Skip to content
type1.science

Inhaled insulin (Afrezza)

MannKind

A dry-powder human insulin inhaled at the start of a meal. Its standout trait is speed — it reaches the blood and starts lowering glucose within minutes and clears within ~1.5–3 hours, faster than any injected insulin — at the cost of coarse cartridge dosing, a mandatory lung-function screen, and US-only access.

Available nowRegulator-approvedinhaledultra-rapidmealtime

The scorecard

Onset speed92

Mealtime convention (faster onset = better): insulin peaks in blood ~10–20 min and glucose-lowering begins ~12 min — the fastest-onset insulin available to consumers.[1]

Time to peak90

Mealtime convention (faster peak = better): peak glucose-lowering effect at ~35–55 min (dose-dependent), far earlier than the ~90–130 min of subcutaneous rapid analogs.[1]

Short tail88

Mealtime convention (shorter tail = better): effect returns to baseline by ~90–270 min (dose-dependent); the short, clean tail sharply cuts insulin stacking and late post-meal lows.[1]

Consistency50

Within-patient PK variability is low (AUC ~16%, Cmax ~21%), but dosing is coarse (4/8/12-unit cartridges only) and lung deposition varies with inhalation technique and lung status.[1]

Exercise flexibility72

The genuinely short tail means less active insulin lingering into exercise, reducing delayed-hypoglycemia risk versus multi-hour subcutaneous analogs.[2]

Access & cost32

Access convention (cheaper/more available = better): US-only, brand-only (no biosimilar), requires baseline + periodic spirometry, contraindicated in asthma/COPD and not advised in smokers, and cannot integrate with pumps; high out-of-pocket cost without coverage, though savings caps substantially lower it for many.[7]

Insulins are scored relative to their role peers (see tags: rapid, ultra-rapid, basal, inhaled). A basal insulin's onset score compares it to other basals, not to mealtime insulins.

Editor’s take

The closest thing to truly fast insulin we have, and a working preview of what closing the insulin-speed gap would feel like at the dinner table. Coarse dosing, the lung-function gate, US-only access and pump-incompatibility keep it niche — but on raw speed it embarrasses every injectable, and head-to-head meal challenges confirm flatter post-meal glucose than rapid analogs.

The full picture

Afrezza is Technosphere inhaled insulin (TI) — recombinant human insulin adsorbed onto microscopic carrier particles of fumaryl diketopiperazine (FDKP) and inhaled as a dry powder through a small breath-powered inhaler.1 It is a mealtime (bolus) insulin: you take it at the start of each meal to cover the carbohydrate, alongside a separate long-acting basal insulin.1 By delivering insulin across the enormous surface of the lungs, it bypasses the slow subcutaneous route entirely — and that gives it a profile no injection can match.

How fast, exactly? After inhalation the FDKP particles (median ~2.0–2.5 µm) dissolve instantly at the lung's neutral pH and insulin is absorbed within minutes.2 Serum insulin peaks at roughly 10–20 minutes (versus ~50 minutes for injected insulin lispro).12 The glucose-lowering effect begins at about 12 minutes, reaches its maximum at roughly 35–55 minutes depending on dose, and returns to baseline by about 90–270 minutes — a much shorter, cleaner tail than the 5–7 hours of subcutaneous analogs.1 Half the maximal effect is reached in ~19 minutes versus ~50 minutes for lispro.2 In short: fastest on, fastest off of any insulin a person with diabetes can buy.2

Absorption variability and dosing. Within a single person, the pharmacokinetics are actually quite reproducible (intrapatient variability ~16% for total exposure, ~21% for peak).1 The real limitation is granularity: cartridges come only in 4-, 8-, and 12-unit doses (blue, green, yellow), so fine titration isn't possible, and the fraction reaching the deep lung depends on inhalation technique and lung health.1 Roughly 24% of an inhaled dose is bioavailable relative to subcutaneous regular insulin, which is already accounted for in the labeled unit conversions.2

Exercise. The short tail is a genuine advantage around activity — far less "insulin on board" lingers into and after exercise, lowering the risk of delayed hypoglycemia that dogs multi-hour subcutaneous analogs.2

Does it work in practice? In the 17-week INHALE-3 randomized trial (123 adults with type 1 diabetes, ~half on automated insulin delivery beforehand), TI plus degludec was noninferior on HbA1c to usual care — but with a striking split: HbA1c improved by >0.5% in 21% of TI users (vs 5% on usual care), yet worsened by >0.5% in 26% (vs 3%).3 It rewards engaged dosers and punishes set-and-forget use. A companion meal-challenge study found TI produced a smaller, earlier, lower post-meal glucose peak than subcutaneous rapid analogs (adjusted reduction of −12 mg/dL in glucose-above-180 area, P=0.02).4

Approvals, safety, access. The FDA first approved Afrezza in 2014, and in May 2026 expanded it to children aged 6 and older — making it the first and only inhaled mealtime insulin for kids.15 It carries a boxed warning for acute bronchospasm and is contraindicated in chronic lung disease (asthma, COPD); every patient needs spirometry (FEV1) at baseline, at 6 months, and annually, and it is not recommended in smokers.1 A small, mostly non-progressive ~40 mL excess decline in FEV1 over two years and a transient cough (~27% of users) are the signature respiratory effects.1 It is US-only, brand-only with no biosimilar, and cannot be loaded into a pump or closed loop; it carries a high out-of-pocket cost without coverage, though Medicare and commercial savings caps and a manufacturer program substantially lower it for many.6

What's coming. The most active frontier is pairing Afrezza's speed with automated insulin delivery: AID systems still can't kill the post-meal spike because the subcutaneous insulin they pump acts too slowly, and an ultra-rapid inhaled bolus at mealtime could plug exactly that hole — a combination now under formal review, with the main hurdle being how to teach the algorithm to account for insulin it didn't deliver.7 Afrezza remains the clearest real-world argument that the meal response can be fast, and a standing case for closing the insulin-speed gap.

References

  1. MannKind Corporation. AFREZZA (insulin human) inhalation powder — US Prescribing Information (revised 05/2026): Indications, Dosage and Administration, Contraindications, Boxed Warning, Warnings and Precautions, Clinical Pharmacology (12.2, 12.3). https://afrezza.com/pdf/full-prescribing-information ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9

  2. Heinemann L, Baughman R, Boss A, Hompesch M. Pharmacokinetic and Pharmacodynamic Properties of a Novel Inhaled Insulin. J Diabetes Sci Technol (2017). https://doi.org/10.1177/1932296816658055 ↩ ↩2 ↩3 ↩4 ↩5 ↩6

  3. Hirsch IB, Beck RW, Marak MC, et al. A Randomized Trial Comparing Inhaled Insulin Plus Basal Insulin Versus Usual Care in Adults With Type 1 Diabetes (INHALE-3). Diabetes Care (2025). https://doi.org/10.2337/dc24-1832 ↩

  4. Hirsch IB, Beck RW, Marak MC, et al. A Randomized Comparison of Postprandial Glucose Excursion Using Inhaled Insulin Versus Rapid-Acting Analog Insulin in Adults With Type 1 Diabetes Using Multiple Daily Injections or Automated Insulin Delivery. Diabetes Care (2024). https://doi.org/10.2337/dc24-0838 ↩

  5. MannKind Corporation. FDA Approval of Afrezza, the First and Only Inhaled Mealtime Insulin for Children and Adolescents Aged 6 and Older (29 May 2026). https://www.globenewswire.com/news-release/2026/05/29/3303734/29517/en/mannkind-announces-fda-approval-of-afrezza-the-first-and-only-inhaled-mealtime-insulin-for-use-in-children-and-adolescents-aged-6-and-older-living-with-diabetes.html ↩

  6. GoodRx. Afrezza Prices, Coupons & Savings Tips. https://www.goodrx.com/afrezza ↩

  7. Cengiz E, Beck RW. Use of Inhaled Insulin with Automated Insulin Delivery Systems. Diabetes Technol Ther (2026). https://doi.org/10.1177/15209156251407710 ↩

What's next for this

  • →Pairing Afrezza's ultra-rapid speed with automated insulin delivery (AID) to blunt post-meal spikes; combination now under formal review, main hurdle is teaching the algorithm to account for insulin it didn't deliver

Sources

  1. [1]AFREZZA (insulin human) inhalation powder — US Prescribing Information (rev. 05/2026) · regulatory — Initial US approval 2014; indication ages 6+ (05/2026). PK: tmax 10–20 min, terminal t½ 120–206 min, baseline by 60–240 min, intrapatient AUC var ~16% / Cmax ~21%. PD (Table 7): first effect ~12 min, peak ~35–55 min, baseline ~90–270 min, AUC-GIR var ~28%. Cartridges 4/8/12 U. Boxed warning: acute bronchospasm; contraindicated in chronic lung disease; FEV1 spirometry at baseline/6 mo/annually; not for DKA; not advised in smokers.
  2. [2]Heinemann L, Baughman R, Boss A, Hompesch M. Pharmacokinetic and Pharmacodynamic Properties of a Novel Inhaled Insulin. J Diabetes Sci Technol (2017) · peer-reviewed — PMC5375067. Gen2 tmax ~8–15 min vs ~50 min lispro; T50% ~19 min vs ~50 min; relative bioavailability ~24% vs SC RHI; FDKP particle ~2.0–2.5 µm; faster on/off than any SC insulin.
  3. [3]Hirsch IB, Beck RW, Marak MC, et al. A Randomized Trial Comparing Inhaled Insulin Plus Basal Insulin Versus Usual Care in Adults With Type 1 Diabetes (INHALE-3). Diabetes Care (2025) · peer-reviewed — PMC11870290. N=123, 17 wks, TI+degludec vs usual care (48% AID, 45% MDI). HbA1c noninferior; >0.5% improved 21% vs 5%, but >0.5% worsened 26% vs 3% — responder/non-responder split. Brief cough most common.
  4. [4]Hirsch IB, Beck RW, Marak MC, et al. A Randomized Comparison of Postprandial Glucose Excursion Using Inhaled Insulin Versus Rapid-Acting Analog Insulin in Adults With Type 1 Diabetes (MDI or AID). Diabetes Care (2024) · peer-reviewed — PMC11362108. N=122 meal challenge. Smaller post-meal excursion with TI vs RAA (AUC >180 mg/dL adjusted diff −12 mg/dL, 95% CI −22 to −2, P=0.02); lower, earlier peak.
  5. [5]MannKind Announces FDA Approval of Afrezza, the First and Only Inhaled Mealtime Insulin for Use in Children and Adolescents Aged 6 and Older (29 May 2026) · regulatory — US-only pediatric expansion to ages 6+, supported by the INHALE-1 trial.
  6. [6]Cengiz E, Beck RW. Use of Inhaled Insulin with Automated Insulin Delivery Systems. Diabetes Technol Ther (2026) · peer-reviewed — Review of combining ultra-rapid inhaled insulin with AID to attack postprandial spikes the subcutaneous loop reacts to too slowly; harmonization challenges remain.
  7. [7]Afrezza Prices, Coupons & Savings Tips (GoodRx) · news — High cash/retail cost; Medicare Part D and commercial savings cards plus a manufacturer Patient Direct program substantially reduce out-of-pocket cost for many. No biosimilar.