Type of insulin Onset (H) Peak (h) Duration (h)
Aspart (Novolog) 0.25 1-3 3-5
Glulisine (Apidra) 0.25 1 2-4
Lispro (Humalog) 0.25 1 3.5-4.5
Regular (Humulin R, Novolin R) 0.5-1 2-3 6-8
NPH (Humulin N, Novolin N) 0.5 2-4 14-18
Lente 1-3 6-15 18-26
Ultralente 6-19 12-16 20-24
Glargine (Lantus) 1.5h none 24
Detemir (Levemir) 1 none 24


I. Total daily dose-typical requirements

  1. Type I: 0.5-0.8U/kg/d (up to 1.5 in adolescents)
  2. Type II: 1-1.5U/kg/d; some may require >100U/d if on insulin monotherapy

II. Insulin monotherapy

  1. Conventional regimen: divided into 2 doses (12h apart): 2/3 before breakfast; 1/3 before supper
  1. Initiate at 0.3-0.5U/kg/d
  2. Reg. insulin should be 1/3 of am dose, 1/2 of pm
  3. The rest should be NPH or Lente
  4. am Reg covers lunch, am NPH covers dinner, PM reg covers nt-snack, PM NPH covers bkfst
  5. Adjust dose based on am fasting and pre-dinner reading
  6. Prebreakfast hyperglycemia is common ("Somogyi effect")
  7. Sig. risk of nocturnal hypoglycemia
  1. Intensive regimen: 3 doses
    1. Initiate at 0.3-0.5U/kg/d
    2. NPH/Reg in am (2/3 of daily dose), Reg at dinner, NPH at bedtime
    3. Better control of breakfast & dinner postprandial glycemia than w/conventional
    4. Better prevention of prebreakfast hyperglycemia with decreased risk of nocturnal hypoglycemia
  1. SQ insulin pumps--used more commonly in Type 1 diabetics
  2. If pt is unreliable, consider just BID NPH w/no regular
  3. Some have used BID ultralente
  4. Regular insulin adjustments before meals: add 1U for each 40mg/dl above 100

III. Combination insulin/oral hypoglycemic regimens in Type 2 DM:

  1. One recommended regimen (Diab. Med. 9:826, 1992; cited in AAFP monograph)
    1. Start with 5-10U NPH @ HS (10:00-11:00 pm)
    2. Check a.m. fasting FSG--set goal, e.g. 100-150 (may also want to check glu at 3-4am when peak effect will be occurring, until dose established)
    3. Adjust until goal is met (incr. insulin 2-3U Q3-4d), then
    4. Start checking pre-dinn. FSG--if high, give an oral hypoglycemic in a.m. on top of HS insulin
  2. The following 3 regimens were ass'd with no sig. differences in HbA1c, weight changes, or lipid levels in a 6mo randomized trial in 95 pts, mean age 68, with Type 2 DM poorly controlled with maximal dose glyburide (Diab. Care 19:1326, 1996-JW):
    1. BID insulin (reg/NPH) and no glyburide
    2. HS insulin (NPH) + BID glyburide
    3. AM insulin (NPH) + BID glyburide
  3. BID NPH  (before breakfast and supper) was associated with sig. better improvements in HbA1c than QD ultralente (before dinner) in a 6mo trial of 60 pts with type 2 DM poorly controlled on oral drugs (Diab. Care 23:1612, 2000--JW)
  4. In a 24-week randomized trial in 756 overweight type 2 diabetes with HbA1c 7.5-10% despite oral therapy with 1 or 2 drugs, insulin glargine vs. NPH insulin, either in a single HS dose, starting at 10U, titrated to achieve fasting plasma glucose < 100 was ass'd with similar drops in HbA1c and fasting glucose but sig. lower incidence of hypoglycemia in glargine group (Diab. Care 26:3080, 2003--JW)
  5. In a 24-week randomized trial in 188 pts with type 2 diabetes and HbA1c > 8.0% despite therapy with 2 oral agents randomized to a 3rd oral med vs. (BID 70/30 insulin + metformin), no sig. diff. in mean reduction in HbA1c levels (Diab. Care 26:2238, 2003--JW)
  6. In a 24-week randomized trial in 695 pts with type 2 diabetes on oral agents only with HbA1c 7.5%-10.5%, taken off their oral meds and switched to glimepiride with either glargine Qam, glargine QHS, or NPH QHS, with adjustment of insulin doses to achieve fasting normoglycemia, HbA1c reductions were sig. greater with am glargine than the other two groups (1.2% vs. 1.0% and 0.8%, respectively); incidence of symptomatic hypoglycemia less with HS glargine than the other two groups (43% vs. 56% with am glargine and 58% with HS NPH) (Ann. Int. Med. 138:952, 2003--JW)
  7. In a study in 371 pts with poorly controlled type 2 DM despite oral therapy, randomized to (metformin + glimepiride + Qam insulin glargine) vs. (insulin 70/30 BID), the metformin/glimepiride/glargine group had sig. greater decrease in HbA1c at 24wks (1.6% vs. 1.3%), and a lower incidence of hypoglycemic events (Diab. Care 28:254, 2005--JW)
  8. In a study of 233 pts with poorly controlled type 2 DM despite oral therapy, all on metformin, randomized to (QHS insulin glargine) vs. (insulin 70/30 BID); at 24wks, the 70/30 group had sig. greater mean reductions in HbA1c (2.8% vs. 2.4%) and also higher incidence of hypoglycemia (Diab. Care 28:260, 2005--JW)
  9. In a study in 217 pts with HbA1c 7.5%-11% on treatment with a sulfonylurea + maximum-dose metformin, randomized to addition of  insulin glargine (starting at 10U/d and titrating up) vs. rosiglitazone (up to 8mg/d), at 24wks there was no sig. diff. in decline in HbA1c or severe hypoglycemia; weight gain and incidence of edema and mild-mod hypoglycemia were sig. greater in rosiglitazone group (Diab. Care 29:554, 2006--JW)
  10. In a study in 708pts with type 2 DM and inadequate glycemic control (HbA1c 7-10%) on (metformin + sulfonylurea) randomized to one of the following regimens for the first year

    • Insulin aspart (NovoRapid) TID with meals

    • "Biphasic" insulin aspart (NovoMix 30) BID

    • Insulin Detemir QD (could increase to BID if needed)

    In year 2 of the study, sulfonylureas were replaced by a 2nd insulin if glycemia not controlled (true for 90% of study subjects) and if in the biphasic group, a mid-day prandial dose was added.  After 3y, the biphasic group had sig. lower likelihood of achieving HbA1c of 6.5% or less (31.9% compared with 44.7% with the meals-only initial regimen and 43.2% with the basal-only initial regimen).  Subjects in the basal group had sig. lower weight gain (3.6kg) compared with the biphasic (5.7kg) or meals-only (6.4kg) groups and lower incidence of hypoglycemic events (1.7% vs. 3.0% and 5.5%, respectively).

    (Treat to Target in Type 2 Diabetes" ("4-T") Trial; NEJM 361:1736, 2009-FP News & JW)