Abstract

STUDY OBJECTIVE--The geographical pattern of mortality in The Netherlands is dominated by an area of relatively high mortality in the southern part of the country. The aim was to analyse the background of this geographical mortality pattern in the early 1980s, and its evolution over time since the early 1950s. DESIGN--Mortality data by district (n = 39), cause of death (13 large causes, "symptoms and ill defined conditions", all other causes), and time period (1950-54, 1960-64, 1970-74, 1980-84) were available from the Netherlands Central Bureau of Statistics. Standardised mortality ratios were calculated, and the logarithms of these were related to three sociodemographic characteristics using multiple, ordinary least squares regression analysis. SETTING--This study used data for the whole Dutch population. MAIN RESULTS--Although the geographical mortality pattern has been rather stable over the last decades, a clear tendency towards convergence is also apparent. Approximately 90% of the current excess mortality in the southern part of the country is due to cardiovascular diseases. The results of regression analysis show that the excess mortality is primarily related to the high percentage of Roman Catholics in this part of the country, and additionally to a slightly lower average income. In The Netherlands, a higher percentage of Roman Catholics in the population is linked with higher all cause mortality rates, as well as with higher mortality rates for lung cancer, ischaemic heart disease, cerebrovascular disease, arterial disease, and chronic non-specific lung disease. Survey data show that these associations are partly due to a higher prevalence of smoking among Roman Catholics. As in many other countries, a lower average income is linked with high all cause mortality rates in The Netherlands. Cause specific data show negative associations for stomach cancer, ischaemic heart disease, cerebrovascular disease, chronic non-specific lung disease, and traffic accidents. Since the early 1950s the association between geographical mortality patterns and the percentage of Roman Catholics in the population has gradually become less strongly positive. This suggests that the convergence of the mortality rates in the South towards the national average may be related to a gradual lessening of differences in lifestyle between population groups. CONCLUSIONS--Both cultural and economic factors are important in the explanation of geographical mortality patterns in The Netherlands.