Cause-of-death matching:Any-listed cause — T59.0 anywhere in the cause-of-death chain (up to 20 contributing cause fields per death record), not limited to the underlying cause of death.
Comparison to JAMA:Yockey & Hoopsick (JAMA Network Open 2025) reported ~1,240 deaths using underlying-cause-only methodology, which counts T59.0 only when it is the single primary cause. Our any-listed methodology captures cases where N₂O toxicity was a contributing factor but another cause (e.g., cardiac arrest, asphyxiation) was coded as primary.
Result:Any-listed methodology yields 1,420 deaths (2010–2023) vs. JAMA-reported ~1,240. The difference reflects cases where N₂O contributed to death but was not the sole coded cause.
KEY FINDINGS
All 51 U.S. jurisdictions (50 states + District of Columbia) have recorded nitrous oxide deaths — zero exceptions.
45 states have death counts suppressed per NCHS confidentiality threshold (fewer than 10 deaths per state-year cell). Only federal-level microdata access reveals the full national scope.
6 states with unsuppressed totals: California (281), New York (53), Florida (48), Texas (39), Colorado (22), Pennsylvania (11).
Any-listed methodology yields 1,420 deaths (2010–2023) vs. JAMA-reported ~1,240 using underlying-cause-only methodology.
2024 provisional data: 186 deaths — the most recent year available, not yet published in peer-reviewed literature.
SUPPRESSION EXPLAINER
CDC’s National Center for Health Statistics applies cell suppression to all publicly released mortality data. Any count fewer than 10 in a given state-year cell is replaced with “Suppressed” to prevent potential identification of decedents.
This means that for 45 of 51 jurisdictions, the public-facing CDC WONDER system shows no numeric death count for N₂O. A researcher querying CDC WONDER would see “Suppressed” for nearly every state — creating the false impression that nitrous oxide deaths are isolated to a handful of large states.
The reality is the opposite: every single jurisdiction has recorded deaths. Seeing the full picture requires access to the restricted-use microdata files, which contain individual death records without cell suppression. These files confirm that N₂O mortality is a genuinely national phenomenon, not a regional one.
This suppression dynamic is directly relevant to federal legislation: state legislators reviewing CDC WONDER may conclude their state has no N₂O death problem, when in fact every state does.
YEARLY TREND TABLE
YEAR
DEATHS
NOTES
2010
26
2011
29
2012
34
2013
41
2014
45
2015
52
2016
131
First major spike (+152% YoY)
2017
99
2018
133
2019
168
2020
155
COVID-19 pandemic year
2021
149
2022
169
2023
189
Highest confirmed year
2024*
186*
Provisional — not all deaths may be coded yet
2010–2023 Total
1,420
Any-listed cause (JAMA comparable: ~1,240 underlying-cause-only)
2010–2024 Total
1,606
Includes 2024 provisional
CITATION
Primary peer-reviewed source: Yockey RA, Hoopsick RA. Nitrous Oxide–Related Deaths in the United States, 2010–2023. JAMA Network Open. 2025;8(2):e2457955. doi:10.1001/jamanetworkopen.2024.57955
2024 provisional data may increase as death certificate coding completes. Final 2024 counts are typically released 12–18 months after the calendar year.
Any-listed methodology counts T59.0 as a contributing cause of death, yielding higher totals than underlying-cause-only counts. Both methodologies are valid; they answer different questions. JAMA’s underlying-cause methodology asks “how many people died of N₂O?” Our any-listed methodology asks “how many deaths involved N₂O?”
ICD-10 T59.0 covers “toxic effects of nitrogen oxides” broadly, which may include rare industrial or environmental exposures in addition to recreational nitrous oxide use. However, the age distribution (median age 22) and trend pattern strongly suggest the vast majority are recreational.
State-level public data is limited by NCHS suppression rules. The 6-state unsuppressed totals shown above are from CDC WONDER; remaining states require restricted-use microdata access.
