Anaesthesia, 1982, Volume 37, pages 285-288

Nitrous oxide analgesia and altitude

M . F. M . J A M E S , E . D. M . M A N S O N AND J . E. D E N N E T T

Summary
The analgesic ej/ectiveness of 50% nitrous oxide and oxygen or oxygen-enriched air, measured by
variations in pain threshold, was studied at an altitude of 1460 m and simulated altitudes of sea level
and 3300 m.
At sea level pressures, SO% nitrous oxide exerted a similar anulgesic effect to that found for morphine,
raising the pain thresholds by a mean o f 71.5%. At 1460 m, the increase in puin threshold produced
was 40% and at 3300 m the increase in pain threshold was only 19%.
The difference between the analgesic elfects qfnitrous oxide at each ultitude was statistically significant.
It is concluded that moderate altitudes signijkantly reduce the effectiveness of nitrous oxide in a manner
directly related to the partial pressure of nitrous oxide at each altitude.

Key words
Analgesia; measurement.
Hypobaria; altitude.

There is a widely held clinical impression that the
effectiveness of nitrous oxide is considerably
reduced at moderate altitudes as a consequence of
the reduction in barometric pressure and thus
of the partial pressure of any given concentration
of the gas. There has, however, been very little
scientific study of the problem. Safar & Tenicela'
concluded that, a t altitudes of 3000 m and above.
nitrous oxide could not be employed effectively
without the risk of hypoxia and advised against
its use.
Powell & GingrichZ studied the use of 70%
nitrous oxide in oxygen using a relaxant/
ventilation technique on 14 patients at an altitude
of 1 mile (1600 m). They found an arterial Poz
of less than 9 kPa in two patients and a further
two who appeared inadequately anaesthetised,
one of whom recalled the procedure. They con-
cluded that unsupplemented nitrous oxide in
oxygen was not an adequate or safe anaesthetic
at their altitude. At the altitude in Salisbury,
Zimbabwe (1460 m) the use of unsupplemented
nitrous oxide/oxygenjrelaxant techniques for
Caesarean section was abandoned after a short
study showed an incidence of awareness of five
cases out of 30.
In an attempt to evaluate the problem,
Cleaton-Jones et aL3 studied a number of
variables in volunteers given nitrous oxide at sea
level and at 1700 m and found only marginal
differences. However, their study was conducted
by different observers with different subjects at

M.F.M. James, MB, ChB, FFARCS, Senior Lecturer, E.D.M. Manson, MB, ChB, Registrar, J.E. Dennett, SRN,
Nursing Sister, Department of Anaesthetics, Godfrey Huggins School of Medicine, P.O. Box A 178, Avondale,
Salisbury, Zirnbabwc.
0003-2409/82/030285 + 04 SOZ.OO/O @ 1982 The Association of Anaesthetists of Gt Britain and Ireland 285
286 M.F.M. James, E.D.M. Munson und J.E. Dennett
each altitude, making direct comparisons diffi-
cult. In addition, analgesia, which is perhaps the
most clinically useful action of nitrous oxide, was
not extensively studied.
It was therefore decided to study the influence
of altered barometric pressure on the degree of
analgesia produced by 50% nitrous oxide using
the same subjects and observers at each pressure
levcl .
Methods
The study was conducted in a large pressure
chamber capable of accommodating with ease six
people and all necessary equipment. Three levels
of barometric pressure were used: 84 kla, this
being the ambient pressure at Salisburys altitude
of 1460 m (altitude 2); and simulated altitudes
of sea level, a barometric pressure of 100 kPa
(altitude 1) and 3300 m, a barometric pressure of
69 kPa (altitude 3).
Pain thresholds were determined by the
application of pressure to the bare area of the
tibia using a spring balance after the method of
Dundee & Moore.4 Each 1 Ib division on the
balance scale was arbitrarily designated as one
pain unit.
Twenty healthy adult volunteers of both sexes,
all normally resident at 1460 m altitude, were
studied. At each pressure level pain threshold
measurements were made whilst subjects
breathed air, 507; nitrous oxide in oxygen (gas
A) and 507 nitrous oxide in oxygen-enriched air
(gas B). The composition of gas B was adjusted
so that at each altitude level it contained a partial
pressure of oxygen of 21 kPa and the balance as
nitrogcn (Table 1). Gas B was included in the
study to reveal any variation as a result of altered
oxygen tension and to investigate any possible
influence of nitrogen.
All gases were inhaled from a Mapleson A
Table 1. Composition of gas A and gas B at each altitude level
Gas A
Barometric
pressure Nitrous oxide
kPa k Pa
~~
Altitude 1 101.5 50.5
(sea level)
Altitude 2 85.0 42.5
(I460 m)
Altitude 3 69.0 34.5
(3300 m)
system which was equipped with a n anti-pollution
hooded expiratory value from which the expired
gases were collected in Douglas bags to prevent
contamination of the chamber.
Tests of pain threshold were made after the
subjects had inhaled each gas for 10 minutes,
previous studies having shown that the maximum
effect or nitrous oxide is achieved in this
At the time of testing expired carbon dioxide
levels were measured using a Beckman infrared
analyser. Both the altitude sequence and the
order in which gases A and B were inhaled were
varied from subject to subject according to a
Latin square design and neither the volunteers
nor the investigator testing pain thresholds were
aware of the altitude or gas sequence used.
Results
The results were tested using the analysis of
variance for repeated measurements. Whilst there
was considerable inter-subject variation in resting
pain threshold, there was no significant difference
between the mean baseline values at the three
different altitudes.
At sea level pressure nitrous oxide produced
a significant increase in pain threshold of a mean
of 71.5: from the baseline (Table 2) for gas A
and gas B (p<O.Ol). At this barometric
pressure three subjects complained of nausea and
two other subjects became semi-comatose during
the inhalation period.
At a pressure of 85 kPa (1460 m altitude)
nitrous oxide again produced a significant
increase in pain threshold of a mean of 40h for
the two gas mixtures but this increase was
markedly less than that found at sea level. There
were no side-effects at this altitude. When the
barometric pressure was reduced to 69 kPa (a
simulated altitude of 3300 m) nitrous oxide
produced only a 197; increase in pain threshold
Gas B
Oxygcn Nitrous oxide Oxygen Nitrogen
k Pa kPa kPa kPd
50.5 50.5 21.3 29.2
42.5 42.5 21.3 21.2
34.5 34.5 21.3 13.7
 Nitrous oxide analgesia and altitude 287

Table 2. Changes in pain threshold (SEM) produced by each gas at each altitude level, and the mean carbon
dioxide levels. All figures other than percentages and carbon dioxide levels are expressed in arbitrary pain units
(see text)

Mean 7; Mean expired

Air Gas A Increase Gas B Increase increase PCO, (kPa)

Altitude 1 8.6 14.8 + 6.2 14.7 + 6.1 71.5 4.59

(sea level) (4.26) (3.01) (p < 0.01) (3.51) (p i0.01)
Altitude 2 8.8 12.3 + 3.5 12.4 t 3.6 40.0 4.51
(1460 m) (3.13) (3.12) (p < 0.01) (2.25) (p < 0.01)
Altitude 3 9.0 10.5 + 1.5 10.9 + 1.9 19.0 4.59
(3300 m) (3.59) (2.03) (p>0.05) (2.40) (p > 0.05)

which failed to reach statistical significance

(p > 0.05).
There was a significant difference between the
elevation of pain threshold produced a t sea level
and that found at the other two levels (p < 0.01
for altitudes 2 and 3). There was also a significant
difference between the analgesia produced at
1460 and that at 3300 m (p < 0.05; Fig. 1. Table
2). At no altitude level was there any difference
between the analgesia produced by nitrous oxide
in oxygen (gas A) or that of nitrous oxide and
oxygen in air (gas B). Expired carbon dioxide
levels were unchanged at each altitude (Table 2).
Subjects commented that the maximum sub-
jective effects of nitrous oxide were achieved
rapidly and either remained the same or in some
cases appeared to diminish towards the end of
the inhalational period.
Altitude level
Fig. 1. Histogram (SD) showing effect of altitude on
changes in pain threshold for each gas mixture.
* p < 0.01 (altitude 1 and 2 for both gases); f p < 0.01
(altitude 1 and 3 for both gases); f p < 0.05 (altitude
2 and 3 for both gases). 0 Gas A; gas B.
Discussion
From these results it would appear that altitude
does exert a considerable influence on the
analgesic effectiveness of nitrous oxide. It is also
likely that this effect is due mainly to the
reduction in partial pressure that occurs at
altitude. As may be seen from Table 1, the partial
pressures of nitrous oxide a t the three altitude
levels studied approximate to sea level concentra-
tions of 50, 42 and 34%. The fact that nausea
and semi-comatose states occurred only at sea
level is in reasonable agreement with Parbrook's
classification of the levels of nitrous oxide
analgesia.'
In addition, the degree of analgesia produced
by SO"/, nitrous oxide at sea level was slightly
greater than that produced by morphine 0.2
mg/kg in a previous study using the same testing
apparatus.g This is in agreement with other
studies which found SO% nitrous oxide to be
superior to 100 mg pethidines and 25% nitrous
oxide to be better than morphine 0.16 mg/
kg.*
The only study of which we are aware in which
direct comparisons were made of the analgesic
effects of various concentrations of nitrous oxide
was by Whitwam et d SIn their study 33%
nitrous oxide produced an increase in pain
threshold of 23.5%, 407i nitrous oxide raised the
pain threshold by 27.2% and 50% nitrous oxide
produced an increase in pain threshold of 74.1%
when the nitrous oxide was increased in stepwise
manner. These figures are in reasonable agree-
ment with ours a t each altitude level (Table 2).
However, when 33 or 50% nitrous oxide was
administered ab initio in their study, no difference
was found between the analgesia produced at
either concentration. We were unable to confirm
this. Our finding that, in some cases, the
subjective effect of nitrous oxide appeared to
288 M . F . M . James. E.D.M. Manson and J.E. Dennett
decrease with time also agrees with previous
It is unlikely that factors other than altered
partial pressure or nitrous oxide affected the
results. The fact that we obtained almost identical
results with each gas mixture virtually excludes
alterations in partial pressure of oxygen or
nitrogen from playing a significant role. This
latter observation is confirmed by the constancy
of baseline readings at each altitude.
Since hypoxia could not occur with any gas
mixture, it is not surprising that expired carbon
dioxide levels should remain constant and thus
changes in the partial pressure of carbon dioxide
did not affect the results. It should be noted,
however, that all our subjects were acclimatised
to moderate altitude where the normal arterial
carbon dioxide tension is of the order of 4.5 kPa.
It is conceivable that results slightly different from
those found in this acute experiment might be
found in people chronically adapted to each
altitudc, but as our results are in accord with
clinical impressions of the effectiveness of nitrous
oxide at altitude, such differences are likely to
be minimal.
It is concluded therefore that the effectiveness
of any given concentration of nitrous oxide as
an analgesic is markedly reduced by altitude as
a result of the reduction in partial pressure.
Anaesthetists working at even moderate altitudcs
should allow for this effect in their use of this
agent.
References
1 SAFARP, TENICELA
R. High altitude physiology in
relation to anesthesia and inhalation therapy.
Anesfhesiology 1964; 25: 5 15-3 1.
2 POWELLJN, GINGRICHTF. Some aspects of
nitrous oxide anesthesia at an altitude of one mile.
Anesthesia and Anuigesia; Currenf Rcscurchcs 1969;
48: 68C.5.
3 CLEATON-JONU: P. MOYESDG, WHITTAKER i\M.
Clinical effects of nitrous oxide and oxygen
mixtures at sea-level and at 1700 metres altitude.
Anaesthesia 1979: 34: 859-62.
4 DUNDEEJW, MOOREJ. Alte~.ationsin response to
somatic pain associated with anaesthesia, I. An
evaluation of a method of analgesirnetry. Brifi.c.h
Journal of Anaesthesia 1960; 32: 39&406.
5 DUNDEEJW, MOOREJ. Alterations in response lo
somatic pain associated with anaesthesia. IV. The
effect of sub-anaesthetic concentrations of in-
halation agents. British Journal nf Anae?thesiu
1960; 3 2 453-9.
6 PARBROOK G D , REESGAD. ROBERTSOY GS. Relief
of post-operative pain: comparison of a 250/,,
nitrous-oxide and oxygen mixture with morphine.
British Mecliccd Journul 1964; 2: 48&2.
7 PARBROOKG D . The levels of nitroua oxide
analgesia. Brifish Journal of Anaesthesia 1967; 3 9
97482.
X WHITWAM J G , MORGANM, HALLGM, PETRIEA.
Pain during continuous nitrous oxide administra-
tion. British Journui ofAnaesthesia 1976; 48:425-9.
9 JAMESMFM. DIJTH16 AM. DUFN BL, MCKEAG
AM. RICECP. Analgesic effcct of ethyl alcohol.
Brifish Journril o/ Anursthesia 1978; 50: 139-41.
10 KEKHF. HOSKIMMR, BROWNM G , EWINC DJ,
IRVINGJB, KIRBY BJ. A double-blind trial of
patient-controlled nitrous oxidejoxygen analgesia
in myocardial infarction. Lancer 1975: 1: 1397-400.