Center for Asbestos Safety

Education and Risk of Cancer

In several different countries, low socioeconomic status (SES) has been linked with a higher risk of morbidity and mortality in varying age groups. It has been shown that education, which is an indicator of socioeconomic status, is inversely related to the occurrence of cancer at a number of (but not all) anatomic sites, which in general implies that higher the educational attainment level, the less is the cancer risk.

Several different types of demographic, behavioral and biological factors including, energy balance, smoking, cancer screening, age at first birth, and hormone use, potentially determine the casual association between education and cancer. It has been revealed in recent studies that inflammation biomarkers, most likely casual in relation to cancer and overall mortality, are linked inversely with education. It has been shown that multivariate adjustment for “unhealthy” behaviors completely eliminates the link between education and occurrence of cancer. While analytic maneuvers like these can potentially explain the association between education and cancer, they do not counteract its public health importance.

In the potentially large cohort of men and women (aged 50 to 71) in the United States, significant inverse education gradients exist for incident cancer. An increased risk among the least educated in comparison to the most educated has been revealed in fully adjusted models, especially in case of combined smoking-related cancers (including those of the head and neck, lung, esophagus, pancreas and bladder). Moreover, inverse education gradients have also been identified to exist for stomach and rectum cancer (men only), and colon (women only). Some direct relationships have also been established between education and cancer risk, especially in case of certain types of cancers such as melanoma of the skin (both men and women), invasive breast and endometrial cancer (women), and localized prostate cancer (men).

A large prospective cohort, the NIH-AARP cohort has detailed information on several different types of covariates, which allows better control for multiple risk factors at individual levels during the analysis of first primary rare and other types of malignancies occurring both in men and women. Similar results have been achieved in other prospective studies carried out in Europe. However, in these studies, risk factors were not controlled, only common cancers were analyzed, or data presented was only for women. Results were also similar in other studies conducted in the United States on the association between education and cancer mortality. The availability of incidence data (registry-based) in the cohort directed the analysis on probable cancer causation, significantly circumventing the complicating influence that treatment factors have on cancer mortality outcomes.

The smoking-adjusted analysis has proved helpful in two different ways. Firstly, in case of specific sites, predominantly a combination of lung and smoking-related cancers, adjustment for smoking results in significant attenuation of the inverse association between education and cancer risks in both men and women. Considering the fact that smoking is evidently linked to education and that it is a well-known cause of these types of cancers, this relative attenuation of risk indicates strongly that smoking is the main intermediate factor influencing the association pathway between education and cancer. Secondly, while the smoking adjustment attenuates the education-cancer relative risks, they do not slip back to the null. Even with the smoking adjustment, risks related to cancers of the lung, esophagus, and other smoking-related cancers remains approximately doubled for the least educated in comparison to the most educated men. This may indicate smoking-related residual confounding or the presence of non-smoking related casual factors (either psycho-social or biological) influencing the education-cancer pathway. Given that education, even when other factors such as smoking have been accounted for, should almost always predict, for instance, the potential development of esophageal cancer in men, has proved a tantalizing scenario and has become a target for etiologic research.

After adjusting for age and smoking, the addition of other covariates in the regression analysis models resulted in little additional attenuation in the association between education and cancer. While residual confounding for these imperfectly measured variables such as total energy intake, physical activity, and alcohol consumption cannot be denied, these other factors explain relatively little about the link between education and cancer.

Information on H. pylori infection status was unavailable for incorporation in the multivariate analyses of gastric cancer. However, in a large nested case control study conducted in Europe, investigations by Nagel et al have revealed the inverse association of gastric cancer, although non-significant, even when the H. pylori was controlled.

A significant inverse gradient for pleural cancer in men has been revealed through collected data. This finding derived from a prospective cohort study, potentially due to the large size of the study, appears unaffected by smoking and may indicate occupational or environmental exposure to asbestos amongst study participants. It is important to note that asbestos use was widespread in the U.S. until the Occupational Safety and Health Administration (OSHA) regulations were implemented in 1971. Since study participants at that time were approximately aged around 26-47 years, it is possible that they may have undergone occupational or environmental exposure.


Center for Asbestos
Safety in the Workplace