Aims and background: The aim of the study was to evaluate mortality lung cancer trends, as an indicator of female smoking trends, in women resident in different urbanization areas.
Methods and study design: Data on the 5782 female lung cancer deaths that occurred in Tuscany, Italy, during the period 1987-2002 were analyzed, using age-period-cohort models by areas at different urbanization levels. Trends were examined with a log-linear regression model, calculating the yearly estimated percent change. Empirical bayesian estimators of the ratios between observed and expected deaths by municipality were calculated for the most recent period and mapped.
Results: The age-adjusted lung cancer mortality rates increased from 1987 to 2002: estimated percentage change values were equal to 24.5% in the urban areas (P < 0.001) and 17.2% in the rural areas (P = 0.023). The age-period-cohort model analyses showed a statistically significant drift and non-linear cohort effects. The higher risk was observed for the birth cohort of women born around 1955 (RR, 5.25; 95% CI, 2.83-9.72). In the rural areas, no significant effects were observed, and the age model showed the best fit. In recent years, the risk appeared concentrated in 9 Tuscan municipalities, accounting more than 35% of the female urban population.
Conclusions: The observed significant cohort effect in the age-period-cohort analyses for the urban areas reflects the social impact of living in these areas to induce smoking-related disease like lung cancer in women. The risk appeared particularly relevant in more recent and urbanized generations (women born around 1955), thereby suggesting urgent effective campaigns against smoking, gender dedicated, especially in urban areas.