Can Urinary Cotinine Predict Nicotine Dependence Level in Smokers?

Background: Although nicotine dependence plays a role as a main barrier for smoking cessation, there is still a lack of solid evidence on the validity of biomarkers to determine nicotine dependence in clinical settings. This study aimed to investigate whether urinary cotinine levels could reflect the severity of nicotine dependence in active smokers. Materials and Methods: Data regarding general characteristics and smoking status was collected using a self-administered smoking questionnaire. The Fagerström test for nicotine dependence (FTND) was used to determine nicotine dependence of the participants, and a total of 381 participants were classified into 3 groups of nicotine dependence: low (n=205, 53.8%), moderate (n=127, 33.3%), and high dependence groups (n=49, 12.9%). Stepwise multiple linear regression model and receiver operating characteristic (ROC) curves analyses were used to determine the validity of urinary cotinine for high nicotine dependence. Results: In correlation analysis, urinary cotinine levels increased with FTND score (r=0.567, P<0.001). ROC curves analysis showed that urinary cotinine levels predicted the high-dependence group with reasonable accuracy (optimal cut-off value=1,000 ng/mL; AUC=0.82; P<0.001; sensitivity=71.4%; specificity=74.4%). In stepwise multiple regression analysis, the total smoking period (β=0.042, P=0.001) and urinary cotinine levels (β=0.234, P<0.001) were positively associated with nicotine dependence, whereas an inverse association was observed between highest education levels (>16 years) and nicotine dependence (β=-0.573, P=0.034). Conclusions: The results of this study support the validity of using urinary cotinine levels for assessment of nicotine dependence in active smokers.


Introduction
Cigarette smoking can cause various cancers and cardiovascular disease, and smoking cessation can reduce the risk of these diseases (USDHHS, 2004). Many smokers experience withdrawal symptoms while attempting to quit smoking; therefore, nicotine dependence should be considered as one of the substantial barriers to success in smoking cessation (Fagerstrom and Furberg, 2008;Hagimoto et al., 2010). The Fagerström test for nicotine dependence (FTND), a short, 6-question, convenient, self-report measure, has been widely used to assess the degree of nicotine dependence (Heatherton et al., 1991). However, there is insufficient evidence for application of objective measurements, such as urinary cotinine levels, to assess nicotine dependence.
Although nicotine is a desirable biomarker for smoking status, there are limitations to its utility: a short half-life, complex measurement, and variability in plasma levels (which might be higher in subjects with impaired kidney excretion function) (Molander et al., 2000). Cotinine, a metabolite of nicotine, can be measured in saliva, plasma, and urine. Cotinine has relative advantages over other biomarkers-a longer half-life (about 18-20 hours), convenience of measurement, and little daily variability in chronic smokers; therefore, its level has been regarded as an useful objective measure for smoking status or exposure (Jarvis et al., 1987;Lerman et al., 1993). Particularly, measurement of cotinine in the urine is easier and less invasive than of cotinine in the plasma. Furthermore, unlike nicotine, urinary excretion of cotinine is rarely affected by the amount or pH of the 5484 urine (Verification., 2002). Nicotine dependence could refer to the result of long habituation of cigarette smoking. Interestingly, several studies found that levels of cotinine, a metabolite with limited residual period, were associated with some specific smoking behaviors reflecting the severity of nicotine dependence. The amount of time before smoking the first cigarette after waking in the morning, one item of the FTND, was associated with cotinine levels in daily smokers (Heatherton et al., 1991;Muscat et al., 2009;Fu et al., 2012). How quickly to smoke after awakening in the morning has been regarded as a strong marker of nicotine dependence, because highly dependent smokers have lower tolerability to depleted plasma nicotine levels on waking (Heatherton et al., 1991;Baker et al., 2007). Other reports showed a dose-dependent relationship between cotinine levels and the number of cigarettes smoked per day, another measure of the FTND (Law et al., 1997;Blackford et al., 2006;Fu et al., 2012). Moreover, several previous reports showed a significant relationship between the degree of nicotine dependence (as assessed by the total FTND score) and cotinine levels (Heatherton et al., 1991;Payne et al., 1994;Pomerleau et al., 1994). However, as those studies primarily focused on a simple linear correlation, there is a lack of data showing how cotinine levels independently predict severity of nicotine dependence after adjustment of potential confounding factors. Furthermore, the standard cut-off values for cotinine levels as an indicator of high nicotine dependence have not been set yet, whereas there have been mainly several reports on urinary cotinine cut-off points discriminating non-smokers/smokers or passive/ active smokers (Zielinska-Danch et al., 2007).
Therefore, the aim of this study is to investigate if the urinary cotinine levels are independently associated with the degree of nicotine dependence, and to determine the validity of using urinary cotinine levels to predict high nicotine dependence in active smokers.

Study population
We collected health data from smokers who had reported for a check-up at the Center for Cancer Prevention and Detection in the National Cancer Center, Republic of Korea from June 1 to December 31, 2007. Study inclusion criteria were as follows: adult ages ≥18; current smoker with total number of lifetime cigarettes smoked ≥400; total lifetime smoking period ≥6 months; and absence of serious physical or mental illness, such as malignancy, chronic obstructive pulmonary disease, chronic kidney disease, depression, and other psychiatric diseases with a high probability of substance abuse. Smoking status and nicotine dependence were determined using self-administered smoking questionnaires that assessed the number of cigarettes smoked per day, total smoking period, total number of lifetime cigarettes smoked, age at smoking initiation, history of smoking cessation, intent to quit smoking, and the presence of a partner that smoked, as well as the measures in the FTND. Based on the FTND score, nicotine dependence status of all study subjects were classified into 3 grades: low (0-3), moderate (4-6), and high (7-10) dependence. This study was approved by the Institutional Review Board at the National Cancer Center at Goyang, Korea.

Measurement of urinary cotinine
Urine specimens were collected in plastic containers and stored in a refrigerator at 4°C. On the day of collection, cotinine levels were assessed using a Cotinine Enzyme Immunoassay (Diagnostic Reagents Inc., CA, USA) on a 502X Multiple Chemistry Unit automated analyzer (A & T Co., Tokyo, Japan) according to manufacturer's instructions. Briefly, 35 μL of urine or calibrator was added to 125 μL of reagent, which included a cotinine-specific monoclonal antibody and enzyme substrate. Next, 125 μL of cotinine-labeled enzyme (glucose-6-phosphate dehydrogenase) was administered and incubated at 37°C. Spectrophotometric measurements were obtained at 340 nm.

Statistical analysis
All statistical analyses were performed using the Statistical Package PASW 18.0 (Chicago, IL, USA). Continuous variables were expressed as mean±standard deviation (SD). To compare urinary cotinine levels of each group, the Student t-test or one-way analysis of variance (ANOVA) were performed. The intergroup differences of median urinary cotinine levels were assessed by the Mann-Whitney U test. Pearson's or Spearman's correlation coefficients were used to investigate correlations between smoking status profile, the FTND score, and urinary cotinine levels. To assess the association between urinary cotinine levels and other variables with high nicotine dependence, stepwise multiple linear regression model was used with adjustment for age, sex, smoking characteristics, and socioeconomic status. Using receiver operating characteristic (ROC) curves, the accuracy and optimal cut-off values of urinary cotinine in the prediction of high nicotine dependence were assessed by the area under the curve (AUC) range of 0.5-1. All confidence intervals (CIs) were set at 95% and statistical significance was determined as P<0.05.

Results
A total of 381 subjects were included in the final study: 352 men (92.4%) and 29 women (7.6%). The mean age was 45.4±7.8 years. The mean FTND score and urinary cotinine level of all subjects were 3.5±2.4 and 786.3±576.3 ng/mL, respectively. General and socioeconomic characteristics and smoking status of the study subjects are summarized in Table 1. Furthermore, the mean FTND score and urinary cotinine level for each subgroup based on characteristics and smoking status are shown. Subjects with little or no intention to quit smoking, who were unmarried, or had been divorced, separated, or bereaved, had significantly higher FTND scores and urinary cotinine levels. Table 2 shows the results of the correlation analysis among age, education, average monthly income, smoking status, FTND scores, and urinary cotinine levels. Particularly, the number of cigarettes smoked per day showed a marked correlation with the FTND scores and
Of the 381 subjects, 205 (53.8%) were classified into the low, 127 (33.3%) into the moderate, and 49 (12.9%) into the high nicotine dependence groups. Urinary cotinine levels significantly increased with the degree of nicotine dependence (P < 0.001): low=523.3±449.0 ng/mL; moderate=982.6±510.4 ng/ mL; and high=1378.3±577.7 ng/mL nicotine dependence group (Figure 1). In a post-hoc analysis, there were significant differences between the comparisons of urinary cotinine levels of each nicotine dependence groups.
In ROC analysis (Figures 2), urinary cotinine levels were reasonably accurate for prediction of the high nicotine dependence group (AUC=0.82; 95% CI=0.76-0.87; P < 0.001). Using this analysis, the optimal cutoff value of urinary cotinine levels for high nicotine dependence was identified as 1,000 ng/mL. For this cutoff value, the sensitivity, specificity, positive predictive value, and negative predictive value (%) for prediction of the high nicotine dependence were 71.4, 74.4, 58.2, and 83.9 respectively.
In the stepwise multiple linear regression model (Table 3), age, marital status, education (13-16 years), average monthly income, age at smoking initiation, history of smoking cessation, intention to quit smoking, and the presence of a smoking partner were excluded. Finally, education (>16 years), total smoking period, and urinary cotinine levels were included with a satisfactory explanation power (R2=0.348, P < 0.001). The results of the collinearity diagnostics on those variables were acceptable, because their variance inflation factor (VIF) ranged from 1.001 to 1.003. The total smoking period (β=0.042, P=0.001) and urinary cotinine levels (β=0.234, P < 0.001) were positively associated with nicotine dependence, whereas an inverse association between highest education levels (>16 years) and nicotine dependence was observed (β=-0.573, P=0.034).

Discussion
Our results indicate that the level of urinary cotinine is independently associated with the degree of nicotine dependence and that cotinine is a valid biomarker for prediction of high nicotine dependence. To the best of our knowledge, this study is the first to suggest an available cut-off value of urinary cotinine as an indicator FTND, the Fagerström test for nicotine dependence; VIF, the Variance Inflation Factor; Performing a stepwise multiple regression analysis, age, marital status, education (13-16 years), average monthly income, age at smoking initiation, history of smoking cessation, intention to quit smoking, and partner's smoking were excluded in the final model interquartile range (25-75 percentile) by nicotine dependence groups (low, FTND score=0-3; moderate, FTND score=4-6; and high, FTND score=7-10). Urinary cotinine levels significantly increased as the degree of nicotine dependence. *P was obtained by the Mann-Whitney U test

Figure 2. Receiver Operating Characteristics (ROC) Curves Analysis of Urinary Cotinine Levels for High Nicotine Dependence. ROC curves analysis
showed that urinary cotinine levels were reasonably accurate for prediction of the high nicotine dependence group (high nicotine dependence is defined as FTND score≥7) (area under curve=0.82; 95% confidence interval=0.76-0.87; P<0.001; sensitivity=71.4%; specificity=74.4%; positive predictive value=58.2%; and negative predictive value=83.9%) of high nicotine dependence. Because cotinine is an alkaloid found in tobacco and a metabolite of nicotine, its levels in the body have predominantly been used to verify smoking status in individuals. However, our data potentially suggest an extended application of urinary cotinine levels, particularly for assessment of nicotine dependence. We suggest that the optimal cut-off value for the urinary cotinine level that can predict high nicotine dependence is 1,000 ng/mL (sensitivity: 71.4%, specificity: 74.4%).
Urinary cotinine levels and the FTND scores showed similar patterns of change with respect to the socioeconomic status of smokers and smoking-related factors. In the correlation analysis, urinary cotinine levels were associated with both the number of cigarettes smoked per day and the FTND scores, which was consistent with previous studies (Heatherton et al., 1991;Payne et al., 1994;Law et al., 1997;Caraballo et al., 1998;Park et al., 2004;Blackford et al., 2006). In contrast to previous reports that showed that the duration of smoking was not associated with the FTND scores (John et al., 2003;Park et al., 2004), the degree of nicotine dependence increased according to the total smoking period in our study. This is an interesting finding because FTND scores have been thought to reflect mainly physical dependence, and the scale did not include any component related to smoking persistence. Furthermore, this result could partly support the "hardening hypothesis" that more dependent smokers maintain smoking, despite the overall decline of smoking prevalence (Fagerstrom and Furberg, 2008).
Our results also showed an inverse correlation between the age of smoking initiation and the FTND scores, consistent with previous findings (Breslau et al., 1993;Park et al., 2004). A possible explanation for this is that an earlier onset of smoking could induce the development of nicotine habituation and tolerance in a younger age, contributing to subsequent higher nicotine dependence and difficulty in smoking cessation (Breslau and Peterson, 1996;Khuder et al., 1999). However, this inverse correlation was not observed in the step-wise multiple regression analysis.
As observed in previous research (Kraft et al., 1998;Siahpush et al., 2006;Cheah and Naidu, 2012), higher education levels had an inverse association with nicotine dependence. This finding could explain the observation that low education levels could be a barrier for smoking cessation (Breslau and Peterson, 1996;Khuder et al., 1999), although the causal relationship should be confirmed in further prospective research. Consistent with a recent study that showed that lower intention to quit smoking was associated with higher nicotine dependence (Feng et al., 2010), our study found that willingness to quit smoking was inversely correlated with FTND scores and urinary cotinine levels, respectively. This finding could further support our study hypothesis that urinary cotinine levels reflect nicotine dependence status in active smokers because increased intention to quit smoking often leads to a substantial increase of smoking cessation attempts. Thus, more comprehensive approach to regard the association of urinary cotinine levels with the socioeconomic status and smoking behaviors of smokers might be required in future studies.
Serum cotinine is a reliable indicator of nicotine intake. However, previous reports have shown that nicotine is eliminated in the serum, urine, and saliva at a similar rate and suggested that cotinine levels in the urine could provide a comparable indication to serum levels of nicotine intake (de Weerd et al., 2002;Shin et al., 2002). In another study that investigated the association of urine and serum cotinine levels with the amount of time prior to smoking the first cigarette of the day, a similar parallel tendency and association of urinary and plasma cotinine levels with the smoking behavior was observed (Muscat et al., 2009). Therefore, considering the simplicity and non-invasiveness of obtaining urine samples, measurement of urine cotinine levels might be a preferable alternative to measurement of plasma cotinine levels in clinical settings.
There are several limitations to this study. First, because the study was retrospective, the major findings should be confirmed by additional studies with higher evidence levels. Second, the mean age of the study subjects was 45.4 years, and the percentage of women in the study was low (7.6%). The high average age of subjects could be attributed to their selection from a regular health check-up program in a single cancer prevention center. Consequently, these data should be interpreted with caution and not generalized to other populations, especially women and young smokers. All information pertaining to smoking status was acquired by self-administered smoking questionnaires; therefore, the potential recall bias should be considered. Finally, because serum cotinine levels in cigarette smokers have been shown to be dependent on race and ethnicity (Caraballo et al., 1998), the application of the cut-off value for the urinary cotinine level in this study should be restricted to Asians. Accordingly, determination of an optimal cut-off value to detect high nicotine dependence for other races should only follow sufficient consideration of results from additional inter-racial comparative studies.
The present study suggested that urinary cotinine levels could reflect not only smoking status, but also the degree of nicotine dependence. In addition, an available optimal cut-off value of the urinary cotinine to determine high nicotine dependence was also assessed, which further support a clinical availability of urinary cotinine measurement as a screening biomarker ahead of smoking cessation treatment in active smokers. Further studies are needed for a more broad utilization of urinary cotinine measurement.