Adjust age as confounding factor I have a continuous response variable (concentration) and a categorical explanatory variable (healthy/ill), and probably two confounding factors: age (continuous) and gender. 
What would you recommend me to adjust both confounding factors?
BTW, I am using R for it, which package/function should I use?
 A: Since concentration is presumably a proportion value, between zero and one, a standard linear regression is probably inappropriate here (unless the outcome values are very tightly packed away from the boundaries).  For this kind of response variable, it is common to use a transformation-based regression model such as logistic regression or beta regression.  The details of these methods are beyond the scope of your question, which pertains to the adjustment for confounding factors.  For this issue, you need to decide whether you just want to add terms for baseline effect for these factors, like this:
~ age + gender + healthy

or whether you want to also include interaction terms, like one of these:
~ age + gender + age:gender + healthy

~ age*gender + healthy

~ age*gender*healthy

The last case is the most general form, since it includes interactions between all of the explanatory variables.  This will give you a more complicated interpretation, but it will allow differences in th effect of the healthy variable for people of different age and gender.
A: As Daniel mentioned in the comments above, a linear regression model might be most appropriate.
However, when entering a linear model into R, it's important to note that the sums of squares used are sequential. That is, you should put age and gender into your model before your covariate of interest.
model <- lm(concentration ~ age + gender + healthy, data = mydf)

Include any interactions, etc. as you seem appropriate.
