The health situation of Turkish inhabitants of Rotterdam and Antwerp

The health situation of Turkish inhabitants of Rotterdam and Antwerp

Health Policy, 16 (1990) 75-64 Elsevier HPE 0355 The health situation of Turkish inhabitants of Rotterdam and Antwerp H.P. Uniken Venemal, R. Peeter...

725KB Sizes 1 Downloads 14 Views

Health Policy, 16 (1990) 75-64 Elsevier

HPE 0355

The health situation of Turkish inhabitants of Rotterdam and Antwerp H.P. Uniken Venemal, R. Peeters2 and P. van Haastrecht’ ‘Department 0 f Ep idemiology, Health Department of Metropolitan Rotterdam, The Netherlands and *University of Antwetp, Belgium Accepted

19 May 1990

Summary Both in Antwerp (Belgium) and in Rotterdam (The Netherlands) a health interview survey was held among Turkish and autochthonous (adult) citizens. In this article the data on the Turkish respdndents in Antwerp and Rotterdam are compared. Turkish respondents in Antwerp, in general, give more positive answers to most questions that concern their own health, especially to those questions that concern the more ‘subjective’ indicators of health (a general rating of the own health and a questionnalre that consists of ‘vague’ physical complaints). Chronic diseases are reported slightly more often by Turkish women in Antwerp. Turkish people in Rotterdam visit a doctor far more frequently than the Turkish people in Antwerp. Some possible explanations for these differences are given. Because of the absence of ‘hard’ data these explanations are of a speculative nature, and can be seen as suggestions for further research. Health status of migrants; Rotterdam; Antwerp; Turklsh

Introduction Rotterdam has approximately 24000 Turkish citizens. There is also a large number of people from Surinam, Morocco and many other countries. In all sectors of local policy, special attention is paid to the situation of these migrant groups. For local health policy-planners the problem was that little was known about the health situation of these migrant groups. Therefore a study was set Address for correspondence: Dr. H.P. Uniken Venema, Department 70032, 3000 LP Rotterdam, The Netherlands.

of Epidemiology,

0168~8510/90/$03.50 Q 1990 Hsevier Science Publishers B.V. (Biomedical Division)

GGD, Postbus

76

up to investigate the health situation and illness behaviour of the Turkish inhabitants of Rotterdam and a control group of Dutch citizens [l]. At the same time, a similar project was carried out among Turkish and autochthonous inhabitants of Antwerp (Belgium) [2]. Antwerp has approximately 5000 Turkish inhabitants. Both projects used the same methodology, and the two research groups cooperated intensively. In this article we report on the comparison of the health situation of Turkish respondents in Rotterdam and Antwerp. An answer is sought to the question whether there are differences in the perceived health and the illness behavior-u of Turks in Antwerp and Rotterdam and, if so, to what extent. Since the two groups seem to be reasonably comparable with regard to potential confounders (see the section on comparability) it can be assumed that differences in health between the two groups are related to differences in the local situation. The next question then is to analyze which aspects of the local situation are more or less beneficial to the health of the migrants. Although it was known beforehand that these projects would not provide empirical data to answer this question, it was seen as a challenge to formulate some hypothetical explanations. Literature on determinants of health and illness behaviour served as a guideline for providing these ideas. Met hods In both cities the Turkish participants were randomly selected from the register of population. In Rotterdam, 248 Turks participated in the project; in Antwerp this number was 244. All participants were between 16 and 69 years of age. The selection was stratified in the sense that 50% of the respondents were male and 50% female. The only difference between Antwerp and Rotterdam with regard to the selection of the Turkish study group was the exclusion of children living with their parents in Antwerp. In Rotterdam, all Turkish citizens falling in the age range were elegible. Because of the possible confounding effect of this difference in the selection method for this comparison between Antwerp and Rotterdam (it could, for example, cause age differences between the groups), the data on children living with their parents were excluded from the Rotterdam results. The total number of respondents used for this comparison and their gender distribution is presented in Table 1. In both projects an autochthonous comparison group was included of the same size as the Turkish group. In Rotterdam the Dutch participants were also selected Table 1 Turkish respondents

in Antwerp and Rotterdam

Antwerp

Rotterdam

Male Female

121 123

E

Total

244

221

77

from the population register, but matched individually to the Turkish respondents according to age, gender and address. After data collection it was checked whether there were large differences in socioeconomic status, and, if so, Dutch respondents were excluded from the comparison with Turkish respondents. In Belgium, selection of the autochthonous group was composed by including all the autochthonous respondents that were available, living in the same neighbourhoods as the Turkish respondents. The group thus selected was less comparable, especially with regard to age, to both the Turkish group in Antwerp and the Dutch group in Rotterdam. Therefore the data on the autochthonous groups can only be used with the greatest of care. All persons selected received a letter, inviting them to participate in the study. If no refusal was returned, everybody was visited by an interviewer at a time suggested in the letter. If the respondents were not at home, the interviewer returned at least two more times. Non-response for the Turkish participants was 13% in Rotterdam and 6% in Antwerp. Data were collected by means of a health interview that was held by interviewers of the same nationality and sex as the respondents. All interviews took place at the homes of the respondents. For practical reasons it was not possible to use the same interviewers in both cities, however, research teams in both cities cooperated in the preparation of the training and instruction of the interviewers. The interview included the following items: (1) ‘background characteristics’ of respondents (work, income, housing situation, knowledge of the Dutch/Flemish language, background in Turkey, etc); (2) health situation: - the presence of chronic diseases; - disease in the two weeks preceding the interview; - general impression of the own health; - Questionnaire for experienced health (VOEG); (3) illness behaviour: - actions that were undertaken in case of physical disorders - use of medical services. In Rotterdam, special attention was paid to the reliability of the data. By means of non-lineair canonic correlation analysis it was tested whether aspects of the situation in which the interview took place (interviewer, month of the interview, presence of others, etc.) influenced the data. Also a ‘re-test’ was carried out for a limited number of respondents (randomly selected). This re-test consisted of a second interview, held a few weeks after the first interview, that contained only those questions of which the answers should remain the same over a time period of a few weeks. A discussion of the results is given elsewhere [l], for the purpose of this article it suffices to state that the data reported in this article all ‘survived’ this test of reliability. Also, an effort was made to investigate the relationship between dependent variables and background variables. The odds ratio’s were computed for the

78

different background variables, controled for age and sex. Probably because of the very strong confounding effect of the variable age, this yielded hardly any results. Most of the background variables were so strongly related to age that, after controlling for age, most respondents fell in the same category of the independent variable.

Comparability Before comparing the health situation of the Turkish groups in Antwerp and Rotterdam it is necessary to check whether there are any substantial differences in ‘background characteristics’. The most important step to achieve this was to select both groups in the same way. After the data were collected it was possible to check whether the intention to end up with to comparable groups was realised. Of special concern were the differences in those background characteristics that relate strongly to health: age, sex and socioeconomic status. Fortunately, there were hardly any differences between the two groups in these aspects, with one exception: The Turkish respondents in Antwerp had, on the average, somewhat higher (family) income level. A first possible explanation for this difference is that the Belgium families were somewhat larger, so that there are more persons to contribute to the family income, but also more to consume it. So it may very well be that the income per capita is the same for both groups. Also, the reliability of the measurement of exactly this variable is questionable. Respondents were asked to give the netto family income by adding up all sources of income. It was the experience of the interviewers that this proved to be difficult for most of the respondents. Also it is possible that the respondents were reluctant to name all sources of income for fear of tax control. The respondents in both cities came from approximately the same regions in Turkey, and the average length of stay was similar.

Results Health status

Data were collected on four aspects of health. The results on all four of them will be given below. When interpreting the results it must be remembered that all information is ‘subjective’ in the sense that it is based on judgements of respondents. Diseases in the two weeks preceding the interview All respondents were asked whether they had been ill in the two weeks preceding the interview. The percentage of respondents who reported illness during this period was approximately the same for the Turkish respondents in Antwerp and Rotterdam and for men and women (between 41 and 44% of the respondents).

19

Chronic diseases A list of 25 chronic diseases was presented to the respondents. For every disease the respondents were asked to indicate whether or not they were suffering from that particular disease. When the average number of diseases is compared in both groups, the Turks in Belgium have somewhat higher scores than the Turks in The Netherlands (2.23 in Antwerp, 1.99 in Rotterdam). Women in Antwerp have higher scores than men in Antwerp; women in Rotterdam have lower scores than the Rotterdam men. So, especially the women in Antwerp are in a somewhat less favourable position. General impression of the own health All respondents were asked to judge their own health as good, reasonable or bad. In Rotterdam considerably more respondents answered this question with ‘bad’ (Table 2). Turkish women in Rotterdam have a more favourable impression of their own health than Turkish men in the same city. In Antwerp there are no such differences between the sexes. Questionnaire for experienced health (VOEG) The version of the questionnaire for experienced health that was used in this study consists of 19 items that measure ‘vague’ physical complaints (headache, stomach ache, nervousness, fatigue, etc). This questionnaire is used in many Dutch studies and is believed to measure both physical health status and general feelings of ‘unwell-being’. Turkish respondents in Rotterdam had higher (= less favourable) scores on this questionnaire. Their average score was 6.35, whereas the Turkish people in Antwerp had an average score of 5.85. Again, in Rotterdam the results are better for Turkish women than for Turkish men; in Antwerp the Turkish men have the best results. Illness behaviour

All responReactions to diseases in the two weeks preceding the interview dents who reported illnesses during the two weeks preceding the interview were asked whether they had done anything to cure the disease. Twenty-nine percent of the Turkish respondents in Antwerp and 79% of the respondents in Rotterdam had visited a doctor. Most Turkish people in Antwerp took care of the disease themselves. Table 2 General Impresslon

of the own health by Turkish respondents Antwerr,

GOOCI

Reasonable Bad Total (100%) x2 = 29.56; df= 2; P = 0.00.

63% 32% 5% 242

In Antwerp and Rotterdam

Rotterdam 51% 27% 22% 221

so Table 3 Number of contacts with the general practitioner during the thrw month8 proceding the interview Antwerp No cOntact E3x 4 and more Total (100%)

69% 13% 15% 2% 244

Roaerdam 35% 25% 24% 16% 215

x2 = 52.56; df= 2; P = 0.00.

Visits to the general practitioner in the three weeks prior to the interview

Ia

line with the results presented above, Turkish respondents in Rotterdam visit their general practitioner more frequently than the respondents in Antwerp (Table 3). It can be concluded that there are several important differences between the health and illness behaviour of Turkish people in Antwerp and Turkish people in Rotterdam. On more ‘subjective’ indicators of health (the VOEG and the general impression of the own health), Turkish people in Antwerp seem to be better off. Turkish men in Rotterdam have the worst results. On the more ‘objective’ indicators (chronic diseases and diseases in the two weeks preceding the interview) there are less marked differences. Chronic diseases, however, are reported slightly less often by the Turkish respondents in Antwerp. Differences in illness behaviour are evident; Turkish people in Rotterdam make far more use of the services of the general practitioner.

Discussion It is not possible to formulate clear-cut explanations for the differences found between the Turkish citizens in Antwerp and Rotterdam. Many factors may be of influence. Which factors are of special relevance here can only be speculated on. In looking for explanations for the differences in health between the Turkish inhabitants of Antwerp and Rotterdam it would be very helpful to know whether the same difference exists between groups of the autochthonous population of low socioeconomic status. In the second paragraph we explained that in both projects autochthonous comparison groups were included. However, the way they were selected differed between the two cities, so these data can only be used with the greatest of care. It can be said though, that both with regard to subjective health and to frequency with which the doctor was visited, there are no apparent differences between the two autochthonous groups. For both variables the results of the autochthonous groups resemble those of the Tur’kish group in Antwerp. Therefore it seems reasonable to look for factors that may specifically influence the health situation of Turkish people. In the literature a great many factors are formulated that influence disease and illness behaviour (for example Refs. 4 and 5). Elsewhere we discussed the possible

81

influence of many of these factors on the health situation of Turkish people in Rotterdam and Antwerp [l]. Here we will only mention the most important factors. Because the largest differences between the two groups concern the ‘subjective’ health indicators and illness behaviour, we will focus on those factors that may influence these two aspects. Relatively little importance is given to the more ‘objective’ indicators of health. Socio-economic status

It is generally recognized that there is a relationship between socioeconomic status and health [6]. People of higher socioeconomic status are, in general, in a better health condition. Other than the processes that may cause these socioeconomic differences in autochthonous populations, belonging to a low socioeconomic status group can be extra ‘stressful’for migrant populations. Most of them came to Western Europe with the explicit goal to improve their socioeconomic position. Failure to achieve such ‘upward social mobility’ may bring about special problems of acceptance [3,7]. As far as the level of education and occupational level are concerned, there are no marked differences between the Turkish respondents in Antwerp and Rotterdam. With regard to income, however, this seems to be different. The Turkish respondents in Antwerp have a higher family income than the Turkish citizens in Rotterdam. In the third paragraph it is argued that the validity of these data is dubious. Therefore it remains questionable whether the level of income is a factor that accounts for some of the differences in health. Housing situation

In general the housing situation in Antwerp is better than the housing situation in Rotterdam. In Antwerp it is cheaper to rent or to buy a house; this is also reflected in the housing situation of the Turkish people in both cities. Since it is known that in Turkish families it is seen as the responsibility of men to take care of adequate housing, it can be expected that a problematic housing situation (as is often the case in Rotterdam) causes specific problems for Turkish men. Social networks

In literature on the health of migrant groups great importance is attached to social networks. It is argued that the presence of tight social networks of migrants is conducive to health [3,8,9]. Unfortunately there is little information about social networks of Turks in Antwerp and Rotterdam. What we do know, however, is that the Turkish people in both cities very often have family members living in the same city, but that Turkish people in Antwerp have far more frequent contacts with those family members. More than half of the Antwerp Turks meet their family daily; this applies to only

82

22% of the Turks in Rotterdam.

Social networks can also influence illness behaviour [lO,ll]. People with tight social networks, based on family ties, are more likely to seek help within the own lay circuit than people with less tight networks [12,13]. Since Turkish people in Antwerp have more frequent contacts with family members it can be speculated that their networks are stronger than those of the Turkish population in Rotterdam. The consequence of this can be that for Turks in Antwerp there is less need to consult a doctor in case of perceived illness; they can first turn to their own lay network. Culture shock

Differences in culture between the country of origin and the new country can be a source of emotional problems. The larger the differences, the larger the ‘culture shock’ (see also Refs. 8 and 14). Murphy (1977) states that the degree to which old values are challenged in the new society is of special importance here. Is there any reason to believe that the ‘distance’ between the old and the new culture is different for Turks in Antwerp and in Rotterdam? There are some indications that this may be the case. Though The Netherlands and Belgium are very close to each other, there are some differences in ‘culture’; the most important one being that in The Netherlands everything is far more organised. There are more rules to adhere to; whereas in Belgium there is more freedom to do ‘as one pleases’. For two reasons it can be hypothesized that the situation in Antwerp is more beneficial to Turkish people: in the first place because the situation resembles the situation in Turkey more closely; and in the second place because, since there are not so many rules, there is less chance that their own behaviour patterns will discord with prevailing rules. Attitude towards ethnic minorities

The attitude towards ethnic minorities can be expressed both by official government policy and in day-to-day contact. The formal goals in policy towards ethnic minorities are approximately the same for both countries. Yet in practice far less concrete actions have been undertaken in Belgium. In The Netherlands a great number of facilities have been set up, i.e., in the field of health care. Consequently one may expect that Dutch society is more beneficial for Turks than Belgium. We saw, however, that the health situation of the Turks in Rotterdam is certainly not better. The consequence may be that Turkish people have learned to rely entirely on (Dutch) institutions in cases of trouble; whereas people in Belgium try to find their own answers to their problems. About the day-to-day contact between Turks and the autochthonous population in Antwerp and Rotterdam, little comparable data are available. The only indication about the attitude of autochthonous citizens towards ethnic minorities is the percentage of people that vote for a racist party. This percentage is higher in Antwerp

(6.7% in 1985 versus 3.4% in Rotterdam). In Antwerp this percentage has increased markedly in the last elections (to 17%). Evidently there is no support for a theory that explains the better health situation of Turks in Belgium by the attitude of the autochthonous population. Organisation of health care institutions

There are several differences between the health care systems in Belgium and The Netherlands. The most important ones are: - in Belgium patients have to contribute financially; - Medical specialists in Belgium are directly accessible; whereas in the Netherlands specialists can only be consulted via the general practitioner; - In Belgium more medicines are available without a prescription. In general it can be concluded that the Belgian system bears more resemblance to the Turkish system (for information about the latter see for example van Haastrecht, unpublished data), especially the greater availability of medication and the absence of the general practitioner as ‘gatekeeper’ for almost every other sector of health care. This, together with the difference in ‘culture’ mentioned before, may enlarge the chances for the Turkish people in Belgium to maintain the behaviour they had learned in Turkey, which is definitely not to go to a doctor very often (Van Haastrecht, unpublished data). In general, Turkish respondents in Antwerp give more positive answers to most questions that concern their own health, especially to those questions that concern the more ‘subjective’ indicators of health. The Turks in Rotterdam visit a doctor far more frequently than the Turks in Antwerp. Taken into account (1) the more accessible health care system in The Netherlands (no financial thresholds, interpreters’ services); (2) a more intensive care system for migrants in general (i.e., social benefits, language programs); and (3) seemingly a relatively positive attitude towards ethnic minorities in Rotterdam, one would rather have expected an outcome in favour of the Rotterdam Turks! It is impossible to give a clear-cut explanation for the differences in health between the Turkish inhabitants of Rotterdam and Antwerp. It has been speculated that several factors can be of influence. Most of the factors are interrelated. All of them have to do, in one way or another, with the possibilities offered by the host society to the migrant groups. It can be speculated that more interference by the host community with the behaviour of the migrants does not necessarily lead to better health. The conclusion that the relation between migrants’ health and possibilities offered by a host society is a extremely complex one, seems to be more than justified. Further research, both qualitative and quantitative, to unravel this complex relationship is highly recommended. The way in which the health policy for migrants is organised and it’s consequences should also be scrutinized.

84

References 1 Uniken Venema, H.P., Toen ik hier kwam was ik kemgezond; de gezondheid van Turken in Nederland. Bohn, Scheltema en Holkema, Utrecht, 1989. 2 Eylenbosch, J.W. and Peeters, R., Omgaan met gezondheid en ziekte, ESOC, Universiteit van Antwerpen, 1986. 3 Peeters, R., Ziekte en gezondheid bij Marokkaanse immigranten, Katholieke Universiteit Leuven, 1983. 4 Kohn, R. and White, K.L., Health care, an international study. Report of the WHO international collaborative study of medical cam utilization, Oxford University press, 1977. 5 Andersen, R., A behavioural model of families use of health services, Research Series 25. Chicago Center of Health Administration Studies, 1986. 6 Mackenbach. J.P. and Van der Maas, J.P., Sociale ongebjkheid en verschillen in gezondheid, Instituut voor Maatschappebjke gezondheidszorg. Erasmus Universiteit Rotterdam, 1987. 7 Mak, G. and Schrameyer, F., Migranten en geestelijke gezondheidszotg, NCGV 55, Uttecht, 1983. 8 Eppink, A., Migranten in het buitenland, buitenlands wetenschappelijk onderzoek naar de gezondheid van migranten, NCGV 113, Utrecht, 1987. 9 Shuval, J.T., The connibution of psychological and social phenomena to an understanding of the aetiology of diseases and illnesses, Social Science and Medicine, 15a (1981) 337-342. 10 Kroeger, A., Anthropological and sociomedical health care research in developing countries, Social Science and Medicine, 3 (1983) 147-161. 11 Kleinman, A., Patients and healers in the context of culture, University of California Press, Berkeley, Los Angeles, London, 1980. 12 McKinlay, J.B., Some approaches and problems in the study of the use of services - an overview. Journal of Health and Social Behaviour, 13 (1972) 115-152. 13 Suchman. E.A., Stages of illness and medical care, Journal of Health and Human Behaviour 6 (1965) 114-128. 14 Kasl, S.V. and Berkman. L.. Health consequences of the experience of migration. Annu. Rev. Public Health 4 (1983) 69-90. 15 Murphy, H.B.M., Migration, culture and mental health, Psychological Medicine, 7 (1977) 677-684.