In most research, two indicators are used to
compare the level of longevity and particularly
extreme longevity in a given population. These
indicators are the life expectancy and the
prevalence in terms of number of centenarians
in proportion of the total population. These
two indicators are in fact inadequate as both
are biased by age structure distortion effects
and migration flows impacts. The life
expectancy is calculated by using age at death
in the population during a given period.
Surviving probabilities at old ages or life
expectancy calculated on a cohort basis are
definitively better indicators in order to
compare extreme longevity. As far as the

prevalence of centenarians is concerned, this
indicator is clearly influenced by the past
history of the concerned population in term of
fecundity, mortality and migration. As an
example, there is a huge risk of overestimating the level of longevity for a population through high
inflows of old persons and high outflows of
young persons. This short contribution will propose more appropriate indicators in order to compare more correctly extreme longevity and apply these indicators to some European populations.