Preview

PDF, English Download (5MB) | Terms of use

Abstract

Objectives: This dissertation investigates temporal change in all-cause and cause-specific mortality in a total of 744,263 immigrants aged 15 and over from the Former Soviet Union that arrived in Israel from 1990 to 2003, relative to the Israeli population. Mortality rates of migrants are analysed using administrative data. In the migrant health literature, temporal change in relative risk is usually examined over duration of residence, through an “environmental transition” paradigm. The existing literature is critically examined , showing the weaknesses of this approach. An alternative analytic scheme is developed here which provides a more realistic depiction of change in migrant mortality. It examines simultaneously change through duration of residence, calendar time, and year of arrival. This scheme guides a review of research on social and economic changes in this immigrant population in Israel in the 1990s. They changed their demographic and socio-economic characteristics over calendar time, over arrival cohort and over duration of residence, while changing the society they had entered. These characteristics had potential effects on factors affecting relative mortality patterns. Change in relative mortality is examined in detail in each dimension, separately and in combination. This study is the first attempt to simultaneously evaluate the effects of arrival cohort, duration of residence and calendar period on comparative migrant mortality in a true longitudinal setting. It is also the first to study differences in migrant mortality by geographic origin in the Former Soviet Union and by ethnic group (Jews and non-Jews). Methods and Results: Migrant mortality was analysed using demographic and epidemiological methods (life expectancy decomposition, Standardized Mortality Ratios (SMR), and Poisson regression for all causes of death and for separately for cardiovascular disease, neoplasms, external causes, and all other causes). Patterns are examined by age groups (15-59 and 60+ years), education, geographic origin in the Former Soviet Union (Moscow and St. Peterburg, European republics, Asian and Caucasian republics) and ethnic group (Jews and non-Jews). Life table decomposition showed differing trends over calendar period of life expectancy at age 15 for adults and elderly, for males and females, and for central and peripheral areas of the country of origin. Advantages and disadvantages with other Israelis differed by these characteristics, and by the contribution of particular cause of death groups. SMR analysis by cause group revealed that apparent declines in relative mortality over duration of residence are often an artefact of large differences in arrival cohort SMRs, for males in particular. Increasing SMRs by arrival cohort are found for adults of both genders, but are small for elderly males and non-existent for elderly females. The 1990-1991 cohort displays SMRs near 1.0 or below for all causes and for the major cause of death groups, (except for external causes). Patterns by duration differ between the cause of death groups. The effects of age, calendar year and time of migration were examined simultaneously using Poisson models, adjusted for education, ethnic group and region of origin, for all causes and for cause-of-death groups. They showed differing effects of arrival year and duration of residence between the genders and age groups. The only exception to this is external causes for females, for which the best model does not include temporal trends. For males, the best models identified arrival cohort and calendar year effects on SMR (all cause mortality and most cause of death groups). Arrival cohort effects were weaker or non-existent at elderly ages, while at adult ages SMRs increase substantially over successive arrival years. For females combined effects of duration of residence and calendar year on all causes, all neoplasms, and cardiovascular disease are found. Relative risk declines by duration of stay, but only at the adult ages. For the all other causes group the arrival cohort and calendar year effects are similar for both genders. Increasing relative risk by calendar year for both genders are surprising (with the sole exception of declining risk for all external causes for females). Arrival cohort effects are not accounted for by composition by age, education, geographic origin or ethnicity. The effect of ethnic group contrasts between the sexes. Male non-Jews have much higher relative mortality for all cause of death groups non-Jewish females have substantially lower relative risk for most cause of death groups. Discussion: This migrant population displayed changes in relative mortality by each of the time dimensions which were examined, and in different combinations. This demonstrates that all should be examined together in order to obtain valid results. A limitation of this study is the absence of data on morbidity and behavioural and other risk factors. The findings conflict with expectations of the Healthy Migrant Effect literature. This “diaspora” migration was neither unselective of “unhealthy”. Health advantages of the elderly, of non-Jewish females, of immigrants from Moscow and St. Petersburg show that “diaspora” migration can create positive health selection. These migrants in Israel invert the age pattern of relative risk found in the Healthy Migrant literature for other countries. The frequently observed negative duration of stay pattern is also absent. The early “panic phase” of 1990-91 brought immigrants who displayed better health. Arrival cohorts after 1991 show (increasing) disadvantages, perhaps due to the increasing dominance of “pull” factors in migration incentives. Patterns in neoplasm mortality may show treatment-motivated migration. “Healthy” and “sick” immigrant effects existed side by side. The advantage of the elderly points to strong positive selection but may also reflect common life-course factors with Other Israelis. Calendar year trends may indicate the long-term effects of past exposures in the FSU compared to those of Other Israelis. Early exposures may also explain similarities and differences in cardiovascular mortality patterns, raising questions concerning the role of medical technology in the “cardiovascular revolution”. Ethnic differences in mortality appear to reflect similar differences found in the Former Soviet Union, but they are smaller, suggesting that non-Jewish migrants are also positively selected. The discussion explores possible behavioural and genetic reasons for the ethnic patterns, by cause of death and age group. These are discussed with reference to a so-called “Jewish pattern of mortality”. Conclusions: The three-dimensional analytic scheme developed in this study successfully challenges the “environmental transition” model, Migrant populations, and FSU migrants in particular, live in a complex temporal flux in which, in addition to the adjustment processes measured by duration of stay, changing circumstances over calendar time and changing composition by arrival cohort both play potentially significant roles. This study has demonstrated the crucial role played by the circumstances which created and altered the migratory wave itself, alongside specific conditions on arrival in Israel. These are factors that cannot be reduced to exposures in either a pre-existing origin or destination environment. Calendar year effects, often ignored, show that socio-political and economic conditions and policies in Israel may have had a differential impact on the relative mortality of FSU immigrants as a group compared to Other Israelis, regardless of duration of stay or arrival cohort.