Exposure to tobacco smoke based on urinary cotinine levels among Israeli smoking and nonsmoking adults: a cross-sectional analysis of the first Israeli human biomonitoring study

Environmental tobacco smoke (ETS) is a combination of smoke emitted from a burning tobacco product and the smoke exhaled by the smoker, which is also called secondhand smoke [1].

The adverse effects of ETS among nonsmoking adults are well proven and mirror those associated with active smoking [2]. ETS is causally associated with lung cancer among never-smokers and among nonsmokers [3]. ETS increases the risk of cardiovascular disease by approximately 30% [4], and also increases the risk of respiratory diseases [5].

Various methods are available to measure ETS: self-reporting, environmental measurements and human biomonitoring of various biomarkers in different biological media [6]. Nicotine is a specific biomarker of exposure to tobacco smoke, either active or ETS, but due to its short half life (1-3 hours) it has limited value as a marker of exposure. Cotinine, the primary proximate metabolite of nicotine, is used most frequently as a biomarker of tobacco smoke exposure, as its half-life is longer (approximately 16-18 hours) and levels remain fairly constant during the day. The cotinine level provides a valid and quantitative measure of average recent human ETS exposure and is therefore the preferred biomarker of exposure to tobacco smoke in active smokers and in nonsmokers exposed to ETS [7]. Dietary intake of nicotine from food like fruits and vegetables is possible but likely to be negligible [8]. Urinary cotinine, especially when corrected for creatinine concentration, is highly correlated with plasma cotinine [9].

In Israel, according to the recent national survey, the overall active smoking rate in the adult (21 years and older) population, based on self-report, is 20.6% [10]. Smoking rates vary by sex and ethnicity: being highest among Arab males (43.8%) and lower for Jewish males (23.7%), Jewish females (15.9%) and Arab females (6.7%). There was a recent study on ETS exposure in public places in Israel based on air quality measurements [11]. However, assessment of ETS exposure of the Israeli population by human biomonitoring has never been conducted, thus the extent of exposure to ETS is unknown, both for the entire population and for specific sub-groups.

The goal of this study was to examine exposure to tobacco smoke among smoking and nonsmoking Israeli adults and to identify differences in ETS exposure among nonsmokers by socio-demographic factors, in order to plan tobacco control activities and to serve as a baseline for future monitoring.

All participants, smokers and nonsmokers, had urinary cotinine levels > LOD. The distribution of cotinine concentration levels (μg/liter) among smokers and nonsmokers reflects clear differences (see Figure 1). Urinary cotinine concentrations were significantly higher among smokers (GM = 89.7 μg/g creatinine; 47.4-169.6) than among nonsmokers (GM = 1.3; 1.1-1.7). Among the smokers, the majority (84.6%) smoked cigarettes (or cigarettes and other forms) and the minority were exclusive waterpipe users (14.3%) or cigar smokers and pipe users (1.1%). Cotinine levels were significantly higher among smokers who reported >20 cigarettes/day (GM = 328.2; 67.7-1591.7, PR = 5.86; 95% CI 1.11-30.85, p = 0.040) as compared to smokers who reported <10 cigarettes/day (GM = 53.0; 18.3-153.0), and not significantly higher as compared to those who reported 10–20 cigarettes/day (GM = 73.5; 22.3-242.9). Among the group of exclusive waterpipe smokers, all had levels above LOQ and cotinine levels were relatively high (GM = 53.4; 95% CI 12.3-232.7) and not significantly different than in cigarette smokers (GM = 89.3; 42.5-187.6).

There was no significant difference in the proportion ≥ LOQ or ≥4 μg/l or the GM by country of birth, ethnicity or urbanicity.

As a sensitivity analysis, we further excluded 5 participants with self-reported smoking status as nonsmokers and creatinine-adjusted levels of cotinine >50 μg/g (but <150 μg/g). Findings were similar in the mutually adjusted model: associated with being male (PR = 1.28; 95% CI: 1.00-1.63, p = 0.048), inversely associated with age (PR = 0.99; 0.98-1.00, p = 0.014, per one additional year) and associated with having lower educational status (PR = 1.55; 1.03-2.35, p = 0.037).

The present study, the first to evaluate exposure to tobacco smoke using a biological marker in an Israeli general population, demonstrates widespread and high ETS exposure in the non-smoking adult population in Israel. Despite limited power due to small sample size, ETS exposure was found to be significantly more common among males, younger study participants and those with lower education.

We found clear differences in urinary cotinine levels between smokers and nonsmokers, with higher levels by smoking frequency. This expected finding gives some validation to our study methods such as data collection and laboratory analysis. Cotinine levels among exclusive waterpipe users were similar to those in cigarette smokers, in line with known limited data indicating that daily use of waterpipe produces nicotine absorption of a magnitude similar to that produced by daily cigarette use [21]. This observation gives further evidence to the addictive potential of this common habit in our population. There is a gradual increase in evidence related to the harmful effects of waterpipe smoking, as was suggested by our previous findings in Israel [22]. The effects of waterpipe passive smoking are largely unknown, although a recent Israeli study showed some physiological effects [23].

In our study all nonsmokers had levels of urinary cotinine above LOD, 62.2% had levels ≥1 μg/liter and 16.2% had levels ≥4 μg/liter. Our results are comparable to those in a previous study conducted in the German population in 1998, which found urinary cotinine levels ≥4 μg/liter in 23% of the nonsmokers and even higher in males (26%) as compared to females (19%) [18]. In a recent study in the Canadian general population 15% of those aged 20-39 years and 11% of those aged 40-59 years had urinary cotinine levels >1.1 μg/liter [19]. In our population, using the same cut-off, rates were at least four times higher in both age groups (67% and 45%, respectively). Median or geometric mean levels could not be compared as those were not reported in the Canadian or German studies due to a high proportion below the LOD, while in our study all nonsmokers had levels above LOD. In a finding similar to that in our study, widespread ETS exposure was previously found among nonsmokers in another Middle-Eastern country–Syria [24]. In that study, a moderate correlation was found between salivary cotinine levels and self-reported exposure measures such as house smoking policy. Our study gives further evidence to the widespread and high level of ETS exposure in the Middle East region.

The proportion of population exposed to ETS in our study, based on urinary cotinine concentrations ≥ LOQ, was much higher in males compared to females, a finding consistent with previous studies in the adult population [18, 19, 25, 26]. Lower education and younger age were also important risk factors for exposure and remained so after mutual adjustment, consistent with studies in other countries [26–29].

In order to supplement our study, we analyzed data on self-reported ETS among nonsmokers, based on the 2010 Israel Central Bureau of Statistics Social Survey [30]. The survey was conducted by interviewing ~7500 Israelis during 2010, a random sample of the adult Israeli population. In this survey, 69.3% of nonsmokers aged 20-74 reported exposure to ETS, higher (74.3%) for males compared to females (65.5%). Among those who reported exposure, 33.6% reported ETS at the workplace, 26.8% at home, and 77.3% in other places. A much higher percentage of males reported ETS at work (46.7%) than females (22.3%). In contrast, a higher proportion of females reported ETS at home (35.4%) compared with males (16.8%). This analysis gives further evidence to the widespread ETS exposure in the Israeli population found in our study. Additional support for widespread ETS exposure stems from a study conducted among Israeli adolescents, based on self-reported exposure [31].

Public health and health promotion efforts regarding tobacco smoke exposure should be directed to the young male population, as not only are active smoking rates high, but ETS is also high and common among nonsmokers in this population. This group needs tailored interventions, to address social needs as well as attitudes to ETS which may include more tolerance to exposure at work.

The main finding of the study, widespread and high ETS in all groups of the Israeli population, has served to raise support and political will for taking further steps to reduce ETS in the public and private arenas in Israel. The present study was instrumental in expanding smoke-free legislation implemented in Israel in July 2012. This new legislation included prohibition of smoking, for the first time, in open spaces such as outside entrances to health facilities, railway platforms, etc. [32].

There are major problems in implementation of smoke-free legislation in bars and pubs in Israel due to social norms [33], despite the fact that the majority (67%) of Israelis support completely smoke-free bars and pubs, according to a previous survey [34]. The new popular trend of e-cigarettes (battery-operated products that deliver nicotine, flavor and other chemicals via a vapor that is inhaled by the user) in Israel and elsewhere should be taken into account when planning tobacco control policy, as it may reduce ETS but may give legitimacy to smoking in public places [35]. There is evidence that mobile health (m-Health) efforts for smoking cessation can be effective and mobile phones should be explored as a method for preventing ETS as well [36]. We are currently developing a text messaging program for Israeli smokers as well as a mobile phone application for reporting violations of the smoke-free law in public places in Israel.

National human biomonitoring studies on ETS exposure using cotinine measurements have been used by other countries to demonstrate the positive effects of smoke-free legislation and identify its impact on different population groups [25, 26]. Cotinine biomonitoring in the general population is an important tool for tobacco control. Conducting national biomonitoring programs can be used to determine the average (and range of) ETS, establish baseline levels of exposure and assess trends over time, provide information for directing public health priorities for environmental health policy, including ETS, support and direct future research and identify potentially vulnerable groups. Urinary cotinine monitoring can also be used to evaluate the success of public health interventions. Such programs exist in the United States, Canada, Belgium, France, Germany and the European Union. Active tobacco smoking and ETS are important exposures in themselves, as well as confounders/mediators in many other environmental exposures, and should be assessed in depth in any national human biomonitoring program. We plan to conduct a follow up biomonitoring study in Israel in 2015 on exposure to ETS (as well as pesticides) in Israel, in order to determine whether the policies described above have been effective in reducing public exposure to ETS in the general population.

Two major strengths of the current study include the representation of a wide geographical distribution and different socio-demographic groups and the use of a valid, sensitive and accurate quantitative biochemical assessment of ETS. Cotinine was shown in the past to be derived from exposure to tobacco smoking, while other sources, such as dietary sources, were shown to be negligible [8]. Indeed, we did not find an association between urinary cotinine and fruits or vegetables consumption in our population (data not shown).

Nonetheless, some limitations of the study should be considered. An important methodological limitation of the study is the “convenient” non-random sampling technique. It is not certain to what extent the study population is representative of the general adult population in Israel. For example, the Arab population was over-represented in the study (25%) compared to 18% in the Israeli population of the same age [37]. Hence, interpretation of the findings should be done cautiously, although selection bias is an unlikely explanation for the widespread high ETS exposure found. A second limitation of our study is the lack of self-reported ETS status and sources, which could have helped in elaboration of ETS patterns. However, urinary cotinine levels give better estimates of ETS and we were able to confirm our findings by data from the 2010 Social Survey. Another limitation is the possibility of that cotinine levels found in our nonsmoking population derived not only from ETS, but, in some cases, from misclassification of smoking status or use of nicotine replacement therapy as a source of exposure. We minimized this possibility by exclusion of those with cotinine levels >150 μg/g and by our sensitivity analysis further excluding cotinine levels >50 μg/g. Unfortunately, we did not collect data on time elapsed since last cigarette smoked among self-reported smokers which might explain why some self-reported smokers (who did not smoke recently) had relatively low urinary cotinine levels as well as the higher smoking rate in our study population compared to the general Israeli adult population. Self-reported smoking status is considered valid, especially for those aged 20 and above [38], so it is unlikely that misclassification by smoking status significantly affected our results. Finally, our relatively small sample size gives limited power to identify differences in magnitude of exposure as expressed by GM, or between specific sub-groups such as those based on ethnicity.

This study shows widespread ETS exposure in the nonsmoking adult Israeli population, especially among males, the younger population and those with lower education. These findings demonstrate the importance of human biomonitoring, were instrumental in expanding smoke-free legislation implemented in Israel on July 2012 and will serve as a baseline to measure the impact of the new legislation.