🎚️ Gradient Factors

Silent Bubbles

Doppler ultrasound studies have revealed something surprising: bubbles form in divers after most dives, even when following conservative profiles with no DCS symptoms.

This means the M-value line on our pressure-pressure diagram isn't a sharp "safe/unsafe" boundary. It's more like a probability gradient — the closer you get, the more bubbles form, and the higher your risk.

DCS risk varies enormously between individuals and even between dives for the same person. Factors that increase risk include:

• Dehydration
• Poor physical fitness
• Age
• Body fat percentage
• Cold exposure
• Heavy workload
• Patent Foramen Ovale (PFO)
• Alcohol consumption

The term "undeserved hit" describes cases where divers followed all the rules but still got DCS. This happens because we're dealing with probabilistic biology, not precise physics.

The Fish Collector's Observation

In the eighties, Richard Pyle noticed that his post-dive fatigue varied markedly from one dive to another. Because he kept meticulous records, he systematically searched for what distinguished the dives after which he felt tired from those after which he did not.

The repeated association suggested that brief stops deeper than his prescribed decompression might matter. It was a compelling personal observation, not a controlled experiment.

Pyle began adding very brief intermediate stops deliberately and later described a simple empirical procedure. He treated the repeated improvement in how he felt as meaningful, while acknowledging that he was not a decompression physiologist.

The idea met understandable skepticism. Conventional dissolved-gas schedules generally sent divers toward a shallower first stop, and extra time at depth could continue loading slower tissues.

Pyle stops are historically important because they challenged the prevailing ascent pattern. His old rule of thumb belongs to the linked historical account, not to current dive-planning guidance.

📄 Read Richard Pyle's account →

The Varying Permeability Model (VPM) had been published in 1986, before Pyle's 1997 article. When Pyle encountered David Yount's work in 1989, he recognized that this independent bubble model predicted a similarly deep start to decompression. The Reduced Gradient Bubble Model (RGBM) later followed another bubble-based path.

Gradient Factors (GF) offered a different route within Bühlmann's dissolved-gas model: a low GF_low can move the first stop deeper. GF are not a bubble model, however, and were not created as proof of Pyle's procedure.

The convergence was persuasive, and deep-stop profiles spread rapidly through technical diving. Their popularity, plausible theory, and many apparently successful dives advanced faster than controlled evidence of clinical benefit.

What Are Gradient Factors?

Gradient Factors (GF) are a way to add conservatism to Bühlmann's algorithm by reducing the allowed supersaturation. Instead of allowing tissues to reach 100 % of the M-value, you set a lower percentage.

A gradient factor of GF 80 % means you only allow the tissue to reach 80 % of the way from ambient pressure to the M-value.

Dive computers and planning software use two gradient factors:

The algorithm interpolates between these values during the ascent:

• At your deepest required stop → GF_low applies
• At the surface → GF_high applies
• In between → linear interpolation

The NEDU Study That Changed the View of Deep Stops

The project covered 2005–2011; the controlled chamber dives were conducted from November 2005 to August 2006, and NEDU issued its final report in July 2011.

The U.S. Navy Experimental Diving Unit (NEDU) conducted a large comparative study: same dive (170 fsw/30 min), same total decompression time, but different profiles:

The deep-stop schedule produced significantly more DCS: 11 cases in 198 dives versus 3 in 192.

This result does not prove that every brief deep stop is harmful. It shows that shifting a substantial share of the same decompression time deeper worsened outcomes on this air profile; it cannot automatically be generalized to every trimix profile.

NEDU and later reviews shifted guidance and practice away from aggressively deep-skewed decompression, especially for air and nitrox. This was a retreat from the assumption that deeper is automatically safer, not proof that every deeper first stop is harmful.

📄 Read the primary NEDU TR 11‑06 report →

The problem with deep stops is subtle but important:

Deep stops reduce the pressure gradient for off-gassing. The deeper you stay, the smaller the difference between tissue pressure and ambient — and that gradient is what drives gas elimination.

The Nuanced Reality

This doesn't mean all deep stops are bad:

• NEDU did not directly test Pyle's very brief intermediate stops; their clinical benefit remains unproven
• NEDU tested redistribution of the same total decompression time, not simply a slower ascent
• A very low GF_low can overemphasize deep stops and reduce shallow off-gassing time
• Air/nitrox and helium profiles require separate evaluation; no single GF preset is universally safest

Reference table courtesy of Jirka Hovorka.

CMAS published a 2025 fact sheet summarising current evidence on gradient factors. Its advice differs from some of the values in the reference table above — worth knowing both perspectives before you pick your settings.

1. For air/nitrox, use symmetric GFs

CMAS recommends GF_low = GF_high on nitrogen dives — typically 90/90, dropping to 85/85 or 80/80 when risk factors apply (age, fatigue, cold, exertion). Asymmetric GFs force deep stops, and multiple independent studies found this increases DCS incidence on air/nitrox dives:

Source: Alain Foret, President of the CMAS Technical Committee — recommendations on Gradient Factors (2025)

• French Navy (2005) — deep stops raised DCS risk on air deco dives study ↗
• U.S. Navy NEDU (2011) — "redistribution of stop time from shallow to deep stops increases incidence of DCS in air decompression dives" study ↗
• Belgian Military (2023) — default 30/70 preset was unsafe for 60 m air dives study ↗

2. For helium (trimix/heliox), asymmetric GFs are still justified

On helium mixes, forcing some deeper stops does help (inner-ear DCS, fast helium tissues). CMAS suggests GF_low 30–50, GF_high 70–80 (e.g. 30/70, 50/80). This aligns with the "Deep dive" row of the table above.

3. On a no-deco dive, GF_low doesn't matter

Non-obvious but important: GF_low only kicks in at the first deco stop depth. If your dive has no mandatory stops, setting 45/95 behaves exactly like 95/95 — because there is no "first stop" for GF_low to anchor. Don't expect conservative GF_low to help you on a pure recreational profile.

4. Check your computer's default preset

Many dive computers ship with GF 30/70 out of the box — a preset chosen for tech/trimix diving. CMAS argues the same preset is not right for air/nitrox.

📄 Read the CMAS recommendations →

• Some conservatism is good — GF 100/100 is too aggressive
• GF 70/85 is a sensible starting point for most divers
• Very low GF_low (like 30 %) can be counterproductive
• More conservatism isn't always better — it's about balance
• Adjust based on personal factors, dive conditions, and experience