Carbon Monoxide in Breath in Relation To Smoking and Carboxyhaemoglobin Levels

Nicholas J Wald, Marianne Idle, Jillian Boreham,and Alan Bailey
From ICRF Cancer Epidemiology and Clinical Trials Unit, Radcliffe Infirmary, Oxford, and BUPA Medical Centre, London

ABSTRACT Carboxyhaemoglobin (COHb)levels were studies in 11,249 men. The distribution among the 2,613 men who smoked cigarettes was well separated from that in 6,441 non-smokers (including ex-smokers). The distribution for 2005 cigar and pipe smokers was intermediated, though some of the highest COHb levels occurred in cigar smokers. Using a COHb cut-off level of 2%, 81% of cigarette smokers, 35% of cigar and pipe smokers, and 1-0% of non-smokers had raised COHB levels. In a subsidiary experiment, alveolar air samples were collected from 162 smokers and 25 non-smokers using a simple breath sampling technique. Carbon monoxide concentrations in alveolar breath were highly correlated with COHb levels (r = 0.97) indicating that COHb levels can be estimated reliably by measuring the concentration of carbon monoxide in breath. Alveolar carbon monoxide measurement is thus a simple method of estimating whether a person is likely to be a smoker.

The level of carboxyhaemoglobin (COHb) in the blood has been shown to be a useful marker of tobacco smoke absorption, 1-6 and for this reason there are epidemiological and clinical reasons for performing COHb measurements. The precise distributions of COHb levels in smokers and non-smokers, and the extent to which they overlap, are not well known. We therefore investigated the measurement of carbon monoxide (CO) in alveolar breath using a portable analyzer, as an indirect measure of the COHb level whichavoids the inconvenience of taking blood samples.


Carboxyhaemoglobin and Smoking
The main study population consisted of 11,249 men aged 35-64 years who attended the BUPA Medical Centre in London for a comprehensive health screening examination. On arrival at the Medical Centre, between 1100 and 1700 hours, men were asked about their usual and recent smoking habits, and were not forewarned about the study. The time of smoking each cigarette, cigar, or pipe since waking was recorded. Each man supplied a sample of venous blood and COHb value was determined using the IL182 CO-Oxomiter, as previously described.

Table 1 shows the number of men according to smoking category, together with the definitions used when classifying smoking habits.

table_1Alveolar Carbon Monoxide and Carboxyhaemoglobin
The investigation of alveolar CO and COHb was performed on a subgroup of 187 men (162 smokers and 25 non-smokers) in the main study. Three samples of alveolar breath were collected at two-minute intervals within about five minutes after blood was taken for the COHb estimation. Without a previous deep inspiration, each subject held his breath for 20 seconds, then exhaled steadily via a respiratory valve and plastic tube 1 m long and 17 mm internal diameter into a three-liter anesthetic balloon. The gas in the proximal end of the plastic tube was sampled by means of a small side tube 5 mm away from the mouthpiece. CO measurements were performed using an Ecolyser (Energetics Science Inc, NY). This is a portable instrument (33cm X 19 cm X 18 cm, and weighing 4 kg), which can be operated on the mains or by means of its own rechargeable batteries. The CO measurement is based on the rate of conversion of CO is based on the rate of conversion of CO to CO2 when passed over a catalytically active electrode in an aqueous electrolyte. All alveolar breath samples were analyzed immediately after collection to avoid mixing with dead space gas. The mean of the last two readings was recorded, to the nearest 0.25 ppm. The calibration of the instrument was checked between each set of reading using a sample of known concentration CO gas.

It is known that alcohol vapor can affect the response of the Ecolyser and this was confirmed. Subjects were asked if they had drunk alcohol on the day they were seen and were excluded if they had done so.


Carboxyhaemoglobin and Smoking
table_2 Table 2
shows the COHb distribution in non-smokers (including both lifelong non-smokers and ex-smokers) and men who smoked cigarettes only. There is very little overlap in the two distributions; for example, 81% of the cigarette smokers had COHb levels equal to or greater than 2%, but only 1% of non-smokers had COHb levels as high as this.

The distributions of COHb levels from lifelong non-smokers and ex-smokers were very similar, although a slightly higher proportion of ex-smokers had COHB levels greater than or equal to 2.0% (1.2% and 0.7% of ex-smokers and lifelong non-smokers respectively), suggesting that men claiming to be ex-smokers were more likely to have smoked within a day or so of the test than men claiming to be lifelong non-smokers. For example, the only four “non-smokers” with COHb levels greater than 5.0% were ex-smokers, two of whom reported having smoked within 24 hours of the test!

Figure 1 shows the cumulative COHb distributions for men who smoked cigarettes only, pipes only, cigars only, and non-smokers. The COHb levels of pipe and cigar smokers were similar and, in general, closer to the non-smoker levels than to those of the cigarette smokers, although 5% of cigar smokers with the highest COHb levels had values very similar to those of the 5% of cigarette smokers with the highest levels.

table_3 Table 3 shows the percentage of non-smokers, cigarette smokers (including those who also smoked pipes and /or cigars), and the remaining pipe and/or cigar smokers with COHb levels greater than or equal to the specified cut-off levels.

Alveolar Carbon Monoxide and Carboxyhaemoglobin
Figure 2
shows the relationship between alveolar breath CO concentration and COHb in 162 smokers and 25 non-smokers. A linear regression was performed, the equation for the regression line  being CO (ppm) = 5.09 X COHb (% haemoglobin saturation) + 2.34. The correlation coefficient was 0.97. This high degree of correlation indicated that a COHb level could be estimated reliably from an alveolar level, and the equation for this can be obtained from the same data, using a linear regression of COHb on CO, giving COHb (%haemoglobin saturation) = 0.18 CO (ppm) – 0.14.



The small overlap in the distribution of COHb levels from cigarette smokers and non-smokers was striking, and to some extent surprising in view of the short half-life of COHb in blood.8 The men were seen at the medical center after 1100 hours and it is possible that if COHb estimations had been performed earlier in the day, the discrimination between smokers and non-smokers would have been less clear.


Table 4 shows the improbability of being a smoker for men with COHb cut-off levels, according to the percentage of cigarette smokers in the population assuming all smokers to be cigarette smokers. The probability that a man with a COHb level of greater than or equal to 2.5% is a smoker is greater than 98% for every smoking prevalence shown.

The correlation between alveolar CO concentration and COHb is very close, and the correlation coefficient (0.97) is consistent with the results of previous studies.3 9 10 The relationship between the two variables appears to be approximately linear, the regression equation differing only slightly from that obtained by other workers, 1 11 and is in close agreement with that reported by Jones et al.10

Our study demonstrated that the Ecolyser is a simple method for measuring alveolar CO levels, provided precautions are taken to eliminate the possibility of false readings because of alcohol interference. The instrument, unlike the Infra-Red Gas Analyzer, is not sensitive to CO2 or water vapor, two factors which make it suitable for use on breath samples. The method is rapid and easy to perform, and enables a person’s COHb level to be estimated while avoiding the discomfort and inconvenience of collecting blood. The use of the Ecolyser has previously been shown to discriminate well between smokers and non-smokers.12 The CO results measured on an Ecolyser can be seen by the subject, which is likely to make it useful as a method of reinforcing medical advice to stop smoking – for example, in vascular clinics, in internal clinics, and perhaps even in general practice. It is also likely to be useful when judging the reliabilityof statements made about a man’s smoking habits, and reference to table 3 will provide an approximate indication of whether a person is telling the truth.


We thank the Medical Research Council for part of our financial support. Jillian Boreham is a Liang Research Fellow in Preventative Medicine.


1. Ringold A, Coldsmith JR, Helwig HI, Finn R, Schuette F. Estimating recent carbon monoxide exposures. Arch Environ Health 192; 5: 308-18.

2. Cohen SI, Perkins NM, Ury HK, Goldsmith JH. Carbon monoxide uptake in cigarette smoking. Arch Environ Health 1971; 22: 55-60.

3. Rea JN, Tyrer PJ, Kasap HS, Beresford SAA. Expired air, carbon monoxide, smoking, and other variables. Br J Prev Soc Med 1973; 27: 114-20.

4. Goldsmith JR, Aronow WS. Carbon monoxide and coronary heart disease: a review. Environ Res 1975; 10:236-48.

5. Wald N, Howard S, Smith PG, Bailey A. Use of carboxyhaemoglobin levels to predict the development of diseases associated with cigarette smoking. Thorax 1975; 30:133-40.

6. McIlvaine PM, Nelson WC, Bartlett D. Temporal variation of carboxyhaemoglobin concentrations. Arch Environ Health 1969; 19:83-91.

7. Wald N, Idle M, Bailey A. Carboxyhaemoglobin levels and inhaling habits in cigarette smokers. Thorax 1978; 33:201-6.

8. Coburn RF, Forster RE, Kane PB. Considerations of the physiological variables that determine the blood carboxyhaemoglobin concentration in man. J Clin Invest 1965; 44:1899-10.

9. Goldsmith JR, Landaw SA. Carbon monoxide and human health. Science 1968; 162: 1352-9.

10. Jones RH, Ellicott MF, Cadigan JB, Gaensler EA. The relationship between alveolar and blood carbon monoxide concentrations during breath-holding. J Lab Clin Med 1958; 51:553-64.

11. Rawbone RG, Coppin CA, Guz A. Carbon monoxide in alveolar air as an index of exposure to cigarette smoke. Clin Sci Mol Med 1976; 51:495-501.

12. Coburn RF, Forster RE, Kane PB. Considerations of the physiological variables that determine the blood carboxyhaemoglobin concentration in man. J Clin Invest 1965; 44:1899-10.

13. Vogt TM, Selvin S, Widdowson G, Hulley SB. Expired air carbon monoxide and serum thiocyanate as objective measures of cigarette exposure. Am J Public Health 1977; 67:545-9.