Determinants of Continued Smoking After Cancer Diagnosis: The Role of Household and Social Exposure
Arnav Pemmaraju, Melissa Neumann, Sheeja Rajan, Narifa Aziz, Kevin Sullivan, Nagashree SeetharamuBackground
Continued use of combustible tobacco after a cancer diagnosis is associated with increased morbidity and mortality. Despite this, tobacco cessation remains difficult for many patients, underscoring the need to better understand clinical and social factors associated with continued smoking after diagnosis.
Methods
In this IRB-exempt study, we administered an anonymous, structured questionnaire using elements of the NCI-AACR Cancer Patient Tobacco Use Questionnaire (C-TUQ) and the Fagerström Test for Nicotine Dependence (FTND), along with additional domains assessing social exposure and behavioral factors to patients with lung or head and neck cancer and any lifetime tobacco exposure treated at a cancer center between June 2022 and December 2023. The survey captured demographics, cancer type and stage, detailed smoking history, household and partner tobacco use, primary care provider (PCP) involvement, motivation to quit, and perceptions of tobacco-related cancer risk. Descriptive statistics summarized patient characteristics. Comparisons were made between individuals who continued smoking after diagnosis and those who successfully quit. Multivariable logistic regression was performed among respondents with complete covariate data (n=60) to evaluate independent associations between smoking cessation and age, cumulative tobacco exposure, and social network factors.
Results
Of 200 respondents, 90 reported smoking at diagnosis; 52 continued and 38 had successful cessation. Continued users were older (72.9±11.0 vs 67.9±10.3 years; p=0.031) and reported a longer smoking history (median 45.1 vs 30.8 years; p<0.0005). A greater share of continued users had a tobacco-using partner (65.4% vs 42.1%; p=0.041). Household income, gender, race, family structure, marital status, and age at smoking initiation were similar between groups. Half of those who quit had done so within three months of the survey; 53.8% reported increased motivation to quit post-diagnosis. Only 6.9% of patients were enrolled in a cessation program. In an exploratory multivariable framework, partner smoking contributed the largest independent increase in the odds of continued smoking, exceeding age, duration, and pack-years, supporting a strong social-network influence. These findings have informed the next phase of the study, which will be a randomized controlled trial comparing dyadic (patient and a partner in their social network) or “Quit Together” approach to standard individual-based cessation programs. Positive results from this trial can shift norms, increase accountability, and foster smoke-free home environments, while providing secondary prevention for patients diagnosed with cancer and primary cancer prevention for their partners. Conclusion: Older age, higher cumulative exposure, and tobacco use in close social networks were associated with continued smoking. Exploratory simulation based on observed subgroup statistics suggests that partner/household smoking provides the largest incremental explanatory power among covariates highlighting the imminent need for social-network-based tobacco cessation interventions.