The study of cancer in specific populations can offer clues useful in determining the extrinsic and intrinsic factors influencing cancer in all populations. Extrinsic factors are sometimes called "environmental" in the broadest sense of the word. They are modifiable or mutable. Intrinsic factors are more inherent to the individual. They are almost always genetic and are immutable or unchangeable. Targeting research on specific populations is and should be a significant ethical issue.
Incidence and cancer outcomes vary among populations, no matter how the populations are defined. The American medical literature commonly categorizes populations by race or ethnicity. A few studies have even categorized populations by socioeconomic status, area of geographic origin, or religion. Since 1994, the National Institutes of Health (NIH) has mandated that virtually all NIH-funded clinical research be conducted in such a way that "differences amongst the races and ethnicities can be determined," and it urges researchers to publish results by race, ethnicity, and gender.
It is unfortunate that we in medicine rarely think in rigorous fashion about how we divide and describe populations and what these divisions might imply. Very often, epidemiologic and clinical research using population categorizations will demonstrate weak correlates with extrinsic influences on biology. When conducted using racial and ethnic categorizations, the findings can easily be mistaken as surrogates for, or evidence of intrinsic influences.
While it is true that the incidence and mortality of disease often vary by race and ethnicity, neither is a biological categorization. Despite common belief, race does not equate with genetic homogeneity; rather, it is a sociopolitical construct. Racial data, as such, often correlate with many sociopolitical factors that influence health. Ethnicity, while also sociopolitical, is somewhat more scientific, as it can equate with culture and other environmental influences on health status.
The burgeoning fields of genetic and molecular epidemiology make it imperative that population categorization be better understood. Single-nucleotide polymorphisms, other genetic polymorphisms, and mutations can have varying prevalences or frequencies among populations, however they are defined. These differences in risk best relate to familial inheritance.
Indeed, virtually every genetic difference or polymorphism that has been correlated or associated with race should be considered familial and not racial. A specific gene or series of genes can be conserved among families. Family as a categorization clearly preserves genetic markers more efficiently than race or ethnicity. A complicating factor is that race and ethnicity are also commonly preserved by family.