risk factor

[GoogleCodeExporter]

Hi,

The InfluenzO ontology tried to define risk factor, for example "older age
is a risk factor for heart disease". I believe this should probably live in
OGMS.

Looking at an online example -
http://www.webmd.com/heart-disease/risk-factors-heart-disease
What Are the Risk Factors for Heart Disease?

There are several risk factors for heart disease; some are controllable,
others are not. Uncontrollable risk factors include:

    * Male sex
    * Older age
    * Family history of heart disease
    * Post-menopausal
    * Race (African Americans, American Indians, and Mexican Americans are
more likely to have heart disease than Caucasians)

Controllable risk factors include:

    * Smoking.
    * High LDL, or "bad" cholesterol and low HDL, or "good" cholesterol.
    * Uncontrolled hypertension (high blood pressure).
    * Physical inactivity.
    * Obesity (more than 20% over one's ideal body weight).
    * Uncontrolled diabetes.
    * High C-reactive protein.
    * Uncontrolled stress and anger.

Trying to sort out those "factors":
Male, older age, race sound like qualities. Post-menopausal is a
disposition, while family history is probably a genetic predisposition.
Smoking, inactivity are probably processes, hypertension, stress, anger,
high cholesterol, diabetes would be disposition.

It sounds like "risk factor" should be a defined class, risk factor for
disorder of type X similar to "a variable increasing the risk of an
organism developing the disorder X"
I am unhappy with the use of the word variable in this definition, but
don't have a better suggestion - as per the examples above this "variable"
can be a number of things.

It resembles the current ogms:predisposition class (predisposition to
disease of type X defined as "A disposition in an organism that constitutes
an increased risk of the organism's subsequently developing the disease X.") 

Do you think it would be adequate to say that a risk factor can cause a
predisposition? For example, high cholesterol, by clogging somebody's
arteries, causes a predisposition to heart attack.

Any suggestions/comments on how to address this?

Thanks,
Melanie

Original issue reported on code.google.com by mcour...@gmail.com on 9 Dec 2009 at 7:14

[GoogleCodeExporter]

My general suggestion is to not try to account for terms such as 'risk factor' 
(which are purely in the realm of 
statistics) in an ontology. 

Original comment by chime...@gmail.com on 9 Dec 2009 at 10:59

[GoogleCodeExporter]

is statistics outside the realm of ontology?

Original comment by alanruttenberg@gmail.com on 9 Dec 2009 at 11:06

[GoogleCodeExporter]

I believe so.  Or at the very least it is outside the realm of a pragmatic use 
of ontology.  I wonder what the use 
case would be that motivates needing to explicitly represent a risk factor, 
rather than - for example - 
capturing a clinical guideline as an axiom involving the risk factors and their 
(statistically significant) 
manifestations such as:

Class: PatientWithHistoryOfHypertension
SubClassOf: PatientThatShouldMonitorBloodPressure

Where the semantics of the underlying language captures the relationship, 
rather than some explicit 
designation of PatientWithHistoryOfHypertension as a risk factor.

Original comment by chime...@gmail.com on 9 Dec 2009 at 11:18

[GoogleCodeExporter]

I think that 'risk factor' finds its way outside of statistical analyses enough 
to at
least be considered for ogms.  However, like 'sign' and 'symptom', 'risk 
factor' is
one of those terms that is bound to span several BFO types, and, as such, will 
be
really difficult to provide a useful ontological definition for.

I'm starting to think that the proper place for such things is as relations:

S sign-of D
S symptom-of D
R risk-factor-of D

signs, symptoms, and risk factors are inherently relational in nature (i.e., 
relating
some entity to a disease). Smoking is a process in a more fundamental way than 
it is
a risk factor. Smoking is only a risk factor only in relation to (say) heart 
disease.

Original comment by albertgo...@gmail.com on 9 Dec 2009 at 11:27

[GoogleCodeExporter]

I wonder if risk factors should be treated in a manner similar to dependent and 
independent variable in IAO

Original comment by alanruttenberg@gmail.com on 11 Dec 2009 at 4:36