These results should be interpreted with certain limitations in
mind. One obvious limitation was the low number of CHD events,
especially in those receiving LLT, which affected the precision of the
hazard ratio estimates. Moreover, information on the duration and
titration of the treatment was unavailable at baseline examination.
The allocation of LLT was not controlled randomly and may have
been subject to an indication bias. Finally, the generalizability of the current results to non-Caucasian populations and to frailer elderly
subjects needs to be evaluated