Migration has probably become one of the most important determinants of global health and social development. People are moving in greater numbers than ever before and they are doing so at a faster pace and over greater distances. Even under ideal conditions the uprooting, displacement and resettlement involved in migration – be it voluntary or involuntary – can pose complex challenges. Migration has implications for those who move, those who are left behind and those who host migrants. Its implications for the health care systems of communities and countries involved can also be far-reaching.
Massive population movement is not new to countries of the European Union (EU). Emigration from them has historically been an important and highly effective economic, political and social means of adjustment, and most countries have benefited from migration at one point or another in recent times. The emergence of the EU as a powerful economic entity, however, is now producing changes in the direction of migration, and the member states are beginning to be net importers rather than exporters of human resources. This is unlikely to change any time soon. Indeed, as the principles of free movement become more established, and as EU membership expands, the potential for people to move within and into the EU will increase. The challenge of making population movement and resettlement healthy and socially productive will become an even greater ethical and pragmatic necessity.
In 1997, the International Centre for Migration and Health (ICMH) was asked by the European Commission to review the health implications of migration into EU countries. The review covered a wide spectrum of health issues and problems, ranging from communicable diseases to mental health and family formation. This paper provides a brief and updated overview of some of the more salient findings and conclusions of that review, and highlights the growing need to adress the issue of migration and health in the European context.
This is not meant to be an exhaustive analysis of the situation. Indeed that would be difficult given the variation in available information on the subject. In some countries little information is systematically gathered in this subject area, and few national health statistics anywhere currently distinguish between migrants and others. Although some countries are beginning to address the issue from an ethnic minority perspective, for now there is a marked paucity of routinely gathered data in this area. Much of the available information comes from studies by research institutes and individual researchers. These constitute a relatively sound basis for assessing the broader dynamics of health and migration, but many are small-scale studies and have been designed using distinct methodologies and reflecting different interests. Thus drawing generalizations from them requires caution. We consider some of the trends that are emerging in the area of communicable and noncommunicable diseases, reproductive health, psychosocial and psychiatric issues, and family life.
Of all the public health problems associated with population movement, communicable diseases have historically most captured the attention of policy makers and the public. Tuberculosis has probably been the most worrisome of these for immigration authorities. It may well come to the forefront again, for in 1993 the World Health Organisation (WHO) announced that TB had become a global emergency, with as many as 10 million new cases predicted by the early part of the 21st century.
TB is typically a disease of poverty. It thrives in communities characterized by low education levels, poor nutrition, inadequate housing, overcrowding (Almeida & Thomas 1996) and limited access to preventive and curative medical services (Rieder et al. 1994). Improvements in all these conditions in Western Europe over the course of the last 50 years have helped to significantly reduce the problem, and the incidence of new cases in 9 EU countries fell from 34.8/100 000 in 1974 to 14.3/100 000 in all 15 EU member states by 1995 ([WHO (1997a) ]WHO 1997a,b). Similar health benefits have not been possible everywhere, however, and in economically less developed countries and societies marked by war and social upheaval TB has remained a pervasive problem.
Because much of the migration within and into the EU involves people moving from less to more economically developed regions, a number of EU countries are seeing changes in their TB profiles. In Denmark, where the incidence of new cases has increased over the past five years, the proportion of foreign-born cases has risen from 18% in 1986 to 60% in 1996 (Prinsze 1997). In England and Wales the National Survey of Tuberculosis in England & Wales (Anonymous 1983) estimated that 40% of all TB cases occur among people arriving from the Indian subcontinent (Karmi 1997), and in the Netherlands (De Jong & Wesenbenk 1997), where the incidence of reported TB rose by 45% between 1987 and 1995, at least 50% of these cases involve migrants. Similar trends are being noted in Germany (Huismann et al. 1997) and France (Gliber 1997) where migrants are three and six times, respectively, more likely to be diagnosed with TB than nonmigrants.
There is little evidence that this presents a problem for host communities, but the risk of spread within migrant communities themselves may be considerable. Not only do many migrants come from backgrounds where TB is still common, but many of them also move into living conditions that offer little protection from TB. Even in postindustrial countries such as the Netherlands (De Jong & Wesenbenk 1997), Austria (Hammer 1994) and France (Gliber 1997) many migrants enter substandard housing and social environments that involve serious overcrowding and poor sanitation, conditions which enhance the risk of TB spread (WHO 1996; Gaspar & Silès 1997). Since many of them also remain socially excluded from host societies, they fail to effectively use the health care services available to them. The situation is especially precarious for the growing number of migrants who arrive on short temporary work permits, and even more so for those who move unofficially.
The importance of social and economic conditions, especially housing, in the spread of TB has been highlighted by health planners in Italy (Carchedi & Picciolini 1997) who have referred to substandard immigrant housing and poor access to health services as important precursors to TB, including drug-resistant varieties. The problem is of course not limited to Italy. A study of migrant agricultural workers in a farming region of southern Spain found that 85% were living in makeshift, highly overcrowded rooms opposite the greenhouses and warehouses they worked in. Over 75% of the dwellings had no running water or toilet facilities, 70% no electricity, 95% no heating or air conditioning, and 65% no refuse collection (CITE 1997). A report on Cape Verde migrants in Lisbon equally highlighted the problem of poor housing and the lack of basic amenities; over a third of their houses had no piped water, 13% had no toilet facilities and 26% no organized sewage disposal (CITE 1997). A 1991 study (Gardete & Antunes 1993) found that 13% of 622 new cases of TB involved migrants living in these conditions. Interviews with Moroccan patients hospitalized with TB in France (Nejmi 1983) have similarly reflected the extent to which poor housing has come to characterize the situation of migrants and its potential implication for the spread of TB within migrant communities.
Of all the diseases that can be sexually transmitted few have raised the concern that has surrounded HIV/AIDS. The possible role of population movement in the spread of HIV/AIDS has begun to attract attention (Carballo & Siem 1996; Carballo et al. 1996; IOM/UNAIDS 1998) but the relative role it has played in EU countries remains unclear. In Belgium (Muynck 1997) and Italy (Carchedi & Picciolini 1997), for example, the prevalence rate for AIDS among migrants is lower than it is among nationals, but in Germany (Huismann et al. 1997) migrants now comprise 14% of all AIDS cases, a disproportionately high number. Most of these cases come from the Americas, Africa and Asia, while carriers from Turkey and Eastern Europe, where HIV/AIDS is still less common, are far less numerous (Huismann et al. 1997). In Sweden rates of HIV and other sexually transmitted diseases (STDs) among migrants, especially from Africa, also appear to be exceeding national rates (Janson et al. 1997).
Just as with hepatitis B, the risk of STDs may be elevated where the sex-selective nature of labour migration has produced predominantly male migrant groups whose patterns of sexual behaviour involve considerable recourse to sex workers. In Belgium (Muynck 1997), STD morbidity is significantly higher among unmarried male immigrants than it is among Belgian males in general.
Cardiovascular diseases (CVDs) account for about 50% of all deaths and 33% of disabilities in the EU (WHO 1994). In the United Kingdom (Balajaran & Raleigh 1992), people from South Asia (primarily India) appear particularly prone to coronary heart disease (McKeigue et al. 1993; McKeigue & Sevak 1994), and both men and women from South Asia have 30–40% higher rates than people of English background and people from other parts of Europe (Balajaran 1991). UK data for stroke incidence also point to major ethnic variations. Significantly higher than average rates are reported for Caribbean people (Balajaran 1991), and although mortality from stroke in all groups has declined in recent years, stroke-related mortality among men from the Indian subcontinent has declined at a much lower rate than for other groups (Cruickshank et al. 1980). Diabetes mellitus, a contributing factor in stroke, is also found twice as often among Caribbean and South Asian people as among the British population overall (Cruickshank 1989) and in Belgium noninsulin-dependent diabetes is three times more common among immigrants than among the indigenous population (de Muynck & Schiliemans 1986; Muynck 1997). In Sweden, where above-average rates of obesity and CVD are reported for people originating from Finland (Järhult et al. 1992), traditional Finnish dietary habits may be a contributing factor. Hypertension-mortality also reflects ethnic differences. People of Polish origin in the UK (Balajaran 1991), for example, have excess hypertension-linked mortality, and mortality rates among people from Ireland and Scotland (where hypertension-related morbidity and mortality is generally higher than in other parts of the UK) tend to reflect the profiles of their regions of origin.
Breast cancer is now the second leading cause of death among women in the EU (Geddes et al. 1993; Muir 1993; Eurostat 1995a). Breast cancer patterns reflect national background, length of residence and lifestyle-related factors. In the UK, rates for cancer of the breast, uterus and ovary have tended to be consistently lower among Italian-born women. Data from Scotland on digestive tract cancers (Black 1993) also show lower rates for Italian female immigrants, but their rates of stomach cancer, on the other hand, tend to be higher. Time sequence data nevertheless indicate that patterns of breast cancer among immigrant groups are gradually beginning to approximate those of the receiving populations, especially in the case of people who immigrated early in life.
Reproductive health of migrants presents one of the most important, and still unmet, public health challenges. In the UK, several studies (Balajaran & Botting 1989; Bundey et al. 1991; Chitty & Winter 1989) report that babies of Asian mothers tend to have lower birthweight than other babies; perinatal and postneonatal mortality rates are also higher among babies born to immigrants from Pakistan and the Caribbean. The situation in Belgium shows similar trends: in 1983 the highest perinatal and infant mortality rates were recorded for babies born to women from Morocco and Turkey (Muynck 1997). Despite overall national improvements in maternal and child health between 1983 and 1993, high perinatal and infant mortality rates persisted in the Turkish community, and in 1993 they were still 3.5 times higher than for the Belgian population as a whole (Muynck 1997). In Germany (Huismann et al. 1997), rates of perinatal and neonatal mortality also tend to be higher in foreign-born groups, especially babies born to Turkish mothers. Much the same has been observed among immigrant groups in Spain, where premature births, low birthweight and complications of delivery are all cited as common problems among women from sub-Saharan Africa, Central and South America.
Among African women giving birth in Spanish hospitals, the incidence of premature births is almost twice as high as for Spanish women, and low birthweight is approximately 11.5% compared to 5.5% (Jansa et al. 1994). African women also tend to have higher rates of miscarriage (Teixidor et al. 1993). In the case of babies born to mothers from Central and South America, over 8% are underweight and 6.3% are premature (Teixidor et al. 1993). Oller et al. (1997, cited in Gaspar & Silès 1997) note that unwanted pregnancy, poor knowledge about contraception and where to get it are common problems, and requests for abortion tend to be twice as common as among Spanish women, especially by women from subsaharan and North Africa.
Dietary habits are strongly culturally defined human behaviours and also extremely unstable. Migration often necessitates fundamental changes in what food people eat and how it is prepared, forcing old nutritional habits and customs to be adapted to new lifestyles. Breast-feeding is a case in point. Breast-feeding is often taken for granted in traditional societies but far less common in industrial and postindustrial societies, where work pressures and the influence of formula foods have helped erode breast-feeding. The impact of this on newcomers can be important. Vietnamese women in the UK have been reported as saying they lack the confidence needed to initiate and maintain breast-feeding, especially when they live in communities with few other Vietnamese families or experienced women they can turn to for support (Sharma et al. 1994). Assumptions by local health care providers that women from traditional societies will persist with breast-feeding irrespective of other factors have not been borne out. More importantly, one report pointed out that assumptions of this kind can lead to migrant women not being allocated the support they need. In one case this led to a decrease in the prevalence of breast-feeding among Bangladeshi mothers (Hilder 1993), which may have reduced their duration of lactational amenhorrea and exposed those not using other contraceptive methods to unplanned pregnancy.
Nutrition-related problems in the children of migrants are not uncommon. Inappropriate use of breastmilk substitutes and poor weaning has been linked to vitamin deficiency rickets and ferropenic anaemia in infants of migrants in Spain (Jansa & et al. 1994). In Sweden under-nutrition, stunting and anaemia are frequent problems among refugee children (Janson et al. 1997) and in the Netherlands vitamin D deficiency together with protein-energy malnutrition is reported to be a fairly common problem among immigrant children (Meulmeister et al. 1990). In Germany, on the other hand, obesity among children of non-German background may be indicative of a tendency to parental overcompensation (Huismann et al. 1997).
Occupational health and safety has become a major concern in the context of migrant labour everywhere. Migrants tend (at least initially) to move into low-skill, temporary employment involving poor environmental conditions. Because their work is often considered too short-term or menial to justify investments in training, they are exposed to avoidable accidental injuries. The problem is exacerbated by lack of good communication with employers and supervisors, frequent lack of familiarity with the machinery used, and different attitudes to safety. In Germany (where, incidentally, migrants make far fewer insurance claims than nationals), immigrants have occupational accidents twice as often as native workers (Huismann et al. 1997) and in France over 30% of accidents resulting in permanent disabilities involve non-French workers (Gliber 1997). Data from Belgium (Peeters et al. 1982) similarly indicate that Moroccan and Turkish migrants employed in heavy industries have a higher incidence of accidents than nationals, and also that they have more secondary psychological sequelae as a result. In the agricultural sector exposure to pesticides and other chemical products is often a chronic problem and in Spain it has been linked to depression, neurological disorders and miscarriages in migrant agricultural workers. Other health problems, including dehydration and heart complaints linked to high temperatures among workers in greenhouses, are also common (Castelló 1992; Parron et al. 1992).
Migrants also appear to be more vulnerable to other types of accidents. In Germany, non-German children in the 5–9 year-old age bracket reportedly have more traffic and domestic accidents than German children (Korporal & Geiger 1990). Children of Moroccan and Turkish migrants in the Netherlands (De Jong & Wesenbenk 1997) also have more domestic accidents, including poisonings and burns and more traffic accidents than Dutch children. In France poor-quality housing is an important risk factor in the frequent incidents of lead poisoning among children of migrants living in old and poorly maintained houses.
Psychosocial health issues are often forgotten in the overall scheme of things. They are nevertheless an important component of the problem faced by migrants. Some EU countries have adopted resettlement policies that stress geographical dispersal of minorities and migrants in order to achieve faster integration into mainstream society. There is little evidence this has been effective, and the isolation that follows can instead be highly detrimental to the mental health and social integration potential of newcomers. This has been the case with Vietnamese refugees in Finland (Liebkind 1996), where younger immigrants were more able than older people to adopt Western values and behaviours, but where doing so prompted anxiety and depression among mothers who saw themselves as losing their children.
Because migration often means separation of spouses, marital problems are common. Even when families are eventually reunited, separation and divorce often follow, particularly when one of the partners cannot find work. Reports (de Jong 1994) suggest that in families which benefited from reunification schemes, men often find themselves sexually and emotionally distant from their spouses and relatives. Some have gone on to develop new relationships, while others had simply idealized their families and family ties in ways irreconcilable with reality. Others seem to lose face when relatives complain about being brought into more precarious economic conditions than expected from the letters and telephone calls they received while still at home. If and when migrant couples do separate, finding culturally sensitive social support can be difficult and severe problems of loneliness, fear and low self-esteem are common. Children of migrants who separate are especially affected and women lose out, especially in countries where their social status in the local immigrant community is tied to spouse and family.
Problems also arise if children who learn the host language more quickly than their parents are seen as deserting their families and cultural heritage. In the absence of supportive social networks to help children resolve this problem, they are easily forced by parents into culturally marginal situations. In Sweden a study (Neiderud 1989) reported that Greek parents actively avoided using social services that might have relieved the load of childcare because of fears that their children would integrate too much and too quickly into Swedish society.
Swedish data (Svedin et al. 1994) also indicate that children of immigrant parents who divorced while in Sweden had a much higher symptom load than children of Swedish divorcees. Because marital problems among immigrants often lead to one of the partners returning home to the country of origin with them, children are then deprived of the affective role of the other parent. The absence of fathers is reported to have adversely affected migrant children and adolescents returning to Finland from Sweden, and the boys in question were more prone to psychiatric disorders than local nonmigrant children (Moilanen & Myrhman 1989).
Psychosomatic problems are also common, and in the initial period of resettlement, Moroccan immigrants in Belgium are five times more likely to develop peptic ulcers than Belgian nationals (Muynck 1997); stress-related ulcers are a also a frequent source of morbidity among immigrants in Germany (Huismann et al. 1997). Other stress-related symptoms include frequent and severe headaches, anxiety attacks, dermatitis, and sleeping disorders, all of which can be detrimental to successful social and occupational integration. In the Netherlands, Turkish immigrants are reported to have high neurosis scores, more gastrointestinal complaints, and a higher risk of abusing alcohol than nationals. The prevalence of ulcers among both Moroccan and Turkish men is high, and immigrants from the Antilles, Morocco, Turkey and Surinam are 5–10 times more likelyto suffer from chronic tension headaches than their Dutch counterparts (van Wieringen et al. 1986). In Sweden (Janson et al. 1997), chronic anxiety, sleep disorders, and frequent headaches are also common problems among recently arrived refugees, and in Spain (Gaspar & Silès 1997) there are reports of frequent hypochondria and paranoia among immigrants. Psychosomatic disorders are additionally complicated by cultural differences in how people perceive health, the body, and the causality of disease. Frequent references to muscular pain, heartache, tightness in the chest or shortness of breath, for example, often reflect culturally prescribed ways of referring to stress rather than actual symptoms. Health care staff working from purely biophysical models of diagnosis do not always appreciate this and misdiagnosis is a serious problem (Mirdal 1985).
Alcohol and drug abuse may be a growing problem in the context of migration. In the United Kingdom Indian male immigrants, especially Sikhs, are manifesting new patterns of alcohol abuse, as reflected in cirrhosis of the liver-related mortality rates twice as high as those of English males (Balajaran et al. 1984; Cochrane & Bal 1989). Irish male migrants have traditionally been characterized as heavy drinkers, and indeed do have higher rates of alcohol-related mental illness than other groups. However, new evidence (Greenslade et al. 1995; Mullen et al. 1996) suggests that they are no more likely to abuse alcohol than other groups, but that those who do, do so more heavily and with more apparent consequences.
Drug abuse is probably a more serious problem. A recent study (ICMH 1998) highlighted the extent to which migrants in some settings are gratuitously offered drugs in their daily lives and how difficult it is for them to find culturally acceptable/sensitive help in prevention. In the United Kingdom (Lipsedge & Turner 1990; Lipsedge et al. 1993), the influx of drug users to the UK in search of better living conditions has become an evident problem. In Amsterdam (de Jong 1994) about a half of methadone bus programme users are foreign-born, and so are about 25% of young women who prematurely leave drug Youth Advice Centres. Approximately 45% of detainees in youth penitentiaries in the Netherlands are said to be children of migrants, but this may of course reflect arrest trends rather than real patterns of drug use.
The reasons for drug use among children of immigrants vary considerably. In France (Ait Menguellet 1988; Yahyaoui 1992; Bendahman 1993; Yakoub 1993; Boylan 1995), it is seen as a manifestation of social marginalization, and as an expression of anger/frustration at the difficulties of integration. A study of psychological stress and coping among Greek immigrant adolescents in Sweden (Giannopoulou 1988) tends to confirm this, and similar observations have been reported from Germany (Akbiyik 1990). Patterns of drug use among migrants and refugees, however, are not necessarily very different from those of local populations, and there is growing evidence that what differentiates them is the immigrants' capacity to make appropriate use of health and social services (K.E. Sparrow, personal communication).
Among rural Turkish workers in Amsterdam (Sayil 1984), where only a minority are able to speak Dutch, mental health problems, including anxiety and neurosis, are reportedly common and regrets are frequently voiced about the decision to leave home and family because of contractual arrangements. The pattern of emerging mental health problems nevertheless varies with time. In Germany approximately 13% of immigrants seen for depressive disorders seem to develop them during the initial 12 months away from home. Another 25% tend to develop problems during the following two to five years, many of them referring to a ‘longing for home’ and reporting exaggerated memories of family, the way they lived and the scenes they saw as children (Huismann et al. 1997). Selective recall and fantasies about home and return are often described as the ‘migrant's opium’, and while they are not thought to be necessarily serious, they can be psychologically debilitating (de Jong 1994). Nowhere is this more true than in the case of forced migration where families are violently dispersed and where unresolved fears about relatives are common and persist for long periods.
Schizophrenia is one of the most dramatic and serious of mental health problems. It is highly disruptive of family life, affecting not only patients but also immediate caregivers. In the UK (London 1986), people of Caribbean, Irish and Polish origin have the highest hospital admission rates for schizophrenia. In fact Caribbean immigrants tend to be diagnosed with schizophrenia three to six times more often than non-Caribbean people (Dean et al. 1981; Cochrane & Bal 1987; McGovern & Cope 1987; Littlewood & Lipsedge 1988), and in 16–29-year-olds it is up to 12 times more common than among non-Caribbean adolescents (Karmi 1997). Schizophrenia is also more frequently diagnosed in people from India and Pakistan, especially among men, than among native British; even so, rates of schizophrenia among Caribbean people are still three times higher than for people from South Asia (Cochrane & Bal 1989). In Germany (Lazaridis 1985) schizophrenia is also a frequent first diagnosis; however, reports suggest that it is often revised on review and redefined as neurosis or psychopathic problems. Drug abuse-related schizophrenia (together with other psychotic diseases) is also a fairly common problem among migrants from Surinam and the Antilles in the Netherlands (Selten & Sijben 1994).
The relatively high incidence of depression among migrants has been associated with high rates of suicide. In the Netherlands, where the unemployment rate among migrants in 1994 was 31% compared to 13% for Dutch nationals, the suicide rate among children of migrants was significantly higher than in the general population (de Jong 1994). In Rotterdam (de Jong 1994) children of Turkish immigrants are five times as likely as Dutch children to commit suicide, and young people of Moroccan background are three times as likely to do so. In the United Kingdom (Patel & Gaw 1996; Karmi 1997) suicide among South Asians is twice the national average, and in the 25–34-year-old age group, 60% higher. Suicide among women, especially women in the 15–24-year-old age range, is more frequent than among men, as is attempted suicide.
Psychiatric morbidity among children may be linked to a range of family, personal and environmental circumstances, including culture conflict, job insecurity, regrets about leaving home, family disruption and uncertain future opportunities. Adults often leave children behind with grandparents, or when they move together have to take jobs that involve work schedules that keep them away from home and children during nonschool hours. Affective relations often suffer, and some studies (Ostek 1983) have reported that up to 20% of preschool age Turkish children in Germany are looked after by siblings, and over 50% of children under 15 by persons other than their mothers.