Summary Time-related trends in the incidence of non-Hodgkin's lymphoma (NHL) in Denmark were analysed for the period 1943-89. A total of 13 822 patients (7565 men and 6257 women) were included in the study. In men, world-standardised incidence rates per 100 000 population increased from 2.5 in 1943-47 to 9.3 in 1988-89. In women, a similar increase was seen, i.e. from 1.9 in 1943-47 to 6.5 per 100 000 population in 1988-89. For all birth cohorts, the male-to-female incidence ratio was highest among young subjects and fell significantly after the age of 29 years. Trends in age-specific incidence were analysed separately for two periods, i.e. 1943-77 and 1978-89, reflecting (Hakulinen et al., 1986;Martinsson et al., 1992;Cartwright, 1992;Zheng et al., 1992;Devesa and Fears, 1992;Carli et al., 1994). Most reports have described annual increases of 2-5%, some suggesting higher rates of increase in men than in women (Martinsson et al., 1992;Devesa and Fears, 1992). Identification of the point in time when this widespread trend started could yield important information as to possible causes of NHL. The vast majority of previous reports, however, have been restricted to the most recent decades, and consequently little is known about the temporal trends before the 1960s.The main objective of the present study was to describe the long-term trends in the incidence of NHL in a wellmonitored, homogeneous population. This was achieved by using data from the Danish Cancer Registry, which has collected national cancer data over half a century. Detailed analyses of changes in age-specific incidence rates between an early and a more recent period were performed in an attempt to separate changes possibly attributable to the AIDS epidemic from those caused by other factors.
Materials and methodsThe Danish Cancer Registry has recorded almost all diagnoses of cancer in Denmark since 1943. In addition to contemporary coding systems, a national modification of the seventh revision of the International Classification of Diseases (ICD-7) has been used continuously in order to generate comparable data over time (Storm, 1991 (<30, 30-39,..., >80 years) for 5 year calendar intervals in the early period (1943-47, 1948-52..,1973-77) Table I. During the 47 year study period, world-standardised incidence rates increased from 2.5 to 9.3 per 100 000 in men and from 1.9 to 6.5 per 100 000 in women, corresponding to an overall increase in both sexes of around 250% (Figure 1).
Age-specific incidence ratesWomen were on average 3-5 years older than men at the time of diagnosis. The mean age at diagnosis increased considerably, from 52.2 years in men and 55.6 years in women during the period 1943-47 to 60.6 years and 65.7 years, respectively, in 1988-89. Incidence rates were consistently higher in men than in women (Figure 2). Throughout the 47 years under study, the male-to-female ratio of age-specific incidence rates (M/F ratio) declined with increasing age. Also, a remarkableCorrespondence: H Hjalgrim,