How does Migratory Status Affect Non-Communicable Disease Risk in a Global Context? A Systematic Review

Critical Appraisal

Eikemo TA, et al. [7] was focused in that it specifically targeted risk ratios and NCD rates in a specific group of migrants in Greece. The authors used a reasonable sample size (1332) that included migrants of non-Albanian origin. Their methods are somewhat convoluted by the fact that they grouped NCDs into categories and tabulated overall prevalence for several different afflictions within each category. The study is also subject to reporting bias, given that prevalence was based on self-reporting. This is particularly relevant given the study aims to study NCD reporting itself. Aung TNN, et al. [8] focused on a more general notion of risk factors and NCDs, but again these were based on self-reported measures, not objectively measured. Their methodology seems sound, however, especially considering their simple odds ratio analysis, and Pearson correlation calculations.

Naicker A, et al. [9] focused on the Indian migrant subset. Their data is particularly sound due to the fact it was clinically obtained. The “random sampling” method is not explained, however, casting a shadow of doubt over their 250 individuals’ results. At the same time, their exclusion criteria of individuals diagnosed with hypertension or diabetes mellitus is not explained. The study, however, astutely considered Asian parameter cut-offs instead of Western standards, which the authors posit more accurately reflects the Indian subpopulation. Bilal PI, et al. [10] aimed to investigate associations between acculturation, perceived ethnic discrimination (PED) and elevated blood glucose level (EBGL). Their sample size is not exceedingly large, but the clinical measurement of blood glucose was conducted appropriately. The PED questionnaire is a novel idea, but does present a level of subjectivity to the data, given that it uses a 1-5 scale for its 25 questions. This in turn poses difficulties regarding statistical analysis.

The sample used by Shah SM, et al. [11] consisted of 1375 individuals, with a participation rate of 76%, so drop- off rates were successfully mitigated. One of the benefits of its migrant participation was the fact that Indian, Pakistani and Bangladeshi individuals were clinically sampled, and by law all migrant workers need to undergo a communicable disease screening test, in this case encompassing TB and HIV. Their sociodemographic survey was offered in several languages, mitigating response bias. The sample used for the Italian risk factor surveillance study was by far the largest of all the articles reviewed (over 200,000) [12]. Their investigation was restricted to country of origin, assimilation process and risk factors, making it a highly relevant focused study. With coverage of over 90% of the Italian population, this is considered highly reliable data.

The English obesity study by Akhter N, et al. [13] is highly focused, as it investigated predisposition towards obesity- related NCDs exclusively, in first and second-generation Bangladeshi migrants in England. The fact that their questionnaire was delivered face-to-face also guarantees that the data are likely to be reliable. Their study, however, might be subject to region-specific biases, as it took place in five boroughs of London and Northeastern England. The study by Joy EJ, et al. [14] is unique in that it was the only study included in this review that represented a sample of internal migrants. I decided to include this study because of two reasons: firstly that it posed a novel question, i.e. dietary patterns and their association with micronutrient under nutrition; secondly, that as explained previously, interstate migration in India is for this review’s intents and purposes akin to international travel if one contemplates both the distance involved, and the cultural, religious and linguistic differences that abound across Indian states. The Indian Migration Study data is a treasure trove of valuable data that should not be ignored, especially considering how dietary factors seem to influence health outcomes as shown in the results section. Their sample size is quite large, and their methodology is rather simple, particularly when determining the dietary patterns themselves.

I believe that the quantitative data presented by the studies in question are generally sound, with some caveats I will discuss in the synthesis below. The information is relevant and up to date, and sources paint a coherent picture in terms of how migratory status affects non-communicable disease risk in a global context.

Synthesis

The eight studies reviewed identify either risk factors or social determinants of health that have significant impacts on migrant health both at present and into the next decades, especially considering a post-COVID world where migration significantly ticks up. A few overarching themes have been identified, all with a quantitative backing, and will be developed as follows.

Risk Factors

Lifestyle factors among migrants were identified, wherein large percentages of the migrant population in Thailand are smokers, drinkers, obese or do not exercise [8]. These are suggested to play a future role in the host country’s premature death burden. Generalizations regarding lifestyle factors are difficult to make, given a higher prevalence of smoking in South East Asia [15]. Even though studies relied on various methodologies and took place in different continents, an emergent theme of particular determinants of health and risk factors affecting migrant populations is evoked. For instance, almost the entire sampled population surveyed by Naicker A, et al. [9] were centrally obese; almost two thirds had blood pressure values above the cut-off; over a third of men and women had higher fasting blood glucose levels; and almost the entire sample recorded raised triglyceride levels. Over half of participants investigated by Shah SM, et al. [11] never had their blood pressure measured, and nearly a third had hypertension; notably, hypertension awareness increased with the increase in the duration of their residency for more than 5 years.

Behavior

Lack of exercise, alcohol consumption, smoking and central obesity were identified as major lifestyle problems [8]; one must consider the harsh conditions often faced by such workers in Northern Thailand, and whether the sample was a representative one for the purposes of this review. While the important health determinant that is socioeconomic status was lower for migrants than native-born Italians [12], the authors also detected a lower attendance in preventive services among that population. Notably, however, this population as a whole appears not too different to native- born Italians, in terms of health attitudes and behaviors perhaps because they are mostly European migrants. As determined by Joy EI, et al. [14], study participants that had increased odds of a high waist-to-hip ratio were consumers of the wheat and pulses dietary regime, while those with the greatest odds of obesity were consumers of the rice and fruit pattern, implying a strong dietary behavioral basis to NCD prevalence, at least within Indian subpopulations.

Discrimination and Integration

Albanian immigrants, particularly men could suffer increased discrimination, in turn affecting their health and their NCD reporting [7]. Participants with a lower mean marginalization score were more likely to smoke. 22.5% of the population studied by Bilal PI, et al. [10] had high level PED and EBGL concomitantly. Traditional and integration modes were thus strongly correlated with EBGL and diabetes. Interestingly, PED served as a “mediating factor” between EBGL and integration mode of acculturation, a measure of how assimilated an individual is in their adopted country. Those in that mode of acculturation were 4.2 times more likely to exhibit an EBGL. Those in the traditional acculturation modes were 80% less likely to show EBGLs, which suggests a link between health and how one perceives their host country. The study’s generalizability is hampered by its small sample size and ethnic composition, however. Migrant respondents as surveyed by Campostrini S, et al. [12] seem to be “more similar” to Italians as their stay increases. The link between discrimination, both perceived and otherwise, as well as integration/acculturation warrants further study. It would also be interesting to assess whether these regional perceived discriminatory trends would apply in other regions, or are wholly European interactions.

Gender Inequality

In one of the strongest overarching data themes, gender inequalities were presented by most authors in a clear and explicit manner. To begin with, gender inequality disfavoring women in reporting NCDs both in migrants and Greek-borns was apparent. Women bear a greater risk of not reporting NCDs, particularly Albanian women, or migrants of non- Albanian origin in Greece. Risky health behavior, as well as economic and social factors is thought to partially explain these differences in reporting [7]. The results of Eikemo TA, et al. [7] suggest that gender and the occurrence of NCDs have a relationship subject to migration status as well as ethnic origin, and as such confirm that immigrants are a diverse group across social categories. Importantly, gender and migration do not represent “fixed attributes with specific health impact”, so further research is necessary to elucidate these interactions. It also begs the question of whether pregnancy influences said risks.

Being female and uneducated was associated with a lack of exercise, while alcohol consumption was associated with male Myanmarese workers [8]; at the same time, central obesity was associated with uneducated females over 40 years of age in the same study. Campostrini S, et al. [12] also demonstrates gender-related differences, particularly regarding attending preventive services: the prevalence is higher for Italian women and migrants. Smoking, on the other hand, is more prevalent in migrant males.

The sample surveyed by Akhter N, et al. [13] showed a higher proportion of women (46% vs 10%) had a large waist circumference. According to their analysis, 42% of females had a high or very high risk of obesity-related NCDs, as opposed to 12% of males. At the same time, females had significantly higher odds of developing obesity-related NCDs, and this was particularly the case in Northeastern England. Similarly, second-generation Bangladeshis had double the odds of developing such diseases, perhaps suggesting strong acculturation, as they reflect current obesity trends in the UK. Most notably, the authors found that participants who walked over 20 minutes daily were 32% less predisposed to obesity-related NCDs [13]. The article by Joy EJ, et al. [14] also exposed gender-related patterns, including an inadequate calcium intake in 13% and 42% of females in the rice and fruit, and rice & low diversity dietary patterns respectively. All in all, findings presented by the eight selected articles provide insight into the research question. Considering the question was restricted to a global context yet most research centered on individual communities, there is ample possibility for further study concerning risk factors and NCD association analysis in a collation of migrants. This would particularly call for an international meta-analysis approach.

Key Findings

Peer-reviewed, primary empirical research evidence tells us the following about non-communicable disease risk in a global context:

• Migrant populations are heterogeneous
• Certain migrant subgroups show increased risk of developing NCDs in some circumstances
• Behavior plays a crucial role in how these risks are distributed
• Perceived discrimination is a novel factor to consider when establishing risk associations
• Gender-related differences exist in most migrant populations
• Gaps in the research exist

In Context

Identifying and quantifying risk factors for migrants is a vital aspect of health science and requires continuous investigation and monitoring. This review identifies several such risk factors, namely: sex, smoker status, obesity, PED, integration, alcohol consumption, blood pressure, blood glucose level, and notably one protection factor: walking 20 minutes or more daily. While research usually focuses on risk factors, protective factors are just as important, and research such as that undertaken by Akhter N, et al. [13] was refreshing in that it considered such a notion. One notable absence from this particular set of articles was the issue of refugees, and the effect this circumstance has on both risk factors and health outcomes. This was a consequence of my selection criteria, but it should be noted that it likely would be relevant for a meta-analysis. The findings of this review are important in that they provide fertile ground for comparison and appropriate planning, as some of the conclusions might be applicable in different countries or localities, at any age, and with particular importance given to gender disparities. Risk factors can easily be incorporated into modeling and subsequently applied as part of public health measures, such as Smit M, et al. [16] and Brinks R, et al. [17] Suggested, albeit in a slightly different milieu. Finally, as suggested by Aung TNN, et al. [8], these factors could create a wave of premature deaths in the host country, with wide ranging future implications for all involved parties.