Methods. The changes in patterns of smoking initiation in Italian male
s and females in 1955, 1965, 1975, and 1985 were analyzed using data f
rom the 1990-1991 Italian National Health Survey, based on a sample of
18,483 males and 19,320 females ages 19 to 59. The sample was randoml
y selected within strata of geographic area and size of place of resid
ence and of household in order to be representative of the general Ita
lian population. The age of each subject was reconstructed for each ca
lendar year of interest (1954 to 1956, centered on 1955, 1964 to 1966,
1974 to 1976, and 1984 to 1986). For each calendar period and year of
age of interest (14 to 24), the numerator of the smoking initiation r
ate was the number of subjects who started smoking and the denominator
was the number of subjects at risk, i.e., nonsmokers of the same age.
The same procedure was repeated for three separate age groups (14 to
17, 18 to 20, and 21 to 24). Results. Among males, there was a trend t
oward earlier age at start of smoking and higher initiation rates betw
een 1955 and 1965, and a subsequent decline in initiation rate, partic
ularly at age 18 or over and during the last calendar decade studied.
In females, between 1955 and 1975 there was an increase of over fourfo
ld in initiation rates and systematic tendencies toward earlier age at
starting; only over the past decade have the starting rates begun to
decline. The peak rate was reached at age 18-20 for males in 1965 (53.
3%) and for females in 1975 (23.9%). Conclusion. The recent declines i
n smoking initiation have been relatively limited among the younger ag
e groups, and this reflects an earlier average age at smoking initiati
on. (C) 1995 Academic Press, Inc.