Basal Bolus Insulin Regimens: Carb Ratios, Correction Factors, and Pump Settings Explained

What Is a Basal Bolus Insulin Regimen?

A basal bolus regimen mimics the pancreas by delivering two distinct types of insulin for two distinct physiological jobs. The basal component suppresses hepatic glucose output overnight and between meals. The bolus component covers glucose that rises after eating, with doses adjusted in real time based on carbohydrate content and the current blood glucose reading.

Before basal bolus therapy became standard, premixed insulin was common. A 2014 meta-analysis in Diabetes Care (N=2,100 pooled patients) found that basal bolus regimens achieved a mean A1C reduction of 0.24 percentage points greater than biphasic premixed insulin, with an odds ratio for reaching A1C <7.0% of 1.4 in favor of basal bolus therapy. [1] The American Diabetes Association (ADA) 2024 Standards of Medical Care in Diabetes state: "Multiple daily injection (MDI) therapy with basal and bolus insulin is the preferred insulin regimen for most people with type 1 diabetes and for type 2 diabetes patients who require intensive glycemic management." [2]

Two drug classes supply the components. Basal insulin is typically glargine U-100 or U-300 (Lantus, Toujeo), detemir (Levemir), or degludec (Tresiba). Bolus insulin is one of the rapid-acting analogs: aspart (NovoLog), lispro (Humalog), or glulisine (Apidra), all with an onset under 15 minutes. [3]

How to Set the Starting Basal Insulin Dose

Start with body weight, not A1C. The standard formula is 0.2 units per kg per day for insulin-naive type 2 patients, or 0.3 to 0.4 units per kg per day for type 1 patients who have lost significant C-peptide. [4] A 90 kg type 2 patient, for example, begins at 18 units of glargine at bedtime.

Titration matters more than the starting dose. The INSIGHT trial titrated glargine once weekly using a simple algorithm (increase by 2 units if fasting glucose exceeded 130 mg/dL on three consecutive days) and achieved a mean fasting plasma glucose of 99 mg/dL at 24 weeks without increasing nocturnal hypoglycemia beyond placebo rates. [5] Clinicians at HealthRX use a similar 2-unit-per-3-days rule for outpatient titration.

Basal dose should represent 40 to 50% of TDD once optimized. If basal is creeping above 50% and postprandial glucose remains elevated, that is a signal to add or increase mealtime bolus doses rather than push basal further.

One practical caution: degludec (Tresiba) has a half-life of 25 hours and takes 3 to 4 days to reach steady state. [6] Do not re-titrate more frequently than every 3 days to avoid stacking.

Insulin-to-Carb Ratio: The 500 Rule and How to Refine It

The insulin-to-carb ratio (ICR) tells you how many grams of carbohydrate one unit of rapid-acting insulin covers. An ICR of 1:15 means one unit covers 15 grams of carbohydrate.

The 500 rule provides the starting ICR: divide 500 by the patient's TDD. A patient using 50 units per day (25 basal plus 25 bolus) gets an ICR of 500/50 = 1:10 (one unit per 10 g carbohydrate). [7]

Real-world refinement is essential. Carbohydrate sensitivity varies by time of day. Insulin resistance tends to be highest in the morning due to cortisol and growth hormone peaks. A patient with ICR 1:10 at dinner may need 1:7 or even 1:6 at breakfast. [8] The DCCT trial, which enrolled 1,441 patients with type 1 diabetes over 6.5 years, showed that intensive MDI therapy including carbohydrate counting reduced the risk of sustained retinopathy progression by 76% and microalbuminuria by 39% compared with conventional two-injection therapy. [9] That evidence base is why carbohydrate counting is taught as a core skill, not an optional upgrade.

To refine an ICR after starting therapy, log: pre-meal glucose, grams of carbohydrate consumed, units of bolus given, and glucose 2 hours post-meal. If 2-hour glucose is more than 40 mg/dL above pre-meal glucose, the ICR is too generous (fewer carbs per unit). If 2-hour glucose drops more than 40 mg/dL, the ICR is too aggressive. [7]

Insulin Correction Factor: The 1,700 Rule

The correction factor (CF), sometimes called insulin sensitivity factor (ISF), tells you how many mg/dL one unit of rapid-acting insulin will lower blood glucose. It is calculated as:

CF = 1,700 divided by TDD (for mg/dL units). [10]

Using the same 50-unit TDD example: CF = 1,700/50 = 34 mg/dL per unit. If the target glucose is 100 mg/dL and the current glucose is 202 mg/dL, the correction bolus is (202 minus 100) divided by 34 = 3 units.

For patients using mmol/L, the equivalent formula uses 94 instead of 1,700.

Correction boluses carry a specific risk that ICR-only dosing does not: stacking. Rapid-acting analogs remain active for 3 to 4 hours. Giving a second correction dose before the first has fully acted can cause severe hypoglycemia. Most insulin pumps calculate "insulin on board" (IOB) automatically and subtract it before recommending a correction. Patients on MDI must do this calculation mentally or use a bolus calculator app. [11]

The FDA has cleared several diabetes decision-support apps that calculate IOB and correction bolus, including the Omnipod 5 controller and the Control-IQ algorithm in the Tandem t:slim X2, both of which have Level A evidence supporting their glycemic outcomes. [12]

Mealtime Bolus Dosing: Timing and Carb Counting Basics

Timing is as important as dose size. Rapid-acting analogs injected subcutaneously take 10 to 20 minutes to begin working. Glucose from a mixed meal starts rising within 15 minutes of the first bite. The practical rule: inject 15 minutes before eating for most meals. [13]

Patients with gastroparesis or unpredictable oral intake should dose after eating or use extended bolus delivery on a pump (discussed below).

Carbohydrate counting accuracy directly determines bolus accuracy. A systematic review in Diabetic Medicine (2019, 12 studies, N=899) found that patients trained in structured carb counting achieved A1C 0.35 percentage points lower than patients using generalized meal plans. [14] The most common error is undercounting restaurant portions. Pasta portions at chain restaurants average 70 to 90 grams of carbohydrate per serving, not the 45 grams many patients assume. [15]

Practical counting steps:

1. Read the nutrition label; identify total carbohydrates (not net carbs).
2. Weigh or measure the portion actually consumed.
3. Calculate: grams eaten divided by ICR = bolus units.
4. Add correction units if pre-meal glucose exceeds target.
5. Subtract IOB if a correction was given within the past 3 hours.

Fiber above 5 grams per serving may be subtracted from total carbohydrates per ADA guidance, because high-fiber foods blunt postprandial glucose rise. [2]

Insulin Pump Settings: Translating MDI Doses to Continuous Subcutaneous Insulin Infusion

An insulin pump (CSII) delivers only rapid-acting insulin continuously at a programmed basal rate, replacing the long-acting injection. Converting an MDI regimen to pump settings requires three foundational numbers: TDD, basal rates, and ICR/CF. [16]

Step 1. Calculate pump TDD. Reduce the MDI TDD by 20% when starting pump therapy to account for improved insulin absorption from continuous low-dose subcutaneous delivery. A patient on 60 units MDI per day starts pump therapy at 48 units per day.

Step 2. Set initial basal rate. Allocate 50% of pump TDD to basal. 48 multiplied by 0.5 equals 24 units per day, delivered as 1.0 unit per hour across all 24 hours initially. [16]

Step 3. Program time-based basal segments. Once continuous glucose monitor (CGM) data accumulates over 2 to 4 weeks, basal rates are adjusted by time block. Dawn phenomenon patients typically need a higher rate between 03:00 and 08:00. Exercise windows need lower rates starting 90 minutes before activity. [17]

Step 4. Enter ICR and CF into the pump bolus calculator. Most pumps accept separate ICR values for morning, afternoon, and evening. This is where the 500 rule estimates are entered and then refined as described above.

The table below summarizes starting parameters for a 70 kg type 1 patient with a TDD of 42 units:

| Parameter | Calculation | Starting Value | |---|---|---| | Basal rate (flat) | 42 × 0.5 / 24 | 0.875 units/hr | | ICR (500 rule) | 500 / 42 | 1 unit per 12 g carb | | Correction factor | 1,700 / 42 | 40 mg/dL per unit | | Target glucose | ADA 2024 guideline | 100 mg/dL |

Advanced pump features include automated insulin delivery (AID) systems that use CGM data to modulate basal rates automatically. The CLOSED-LOOP trial (N=168 to 24 weeks) found that the Control-IQ AID system increased time-in-range (70-180 mg/dL) from 61% to 71% compared with sensor-augmented pump therapy alone (P<0.001). [18]

Recognizing and Managing Hypoglycemia on a Basal Bolus Regimen

Hypoglycemia is the principal risk of intensified insulin therapy. The DCCT found that intensive therapy tripled the rate of severe hypoglycemia (61.2 vs. 18.7 episodes per 100 patient-years) compared with conventional therapy. [9] This underscores the need for structured patient education alongside dose optimization.

Glucose <70 mg/dL meets the ADA Level 1 hypoglycemia threshold. [2] Treatment follows the 15-15 rule: consume 15 grams of fast-acting carbohydrate (4 glucose tablets, 120 mL of juice), recheck in 15 minutes, and repeat if glucose remains <70 mg/dL.

Nocturnal hypoglycemia deserves special attention. Glargine U-300 and degludec produce significantly flatter pharmacokinetic profiles than glargine U-100, with the EDITION 1 trial (N=807) showing a 25% reduction in nocturnal confirmed hypoglycemia for glargine U-300 versus glargine U-100 at the same A1C. [19]

Patients and caregivers should be trained on glucagon rescue. Nasal glucagon (Baqsimi 3 mg intranasal) received FDA approval in 2019 and demonstrated non-inferiority to intramuscular glucagon in a crossover trial: 98.7% vs. 100% recovery within 30 minutes. [20]

Monitoring, A1C Targets, and When to Adjust Doses

No dose adjustment should occur without data. At minimum, patients need pre-meal and 2-hour post-meal glucose readings or CGM data covering 14 or more days. The ADA 2024 guidelines recommend A1C <7.0% for most non-pregnant adults, with individualization to <6.5% for those with short disease duration and low hypoglycemia risk, or <8.0% for those with a history of severe hypoglycemia or limited life expectancy. [2]

CGM time-in-range (TIR, 70-180 mg/dL) is increasingly used alongside A1C. A TIR of 70% correlates approximately with an A1C of 7.0%. [21] The ADA and the Advanced Technologies and Treatments for Diabetes (ATTD) consensus group recommend a TIR goal of at least 70% for adults with type 1 or type 2 diabetes on intensive insulin therapy. [21]

Dose adjustment triggers:

• Fasting glucose consistently above 130 mg/dL on 3 consecutive mornings: increase basal by 2 units.
• Postprandial glucose more than 180 mg/dL two hours after a counted meal: tighten ICR (reduce carbs per unit by 1 to 2 g).
• Recurrent hypoglycemia at a specific time of day: reduce the dose (basal rate segment, or bolus ICR) active 2 to 3 hours before that window.
• A1C above target with TIR below 70%: review carb counting accuracy before increasing insulin doses.

Special Populations: Pregnancy, Renal Impairment, and Older Adults

Pregnancy requires tighter targets (fasting <95 mg/dL, 1-hour post-meal <140 mg/dL) and more frequent dose revisions as insulin resistance increases with gestational age. [22] The ACOG Practice Bulletin No. 201 recommends MDI or CSII for all pregnant patients with type 1 diabetes. [22]

Renal impairment reduces insulin clearance. GFR below 45 mL/min/1.73 m² may require a 25 to 50% TDD reduction to avoid hypoglycemia, with more frequent monitoring. [23]

Older adults face particular hypoglycemia hazard because of impaired counter-regulatory responses and fall risk. The ADA recommends an A1C target of <8.0% for older adults with multiple comorbidities, with basal-only or basal-plus regimens preferred over full MDI to reduce dosing complexity. [2]

Working with a Clinician to Personalize Your Regimen

Starting numbers derived from the 500 rule and 1,700 rule are starting points. Glucose physiology varies enough between individuals that most patients need 4 to 8 weeks of systematic logging before their ICR and CF converge on accurate values. A certified diabetes care and education specialist (CDCES) can cut that calibration time substantially. The American Association of Diabetes Care and Education Specialists recommends at least three structured education visits in the first year after starting MDI therapy. [24]

Telehealth-based diabetes management has demonstrated non-inferior glycemic outcomes versus in-person care. A 2022 randomized trial published in JAMA Internal Medicine (N=317) found that remote MDI management achieved a mean A1C reduction of 0.8% at 12 months, identical to the in-person cohort, with higher patient satisfaction scores in the telehealth arm. [25]

Patients who upload CGM or glucometer data before every visit allow clinicians to identify pattern problems, not just current A1C. At HealthRX, dose adjustments are made based on a minimum of 14 days of CGM data showing at least 70% sensor wear time, a threshold consistent with the ATTD 2023 consensus statement. [21]