trying to run tvc with stcox for age##age - problems with collineratity

[Denise Vella]

hello i am trying to identify my tvc factors . to improve my first model i had to add this to age: age##age with a final model looking like this (section a)

However, when i try to find tvc - if i use ## (section b) i get problems with collinearity, with a list of all the ages from 20-90...this takes a long time with a 180,000 dataset

is tvc unable to tolerate age##age?

how can i work around it ?

///section a
stcox c.pain i.drug c.age##c.age, hr 

///trying to find tvc

///problems with colinearity - section b 
stcox pain drug age##age, tvc(pain drug age##age) texp(ln(_t)) hr 


///no problems
stcox pain drug age, tvc(pain drug age) texp(ln(_t)) hr