Immigration and diabetes
There are more people on the move around the world than ever before, so it is crucial to understand the specific factors that affect immigrants’ health. An expert panel discussed the latest research on immigration and diabetes at the recent American Diabetes Association’s 83rd Scientific Sessions. Dr Susan Aldridge reports.
One theme that often comes up when discussing the health of migrants is the ‘healthy immigrant’ paradox, whereby a recent migrant has better health than a native of the country they are arriving in. This benefit tends to disappear over time, with future generations’ health becoming more like that of native people.
Megha Shah, Assistant Professor at Emory University School of Medicine, took up the discussion with reference to the Hawaii-Los Angeles-Hiroshima study. Started in 1970, this study showed how adapting to a Western diet has led to an increase in diabetes among Japanese people emigrating to the US over the last 50 years compared with those who remained in their native country. “Native Japanese people have a certain lifestyle, but when they emigrate, their risk of diabetes and obesity increases,” said Professor Shah. “But that’s only one community and we have much variation in diabetes risk depending on where someone comes from.”
The immigrant paradox does not always hold true. There are particular risk factors – namely acculturation, social context and environmental exposures – that the next generations of an immigrant family experience compared with the first generation and these will affect their health going forward.
Professor Shah went on to give some multigenerational findings in the US that highlight the heterogeneity in the health of immigrant populations. The Sacramento Area Latino Study on Aging (SALSA) showed that second and third generation Mexican adults have a greater risk of diabetes, with acculturation having little impact. However, in the California Men’s Health Study, involving a multi-ethnic cohort, there was a gradient of risk with increasing years in American culture – so there was a greater risk of diabetes for the second generation than for the first.
Finally, there was the MASALA (Mediators of Atherosclerosis in South Asians Living in America) study, which showed the impact of three different acculturation styles – separation (preference for South Asian lifestyle), integration (equal preference for South Asian and US lifestyles) and assimilation (preference for US lifestyle) – upon type 2 diabetes risk. Women in the integration group had the lowest risk of type 2 diabetes and prediabetes, at 16.4% and 29.7%, while those in the separation group had the highest risk at 29.3% and 31.5%. No such differences were observed among male participants.
Professor Shah is involved in the Community Engaged Needs Assessment of South Asians in Atlanta (CENSAA) study where they are partnering with the Indian community to collect data and explore their health needs related to cardiometabolic disease risk, looking at differences between those born in India (473 participants) and those born in Atlanta (837 participants).
A survey was launched, which showed that obesity was higher in those born in the US, although they were more likely to do the recommended 150 minutes a week of physical activity, while those born in India have more hypertension and more diabetes. They went back to the respondents for more detail and found health belief/attitudes, sense of identity and interaction with healthcare systems seemed to be key factors for the prevalence of cardiometabolic disease. “We need future cohort studies to be more diverse, spanning generations,” concluded Professor Shah. “We need to understand diabetes trajectories through the generations to provide etiological insights into the development of diabetes and inform prevention efforts.”
Tailoring prevention studies
Dr Mary Beth Weber of the Emory Global Diabetes Research Center then spoke about how diabetes prevention efforts could be tailored to take account of cultural influences. For example, we know that South Asians describe food at social events as being very important. Much of it can be very rich, but turning it down on health grounds can lead to tension, eroding the social support that is so important. “Culture is important in driving behaviours, particularly those of interest for diabetes prevention,” she said. “It impacts self-care behaviour and adherence. So, when designing interventions, do consider cultural aspects throughout the whole process.”
There are several frameworks that can guide researchers with this, such as the ADAPT guidance, which emphasises stakeholder involvement early and often to adapt an intervention to a new context (such as taking cultural influences into account). However, there are still large gaps in provision. A 2015 review of cultural adaptations to the US Diabetes Prevention Programme (DPP) found that very few actually used cultural adaptation frameworks and there was a limited rationale for modifications made. “If you take nothing else away from today, here is the number-one thing we need to do better – we need to document what we are doing to make programmes culturally appropriate because documentation is frequently absent, vague or superficial,” said Dr Weber. “So do report on how or why adaptations were made and link them to the evaluation, so we can see which adaptations are the most effective. I argue that this would be worth its weight in gold for moving this field forward.”
In addition, it is important to let the community that will be taking up the programme be your guide. “Speak early and often,” Dr Weber continued. “Educate, if needed, but listen too. Work at the ‘speed of trust’, taking the time to build that trust. That needs to be built into the timeline of studies.”
It is also important to understand your intervention and why it might work for your target population, using feedback from stakeholders. This includes understanding what the core components are that might be driving change. One example of a good prevention programme that reflects these principles is the SHAPE programme, which is an adaptation of the DPP for South Asians, with a focus on portion sizes of foods such as rice and bread. Dr Weber concluded by urging delegates to be sure to write up and publish similar examples to share with the healthcare and wider community.
Mind your language
Alicia Fernandez, Professor of Medicine at the University of California San Francisco, began her talk on the impact of English proficiency on diabetes care with a simple question to the audience. In a healthcare encounter via a professional interpreter, is patient comprehension, on average, similar to that during an encounter with a language-concordant physician? The answer is no – comprehension is a lot less when an interpreter is involved.
A study from a New York emergency room compared consultations with two state-of-the-art interpretation systems and a language-concordant physician. When patients were asked if they understood the explanation for their condition, 33% and 39% said yes for the interpreters compared with 59% for the physician. When it came to understanding the instructions for aftercare following discharge, the figures were 33%, 38% and 63%, respectively.
Dr Fernandez and colleagues have looked at this issue in Kaiser Permanente Latino patients who have diabetes and have either good English or limited English. The two groups were compared and then analysed by whether those with limited English were seen by Spanish-speaking doctors. “There was a big difference, in that those seen by a doctor who did not speak Spanish were more likely to feel lack of trust, that they were treated poorly, not shown respect and that the doctor was not listening,” she said.
The use of an interpreter instead of a language-concordant physician also has a clinical impact, leading to more medication errors, less understanding of medication labels and less understanding of follow-up and discharge instructions.
Dr Fernandez also referred to the DISTANCE study, which involved 20,000 Kaiser Permanente patients speaking five languages and including 3193 Latino participants. Ten per cent of White participants had poor glycaemic control, compared with 18% of Latino participants with good English and 21.4% with limited English. And the proportions with poor glycaemic control were 16.1% and 27.8%, depending on whether or not they consulted with a Spanish-speaking physician. No such effect was seen for blood pressure or LDL-cholesterol levels.
Clearly this issue is worth further investigation but, as Dr Fernandez put it, it is difficult to set up a randomised controlled trial on the basis of patients either seeing, or not seeing, a language-concordant physician, as they would presumably all prefer the former. However, her team was able to carry out a trial looking at the impact of switching from a language non-concordant to language-concordant doctor over the course of the study (as patients are used to seeing different physicians in clinic). At the end, the group that switched to the concordant doctor had much better glycaemic control.
“Switching to a doctor that speaks Spanish improves glycaemic control, but there was no impact of switching on blood pressure or LDL-cholesterol,” noted Dr Fernandez. “So there is something about glycaemic control that is particularly sensitive to language barriers.”
Analysis of consultations shows that less counselling goes on in discordant languages. Giving examples from her own conversations with Chinese patients who had limited English, she said the conversation is more limited and there are fewer questions, less empathy and less attention given to the patient’s point of view. So this might feed into poorer glycaemic control with less discussion about insulin and diet and, overall, less trust, comprehension and satisfaction.
Another angle to consider is that with low English literacy often comes low health literacy. Most older immigrants tend to have limited education compared with those who are younger. This can feed into comprehension of health messaging. For example, a survey on knowledge about sugar-sweetened beverages among Spanish speakers showed that while many knew sodas were high in sugar, they did not realise that horchata and agua fresca, which are popular drinks in the community, also contain a lot of sugar.
So what is the way forward? “We obviously don’t all speak all languages, so check back more on comprehension,” said Dr Fernandez. “I will force myself to make an empathic statement and make sure the interpreter makes it too. And always be sure to elicit the patients’ point of view. Arranging for language-concordant diabetes and other health educators is probably one of the most important things we can do.”
Finally, what about technology? A study of Google Translate translating into Chinese and Spanish and then back into English found potentially life-threatening errors. “If you are going to use this technology, go for simple sentences and plain language, and do not use without oversight from someone,” she warned. Furthermore, machine interpretation by phone technology is not yet ready for clinical use.
In conclusion, language barriers create difficulties for people with diabetes and their clinicians. They make healthcare less patient-centric, leading to more diabetes distress and may also affect glycaemic control. More needs to be done to facilitate language access and language-concordant diabetes care.
Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.