GLYCEMIC CONTROL IN DIABETES MELLITUS

Note-Glycemic control is particularly important in women of childbearing age at risk for (or planning) pregnancy

I. Targets
  1. ADA Recommentations (figures are for whole blood glucose):
    1. Preprandial glucose 80-120 (change tx if > 140 or < 80)
    2. HS glucose 100-140 (change tx if > 160 or < 100)
    3. HbA1c < 7 (change tx if > 8)

II. Clinical impact of glycemic control

  1. Decreases risk of hyperosmolar coma
  2. Decreases symptoms (blurred vision, polyuria, polydipsia, etc.)
  3. Improves lipid profiles
  4. UK Prospective Diabetes Study--Demonstrated reduced risk of microvascular complications with tight control in newly-diagnosed patients with type 2 DM
    1. "UKPDS 33"--Lancet 352:837, 1998--JW

      1. 3,867 pts (median age 54y) with newly diagnosed type 2 DM randomized to intensive (randomly assigned to chlorpropamide, glibendamide, glipizide, or insulin, adjusted to keep fasting glucose < 108 mg/dl) vs. conventional (meds as needed to keep fasting glucose < 270 mg/dl or to treat symptomatic hyperglycemia) treatment.

      2. Over median 10y f/u, median HbA1c was 7.0% in intensive group and 7.9% in conventional group.

      3. Primary outcome defined as any of diabetes-related adverse events, including cardiac, cerebrovascular, microvascular, renal, ophthalmologic, and glycemic events.  RR of primary outcome was 0.88 in intensive group (sig.); only specific outcomes that were significantly reduced were sudden death (RR 0.54), receipt of retinal photocoagulation (RR = 0.71), and cataract extraction (RR = 0.76). RR of any microvascular endpoint (renal failure, receipt of retinal photocoagulation, or vitreous hemorrhage) was 0.75 (sig.)

      4.  RR’s for DM-related death and all-cause mortality were 0.90 and 0.94 respectively; neither of the latter being significant.  No differences seen in any of the endpoints based on which tx used in the intensive group.  Sig. more hypoglycemic episodes and weight gain in intensive group

    2. "UKPDS 34"--Lancet 352:854, 1998--JW
      1. 1,704 overweight pts with type 2 DM randomized to intensive tx with metformin vs. conventional tx.
      2. Intensive group had sig. reduction in DM-related endpoints
      3. In comparison between metformin and other drugs (sulfonylureas or insulin), metformin group had fewer DM-related endpoints, less weight gain, and less hypoglycemia
    3. "UKPDS 35"--BMJ 321:405, 2000--JW
      1. 3,642 pts with newly dx'd type 2 DM randomized to intensive vs. conventional hypoglycemic Rx
      2. Compared subgroups of pts with final HbA1c of > 10%, 7-8%, and 6%; endpoints included any DM complication as well as DM-related death
      3. For all endpoints, risk was lower for lower HbA1c levels; there was no apparent "threshold" below which further lowering did not appear to be beneficial
  5. Diabetes Control and Complications Trial (NEJM 329:977-986, 1993)-Demonstrated reduced risk of microvascular complications with tight control in type 1 DM
    1. Compared "intensive" vs. "conventional" therapy for Type 1 DM & impact on vascular & neurologic complications in 1,441 pts 13-39y at enrollment followed for mean 6.5y. 
    2. Intensive: 3 or more daily insulin injections OR insulin pump, SMBG at least QID, target sugars 70-120 preprandial and < 180 postprandial
    3. Conventional: 1-2 daily insulin injections
    4. Intensive group had significantly better glycemic control by HbA1c & mean glucoses and significantly lower incidence of...
      1. New-onset retinopathy (RR 0.24)
      2. Progression of retinopathy (RR 0.46)
      3. New-onset microalbuminuria (RR 0.66)
      4. New-onset neuropathy (RR 0.31)
      5. HIGHER incidence of severe hypoglycemia (RR 3.3)
      6. HIGHER incidence of reaching > 120% ideal body weight (RR 1.33)
      7. NO sig. diff. in overall mortality, incidence of DKA, or quality of life (assessed through as pt questionnaire)
    5. In a follow-up report on 93% of the pts in the DCCT, over mean 17y f/u, intensive therapy pts had significantly reduced incidence of... (NEJM 353:2643, 2005--abst)
      1. Any cardiovascular disease (RR 0.58)
      2. Nonfatal MI, CVA, or cardiovascular death (RR 0.43)
  6. "ACCORD" Trial (NEJM 358:2545, 2008-abst)
    1. In a study of 10,251 Type 2 diabetics with (established CV disease) or (CV risk factor) randomized to intensive glucose-lowering therapy (target HbA1c < 6.0%) vs. "standard" therapy (target HbA1c 7.0-7.9%), after mean 3.5y f/u the study was discontinued because of the finding of sig. higher overall mortality in the intensive-therapy group (HR 1.22); there was no difference in the incidence of (nonfatal MI, nonfatal CVA, or cardiovascular death),
  7. "ADVANCE" Trial (NEJM 358:2560, 2008-abst)
    1. In a study in 11,140 Type 2 diabetics randomized to "intensive" vs. "standard glycemic control efforts (the former involving use of glicazide + other drugs to target HbA1c of < 6.6%), after median 5y f/u, the intensive control group had sig. lower incidence of major microvascular events (9.5% vs. 10.9%) but not major macrovascular events, cardiovascular death, or all-cause mortality
  8. "VADT" Trial (NEJM 360:129, 2009-abst)
    1. In a astudy in 1,791 Type 2 diabetics randomized to standard vs. intensive glycemic control targets (the latter consisting of a targeted 1.5% absolute reduction in HbA1c), over 5.6y  median f/u, the time to occurrence of (MI, CVA, cardiovascular death, HF, surgiery for vascular disease, inoperarable coronary disease, or amputation for ischemic gangrene) was not sig. diff between the two groups; neither was the incidence of microvascular complications.
  9. In a meta-analysis of five randomized trials (including ACCORD, ADVANCE, and VADT) with over 3-10y f/u and involving more than 33,000 pts with type 2 DM, "intensive" glucose-lowering tx was associated with sig. lowering of HbA1c and risk of non-fatal MI (RR 0.83) and CHD events (RR 0.85) but no effect on CVA or all-cause mortality (Lancet 373:1765, 2009-JW)
  10. In a meta-analysis of five studies each involving over 500 subjects, "intensive" glycemic control was associated with sig. reduced risk for cardiovascular events (RR 0.90) but not cardiovascular- or all-cause mortality, and was associated with a sig. increase in incidence of severe hypoglycemia (RR about 2, absolute risk increase 3.9% over 5y) (Ann. Int. Med. 151:394, 2009-JW)

III. Temper attempts at normoglycemia with caution to avoid hypoglycemia, particularly in patients on insulin and sulfonylureas

IV. Antidiabetic Medication-Click link for details

V. Dietary interventions

  1. Medical nutrition therapy--basically just emphasizes balance between fat, carbohydrate, and protein and not too much saturated fat
    1. 50-60% of calories from carbohydrate, mostly complex rather than simple
    2. 12-20% of calories as protein (less with renal disease)
    3. Restrict saturated fats to 10% of calories
  2. High intake of dietary fiber improves glycemic control in diabetics
    1. In a randomized crossover study, 13 pts with type 2 DM at either high fiber (25g/d insoluble + 25g/d soluble) vs. moderate-fiber diet (8g/d soluble + 16g/d insoluble)--at 6wks, overall daily plasma glucose levels were 10% lower in the high-fiber group (NEJM 342:1392, 2000--JW)
  3. "Mediterranean diet" for glycemic control in diabetics
    1. In a a study in 215 adults with newly-diagnosed type 2 DM, not yet on medication, all with BMI > 25 kg/m2 and HbA1c < 11% randomized to a low fat "AHA" diet (no more t han 30% of calories from fat/no more than 10% of calories from saturated fat) vs. a low-carbohydrate "Mediterranean-style" diet (lots of vegetables, very little red meat, at least 30% of calories from fat with 30-50g olive oil daily, no more than 50% of calories from complex carbohydrates) with identical daily caloric content.  At 18mos, sig. fewer in the Mediterranean group needed medicaation to achieve HbA1c < 7% (12% vs. 24%); Mediterranean group also had sig. greater weight loss (by 2kg). (Ann. Int. Med. 151:306, 2009-JW)
    2.  

VI. Monitoring glycemic control--"Daily monitoring is especially important for patients treated with insulin or sulfonylureas to monitor for hypoglycemia...optimal frequency [of glucose monitoring] in Type 2 diabetes is unknown" (ADA)

  1. Home glucose monitoring
  1. Whole blood glucose levels are 10-15% lower than plasma glucose levels
  2. Good idea to check Qac & Qhs when changing therapy
  3. Post-lunch glucose levels have greatest correlation with overall glycemic control
    1. 66 pts 40-78yo with type 2 DM on diet therapy or oral meds. All had plasma glucose measured x 4 on a single day--at 8am and Q3h x3. All had their usual breakfasts at 8am and their usual lunches around 12pm. Also had HbA1c checked < 10 afterward. In multiple linear regression analysis, only the "postlunch" (2pm) and "extended postlunch" (5pm) plasma glucose values were sig. ass'd with HbA1c and the correlation with HbA1c was strongest for the postlunch value. A postlunch plasma glucose < 151 was predictive of HbA1c < 7.0 and a postlunch plasma glucose < 207 was predictive of HbA1c < 8.6% ("expected values" derived from regression lines). Using these thresholds, the postlunch plasma glucose had a sensitivity of 73% and a specificity of 92% for predicting "poor control" of DM (defined as HbA1c > 8.4%). In contrast, the prebreakfast plasma glucose (8am, > 174mg/dl) was only 69% sensitive and 85% specific for predicting poor diabetic control. (Diab. Care 20:1822, 1997)
  1. HbA1c-reflects mean blood glucose over previous 2-3mos--ADA recc's checking at least 2x/yr in pts with stable glycemic control, more frequently in pts with change in tx or not meeting goals of glycemic control
  2. Glycated Serum Protein--reflects mean blood glucose over previous 1-2wks