[R] Calculating a P for trend

[Lisa van der Burgh]

Hi all,



I have a question about calculating a P for trend on my data. Let’s give an example that is similar to my own situation first: I have a continuous outcome, namely BMI. I want to investigate the effect of a specific medicine, let’s call it MedA on BMI. MedA is a variable that is categorical, coded as yes/no use of the medication. A also have the duration of use of the MedA, divided in three categories: use of MedA for 1-30 days, use of MedA for 31-60 days and use of MedA for 61-120 days (categories based on literature). I have performed a linear regression analyses and it seems like there is some kind of trend: the longer the use of MedA, the higher the BMI will be (the betas increase with time of use). So an exemplary table:






Outcome: BMI


Beta


MedA use duration





Use for 1-30 days


0.060


Use for 31-60 days


0.074


Use for 61-120 da


0.081




So, I have created three variables and I modelled them in Rstudio (on a multiple imputed dataset using MICE):



mod1 <- with(imp, lm(BMI ~ MedA_1to30))

pool_ mod1 <- pool(mod1)

summary(pool_ mod1, conf.int = TRUE)



mod2 <- with(imp, lm(BMI ~ MedA_31to60))

pool_ mod2 <- pool(mod2)

summary(pool_ mod2, conf.int = TRUE)



mod3 <- with(imp, lm(BMI ~ MedA_61to120))

pool_ mod3 <- pool(mod3)

summary(pool_ mod3, conf.int = TRUE)



Now that I have done this, I want to calculate a p for trend. I do know what a P for trend measures, but I do not know how to calculate this myself. I read something about the partial.cor.trend.test() function from the trend package, but I do not know what I should fill in. Because I can only fill in an x and y, but I have three time variables. So I do not know how to solve this. Can somebody help me?



If more information is necessary, I am happy to give it to you!


[[alternative HTML version deleted]]

[Bert Gunter]

I would suggest that if at all possible, you find a local statistician
(your instructor??) with whom to consult. Much of what you are doing
appears likely to result in irreproducible nonsense.

This list is concerned with R programming, not statistics, although they
sometimes do intersect. So I think your post is off topic here (but others
may disagree).

Cheers,
Bert


Bert Gunter

"The trouble with having an open mind is that people keep coming along and
sticking things into it."
-- Opus (aka Berkeley Breathed in his "Bloom County" comic strip )

Post by Lisa van der Burgh
Hi all,
I have a question about calculating a P for trend on my data. Let’s give
an example that is similar to my own situation first: I have a continuous
outcome, namely BMI. I want to investigate the effect of a specific
medicine, let’s call it MedA on BMI. MedA is a variable that is
categorical, coded as yes/no use of the medication. A also have the
duration of use of the MedA, divided in three categories: use of MedA for
1-30 days, use of MedA for 31-60 days and use of MedA for 61-120 days
(categories based on literature). I have performed a linear regression
analyses and it seems like there is some kind of trend: the longer the use
of MedA, the higher the BMI will be (the betas increase with time of use).
Outcome: BMI
Beta
MedA use duration
Use for 1-30 days
0.060
Use for 31-60 days
0.074
Use for 61-120 da
0.081
So, I have created three variables and I modelled them in Rstudio (on a
mod1 <- with(imp, lm(BMI ~ MedA_1to30))
pool_ mod1 <- pool(mod1)
summary(pool_ mod1, conf.int = TRUE)
mod2 <- with(imp, lm(BMI ~ MedA_31to60))
pool_ mod2 <- pool(mod2)
summary(pool_ mod2, conf.int = TRUE)
mod3 <- with(imp, lm(BMI ~ MedA_61to120))
pool_ mod3 <- pool(mod3)
summary(pool_ mod3, conf.int = TRUE)
Now that I have done this, I want to calculate a p for trend. I do know
what a P for trend measures, but I do not know how to calculate this
myself. I read something about the partial.cor.trend.test() function from
the trend package, but I do not know what I should fill in. Because I can
only fill in an x and y, but I have three time variables. So I do not know
how to solve this. Can somebody help me?
If more information is necessary, I am happy to give it to you!
[[alternative HTML version deleted]]
______________________________________________
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

[[alternative HTML version deleted]]

______________________________________________
R-***@r-project.org mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.