Abstract

Secular and cohort trends in mortality from cancer in Scotland during 1953-93, and incidence during 1960-90, were analysed using individual records from the national mortality and registration files. For certain cancer sites, the secular analyses of mortality were extended back to 1911 by use of published data. Mortality from cancer at older ages in Scotland has increased over the last 40 years. In each sex, this trend has been dominated by the effects of smoking: all-cancer rates and rates of lung cancer, now the most common fatal cancer in men and in women in Scotland, reached a peak in the cohort of men born at the turn of the century and the cohort of women born in the 1920s. For much of the period, the Scottish all-age rates of lung cancer were the highest reported in the world; they are now decreasing on a secular basis in men, but are still increasing in women. There have also been large increases at older ages in the incidence and mortality rates for cancer of the prostate in recent years. bladder cancer, nervous system cancer, non-Hodgkin's lymphoma, myeloma and leukaemia; for each there is likely to be a considerable artefactual element to the increase, with differing degrees of possibility that there may in addition be an element of real increase. Substantial decreases in mortality at all ages have occurred for stomach and colorectal cancers and substantial increases at all ages for pleural cancer and melanoma. Rates of mortality from breast cancer, the most common cancer in women in Scotland, have generally increased over the past 80 years; a temporary cessation in this upward trend occurred in the years during and after the Second World War, and recently rates have turned downward, probably at least in part because of better treatment. Mortality from ovarian cancer, the second most common reproductive-related female tumour in Scotland, has also increased at older ages. At younger ages, mortality from cancer in Scotland has decreased, especially in men, whereas incidence has not. This divergence, which has been a consequence of better treatment, has occurred especially for cancers of the testis and ovary, Hodgkin's disease and leukaemia. There have been increases at young adult ages, however, in both mortality from and incidence of oral and pharyngeal, oesophageal and laryngeal cancers in men, and melanoma and non-Hodgkin's lymphoma in each sex. Cervical cancer rates at young ages also increased, but this trend has reversed for incidence in the most recent birth cohorts. Incidence rates have also increased for testicular cancer in young adults and leukaemia in children. With the possible exceptions of non-Hodgkin's lymphoma and childhood leukaemia, the increasing rates are likely largely to reflect real rises in incidence, and they highlight the need for investigation of the causes of these cancers, and, when causes are known, for preventive action.

Highlights

  • We have presented the material for mortality before that for incidence for each site, and for sites with poor survival we have often presented as Figures only the mortality data, because these are available for a longer period and are less likely to be affected by changes in completeness over time than are the incidence data

  • Cancer mortality rates in Scotland have increased at older ages but decreased at younger ages

  • At young ages the decreasing trends in mortality from testicular and ovarian cancers, Hodgkin's disease and leukaemia are an encouraging result of advances in treatment

Read more

Summary

MATERIALS AND METHODS

Cause-specific mortality data have been published for Scotland since 1855, the first year for which cancers were tabulated by site was 1901 (Registrar General in Scotland, 1904). In order to consider changes over a longer period, we used published annual mortality data back to 1911, which have been aggregated by Division of Epidemiology (1976) From this source we extracted rates for 1911-54, and joined these with the rates for 1955-93 that we had calculated from the computer files. To analyse cancer trends in relation to birth cohort, we used the date of birth information in the original individual records to calculate age-specific mortality rates by actual year of birth, rather than the quinary quinquennial estimates by the method of Case (1956) that are usually used. Information on accuracy of Scottish national cancer registration data is available only for 1990 (Brewster et al, 1994), the last year of incidence included in this Supplement. Reabstraction of data for over 2000 cases in that year showed a good level of accuracy, with 5% of cases coded to the wrong ICD code, and 2.8% of registrations

65-84 Males Females
RESULTS AND DISCUSSION
CONCLUSIONS
72. IARC: Lyon
C Cancer Research Campaign 1998
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call