What is the starting dose of NPH (Neutral Protamine Hagedorn) insulin for a patient taking prednisone 25 mg?

NPH Insulin Starting Dose for Steroid-Induced Hyperglycemia

For a patient receiving prednisone 25 mg daily while on tube feedings, start NPH insulin at 0.1-0.2 units/kg/day administered in the morning, which translates to approximately 7-14 units for a 70 kg patient. 1, 2

Rationale for NPH Insulin Selection

• NPH insulin is specifically recommended by the American Diabetes Association for steroid-induced hyperglycemia because its intermediate-acting profile peaks at 4-6 hours after administration, which aligns perfectly with the peak hyperglycemic effect of morning prednisone 1, 2, 3
• Prednisone 25 mg is considered a medium-dose glucocorticoid that causes disproportionate hyperglycemia during daytime hours (midday to midnight), with blood glucose often normalizing overnight 2, 4
• Long-acting basal insulin alone (like glargine) may under-treat daytime hyperglycemia and cause nocturnal hypoglycemia in this setting 4

Specific Dosing Algorithm

Step 1: Calculate Initial NPH Dose

• Start with 0.1-0.2 units/kg/day for insulin-naive patients 1, 2
• For a 70 kg patient: 7-14 units NPH
• Administer the entire dose in the morning (between 0600-0800 hours) to coincide with prednisone administration 1, 2

Step 2: Adjust for Patient-Specific Factors

• For patients with higher insulin resistance (obesity, pre-existing diabetes): use the higher end of the range (0.2 units/kg) 1
• For elderly patients or those at high hypoglycemia risk: use the lower end (0.1 units/kg) 2
• For high-dose glucocorticoids (>40 mg prednisone equivalent): insulin requirements may increase by 40-60% above standard dosing 2, 3

Step 3: Add Prandial Coverage

• Start rapid-acting insulin at 0.1 units/kg or 4 units before each meal if the patient is eating 5
• Use a carbohydrate ratio of approximately 1:10 (1 unit per 10g carbohydrate) initially 2
• Correction scale: 1 unit for every 40-50 mg/dL above target (150 mg/dL), with more aggressive correction in afternoon/evening when steroid effect peaks 2

Step 4: Special Considerations for Tube Feedings

• For continuous tube feedings: the total daily nutritional insulin component should be calculated as 1 unit for every 10-15 g carbohydrate per day 5
• Maintain basal insulin coverage even if tube feedings are interrupted to prevent hyperglycemia 5
• Consider splitting NPH to twice daily (every 12 hours) if receiving continuous enteral nutrition 3

Monitoring and Adjustment Protocol

Initial Monitoring

• Check blood glucose every 2-4 hours while hospitalized to guide insulin adjustments 1, 2, 3
• Pay special attention to afternoon and evening values (1200-2400 hours) when steroid effect peaks 2, 4
• Target blood glucose range: 80-180 mg/dL (4.4-10.0 mmol/L) 5, 2

Dose Titration

• For persistent hyperglycemia: increase NPH dose by 2 units every 3 days until target achieved 1, 2
• If hypoglycemia occurs: reduce NPH dose by 10-20% 1, 2
• Adjust daily based on glucose patterns and anticipated changes in glucocorticoid dosing 3

Common Pitfalls to Avoid

• Do not rely solely on fasting glucose to guide NPH dosing—this leads to under-treatment of daytime hyperglycemia and potential nocturnal hypoglycemia 2, 4
• Avoid using only sliding scale insulin—this approach is associated with poor glycemic control and has been discouraged in clinical guidelines 3
• Do not use long-acting insulin alone without adding NPH, as this provides inadequate coverage of daytime steroid-induced hyperglycemia 2
• Remember that insulin requirements decrease rapidly after steroid discontinuation, requiring prompt dose adjustments to avoid hypoglycemia 2, 3

Evidence Quality Note

The recommendation for 0.1-0.2 units/kg/day is supported by multiple high-quality sources including American Diabetes Association guidelines 1, 2 and validated by randomized controlled trials showing NPH-based protocols achieve better glycemic control than glargine-based regimens in steroid-induced hyperglycemia 4, 6, 7. A 2018 randomized trial demonstrated that an NPH insulin-based protocol significantly reduced mean blood glucose (226 vs 269 mg/dL, p<0.0001) compared to usual care 7.