Striking variations by ethnic group in asthma in Scotland

BMC Medicine recently published research investigating variations in outcomes of asthma-related hospitalisations in Scotland, based on ethnicity. To tell us more, Aziz Sheikh and Raj Bhopal explain the impact of their findings on future research.

Asthma causes wheezing and shortness of breath long-term in over 300 million people globally. A systematic review and meta-analysis published in 2005 showed marked ethnic variations in asthma in the UK, with ethnic minority origin individuals being, surprisingly, at increased risk of hospitalisation when compared to the White ethnic group population.

This systematic review was limited as the underpinning studies tended to be small, and used very broad ethnic groupings (e.g. South Asians). The Scottish Health and Ethnicity Linkage Study (SHELS) provided an ideal vehicle to undertake important follow-on work.

SHELS is a Scotland-wide retrospective cohort that has – possibly uniquely – linked national census data with hospital and mortality records; and, in a pilot study, recently linked these data to GP primary care records. The linkage to census data was important because ethnicity tends to be very poorly recorded in National Health Service (NHS) records; it is however a mandatory question in the census and so completion rates of self-reported ethnicity were very high.

Ethnic variations in disease outcomes

SHELS covers over 90% of the Scottish population, making this study on 4.62 million people by far the largest investigation into within country ethnic variations in asthma outcomes ever undertaken. SHELS has answered important questions relating to ethnic variations in disease risk and outcomes across a range of conditions responsible for high morbidity and/or mortality in Scotland. This new study on asthma is an important addition to this body of knowledge.

Our findings demonstrated striking variations in risk of hospital admission and/or death (henceforth, admission for simplicity) for asthma by ethnic group. The higher rate of age-adjusted admissions shown by the 2005 review in people of South Asian origin, has now been shown to be driven primarily by Pakistani-origin males and females.

This kind of ethnic difference is not simply due to migration related factors, or low socio-economic status.


The Chinese ethnic group has a similar migratory history to the Pakistani population and is not wealthier. Notably, we also found that Chinese-origin males and females had by far the lowest rates of asthma admission.

There were (thankfully!) relatively few readmissions and in particular deaths, so we were unable to obtain clear answers about how these varied by ethnic group. The data available suggested that there was far less ethnic variation in the risk of hospital readmission for asthma.

We investigated some factors that may have contributed to these variations in risk of asthma admission. We found that statistical adjustment for country of birth and socioeconomic factors did not greatly affect these findings. Clearly, much more detective work is required to unravel the causes of these ethnic variations.

Further research

We aim to expand our linkages to primary care data as it is here that most asthma care is provided. We are, for example, interested in whether variations in immunisations (e.g. to influenza), differences in prescribing behaviour (e.g. bronchodilator to inhaled corticosteroid ratio) and medication adherence (particularly with preventer treatments) or self-management ownership can help explain these differences in hospitalisation.

We are also interested in investigating whether there is any clustering of poor outcomes at the GP level and if so whether practice organisational characteristics may be important factors.

We want to understand why the Chinese people have such good outcomes and whether there are any potentially transferable lessons to other populations. In addition to the primary care linkage work described above, this can usefully be explored using both qualitative methods and hypothesis testing research designs.

We are planning a new linkage to Census 2011 data which, if successful, will allow us to see if there has been any improvements in asthma outcomes across Scotland overall and reductions in the variations between ethnic groups. This would provide a unique opportunity ‘newer’ ethnic groups in Scotland such as the Gypsy Traveller-, Arab- and Polish-origin populations.

Finally, in addition to describing patterns of asthma by ethnic group, our work aims to stimulate new ideas on the causes and consequences of asthma and how control can be achieved in our increasingly multi-ethnic populations.

View the latest posts on the On Medicine homepage