Conventional cigarette smoking and environmental tobacco smoke have significant health effects [1–4]. Smoking is considered to cause cancer, stroke, and heart disease and also to have a close relationship with the occurrence of gastric ulcers, periodontal disease, sudden infant death syndrome, and metabolic syndrome [5–8]. Levels of cotinine, a metabolite of nicotine [9–11], in the blood track with exposure to tobacco smoke [1]. Promoting smoking cessation should therefore be a major priority in all countries [12] and for all health professionals in all clinical settings; however, it has been pointed out that in 2007 39.9% of Japanese men were current smokers [13]. It has been reported that even light smokers (1–10 cigarettes/day) have some increased risk of adult mortality [14]. Against this background, it has been reported that blood cotinine levels for nonsmokers in the United States population have decreased by about 70% in the past 15 years, indicating that public health interventions to reduce exposure have been successful (http://www.cdc.gov/exposurereport/), but such information in other groups or countries is limited [15,16]. Nicotine metabolite (cotinine and 3′-hydroxycotinine) concentrations in plasma and/or expired carbon monoxide have been examined as validated biomarkers of self-reported cigarette numbers associated with nicotine-metabolizing enzymes cytochrome P450 2A6 (CYP2A6) and CYP2B6 genotypes [17,18].

We reported that white blood cell counts are sensitive biomarkers for smoking exposures and time-dependent recovery in healthy volunteers, and such data might be useful in education and in monitoring cigarette smoking and cessation [3]. The aim of the present study was to carry out a biomonitoring approach using nicotine and its primary and secondary metabolites, which are widely used as biomarkers for tobacco smoke in Japanese males. We report herein that the biomonitoring of cotinine in the urine is a good, easy-to-use marker for plasma levels of the sum of nicotine metabolites in Japanese smokers independent of genetic polymorphism of nicotine-metabolizing enzyme, CYP2A6. The number of cigarettes smoked per day in subjects with light smoking habits could also be predicted by measuring the urinary cotinine concentrations and using linear regression equations established in this study.

The subject population consisted of 92 male Japanese smokers with a mean age of 37 years who had smoked an average of 23 cigarettes per day for 16 years (Table 1). Histograms of the concentrations of nicotine and its metabolites in plasma samples taken from these smokers one hour after smoking the first cigarette on the sampling day are shown in Figure 1. From the same data, the mean plasma levels of nicotine, its metabolites cotinine and 3′-hydroxycotinine, and the sum of the levels of these three compounds in the study population were 20.1, 158, 27.7, and 198 ng/mL, respectively (Table 2). The 95th percentile values of nicotine, cotinine, 3′-hydroxycotinine, and the sum were 33.5, 374, 89.1, and 446 ng/mL, respectively. Similar histograms of nicotine and its metabolite concentrations in urine samples are shown in Figure 2 at 1 hour after smoking the first cigarette on the sampling day; the mean and 95th percentile values of nicotine, cotinine, 3′-hydroxycotinine, and the sum were 846, 1,020, 1,010, and 2,870 ng/mL, and 3,540, 2,450, 3,060, and 7,470 ng/mL, respectively (Table 2).

Plasma levels of 3′-hydroxycotinine (Figure 3C) and urinary levels of nicotine (Figure 4A) and 3′-hydroxycotinine (Figure 4C) were dependent on the CYP2A6 phenotype group estimated from the subjects’ CYP2A6 genotypes, including those with impaired function or whole gene deletion mutations. Concentrations of the secondary metabolite 3′-hydroxycotinine in plasma and urine were significantly lower in expected poor metabolizers; however, urinary unmetabolized nicotine levels in the same group were high. Cotinine levels in plasma (Figure 3B) and urine (Figure 4B) samples were not affected by the CYP2A6 phenotype under daily cigarette smoking conditions. The sum of the metabolite concentrations in plasma (Figure 3D) and in urine (Figure 4D) were also independent of the CYP2A6 phenotype. Urinary concentrations of cotinine and the sum of nicotine metabolites also showed significant correlations (r = 0.75 and 0.65) with their plasma levels (Figure 5), indicating that urinary cotinine concentration was a good marker for plasma levels of nicotine metabolites.

Plasma cotinine levels were significantly correlated with the number of cigarettes smoked on the day before sampling (Figure 6A, r = 0.71), the average number of cigarettes smoked per day (Figure 6B, r = 0.58), and the Brinkman index (Figure 6C, r = 0.48) under the present conditions (n = 91–92).

The sum of nicotine, cotinine, and 3′-hydroxycotinine concentrations in plasma showed similar relationships (Figures 6D–F, r = 0.52–0.69). The relationship between urinary concentrations of cotinine showed low but significant correlation coefficients with the number of cigarettes smoked on the day before sampling (Figure 7A, r = 0.52), the average number of cigarettes smoked per day (Figure 7B, r = 0.47), and the Brinkman index (Figure 7C, r = 0.24) under the present conditions (n = 91–92). Similar correlation coefficients were seen for the sum of nicotine metabolites (Figures 7D–F).

Urinary cotinine excretion in self-reported light smokers was also measured (Table 3). From the slopes of the good linear regression lines shown in Figures 5A and 5B, plasma cotinine levels and the sum of nicotine metabolites were estimated by dividing the urinary cotinine concentration by 6.13 (r = 0.75) or the sum of metabolite concentrations in urine by 13.7. The number of cigarettes smoked on the previous day was then predicted by dividing the estimated plasma cotinine level by 8.24 (the slope in Figure 6A, r = 0.71) or the sum of metabolites in plasma by 10.4 (the slope in Figure 6D). When the urinary cotinine concentration was used, the actual number of cigarettes smoked on the preceding day as obtained by interview was able to be predicted using the positive relationships found in the present study.

It has been reported that single cigarette smoking might be a tool for CYP2A6 phenotyping in those affected by loss-of-function gene mutations [22]. However, under daily smoking conditions, cumulative cotinine concentrations in the urine or plasma may not be affected by the CYP2A6 genotypes, as shown in Figures 3B and 4B, which is consistent with our findings that nicotine could be metabolized by CYP2A6 and CYP2B6 in human liver microsomes [23]. Roles of exptrahepatic CYP2A13 [24,25] could not be ruled out in this nicotine metabolism. Significant, but relatively low correlation coefficients between the concentration of nicotine metabolites in biological fluids and the Brinkman index (number of cigarettes smoked × years) were also seen in the present study (Figures 6 and 7). Cotinine levels may be the best current biomarker for cigarette smoking [16] or secondhand smoke exposure [12], as reported previously.

Smoking is one of the most important behavioral risk factors for premature death worldwide [14]. The present results indicate that biomonitoring of the urinary cotinine concentration is a good, easy-to-use marker for plasma levels of cotinine and is convenient for predicting the number of cigarettes smoked on the preceding day under daily smoking conditions, independent of genetic polymorphism of CYP2A6.