2008-07-12 09:25
(5560 d 10:18 ago)

Posting: # 2018
Views: 6,158

 Serum cortisol AUC [General Sta­tis­tics]

Dear friends,
plz tell me how to calculate serum cortisol AUC.
I have data with columns: subjectid, time, serum cortisol at pre treat, serum cortisol at end of treat.
thanks and regards,

Edit: Category changed. [Helmut]

2008-07-14 14:29
(5558 d 05:15 ago)

@ Yayati
Posting: # 2024
Views: 4,962

 Serum cortisol AUC

Dear Yayati,

the AUC is the area under the curve and can therefore be estimated by any integration method you like regardless if you have serum cortisol or monthly income.
The most common one is the trapezoidal rule (which is based on linear interpolation) but you can also use log-linear interpolation, estimation of a polynomous function with arbitrary order (but less or equal to the number of time points) or some other function which make sense in your mind. Several textbooks give good explanations and examples on this toipc.
However, take care of any kind of pre-treatment adjustment, otherwise your AUC is biased.


2008-07-14 17:34
(5558 d 02:09 ago)

Posting: # 2026
Views: 4,955

 Serum cortisol AUC

Dear JPL and Yayati!

❝ However, take care of any kind of pre-treatment adjustment, otherwise your

❝ AUC is biased.

Yes, you are right. But Yayati in his original post stated only two sampling points (pre- and post treatment). IMHO in such a case AUC is meaningless. I would only compare pre- and post-treatment - and not even think about calculating an 'AUC'. In the case of cortisol a baseline correction is debatable; maybe a transformation of the post-treatment data into a percentage of pre-treatment makes sense. For a discussion of baseline values see Senn (Cross-over Trials in Clinical Research, Section 3.16, Wiley, 2nd ed. 2006).

It's not considered good statistical practice to gather data and seeking a method for evaluation afterwards. We should invest the better part in planning.

Regards, Jaime

2008-07-16 07:36
(5556 d 12:07 ago)

@ Yayati
Posting: # 2040
Views: 4,863

 distribution of Serum cortisol AUC

Dear friends,
Serum cortisol AUC follows normal disrtibution or log-normal distribution?
How to calculate change from baseline serum cortisol AUC?

Edit: Linked to previous thread. [Helmut]
UA Flag
 Admin contact
22,764 posts in 4,776 threads, 1,628 registered users;
19 visitors (0 registered, 19 guests [including 5 identified bots]).
Forum time: 19:44 CEST (Europe/Vienna)

The object of statistics is information.
The objective of statistics is the understanding of information
contained in data.    Irwin and Marylees Miller

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz