Initiation of Insulin Therapy in Type II Diabetes

 In Endocrinology, Gastrointestinal

Ryan Bradley, ND

Despite steps taken to increase exercise, reduce stress and improve poor dietary habits in an effort to reduce blood glucose, many patients with Type II diabetes remain poorly controlled. Poor glycemic control dramatically increases risk of cardiovascular events and microvascular complications, including end-stage renal disease, in patients with Type II diabetes (Stratton et al., 2000). There is a time and place for high-order therapeutics in the management of patients. Oral diabetes medications have limitations in their effectiveness and all have side effects of varying severity. A recent study by Kahn et al., called A Diabetes Outcome Progression Trial (ADOPT), confirmed that reduced beta-cell function results from sulfonylurea use (e.g., glyburide), yet increased congestive heart failure results from both metformin and thiazoladinedione use compared to glyburide (Kahn et al., 2006). Recommendations by the American Diabetes Association state that metformin should remain as the first-line oral agent for Type II diabetes, but some patients do not tolerate the nausea and other gastrointestinal side effects of the medications, leaving few options (Nathan et al., 2006). Newer injectable incretin analogs exist, such as the glucagon-like peptide 1 (GLP-1) agonist exenatide. These are useful and rather “clean” in terms of side effects, but are relatively expensive. Exogenous insulin remains a viable choice in diabetes management, and with new advances in insulin preparations and needle technology, insulin is easy for both the prescribing doctor and the patient to use. 

Advantages of Insulin

In the absence of complete correction of the metabolic environment, diabetes is a progressive disease. The chicken and the egg argument applies to diabetes; which comes first, the insulin resistance or the insulin deficiency? Regardless, diabetes can be viewed as a relative insulin deficiency – either a lack of adequate production or a lack of a robust enough insulin response to overcome the resistance that is present. However, unlike thyroid hormone deficiency, patients are not eager to use insulin to replace their own production. For many patients, insulin seems taboo, or perhaps terminal, rather than a desirable treatment. Much of the reservation comes from insulin being an injectable therapy and requiring careful self-monitoring. (Although inhaled insulin is now available, long-term studies of safety are unavailable, so it will not be discussed here.)

Insulin treatment has several distinct advantages. Using insulin allows more flexibility in dietary choices, including food like fruits that are high in sugar yet have health benefits; fruit is difficult to include in a severely carbohydrate-restricted diet without negative impact on blood glucose. One review suggested that switching from oral agents to the use of exogenous insulin also increased quality of life in patients with Type II diabetes (DeWitt and Hirsch, 2003). Many patients feel physically better once they start insulin therapy because their skeletal muscle tissue is finally able to absorb glucose effectively. In many patients, muscle strength improves and overall fatigue declines. Most importantly, exogenous insulin can help patients stay in good glycemic control when the diet or exercise slips off, rather than blood glucose skyrocketing. Finally, for many patients in moderate glycemic control, a single injection of basal, or background, insulin can afford them a “honeymoon” period of good blood glucose control without requiring mealtime insulin. In my observations, when patients use basal insulin, follow a low glycemic index/load diet and exercise frequently, they can maintain excellent glycemic control with very low basal insulin requirements for a very long time.

Basal and Prandial Insulins

The goal of insulin replacement therapy is to mimic the body’s two-phase insulin response. We always have background insulin in circulation regardless of whether or not we have consumed food. Insulin levels are constantly in flux, balancing the effects of glucagon, catecholamines and cortisol. Assuming normal insulin production, the more robust insulin response occurs following meals proportional to the carbohydrate content and degree of absorption of carbohydrates in the meal. Therefore, the goal of insulin replacement therapy is to recreate the insulin response, reducing blood glucose and high glycemic variability. Replacing basal insulin then allows for the remaining endogenously produced insulin to be used during mealtime.

When and How to Start

I start patients on basal insulin when they have a hemoglobin A1c (HbA1c) greater than 7% after a six-month trial of lifestyle change and supplementation, especially if the patient is already taking metformin. If I am seeing the patient for the first time and he or she has been on multiple oral agents already or remains on them, and HbA1c is greater than 10%, I recommend insulin immediately. Some pioneers in the field recommend earlier initiation of basal insulin (some doctors start basal insulin at diabetes diagnosis) to achieve glycemic control and then work on lifestyle change to gradually reduce insulin over time. A large trial called ORIGIN (Outcome Reduction with Initial Glargine Intervention) is underway which, when completed in 2008, will provide useful data on the impact of early basal insulin in people with diabetes and pre-diabetes.

New basal insulins are extremely stable in their absorption following injection in the subcutaneous tissue. Twenty-four-hour basal insulins like insulin glargine and insulin detemir have an amazingly flat, stable release with little risk of hypoglycemia. These insulin preparations typically require only one nighttime injection, although some doctors prefer two injections at half the dose – one first thing in the morning and the other right before bed.

A published recommendation is to begin basal insulin dosing at 0.2units/kg body weight and then titrate up based on fasting blood glucose (FBG) levels (Chan and Abrahamson, 2003). The rules of thumb are:

  • Initiate a 24-hour basal insulin at 0.2units/kg body weight for three days, having the patient check his or her fasting morning glucose levels with a glucometer.
  • If FBG is >160, increase insulin dose by 6 units.
  • If FBG is >140, increase insulin dose by 4 units.
  • If FBG is >120, increase insulin dose by 2 units.

I use three-day intervals before increasing insulin doses during titration. Every patient’s insulin requirements are different, so dose titration is required. It can be helpful to have the patient keep a log of his or her blood glucose during titration. Usually you and the patient can determine the patient’s basal insulin ratio (i.e., 1 unit drops blood glucose Xmg/dL); this ratio varies individually.

Most recommendations suggest that a FBG <120mg/dL is adequate; however, I usually titrate to FBG <100mg/dL. Severely insulin-resistant patients can require upward of 80-100 units; however, most patients are in the 20-50 units range. Obviously, insulin requirements are proportional to total carbohydrate intake as well as insulin resistance, so dietary and exercise recommendations still apply for reducing insulin requirements. Insulin can always be titrated back down!

When to Start Prandial Insulin

Because insulin reserve progressively declines (unless the environment is corrected), patients are not always able to continue on basal insulin alone; many require mealtime insulin to avoid postprandial hyperglycemia. Postprandial hyperglycemia and glycemic variability contribute to risk for microvascular complications in diabetes (Hanefeld and Temelkova-Kurktschiev, 2002).

If a patient’s HbA1c returns at >7% despite the FBG consistently at <120mg/dL, this is a sign that mealtime insulin coverage is required (an HbA1c of 7% corresponds to an average blood glucose of ~170mg/dL). Ask the patient to check and record two-hour postprandial glucoses. If postprandial blood glucose is consistently >150mg/dL, a mealtime insulin is required for better control.

Several rapid-acting mealtime insulins exist. I tend to use insulin lispro. Other examples include insulin glulisine and insulin aspart.

There are several ways to begin mealtime insulin. One rule of thumb is that approximately 50% of our insulin is basal and 50% is prandial. Therefore, you can take the patient’s basal insulin dose and divide the units by three for each meal. This is an oversimplification for most patients, as it assumes an equal distribution of carbohydrates in all three meals (and a similar activity level).

A better, more precise way to titrate prandial insulin is by having patients find their rapid-acting insulin-to-carbohydrate ratio; this requires learning to “count” carbohydrates in their meals at a glance and anticipate insulin requirements (15g carbohydrate = one serving). See the accompanying sidebar for instructions.

The matrix of other micro- and macro-nutrients impact carbohydrate absorption and, therefore, impact insulin requirements for rapid-acting insulin, so there is a bit of art in the use of mealtime insulin. I encourage patients to monitor and record meals and blood glucoses frequently during the initial titration phase. The long-term idea is to not only get postprandial blood sugar to goal, but also to reduce the variability in blood glucose throughout the day; this is still best accomplished with a low glycemic index and low glycemic load diet. Consistency in dietary content and meal timing helps reduce variations in insulin requirements, but can get boring for patients who enjoy food!

What About Exercise?

Exercise makes the management of prandial insulin in patients with diabetes more complex, but it is a complexity to be encouraged. Once again, consistency in exercise during the insulin titration phase is key to learning how the individual patient responds. I typically recommend patients evaluate how 30 minutes of both aerobic and anaerobic exercise impact their blood sugars. For most people, 45-60 minutes of exercise following meals reduces their prandial insulin needs by approximately 50%. To test this conservatively, have the patient administer 50% of his or her estimated prandial insulin requirement for a given meal, eat and then exercise aerobically for 30 minutes and test blood sugars at the same two-hour postprandial time point. (I recommend repeating this with anaerobic exercise, as the responses are not the same for most people.) If blood glucose is still too high, the patient requires more insulin with meals or a longer workout (you should be able to compare previous mealtime requirements with 50% of normal insulin plus exercise to determine the required adjustments). As patients exercise more consistently, insulin requirements drop, so additional adjustments are needed frequently. Additionally, some insulin-dependent patients choose to eat only protein-based foods and snacks for days when they are going to be exercising the majority of the day (e.g., day hikes).

Insulin Supplements

Insulin supplements are small doses of insulin used to offset miscalculations of insulin needs following meals or exercise. For example, a patient reduced his or her mealtime insulin by 50% in anticipation of 30 minutes of exercise, but then only exercised for 15 minutes; or maybe the patient miscalculated his or her mealtime insulin needs and was left with hyperglycemia following exercise. Rules of thumb for insulin supplements are to use one unit per 50mg/dL above goal for a patient with Type II diabetes. For a patient with Type I diabetes, the rule of thumb is one unit per 20mg/dL above goal (DeWitt and Hirsch, 2003). These may need to be adjusted based on the individual. Supplements can be used in addition to mealtime insulin (e.g., preprandial glucoses are above goal, so the supplement of insulin can be added to the mealtime insulin for the next meal), following exercise (e.g., post-exercise glucose is higher than expected) and postprandially but prior to the next meal (e.g., three hours postprandial glucose is too high and the next meal is still hours away).

Other Issues

  • Reducing blood glucose too quickly There is some evidence, although not definitive, that too rapid of blood glucose reduction may worsen retinopathy in patients with this complication. There also are hypotheses that rapid reductions in blood glucose may induce a relative metabolic ischemia in the cardiac muscle. Finally, patients certainly experience relative hypoglycemia (i.e., they become symptomatic) if blood sugar is reduced too rapidly. Because of these factors, I try to start slowly, beginning with basal insulin only and then adding prandial insulins as needed. Also, if patients are symptomatic, it is prudent to slow the dose titration phase.
  • Advising patients about hypoglyemia Obviously, it is critically important to advise patients of the symptoms of hypoglycemia and get their commitment to monitor their glucose regularly! Sometimes if a patient is hypoglycemic, they are unable to handle the glucometer and check their blood to confirm. It is better to avoid the hypoglycemia and correct their glucose later (using a supplement) than to wait too long.
  • Prandial first? Some patients first present with uncontrolled postprandial glucose when their fasting glucoses are within target. In this situation, it is reasonable to begin a patient on prandial insulin only. However, often a lower glycemic load diet and better attention to resistance training can help reduce their prandial hyperglycemia as effectively. Since post-prandial hyperglycemia is directly correlated with skeletal muscle insulin resistance and reduced glucose disposal, reducing insulin resistance in the skeletal muscle via resistance exercise greatly reduces post-prandial glucose.
  • Pre-mixed insulin Pre-mixed insulin contains a combination of basal insulin and prandial insulin in one preparation. Although easier for patients because of the need for fewer injections, pre-mixed insulins do not provide consistent basal insulin replacement (due to shorter duration of action), and often result in some periods of inadequate insulin coverage. Once patients learn the increased flexibility basal/prandial replacement provides, they often forego the pre-mixes.
  • Insulin is not an excuse for complacency Unfortunately, some patients become less activated on their lifestyle change efforts when insulin is introduced. It is important to educate them that all lifestyle recommendations still apply; insulin should be dosed around a healthy lifestyle, not used to compensate for bad behaviors. Severe insulin resistance in an insulin-dependent patient is a considerable clinical challenge, and is not always one with a solution.

Insulin is an effective and sometimes necessary treatment for patients with diabetes. Used effectively, insulin reduces risk of cardiovascular disease and complications, and increases flexibility in diet and exercise choices. Modern insulin is stable and easy to titrate for individual patients. Insulin management, even if not the initial insulin prescription, is easy to improve and is within the skill set of a licensed ND.

Determining Insulin-to-Carbohydrate Ratio

I have the patient use a standard serving of rice as a carbohydrate control; 1/8 cup of rice is approximately 15g of carbohydrates. The process:

  • Day 1: Patient checks blood glucose before eating; eats 1/8-cup rice only; rechecks two hours later and records reading
  • Day 2: Patient checks glucose before eating and administers one unit of rapid-acting insulin; eats 1/8-cup rice; rechecks blood glucose two hours after eating and records the reading. This gives the patient the amount of glucose reduction achieved from that one unit of insulin.
  • Use the insulin-to-carbohydrate ratio to estimate mealtime needs for future meals.

Example: Day 1: Susie checks her glucose before eating: 110mg/dL; eats rice and rechecks after two hours to find it 200 mg/dL. Her blood sugar rose 90mg/dL from one serving of carbohydrates. Day 2: Susie checks her glucose before eating: 105mg/dL; administers one unit rapid-acting insulin and eats her rice, then rechecks her glucose after two hours to find it 165mg/dL. (You would have expected blood glucose to be 195mg/dL without the insulin, based on Day 1.) For Susie, one unit drops her glucose approximately 30mg/dL and covers 1/3 of one serving of carbohydrates. If she wanted to cover her one serving of rice completely, three units would have been required. Her ratio is about three units of rapid-acting insulin for one 15g serving of carbohydrates.


Ryan Bradley, ND is adjunct clinical faculty for the Diabetes & Cardiovascular Wellness Program at the Bastyr Center for Natural Health in Seattle. He also works as a clinical research fellow at Bastyr University, where he is conducting clinical trials of nutrients in Type II diabetes. Politically active, he serves as chairman of the Guidelines Development Subcommittee for the American Association of Naturopathic Physicians (AANP) and is a director of the Washington Association of Naturopathic Physicians (WANP). Concerned with improving healthcare quality, Bradley served on the Diabetes Clinical Improvement Team with the Puget Sound Health Alliance, a regional healthcare quality improvement coalition.


Stratton IM et al: Association of glycemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study, BMJ 321:405-412, 2000.

Kahn SE et al: Glycemic durability of rosiglitazone, metformin, or glyburide monotherapy, N Engl J Med 355(23):2427-43, 2006.

Nathan DM et al: Management of hyperglycemia in type 2 diabetes: A consensus algorithm for the initiation and adjustment of therapy: a consensus statement from the American Diabetes Association and the European Association for the Study of Diabetes, Diabetes Care 29(8):1963-72, 2006.

DeWitt DE and Hirsch IB: Outpatient insulin therapy in type 1 and type 2 diabetes mellitus: scientific review, JAMA 289(17):2254-64, 2003.

Chan JL and Abrahamson MJ: Pharmacological management of type 2 diabetes mellitus: rationale for rational use of insulin, Mayo Clin Proc 78(4):459-67, 2003.

Hanefeld M and Temelkova-Kurktschiev T: Control of post-prandial hyperglycemia – an essential part of good diabetes treatment and prevention of cardiovascular complications, Nutr Metab Cardiovasc Dis 12(2):98-107, 2002.


Recommended Posts

Start typing and press Enter to search