Large ethnic differences in COPD prevalence – what is the role of smoking intensity?

Problem

COPD risk has been found to be lower in black people (OR 0.44; 0.39 to 0.51) and Asian people (0.82; 0.68 to 0.98) than white people, when adjusting for age, sex, deprivation and smoking status. Ethnic variation in smoking intensity was assessed for its role in these differences.

Approach

Primary care data from Lambeth, Newham, Tower Hamlets and Hackney were used for this retrospective, cross-sectional study. Mean values for cigarettes per day (CPD) were compared between ethnic groups. Adjusted mean CPD was calculated, adjusting for age, sex and deprivation. COPD risk was compared between ethnic groups using multiple logistic regression, controlling for age, sex, deprivation, and either smoking status (never, current, ex-smoker) or smoking intensity (never, light 0 – 9 CPD, moderate 10 – 19 CPD, or heavy smoker 20+ CPD). Smoking intensity categories were used rather than raw CPD values so that never smokers could be meaningfully accounted for in the analysis.Sensitivity analyses: analyses were repeated assessing spirometry-confirmed COPD, excluding unconfirmed COPD; analyses were also repeated using imputed values for missing smoking intensity data.

Findings

1,000,388 over-18s were included. 82% of current smokers and 33% of ex-smokers had CPD recorded. Adjusted mean CPD was significantly lower in black smokers (7.0; 7.0 to 7.1) and Asian smokers (7.2; 7.1 to 7.2) than white smokers (10.3; 10.3 to 10.3).When controlling for smoking status (never, current, ex-smoker), risk of COPD was significantly lower in black (0.42; 0.39 to 0.45), Asian (0.77; 0.73 to 0.82), mixed (0.58; 0.51 to 0.67) and Chinese (0.52; 0.46 to 0.59) ethnic groups compared to white people. Adjusting for smoking intensity (never, light, moderate, heavy smoker) gave very similar results for COPD risk for all ethnic groups compared to white people. Sensitivity analyses assessing spirometry-confirmed COPD and then with imputed missing values gave very similar results.The white Irish subgroup was the only subgroup more likely than white British to have COPD on subgroup analysis, when adjusting for both smoking status or intensity.

Consequences

COPD risk is lower in all minority ethnic groups than the white group. Ethnic minority smokers in London are lighter smokers than white smokers. Adjusting for smoking intensity rather than smoking status did not significantly affect ethnic COPD risk. This suggests that the observed ethnic differences in COPD prevalence are not explained by ethnic differences in smoking intensity. Sensitivity analyses gave similar results, suggesting that misdiagnosis and missing data were not the causes of ethnic differences.Causes of ethnic differences in risk of COPD other than smoking intensity must be sought. These might include biological differences in metabolism and addictive potential of nicotine, ethnic differences in smoking behaviour (eg number of inhalations, depth of inhalation), and biological differences in susceptibility to noxious effects of cigarette smoke.