Commentary: Mona Okasha

Mona Okasha takes you through the paper and explains what it means

Descriptive epidemiology is the area of medicine concerned with changes in patterns of disease, according to time, place, and populations. These changes may seem far removed from clinical practice, but an understanding can provide important clues about determinants of health and how to help prevent disease. The paper describes changes in the incidence of lung cancer in Leicestershire over time, according to ethnic group and deprivation.

Data for this study were obtained from the Trent cancer registry. Regional cancer registries aim to collect information on all people in their area who are diagnosed as having cancer, and submit it to the national cancer registry.

The study was of people who had had lung cancer diagnosed in Leicestershire from 1 January 1990 to 31 December 1999. The authors were particularly interested in lung cancer in south Asian—Indian, Pakistani, and Bangladeshi—people. Ethnic grouping (south Asian or not) was assigned according to name. Incidence of lung cancer is the number of new cases divided by the total population at risk in a given time period. This was calculated separately for south Asians and others. Then one was divided by the other to calculate the relative incidence or incidence ratio. For similar incidences in both groups, the ratio would be close to one. If the incidence was higher in south Asian people than in the other group then the ratio would be more than one. Confidence intervals estimate the precision of the incidence ratio.

The authors found that the incidence ratio of lung cancer comparing south Asians with others was 0.41—that is, the risk of lung cancer is lower in south Asian people. The patterns of disease over time, however, were different between the two groups. The incidence of lung cancer in south Asian men increased between 1990-4 and 1995-9, but that of other men decreased over the same time period. Among women, lung cancer became more common over the specified time period in both south Asian people and in others.

The authors looked at whether the changes in the incidence of lung cancer over time differed depending on deprivation category. These results are shown in the figure. This time, the incidence ratio compares the incidence in the later time period (1995-9) with the incidence in the earlier time period (1990-4). An incidence ratio of greater than one, therefore, indicates an increasing incidence over time; an incidence ratio of less than one indicates a falling incidence over time. The greatest decrease in incidence was evident in the more deprived men, who were not south Asian, whereas the greatest increase in incidence was in south Asian men, the least deprived. Since the confidence intervals (vertical bars) in the graph are wide, however, the estimated incidence ratios for south Asian men are imprecise. In women, there seemed to be no difference in change in incidence according to deprivation category.

The authors found evidence of an increasing incidence of lung cancer in south Asian men and a decreasing incidence of lung cancer in men from other ethnic groups. In women, the data suggested increasing incidence in both ethnic groups, but the confidence intervals were wide for south Asian women.

The message on the packets of dates

Possible explanations

Why might lung cancer incidence be rising among south Asian men? One of the most obvious reasons may be to do with smoking behaviour. A recent survey found that smoking rates among Pakistanis and Indians are similar to the white British population.¹ However, Bangladeshi men (49%), particularly for those aged 30-49 years (54%) and 50-74 years (70%), smoke heavily.

Diverging patterns of smoking in men between ethnic groups may explain the diverging lung cancer rates seen in this study. But that leads to question why smoking continues to be so popular among Bangladeshi men. The health and lifestyles survey identified a serious lack of knowledge of the health risks of smoking, and British south Asian people—with only about half of Bangladeshis, Pakistanis, and Indians—reported that they were aware that smoking is associated with lung cancer.¹

The NHS Asian Tobacco Information Campaign, set up in 2001, funds 19 local projects to the tune of £1m ($1.6m; €1.5m) to raise awareness of the health risks of smoking among south Asian communities in Britain. One study in Bradford distributed 5000 packets of dates to people attending mosques, with details about a stopping smoking service in Urdu and Bengali and a message about smoking and health. The campaign also funds tobacco helplines in five Indian languages.

The tobacco industry is notorious for advertising targeted at vulnerable groups. This has been exposed in relation to young people. Strategies to promote smoking cigarettes among more deprived people include loyalty schemes, which encourage smokers not to quit. From the results that Smith and colleagues report, it may be that tobacco advertising is targeted at south Asian people. As the authors emphasise, programmes to encourage quitting need to include all sectors of the community, and the perception that preventing lung cancer is not a health priority for south Asians must be reversed.

1. Black and Minority Ethnic Groups in England: The Second Health and Lifestyles Survey. Health Education Authority: 2000.