Risk of decompression sickness in extreme human breath-hold diving
Authors: J. R. Fitz-Clarke
DOI / Source: Undersea Hyperbaric Medicine 2009; 36(2):83–91
Date: 2009
Reading level: Intermediate
Why This Matters for Freedivers
This paper tackles a question almost every deep diver has heard: “Can you really get DCS from a single breath.” The answer here is: yes, it’s possible, and the risk rises sharply after about ~100 m, reaching a few percent for the deepest sled-style dives. That’s a big deal for planning safety: deep attempts should be treated like a real decompression-risk activity, with oxygen availability and a serious plan for neurological symptoms.
Synopsis
A lot of freedivers assume DCS can’t happen on a single breath because the dive is “too short” and “there isn’t enough nitrogen.” Fitz-Clarke challenges that by looking at the deepest breath-hold dives ever documented and asking a very practical question: what’s the risk of DCS if you do one extreme dive.
First, the paper builds a real-world database: all confirmed human breath-hold dives to 100 m or deeper over a 30-year period (1976–2006). That ends up being 192 dives by 24 divers, across different disciplines (including sled dives). In that set, there were two confirmed DCS cases after single deep dives (plus fatalities unrelated to DCS mechanisms). One of the DCS cases was severe and clearly neurological (vertigo, vomiting, later weakness), requiring multiple chamber treatments and only partial recovery.
Then comes the “how do we turn that into a risk estimate” part. The author adapts probabilistic decompression risk models (normally used for scuba / submarine escape exposures) and adjusts them to behave more like freediving physiology. The key tweaks are: - Lung compression and lung collapse: as you go deeper, your lungs shrink, and at very great depth gas exchange becomes limited and eventually stops increasing nitrogen uptake once the lungs are effectively collapsed. - Diving reflex effects: cardiac output drops with depth and time (bradycardia + vasoconstriction), which changes how fast nitrogen can be taken up. - Brain stays “well supplied”: the model assumes the brain keeps relatively stable blood flow compared with the periphery, which lines up with the observation that breath-hold DCS tends to be mainly neurological.
The headline result is simple and memorable: - Up to about 100 m, predicted DCS risk is negligible for a single dive. - Beyond that, risk rises nonlinearly (it ramps up faster than you might expect). - At extreme depths, risk plateaus (around ~5–7%) because once the lungs are collapsed, you can’t keep absorbing more nitrogen from that single breath in the same way.
One interesting note: the paper addresses the common rumor that an unusually long deep dive time (over 5 minutes) might have “caused” one severe DCS case. The model suggests time matters, but within that extreme-depth zone it changes risk only modestly compared with the depth effect.
The practical takeaway is not “deep diving is impossible.” It’s that once you enter extreme depth territory, the risk is no longer theoretical, and it’s not tiny. The author even mentions that this kind of event led to operational guideline changes (like prophylactic oxygen breathing after deep sled dives).
Abstract
The risk of decompression sickness (DCS) in human breath-hold diving is expected to increase as dives progress deeper until a depth is reached where total lung collapse stops additional nitrogen gas uptake. We assembled a database of all documented human breath-hold dives to 100 metres or greater, including both practice and record dives. Between 1976 and 2006 there were 192 such dives confirmed by 24 divers (18 male, 6 female). The deepest dive was to 209 metres. There were two drowning fatalities, and two cases of DCS. Depth-time risk estimates for DCS were derived for single breath-hold dives by modifying probabilistic decompression models calibrated with data from short deep no-stop air dives and submarine escape trials using maximum-likelihood estimation. Arterial nitrogen levels during apnea were adjusted for lung compression and decreased cardiac output. Predicted DCS risk is negligible up to about 100 metres, beyond which risk increases nonlinearly and reaches a plateau around 5 to 7 percent when total lung collapse occurs beyond 230 metres. Results are consistent with data available from deep breath-hold dives.