Differential contribution of smoking to lung cancer by education in Czechia: prospective study
Abstract
Background: Czechia has amongst the highest prevalence of smoking in Europe but little enough is known about how much smoking contributes to lung cancer in different socioeconomic groups.
Objectives: To estimate the relative and absolute contribution of smoking to lung cancer incidence by sex and educational categories.
Methods: We used data from a prospective cohort of adults aged 45-69 at baseline (2002-05) from Czech cities. Cox regression was used to estimate hazard ratios (HRs). The relative and absolute contribution of smoking to lung cancer burden within educational categories (in men and women separately) was estimated by Population Attributable Risk Fraction (PARF) and Population Attributable Risk (PAR), respectively.
Results: Among 8,674 subjects with valid data, 238 new cases of lung cancer occurred during average 18 years of follow up. Smoking was strongly associated with the incidence of lung cancer, with substantially stronger association in men (HR 26.7, 95%CI 9.8- 73.3) than in women (HR 6.28, 95%CI 3.81-10.34), and smoking explained a substantial part of the educational gradient in lung cancer incidence. PARF calculations suggested that 91% of incident cases in men and 53% in women can be attributed to smoking but there was no educational gradient in relative impact. By contrast, the absolute impact showed a pronounced social gradient. In men, PAR ranged from 0.86 per 1,000 person-years in men with university education to 2.96 in those with primary education or less; the corresponding PAR values in women were 0.62 and 0.91 per 1,000 person-years.
Conclusions: The study confirms the strong association between smoking and lung cancer incidence and highlights the unequal absolute burden of smoking-related lung cancer across socioeconomic and sex groups, which is largely driven by smoking prevalence. The findings demonstrate that relying solely on relative measures may obscure inequalities in the burden of lung cancer.
Key messages: • There are pronounced inequalities in the absolute impact of smoking on lung cancer risk.
• These inequalities would be obscured by focusing solely on relative impact.
Related
- Regional and sex differences in the contribution of smoking-related mortality to changes in adult life expectancy in Brazil, 2000-23
- Smoking and dementia risk: an updated systematic review and meta-analysis of 61 prospective cohort studies and implications for tobacco control
- Managing public education on smoking behavior in families with stunted children: Between challenges and opportunities in Indonesia.
- Smoking cessation outcomes by employment status and education attainment after intensive smoking cessation intervention
- Smoking, education level and excess premature deaths in the USA
BibTeX
@article{bobak2026differential,
title = {{Differential contribution of smoking to lung cancer by education in Czechia: prospective study}},
author = {M Bobak and A Castedo Garcia and A Dalecka and M Marmot},
journal = {The European Journal of Public Health},
year = {2026},
month = oct,
doi = {10.1093/eurpub/ckag169.114},
url = {https://doi.org/10.1093/eurpub/ckag169.114},
note = {CC BY-NC 4.0},
}
RIS
TY - JOUR
TI - Differential contribution of smoking to lung cancer by education in Czechia: prospective study
AU - M Bobak
AU - A Castedo Garcia
AU - A Dalecka
AU - M Marmot
PY - 2026
DA - 2026/10/01
JO - The European Journal of Public Health
DO - 10.1093/eurpub/ckag169.114
UR - https://doi.org/10.1093/eurpub/ckag169.114
AB - Background: Czechia has amongst the highest prevalence of smoking in Europe but little enough is known about how much smoking contributes to lung cancer in different socioeconomic groups. Objectives: To estimate the relative and absolute contribution of smoking to lung cancer incidence by sex and educational categories. Methods: We used data from a prospective cohort of adults aged 45-69 at baseline (2002-05) from Czech cities. Cox regression was used to estimate hazard ratios (HRs). The relative and absolute contribution of smoking to lung cancer burden within educational categories (in men and women separately) was estimated by Population Attributable Risk Fraction (PARF) and Population Attributable Risk (PAR), respectively. Results: Among 8,674 subjects with valid data, 238 new cases of lung cancer occurred during average 18 years of follow up. Smoking was strongly associated with the incidence of lung cancer, with substantially stronger association in men (HR 26.7, 95%CI 9.8- 73.3) than in women (HR 6.28, 95%CI 3.81-10.34), and smoking explained a substantial part of the educational gradient in lung cancer incidence. PARF calculations suggested that 91% of incident cases in men and 53% in women can be attributed to smoking but there was no educational gradient in relative impact. By contrast, the absolute impact showed a pronounced social gradient. In men, PAR ranged from 0.86 per 1,000 person-years in men with university education to 2.96 in those with primary education or less; the corresponding PAR values in women were 0.62 and 0.91 per 1,000 person-years. Conclusions: The study confirms the strong association between smoking and lung cancer incidence and highlights the unequal absolute burden of smoking-related lung cancer across socioeconomic and sex groups, which is largely driven by smoking prevalence. The findings demonstrate that relying solely on relative measures may obscure inequalities in the burden of lung cancer. Key messages: • There are pronounced inequalities in the absolute impact of smoking on lung cancer risk. • These inequalities would be obscured by focusing solely on relative impact.
SN - 1101-1262
SN - 1464-360X
ER -