Social epidemiology

Does Selective Migration Explain the Latino Paradox?: A Comparative Analysis of Mexicans in the U.S. and Mexico

by Georgiana Bostean

Published 5/23/2012 (OnlineFirst) in Journal of Immigrant and Minority Health

Latino immigrants, particularly Mexican, have some health advantages over U.S.-born Mexicans and Whites. Because of... more Latino immigrants, particularly Mexican, have some health advantages over U.S.-born Mexicans and Whites. Because of their lower socioeconomic status, this phenomenon has been called the epidemiologic “Hispanic Paradox.” While cultural theories have dominated explanations for the Paradox, the role of selective migration has been inadequately addressed. This study is among the few to combine Mexican and U.S. data to examine health selectivity in activity limitation, self-rated health, and chronic conditions among Mexican immigrants, ages 18 and over. Drawing on theories of selective migration, this study tested the “healthy migrant” and “salmon-bias” hypotheses by comparing the health of Mexican immigrants in the U.S. to non-migrants in Mexico, and to return migrants in Mexico. Results suggest that there are both healthy migrant and salmon-bias effects in activity limitation, but not other health aspects. In fact, consistent with prior research, immigrants are negatively selected on self-rated health. Future research should consider the complexities of migrants’ health profiles and examine selection mechanisms alongside other factors such as acculturation.

Perceived Discrimination and Smoking Among Rural-to-Urban Migrant Women in China

by Sanghyuk Shin

Published online first in the Journal of Immigrant and Minority Health, March 3, 2012

Smoking may be a coping mechanism for psychosocial stress caused by discrimination. We conducted a cross-sectional... more Smoking may be a coping mechanism for psychosocial stress caused by discrimination. We conducted a cross-sectional survey of rural-to-urban migrant women working as restaurant/hotel workers (RHWs) and those working as sex workers (FSWs) in 10 Chinese cities to investigate whether perceived discrimination is associated with smoking. We interviewed RHWs at medical examination clinics and FSWs at entertainment venues. Modified Poisson regression was used to estimate prevalence ratios. Of the 1,696 RHWs and 532 FSWs enrolled, 155 (9.1%) and 63 (11.8%) reported perceived discrimination, respectively. Perceived discrimination was independently associated with ever tried smoking (prevalence ratio [PR], 1.71; 95% confidence interval [CI], 1.31–2.23) and current smoking (PR, 2.52; 95% CI, 1.32–4.79) among RHWs and ever tried smoking (PR, 1.36; 95% CI, 1.16–1.61) and current smoking (PR, 1.63; 95% CI, 1.28–2.06) among FSWs. Perceived discrimination is associated with higher prevalence of smoking among rural-to-urban migrant women in China.

“Hearts for Sale: The French Romance and the Sexual Traffic of Musical Mimicry,” 19th-Century Music 35:1 (2011), 34-71.

by William Cheng

How has social status been measured in health research? A review of the international literature

by Christo Albor

I contributed to this paper, and was a co-author. The paper was led by Baltica Cabieses, and the other co-authors were Pedro Zitko, Rafael Pinedo, and Manuel Espinoza. Published in Rev Panam Salud Publica 29(6), 2011

Social status (SS) is a multidimensional variable that is used widely in health research. There is no single optimal... more Social status (SS) is a multidimensional variable that is used widely in health research. There is no single optimal method for estimating social status. Rather, in each case the mea- surement may vary depending on the research subject, the base theory considered, the population of interest, the event of interest and, in some cases, the available information. This literature review develops the following topics related to SS measurement, based on the international scientific sources available electronically: i) identification of the role of SS in the context of social epidemiology research, ii) description of the principal indicators and methodological approaches used to measure SS in health research, and iii) analysis of the dis- tinct difficulties of SS measurement in specific populations such as ethnic groups, women, children, the elderly, and in rural vs. urban contexts. The review finally makes it possible to describe some of the implications of SS measurement in Latin American countries.

Examining the differential association between self-rated health and area deprivation among white British and ethnic minority people in England

by Christo Albor

The lead author was Laia Becares (Manchester). I co-authored this along with James Nazroo (Manchester), Tarani Chandola (Manchester), and Mai Stafford (MRC Unit for Lifelong Health and Ageing). It was published in 2012 at Social Science & Medicine 74(4).

Recent discourses in the area of neighbourhood effects on health have advocated for a relational perspective of space... more Recent discourses in the area of neighbourhood effects on health have advocated for a relational perspective of space and place, focussing on the mutually reinforcing and reciprocal relationship between the environment and the individual. An example of such relationship is that of the interaction between area deprivation and individual ethnicity on reports of self-rated health, which we explored using cross-sectional data from the 2007 Citizenship Survey linked to the 2001 UK census. We aimed to examine whether the association between area deprivation and poor self-rated health differs for ethnic minority groups, as compared to white British people. Following from this, we also examined whether possible differential associations were mediated by ethnic density effects and perceptions of and satisfaction with neighbourhood characteristics. Results of random effects multilevel logistic regression models showed the detrimental association between area deprivation and self-rated health to be of greater magnitude and stronger for white British people than for ethnic minority people. This differential association was not mediated by ethnic density effects or perceptions of and satisfaction with neighbourhood characteristics.

Highlights
► A detrimental association between area deprivation and poor self-rated health was found across all ethnic groups.
► However, area deprivation was less strongly associated for ethnic minority people, as compared to white British people.
► Local service satisfaction, neighbourhood disorder, and collective effects did not mediate this differential association.

Ethnic density effects on physical morbidity, mortality and health behaviours: A systematic review of the literature.

by Christo Albor

This was a systematic review in which I contributed. It was led by Laia Becares (Manchester). Other co-authors were Mai Stafford, James Nazroo, Richard Shaw, Karl Atkin, Kate Pickett, Kath Kiernan and Richard Wilkinson. It is currently in press with the American Journal of Public Health (2012)

A review of the impact of ethnic density on adult mental disorders

by Christo Albor

In press with the British Journal of Psychiatry (2012). The lead author is Richard Shaw. I am a co-author along with Karl Atkin, Laia Bécares, Mai Stafford, Kathleen Kiernan, James Nazroo, Richard Wilkinson, and Kate Pickett

Background
The ‘ethnic density hypothesis’ is a proposition that members of ethnic minority groups may have... more Background
The ‘ethnic density hypothesis’ is a proposition that members of ethnic minority groups may have better mental health when they live in areas with higher proportions of people of the same ethnicity. Investigations into this hypothesis have resulted in a complex and sometimes disparate literature.

Aims
Systematically identify relevant studies, summarise their findings in a narrative synthesis, and discuss potential explanations of the associations found between ethnic density and mental disorders.

Method
A narrative review of studies published up to January 2011, identified through a systematic search strategy. Studies included have a defined ethnic minority sample; some measure of ethnic density defined at a geographical scale smaller than a nation or a US state; and a measure ascertaining mental health or disorder. Analysis was by narrative synthesis with the aid of semi-quantitative visualisation of data.

Results
Thirty four papers from 29 datasets were identified. Protective associations between ethnic density and diagnosis of mental disorders were most consistent in older US ecological studies of admission rates. Among more recent multilevel studies, there was some evidence of ethnic density being protective against depression and anxiety for African Americans and Hispanic adults in the US. However, Hispanic, Asian-American and Canadian ‘visible minority’ adolescents suffer higher levels of depression at higher ethnic densities. Studies in the UK showed mixed results, with evidence for protective associations most consistent for psychoses.

Conclusions
The results of this review indicate that the most consistent associations of ethnic density are found for psychoses. Ethnic density may also protect against other mental disorders, but presently, as most studies of ethnic density have limited statistical power, and given the heterogeneity of their study designs, our conclusions can only be tentative.

Income and oral health relationship in Brazil: is there a threshold?

by Roger Keller Celeste

Objectives: We explored the relationship between income and two oral health outcomes in Brazil, in order to assess the... more Objectives: We explored the relationship between income and two oral health outcomes in Brazil, in order to assess the shape of this relationship.
Methods: Individual-level data from a national oral health survey were obtained for 22 634 15- to 19-year-old subjects from 330 municipalities. Relationships between income (equivalized household income) and oral health were smoothed using the locally weighted ordinary least squares regression
(LOWESS) technique in order to assess the relationship between material circumstances and oral health. We also ranked individuals based on equivalized household income, supplemented by information from total household income, interviewees’ earnings, number of cars in the household and
years of education, in order to assess the relationship between social position and oral health.
Results: The relationship between oral health and equivalized household income showed a threshold and, assuming causality, income levels higher than R$850 per month did not improve oral health further. The relationship between oral health and social position was linear. Correlations of
oral health with the ranking variable (social position) were stronger than with equivalized household income, regardless of the income level, and did not decrease after controlling for income.
Conclusions: The relation of oral health in teenagers with equivalized household income (material circumstances) showed a threshold, but the relation with a ranking variable (social position) was linear. Maybe differences in oral health between individuals are influenced by both their material circumstances (up to a certain level of income) and their social position in relation to others, i.e. social status (at any income level).

Harmful social and biological risk factors accumulated early in life influences the oral health of 6-year-old children

by Roger Keller Celeste

Racial discrimination and health: A systematic review of scales with a focus on their psychometric properties

by Roger Keller Celeste

The literature addressing the use of the race variable to study causes of racial inequities in health is characterized... more The literature addressing the use of the race variable to study causes of racial inequities in health is characterized by a dense discussion on the pitfalls in interpreting statistical associations as causal relationships. In contrast, fewer studies have addressed the use of racial discrimination scales to estimate
discrimination effects on health, and none of them provided a thorough assessment of the scales’ psychometric properties. Our aim was to systematically review self-reported racial discrimination scales to describe their development processes and to provide a synthesis of their psychometric properties. A computer-based search in PubMed, LILACS, PsycInfo, Scielo, Scopus and Web of Science was conducted without any type of restriction, using search queries containing free and controlled vocabulary. After initially identifying 3060 references, 24 scales were included in the review. Despite the fact that discrimination stands as topic of international relevance, 23 (96%) scales were developed within the United States. Most studies (67%, N ¼ 16) were published in the last 12 years, documenting initial attempts at scale development, with a dearth of investigations on scale refinements or cross-cultural adaptations. Psychometric properties were acceptable; sixteen of all scales presented reliability scores above 0.7, 19 out of 20 instruments confirmed at least 75% of all previously stated hypotheses regarding the constructs under consideration, and conceptual dimensional structure was supported by means of any type of factor analysis in 17 of 21 scales. However, independent researchers, apart from the original scale developers, have rarely examined such scales. The use of racial terminology and how it may influence self-reported experiences of discrimination has not yet been thoroughly examined. The need to consider other types of unfair treatment as concurrently important health-damaging exposures, and the idea of a universal instrument which would permit cross-cultural adaptations, should be discussed among researchers in this emerging field of inquiry.

Explicit discrimination and health: development and psychometric properties of an assessment instrument

by Roger Keller Celeste

Rev Saúde Pública 2012;46(2):269-78

OBJECTIVE: To develop an instrument to assess discrimination effects on health outcomes and behaviors, capable of... more OBJECTIVE: To develop an instrument to assess discrimination effects on health outcomes and behaviors, capable of distinguishing harmful differential treatment effects from their interpretation as discriminatory events.

METHODS: Successive versions of an instrument were developed based on a systematic review of instruments assessing racial discrimination, focus groups and review by a panel comprising seven experts. The instrument was refi ned using cognitive interviews and pilot-testing. The fi nal version of the instrument was administered to 424 undergraduate college students in the city of Rio de Janeiro, Southeastern Brazil, in 2010. Structural dimensionality, two types of reliability and construct validity were analyzed.

RESULTS: Exploratory factor analysis corroborated the hypothesis of the instrument’s unidimensionality, and seven experts verified its face and content validity. The internal consistency was 0.8, and test-retest reliability was higher than 0.5 for 14 out of 18 items. The overall score was higher among socially disadvantaged individuals and correlated with adverse health behaviors/conditions, particularly when differential treatments were attributed to discrimination.

CONCLUSIONS: These fi ndings indicate the validity and reliability of the instrument developed. The proposed instrument enables the investigation of novel aspects of the relationship between discrimination and health

The relationship between levels of income inequality and dental caries and periodontal diseases

by Roger Keller Celeste

Free download from journal homepage.

The objective of this study was to evaluate the association between income inequality at a lagged time of 2 and 11... more The objective of this study was to evaluate the association between income inequality at a lagged time of 2 and 11 years with two short latency outcomes (untreated dental caries and gingivitis) and two long latency outcomes (edentulism and periodontal attachment loss > 8mm). We used data from the Brazilian oral health survey in 2002-2003. Our analysis included 13,405 subjects aged 35-44 years. Different lagged Gini at municipal level were fitted using logistic and negative binomial multilevel analyses. Covariates included municipal per capita income, equivalized income, age, sex, time since last dental visit and place of residence (rural versus urban). Crude estimates showed that only untreated dental caries was associated with current and lagged Gini, but in adjusted models only current Gini remained significant with a ratio of 1.19 (95%CI: 1.09-1.30) for every ten-point increase in the Gini coefficient. We conclude that lagged Gini showed no association with oral health; and current income Gini was associated with current dental caries but not with periodontal disease.

The individual and contextual pathways between oral health and income inequality in Brazilian adolescents and adults

by Roger Keller Celeste

We evaluate the association between income inequality (Gini index) and oral health and in particular the role of... more We evaluate the association between income inequality (Gini index) and oral health and in particular the role of alternative models in explaining this association. We also studied whether or not income at the individual level modifies the Gini effect.

We used data from an oral health survey in Brazil in 2002–2003.
Our analysis included 23,568 15–19 and 22,839 35–44 year-olds nested in 330 municipalities. Different models were fitted using multilevel analysis. The outcomes analysed were the number of untreated dental caries (count), having at least one missing tooth (dichotomous) and being edentulous (dichotomous). To assess interaction as a departure from additivity we used the Synergy Index. For this, we dichotomized the Gini coefficient (high vs low inequality) by the median value across municipalities and the individual income in the point beyond which it showed roughly no association with oral health.

Adjusted rate ratio of mean untreated dental caries, respectively for the 15–19 and 35–44 age groups, was 1.12 and 1.16 for each 10 points increase in Gini scale. Adjusted odds ratio of a 15–19 year-old having at least one missing tooth or a 35–44 year-old being edentulous was, respectively, 1.19 and 1.01. High income inequality had no statistically significant synergistic effect with being poor or living in a poor municipality. Higher levels of income inequality at the municipal level were as

sociated with worse oral health and there was an unexplained residual effect after controlling for potential confoundings and mediators. Municipal level income inequality had a similar, detrimental effect, among individuals with
lower or higher income.

How much of the income inequality effect can be explained by public policy? Evidence from oral health in Brazil

by Roger Keller Celeste

Objectives: To evaluate the association between income inequality, a public policy scale and to oral health.
more Objectives: To evaluate the association between income inequality, a public policy scale and to oral health.

Methods: Analysis, using the Brazilian oral health survey in 2002–2003, included 23,573 15–19-year-old subjects clustered in 330 municipalities. Missing and decayed teeth and malocclusion assessments were the outcomes. Gini coefficient and a novel Scale of Municipal Public Policies were the main exposure variables. Individual level covariates were used as controls in multilevel regressions.

Results:Anincrease from the lowest to the highest Gini value in Brazilwasassociated with an increase in the number of missing (rate ratio, RR = 2.11 confidence interval 95% 1.18–3.77) and decayed teeth (RR = 2.92 CI 95% 1.83–4.65). After adjustment for public policies and water fluoridation, the Gini effect was non-significant and public policies explained most of the variation in missing and decayed teeth. The public policy scale remained significant after adjustment with a rate ratio of 0.64 for missing and 0.72 for decayed teeth. Neither Gini nor public policies were significantly related to malocclusion. The public policy effect
on missing and decayed teeth was stronger among those with higher education and income.

Conclusions: Income inequality effect was explained mainly by public policies, which had an independent effect that was greater among the better-off.

Trends in socioeconomic disparities in oral health in Brazil and Sweden

by Roger Keller Celeste

Objectives: To describe the dynamics of trends in socioeconomic
disparities in oral health in Brazil and Sweden... more Objectives: To describe the dynamics of trends in socioeconomic
disparities in oral health in Brazil and Sweden among adults, to assess whether trends follow expected patterns according to the inverse equity
hypothesis.

Methods: In Sweden, we obtained nationally representative data for the years 1968, 1974, 1981, 1991 and 2000, and in Brazil, for 16 state capitals in 1986 and in 2002. Trends in the prevalence of ‘edentulism’ and of ‘teeth in good conditions’ were described in two groups aged 35–44 with lower and higher economic standards, respectively.

Results: There was an annual decline in disparities in ‘edentulism’ of 0.4 percentage points (pp) (95% CI = 0.2–0.7) in Brazil and 0.7pp (95% CI = 0.5–0.9) in Sweden, as a result of improvements in both income groups. Concerning ‘teeth in good conditions’, in Brazil, there was improvement only in the higher income group and absolute disparities have increased (0.5pp annually), while in Sweden, there was a nonsignificant decrease (0.3pp annually) with improvements in both groups. Since 1991 in Sweden and in 2002 in Brazil, our measures of socioeconomic disparities in ‘edentulism’ were not statistically significant. Trends did not differ by sex or dental visit.

Conclusions: Despite improvements in both income groups and a decrease in disparities in ‘edentulism’, the poorer group in Brazil has seen no improvement in ‘teeth in good conditions’ and disparities have increased. It appears that Brazil and Sweden reflect different stages of trend for ‘teeth in good conditions’ and the same stages for ‘edentulism’, represented by the inverse equity hypothesis.

Trends in socioeconomic disparities in the utilization of dental care in Brazil and Sweden

by Roger Keller Celeste

Aims: To describe trends in socioeconomic disparities in utilization of dental care.
Methods:We obtained... more Aims: To describe trends in socioeconomic disparities in utilization of dental care.
Methods:We obtained cross-sectional data from Sweden in the period 1968–2000 and from Brazil in 1986 and 2002 for 16 state capitals. The outcome was the percentage of people who reported that they had visited the dentist in the last 12 months, calculated for a higher and a lower income group and stratified by sex, age (two groups: young and adults) and dental status. Adjusted prevalence differences and prevalence ratios were produced using Poisson regression.
Results: In Brazil, there was a decline in use of dental care among the 15–19 year olds in the period 1986–2002, but not among the 35–44 year olds. In Sweden, there was a decline among the young and adults between 1991 and 2000. Overall, socioeconomic disparities in use of dental services between the higher and the lower economic groups showed a decline in both countries. The reduction in disparities among young Brazilians was 1.1 percentage points per year (p<0.01), but among the other age groups the decline was not significant (p40.01). In the last surveys, the gap remained in both countries and age groups (p<0.01).
Conclusions: The recent decline in utilization of dental care and in the socioeconomic gap may mirror improvements in oral health. However, there are still relevant and persistent disparities in utilization of dental care in both countries, with a
higher proportion of people of higher socioeconomic status visiting the dentist.

Lifespan variation by occupational class: Compression or stagnation over time?

by Alyson van Raalte

Co-authored with Pekka Martikainen and Mikko Myrskylä. MPIDR working paper.

Perturbation analysis of indices of lifespan variability

by Alyson van Raalte

Co-authored with Hal Caswell. MPIDR working paper.

The contribution of educational inequalities to lifespan variation

by Alyson van Raalte

Co-authored with Anton E Kunst, Olle Lundberg, Mall Leinsalu, Pekka Martikainen, Barbara Artnik, Patrick Deboosere, Irina Stirbu, Bogdan Wojtyniak, and Johan P Mackenbach. Published in 'Population Health Metrics'.

The association between educational level and age at the menopause: a systematic review

by Roger Keller Celeste