We wish to thank Tillin et al. (1) for their comments on our study (2). We agree that the absence of information on smoking status from population-level health care administrative data limits our ability to adjust for this important cardiovascular risk factor, and that this and other unmeasured risk factors could confound the relationships we found between ethnicity and cardiovascular complications or mortality. However, as noted in the article, we conducted sensitivity analyses to see how imbalanced an unmeasured confounder would have to be to explain the observed results. A risk factor that quadrupled the risk for mortality would have to have a prevalence of 5% in the South Asian population and 30% in the European population to explain the observed reduced hazard for mortality. Compared with this hypothetical risk factor, cigarette smoking is both less potent as a risk factor and less imbalanced in prevalence in the Ontario population (3,4); hence, it is unlikely that it explains the striking difference in mortality observed in our study.
The discrepancy between our results and those of Tillin et al. in the SABRE study (5) highlights an important message often lost in studies of minority populations: broad ethnic categories such as South Asians are not homogeneous, and there may be important differences in diabetes and other cardiovascular risk factors within members of these groups (6,7). The differing migration histories to the U.K. and to Canada mean that the South Asian populations in each country have different regional origins, generation status, cultural and religious traditions, socioeconomic position, and acculturation and integration after migration. Hence, it should not be surprising that their health status and disease risk are also different. Comparing and contrasting research findings from different countries can shed light onto the biological, behavioral, socioeconomic, and health care system differences between different immigrant populations to better understand their health risks.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.