Contextual effects on health: systematic review of studies using natural experiments among migrants

Abstract Background Many studies on contextual health effects suffer from compositional bias and selective migration into neighbourhoods. Longitudinal natural experiments have the potential to overcome these limitations, and there are several opportunities for this research design in the migration context. We aimed to synthesize evidence from natural experiments among migrants studying the effect of contextual factors on health and healthcare. Methods Peer-reviewed literature in English or German was systematically searched in four major databases in December 2021. Following systematic abstract- and fulltext-screening, 32 studies were included for analysis. Evidence on contextual impacts on physical and mental health, mortality, and healthcare was narratively synthesized and quality appraisal conducted. Results We found four types of contextual health effects: factors of the place of residence in receiving countries (n = 6), migration-context interactions (n = 10), policy environments (n = 15) and cultural factors (n = 1). Results show the negative impacts of post-migratory contexts on physical health and mortality and the favourable impacts on child health. Impacts on mental health are mixed. Analyses of policy contexts indicate the negative impacts of restrictive migration and social policies on healthcare utilization, mental health and mortality as well as the positive effects when restrictions are lifted. Conclusions Natural experiments can serve as powerful tools in disentangling the effect of context on health and reducing bias through self-selection. Results demonstrate the negative impacts for health which lie at the nexus of migration and neighbourhood disadvantage. At the same time, studies uncover the potential of health, welfare and visa programs to counteract such disadvantages and create healthy post-migratory contexts. With careful consideration of causal pathways, results from migration contexts can serve as a magnifying glass for effects of context in other population groups. Key messages • Natural experiments can serve as powerful tools in disentangling the effect of context on health and reduce bias through self-selection. • Results show the negative impacts for health that lie at the nexus of migration and neighborhood disadvantage, as well as the potential of inclusionary policies to counteract them.


Background:
Many studies on contextual health effects suffer from compositional bias and selective migration into neighbourhoods. Longitudinal natural experiments have the potential to overcome these limitations, and there are several opportunities for this research design in the migration context. We aimed to synthesize evidence from natural experiments among migrants studying the effect of contextual factors on health and healthcare.

Methods:
Peer-reviewed literature in English or German was systematically searched in four major databases in December 2021. Following systematic abstract-and fulltext-screening, 32 studies were included for analysis. Evidence on contextual impacts on physical and mental health, mortality, and healthcare was narratively synthesized and quality appraisal conducted.

Results:
We found four types of contextual health effects: factors of the place of residence in receiving countries (n = 6), migrationcontext interactions (n = 10), policy environments (n = 15) and cultural factors (n = 1). Results show the negative impacts of post-migratory contexts on physical health and mortality and the favourable impacts on child health. Impacts on mental health are mixed. Analyses of policy contexts indicate the negative impacts of restrictive migration and social policies on healthcare utilization, mental health and mortality as well as the positive effects when restrictions are lifted.

Conclusions:
Natural experiments can serve as powerful tools in disentangling the effect of context on health and reducing bias through self-selection. Results demonstrate the negative impacts for health which lie at the nexus of migration and neighbourhood disadvantage. At the same time, studies uncover the potential of health, welfare and visa programs to counteract such disadvantages and create healthy post-migratory contexts. With careful consideration of causal pathways, results from migration contexts can serve as a magnifying glass for effects of context in other population groups.

Key messages:
Natural experiments can serve as powerful tools in disentangling the effect of context on health and reduce bias through self-selection.
Results show the negative impacts for health that lie at the nexus of migration and neighborhood disadvantage, as well as the potential of inclusionary policies to counteract them. Background: Regular mammographic screening can reduce breast cancer morbidity and mortality. Participation rates are suboptimal in Australia's fully funded biennial breastscreening program (BreastScreen) for women aged 50-74. Despite obesity being a well-established risk factor for post-menopausal breast cancer, cross sectional data suggests obesity may be a risk factor for non-participation in recommended screening, due to adverse screening experiences. This research aimed to ascertain the link obesity and non-participation by using data linkage of routinely collected data.

Methods:
Data for women age eligible for breast screening were linked between the NSW Cancer Registry and the Australian Longitudinal Study of Women's Health (ALSWH) to create a cohort of women who either participated in screening as recommended or not. Women from the 1946-1951 ALSWH birth cohort were included in the study. These women reported BMI via 8 survey waves. The primary outcome was adherence to breast screening measured by frequency of screening over the follow-up period (1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016). Unadjusted risk ratios were calculated using mixed-effects logistic regression for the association between BMI and screening participation.

Results:
The study included 2804 linked records of age eligible women (mean age of 52.37 [SD 5.47]). 22.8% of the cohort were obese (BMI>30kg/m2). Obesity was significantly associated with non-recommended screening participation (screening within 3 years of last breast screen); odds ratio 1.63 (95% confidence interval 1.32 to 2.00, p < 0.0001).

Conclusions:
Obesity has a significantly impact on recommended participation in a nationally provided breast screening program, despite obesity being a risk factor for post-menopausal breast cancer. Optimising participation among higher risk and underscreened women in under utilised breast cancer screening programs is warranted. Development of targeted interventions to increase screening participation among these higher risk women is needed. Key messages: Women living with obesity and less likely to participate in recommended breast screening. Targeted interventions are needed to optimise participation in breast screening to ensure these higher risk women are not at higher risk of adverse outcomes due to breast cancer.