Extreme Population-level Events: Do They Have an Impact on Cancer?

The paper by Sadetzki et al., reported in this issue¹, showcases the impact of a horrendous event – the Holocaust – on subsequent health, decades after the occurrence. This study not only brings to light additional effects of the Holocaust in the Jewish population, but brings to mind similarities in cancer development and extreme population-level tragedies. We highlight two additional examples of calamities – country-wide famines and population-level discrimination – as examples of this phenomenon to add evidence to the causal pathway identified by Sadetzki et al.

First, however, important factors to consider are the “event” itself – is it a constant event (e.g., discrimination) or episodic/defined in time (e.g., a famine). The time point at which the event occurred in the life-course is also important; for example, events may occur early in life, late in life, or continuously. The type of event is also important; does it have a physical nature, such as a famine, or an emotional factor, such as discrimination, or both, as occurred in the Holocaust? The focus of the event is also important: did it impact the entire population, as in a famine, or emphasize a specific group, as the Jews during the Holocaust? Finally, was the duration episodic (e.g., a few months or years) or continuous (over the lifetime)? Individuals can be impacted differently by the same event due to many of these factors, as well as genetic “heartiness.”

Cancer is a dynamic process that increasingly is thought to be linked to the life-course; that is, exposures to carcinogens during one stage of life may influence cancer risk at a subsequent stage^2–4. One area that has been extensively studied in animal models is calorie deprivation^5,6. In general, such studies indicate that animals who experience caloric restrictions live longer and have fewer diseases than animals fed higher calorie diets^5,7. There also is some evidence that this is true in humans. Patients with eating disorders that involve calorie restriction have reduced risk of subsequent breast cancer⁸. Ecological studies note that individuals who live in countries where fewer calories are consumed have lower death rates from cancer and other chronic diseases⁹.

Exposures during the life-course, however, may not only be due to the choices that individuals make, but also may be due to external factors. War and its consequences are such external factors. Unlike the animal studies and some limited human research, war provides examples of real-life situations that can be examined for later consequences. World War II is an example of a time when exposures around caloric intake varied by country. In Europe, countries occupied by Germany included Norway and the Netherlands, and both have examined dietary intake with subsequent cancer outcomes. In Norway, occupied by Germany from 1940 to 1945, food restrictions were enacted early in the occupation. A 2002 analysis of Norwegian women who lived in food-producing areas compared to non-food-producing areas during World War II found decreased incidence of breast cancer among women who lived in the non-food areas, lending some credence to the suggestion that caloric restriction is protective against some cancers¹⁰.

The Dutch experience indicates the obverse. In a short, but severe famine that lasted approximately eight months (September, 1944 – April, 1945), the Dutch faced a cut in daily adult food rations from 1500 kilocalories to 700 kilocalories a day. Examination of a large cohort of Dutch women who lived during the famine began in 1974. Women born between 1911 and 1945 were eligible to participate in the project that called for the early detection of breast cancer¹¹. Women were surveyed as to their personal experiences during the Dutch famine. Subsequently, the cohort was divided into women who were severely exposed, moderately exposed, or unexposed. Using a case-cohort design of approximately 15% of the total cohort (n = 2352), the authors found a significant increase in breast cancer risk for all women who were severely exposed (4.05 per 1000 person-years) or moderately exposed (2.80 per 1000 person years) compared to women who were not exposed (2.48 per 100 person years). Moderately exposed women had a 13% (hazard ratio [HR] =1.13, 95% CI = 0.92 to 1.38) higher risk of breast cancer and severely exposed women had a 48% (HR = 1.48, 95% CI = 1.09 to 2.01) higher risk of breast cancer than women not exposed. Further analysis indicated that those women who were older during the famine and who were severely exposed had the highest rates of breast cancer¹². This may have been because rations were distributed on the basis of age so that young people could be protected. Other cancers were not reported.

In the Sadetzki et al paper in this issue, exposure to the horrific stressors of the Holocaust is associated with increased risk of subsequent cancers, although there was not an association with breast cancer¹. The stressors of Holocaust survivors went far beyond those of victims of the Norway food restrictions and those that suffered from the Dutch famine. Stressors were emotional and physical and the duration of the abuses was both long and continuous. Such stressors are very likely to have an impact on subsequent disease outcomes such as cancer. The combination of emotional and physical insults was far greater than those experienced by other victims of environmental exposures; yet, there are some similarities between those stressors and the ones experienced by underserved groups today.

The associations that Sadetzki and colleagues found between the extreme deprivation experienced by Holocaust survivors and cancer also may have parallels with racial minority groups in the United States who experience serious social deprivation over time. Although the size of this group is impossible to estimate because they remain largely invisible to society, it is a subset of the 43.1 million persons who live below the federal poverty line (e.g., those families of four who earnings of $24,257 or less in 2015)¹³. This group includes over 1.5 million families whom Edin and Schaeffer identify as living on less than $2.00 per day and food stamps¹⁴. Racial minority groups are much more likely to experience extreme poverty compared to non-Hispanic whites¹⁵.

Lutgendorf and Sood propose pathways through which biobehavioral risk factors like poverty and racial discrimination affect cancer progression¹⁶. In a 2016 paper, they and co-authors suggested that environmental stressors trigger norepinephrine or epinephrine and cortisol, which profoundly dampen immune responses and perturb homeostasis to accelerate cancer progression and increase the risk of other diseases¹⁷. Studies using animal models support these associations. Hermes et al., for example, produced psychosocial stressors in Sprague-Dawley rats by manipulating their living conditions, increasing the tumor burden of these animals to 84 times of control animals, with a 3.3 relative risk of ductal carcinoma in situ and invasive ductal carcinoma, compared to control animals¹⁸.

Racial and ethnic minorities in the United States are disproportionately exposed to social stressors like neighborhood violence^19,20. Goldmann et al. found 87.2% of a randomly-selected sample of group of African- American in Detroit had experienced at least one traumatic event such as assaultive violence during their lifetimes²¹. Gehlert and colleagues found that African-American breast cancer survivors living in predominantly African-American neighborhoods in Chicago exhibited a constellation of psychosocial characteristics including high rates of sexual abuse during childhood and adolescence, felt loneliness, a psychological manifestation of social isolation, and depression compared to other survivors^22,23. Over 31% of their sample of survivors reported sexual abuse during childhood or adolescence. Recently, Conroy and colleagues found an association between education and neighborhood socioeconomic status (nSES) and odds of breast cancer, with women living in the highest versus lowest nSES quintile having a two-fold greater odds of breast cancer²⁴. Work by Theall et al. in which they found an association between exposure to neighborhood violence and physiological and cellular markers of stress²⁵, suggests a link between neighborhood stressors supports the pathways proposed by Lutgendorf and Sood^16,17.

Sadetzki and colleagues’ findings suggest an association between the dreadful experiences of Holocaust survivors and cancer incidence. We see a parallel with the disproportionate exposure to stressors at the neighborhood and community level experienced by some racial minorities in the United States. Although the body of knowledge to support this link has grown in the past decade^21–25, additional research is needed from teams of social, behavioral, and biological scientists working to untangle these complex phenomena.

What is the message for future researchers? Although a 2009 epidemiological study failed to find a an independent association between stressful life events and cancer among women enrolled in the Women’s Health Initiative, its authors noted the compatibility of their results with a model of interacting psychosocial factors in breast cancer risk²⁶. Since that time, models have been proposed in which factors at various levels of influence interact to contribute to cancer and other disease and their progression. Investigators from the Center for Population Health and Health Disparities, for example, developed a conceptual framework for cancer disparities, with influences at the proximal (biological and genetic factors such as ancestry), intermediate (immediate social and physical contexts and social relationship in which the distal effects are experience, such as neighborhoods and communities), and proximal levels (population-level social conditions such as variation in rates of disease or poverty), interacting in complex ways to produce cancer disparities²⁷. Transdisciplinary team science research has the potential to produce multi-level interventions to address the root causes of cancer disparities and stem the growing tide of cancer disparities in the United States. It is incumbent on future researchers to recognize the complexity of disease outcome. Simple associations in a time-limited study may only be part of an intricate constellation of factors that affect disease outcomes. Recognition of such complexity is warranted in order to truly understand and reduce risks of premature death and suffering from diseases like cancer.