Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: Is it valid to use the individual ratios (i.e. Xi/Yi) in the dependent or independent part of a regression model?
From
Phil Clayton <[email protected]>
To
[email protected]
Subject
Re: st: Is it valid to use the individual ratios (i.e. Xi/Yi) in the dependent or independent part of a regression model?
Date
Mon, 28 May 2012 06:30:14 +1000
I'm not convinced. The first paper you cited measured 24-h urine concentration, not spot urine, and the second paper presents no data directly supporting its argument.
There is also (in my opinion) an argument for using the same methodology as others in the field, and most people would use an ACR. Why don't you try modelling it in different ways and see whether your conclusions are sensitive to the choice of model?
Phil
On 28/05/2012, at 2:17 AM, [email protected] wrote:
> In general, your comments are correct (esp. for an individual
> patient). However, low urinary concentration is an independent
> predictor of CKD progression
> (http://www.ncbi.nlm.nih.gov/pubmed/12722030?dopt=Abstract) and there
> has been some suggestions that the analyte concentration (unadjusted
> for creatinine) should be included in the analysis with urinary
> creatinine added as a separate independent variable in a population of
> groups (http://www.ncbi.nlm.nih.gov/pubmed/15687057).
>
> Jinn-Yuh
>
>
> 2012/5/27 Phil Clayton <[email protected]>:
>> On 27/05/2012, at 9:35 PM, David Hoaglin wrote:
>>
>>> As an explanatory variable, ACR is one function
>>> of urinary albumin and urinary creatinine; but you could reasonably
>>> consider other functions, such as the linear combination of urinary
>>> albumin and urinary creatinine that arises from using those two as
>>> explanatory variables or the nonlinear function in which the
>>> explanatory variables in that part of the model are urinary albumin,
>>> urinary creatinine, and their product
>>
>> What we're interested in, biologically, is the 24-h urinary albumin excretion.
>>
>> The reason the albumin is divided by creatinine is that the ACR is used to estimate the 24-h urinary albumin excretion from a single urine specimen rather than asking someone to collect their urine for a day. The urinary concentration can vary several fold (eg if you're dehydrated it goes up) which changes the albumin concentration in that specimen - but it changes the creatinine concentration by a similar amount, and we know the normal 24-h excretion of creatinine, so we divide the albumin by the creatinine to estimate the 24-h albumin excretion.
>>
>> So biologically ACR should be a ratio and must be nonnegative. You could include it in a regression model as a surrogate for the 24-h urine albumin excretion, but would need to careful how to model it as it generally has a nonlinear effect. For example it is commonly modelled as a categorical variable - normal, microalbuminuria, macroalbuminuria.
>>
>> Phil
>> *
>> * For searches and help try:
>> * http://www.stata.com/help.cgi?search
>> * http://www.stata.com/support/statalist/faq
>> * http://www.ats.ucla.edu/stat/stata/
>
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/