Study design and participants

This cross-sectional study was conducted as an ancillary survey of an ongoing multicenter epidemiological study of workers in Japan¹². In April 2024, a survey was conducted during annual health checkups at a participating company’s smoke-free factory, where the use of all tobacco products, including cigarettes and heated tobacco products, is prohibited in all areas, including outdoors, and workers are not permitted to leave the premises for these purposes during work hours or lunch breaks. Participants were asked to complete a self-administered questionnaire regarding smoking habits, SSA exposure, and occupational and behavioral factors, and to provide urine samples for biomarker assessment. Anthropometric measurements, biochemical data, and medical history were obtained from health checkup records. Workers were included in the present analysis if they underwent the periodic health checkup and agreed to participate in the study. Those with missing information on smoking status or without a urine sample were excluded. For the main analysis among non-smokers, users of any tobacco product were further excluded. The study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Ethics Committee (approval number: JIHS-S-001140). Written informed consent was obtained from all participants before data collection.

Measurement of urinary creatinine and cotinine

Participants were instructed to collect first-morning urine samples in designated collection tubes and submit them at the survey reception desk during the health checkup. Urinary creatinine concentrations were measured using an enzymatic method at a commercial laboratory (LSI Medience Corporation, Tokyo, Japan). The remaining aliquots were transported to the National Institute of Public Health and stored at -80°C until biochemical analysis.

Urinary nicotine metabolites, including cotinine, were measured by liquid chromatography–tandem mass spectrometry (LC-MS/MS), with the addition of an enzymatic hydrolysis step to convert cotinine glucuronide to free cotinine. Briefly, urine samples were incubated for 6 h at 37°C with 1000 units of β-glucuronidase from Helix pomatia (SigmaAldrich, St. Louis, MO, USA; catalog no. G7017) in 100 mM ammonium acetate buffer (pH 4.50). The hydrolyzed samples were then processed and analyzed¹³. Accordingly, cotinine concentrations are reported as total cotinine, comprising free cotinine and cotinine glucuronide. The limit of quantification for urinary cotinine was 0.01 ng/mL. To account for urinary dilution, quantified cotinine concentrations were normalized to urinary creatinine concentrations and expressed as ng/mg creatinine. Participants with cotinine concentrations below the limit of quantification were classified as having undetectable cotinine.

We visually inspected the distribution of urinary cotinine concentrations among all participants, including smokers. The cotinine concentration corresponding to the lowest frequency between the peaks was used as the cutoff separating typical smokers from non-smokers. Among self-reported non-current smokers with detectable cotinine concentrations below this cutoff, participants were further categorized into tertiles according to cotinine concentration. Participants with cotinine concentrations greater than the cutoff were classified as the very high exposure group. Thus, four urinary cotinine categories (Tertile 1 [low], Tertile 2, Tertile 3 [high], and very high exposure) were created among self-reported non-current smokers. Tertile 3 and the very high exposure group were combined and defined as the high exposure group for regression analyses.

Smoking, SSA exposure, and other variables

Information on demographic, occupational, and behavioral factors was obtained from questionnaires and health checkup records. Participants were asked about current and past smoking habits, including use of combustible cigarettes, HTPs, and other tobacco products. They were also asked about the frequency of SSA exposure separately for combustible cigarettes and HTPs in any settings, including homes, workplaces, restaurants, bars, building entrances, and other public spaces, with response categories ranging from ‘never’ to ‘daily’. Participants who reported exposure to either tobacco type at least once per month were considered to be aware of SSA exposure in daily life. Background factors considered in the analysis were sex, age, education level, marital status, occupation, work shift, chronic condition limiting daily activity, former smoking, and alcohol use.

Statistical analysis

The proportions of participants with self-reported SSA exposure, detectable urinary cotinine concentrations, and high detectable urinary cotinine concentrations (≥0.207 ng/mg creatinine) among self-reported non-current smokers were calculated and expressed as percentages. These proportions were evaluated according to sex (male, female), age (<40 or ≥40 years), education level (<13 or ≥13 years), marital status (married, other), occupation (non-manual, manual), work schedule (day shift, night/rotating shift), chronic condition limiting daily activity (no, yes), smoking status (never, former), and alcohol use (<1 or ≥1 day/week). Differences between groups were tested using the chi-squared test. Univariable logistic regression analysis was used to estimate crude odds ratio (OR) and 95% confidence interval (CI) for having high detectable nicotine exposure, defined as belonging to the highest tertile group and the very high exposure group combined, in relation to each covariate. We also ran a multivariable logistic regression model including all the abovementioned variables to estimate the adjusted odds ratio (AOR). To evaluate discrepancies between biomarker-based nicotine exposure and self-reported SSA exposure, we calculated the proportion of participants who were aware of SSA exposure according to urinary cotinine category and tested the trend among those with detectable cotinine using the Cochran–Armitage trend test. All analyses were performed using Stata version 18.0 (StataCorp LLC, College Station, TX, USA), and a two-sided p<0.05 was considered statistically significant.

Prevalence of SSA

The high prevalence of detectable cotinine observed in the present study is consistent with biomarkerbased studies^6,7, suggesting that low-level nicotine exposure persists despite progressively strengthened smoking restrictions in public places and workplaces⁴. In a nationally representative survey of US non-smoking adults using NHANES 2013–2020 data, Wang et al.⁶ reported detectable serum cotinine concentrations in 51.2% of participants⁶. In Korea, urinary cotinine was detected in over 80% of non-smoking adults in nationally representative survey data collected between 2009 and 2018⁷. The present study extends these findings by examining urinary cotinine concentrations among workers employed at a smoke-free factory under current conditions in which HTPs are widely used in Japan.

The prevalence of detectable cotinine in the present study and the above Korean study⁷ was much higher than that reported in the US study⁶. This could be ascribed to the use of urinary cotinine measurements. Cotinine is an established biomarker of environmental tobacco smoke exposure⁵, and its relatively long elimination half-life of approximately 16–20 hours in body fluids allows cotinine measurements to reflect recent nicotine exposure over the preceding one to several days¹⁴. Urinary cotinine concentrations are generally several-fold higher than those in blood or saliva, which facilitates the detection of low-level nicotine exposure¹⁵. The Korean study excluded non-smokers who had cotinine levels comparable to smokers⁷, whereas we included workers with very high cotinine levels, likely accounting for the higher prevalence in our study than the Korean study.

Factors associated with SSA

In the present study, detectable urinary cotinine was highly prevalent across participant subgroups, suggesting that low-level SSA exposure may be widespread even among non-smoking workers in a smoke-free workplace. In contrast, high urinary cotinine concentrations were not uniformly distributed: female sex, manual work, and former smoking were independently associated with higher odds of high urinary cotinine concentrations, whereas higher level of education was associated with lower odds. Some of these patterns are consistent with previous biomarker-based studies. In a Korean national study, cotinine-defined SHS exposure was more prevalent among adults engaged in blue-collar work⁷, and in a US national study, cotinine-defined SHS exposure was more prevalent among adults with lower level of education⁶. These differences may reflect social and environmental disparities in exposure opportunities, including residential environments, household smoking, social networks, and the ability to avoid settings where tobacco smoke or HTP aerosol is present. Given the strict factory-wide smoking ban in the present study, exposure within the workplace was unlikely. Therefore, the higher odds observed in these subgroups may reflect unmeasured SSA exposure outside the factory, such as at home, during commuting, in social settings, or in venues where smoking or HTP use is allowed. Former smokers may also be more likely than never smokers to live with or socialize with smokers, which could increase opportunities for unrecognized exposure.

Awareness of SSA exposure

In the present study, there was a large gap between low self-reported awareness and high biomarker-based exposure of SSA, with nearly 80% of those with detectable urinary cotinine being unaware of their exposure. This finding is consistent with previous biomarker-based studies showing that many individuals with measurable nicotine exposure do not recognize their SSA exposure. In a 2018 Korean national survey, substantial discrepancies were observed between self-reported indoor SSA exposure (3.7% at home, 15.4% at public places, and 12.2% at workplaces) and urinary cotinine-based assessment (80.0%)⁷. In the US nationally representative survey, 22.0% self-reported SSA exposure, whereas 51.2% had detectable serum cotinine, with 67.6% of those with detectable serum cotinine being unaware of their exposure⁶. Nicotine exposure in daily life may occur intermittently in settings outside the workplace, such as building entrances, outdoor smoking areas, restaurants, bars, and commuting spaces, which people may not be aware of. The widespread use of HTPs in Japan, especially in younger working-age adults⁸, may partly explain this large gap. Compared with combustible cigarettes, HTPs generate aerosol rather than smoke and are often perceived as producing less visible emissions and less residual odor^9,16, potentially making passive exposure less noticeable to bystanders while still delivering nicotine-containing emissions.

Strength and limitations

A major strength of the present study is the use of urinary cotinine, an objective and sensitive biomarker of nicotine exposure, which enabled the detection of low-level exposure. In addition, the study was conducted among workers employed at a workplace with a strict smoke-free policy, providing insight into SSA exposure outside the workplace, which has been a major location where SSA exposure occurs.

However, several limitations should be considered. First, smoking status and secondhand smoke or aerosol exposure were self-reported and therefore may be subject to information bias and misclassification. In particular, some current smokers may have been misclassified as non-smokers because of inaccurate reporting of smoking status, potentially contributing to the high prevalence of detectable urinary cotinine. Second, although urinary cotinine is a useful biomarker of recent nicotine exposure, it cannot identify the specific timing, location, or source of exposure. Therefore, we were unable to determine whether the detected cotinine reflected exposure in the workplace, at home, during commuting, or in other daily life settings. Third, although multivariable logistic regression analyses were performed to adjust for measured background factors, residual confounding by unmeasured or imperfectly measured factors cannot be excluded. In addition, because this was a cross-sectional analysis, the observed associations between background factors and high urinary cotinine concentrations should not be interpreted as causal relationships. Rather, these findings should be regarded as exploratory evidence for identifying subgroups with higher levels of cotinine-detected nicotine exposure. Lastly, the study population was drawn from a workplace with a strict factory-wide smoking ban, which may limit generalizability of our findings to workplaces with less stringent smoking policies.