INITIATING BEDTIME

Sharon A. Watts, ND, RN-C, CDE, and Rossana D. Danese, MD

Patients and practitioners alike may be reluctant to begin a regimen of injectable insulin—even when oral agents fail to control blood . But such therapy can be highly effective if combined with carefully timed administration and patient education.

n the year 2000, the World mends that patients with type 2 DM SUPPORTING EVIDENCE Health Organization estimated maintain levels of glycosylated he- In 1992, Yki-Jarvinen and col- 5 that over 176 million people moglobin (HbA1C) below 7%. leagues conducted a short but im- I worldwide had type 2 To assist patients in meeting this portant trial to determine the mellitus (DM)—and predicted that ADA goal, we must consider all optimal mode of insulin administra- by 2030, that number would more facets of glycemic control, focusing tion for diabetic patients in whom than double.1 In the United States, not only on pharmacologic treat- oral therapy with and DM is the sixth leading cause of ment but also on patient education no longer controlled death among people aged 25 years regarding diet, exercise, and disease blood glucose effectively.6 The 153 and older,2 and its associated self-management as well. And when study participants were divided health care costs exceed $130 bil- a patient’s DM fails to be controlled randomly into five groups, each lion per year.3 by oral agents alone, we must be given one of the following treat- The 1995 United Kingdom Pro- willing to initiate insulin treatment— ment regimens: an oral hypo- spective Diabetes Study (UKPDS), a vital step frequently resisted by glycemic agent plus neutral one of the most important studies practitioners and patients alike. protamine Hagedorn (NPH) insulin on DM to date, demonstrated con- In this article, we guide primary given at 7:00 AM; an oral hypo- clusively what many clinicians had care providers in the timely and ef- glycemic agent plus NPH insulin known from clinical observation: ficient initiation of bedtime insulin. given at 9:00 PM; NPH insulin and that DM is a progressive disease We review existing data on this given at a 70:30 dose and that intensive management can treatment modality, explain the ra- ratio before breakfast and dinner; delay the onset of its devastating tionale behind its timing, and dis- NPH insulin given at 9:00 PM and complications.4 In light of these cuss the importance of patient regular insulin given before meals; findings, which have been sup- education in promoting tight or continued oral hypoglycemic ported by a number of subsequent glycemic control. therapy (the control group). studies, the American Diabetes Association (ADA) now recom- For patients with type 2 DM, the addition of

Dr. Watts is a family nurse practitioner, diabetes educator, and study coordinator at the Louis Stokes bedtime NPH insulin is the optimal treatment Cleveland VA Medical Center, Cleveland, OH. Dr. Danese, a fellow of the American College of to reduce blood glucose with the least amount Endocrinology, is a staff physician at the Louis Stokes Cleveland VA Medical Center and an assis- tant professor of medicine at Case Western Re- of weight gain. serve University, Cleveland, OH.

Continued on next page

NOVEMBER 2003 • FEDERAL PRACTITIONER • 33 BEDTIME INSULIN

Continued from previous page

At the end of three months, the metformin, and bedtime plus morn- tion plus daytime , two in- mean HbA1C level decreased simi- ing insulin, respectively (P < .001 jections of insulin daily, or multiple larly and significantly in the four in- compared with all other groups). insulin injections daily. The mean sulin groups (1.7%, 1.9%, 1.8%, and Furthermore, the group receiving HbA1C differed significantly be- 1.6%, respectively) compared with bedtime insulin plus metformin tween the standard and intensive the control group (0.5%, P < .001). had the greatest reduction in groups (P < .001): After six months, Weight gain was significantly less in the group receiving NPH insulin at 9:00 PM (1.2 kg) than in the other The data suggested that a regimen based on bed- insulin treatment groups (2.2 kg in the 7:00 AM NPH group, 1.8 kg in time insulin targeted to fasting glucose levels the 70:30 NPH/regular insulin be- fore breakfast and dinner group, could be maintained for more than two years. and 2.9 kg in the NPH at 9:00 PM and regular insulin before meals group; P < .05). The authors con- HbA1C—from 9.7% to 7.2% at one it was at or below 7.3% in the inten- cluded that, for patients with type 2 year (P < .001 compared with base- sive therapy group, compared with DM, the addition of bedtime NPH line and P < .05 compared with the 9% to 9.6% in the standard treat- insulin is the optimal treatment to other groups)—and significantly ment group, and it remained 2% reduce blood glucose with the least fewer symptomatic and biochemi- lower than the standard group for amount of weight gain. cal episodes of than the duration of the trial. The au- Yki-Jarvinen and colleagues fol- the other groups (P < .05). The au- thors noted that most of the de- lowed up this study in 1999 with a thors concluded that bedtime in- crease in HbA1C in the intensive one-year, randomized clinical trial sulin with metformin was superior therapy group was obtained with designed to help determine which to other regimens in helping pa- a single evening injection of inter- agents, when used in conjunction tients achieve glycemic control mediate insulin (P < .05), alone or with bedtime insulin, resulted in the with less weight gain and fewer with daytime glipizide. The data least weight gain by patients whose hypoglycemic events. suggested that a regimen based on type 2 DM was controlled insuffi- Meanwhile, in 1995, VA re- bedtime insulin targeted to fasting ciently with monother- searchers conducted the VA Coop- glucose levels could be maintained apy.7 The 96 patients included in the erative Study on Glycemic Control for more than two years. study had a mean HbA1C value of and Complications in Type 2 Dia- 9.9% and a fasting plasma glucose betes Mellitus (VA CSDM). The pri- CONSIDERING GLITAZONES level of 11.9 mmol/L. All patients re- mary objective was to evaluate It’s important to note that the previ- ceived intermediate-acting bedtime whether intensive glucose lowering ous studies were initiated prior to NPH insulin. The cohort was di- therapy could be sustained over a the FDA approval of the thiazo- vided further into four groups that two-year period, and the study lidinediones and, therefore, didn’t would receive either glyburide and yielded convincing scientific evi- include this class of medications. placebo, metformin and placebo, dence for using long-acting NPH There remains a paucity of scien- glyburide and metformin, or a sec- insulin at bedtime to maintain a tific medical research comparing ond morning injection of intermedi- lower HbA1C level in patients with the effectiveness of insulin versus a ate-acting insulin. type 2 DM for whom oral therapy third oral insulin sensitizing agent. At one year, body weight re- is no longer effective.8 It seems that for some patients, mained essentially unchanged in In this prospective trial, 153 par- however, an insulin sensitizer can patients receiving bedtime insulin ticipants were assigned randomly add benefit to the overall glycemic plus metformin (mean change, 0.9 to receive either standard insulin picture. kg) but increased by 3.9 kg, 3.6 kg, treatment (one morning injection Fonseca and colleagues con- and 4.6 kg in patients receiving per day) or one of four types of in- ducted a randomized, double-blind, bedtime insulin plus glyburide, bed- tensive therapy: an evening insulin placebo-controlled trial of 348 time insulin plus glyburide and injection, an evening insulin injec- patients to determine whether met-

Continued on page 37

34 • FEDERAL PRACTITIONER • NOVEMBER 2003 BEDTIME INSULIN

Continued from page 34

Furthermore, there was the addi- Table 1. Advantages of bedtime NPH* insulin tional benefit of lower postdinner administration glucose concentrations in those pa- • Potentially less weight gain than 24-hour NPH insulin coverage or tients treated with addition of (P < .02). As practitioners gain more experience with insulin • Morning euglycemia glargine, it may well become a first- • Ease of transition to insulin therapy line insulin treatment for certain • No difficulty with renal, liver, or heart dysfunction or any other illness patients—especially those predis- • Less expensive posed to hypoglycemia. • Ability to lower HbA † more than 1% to 2% Despite the advantages of insulin 1C glargine, it’s costly. And since it’s a *NPH = neutral protamine Hagedorn. †HbA = glycosylated hemoglobin. 1C 24-hour insulin, it can be difficult to titrate the dose (compared to NPH formin or metformin plus rosig- and a or bedtime insulin, for which the morning fast- litazone was more efficacious in insulin, but the optimal DM manage- ing levels determine adequate dos- reducing blood glucose indexes.9 ment strategy remains unknown. ing). Also, patients who remain None of the patients in this study There are, however, clear advan- relatively active during the day may were receiving insulin and oral hy- tages to initiating bedtime NPH in- not need a 24-hour supplemental in- poglycemic medications concur- sulin (Table 1). And while NPH sulin, since glucose levels dip with rently. Over 26 weeks, the group remains the best initial bedtime in- activity. For patients who need pri- that received the plus metformin showed a statistically significant (P < .001) decrease in Bedtime NPH insulin is inappropriate for mean HbA1C levels but an increase in body mass (from 0.7 kg to 1.9 kg). patients with relatively normal fasting blood While the authors concluded that the combination of metformin and glucose levels (less than 126 mg/dL) and rosiglitazone treatment is effective and safe in reducing blood glucose elevated daytime glucose levels. in type 2 DM, weight gain isn’t un- common with the thiazolidine- diones, especially in patients who sulin therapy, such recent pharma- marily morning (not afternoon) are already overweight or obese. cologic breakthroughs as the blood glucose control, traditional This class of medications also can 24-hour glargine also NPH insulin is appropriate. precipitate edema and congestive are appropriate for bedtime use. heart failure in patients with fluid NPH insulin was compared to USING BEDTIME INSULIN overload. The thiazolidinediones insulin glargine in a 2000 trial. Yki- PROPERLY also are contraindicated for pa- Jarvinen and colleagues randomly Practitioners must be aware that tients with liver disease, and when assigned 426 patients with type 2 bedtime insulin isn’t suitable for all they are used, require frequent liver DM who had poor glycemic control patients who experience elevated function testing. while taking oral agents to receive HbA1C levels while taking oral either NPH insulin or insulin agents. Although fasting blood glu- THE CASE FOR BEDTIME INSULIN glargine at bedtime for one year.10 cose levels tend to worsen with the

In most cases, sulfonylurea and met- Average improvement in HbA1C lev- duration of DM, bedtime NPH in- formin fail to control glucose levels els was about the same for both sulin is inappropriate for patients adequately over the long term. After groups, but there was less noctur- with relatively normal fasting blood such treatment failure, clinicians nal hypoglycemia in the glargine glucose levels (less than 126 mg/dL) often prescribe either a triple oral group: 9.9% versus 24% in those pa- and elevated daytime glucose lev- regimen of sulfonylurea, metformin, tients treated with NPH (P < .001). els. In such cases, rather than tak-

NOVEMBER 2003 • FEDERAL PRACTITIONER • 37 BEDTIME INSULIN

mg/dL—depending on the patient’s age and ability to handle the in- Table2.Suggested initial bedtime NPH* insulin dosage sulin. There’s no limit to the Fasting glucose level NPH insulin amount of NPH insulin that can be BMI† (kg/m2) (mg/dL) (units) taken at bedtime. In the VA CSDM, for example, the mean final dose of > 30 > 300 15–20 bedtime NPH insulin was 61.3 ± 38.1 units.8 < 300 12–15 < 30 > 300 10–12 THE IMPORTANCE OF EDUCATION < 300 5–8 Patient education regarding insulin Lean patient 5 administration is imperative. At the time of insulin initiation, it’s impor- *NPH = neutral protamine Hagedorn. †BMI = body mass index. tant for a trained DM educator to discuss hypoglycemia precautions ing bedtime insulin, patients may Stokes Cleveland VA Medical Cen- with the patient and provide in- consider walking after their main ter (LSCVAMC) in Cleveland, OH, struction on administering the in- meal, changing their diet, or taking we provide practitioners with gen- sulin, so that the patient is more to lower glucose levels. eral dosing guidelines based on likely to adhere successfully to the Patients who experience the So- body mass index (BMI) and fasting treatment program. It also can be mogyi effect, which is more com- blood glucose values. Following helpful to explain to the patient that mon in type 1 than type 2 DM, these guidelines both expedites the insulin is a normal substance within comprise another group for whom lowering of HbA1C values and mini- the body, has minimal adverse bedtime insulin is inappropriate. mizes the patient’s risk for hypo- effects, and frequently is just

The high morning blood glucose glycemia (Table 2). what’s needed to bring a high HbA1C levels that characterize this phe- Generally, overweight or obese level into normal range.11 At the nomenon are a response to stress patients require more insulin to LSCVAMC, we provide our patients hormones activated by a low night- counteract the insulin resistance in- with written instructions reiterating time blood glucose level, brought herent in bodies with more fat that insulin is to be administered at on by either too much sulfonylurea, cells. For this reason, if the BMI is bedtime—not dinnertime. Be sure additional exercise, or too little greater than 30 kg/m2 and fasting to ask patients when they usually food. Practitioners need to rule out blood glucose levels are less than go to bed because many older peo- the occurrence of the Somogyi ef- 300 mg/dL, our team recommends ple have very early bedtimes, and fect by carefully checking the pa- starting with 12 to 15 units of NPH this should be considered in deter- tient’s history for night sweats and insulin at bedtime. If the patient’s mining the starting dose. nightmares and recording several BMI is greater than 30 kg/m2 and To lessen patients’ risk of falling 3:00 AM blood glucose levels. fasting blood glucose levels are during the night, caution them to even higher than 300 mg/dL, larger keep a glucose source (containing DETERMINING THE DOSE amounts (up to 20 units) of bed- 15 g of carbohydrates) and meter at Of course, all decisions regarding time insulin can be given. When the bedside in case they develop hypo- DM medical therapy must take into BMI is lower than 30 kg/m2 or if the glycemic shaking, nightmares, or consideration the patient’s lifestyle, patient is especially lean, the more night sweats. We instruct patients age, mental status, and level of ad- traditional dose of 5 to 10 units of that the effects of NPH insulin gen- herence. When the patient and NPH insulin at bedtime may be ap- erally peak eight hours after admin- practitioner agree that oral therapy propriate. istration, and that they should avoid is no longer adequate and bedtime After insulin is started and the skipping breakfast, wear a bracelet insulin should be initiated, how- patient responds safely to the initial that identifies them as having DM, ever, certain physiologic parame- dose, the dose can be titrated up 2 and check blood glucose levels be- ters may be used to determine units every two to four days until fore driving—particularly in the initial insulin doses. At the Louis fasting levels are around 120 to 150 morning.

Continued on page 49

38 • FEDERAL PRACTITIONER • NOVEMBER 2003 BEDTIME INSULIN

Continued from page 38

and then titrated by the patient to 26(suppl 1):S37. A PROGRESSIVE DISEASE, 6. Yki-Jarvinen H, Kauppila M, Kujansuu E, et al. A CONTINUUM OF CARE achieve timely control. With more Comparison of insulin regimens in patients with DM is a disease that’s progressive in practitioners taking the initiative to non-insulin-dependent diabetes mellitus. N Engl J Med. 1992;327:1426–1433. nature. When double or triple oral move patients to this next level of 7. Yki-Jarvinen H, Ryysy L, Nikkila K, Tulokas T, Vanamo R, Heikkila M. Comparison of bedtime therapy is no longer effective or is DM management, we can hope to insulin regimens in patients with contraindicated, insulin shouldn’t see better control of this devastat- mellitus. A randomized, controlled trial. Ann In- tern Med. 1999;130:389–396. be looked on as the last resort but ing disease and its costly complica- 8. Abraira C, Colwell JA, Nuttall FQ, et al. Veterans rather as an effective, relatively in- tions in the near future. ● Affairs Cooperative Study on glycemic control and complications in type II diabetes (VA expensive, and—provided that the CSDM). Results of the feasibility trial. Veterans patient is well educated on proper REFERENCES Affairs Cooperative Study in Type II Diabetes. 1. World Health Organization. Total of people with Diabetes Care. 1995;18:1113–1123. administration—safe choice for diabetes. Available at: www.who.int/ncd/dia 9. Fonseca V, Rosenstock J, Patwardhan R, Salz- glycemic management. Initiating /databases4.htm. Accessed October 7, 2003. man A. Effect of metformin and rosiglitazone 2. National Institute of Diabetes and Digestive and combination therapy in patients with type 2 dia- bedtime insulin with oral agents Kidney Diseases. National diabetes statistics. betes mellitus: A randomized controlled trial. can have a tremendous impact on Available at: diabetes.niddk.nih.gov/dm/pubs JAMA. 2000;283:1695–1702. /statistics/index.htm. Accessed October 7, 2003. 10. Yki-Jarvinen H, Dresselr A, Ziemen M, for the high fasting glucose levels, making 3. American Diabetes Association. Direct and indi- HOE 901/3002 Study Group. Less nocturnal hy- it easier to achieve target daytime rect costs of diabetes in the United States. 2002. poglycemia and better post-dinner glucose con- Available at: www.diabetes.org/main/info/facts trol with bedtime insulin glargine compared with blood glucose levels and HbA1C val- /facts_costs.jsp. Accessed October 7, 2003. bedtime NPH insulin during insulin combination ues with less weight gain. Insulin 4. UK Prospective Diabetes Study Group. UK therapy in type 2 diabetes. HOE 901/3002 Study prospective diabetes study 16. Overview of 6 Group. Diabetes Care. 2000;23:1130–1136. doses can be determined initially years’ therapy of type II diabetes: A progressive 11. Berger M, Jorgens V, Muhlhauser I. Rationale for by considering the patient’s BMI disease. Diabetes. 1995;44:1249–1258. the use of insulin therapy alone as the pharmaco- 5. American Diabetes Association. Clinical practice logical treatment of type 2 diabetes. Diabetes and fasting blood glucose levels recommendations 2003. Diabetes Care. 2003; Care. 1999;22(suppl 3):C71–C75.

FEDERAL HEALTH MATTERS

Continued from page 40 Veterans Benefits Act of 2003 (H.R. children with spina bifida born to of the Atlantic to find a treatment 2297), was approved by the full Vietnam-era veterans who served for another of the world’s most elu- House of Representatives on Octo- near Korea’s demilitarized zone be- sive and deadly diseases: smallpox. ber 9. According to Smith, this leg- tween October 1, 1967 and May 7, As with the previous projects, the islation “would provide significant 1975. Other provisions of the legis- Smallpox Research Grid enlisted new support to veterans, particu- lation relate to vocational and self- the help of personal computer larly to disabled veterans and sur- employment training for veterans, users all over the world, who do- viving spouses of veterans.” home loans for reservists, and state nated their computers’ spare pro- Specifically, severely disabled cemetery grants. cessing power to help screen 35 veterans could receive larger million potential drug molecules grants to obtain specially adapted against eight models of the small- automobiles and housing, and sur- pox protein to identify those that viving spouses who remarry after International Tech potentially could bind to the pro- age 55 would retain their Team Tackles tein, rendering it inactive. With the widow/widower benefits. In addi- help of this virtual supercomputer, tion, the bill would grant presumed Smallpox the field of drug candidates has service connection for cirrhosis of Building on the success of the pre- been narrowed dramatically in a the liver in former prisoners of war vious cancer and anthrax research matter of weeks, instead of years. (POWs), lift the requirement that a projects, investigators from Oxford On September 30, the team pre- POW’s imprisonment period ex- University and United Devices sented the list of candidates to rep- ceed 30 days for service connec- (Austin, TX) have joined forces resentatives from the DoD and tion to be granted for several health with a host of other academic and United Kingdom for use in further conditions, and expand benefits for corporate partners from both sides research and drug development. ●

NOVEMBER 2003 • FEDERAL PRACTITIONER • 49