Nicotine cut-off value in human hair as a tool to distinguish active from passive smokers: A cross-sectional study in Japanese men

Abstract

BACKGROUND:

Nicotine concentration in hair is a useful marker of tobacco exposure. Detection of nicotine in the hair of non-smokers indicates passive smoking. Accurate measurement of nicotine among active and passive smokers can help in smoking cessation programs or programs designed to prevent secondhand smoke exposure.

OBJECTIVE:

To establish, using high-performance liquid chromatography-ultraviolet detection (HPLC/UV), a hair nicotine cut-off value to distinguish active from passive smokers.

METHODS:

Hair samples were collected from randomly chosen Japanese men ( $n=$ 192) between 2009 and 2011. Nicotine and cotinine levels in hair were measured using HPLC/UV with column-switching. T-tests and chi-square tests were performed to compare active and passive smokers, while receiver operating characteristic curves were used to evaluate the effectiveness of the cut-off value.

RESULTS:

There were 69 active smokers and 123 passive smokers. The nicotine and cotinine concentrations in hair were significantly higher in active than in passive smokers ( $p<$ 0.01). The area under the curve for nicotine was 0.92. A hair nicotine cut-off value of 5.68 ng/mg, with a sensitivity of 94.2% and specificity of 87.0%, was identified as the optimal cut-off value for separating active from passive smokers.

CONCLUSION:

Nicotine and cotinine concentrations in hair clearly distinguished active from passive smokers.

Keywords

Screening cut-off value nicotine hair samples active smoker passive smoker

1. Introduction

Smoking is a serious public health issue that is widely recognized as a preventable risk factor for serious illnesses, such as cancer [1, 2] and cardiovascular disease [3, 4]. Exposure to passive smoke is also known to lead to numerous health problems, particularly among children [5, 6]. Distinguishing active smoking from passive smoking in the general population via screening tests has a number of benefits, including establishing the effectiveness of smoking cessation interventions. In addition, early screening may prevent exposure to further passive smoking for non-smokers. However, in order to distinguish active from passive smokers in the general population, it is important to biologically monitor smoking exposure.

Although previous studies have reported that nicotine and cotinine levels can be measured in the blood, urine, and saliva of smokers [7, 8, 9], the amount of nicotine and cotinine in these samples may reflect acute exposure to smoking rather than habitual smoking. Furthermore, because levels of nicotine in hair remain stable over a long period, it has been suggested that hair testing may be useful for the biological monitoring of nicotine exposure in the medium and long term [10, 11, 12, 13].

The measurement of nicotine and cotinine concentrations in hair samples is usually conducted using gas chromatography with mass spectrophotometry (GC/MS) or high-performance liquid chromatography with mass spectrophotometry (HPLC/MS). However, because of their expense, these methods are performed only in special laboratories [14, 15, 16]. In light of this, we previously established a method of high-performance liquid chromatography with ultra violet detection (HPLC/UV) with column-switching to measure nicotine and cotinine in hair [12]. Our results showed that this method was advantageous in its simplicity and low cost whilst maintaining a sensitivity similar to that of GC/MS and HPLC/MS. We also detected an average of 1.60 ng/mg of nicotine in hair samples of non-smokers and therefore concluded this method was a useful tool to measure the effect of secondhand smoke on non-smokers. However, the cut-off value for distinguishing active from passive smokers is yet to be determined – such classification would be beneficial in assessing the smoking status of those in smoking cessation programs.

Therefore, this study aimed to assess smoking status by measuring nicotine and cotinine in the hair of Japanese men selected from the general population, using HPLC/UV with column-switching. We also aimed to determine the nicotine cut-off value to distinguish active from passive smokers using a receiver operating characteristics (ROC) analysis, as this test could be used as a screening tool.

2. Materials and methods

2.1 Subjects

A total of 2,000 Japanese men were chosen from government Basic Resident Registries using cluster random sampling; the sample size calculation was based on the findings of a previous pilot study. During the period 2009–2011, a questionnaire about smoking status was administered to patients at their homes, and hair samples were taken from subjects during home visits. Questionnaires were anonymous to protect private information, and informed consent was obtained from each subject. The study was approved by the appropriate ethics committee. Completed questionnaires were collected from 1,355 participants (response rate 68%), and hair samples were collected from 210 participants (collection rate, 15.5%). Eighteen samples were not used in the analysis because of gray hair, dyed hair, or an insufficient quantity of hair, leaving 192 samples for analysis. Active smokers were identified as subjects who answered “Yes” or “Sometimes” to the question “Do you smoke tobacco presently?”. Those who answered “No” to this question or the question "Have you ever smoked tobacco?" were identified as passive smokers.

2.2 Questionnaire

The questionnaire included items on age, height, weight, education, occupation, and annual income. Items relating to health, including past history of cardiovascular disease, knowledge of smoking-related risks, and alcohol consumption history, were also collected. A history of cardiovascular disease was defined as having been diagnosed with hypertension, angina, or another heart disease in the past year. Knowledge of smoking-related risks was assessed by the question “Do you think that smoking causes serious disease?” with response options of “Yes”, “No”, or “Unknown”. Alcohol consumption was categorized into six groups: everyday, several times a week, several times a month, several times a year, no drinking for more than a year, and never had a drink. Because of the relationship between smoking and obesity [17], we also calculated body mass index (BMI) from height and weight (kg/m ${}^{2}$ ). Education was determined by total education (in years) since elementary school. Occupation was categorized into five groups: office work, specialist work (e.g., doctor, teacher, hairdresser), technical work (e.g., construction, agriculture), unemployed (including student), and other. Annual income was categorized into five groups: $<$ 3 million yen (US $30,000), 3–6 million yen (US $30,000–60,000), $>$ 6 million yen (US $60,000), no income, and unknown. Items for smokers included duration of smoking, number of cigarettes smoked per day, and nicotine dependence. The Brinkman Index (BI) was calculated as the number of years of smoking multiplied by the number of cigarettes smoked per day in order to estimate the cumulative smoking dose. Smokers were then classified into two groups: BI $<$ 400 and BI $\geqslant$ 400. A score $\geqslant$ 400 indicated heavy smoking, as used by previous studies [18]. The Tobacco Dependence Screener (TDS) score was used to examine nicotine dependence [19], with those scoring $\geqslant$ 5 meeting the criteria for dependence.

2.3 Hair sampling

We previously developed a hair-cutting kit to collect hair samples by affixing hair to the back of the head with tape (15 cm $\times$ 5 cm) [12] and then cutting at 1 cm from the end of the hair with scissors. Hair was used from the cut point to a length of 5 cm. Each hair sample taken with the hair-cutting kit was placed in a plastic bag and stored at $-$ 80 ${}^{\circ}$ C in a freezer to prevent potential hydrolysis and oxidation from exposure to air.

Figure 1.

Stages of hair preparation for testing.

2.4 Preparation of hair samples

Table 1
Characteristics of active and passive smokers

Variable Active smokers Passive smokers $p$ -value

( $n=$ 69) ( $n=$ 123)

Nicotine (ng/mg) 26.6 $\pm$ 24.7 3.6 $\pm$ 8.4 $<$ 0.01 ${}^{\rm a}$

Cotinine (ng/mg) 1.9 $\pm$ 2.1 0.2 $\pm$ 0.6 $<$ 0.01 ${}^{\rm a}$

Nicotine $+$ Cotinine (ng/mg) 28.4 $\pm$ 25.8 3.8 $\pm$ 8.8 $<$ 0.01 ${}^{\rm a}$

Age (years) 52.0 $\pm$ 15.5 56.7 $\pm$ 16.6 0.05 ${}^{\rm a}$

BMI (kg/m ${}^{2}$ ) 23.4 $\pm$ 3.1 23.0 $\pm$ 3.5 0.37 ${}^{\rm a}$

Education (years) 12.7 $\pm$ 2.7 12.4 $\pm$ 2.7 0.51 ${}^{\rm a}$

Occupation

Office work 15 (21.7) 32 (26.0) 0.65 ${}^{\rm b}$

Specialist work (e.g., doctor, teacher, hairdresser) 8 (11.6) 19 (15.4)

Technical work (e.g., construction, agriculture) 15 (21.7) 27 (22.0)

Unemployed (including students) 29 (42.0) 44 (35.8)

Other 2 (2.9) 1 (0.8)

Annual income

Under US $30,000 (3 million yen) 20 (29.0) 52 (42.3) 0.25 ${}^{\rm b}$

US $30,000–60,000 (3–6 million yen) 21 (30.4) 35 (28.5)

Over US $60,000 (6 million yen) 8 (11.6) 9 (7.3)

No income 10 (14.5) 9 (7.3)

Unknown 10 (14.5) 18 (14.6)

History of cardiovascular disease

Yes 7 (10.1) 25 (20.3) 0.07 ${}^{\rm b}$

No 62 (89.9) 98 (79.7)

Knowledge of smoking-related risks for smokers

Yes 48 (69.6) 103 (86.2) 0.01 ${}^{\rm b}$

No 11 (15.9) 6 (4.9)

Unknown 10 (14.5) 11 (8.9)

Alcohol consumption

Every day 26 (37.7) 32 (26.0) 0.45 ${}^{\rm b}$

Several times a week 19 (27.5) 40 (32.5)

Several times a month 7 (10.1) 21 (17.1)

Several times a year 5 (7.2) 13 (10.6)

No drinking in over a year 7 (10.1) 10 (8.10)

Never had a drink 5 (7.2) 7 (5.7)

Duration of smoking (years) 32.8 $\pm$ 15.4

Number of cigarettes per day, ( $n=$ 65) 17.3 $\pm$ 7.7

BI ( $n=$ 65)

$<$ 400, n (%) 23 (35.4)

$\geqslant$ 400,n (%) 42 (64.6)

TDS score (nicotine dependence)

$<$ 5, n (%) 29 (42.0)

$\geqslant$ 5, n (%) 40 (58.0)

Variable	Active smokers	Passive smokers	$p$ -value
	( $n=$ 69)	( $n=$ 123)
Nicotine (ng/mg)	26.6 $\pm$ 24.7	3.6 $\pm$ 8.4	$<$ 0.01 ${}^{\rm a}$
Cotinine (ng/mg)	1.9 $\pm$ 2.1	0.2 $\pm$ 0.6	$<$ 0.01 ${}^{\rm a}$
Nicotine $+$ Cotinine (ng/mg)	28.4 $\pm$ 25.8	3.8 $\pm$ 8.8	$<$ 0.01 ${}^{\rm a}$
Age (years)	52.0 $\pm$ 15.5	56.7 $\pm$ 16.6	0.05 ${}^{\rm a}$
BMI (kg/m ${}^{2}$ )	23.4 $\pm$ 3.1	23.0 $\pm$ 3.5	0.37 ${}^{\rm a}$
Education (years)	12.7 $\pm$ 2.7	12.4 $\pm$ 2.7	0.51 ${}^{\rm a}$
Occupation
Office work	15 (21.7)	32 (26.0)	0.65 ${}^{\rm b}$
Specialist work (e.g., doctor, teacher, hairdresser)	8 (11.6)	19 (15.4)
Technical work (e.g., construction, agriculture)	15 (21.7)	27 (22.0)
Unemployed (including students)	29 (42.0)	44 (35.8)
Other	2 (2.9)	1 (0.8)
Annual income
Under US $30,000 (3 million yen)	20 (29.0)	52 (42.3)	0.25 ${}^{\rm b}$
US $30,000–60,000 (3–6 million yen)	21 (30.4)	35 (28.5)
Over US $60,000 (6 million yen)	8 (11.6)	9 (7.3)
No income	10 (14.5)	9 (7.3)
Unknown	10 (14.5)	18 (14.6)
History of cardiovascular disease
Yes	7 (10.1)	25 (20.3)	0.07 ${}^{\rm b}$
No	62 (89.9)	98 (79.7)
Knowledge of smoking-related risks for smokers
Yes	48 (69.6)	103 (86.2)	0.01 ${}^{\rm b}$
No	11 (15.9)	6 (4.9)
Unknown	10 (14.5)	11 (8.9)
Alcohol consumption
Every day	26 (37.7)	32 (26.0)	0.45 ${}^{\rm b}$
Several times a week	19 (27.5)	40 (32.5)
Several times a month	7 (10.1)	21 (17.1)
Several times a year	5 (7.2)	13 (10.6)
No drinking in over a year	7 (10.1)	10 (8.10)
Never had a drink	5 (7.2)	7 (5.7)
Duration of smoking (years)	32.8 $\pm$ 15.4
Number of cigarettes per day, ( $n=$ 65)	17.3 $\pm$ 7.7
BI ( $n=$ 65)
$<$ 400, n (%)	23 (35.4)
$\geqslant$ 400,n (%)	42 (64.6)
TDS score (nicotine dependence)
$<$ 5, n (%)	29 (42.0)
$\geqslant$ 5, n (%)	40 (58.0)

Summary measures are given as mean $\pm$ standard deviation for continuous variables and number (percentage) for categorical variables. BMI, body mass index; BI, Brinkman Index; TDS, Tobacco Dependence Screener. ${}^{\rm a}$ $t$ -test result; ${}^{\rm b}$ chi-square result.

Figure 1 shows the stages involved in the hair preparation for testing. Hair samples were placed in test tubes and washed three times using 3 mL of dichloromethane (133-02441, Wako Pure Chemical Industries, Ltd., Osaka, Japan). Hair samples (25 $\sim$ 50 mg) were weighed after being dried with nitrogen stream. According to internal standards, samples were mixed with 1.6 mL of sodium hydroxide (NaOH) (197-02125, Wako Pure Chemical Industries, Ltd., Osaka, Japan) 2.5 M solution and 60 $\mu$ L of N-ethyl norcotinine (1000 ng/mL; Cosmo Bio, Tokyo, Japan) and then incubated at 40 ${}^{\circ}$ C until the hair was completely dissolved. Next, a solvent mixture (4 mL chloroform:isopropyl alcohol 95:5 v/v, 038-02601:166-04836, Wako Pure Chemical Industries, Ltd., Osaka, Japan) was added and the mixture was vortexed for 2 min. The mixture was then centrifuged for 5 min at 2000 rpm and the supernatant was aspirated under a fume hood. Next, 2 mL hydrogen chloride (HCl) (080-01066, Wako Pure Chemical Industries, Ltd., Osaka, Japan) 0.5 M was added to this solution and was vortexed for 2 min. The mixture was centrifuged for 5 min at 2000 rpm, and the supernatant was transferred to another test tube. Afterwards, 0.4 mL NaOH (2.5 M) was added to the test tube followed by 1.6 mL of ammonium chloride (017-02995, Wako Pure Chemical Industries, Ltd., Osaka, Japan [pH 9.5]) and 4 mL of the solvent mixture (chloroform:isopropyl alcohol 95:5 v/v). The sample was then vortexed for 2 min. Next, the mixture was centrifuged for 5 min at 2000 rpm; the supernatant was discarded and the extract was dried under a nitrogen stream. The extract was dissolved with 600 $\mu$ L of ammonium formate (0.5 M [014–03125, Wako Pure Chemical Industries, Ltd., Osaka, Japan]), centrifuged for 1 min at 2000 rpm, and filtered with a 0.45 $\mu$ m filter. Solvent (200 $\mu$ L) was then injected into the HPLC system (JASCO, Tokyo, Japan) and analyzed.

2.5 HPLC/UV with column-switching system

Samples were analyzed for nicotine and cotinine using HPLC/UV with the column-switching method [12]. We used the HPLC LC-2000 Plus Series, the UV-2075 detector (set at 260 nm), and the HV-2080-01 column selection unit (all from JASCO, Tokyo, Japan). The concentrating column was introduced using the Develosil ODS-UG-5 Column (1 cm $\times$ 4.0 mm i.d., 5.0 $\mu$ m particles; Nomura Chemical, Aichi, Japan) and the DP-model 203 pump (Eicom, Kyoto, Japan). The mobile phase of the concentrated column consisted of ammonium formate (50 mM, pH 9.0) with a flow rate of 0.5 mL/min. The analytical column was introduced using the Ascentis Express C18 Column (10 cm $\times$ 3.0 mm i.d., 2.7 $\mu$ m particles; Sigma-Aldrich, Tokyo, Japan) with the PU-2089 pump (JASCO). The mobile phase in the analytical column consisted of ammonium formate (50 mM, pH 4.3): acetonitrile (96:4) at a flow rate of 0.4 mL/min.

2.6 Statistical analysis

Differences in the mean concentrations of nicotine, cotinine, and nicotine plus cotinine as well as age, BMI, and education among active and passive smokers were assessed using $t$ -tests after confirming that these data were normally distributed. Variations in proportions among groups for occupation, annual income, past history of cardiovascular disease, knowledge of smoking-related risks, and alcohol consumption history were assessed using chi-square tests. In addition, the area under the ROC curve (AUC) was calculated based on sensitivities and specificities to distinguish active smokers from passive smokers. All statistical analyses were conducted using SPSS software, version 17.0 (Nihon IBM, Tokyo, Japan). All probability values were 2-tailed with 95% confidence intervals (CI).

Figure 2.

ROC curve for nicotine and nicotine plus cotinine in hair as predictors of smoking. Nicotine: AUC $=$ 0.92 (95% CI, 0.88–0.96), cut-off value $=$ 5.68 ng/mg. Nicotine plus Cotinine: AUC $=$ 0.93 (95% CI, 0.89–0.97), cut-off value $=$ 6.39 ng/mg.

3. Results

3.1 Characteristics of active and passive smokers among the general population of Japanese men

Among the 192 subjects, there were 69 (36%) active smokers and 123 (64%) passive smokers. The characteristics of the participants are shown in Table 1. The mean $\pm$ standard deviation (SD) age of active smokers was 52.0 $\pm$ 15.5 years, younger than passive smokers (56.7 $\pm$ 16.6 years, $p=$ 0.05). Passive smokers were significantly more likely to be report knowledge of smoking-related risks compared to active smokers. However, there were no significant differences between the two groups in relation to BMI, years of education, occupation, income, history of cardiovascular disease, or alcohol consumption. The mean duration of smoking among the active smokers was 32.8 $\pm$ 15.4 years and they smoked on average 17 $\pm$ 7.7 cigarettes per day.

The concentrations of nicotine, cotinine, and the sum of nicotine plus cotinine were significantly higher among active smokers than among passive smokers. The means $\pm$ SDs of nicotine, cotinine, and sum of nicotine plus cotinine in hair among active smokers were 26.6 $\pm$ 24.7 ng/mg, 1.9 $\pm$ 2.1 ng/mg, and 28.4 $\pm$ 25.8 ng/mg, respectively, and among passive smokers were 3.6 $\pm$ 8.4 ng/mg, 0.2 $\pm$ 0.6 ng/mg, and 3.8 $\pm$ 8.8 ng/mg, respectively. Among active smokers, 42 (64.6%) subjects had a BI score $\geqslant$ 400, and 40 (58.0%) had a Tobacco Dependence Screener (TDS) score $\geqslant$ 5, indicating nicotine dependence. There were no differences in questionnaire responses between those who answered the questionnaires but did not provide hair samples ( $n=$ 1355) and those who answered the questionnaires and did provide hair samples ( $n=$ 192).

3.2 Nicotine and nicotine plus cotinine in hair as predictors of smoking

The AUCs for nicotine and the sum of nicotine plus cotinine were 0.92 (95% CI, 0.88–0.96) and 0.93 (95% CI, 0.89–0.97), respectively (Fig. 2). The cut-off values for nicotine and the sum of nicotine plus cotinine were 5.68 ng/mg (sensitivity, 94.2%; specificity, 87.0%) and 6.39 ng/mg (sensitivity, 92.8%; specificity, 87.0%), respectively. In addition, these cut-off values were found to be optimal for distinguishing active from passive smokers.

4. Discussion

Our results show that nicotine and cotinine are clearly detected in the hair of Japanese men who actively smoke. This is consistent with previous studies showing that nicotine from smoking accumulates in the hair [20]. We found that active smokers were less likely than passive smokers to report knowledge of smoking-related risks. This is in contrast to earlier work in Japan where more smokers than non-smokers understood the hazards of smoking [21, 22]. Our results may be indicative of an increased awareness of smoking risks from recent public health campaigns in Japan, or it may be that smokers deny the harmful effects of smoking and are under-reporting their true knowledge of smoking risks.

This study identified the appropriate cut-off value for separating active from passive smokers among adult Japanese men. Such a cut-off value may be useful for establishing adherence to early smoking cessation interventions, and, hence, may improve the effectiveness of such interventions. Our findings also support the notion that nicotine in hair can effectively be used to screen for long-term exposure to smoking via a simple and inexpensive test.

Previous studies have demonstrated that cut-off values vary substantially with different biological samples and may reflect acute exposure to smoking. For example, cut-off values of cotinine to distinguish active from passive smokers range from 13.7 ng/mL in blood [23, 24, 25] to 10.0 ng/ML in plasma [23] and to 3.08 ng/mL in serum [24] samples. Optimal cut-off values for cotinine in urine and saliva also vary widely between different populations [23, 26, 27, 28, 29]. Measurement of exhaled carbon monoxide is a useful technique for monitoring smoking; it can determine whether an individual has smoked in the previous 8 hours [30, 31]. However, it is not suitable as a tool for long-term smoking evaluation because measurement of carbon monoxide cannot evaluate smokers who have not smoked in the preceding 8 hours. Thus, our findings demonstrate that hair is more useful than other biological samples for measuring exposure to tobacco xenobiotics.

Furthermore, although the liver metabolizes 80–90% of nicotine to cotinine [32], there is a great deal of individual variation. Measurement of nicotine in blood, urine, and saliva samples is prone to inaccuracy due to the short half-life of nicotine; therefore, measuring cotinine is more reliable as it has a longer half-life (averaging 16–19 hours) [33, 34, 35]. Moreover, only approximately 10% of nicotine is excreted in the urine [32]. However, our results indicate that nicotine is stable in hair. This is most likely because nicotine has a high affinity for melanin in hair. In addition, the amount of nicotine in hair accurately reflects the amount of nicotine absorbed from the lungs [36]. Thus, the total concentration of nicotine and cotinine in hair can be used to determine smoking status. However, it seems that only nicotine is useful as an index for smokers because cotinine loses its affinity for melanin and seldom accumulates in hair. In addition, findings from the ROC analysis show that the sensitivity and specificity of nicotine plus cotinine for distinguishing active from passive smokers was similar to that of nicotine alone, suggesting that nicotine in hair can serve as an index variable. Furthermore, some reports show that nicotine concentration in hair stabilizes over the long term, thus reflecting smoking exposure concentration [10, 37, 38, 39, 40].

To the best of our knowledge, no previous study has examined the cut-off value of nicotine and/or cotinine in hair using an ROC analysis. The AUC and the cut-off value of nicotine in hair were shown to be 0.92 ng/mg and 5.68 ng/mg, respectively. The ROC curve for nicotine in hair was convex at the upper left, corresponding to a sensitivity of 94.2% and specificity of 87.0%. Thus, nicotine concentration in hair may be more suitable for distinguishing active from passive smokers compared with total nicotine plus cotinine concentration in hair.

Research on smoking status among patients who visit health care facilities may be prone to selection bias [41]; however, our subjects were community-dwelling residents. The results of the present study indicate that nicotine levels in hair may be a practical way to determine smoking status in the general population. The results of a previous study indicated that taking action against smoking at the local level is important [42]. Our methods and cut-off value could be used to screen for long-term exposure among smokers at the local level. Our results also show that it is possible to intervene early to encourage smoking cessation and to prevent additional effects related to passive smoking if active and passive smokers are clearly distinguished in the general population. The classification of active and passive smoking can be carried out directly through hair samples, thus avoiding any misclassification of smoking habits as a source of potential bias.

This study has a number of limitations that need to be considered. First, our sample was restricted to men and therefore we cannot generalize these findings to women. Previous research has shown a sex difference in the susceptibility of developing lung disease following exposure to nicotine [43, 44]. In addition, women have been shown to metabolize the chemical substances contained in cigarettes differently to men [45]. Second, as the gene CYP2A6 polymorphism, that helps to metabolize nicotine to cotinine, reportedly differs between Japanese and non-Japanese individuals [46, 47, 48], the cut-off value may differ in different populations.

In conclusion, we were able to distinguish active smokers from passive smokers in the general population by measuring the hair concentration of nicotine using the HPLC/UV with column-switching method. Our results suggest that nicotine in hair is a more valuable index for smoking evaluation than other biological measures. The optimal cut-off value had both high sensitivity and specificity, suggesting that nicotine in hair can be used as an effective and noninvasive screening tool to assess current smoking status and monitor adherence to smoking cessation programs.

References

McGinnis

J.M.

and Foege

W.H.

, Actual causes of death in the United States, JAMA 270 (1993), 2207–2212.

West

, Smoking: its influence on survival and cause of death, J R Coll Physicians Lond 26 (1992), 357–366.

Kawaminami

Minowa

Okayama

Hayakawa

and Ueshima

, An association (population attributable fraction) between smoking habit and mortality from all causes, cancer and lung cancer: NIPPON DATA80, 1980–1999. National Integrated Projects for Prospective Observation of Non-Communicable Diseases and its Trend in the Aged, Nihon Eiseigaku Zasshi 57 (2003), 669–673 (in Japanese).

Nakamura

Nakagawa

Sakurai

Murakami

Irie

Fujiyoshi

Okamura

Miura

Ueshima

and EPOCH-JAPAN Research Group, Influence of smoking combined with another risk factor on the risk of mortality from coronary heart disease and stroke: pooled analysis of 10 Japanese cohort studies, Cerebrovasc Dis 33 (2012), 480–491.

Sandler

D.P.

Wilcox

A.J.

and Everson

R.B.

, Cumulative effects of lifetime passive smoking on cancer risk, Lancet 1 (1985), 312–315.

Mannino

Siegel

Huston

Rose

and Etzel

, Environmental tobacco smoke exposure and health effects in children: results from the 1991 National Health Survey, Tobacco Control 5 (1996), 13–18.

Heinrich

Hölscher

Seiwert

Carty

C.L.

Merkel

and Schulz

, Nicotine and cotinine in adults’ urine: The German Environmental Survey 1998, J Expo Anal Environ Epidemiol 15 (2005), 74–80.

Higashi

Sashikuma

Itani

and Muranaka

, Simultaneous determination of nicotine and cotinine in urine by gas liquid chromatography, Eisei Kagaku 32 (1986), 276–280.

Yamamoto

Shibata

Mugikura

Igarashi

and Satoh

, Usefulness of urinary cotinine measurement in determining individual smoking status, J Transportation Med54 (2000), 142–146.

10.

Al-Delaimy

, Hair as a biomarker for exposure to tobacco smoke, Tob Control 11 (2002), 176–182.

11.

Klein

Blanchette

and Koren

, Assessing nicotine metabolism in pregnancy – a novel approach using hair analysis, Forensic Sci Int 145 (2004), 191–194.

12.

Tsuji

Mori

Kanda

Ito

Hidaka

Kakamu

Kumagai

Hayakawa

Osaki

and Fukushima

, Development of simple HPLC/UV with a column-switching method for the determination of nicotine and cotinine in hair samples, Health 5 (2013), 687–694.

13.

Tzatzarakis

Vardavas

Terzi

Kavalakis

Kokkinakis

Liesivuori

and Tsatsakis

, Hair nicotine/cotinine concentrations as a method of monitoring exposure to tobacco smoke among infants and adults, Hum Exp Toxico 31 (2012), 258–265.

14.

Man

C.N.

Ismail

Harn

G.L.

Lajis

and Awang

, Determination of hair nicotine by gas chromatography–mass spectrometry, J Chromatogr B Analyt Technol Biomed Life Sci 877 (2009), 339–342.

15.

Marchei

Durgbanshi

Rossi

Garcia-Algar

Ó.

Zuccaro

and Pichini

, Determination of arecoline (areca nut alkaloid) and nicotine in hair by high-performance liquid chromatography/electrospray quadrupole mass spectrometry, Rapid Commun Mass Spectrom 19 (2005), 3416–3418.

16.

Pichini

Altieri

Pellegrini

Pacifici

and Zuccaro

, The analysis of nicotine in infants’ hair for measuring exposure to environmental tobacco smoke, Forensic Sci Int 84 (1997), 253–258.

17.

Eisen

S.A.

Lyons

M.J.

Goldberg

and True

W.R.

, The impact of cigarette and alcohol consumption on weight and obesity: An analysis of 1911 monozygotic male twin pairs, Arch Int Med 153 (1993), 2457–2463.

18.

Shitara

, K Matsuo Hatooka

Ura

Takahari

Yokota

Abe

Kawai

Tajika

Kodaira

Shinoda

Tajima

Muro

and H Tanaka, Heavy smoking history interacts with chemoradiotherapy for esophageal cancer prognosis: A retrospective study, Cancer Sci 101 (2010), 1001–1006.

19.

Kawakami

Takatsuka

Inaba

and Shimizu

, Development of a screening questionnaire for tobacco/nicotine dependence according to ICD-10, DSM-III-R, and DSM-IV, Addict Behav 24 (1999), 155–166.

20.

Chetiyanukornkul

Toriba

Kizu

Kimura

and Hayakawa

, Hair analysis of nicotine and cotinine for evaluating tobacco smoke exposure by liquid chromatography-mass spectrometry, Biomed Chromatogr 18 (2004), 655–661.

21.

Morimoto

Miyamatsu

Okamura

Nakayama

Morinaga

Toyota

Suzuki

Hata

and Yamaguchi

, [Drinking/smoking habits and knowledge regarding heavy drinking/ smoking as a risk factor of stroke among Japanese general population], Nihon Arukoru Yakubutsu Igakkai Zasshi 45(5) (2010), 411–419 (in Japanese).

22.

Sumida

Katsura

Hoshino

and Usui

, Factors relating to smoking of male university students – comparison between smokers and non-smokers, Health Science: Annual reports of School of Health Sciences Faculty of Medicine, Kyoto University 7 (2011), 37–42 (in Japanese).

23.

Jarvis

M.J.

Tunstall-Pedoe

Feyerabend

Vesey

and Saloojee

, Comparison of tests used to distinguish smokers from nonsmokers, Am J Public Health 77 (1987), 1435–1438.

24.

Beineke

Fitch

Tao

Elashoff

M.R.

Rosenberg

Kraus

W.E.

and Wingrove

J.A.

, A whole blood gene expression-based signature for smoking status, BMC Med Genomics 5 (2012), 1.

25.

Benowitz

N.L.

Bernert

J.T.

Caraballo

R.S.

Holiday

D.B.

and Wang

, Optimal serum cotinine levels for distinguishing cigarette smokers and nonsmokers within different racial/ethnic groups in the United States between 1999 and 2004, Am J Epidemiol 169 (2009), 236–248.

26.

Etter

J.F.

Due

T.V.

and Perneger

T.V.

, Saliva cotinine levels in smokers and nonsmokers, Am J Epidemiol 151 (2000), 251–258.

27.

Hegaard

H.K.

Kjærgaard

Møller

L.F.

Wachmann

and Ottesen

, Determination of a saliva cotinine cut-off to distinguish pregnant smokers from pregnant non-smokers, Acta Obstet Gynecol Scand 86 (2007), 401–406.

28.

Okayama

and Sato

, Assessment of smoking status among workers using an improved colorimetric method, Ind Health 42 (2004), 348–351.

29.

Zielińska-Danch

Wardas

Sobczak

and Szołtysek-Bołdys

, Estimation of urinary cotinine cut-off points distinguishing non-smokers, passive and active smokers, Biomarkers 12 (2007), 484–496.

30.

Ripoll

Girauta

Ramos

Medina-Bombardó

Pastor

Alvarez-Ossorio

Gorreto

Esteva

García

and Uréndez

, Clinical trial on the efficacy of exhaled carbon monoxide measurement in smoking cessation in primary health care, BMC Public Health 12 (2012), 322.

31.

Sandberg

Sköld

C.M.

Grunewald

Eklund

and Wheelock

Å.M.

, Assessing recent smoking status by measuring exhaled carbon monoxide levels, PLoS One 6 (2011), e28864.

32.

Nakajima

, Nicotine metabolism and genetic polymorphisms of CYP2A6 gene in humans, Japanese Journal of Electrocardiology 26 (2006), 217–222 (in Japanese).

33.

Benowitz

N.L.

, Cotinine as a biomarker of environmental tobacco smoke exposure, Epidemiol Rev 18 (1996), 188–204.

34.

Jarvis

M.J.

Russell

Benowitz

N.L.

and Feyerabend

, Elimination of cotinine from body fluids: implications for noninvasive measurement of tobacco smoke exposure, Am J Public Health 78 (1988), 696–698.

35.

Scherer

Jarczyk

Heller

Biber

Neurath

and Adlkofer

, Pharmacokinetics of nicotine, cotinine, and 3’-hydroxycotinine in cigarette smokers, Klin Wochenschr 66 (1988), 5–11.

36.

Niwa

and Uematsu

, Management of the smoking behavior by nicotine analysis, J Therapy 79 (1997), 1503–1506 (in Japanese).

37.

Kim

S.R.

Wipfli

Avila-Tang

Samet

J.M.

and Breysse

P.N.

, Method validation for measurement of hair nicotine level in nonsmokers, Biomed Chromatogr 23 (2009), 273–279.

38.

Mizuno

Uematsu

Ishikawa

Yoshimine

and Nakashima

, Clinical outcome of smoking-cessation trial of nicotine chewing gum evaluated by analysis of nicotine in hair, Ther Drug Monit 19 (1997), 407–412.

39.

Uematsu

, Utilization of hair analysis for therapeutic drug monitoring with a special reference to ofloxacin and to nicotine, For Sci Int 63 (1993), 261–268.

40.

Uematsu

Mizuno

Nagashima

Oshima

and Nakamura

, The axial distribution of nicotine content along hair shaft as an indicator of changes in smoking behaviour: evaluation in a smoking-cessation programme with or without the aid of nicotine chewing gum, Br J Clin Pharmacol 39 (1995), 665–669.

41.

Anda

R.F.

Croft

J.B.

Felitti

V.J.

Nordenberg

Giles

W.H.

Williamson

D.F.

and Giovino

G.A.

, Adverse childhood experiences and smoking during adolescence and adulthood, JAMA 282 (1999), 1652–1658.

42.

Tsuji

Tsunoda

Suzuki

Ueno

and Aizawa

, A survey on smoking among primary and junior high school students: for the prevention of smoking in one town, Memoirs of Mejiro University College 44 (2008), 85–96 (in Japanese).

43.

de Torres

J.P.

Cote

C.G.

López

M.V.

Casanova

Díaz

Marin

J.M.

Pinto-Plata

de Oca

M.M.

Nekach

Dordelly

L.I.

Aguirre-Jaime

and Celli

B.R.

, Sex differences in mortality in patients with COPD, Eur Respir J 33 (2009), 528–535.

44.

International Early Lung Cancer Action Program Investigators Henschke

C.I.

Yip

, and Miettinen

O.S.

, Women’s susceptibility to tobacco carcinogens and survival after diagnosis of lung cancer, JAMA 296 (2006), 180–184.

45.

Ben-Zaken Cohen

Paré

P.D.

Man

S.F.

and Sin

D.D.

, The growing burden of chronic obstructive pulmonary disease and lung cancer in women: examining sex differences in cigarette smoke metabolism, Am J Respir Crit Care Med 176 (2007), 113–120.

46.

Ito

Tsuji

Mori

Kanda

Hidaka

Kakamu

Kumagai

Hayakawa

Osaki

and Fukushima

, Effect of CYP2A6* 4 genetic polymorphisms on smoking behaviors and nicotine dependence in a general population of Japanese men, Fukushima J Med Sci 61 (2015), 125–130.

47.

Minematsu

Nakamura

Iwata

Tateno

Nakajima

Takahashi

Fujishima

and Yamaguchi

, Association of CYP2A6 deletion polymorphism with smoking habit and development of pulmonary emphysema, Thorax 58 (2003), 623–628.

48.

Rao

Hoffmann

Zia

Bodin

Zeman

Sellers

E.M.

and Tyndale

R.F.

, Duplications and defects in the CYP2A6 gene: identification, genotyping, and in vivo effects on smoking, Mol Pharmacol 58 (2000), 747–755.