Understanding socioeconomic status as a risk for lung cancer

Lung cancer is the pre-eminent cause of cancer death worldwide. The development of lung cancer occurs due to numerous risk factors, with tobacco smoking the leading etiology. The SYNERGY project was created to thoroughly evaluate smoking in conjunction with occupational exposures as risk factors for lung carcinogenesis. The study was a pooled analysis of case-control studies from 22 centers in Europe and Canada, collecting data on 39,470 individuals from 1985 to 2010. The largest contributors of cases and controls were centers in Germany, France, Italy, Canada, and Sweden (1).

In 2018, an initial analysis of the SYNERGY project was conducted to examine the relationship between socioeconomic status (SES), largely defined by occupation type and status, and lung cancer incidence and further to control for how smoking habits and occupational exposures attenuate this risk. Based on these data, a social gradient for lung cancer incidence among those considered lower SES was identified; adjustments for smoking habits attenuated the gradient by up to 50% for men and 34% for women. Interestingly when evaluating lung cancer by subtype, there was much less of a gradient based on SES for adenocarcinoma (ADC) as compared to squamous cell carcinoma (SCC) and small cell lung cancer (SCLC). Adjusting for smoking history largely attenuated the risk gradient of lung cancer development across SES for ADC, a less exposure driven lung cancer, but either did not or only partially attenuated it for SCC and SCLC (2).

While the original 2018 SES SYNERGY analysis was thorough and successfully leveraged a large and detailed database to explore the association of SES with lung cancer incidence, there were a few limitations. One major limitation was the classification of SES into discrete categories; these categories were validated for other purposes, and they placed large numbers of cases and controls into one quartile or tertile. Among men, ~60% of cases and ~50% of controls fit into the 3^rd quartile for one measure and the 3^rd tertile for the other measure. There was also a concern about data clarity surrounding occupational details, namely categorization of occupations into high risk for lung cancer exposures, unemployment versus part-time work and early retirement, as well issues with being able to categorize SES for individuals who stayed home lifelong. Another concern surrounded misclassification of smoking data, especially for low intensity smoking. In addition, there were concerns around selection bias in the analysis with people in the lower SES category having lower response proportions. Lastly, there are numerous exposures for lung cancer outside of smoking and occupational exposures that were not captured in the data (2).

Given these limitations, the authors set about to refine their initial 2018 analysis to compensate for potential confounders and sources of bias, and either confirm or adjust their initial estimation of association between SES and the development of lung cancer. This new analysis sought to further correct for three areas of bias: (I) Misclassification of smoking status; (II) selection bias from low response rates among lower SES individuals, and (III) unmeasured confounding by genetic predisposition for lung cancer in lower SES individuals, namely CYP2A6 (cytochrome P450 2A6) polymorphisms. These biases were examined using sensitivity analyses with a wide range of plausible parameters. In addition, further analysis was conducted to tease out the effects of smoking for lower SES groups on lung cancer development. While the 2018 SYNERGY analysis controlled for smoking as well as other named variables, the authors planned to use a mediation analysis to better evaluate the true effect of smoking on lung cancer versus other variables with varying levels of interaction.

The mediation analysis was the most technically complex and useful addition in the current study (Figure 1). First the total effect (TE) was calculated by estimating the odds ratios for lung cancer risk for different quartiles of socioeconomic status, adjusting for age and study center, and split by gender. Then smoking related variables were added to the model to calculate the controlled direct effects (CDE) or the risk of SES on lung cancer development isolated from smoking habits. Then an inverse odds ratio weighting was conducted using a model of variables to identify the natural direct effects (NDE), or the risk of SES on lung cancer development if the smoking habits of the population was held constant to that of the top SES quartile of individuals on the study, which identifies exposures and risks other than smoking that lead to lung cancer development. The natural indirect effects (NIE), or the risk of lung cancer development that occurs due to increased smoking present with lower SES, can then be calculated using the TE and the NDE. This leaves the risk of developing lung cancer in lower SES from differences in smoking habits (NIE) versus other unmeasured risks and exposures (NDE) (3).

Figure 1 Mediation analysis diagram. Long dashed lines represent lung cancer risk mediated by non-smoking factors associated with socioeconomic status. Short dashed lines represent lung cancer risk mediated by smoking habits. CDE, controlled direct effects; NDE, natural direct effect; NIE, natural indirect effect; SES, socioeconomic status; TE, total effect.

With the bias analyses and the mediation analysis completed, the results were then combined into an analysis to evaluate the effect the biases had on the association of SES and lung cancer; instead of simple ranges, 2,500 brute force iterations were used to better identify the realistic outskirts of the parameters.

The results showed that the CDE and the NDE, which both represent risks of lung cancer development based on SES isolated from smoking, were very similar and confirm an increased risk of lung cancer in lower SES groups. Compared to the 1^st quartile for SES, the 4^th quartile CDE and NDE were 1.83 and 1.83 for men and 1.48 and 1.56 for women. Models with the addition of occupational data—ever working in a list A or B occupational category corresponding to levels of potential carcinogenic risk—only minimally changed NDE and NIE.

Among the three biases analyzed, selection bias demonstrated the largest potential effect on decreasing CDE and NDE with unmeasured genetic mediator effect and misclassification having smaller effects. After application of all three bias analyses, the OR for the 4^th quartile NDE decreased from 1.83 to 1.51 in men and from 1.56 to 1.27 in women with NIE increasing slightly from 1.31 to 1.34 and 0.98 to 1.02 respectively. After the final analysis, given the NIE OR was smaller than the NDE OR, this continued to indicate that more intense smoking habits among lower SES populations explained <50% of the increased risk for lung cancer. As a caveat, this conclusion holds among all quartiles for men but was only present for the 4^th quartile compared to the 1^st for women and not for either the 3^rd or 2^nd quartiles; this result is somewhat limited by the much smaller proportion of women included in the dataset—women represented 20% of the dataset—and the fact that the categorization of socioeconomic status was based on a measure developed from surveying men, limiting the applicability for female and especially older female cohorts.

This updated analysis by Hovanec et al. (3) is an excellent contribution to the study of SES and its association with lung cancer incidence. The initial analysis was based on a large multinational database, which collected extensive data longitudinally over decades especially surrounding smoking habits and occupational exposures, which are both known to be strong predictors of risk for lung cancer development. After the initial analysis, the authors asked the questions: (I) What could have been missed? (II) How could the data be further analyzed to compensate for confounders? This led to evaluating the three areas of potential bias and conducting a mediation analysis to evaluate the indirect and direct effects of smoking and unmeasured exposures on lung cancer development with more nuance. After these rigorous analyses, the underlying findings of the 2018 SYNERGY analysis were validated, namely that increased risk of developing lung cancer for individuals of lower SES is not solely due to differential smoking behaviors or occupational exposures.

The major limitation of both the 2018 and 2025 studies surrounds the nature of the data to analyze. While the database was extensive and biases were evaluated, certain areas of data were limited. As mentioned earlier, the categorization of individuals based on socioeconomic status utilized the International Socio-Economic Index of Occupational Status (ISEI), which used survey results of men working 30 hours or more per week in 16 countries from 1968 to 1982 that links education, occupation, and income to rank respondents on a scale of 10 to 90 (4). It is a strong surrogate for SES; however, its limitations are based on it evaluating occupations in male dominated fields while not considering household income. It also does not consider variations in incomes within an occupation and potentially across countries, inherited wealth, previous debt, the changing landscape of occupations over time, and scoring of ISEI using the job longest held versus the highest position held. Using this index may also overly associate socioeconomic position with occupation, and SES has other key contributing components that may impact health and healthcare access including education access and quality, neighborhood and built environment, and social and community context (5). Variations in national policy (taxation, social benefits, etc.) also contribute to SES and its impact on health.

As identified by the authors, there are a set of variables that were not captured in the data that make individuals of lower SES more prone to develop lung cancer. In the original 2018 analysis, accounting for smoking habits significantly reduced the increased risk of ADC in lower SES individuals, but variably decreased the risk of SCC and SCLC. Compared to SCC and SCLC, while risk of lung ADC is increased with increased smoking and particulate exposure, it is not as closely linked as SCC and SCLC (6). One potential confounder may be passive smoke inhalation, which is independently linked to lung cancer development; in Canada and many European countries smoking in the workplace and in enclosed public spaces did not start being banned until the early 2000s, after much of the data collection for the SYNERGY database was completed (7).

The pathogenesis of lung cancer is a complex interplay of factors, which can be exacerbated in lower SES populations, especially in the United States where healthcare access can be limited among low SES individuals. As mentioned, smoking, both by the individual (6) and passively (7), is known to increase lung cancer risk. Genetics of different populations also affect the independent risk of lung cancer development (8) and can modify the risk of smoking on lung cancer development (9). Exposures to carcinogens both in the workplace and at home increase the risk of lung cancer. Most notably, carcinogens exist such as indoor air pollution from cooking with poor ventilation (10), radon (11), asbestos (12), silica (13), NO2 (nitrogen dioxide) (14) and PM2.5 (particular matter ≤2.5 µm) (15) as measures of outdoor particulate contamination, benzene both from chemical plant emissions and natural emissions (16), pesticides (17), heavy metals such as arsenic (18), polyfluoroalkyl substances (19), among others. Outside of chemical exposures, conditions such as HIV (human immunodeficiency virus) (20) and previous lung disease (21,22) as well as previous ionizing radiation (23) increase lung cancer incidence.

Considering a broader more comprehensive measure of SES could be helpful, such as the area deprivation index (ADI) for United States populations, a composite of census level data of variables including education, occupation or unemployment, income, housing costs, and rates of households with health issues associated with low socioeconomic status. In a study in the United States evaluating lung cancer incidence with relation to ADI, it noted worsening of the ADI directly correlated with increasing lung cancer incidence. Additionally, while racial minorities were more likely to live in higher ADI areas, with adjustment by ADI, there was no increase in lung cancer incidence among racial or ethnic minorities (24). Similarly, a French version of the EDI (European deprivation index) identified an increased incidence of lung cancer in men and women as social class decreased (25). The EDI, ADI, and the SYNERGY SES analyses—the initial 2018 analysis and the additional validation of the 2025 bias and mediation analysis—indicate that lower SES and associated factors are clearly associated with lung cancer incidence independent of smoking. Future study is needed to elucidate specific key causative factors and develop interventions to address these. As lung cancer incidence is rising in never smoking women, causative factors by sex should also be further explored given the differential associations seen in this study (26). Further, attention to social equity, in general, may decrease the global burden of lung cancer.

Acknowledgments

None.

Footnote

Provenance and Peer Review: This article was commissioned by the editorial office, Annals of Cancer Epidemiology. The article has undergone external peer review.

Peer Review File: Available at https://ace.amegroups.com/article/view/10.21037/ace-2025-14/prf

Funding: None.

Conflicts of Interest: Both authors have completed the ICMJE uniform disclosure form (available at https://ace.amegroups.com/article/view/10.21037/ace-2025-14/coif). M.T. has received grants from the National Cancer Institute and National Institute on Minority Health and Health Disparities, and the Bristol Myers Squibb Foundation unrelated to the current work. M.T. serves as a paid consultant for an advisory board for GO2 Foundation. M.T. serves as an unpaid member for National Comprehensive Cancer Network Smoking Cessation Guidelines Panel. The other author has no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. SYNERGY. Available online: https://synergy.iarc.fr/. Accessed September 1, 2025.
2. Hovanec J, Siemiatycki J, Conway DI, et al. Lung cancer and socioeconomic status in a pooled analysis of case-control studies. PLoS One 2018;13:e0192999. [Crossref] [PubMed]
3. Hovanec J, Kendzia B, Olsson A, et al. Socioeconomic Status, Smoking, and Lung Cancer: Mediation and Bias Analysis in the SYNERGY Study. Epidemiology 2025;36:245-52. [Crossref] [PubMed]
4. Ganzeboom H, De Graaf P, Treiman D. A standard international socio-economic index of occupational status. Social Science Research 1992;21:1-56.
5. Healthy People 2030. Available online: https://odphp.health.gov/healthypeople. Accessed September 1, 2025.
6. Pesch B, Kendzia B, Gustavsson P, et al. Cigarette smoking and lung cancer--relative risk estimates for the major histological types from a pooled analysis of case-control studies. Int J Cancer 2012;131:1210-9. [Crossref] [PubMed]
7. Taylor R, Najafi F, Dobson A. Meta-analysis of studies of passive smoking and lung cancer: effects of study type and continent. Int J Epidemiol 2007;36:1048-59. [Crossref] [PubMed]
8. DeRouen MC, Canchola AJ, Thompson CA, et al. Incidence of Lung Cancer Among Never-Smoking Asian American, Native Hawaiian, and Pacific Islander Females. J Natl Cancer Inst 2022;114:78-86. [Crossref] [PubMed]
9. Stram DO, Park SL, Haiman CA, et al. Racial/Ethnic Differences in Lung Cancer Incidence in the Multiethnic Cohort Study: An Update. J Natl Cancer Inst 2019;111:811-9. [Crossref] [PubMed]
10. Lissowska J, Bardin-Mikolajczak A, Fletcher T, et al. Lung cancer and indoor pollution from heating and cooking with solid fuels: the IARC international multicentre case-control study in Eastern/Central Europe and the United Kingdom. Am J Epidemiol 2005;162:326-33. [Crossref] [PubMed]
11. Tirmarche M, Harrison JD, Laurier D, et al. ICRP Publication 115. Lung cancer risk from radon and progeny and statement on radon. Ann ICRP 2010;40:1-64.
12. Klebe S, Leigh J, Henderson DW, et al. Asbestos, Smoking and Lung Cancer: An Update. Int J Environ Res Public Health 2019;17.
13. Ge C, Peters S, Olsson A, et al. Respirable Crystalline Silica Exposure, Smoking, and Lung Cancer Subtype Risks. A Pooled Analysis of Case-Control Studies. Am J Respir Crit Care Med 2020;202:412-21.
14. Hamra GB, Laden F, Cohen AJ, et al. Lung Cancer and Exposure to Nitrogen Dioxide and Traffic: A Systematic Review and Meta-Analysis. Environ Health Perspect 2015;123:1107-12. [Crossref] [PubMed]
15. Huang F, Pan B, Wu J, et al. Relationship between exposure to PM2.5 and lung cancer incidence and mortality: A meta-analysis. Oncotarget 2017;8:43322-31. [Crossref] [PubMed]
16. Chiavarini M, Rosignoli P, Sorbara B, et al. Benzene Exposure and Lung Cancer Risk: A Systematic Review and Meta-Analysis of Human Studies. Int J Environ Res Public Health 2024;21.
17. Kim B, Park EY, Kim J, et al. Occupational Exposure to Pesticides and Lung Cancer Risk: A Propensity Score Analyses. Cancer Res Treat 2022;54:130-9. [Crossref] [PubMed]
18. Celik I, Gallicchio L, Boyd K, et al. Arsenic in drinking water and lung cancer: a systematic review. Environ Res 2008;108:48-55. [Crossref] [PubMed]
19. Li S, Oliva P, Zhang L, et al. Associations between per-and polyfluoroalkyl substances (PFAS) and county-level cancer incidence between 2016 and 2021 and incident cancer burden attributable to PFAS in drinking water in the United States. J Expo Sci Environ Epidemiol 2025;35:425-36. [Crossref] [PubMed]
20. Park LS, Tate JP, Sigel K, et al. Association of Viral Suppression With Lower AIDS-Defining and Non-AIDS-Defining Cancer Incidence in HIV-Infected Veterans: A Prospective Cohort Study. Ann Intern Med 2018;169:87-96. [Crossref] [PubMed]
21. Ang L, Ghosh P, Seow WJ. Association between previous lung diseases and lung cancer risk: a systematic review and meta-analysis. Carcinogenesis 2021;42:1461-74. [Crossref] [PubMed]
22. Xu H, Yin L, Bian W, et al. Interstitial Lung Disease and Risk of Lung Cancer. JAMA Netw Open 2025;8:e2519630. [Crossref] [PubMed]
23. Egawa H, Furukawa K, Preston D, et al. Radiation and smoking effects on lung cancer incidence by histological types among atomic bomb survivors. Radiat Res 2012;178:191-201. [Crossref] [PubMed]
24. Adie Y, Kats DJ, Tlimat A, et al. Neighborhood Disadvantage and Lung Cancer Incidence in Ever-Smokers at a Safety Net Health-Care System: A Retrospective Study. Chest 2020;157:1021-9. [Crossref] [PubMed]
25. Bryere J, Dejardin O, Launay L, et al. Socioeconomic status and site-specific cancer incidence, a Bayesian approach in a French Cancer Registries Network study. Eur J Cancer Prev 2018;27:391-8. [Crossref] [PubMed]
26. Triplette M, Aldrich MC. Lung Cancer in Special Populations. Semin Respir Crit Care Med 2025;46:479-89. [Crossref] [PubMed]