Heterogeneity and Data Analysis: Heterogeneity #2, Mixture of types

Heterogeneity #2, Mixture of types

If the population in question contains a mix of different types that are identified and separable, then identification of a subject’s type allows them to be treated or investigated separately and differently.

• Phenylketonuria (PKU) is a condition associated with a single genetic locus, but there are different mutations within that locus. People with PKU are a mix of people with different mutations or genetic sub-types.

• “Maternal PKU” arises when a fetus gestates in high-phenylalanine conditions that occur when a PKU mother was not insufficiently compliant with the diet (where compliance may be influenced by a variety of factors).  If maternal PKU is considered a form of PKU, then the population of children with the deleterious symptoms of PKU is a mix of those with a genetic condition that was not followed by the special diet and those without the genetic condition who have.

• A certain population of light-eyed, yellow rats consisted of two strains, each bred separately from some ancestral founding group. The “two strains of light-eyed, yellow rats, each of which bred true by itself… produced nothing but black-eyed rats when crossed with each other” (Wright 1920, 37). (If each strain was bred in the same, uniform laboratory conditions, this would seem to be a case of different kinds of genetic factors producing light-eyes for the two strains in those conditions.)

• The protective effect with respect to heart disease and stroke of taking a daily low dose of aspirin differs on average for men and women. This means the human population can be treated as a mix of female and male types with respect to the protective effects of aspirin. However, 9% of the patients in one study appeared “resistant,” i.e., their samples didn’t show the typical blood-thinning effects of aspirin, meaning that the male and female types are heterogeneous with respect to resistance.  In practice, these sub-types are costly to identify and this is rarely done (Eikelboom 2003).  Note: It is also possible that the heterogeneous factors underlying the sub-types overlap, that is, the male and female types (which were based on differences on average) can eventually be resolved into a number of types not unique to males or females.

Heterogeneity #2, Mixture of types -> #6, Possibility of “underlying heterogeneity,” and vice versa.

• In genetics, homogeneity may be on the surface only, e.g., when it is discovered that different genetic conditions are expressed as the “same” clinical entity.  Conversely, the clinical expression of mutations at a single genetic locus can vary significantly from one person to the next (Kaplan 2000, 18). This may be because the mutations are at different points within the locus or because the same genetic condition develops in different genetic and environmental contexts, i.e., the other genetic and environmental factors vary among the people.

• In medical sociology Brown and Harris (1989) often find common meaning among subjects’ different types of experience.  In other words, Brown and Harris code sameness despite surface heterogeneity.

(continuing a series of posts—see first post; see next post)


One thought on “Heterogeneity and Data Analysis: Heterogeneity #2, Mixture of types

  1. Pingback: Heterogeneity and Data Analysis: Heterogeneity #1, An assortment « Intersecting Processes

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s