-subject effects. Data were controlled for BMI, sex and age, and glucose when necessary.

Patients with PL had higher glucose and triglyceride levels throughout the MMT at all-time points (p < 0.05). While the glucose levels showed an increase and subsequent decrease, the triglyceride levels remained flat throughout the test in both groups. Free fatty acid levels were suppressed compared to baseline during the test, but PL patients had significantly higher FFA from time 30 to time 180 (p < 0.05) and tended to suppress less. While controlling for the differences in glucose levels, GIP levels displayed a large peak at time 30 min in both groups but were significantly higher over the course of the test in the PL group (AUC: 32542, pg/mL x min (20528-57728) vs. 3343 pg/mL x min (1728-4498), p < 0.05). In contrast, GLP-1 levels (also peaking at time 30 min in both groups), were significantly lower in PL throughout the test (AUC: 3017 pg/mL x min (2309-6051) vs. 28387 pg/mL x min (20422-36045), p < 0.05). Ghrelin and PPY levels did not differ significantly between the two groups.

PL patients displayed more profound hyperglycemia and impaired suppression of FFAs. Interestingly, PL patients did not show substantial increases in triglyceride levels during MMT. There was a striking difference in the incretin responses between the two populations despite controlling for glucose, suggesting that MMT may have a role in differential diagnosis PL. Also, altered incretin response should be investigated as a contributor to metabolic perturbations and pathophysiology of PL.

<![CDATA[SAT-627 Racial and Ethnic Differences in Metabolic Disease in Obese Adolescents with Polycystic Ovary Syndrome]]> 90 BMI%ile) adolescent (12-21 years) female participants with PCOS enrolled across 4 protocols. Measurements included fasting hormone and metabolic measures, a 2-hour oral glucose tolerance test and MRI for hepatic fat. Groups were compared by ANOVA, with and without correction for BMI or chi-square tests for proportions. Results: Participants included 39 white (NHW 15.7±0.2 years; 97.7±0.2 BMI%ile), 50 Hispanic (15.2±0.3 years; 97.9±0.3 BMI%ile) and 12 black (NHB 16.0±0.6 years; 98.6±0.4 BMI%ile) adolescents. BMI%ile was different between groups (p=0.04), but age of menarche, free testosterone and hirsutism severity were not. Hepatic markers of insulin resistance were worse in Hispanic youth, including lower sex hormone binding globulin and TG/HDL ratio (p<0.001), although HOMA-IR was worst in NHB (p=0.009) and Hispanic (p=0.036) compared to NHW youth. There were no significant differences in insulin concentrations—either fasting or during the OGTT—although fasting C-peptide was higher in Hispanic (p=0.008) compared to NHW youth. Fasting and 2-hour glucose concentrations were not different between groups. HbA1c was highest in NHB (5.7±0.4%, p<0.001 vs. NHW, p=0.026 vs. Hispanic), then Hispanic (5.5±0.3, p<0.001 vs. NHW), then NHW (5.2±0.3) youth. Fasting triglycerides differed between groups (p=0.029), being highest in Hispanic youth (129 [105,167] mg/dL). The frequency of hepatic steatosis (NHW 42%, Hispanic 62% NHB 25%, p=0.032) and the metabolic syndrome components HDL <40 mg/dL (NHW 61%, Hispanic 82% NHB 50%, p<0.001) and HbA1c 5.7-6.4% (NHW 5%, Hispanic 36% NHB 50%, p<0.001) were different between the groups. Conclusions: Adolescents with PCOS appear to show similar racial and ethnic variation to the general population in terms of metabolic disease components. ]]> <![CDATA[SAT-620 Left Ventricular Myocardial Deformation in T2DM Is Associated with Chronic Hyperglycemia but Not Myocardial Perfusion: A Study Based on Magnetic Resonance Imaging]]> <![CDATA[SAT-623 Glucose Intolerance Modifies the Association Between Insulin-Like Growth Factor-1 and All-Cause Mortality]]> 20 years from the National Health and Nutrition Examination Survey (NHANES)-III (1988-1994) were linked to the National Death Index through 2015. Glucose intolerance was classified as per fasting blood sugar (≥100 mg/dl), hemoglobin A1c (≥5.7%), medication use, or self-reported diagnosis. IGF-1 was categorized into sex-specific quartiles. To analyze the joint impact of insulin and IGF-1, we also categorized participants into four groups: Group I) IGF-1 <230 ng/mL & insulin ≥11 uIU/mL, II) IGF-1 <230 & insulin <11, III) IGF-1 ≥230 & insulin ≥11, and IV) IGF-1 ≥230 & insulin <11. Our primary outcome was all-cause mortality. We used survey design-adjusted Cox regression to estimate the risk of all-cause mortality, adjusting for confounders. Results: Among the 5,283 subjects, 2,214 (42%) had GI. Participants had a mean follow-up of 22.1 years, during which 1,835 (34%) of them died. Those with GI in the highest quartile of IGF-1 had an unadjusted 64% lower risk of all-cause mortality compared to the lowest quartile (GI = unadjusted OR [95% CI]: 0.37 [0.24,0.55]). This association, although protective, was significantly less protective than those with normal glucose tolerance (NGT) (unadjusted OR: 0.16 [0.12,0.23]). After adjusting for confounders, these associations became insignificant (GI = aOR: 1.02 [0.73,1.42], NGT = aOR: 1.04 [0.74,1.46]). When estimating risk of mortality among joint groups of insulin and IGF-1 levels, those in Group I with NGT had 20% increased adjusted odds of mortality (1.30 [1.01,1.71]), while in GI subjects, there was an insignificant increased odds of mortality (1.17 [0.84, 1.62]). Neither subgroup in Group 2 had significant adjusted odds of mortality relative to Group 4. Group 3 subjects with GI had an adjusted, insignificant 24% increased odds of mortality (1.24 [0.91, 1.70]) compared to 70% increased odds in NGT subjects (1.69 [1.18, 2.42]). Conclusion: The differences in odds of all-cause mortality across IGF-1 quartiles in glucose tolerant vs. intolerant individuals suggests that IGF-1 may play less of a protective role in Type II diabetes and prediabetes. Among those with normal glucose, higher insulin levels, regardless of IGF-1 levels, was associated with all-cause mortality. This association did not hold in those with glucose intolerance. ]]> <![CDATA[SAT-618 Nattokinase to Improve Insulin Sensitivity and Weight Loss in Women with Obesity +/- Diabetes]]>