Life expectancy and mortality differences between migrant groups living in Amsterdam, the Netherlands

Abstract

There is an apparent contradiction between the high level of morbidity and the low level of mortality observed in certain groups of migrants living in Europe. This observation should have some consequences for health policy development and the targeting of resources in a city like Amsterdam. In this paper a number of hypotheses to explain the low mortality in migrant groups are discussed. An analysis is made of mortality in Amsterdam using data from the civil registry as to mortality according to age, sex and nationality group of the deceased. Standard demographic techniques such as the standardised mortality ratio (SMR) and life table analysis were employed. Life table analysis shows that life expectancy in Amsterdam is lowest among residents of Dutch descent (73.3 yr for males and 79.1 yr for females) and highest among those of Mediterranean origin (77.6 yr for males and 86.1 yr for females). This appears to contradict previous research based on the SMR, which showed high mortality in migrant groups. To find the cause of this contradiction, the SMR and risk ratios by age are studied. The conclusion of this paper is that on the basis of life table analysis it appears that some immigrant groups living in Amsterdam have a remarkably high life expectancy. Since the SMR is sensitive to demographic differences between groups compared, questions can be raised about previous studies using the SMR. It has been suggested that the high life expectancy in migrant groups is not really caused by good health but by ‘spurious’ phenomena, such as problems in mortality registration. However, in view of the available data it seems likely that some migrant groups do in fact have high life expectancy, although the morbidity in these groups can be quite high. These findings should inform health-related policy.

Introduction

Recent research into the health of migrant populations and ethnic minority groups in European countries has raised a number of epidemiological, theoretical and methodological questions (Weitoft, Gullberg, Hjern, & Rosén, 1999). Some studies, based on surveys and objective morbidity and mortality estimates, have shown that there are many health problems among migrant groups (Uniken Venema, Garretsen, & van der Maas, 1995; Bollini & Siem, 1995; Sundquist, 1995; Stronks, Uniken Venema, & Dahhan, 1999). However, other studies have indicated a remarkably low mortality in migrant groups (Wild & McKeigue, 1997; Razum, Zeeb, Seval Akun, & Yilmaz, 1997; Abraı́do Lanza, Dohrenwend, Ng Mak, & Turner, 1999), as well as lifestyles associated with good health (Bennett, 1993). This is a matter of importance; health policy development must be based on valid and reliable assessments as to migrant health. If migrant groups show a specific susceptibility to a particular health problem, positive policies should be developed which take the migrants’ culture, language and lifestyle into account (Gupta, de Belder, & Hughes, 1995; Bhopal, 1997). High-quality epidemiological knowledge which can provide unambiguous information is essential to effective policy-making and affirmative action (Lillie-Blanton & Laveist, 1996).

Historically, migrant studies and studies of geographical differences in lifestyle and health have been an important impetus in developing the ‘risk factor’ approach (Feinleib, 1995; Marmot, Adelstein, & Bulusu, 1984; Marmot et al., 1975; Salmond, Prior, & Wessen, 1989). These studies make it clear that health and disease result from complex interactions between genetic predisposition, environment, lifestyle, living circumstances and personal history, and that they can never be accounted for simply on the basis of ethnicity (Bennett, 1993). In order to provide the information needed to develop effective policies to address inequalities in health, general epidemiological knowledge will have to be combined with knowledge of patterns of health, lifestyle and living circumstances in the population.

The international literature indicates that a number of factors should be examined in studies of migrant health. Firstly, differences in health between population groups can be quite pronounced, even in the case of close neighbours such as the English, Scots and Irish (Wild & McKeigue, 1997; Balarajan, 1991; Abbotts, Williams, Ford, Hunt, & West, 1997) or the Swedish and Finnish (Sundquist (1997a), Sundquist & Johansson (1997b); Sundquist, 1995). Studies of differences in the causes of specific mortality in population groups most often report differences in the prevalence of cardiovascular diseases (Balarajan, 1991; Klatsky, 1994; Sundquist & Johansson, 1997b) and cancers (Muir, 1996) and in risk factors for these diseases. However, there are also distinct variations between groups of migrants in the prevalence of accidents, suicide, homicide, and perinatal mortality and morbidity (Lumey & Reijneveld, 1996; Shai & Rosenwaike, 1988; Fang, Madhaven, & Alderman, 1997). These differences are often thought of as being influenced by three kinds of factors: biological, environmental and lifestyle. To study differences in these factors for migrants requires a dynamic view, since in a migrant group they will have been subject to much more change than in a ‘normal’ population. Firstly, as far as biological factors are concerned, there are only a few examples of important diseases which are clearly related to a biological predisposition in a particular migrant group. Diabetes and heart disease among immigrants from South Asia have often been discussed (Gupta et al., 1995; Greenhalg, 1997; Middelkoop, Kesarlal-Sadhoeram, Ramsaransing, & Struben, 1999). Taking the dynamic view, there may have been positive health predispositions among migrants due to selection processes. The ‘healthy migrant’ hypothesis states that dynamic individuals without disabilities are most likely to migrate (Feinleib, 1995; Marmot et al., 1984). The ‘unhealthy remigration’ hypothesis (Razum, Zeeb, Seval Akun, & Yilmaz, 1997) states that there is a tendency among migrants with health problems and migrants who have adapted less successfully to return ‘home’ (Abraı́do Lanza et al., 1999). Secondly, environmental factors: most migrants work and live in deprived conditions, characterised by cramped and sub-standard housing and high unemployment rates. Migrants predominantly work in manual and unskilled jobs. Possibly they have access only to inferior health and social care (Bollini & Siem, 1995). Furthermore, the migration process itself may have an additional negative effect (Marmot et al. (1984); Trovato, 1992; Stronks et al., 1999). Among political refugees in particular a drop in social status and a history of persecution are often important factors (Fassil, 2000; Connelly & Schweiger, 2000; Adams & Borman, 2000). However, for other migrants making the change can be a positive experience, as it may constitute a form of upward social mobility (Razum, Zeeb, Seval Akun, & Yilmaz, 1997), possibly producing health benefits (Hart, Smith, & Blane 1998; Lynch et al., 1994). Thirdly, many migrants had a healthy lifestyle in their original country, characterised by a healthy diet, high levels of physical activity, a moderate use of alcohol and low use of tobacco and illicit drugs. An important example of such a lifestyle is the Mediterranean lifestyle (Gjon$\text{c}\text{̡}$a & Bobak, 1997). Often this favourable lifestyle is maintained in the new country of residence to a certain extent and for a certain period of time. However, in the process of acculturation, health lifestyles may well deteriorate (Benfante, 1992; Salmond et al., 1989). Another problem is that some groups of migrants are more vulnerable with regard to illicit drug use, alcohol use, and accidents and violence (Shai & Rosenwaike, 1988). This may be due to a high level of social and economic deprivation in groups of migrants and therefore may be no more serious than in comparable groups of the host population (Trovato, 1992).

The low level of mortality in some migrant groups is surprising because these migrant groups are often some of the most deprived groups in their host country. Morbidity studies show the presence in these groups of all the health problems associated with social deprivation and disadvantageous working, housing and living conditions (Uniken Venema et al., 1995; Bollini & Siem, 1995; Reijneveld, 1998). There seems to be a contradiction between the disadvantageous social position and morbidity in certain groups of migrants on the one hand, and low mortality on the other. If research is going to be able to develop knowledge and inform health policy effectively, it will have to address this apparent contradiction. A study by Wild and McKeigue (1997) into premature mortality showed conclusively that the health problems requiring the most urgent attention in migrant groups living in England were to be found not among people of Asian descent (Gupta et al., 1995), but among Scottish and Irish migrants. However, it has been pointed out that the low mortality observed in some migrant groups might be due not to good health but due to spurious relationships which have little to do with health status (Weitoft et al., 1999). Such relationships could invalidate a conclusion like the one made by Wild and McKeigue. For example, most migrants have to be registered in the host country in order to be entitled to benefits and to be able to maintain residence status and freedom of travel between the host country and the country of birth. This creates a tendency for migrants to remain registered in the host country even if in actual fact they reside wholly or partly in their country of origin. This phenomenon will be observed particularly frequently among people who do not work because of old age, disability or sickness. There is an unofficial remigration of unhealthy individuals which results in the estimated number of migrants in the host country being too high, while many deaths take place abroad, so that registration of deaths in migrant groups is often postponed. The result of all this is a low observed death rate which is in fact not related to good health but to registration problems. Lastly, a problem which has not received much attention in the literature and which we would like to put forward here has to do with the registration of age. One aspect of the problem is that in some cultures dates of birth may not be registered as precisely and reliably as they are in industrialised countries. Another aspect is that if there are legal limitations associated with age, people may overstate or understate their age on arrival in the host country. In particular, marriage partners from abroad and those applying for a working permit, when a minimum age for entry into the country is required, are inclined to overstate their age. This would result in too low an estimate of age-specific mortality in all age categories and a higher estimated life expectancy.

This paper presents a study of mortality in Amsterdam, the Netherlands. The study is informed by previous research, which shows that in Amsterdam life expectancy is lower than that in the Netherlands as a whole. Even more seriously, overtime life expectancy is increasing for the Netherlands whereas in Amsterdam it is deteriorating (Veugelers, Erkens, & Verhoeff, 1998). The opinion has been voiced that the high percentage of migrants and people of foreign origin in the population are the causes of Amsterdam's deteriorating life expectancy. This opinion is based on two observations: firstly, several studies in the Netherlands have shown that groups of immigrants have a high relative mortality as expressed in the standardised mortality ratio (SMR) estimates (Uniken Venema et al., 1995; Reijneveld, 1998; Kocken, Mackenbach, Oers, & Uniken Venema, 1994; Hoogenboezem & Israëls, 1990); secondly, the proportion of migrants and people of foreign origin has increased in Amsterdam and now stands at approximately 45% of the population (O&S, annual).

The present study includes an analysis based on the data made available by the Amsterdam city council as to age-specific number of deaths and population size. The data are available for eight nationality groups, either by country of origin of an individual, or, in the case of smaller nationality groups, by geographical region of origin. However, the analysis was carried out on only five nationality groups, including some groups with strikingly low mortality rates. The differences in mortality between these five nationality groups and the populations of Amsterdam and the Netherlands as a whole are examined by means of life table analysis, a method which consists of applying the current age-specific mortality rates to a hypothetical cohort. The life expectancy measure is the most important outcome measure of life table analysis and will be at the centre of interest in this paper. Life expectancy is the mean number of years a cohort of people might expect to live in accordance with the current age-specific mortality rates. Life expectancy is the most important indicator used in demography to compare periods and regions in terms of health and health care (including prevention) (Pressat, 1969; Shryock & Siegel (1973a), Shryock & Siegel (1973b)). The life expectancy measure is generally considered to reflect differences in mortality well, but on the other hand to be insensitive to age structures in populations, changes in birth rates, and other demographic phenomena.

In migrant studies, the SMR is often used to compare mortality rates in different population groups. The analysis in this paper also made use of the SMR. This analysis follows previous studies in migrant health in which mortality in migrant groups was compared with the national figures for the Netherlands. The SMR is sensitive to differences in age distribution between groups (McMahon & Trichopoulos, 1996; Mausner & Kramer, 1985; Shryock & Siegel (1973a), Shryock & Siegel (1973b)). This paper will study the question to what extent problems related to age distribution may have influenced previous studies.