Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Cancer Epidemiol Biomarkers Prev. Author manuscript; available in PMC 2013 July 1.
Published in final edited form as:
PMCID: PMC3392523

Reproducibility of the Nicotine Metabolite Ratio in Cigarette Smokers



The nicotine metabolite ratio (NMR or 3-hydroxycotinine/cotinine) has been used to phenotype CYP2A6-mediated nicotine metabolism. Our objectives were to analyze (a) the stability of NMR in plasma, saliva, and blood in various storage conditions, (b) the relationship between NMRs derived from blood, plasma, saliva, and urine, and (c) the reproducibility of plasma NMR in ad libitum cigarette smokers.


We analyzed data from four clinical studies. In studies 1 and 2, we assessed NMR stability in saliva and plasma samples at room temperature (~22°C) over 14 days and in blood at 4°C for up to 72 hours. In studies 2 and 3, we used Bland-Altman analysis to assess agreement between blood, plasma, saliva, and urine NMRs. In study 4, plasma NMR was measured on 6 occasions over 44 weeks in 43 ad libitum smokers.


Reliability coefficients for stability tests of NMR in plasma and saliva at room temperature were 0.97 and 0.98, respectively, and 0.92 for blood at 4°C. Blood NMR agreed consistently with saliva and plasma NMRs but showed more variability in relation to urine NMR. The reliability coefficient for repeated plasma NMR measurements in smokers was 0.85.


The NMR is stable in blood, plasma, and saliva at the conditions tested. Blood, plasma, and saliva NMRs are similar while urine NMR is a good proxy for these NMR measures. Plasma NMR was reproducible over time in smokers.


One measurement may reliably estimate a smoker’s NMR for use as an estimate of the rate of nicotine metabolism.

Keywords: Nicotine metabolite ratio (NMR), cotinine, 3-hydroxycotinine, biological stability, chemical stability


The rate of nicotine metabolism, which influences nicotine dependence and smoking behavior (1, 2), is highly variable between individuals (3). Characterization of the rate of nicotine metabolism in smokers has important epidemiologic and pharmacologic applications. Because nicotine is metabolized primarily by hepatic cytochrome P450 (CYP) 2A6, one approach to measuring the rate of nicotine metabolism has been to genotype CYP2A6 (4). However, genotyping CYP2A6 alone does not account for environmental and regulatory influences on enzymatic activity. Cotinine (COT), the major proximate metabolite of nicotine, is further metabolized, essentially exclusively by CYP2A6, to trans-3′-hydroxycotinine (3-HC) (5, 6). The ratio of the product 3-HC to the parent COT (3-HC/COT, also referred to as the nicotine metabolite ratio or NMR) is an indicator of CYP2A6 enzymatic activity. Since cotinine has a relatively long half-life (averaging 16 hr) and 3-HC intrinsically has a shorter half-life (5 hr) (7, 8), the elimination rate of the latter is formation-limited, meaning that the NMR is expected to be stable over time among regular smokers. The NMR can be measured in the saliva, urine, or plasma of smokers and has been validated as a biomarker of nicotine clearance (5, 9).

While the NMR is being used in a growing body of research to explore the influence of nicotine metabolism on smoking behaviors, dependence, and as a predictor of response to smoking cessation treatment (1013), data on its reproducibility in ad libitum cigarette smokers are limited (14, 15). The long-term reproducibility of NMR measured in saliva or plasma of smokers has not been reported. In addition, there are limited data on the stability of the NMR in biological samples at various storage temperatures. Lea and colleagues showed that saliva NMR was unaffected by storage at room temperature or at −20°C for up to seven days (14).

We investigated the stability of NMR in blood, plasma and saliva stored at various temperatures to better understand how storage conditions after sample collection may affect the NMR. We analyzed the relationships between NMRs derived from whole blood, plasma, saliva, and urine. Finally, we assessed the reproducibility of plasma NMR over 44 weeks as well as the precision of plasma NMR with multiple replicate measurements in a well-defined population of ad libitum smokers.


Studies, subjects and experimental protocols

Study 1

Details of this study are presented elsewhere (16). To investigate the stability of the NMR in plasma and saliva at room temperature over 14 days, we obtained previously frozen (−30°C) saliva and plasma samples collected during the same visit from 24 healthy smokers recruited by newspaper advertisement. Saliva and plasma samples were thawed and stored at room temperature (~22°C) and aliquots of each were analyzed for 3-HC and cotinine on days 1, 4, 7, and 14.

Study 2

This study was a clinical trial of alternative therapies for smoking cessation treatment in smokers (results have not been published). Subjects were studied at the Centre for Addiction and Mental Health in Toronto, Canada. After completing an initial telephone interview that determined preliminary eligibility, the subjects attended an intake session during which whole blood, saliva, and urine samples were collected. Plasma was extracted from one of the collected tubes of whole blood by centrifugation. The samples were frozen and stored at −30°C until analysis within 14 days from the collection date. An additional tube of whole blood was collected from 9 randomly selected subjects and one more subject whose urine and saliva were not available for NMR analysis. These 10 samples were stored at 4°C, aliquots were taken at 2, 18, 24, 48 and 72 hours, transferred to −30 °C until all aliquots were collected and then analyzed in duplicate for the concentrations of 3-HC and COT.

Study 3

Study design has been described in a previous paper (5). This was an outpatient study at the General Clinical Research Center at San Francisco General Hospital. In the current article, we assessed the relationship between saliva and plasma NMR in a sample of smokers (n=67) and nonsmokers (n=11) exposed to tobacco smoke. Subjects were healthy volunteers recruited from newspaper advertisements. Afternoon plasma and saliva samples were collected and analyzed for cotinine and 3-HC.

Study 4

This was a clinical trial of reduced-nicotine content cigarettes in which smokers were randomly assigned to a control or research arm after a 2-week baseline period in which they smoked their usual brand of cigarettes. The control group continued to smoke their usual brand of cigarettes for the duration of the study. Subjects were studied in a community-based clinic. Study design and subject inclusion/exclusion criteria have been described elsewhere (17). In the present analysis, we focused on the first 44 weeks of the control arm of this study to investigate the long-term reproducibility of plasma NMR in smokers who smoked ad libitum. Subjects were 43 healthy smokers of 10 or more cigarettes per day recruited from newspaper advertisements. Plasma samples collected at weeks 1, 8, 16, 24, 32, and 44 and urine samples collected at baseline (week 1) were assayed for concentrations of cotinine and 3-hydroxycotinine. To compare plasma NMR measures between two testing laboratories, the NMR was analyzed in 15 split plasma samples at the University of California, San Fransisco, California, USA (UCSF) and the University of Toronto, Toronto, Canada (UT).

Written informed consent was obtained from each participant. Studies 1, 3, and 4 were approved by the Institutional Review Board at the University of California, San Francisco (UCSF), and Study 2 was approved by the Institutional Review Boards at the University of Pennsylvania, the Center for Addiction and Mental Health, and the University of Toronto (UT).

Analytical chemistry

Analyses of plasma and saliva cotinine and 3-HC concentrations as well as urine cotinine, 3-HC, and their respective glucuronide metabolites for Studies 1, 3, and 4 were carried out at UCSF by liquid chromatography-tandem mass spectrometry (LC-MS/MS) as described previously (18). In quality control (QC) tests of 176 QC plasma samples over a time span of ~ 6 years, the coefficients of variation (CV) for assays of plasma cotinine, 3-HC, and NMR were 8.7%, 10.7%, and 8.2%, respectively. Additional QC and precision data for the assays used are presented elsewhere (18). A variation of the method described above was used at UT to measure cotinine and 3-HC in 15 plasma samples from Study 4 and whole blood, plasma, saliva, and urine samples from Study 2. Specifically, 100 μL of biosamples was used and the volume was brought up to 1 mL with water and urine was diluted 1:20 with water prior to analysis. Assays at UT were run in duplicate. For analysis of total concentration of urinary metabolites the urine sample was deconjugated using incubation with β-glucuronidase (type H-1 from Helix pomatia, 20,000 units/mL in 0.5M ammonium acetate buffer pH 5.1) for 21 hours at 37°C.

Statistical analysis

Biomarker data for all studies were approximately log-normally distributed and therefore log-transformed. Also, log-transformed NMR has been shown to be more related to nicotine clearance than raw-scale NMR (19).

Study 1

To analyze the chemical/physical stability of biomarkers obtained from plasma and saliva samples stored up to 14 days at room temperature (~22°C) we computed reliability coefficients, the ratio of between-sample variance to the sum of between- and within-sample variances, for plasma and saliva log-transformed cotinine, 3-HC, and NMRs from mixed-model analyses. We also conducted a repeated-measures mixed-effects analysis to assess changes of log-transformed plasma and saliva cotinine, 3-HC, and NMR over the 14-day period. We included a fixed effect for time in the models.

Study 2

The reproducibility coefficients and repeated measures analysis to assess the stability of log-transformed cotinine, 3-HC, and NMR in refrigerated whole blood were done as described above. We used a Bland-Altman analysis to assess the level of agreement between whole blood, plasma, saliva, and urine log-transformed NMR, cotinine, and 3-HC values, respectively. We defined a range of agreement as mean bias ±2 standard deviations (SD). The Bland-Altman analysis is based on examination of a scatterplot of variable means plotted on the horizontal axis and differences between variables plotted on the vertical axis (20). We used linear regression to examine the strength of the relationships between biomarkers in different biologic fluids.

Study 3

We performed a Bland-Altman analysis to assess the agreement between saliva and plasma log-transformed cotinine, 3-HC, and NMRs obtained from smokers and nonsmokers as described above. We also fit a linear regression for saliva on plasma log-transformed NMR.

Study 4

The reproducibility of plasma log-transformed cotinine, 3-HC, and NMR in smokers measured during six visits over 44 weeks was estimated as described above. To evaluate the sensitivity of reliability coefficient estimates to outlying observations, we excluded data from one subject who had the highest body mass index (BMI) in the sample (47.5) and whose plasma NMR and 3-HC concentrations were an order of magnitude lower than the mean of these variables for the sample.

We characterized changes in the precision of plasma NMR with multiple replicates of NMR measurements (i.e. sampling at different times). The variance of the mean of n replicated measurements taken on a single subject at different times is:

Equation 1

We computed the reliability of the average of n measurements and the variance inflation factor of x replicate measurements relative to y replicate measurements, where x and y can be any finite number. The reliability of the average of n measurements is the ratio of the between-subjects variance to the total variance of the mean for that subject (Equation 1). The variance inflation factor is the variance of x divided by the variance of y, and indicates how much larger a sample using x replicate measurements has to be to achieve the same precision in NMR as one using y replicate measurements.

As a further investigation of changes in plasma log-transformed cotinine, 3-HC, and NMR, respectively, over time, we used mixed effects repeated-measures analysis with subjects as a random effect and included covariates age, gender, BMI, and cigarettes per day (CPD). Covariate data were all obtained at baseline. Age was treated as a continuous variable and CPD was as a categorical variable (1–10, 11–20, and >20 CPD). In addition, we computed Spearman correlation coefficients between baseline plasma cotinine, 3-HC, and NMR and age, BMI, CPD, time to first cigarette (TFC), FTND, and NDSS overall score and individual NDSS factor scores, all obtained at baseline. These analyses were done with and without data from the subject with outlying observations.

The agreement between urinary NMR derived from total 3-HC (NMRtotal(3-HC)) and NMR derived from free 3-HC (NMRfree(3-HC)) were assessed by Bland-Altman analysis. The rationale for this analysis is that the NMR is intended to represent the product/parent drug ratio for the CYP2A6 pathway. The parent is cotinine and 3-HC is the product. 3-HC is further metabolized to 3-HC-glucuronide, so accounting for all of 3-HC generated by cotinine metabolism via CYP2A6 should include both free and conjugated species. All cotinine is not converted to 3-HC, so total COT is not appropriate as the denominator in the NMR. Thus, we used the free form of urine cotinine, as the substrate for CYP2A6, in NMR computations.

We did a Bland-Altman analysis to assess the level of agreement between the two testing sites’ measurements of cotinine, 3-HC, and NMR in split plasma samples. All analyses were carried out using SAS v. 9.3 (SAS Institute, Inc., Cary, NC, USA) and statistical tests were considered significant at α = 0.05.


NMR chemical/physical stability (Studies 1 and 2)

The reliability coefficients for stability tests of log-transformed biomarkers in whole blood samples stored at 4°C were as follows: cotinine, 0.99; 3-HC, 0.96; and, NMR, 0.92. Changes in whole blood log-transformed cotinine (p=0.687), 3-HC (p=0.090), and NMR (p=0.181) over time were non-significant. The reliability coefficients for stability tests of log-transformed biomarkers in saliva at room temperature were as follows: cotinine, 0.99; 3-HC, 0.99; and NMR, 0.98, which were similar to the reliability coefficients for stability tests of biomarkers in plasma at room temperature: cotinine, 0.99; 3-HC, 0.99; and, NMR, 0.97. Changes in both plasma (p=0.032) and saliva (p<0.001) NMR over time were significant. Changes in cotinine and 3-HC concentrations in these biologic fluids over time were relatively small but significant (all p-values<0.001).

Comparison of plasma NMR between testing sites (Study 4)

The two testing laboratories consistently provided similar measures of plasma NMR (Table 1A). Although plasma 3-HC measurements between the labs were on average closer than cotinine measurements (ratio of 0.95 vs. 0.89), we observed some discrepancies between the testing sites in plasma 3-HC measurements while there was consistent agreement in plasma cotinine measurements.

Bland-Altman analysis of agreement between measures

Urine NMR derived from total 3-HC and free 3-HC (Study 4)

The correlation between urine NMRs derived from total and free 3-HC from Study 4 was strong (r=0.99). Urine NMRtotal(3-HC) was on average 1.21 (95% CI: 1.18–1.25) times larger than urine NMRfree(3-HC) and we demonstrate that the two methods of measuring NMR in urine did not consistently agree (Table 1B). The correlations between plasma NMR and urine NMRtotal(3-HC) (r=0.82) and NMRfree(3-HC) (r=0.84) in that sample were similar.

Relationship between whole blood, plasma, saliva, and urine NMRs (Studies 2 and 3)

Based on Bland-Altman analysis, blood-plasma and blood-saliva comparisons consistently provided similar measures of NMR (Study 2, Table 1C). We observed discrepancies in saliva-plasma, urine-blood, urine-plasma, and urine-saliva NMR comparisons (results are for urine NMRtotal(3-HC) but similar observations were made for urine NMRfree(3-HC)).

Figure 1A–C summarize the linear regression of plasma, saliva, and urine NMRs, respectively, on whole blood NMR measured in smokers enrolled in Study 2. Additionally, correlations between log-transformed NMRs in biologic fluids include: urine NMRfree(3-HC) vs. whole blood NMR, r=0.69; urine NMRtotal(3-HC) vs. saliva NMR, r=0.66; urine NMRfree(3-HC) vs. saliva NMR, r=0.46; urine NMRtotal(3-HC) vs. plasma NMR, r=0.64; urine NMRfree(3-HC) vs. plasma NMR, r=0.56; and, urine NMRfree(3-HC) vs. urine NMRtotal(3-HC), r=0.91 (all p-values<0.01).

Regression of plasma 3-hydroxycotinine to cotinine (3-HC/COT) ratio or nicotine metabolite ratio (NMR) in plasma vs. whole blood (plot A); saliva vs. whole blood (plot B); urine vs. whole blood (using urine total 3-HC/free COT) (plot C) in a sample of ...

Saliva and plasma log-transformed NMR from a sample of smokers and nonsmokers in Study 3 were highly correlated (r=0.95) and is shown in Figure 1D. However, Bland-Altman analysis indicated that plasma and saliva log-transformed NMR from this sample were not consistently similar (Table 1D). Similarly, correlations between plasma and saliva log-transformed 3-HC (r=0.95) and plasma and saliva log-cotinine (r=0.97) were high but we also observed disagreement between these measures in saliva and plasma in Bland-Altman analyses (Table 1D).

Plasma NMR reproducibility with multiple measurements over time (Study 4)

Demographic data, cigarette consumption, and various measures of nicotine dependence for participants in Study 4 are presented in Table 2. Controlling for age, gender, and BMI, we observed small changes in CPD over time (p=0.044).

Demographic, cigarette consumption, and nicotine dependence of participants enrolled in Study 4, which investigated the variation of plasma nicotine metabolite ratio (NMR) in ad libitum smokers over 44 weeks

The geometric means and 95% confidence intervals (CI) of plasma NMR at each measurement occasion for the full sample and plasma NMRs for individual subjects are presented in Figure 2A. Geometric means and 95% CI for plasma cotinine and 3-HC are presented in Figure 2B. All participants had biomarker data for five of the six visits while 38 participants had data for all six visits over the 44-week study period. The reliability coefficients for repeated measurements of these variables were as follows: cotinine, 0.68; 3-HC, 0.79; and, NMR, 0.85. When outliers were omitted, the reliability coefficients were reduced primarily because the between-subjects variance decreased while within-subjects variance was unchanged: NMR, 0.70; 3-HC, 0.63; and cotinine, 0.67. To investigate whether the variability in NMR was higher at larger NMRs, we determined the trend in the mean of NMRs over six visits for each individual vs. the standard deviation (SD) of the NMR for that person. The positive linear trend was significant (p = 2.7 × 10−7).

Plasma nicotine metabolite ratio (NMR or 3-HC/COT) for individual subjects and geometric mean and 95% confidence interval (CI) (plot A) and 3-hydroxycotinine (3-HC) and cotinine (COT) concentrations (geometric means and 95% CI) (plot B) in ad libitum ...

Using repeated-measures analysis and controlling for age, gender, BMI, and CPD, we observed some variation in plasma NMR over the 44-week period (p=0.006). BMI (p<0.001) had a significant effect on plasma NMR but the effects of age (p=0.545), gender (p=0.640), and CPD (p=0.165) were not significant (covariates were collected at baseline). Plasma 3-HC also had small but significant changes over time (p=0.013) (Figure 2B). The effects of BMI (p<0.001), age (p=0.049), and CPD (p=0.001) on plasma 3-HC were significant while the gender effect was non-significant (p=0.556). Plasma cotinine did not change significantly over time (p=0.266). Age (p=0.038) and CPD (p<0.001) had significant effects on plasma cotinine while the effects of BMI (p=0.089) and gender (p=0.140) were non-significant. The results of these analyses were consistent when outliers were omitted but the effect of BMI on plasma NMR and 3-HC, respectively, did not achieve significance. Plasma NMR was not significantly associated to demographic, CPD, and nicotine dependence scales (Table 3).

Spearman rank correlation coefficients between baseline demographic and nicotine dependence and baseline plasma cotinine (COT), 3-hydroxycotinine (3-HC), and NMR (3-HC/COT)

Precision of plasma NMR with replication (Study 4)

The variance in plasma NMR measurements decreased with increasing replicates, which translates to increased precision of plasma NMR (Table 4). The variance of plasma NMR measurements was lower (i.e. precision higher) when outliers are excluded. Table 4 further illustrates a comparison of precision obtained when using x replicate measurements of plasma NMR relative to 2 or 3 replicates. For example, if 1 plasma NMR measurement is used rather than 2, the variance in NMR estimate is increased by 8%. Removal of outliers increased the variance inflation of x replicates relative to y replicates. Thus, with outliers removed the variance of plasma NMR when 1 measurement is used instead of 2 increases from 8% to 18%.

Precision of plasma nicotine metabolite ratio (NMR) with replication of NMR measurements


Main observations

We make several observations in this study that may help guide the application and utility of the NMR in epidemiologic studies and in treatment of nicotine addiction. First, we demonstrate that the NMR is stable in whole blood at 4°C over a 72-hr period and in plasma and saliva at room temperature over a 14-day period. Second, we show that whole blood NMR consistently provided similar measures as plasma or saliva NMRs and that urine NMR can serve as a reasonably good proxy for blood, plasma, and saliva NMR. Finally, in a 44 week-study of ad libitum smokers, the longest such study to date, our results indicate that a single measurement of plasma NMR is relatively reliable (reliability coefficient was 0.85 in full sample).

NMR chemical/physical stability

NMR, cotinine, and 3-HC had relatively high stability in biologic fluids at the conditions tested. The small within-sample variability in plasma and saliva cotinine and 3-HC concentrations observed may be attributable to small degradations over time at room temperature. Within-sample variability accounted for a higher fraction of the total variability in whole blood NMR than that of plasma and saliva (8% vs. 3% and 2%, respectively). This may be due to the smaller number of samples used for whole blood stability tests (10 whole blood samples vs. 24 pairs of saliva and plasma samples). To minimize variability in NMR estimates, we recommend that blood, plasma and saliva samples be refrigerated after sample collection as soon as is logistically feasible in field studies and all samples should be kept frozen during storage and/or transportation to analytical laboratories. Limitations of our findings include assessing the effects of only one storage condition per biologic fluid and not assessing the stability of biomarkers in urine. Also, we did not assess the impact of freeze-thaw cycles on the stability of biomarkers at the conditions tested.

Urine NMR derived using free and total 3-HC and plasma NMR

We showed that although urine NMRtotal(3-HC) and urine NMRfree(3-HC) are highly correlated, the two methods did not provide similar measures of NMR. Urine NMRtotal(3-HC) is more biologically reasonable since it incorporates all 3-HC generated from cotinine, however urine NMRfree(3-HC) explains reasonably well the variation in nicotine metabolism and is technically less time-consuming to measure. Both forms of urine NMR were similarly well correlated to plasma NMR, suggesting that both forms would serve as good proxies for plasma NMR.

Relationship between whole blood, plasma, saliva, and urine NMR

Our results suggest that whole blood NMR can be substituted for plasma and saliva NMR since the measures were consistently similar. On the other hand, urine NMRs did not consistently provide similar NMRs as whole blood, plasma, and saliva but it can serve as a reasonably good proxy for these measures based on moderate to high correlations between these variables. We also show that saliva and plasma NMRs are not consistently similar. Nonetheless, plasma and saliva NMRs are strongly related, indicating that saliva NMR can be used to estimate plasma NMR with reasonably high accuracy.

Plasma NMR reproducibility with multiple measurements over time

The variability within subjects accounted for about 15% to 30% of the total variability observed in plasma NMR measurements (full sample and outliers excluded, respectively) and the variability in NMR increased with increasing NMR. Several sources of intra-individual variability in NMR have been proposed. First, since the stability of NMR depends on steady state level of cotinine and therefore smoking, abrupt changes in cigarette consumption and smoking pattern in the days before sample collection could have led to 3-HC not reaching steady state and incorrect estimation of the NMR (14). There was evidence of slight within-subjects changes in cigarette consumption (CPD) during the study (p=0.044). But since within-subjects plasma cotinine (p=0.266) remained stable while plasma 3-HC changed during the study (p=0.013), changes in plasma NMR were probably not related to variation in CPD. Instead, this indicates that there were times when plasma 3-HC was not at steady state. Other sources of intra-individual variation in plasma NMR over time include the effects of food, drugs, or other environmental exposures on CYP2A6 activity and other enzymes involved in the rates of cotinine clearance from other competing pathways such as glucuronidation or renal clearance (3). Variations in laboratory measurements of these biomarkers may also contribute to the overall variation in plasma NMR. We saw a linear trend of increasing within-subject SD of NMRs with increasing mean NMRs. However, it is common to see increasing SD with increasing mean values for analytical chemistry assays.

Precision of plasma NMR

We demonstrate that replication of NMR measurements (i.e. taking samples on separate occasions) is one method to increase the precision of NMR estimates. For example, the variance in plasma NMR is reduced from 0.080 to 0.072 (on log-scale) by increasing the number of replicate measurements from 1 to 3 (Table 4A). We illustrate the extent to which smaller sample sizes are needed to attain the same precision in plasma NMR measurements when the number of replicates is increased. The precision of NMR estimates is sensitive to the presence of outlying observations in a sample. We believe that as a rule one should not exclude observations simply because they are extreme or influential. The presence of outliers in a sample, however, may suggest within-study differences or errors in measurement or data handling that render some observations incommensurable with the others. If such problems can be identified, then it is appropriate to investigate all observations and exclude those that are not comparable on that basis, but never simply because they are outliers. Nevertheless there is value in conducting statistical analyses both including and excluding influential or extreme observations, as a form of sensitivity analysis.

Correlates of plasma NMR

Baseline plasma NMR was not significantly correlated with demographic data, CPD, and nicotine dependence scores (Table 3). Several studies have reported significant correlations between the NMR and cigarette consumption (9, 10, 15), while others have not found significant correlations (11, 21, 22). These conflicting results are not surprising because CPD is not always predictive of nicotine intake due to variations in puff frequency and depth of inhalation (23). On the other hand, the NMR has been shown to be correlated with total cigarette puff volume (24, 25). Plasma NMR and cotinine were not significantly correlated in this study. Further, we did not find significant correlations between plasma NMR and commonly used scores of nicotine dependence, FTND and NDSS. These results are similar to other published studies (11, 13, 22). In contrast, genetic variation in CYP2A6 has been shown to be significantly associated to FTND scores (26). It should be noted that the currently available scales of nicotine dependence have their limitations, and are not likely to be valid predictors of all aspects of smoking behavior (27). For example, these scores are less predictive of quit attempts than the NMR (11, 28).


Our data show relatively high stability of NMR derived from plasma and saliva samples stored at room temperature over a 14-day period and in refrigerated whole blood over a 72-hr period. Whole blood provides similar measures of NMR as plasma and saliva while urine NMR is a reasonably good proxy for plasma NMR measurement in whole blood, plasma, and saliva. Finally, a single measurement of plasma NMR may reliably estimate a smoker’s true homeostatic rate of nicotine metabolism.


We thank Faith Allen for data management and Drs David Conti and WonHo Lee for assistance in statistical analysis.


Studies described in this report were funded by National Institutes of Health grants DA U01 02830, DA 02277, CA 78603, DA 12393, R25 CA 113710, NCRR UCSF CTSI UL1 RR 024131 and by Canadian Institute of Health (CIHR) operating grants MOP#86471, MOP#115145 and TMH-109787 and an operating grant from the Ontario Mental Health Foundation (OMHF). Dr. Wing was supported by a Postdoctoral Research Fellowship from the Centre for Addiction and Mental Health (CAMH), Dr. George was supported in part by the Canada Foundation for Innovation Leader Opportunity Fund (CFI-LOF, #19229) and the Chair in Addiction Psychiatry at the University of Toronto, Dr. Tyndale was supported in part by CAMH and a CRC.


Conflict of interest statements

Drs. St. Helen, Heitjan, and Wing have no conflicts of interest. Dr. George reports that in the past 2 years, he has received contract support from Pfizer, is a consultant to Novartis, Astra-Zeneca, Eli Lilly, Memory Pharmaceuticals, Evotec, Janssen and Pfizer. Dr. Tyndale has consulted for Novartis and McNeil and holds shares in Nicogen Research Inc., a company that is focused on novel smoking cessation treatment approaches; no Nicogen funds were used in this work. Dr Benowitz consults with Pfizer on smoking cessation medications and has provided paid expert testimony concerning nicotine addiction in litigation against tobacco companies.


1. Ray R, Tyndale RF, Lerman C. Nicotine dependence pharmacogenetics: role of genetic variation in nicotine-metabolizing enzymes. J Neurogenet. 2009;23:252–61. [PubMed]
2. Benowitz N. Clinical pharmacology of nicotine: implications for understanding, preventing, and treating tobacco addiction. Clin Pharmacol Ther. 2008;83:531–41. [PubMed]
3. Hukkanen J, Jacob P, III, Benowitz NL. Metabolism and disposition kinetics of nicotine. Pharmacol Rev. 2005;57:79–115. [PubMed]
4. Messina E, Tyndale R, Sellers E. A major role for CYP2A6 in nicotine C-oxidation by human liver microsomes. J Pharmacol Exp Ther. 1997;282:1608–14. [PubMed]
5. Dempsey D, Tutka P, Jacob P, Allen F, Schoedel K, Tyndale RF, et al. Nicotine metabolite ratio as an index of cytochrome P450 2A6 metabolic activity. Clin Pharmacol Ther. 2004;76:64–72. [PubMed]
6. Mwenifumbo JC, Zhou Q, Benowitz NL, Sellers EM, Tyndale RF. New CYP2A6 gene deletion and conversion variants in a population of Black African descent. Pharmacogenomics. 2010;11:189–98. [PMC free article] [PubMed]
7. Benowitz NL, Jacob P., III Nicotine and cotinine elimination pharmacokinetics in smokers and nonsmokers. Clin Pharmacol Ther. 1993;53:316–23. [PubMed]
8. Benowitz N, Jacob P., III Trans-3′-hydroxycotinine: Disposition kinetics, effects and plasma levels during cigarette smoking. Br J Clin Pharmacol. 2001;51:53–9. [PMC free article] [PubMed]
9. Benowitz NL, Pomerleau OF, Pomerleau CS, Jacob P. Nicotine metabolite ratio as a predictor of cigarette consumption. Nicotine Tob Res. 2003;5:621. [PubMed]
10. Derby KS, Cuthrell K, Caberto C, Carmella SG, Franke AA, Hecht SS, et al. Nicotine metabolism in three ethnic/racial groups with different risks of lung cancer. Cancer Epidemiol Biomarkers Prev. 2008;17:3526–35. [PMC free article] [PubMed]
11. Ho M, Mwenifumbo J, Al Koudsi N, Okuyemi K, Ahluwalia J, Benowitz N, et al. Association of nicotine metabolite ratio and CYP2A6 genotype with smoking cessation treatment in African-American light smokers. Clin Pharmacol Ther. 2009;85:635–43. [PMC free article] [PubMed]
12. Lerman C, Tyndale R, Patterson F, Wileyto EP, Shields PG, Pinto A, et al. Nicotine metabolite ratio predicts efficacy of transdermal nicotine for smoking cessation. Clin Pharmacol Ther. 2006;79:600–8. [PubMed]
13. Johnstone E, Benowitz N, Cargill A, Jacob R, Hinks L, Day I, et al. Determinants of the rate of nicotine metabolism and effects on smoking behavior. Clin Pharmacol Ther. 2006;80:319–30. [PubMed]
14. Lea RA, Dickson S, Benowitz NL. Within-subject variation of the salivary 3HC/COT ratio in regular daily smokers: prospects for estimating CYP2A6 enzyme activity in large-scale surveys of nicotine metabolic rate. J Anal Toxicol. 2006;30:386–9. [PubMed]
15. Mooney ME, Li Z, Murphy SE, Pentel PR, Le C, Hatsukami DK. Stability of the nicotine metabolite ratio in ad libitum and reducing smokers. Cancer Epidemiol Biomarkers Prev. 2008;17:1396–400. [PMC free article] [PubMed]
16. Benowitz NL, Hall SM, Stewart S, Wilson M, Dempsey D, Jacob P., III Nicotine and carcinogen exposure with smoking of progressively reduced nicotine content cigarette. Cancer Epidemiol Biomarkers Prev. 2007;16:2479–85. [PubMed]
17. Benowitz NL, Dains KM, Hall SM, Stewart S, Wilson M, Dempsey D, et al. Smoking behavior and exposure to tobacco toxicants during 6 months of smoking progressively reduced nicotine content cigarettes. Cancer Epidemiol Biomarkers Prev. 2012 In press. [PMC free article] [PubMed]
18. Jacob P, Yu L, Duan M, Ramos L, Yturralde O, Benowitz NL. Determination of the Nicotine Metabolites Cotinine and Trans-3′-Hydroxycotinine in Biologic fluids of Smokers and Non-Smokers using Liquid Chromatography-Tandem Mass Spectrometry: Biomarkers for Tobacco Smoke Exposure and for Phenotyping Cytochrome P450 2A6 Activity. J Chromatogr B. 2011;879:267–76. [PMC free article] [PubMed]
19. Levi M, Dempsey DA, Benowitz NL, Sheiner LB. Prediction methods for nicotine clearance using cotinine and 3-hydroxy-cotinine spot saliva samples II. Model application J Pharmacokinet Pharm. 2007;34:23–34. [PubMed]
20. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. The Lancet. 1986;327:307–10. [PubMed]
21. Berlin I, Gasior MJ, Moolchan ET. Sex-based and hormonal contraception effects on the metabolism of nicotine among adolescent tobacco-dependent smokers. Nicotine Tob Res. 2007;9:493–8. [PubMed]
22. Kandel DB, Hu MC, Schaffran C, Udry JR, Benowitz NL. Urine nicotine metabolites and smoking behavior in a multiracial/multiethnic national sample of young adults. Am J Epidemiol. 2007;165:901. [PubMed]
23. West O, Hajek P, McRobbie H. Systematic review of the relationship between the 3-hydroxycotinine/cotinine ratio and cigarette dependence. Psychopharmacology (Berl) 2011:1–10. [PubMed]
24. Strasser AA, Benowitz NL, Pinto AG, Tang KZ, Hecht SS, Carmella SG, et al. Nicotine metabolite ratio predicts smoking topography and carcinogen biomarker level. Cancer Epidemiol Biomarkers Prev. 2011;20:234. [PMC free article] [PubMed]
25. Moolchan ET, Parzynski CS, Jaszyna-Gasior M, Collins CC, Leff MK, Zimmerman DL. A link between adolescent nicotine metabolism and smoking topography. Cancer Epidemiol Biomarkers Prev. 2009;18:1578–83. [PMC free article] [PubMed]
26. Wassenaar CA, Dong Q, Wei Q, Amos CI, Spitz MR, Tyndale RF. Relationship Between CYP2A6 and CHRNA5-CHRNA3-CHRNB4 Variation and Smoking Behaviors and Lung Cancer Risk. J Natl Cancer Inst. 2011;103:1342–6. [PMC free article] [PubMed]
27. Etter JF. Comparing the validity of the cigarette dependence scale and the Fagerström test for nicotine dependence. Drug Alcohol Depend. 2008;95:152–9. [PubMed]
28. Schnoll RA, Patterson F, Wileyto EP, Tyndale RF, Benowitz N, Lerman C. Nicotine metabolic rate predicts successful smoking cessation with transdermal nicotine: a validation study. Pharmacol Biochem Behav. 2009;92:6–11. [PMC free article] [PubMed]
29. Shiffman S, Waters AJ, Hickcox M. The Nicotine Dependence Syndrome Scale: A multidimensional measure of nicotine dependence. Nicotine Tob Res. 2004;6:327–48. [PubMed]