Many people use both combustible cigarette (CC) and electronic cigarette (EC). We compared nicotine dependence among CC, EC, and dual users using questionnaires and urinary cotinine levels.
Data from the Korea National Health and Nutrition Examination Survey (2014–2017) databases were analyzed; 3,917 CC, EC, and dual users were administered the urinary cotinine test, and 1,045 current CC and dual users completed the Heaviness of Smoking Index (HSI) questionnaires. Weighted geometric means of urine cotinine levels were compared between exclusive CC, exclusive EC, and dual users. The distribution rate, based on time to first cigarette (TTFC), cigarettes per day (CPD), and the HSI was analyzed in two weighted groups, exclusive CC and dual users.
Among those who currently use any type of cigarette, 89.4%, 1.4%, and 9.2% were exclusive CC, exclusive EC, and dual users, respectively. Weighted geometric means of urine cotinine were highest in dual users (1,356.4 ng/mL), followed by exclusive CC (1,270.3 ng/mL), and exclusive EC (867.7 ng/mL) with significant differences between all three groups (P<0.05). There were no statistically significant differences in CPD between exclusive CC and dual users (P=0.626). The proportion of TTFC ≤5 minutes was 21.5% and 29.5% in the two groups, respectively (P=0.010); however, HSI differences in the two groups was marginal (P=0.557).
In this study, the urinary cotinine value could distinguish the three groups, CC, EC, and dual users, but the questionnaire using HSI could not distinguish the three groups.
Since their development in China in 2003, battery-powered electronic cigarette (EC) devices that evaporate liquid nicotine solutions have been popular in many countries, and the usage rate varies from country to country. For example, the rate of daily adult EC use is 4.2% in the United Kingdom and 2.8% in South Korea with less restrictive use policies, while it is 2.1% in Canada and 0.4% in Brazil with more restrictive policies [
According to previous cross-sectional studies [
The main reason for smoking cessation failure is associated with the level of nicotine dependence, which also applies to EC users. Nicotine dependence is mainly evaluated using questionnaires and urinary cotinine levels. In previous studies that evaluated the nicotine dependence level of dual users, the difference between CC users and dual users was controversial [
Appropriate evaluation of nicotine dependence is an important part of smoking cessation consultations, the Heaviness of Smoking Index (HSI) consists of two of the Fagerström Test for Nicotine Dependence (FTND) items, cigarettes per day (CPD), and time to first cigarette (TTFC). According to a report by Pérez-Ríos et al. [
Cotinine is a major metabolite of nicotine, with a longer half-life (20 hours) than nicotine (2 hours), and is easily detected in the urine, plasma, and saliva. It is widely used as a biomarker for more accurate assessments of smoking status in research, including epidemiological studies [
This study aims to compare nicotine dependence among exclusive CC, EC, and dual users using the HSI and urine cotinine as a biological marker. The Korea National Health and Nutrition Examination Survey (KNHANES) database, which includes a nationally representative sample, provides the opportunity to assess dependence among the three groups.
This study used raw data from the sixth (2014–2015) and seventh (2016–2017) KNHANES surveys of a representative sample of the Korean population, conducted by the Korea Centers for Disease Control and Prevention since 1998, using a two-stage stratified sampling method [
Among the questionnaire respondents, those who indicated that they used CCs or ECs were included. A total of 1,916 people without smoking history information, and 5,980 people without urine cotinine values were excluded; thus, 23,311 people were analyzed. Current smokers were categorized into three groups: exclusive CC, exclusive electronic cigarette, and dual users. For the analysis of urine cotinine values, 82 people who received nicotine replacement therapy (NRT) within 5 days were excluded. An additional 19,312 never- and former smokers were also excluded. As a result, 3,917 people were included in the urine cotinine analysis with further analyses of smoking cessation attempts and motivation to quit smoking using the transtheoretical model (TTM) [
Respondents who answered, “daily or sometimes” to the question, “Do you currently smoke?” were categorized as CC users. Respondents who answered “yes” to the question, “Have you used electronic cigarettes in the last month?” were categorized as EC users. Respondents who answered “yes” to both questions were categorized as dual users. Given the above classifications, CC and EC users were classified as “current,” “former,” and “never use,” respectively. When the three types of usage in these two groups are combined, respectively, we obtain a total of nine cases, depending on CC and EC usage status. Based on the number of cases above, we classified them into three groups; exclusive CC user, exclusive electronic cigarette user, and dual user.
Urine cotinine analyses were performed in the three groups. Urine cotinine concentration (ng/mL) was measured using gas chromatography-mass spectrometry, using the Clarus 600/600T (Perkin Elmer, Waltham, MA, USA), and the detection limit was less than 0.27399 ng/ mL [
This study excluded exclusive electronic cigarette users from the HSI analysis because the CPD unit measure is based on CC use and does not include a measurement unit for EC use. When asked, “How many cigarettes do you smoke per day?”, the CPD item was divided into four groups: 10 cigarettes or less, 11 to 20 cigarettes, 21 to 30 cigarettes, and more than 30 cigarettes. The groups were scored from 0–3 points, respectively. To evaluate TTFC values, those who did not conduct the TTFC survey were excluded (n=2,820) as the TTFC question was only included in the 2016 survey. The TTFC item was assessed by asking, “How soon after you wake up do you smoke your first cigarette in the morning?” and was also divided into four groups: within 5 minutes, 6 to 30 minutes, 31 minutes to 1 hour, and more than 1 hour. The TTFC was scored from 3–0, respectively. The HSI item is calculated as the sum of CPD (0–3 points) and TTFC (3–0 points) scores and is divided into three categories: mild (0–1 points), medium (2–4 points), and high (5–6 points), according to the criteria set forth by the National Institute on Drug Abuse in the United States [
Additional smoking-related variables, smoking cessation attempts, and motivation to quit smoking were evaluated using the TTM. Smoking cessation attempts were evaluated in response to the question, “Have you stopped smoking for more than a day (24 hours) in an attempt to quit smoking in the past year?” Motivation to quit smoking was evaluated using the question, “Do you have plans to quit smoking in the next month?” Respondents who answered, “I have plans to quit smoking in 1 month” were categorized as “preparation stage” and the rest as “others.”
Age, sex, education level, income level, marital status, and high-risk alcohol use were analyzed. High-risk alcohol use was defined as an average of more than seven glasses for men or more than five glasses for women more than twice a week, according to the KNHANES standards [
Chi-square tests were used to identify the relationship between the type of tobacco product and smoking-related sociodemographic factors such as sex, education, quartile of monthly household income, marital status, and high-risk alcohol drinking. One-way analysis of variance (ANOVA) was used for the average age of each group. Because the urine cotinine values were not normally distributed, the GM was used and an ANOVA test was conducted. The chi-square test was used to analyze the nicotine dependence questionnaires (CPD, TTFC, and HSI) between exclusive CC and dual users. STATA ver. 15.0 (Stata Corp., College Station, TX, USA) was used for statistical analyses, and P-values <0.05 were considered statistically significant.
The respondents’ demographic characteristics are presented in
Regarding participants with education beyond high school, 56.6% were exclusive EC users, 38.8% were exclusive CC users, and 39.1% were dual users. Among dual users, the rate of those in the top 50% income level was 75.3%, which was higher than that of exclusive CC users (61.7%) and exclusive EC users (61.0%). A higher percentage of participants were married in all three groups. The rates among the three groups of high-risk drinking (P=0.131), smoking cessation attempts (P=0.116), and TTM for motivation to quit smoking (P=0.080) were no statistically significant. The weighted GM of urine cotinine was highest in dual users, followed by exclusive CC and exclusive EC users (P<0.001) (
The weighted GM of urine cotinine was highest in dual users (1,356.4 ng/mL), followed by exclusive CC users (1,270.3 ng/mL, standard error=17.9), and exclusive EC users (867.7 ng/mL) (
In the exclusive CC and dual user groups, more than 90% of respondents used less than a pack of CPD. The proportion of the two groups who smoked more than 30 cigarettes was similar, with 1.7% for exclusive CC users and 1.5% for dual users; however, the CPD item score was not significantly different (P=0.626).
Among 29.5% of dual users and 21.5% of exclusive CC users, the TTFC was within five minutes of waking (P=0.010). When the HSI was divided into three categories according to score (mild, medium, and high), approximately 4.3% of CC users and 5.5% of dual users had a high score (
The GM of urine cotinine was highest among dual users, followed by exclusive CC users and exclusive EC users, and the differences between the three groups were statistically significant. The differences in the average CPD and HSI scores were not statistically significant among exclusive CC and dual users. Although the urine cotinine analysis showed differences in nicotine dependence between the three groups, the differences were not clear in the survey questionnaire.
Both the HSI questionnaire and urine cotinine test have validity in assessing nicotine dependence; therefore, the survey results, indicating no difference between the three groups, could be questioned. This may be due to the fact that the KNHANES does not contain a detailed questionnaire on EC use. (e.g., EC use frequency, number of puffs, nicotine concentrations, vaporizing power, flavors, etc.). When we analyzed the correlation coefficient between urine cotinine and CPD, TTFC, and HSI scores (the correlation analysis results of the GM of urine cotinine and CPD, TTFC, and HSI were not provided in the results), we found that the correlation coefficient between urine cotinine and CPD (β=0.4093, P<0.001) was the lowest compared to that of TTFC (β=0.4312, P<0.001) and HSI (β=0.4435, P<0.001). In the HSI questionnaire, the TTFC is likely to be relatively useful in developing nicotine dependence over a short period of time.
The assessment of nicotine dependence in dual users is likely to be affected by the type of cigarette (CC, EC, or dual use) the respondents use at the time of analysis. However, measuring the frequency of use of each cigarette type is important when assessing nicotine dependence. Martinez et al. [
This study has several limitations. First, it is difficult to identify whether dual use causes greater nicotine dependence or vice versa as this was a retrospective, cross-sectional study. Also, since the transition of cigarette use could not be identified, it was not known whether an EC user switched from CCs in the past or started ECs anew. In this study, “Help quit smoking tobacco” was the first reason for the use of ECs analyzed for exclusive EC users (56.3%) and dual users (43.0%). In terms of dual users, 64.7% have tried to quit smoking for more than 24 hours in a year. In future research, longitudinal studies on cigarette use patterns, including transition, are needed to understand the causal effect of dual use on nicotine dependence.
It is meaningful to assess nicotine dependence on each cigarette type using the KNHANES data, considering the growing number of dual users among EC users. Several studies have evaluated EC users’ nicotine dependence; however, nicotine dependence has not been assessed using biochemical markers [
Currently, EC technology is developing gradually, and future devices are likely to deliver high-concentration nicotine more efficiently. Moreover, the number of new types of cigarettes, including electronic cigarettes and heat-not-burn tobacco products, is increasing. Developing new evaluation tools to measure nicotine dependence among dual and poly-users is essential. In a report of Vogel et al. [
In conclusion, this study compared and analyzed nicotine dependence according to urine cotinine values and the HSI questionnaire, using KNHANES data, among exclusive CC, exclusive EC, and dual users. Dual users had significantly higher urine cotinine levels than the others. The differences in nicotine dependence among the three groups were revealed by the urine cotinine analysis; however, the difference was not clear in the survey questionnaire. It may not be possible to distinguish differences in nicotine dependence through the HSI questionnaire alone. However, the TTFC item in the HSI items can be useful in developing nicotine dependence in dual users over a short period of time. The smoking cessation consultations for dual users need to be more strategic. New evaluation tools are needed to assess the amount of smoking and nicotine dependence of dual users or poly-users more efficiently.
No potential conflicts of interest relevant to this article were reported.
Supplementary materials can be found via
Analysis of urine cotinine, cigarette per day, TTFC, and Heaviness of Smoking Index by gender
Flowchart: study population. TTFC, time to first cigarette.
Geometric mean of urine cotinine analysis in current cigarette users. P-value from analysis of variance for urine cotinine geometric mean. *P<0.05.
Characteristics of study populations
Characteristic | Combustible cigarette user | EC user | Dual user | P-value | |
---|---|---|---|---|---|
Sample size (%) | 3,557/7,373,307 (89.4) | 52/115,984 (1.4) | 308/753,899 (9.2) | ||
Age (y) | 43.9±0.3 | 37.6 |
35.3 |
<0.001 | |
<30 | 17.2 (0.9) | 24.8 (7.2) | 38.3 (3.2) | ||
30–39 | 23.2 (1.0) | 43.2 (7.7) | 30.1 (3.1) | ||
40–49 | 25.7 (0.9) | 15.8 (5.2) | 19.4 (2.3) | ||
50–59 | 20.3 (0.8) | 6.9 (3.2) | 8.9 (1.7) | ||
60–69 | 9.1 (0.5) | 6.2 (3.5) | 2.5 (0.7) | ||
≥70 | 4.4 (0.3) | 3.0 (1.6) | 0.7 (0.3) | ||
Sex | 0.372 | ||||
Male | 87.7 (0.6) | 83.9 (5.3) | 89.9 (1.7) | ||
Female | 12.3 (0.6) | 16.1 (5.3) | 10.1 (1.7) | ||
Education | <0.001 | ||||
Less than high school | 18.9 (0.8) | 8.8 (3.3) | 7.0 (1.6) | ||
High school graduate | 42.4 (1.1) | 34.6 (7.0) | 53.9 (3.5) | ||
Beyond high school | 38.8 (1.1) | 56.6 (7.2) | 39.1 (3.3) | ||
Quartile of monthly household income | 0.002 | ||||
Low | 14.8 (0.8) | 10.2 (4.2) | 8.0 (1.7) | ||
Middle-low | 23.6 (0.9) | 28.7 (7.8) | 16.8 (2.3) | ||
Middle-high | 32.5 (1.0) | 38.3 (7.9) | 38.2 (3.4) | ||
Upper | 29.2 (1.0) | 22.7 (6.8) | 37.1 (3.4) | ||
Marital status | <0.001 | ||||
Married | 73.0 (1.0) | 60.0 (7.6) | 51.0 (3.2) | ||
Unmarried | 27.0 (1.0) | 40.0 (7.6) | 49.0 (3.2) | ||
High-risk alcohol drinking |
0.131 | ||||
No | 70.5 (0.9) | 84.7 (5.6) | 70.4 (2.8) | ||
Yes | 29.5 (0.9) | 15.3 (5.6) | 29.6 (2.8) | ||
Smoking cessation attempt |
0.116 | ||||
No | 43.0 (1.3) | 35.3 (4.5) | |||
Yes | 57.0 (1.3) | 64.7 (4.5) | |||
Motivation to quit smoking by TTM |
0.080 | ||||
Preparation stage | 21.4 (0.8) | 26.9 (3.2) | |||
Others | 78.6 (0.8) | 73.1 (3.2) | |||
Urine cotinine (ng/mL) | 1,270.3±17.9 | 867.7 |
1,356.4 |
<0.001 |
Values are presented as unweighted sample size/weighted sample size (%), mean±SE, or % (SE), unless otherwise stated. P-value from chi-square test for categorical variables or analysis of variance for continuous variables.
EC, electronic cigarette; TTM, transtheoretical model; SE, standard error.
SE were missing because of strata with single sampling unit such as mean age and urine cotinine geometric mean in the exclusive EC and dual user group.
Drink more than twice a week and average drinking amount is more than 7 drinks for men, 5 drinks for women.
Question about Smoking cessation attempt; Have you ever quit smoking more than 24 hours in the last year?
Question about smoking cessation plan; preparation stage (quit smoking within 1 month) vs. others (around the next 6 months or don’t want to quit smoking at all).
Comparison of nicotine dependence using Heaviness of Smoking Index between exclusive cigarette smokers and dual users of cigarette and e-cigarette
Variable | Combustible cigarette user | Dual user | P-value |
---|---|---|---|
Sample size (%) | 3,557/7,403,800 (90.67) | 308/761,855 (9.33) | |
Cigarette per day |
0.626 | ||
0 (≤10) | 45.3 (±1.0) | 44.1 (±2.9) | |
1 (11–20) | 47.6 (±1.0) | 47.3 (±3.0) | |
2 (21–30) | 5.3 (±0.4) | 7.1 (±1.6) | |
3 (>30) | 1.7 (±0.3) | 1.5 (±0.7) | |
Sample size (%) | 979/1,995,488 (93.20) | 66/145,593 (6.80) | |
TTFC |
0.010 | ||
0 (>60 min) | 31.0 (±1.9) | 15.6 (±5.0) | |
1 (31–60 min) | 15.3 (±1.5) | 29.3 (±6.7) | |
2 (6–30 min) | 32.1 (±1.9) | 25.7 (±6.0) | |
3 (≤5 min) | 21.5 (±1.8) | 29.5 (±6.0) | |
Heaviness of Smoking Index |
0.557 | ||
Low (0–1) | 37.3 (±2.0) | 30.1 (±6.5) | |
Medium (2–4) | 58.4 (±1.9) | 64.4 (±6.8) | |
High (5–6) | 4.3 (±0.7) | 5.5 (±3.1) |
Values are presented as unweighted sample size/weighted sample size (%) or % (±standard error), unless otherwise stated. P-value from chi-square test for categorical variables.
TTFC, time to first cigarette.
Question about cigarette per day; On average, how many cigarettes do you smoke per day (for current smoker)?
Question about TTFC; How soon do you smoke your first cigarette in the morning (for current smoker)?
Since the TTFC questionnaire was only conducted in 2016, TTFC and Heaviness of Smoking Index analysis is only for respondents to the 2016 TTFC questionnaire out of the 4 years in 2014–2017.
Heaviness of Smoking Index is the sum of cigarette per day (0–3 points) and TTFC (0–3 points), total of 6 points divided into three categories according to the score.