Comparative Effects of Randomized Second-line Therapy for Type 2 Diabetes on a Composite Outcome Incorporating Glycemic Control, Body Weight, and Hypoglycemia: An Analysis of the Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE).

OBJECTIVE: In Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE) (5,047 participants, mean follow-up 5.0 years), differences in glycemic control were demonstrated over time among four randomized therapies added to metformin. Weight gain and hypoglycemia are also important outcomes for people with type 2 diabetes. We compared the effects of the four randomized GRADE medications on a composite outcome incorporating glycemic deterioration, weight gain, and hypoglycemia. RESEARCH DESIGN AND METHODS: The composite outcome was time to first occurrence of any of the following: HbA1c >7.5%, confirmed; ≥5% weight gain; or severe or recurrent nonsevere hypoglycemia. Secondary analyses included examination of individual components of the composite outcome, subgroup effects and potential mediators, and treatment satisfaction. Cumulative incidence was estimated with the Kaplan-Meier estimator. Cox proportional hazards models were used to assess pairwise group differences in risk of an outcome. RESULTS: Risk of reaching the composite outcome (events per 100 participants per treatment year [PTYs]) was lowest with liraglutide (19 per 100 PTYs) followed by sitagliptin (26 per 100 PTYs), glargine (29 per 100 PTYs), and glimepiride (40 per 100 PTYs); all pairwise comparisons were statistically significant. The order was the same for risk of weight gain and hypoglycemia, but risk of glycemic deterioration was lowest with glargine, followed by liraglutide, glimepiride, and sitagliptin. No significant heterogeneity in risk of composite outcome was detected across prespecified covariates. Participants who reached the composite outcome had modestly but significantly lower treatment satisfaction. CONCLUSIONS: Among participants treated with common second-line drug classes for type 2 diabetes, the liraglutide group had the lowest and glimepiride the highest risk of reaching a composite outcome encompassing glycemic deterioration, weight gain, and hypoglycemia. These findings may inform decision-making regarding type 2 diabetes therapy.

INSULIN TIMING: A PATIENT-CENTERED APPROACH TO IMPROVE CONTROL IN TYPE 1 DIABETES.

OBJECTIVE: The goal of insulin therapy in type 1 diabetes (T1D) is to reduce hemoglobin A1C (A1C) to ≤7.0% (53 mmol/mol) with minimal hypoglycemia. We investigated the possibility that "insulin timing" would improve A1C without incurring severe hypoglycemia in volunteers with T1D over a 6-month observation period. METHODS: Forty healthy adult volunteers with T1D were randomly assigned for 6 months to either a control group or an insulin timing group. The primary endpoint was the difference in A1C between the two groups. As a secondary endpoint, both groups were further divided to assess the importance of the baseline A1C in determining the response to timing. The insulin timing algorithm altered the time when the meal dose of insulin was injected or infused from 30 minutes before the meal to 15 minutes after the meal, depending upon the premeal blood glucose concentration. RESULTS: An improvement in mean A1C was observed in the timing group compared with no change in the control group, but this improvement did not reach statistical significance (P>.05). In contrast, when the two groups were analyzed according to baseline A1C, the timing volunteers with baseline A1C values in the upper half (separated by the A1C median of 7.45% [57.9 mmol/mol]) of the timing group had a more robust response to timing (decline in A1C) than the upper half of the control group (P<.05). CONCLUSION: Insulin timing is a patient-centered translational approach that is safe and effective in improving A1C in individuals with T1D with an elevated A1C. ABBREVIATIONS: A1C = hemoglobin A1C ANOVA = analysis of variance CGM = continuous glucose monitoring CSII = continuous subcutaneous insulin infusion MDI = multiple daily injection T1D = type 1 diabetes.

The Effectiveness and Risks of Programming an Insulin Pump to Counteract the Dawn Phenomenon in Type 1 Diabetes.

OBJECTIVE: Continuous subcutaneous insulin infusion (CSII) programming for an early morning increase in insulin delivery is frequently recommended to counteract the rise in glucose prior to breakfast (dawn phenomenon). However, both the effectiveness and safety of this approach have not been tested in the ambulatory setting. Using continuous glucose monitoring, we investigated the safety and effectiveness of early morning CSII programming for management of the dawn phenomenon in subjects with type 1 diabetes. METHODS: We conducted a controlled, observational eight-month longitudinal study of type 1 diabetic patients (n=40). Reproducibility of the dawn phenomenon was determined in subjects treated with multiple daily injections of insulin (MDI, n=12) and those on CSII who did not program an early morning increase in insulin delivery (CSII non-programmers, n=8). The effects of early morning CSII programming were determined by comparing rates of the dawn phenomenon and hypoglycemia in CSII non-programmers versus CSII-users who programmed an early morning increase in insulin delivery (CSII programmers, n=20). RESULTS: The dawn phenomenon occurred in all tested subjects to a variable extent (median rate 56% of nights). CSII programming was not associated with a reduction in the occurrence of the dawn phenomenon (42%) compared to non-programmers (48%) (P=0.47) nor in the magnitude of the dawn phenomenon. Hypoglycemia occurred more frequently in the CSII programmers (37%) compared to non-programmers (18%) (P=0.001). CONCLUSION: The dawn phenomenon occurs unpredictably; therefore, early morning CSII programming for a fixed increase in early morning insulin delivery is ineffective and may be hazardous to the patient.

Is the masked continuous glucose monitoring system clinically useful for predicting hemoglobin A1C in type 1 diabetes?

BACKGROUND: The masked continuous glucose monitoring system (Masked-CGMS) differs from standard CGMSs in three ways: (1) there is no feedback to the user so that no immediate regimen changes can be made; (2) it can only be worn for up to 5 days; and (3) there are no alarms to warn of hyperglycemia or hypoglycemia. Since 2008 masked-CGMS has become popular for identifying reasons that a patient's hemoglobin A1C does not correlate closely with his or her capillary blood glucose measurements. To date only one study addressing the clinical utility of Masked-CGMS for improving A1C in diabetes has been published. No studies are available specifically examining the variability and correlation of Masked-CGMS and A1C. SUBJECTS AND METHODS: We performed 156 Masked-CGMS studies (40 patients studied sequentially a maximum of four times each) in type 1 diabetes patients. We then analyzed the resulting interstitial glucose levels obtained from the Masked-CGMS compared with an A1C measurement performed within 1 week of the Masked-CGMS study. RESULTS: There was a very low correlation between the A1C and the Masked-CGMS-derived mean interstitial glucose level. This statistic did not provide sufficiently predictive information to be clinically useful for changing an individual patient's intensive insulin therapy regimen. CONCLUSIONS: Our data demonstrate that a very weak correlation exists between 5 days of masked CGMS and a concurrently measured A1C level. For the individual type 1 diabetes patient, this relationship would unlikely to be clinically useful in altering the individual's treatment regimen.

Does short-term vitamin C reduce cardiovascular risk in type 2 diabetes?

OBJECTIVE: Vitamin C is a powerful antioxidant that is potentially useful in the prevention of atherosclerosis in diabetic individuals. However, the mechanism(s) of vitamin C's anti-atherosclerotic effects in vivo are unresolved, and clinical trials in nondiabetic individuals have yielded conflicting results. Therefore, we performed 32 studies in a randomized, crossover, dose-response trial in 8 volunteers with type 2 diabetes to determine the effects of vitamin C on serum vitamin C levels, lipids, inflammation, and coagulation. METHODS: Well-controlled, type 2 volunteers received, in randomized order for 2-week periods, each of the following: 1) no supplemental vitamin C, 2) low-dose vitamin C (250 mg/day), 3) medium-dose vitamin C (500 mg/day), and 4) high-dose vitamin C (1,000 mg/day). A high-caloric content lunch was ingested during each study arm to enhance oxidative stress. Serum vitamin C levels and atherosclerotic risk factors including lipids and markers of oxidative stress, inflammation, and hypercoagulation were determined. RESULTS: Serum vitamin C levels increased significantly at all dosages. In addition, the high-caloric content meal resulted in acute elevations of glucose, insulin, and triglycerides for several hours postmeal. However, no significant effect of vitamin C was observed on lipid parameters or any of the surrogate markers of oxidative stress, inflammation, or hypercoagulability. CONCLUSION: Our study suggests that if vitamin C does have anti-atherosclerotic effects in diabetes, it does not exert them through the traditional pathways identifiable by established surrogate markers of cardiovascular risk.

Impaired orthostatic response in patients with type 2 diabetes mellitus after 48 hours of bed rest.

OBJECTIVE: To compare the effect of bed rest on orthostatic responses of patients with type 2 diabetes mellitus and nondiabetic control subjects. METHODS: Six patients with type 2 diabetes and 6 non-diabetic control subjects underwent 48 hours of bed rest and 48 hours of ambulatory activity in randomized order. A 10-minute tilt test was conducted before and after each period of hospitalization, and cardiovascular responses to 80 degrees head-up tilt were analyzed with use of a 2-factorial (study group and bed rest condition) analysis of variance design. We hypothesized that patients with diabetes would experience more severe changes in orthostatic response after bed rest. RESULTS: No significant differences in orthostatic responses were observed before bed rest between control subjects and patients with diabetes. After bed rest, control subjects had a greater (P = .01) increase in heart rate during tilt in comparison with before bed rest (before versus after bed rest, 9 +/- 4 versus 24 +/- 7 beats/min) and maintained their blood pressure during tilt. After bed rest, patients with diabetes did not have a compensatory increase in heart rate and had a greater (P = .02) decline in systolic blood pressure during tilt in comparison with before bed rest (before versus after bed rest, -7 +/- 10 versus -21 +/- 11 mm Hg). Their arm and leg skin vasomotor responses (laser Doppler flowmetry) during tilt were not altered after bed rest and were similar to those in control subjects before and after bed rest. CONCLUSION: Cardiac neuropathy in patients with type 2 diabetes may prevent a compensatory heart rate response after bed rest deconditioning and result in a more severe orthostatic response. A greater decrease in blood pressure with upright tilt is evident after a relatively short period of bed rest.

Metabolic effects of 2 days of strict bed rest.

OBJECTIVE: To examine the possibility of whether 2 days of strict hospitalized bed rest would alter the metabolic profile (including insulin resistance as calculated by the quantitative insulin sensitivity check index or QUICKI) in both normal subjects and patients with type 2 diabetes in comparison with 2 days of normal activity. METHODS: The design of this pilot study was a randomized, crossover protocol that evaluated the effects of strict bed rest versus normal activity in 5 healthy normal subjects and 5 healthy patients with type 2 diabetes. All study participants completed a screening visit for assessment of baseline health. RESULTS: All 10 study subjects completed the protocol without adverse events. Fasting plasma glucose, insulin, and C-peptide levels as well as several known risk factors for atherosclerosis were unchanged in both the subjects without diabetes and the patients with type 2 diabetes after 2 days of hospitalized bed rest. Insulin resistance demonstrated no significant change during the 48 hours of bed rest when compared with the mean value at baseline. CONCLUSION: This study demonstrates that 48 hours of bed rest has no significant effect on insulin resistance or standard metabolic variables in normal subjects and patients with type 2 diabetes. Therefore, achieving good glucose control in patients hospitalized for a period of 2 days or less does not necessitate early ambulation to prevent an increase in insulin resistance.