For the past 10 months, I have been working in internal medicine. The closest endocrinology practice is 75 miles away, and there is a three-month wait for new patient consults. I have many patients with type 2 diabetes who are taking maximally effective doses of metformin plus glipizide ER or glimepiride. This combination of medications is cost-effective for my patients, but when their A1C is no longer at goal (ie, greater than 7%), I need additional therapy.
The majority of my patients are unable to afford nongeneric diabetes medications (thiazolidinediones, dipeptidyl peptidase-4 inhibitors, glucagon-like peptide-1 agonists), so I resort to insulin. I get conflicting information that I should start basal insulin and “fix fasting blood sugars first.”
Unfortunately, many of my patients can’t afford basal insulin, and I know that twice-daily neutral protamine hagedorn (NPH) insulin is not a replacement for it. I’m aware that NPH insulin given in the evening can cause nighttime hypoglycemia, but it is affordable for my patients.
Q: What is the best choice of insulin when a patient can no longer maintain an A1C of < 7% on maximally effective doses of metformin and a sulfonylurea?
Metformin and sulfonylureas are available in generic form, affordable, and effective for glucose management. Unfortunately, with the progressive nature of type 2 diabetes, these agents usually lose their efficacy in a few years (if not sooner). We are then forced to add a nongeneric oral agent or an injectable agent such as insulin or a glucagon-like peptide-1 (GLP-1) agonist.
The GLP-1 agonists are effective for lowering blood glucose and can help with weight loss, but they are not available in generic form. Basal insulin (glargine, detemir) and “rapid-acting” prandial insulin (lispro, aspart, glulisine) more closely mimic normal physiologic insulin secretion than do the “older” insulin preparations (“short-acting” regular, NPH, and premixed NPH/regular insulin). Since basal and rapid-acting insulin preparations are unavailable in generic form, there are many times when we have to rely on the older, somewhat less physiologic types of insulin.
We must remember that patients will take only what they can afford. Otherwise, they may resort to taking less than the prescribed dose of insulin to make their prescription last longer.
Q: What are some guiding principles about which insulin to start when glucose control targets are not being achieved?
Health care providers must not start a patient on insulin based solely on fasting blood sugars, without first evaluating blood sugars at other times of the day. A patient’s fasting hyperglycemia may be a result of the dawn phenomenon, “carry over” from bedtime hyperglycemia, or panhyperglycemia (“high around the clock”).
First, you will need to see blood sugar readings at different times of the day—including fasting, before meals, two-hour postprandial, and bedtime—to isolate when nongoal blood sugars are occurring. Remember, when addressing blood sugars, “you can only fix what you see.”
Caution must be exhibited when putting a patient on basal insulin to “fix fasting blood sugars first” without knowing what is happening at bedtime and throughout the day. The patient may already be having low-normal blood sugars during the day as a result of increased physical or work activity, so giving basal insulin to “fix” fasting blood sugars will likely cause daytime hypoglycemia.
Helpful guidance about what insulin to start is provided by the popular Monnier data (Diabetes Care. 2003;26:881-885), which show the contribution of fasting and postprandial glucose values to A1C. As A1C increases, the proportion attributable to fasting plasma glucose also increases. When A1C decreases below ~7.6%, however, postprandial glucose has the greatest influence on A1C. Understanding these data should allow one to conclude that if A1C is greater than 8%, the fasting glucose has an equal, if not greater, influence on A1C, compared to postprandial glucose.
A frequent problem I encounter is that patients snack in the evening, which causes high blood sugars at bedtime and subsequent high fasting blood sugars. After instructing them not to consume any calories after dinner, I have patients check a bedtime blood sugar, followed by a fasting blood sugar the next morning, for a few days to identify the nighttime glucose trend. If the fasting blood sugar is elevated, but is consistently lower than the high bedtime blood sugar, then the most appropriate treatment would be prandial insulin with dinner and not bedtime NPH.
Rapid-acting prandial insulin is given zero to 10 minutes before the meal, but short-acting (regular) insulin is given about 30 to 45 minutes before the meal, since it has a slower onset of action. If the patient consumes a high-carbohydrate dinner, it would make sense to try a lower (30- to 60-gram) carbohydrate dinner. This may be enough to prevent postprandial hyperglycemia after dinner, and therefore the need for prandial insulin.
If the fasting blood sugar is elevated and consistently equal to or greater than the bedtime blood sugar, giving NPH at bedtime will address this dawn phenomenon. In this case, NPH is best administered around 10 or 11 pm to maximize its glucose-lowering effects in the dawn hours (~ 3 to 8 am), when there is a natural increase in the hormones (growth hormone, cortisol, and glucagon) that increase blood sugars. Caution should be exercised to avoid giving NPH at suppertime or too early before bed, as this will increase the risk for nocturnal hypoglycemia.
Another factor to keep in mind is that patients may snack immediately before or right after they take their bedtime blood sugar (ie, not accurately reflecting the actual peak in blood sugar at night). If this is the case, you may mistakenly think they have the dawn phenomenon. Regarding risk for nocturnal hypoglycemia, it is not recommended to routinely give prandial insulin for hyperglycemia at bedtime. The proper approach to glucose management should be proactive (treat to prevent), not reactive (treat to fix).
Q: Can you provide some brief examples of which insulin to use and why?
1. If the bedtime and daytime blood sugars are on goal but the fasting is high, fix the dawn phenomenon with NPH insulin at bedtime. In this case, fixing fasting hyperglycemia may increase the risk for daytime hypoglycemia if the patient is on a secretagogue (sulfonylurea, glinide), so you may need to decrease the dose of the secretagogue accordingly.
2. If blood sugar is high only after a meal (or meals), use prandial insulin only for that meal.
3. If blood sugars are mild to moderately high “around the clock,” the addition of once-daily basal insulin generally works very well.
Note: Although there are other insulin treatment regimens (basal + prandial and premixed once or twice daily,) it is not possible to fully discuss these more complex regimens in a single column. This article is intended for the “when, why, and how” to add once-daily insulin to the regimen of a patient whose A1C is in the 7% range on maximal doses of metformin plus a secretagogue.