Bone marrow transplantation for MHC class I deficiency corrects T-cell immunity but dissociates natural killer cell repertoire formation from function

The index datewas defined as the date of the first indicator of diabetes for cases and the date of selection for controls. The dose of ICSs was categorized as low (<251 mg/d, fluticasone propionate equivalents), medium (251-500 mg/d), high (>500 mg/d), or no use, and LABAwas classified as use (>_1 mg/d, salmeterol equivalents) or no use and measured on the index date. This study was approved by the Research and Ethics Committee of the Hôpital du Sacr e-Cœur de Montr eal and by the Commission d’acc es a l’information du Qu ebec. For additional methodological details, see this article’s Online Repository at www.jacionline.org. We identified 1001 cases of GD and selected 30,030 controls without GD among the 12,587 pregnancies of women with asthma (see Fig E1 in this article’s Online Repository at www. jacionline.org). The cases were older and more likely to have uncontrolled asthma and a history of GD (see Table E3 in this article’s Online Repository at www.jacionline.org). The generalized estimation equations models (Table I) indicate that the dose of ICSs among nonusers of LABAs was not significantly associated with GD. Moreover, adding LABAs to ICSs did not increase the risk of GD compared with ICSs alone. We also found that women who used a medium ICS dose without a LABAwere not significantly at a higher risk of GD than women who used a low ICS dose plus a LABA. Moreover, the risk of GD was found similar between women who used a high ICS dose without a LABA and those who used a medium ICS dose plus a LABA. Strengths of this study include a large sample size, avoiding recall bias, and fixing cohort entry at week 20 of gestation to avoid index event bias. Fixing cohort entry at pregnancy onset could have biased the odds ratios toward the null in case ICSs or LABAs would be associated with early pregnancy termination. Nevertheless, unmeasured covariates such as ethnicity, obesity, and familial history of GD may cause residual confounding if they are associated with the use of ICSs or LABAs. The indications for antiasthmatic medications may also introduce residual confounding because higher ICS doses or the addition of LABAs is likely to be prescribed to those with more severe asthma. However, despite this potential bias, the adjusted odds ratios associated with medium and high ICS doses and with LABAs remained close to the null and not statistically significant. Our results provide further evidence on the safety of LABAs and ICSs during pregnancy with no significantly increased risk of GD detected among women exposed to those medications. Although not statistically significant, the adjusted odds ratios associated with moderate and high doses of ICS were slightly greater than 1, and it might be due to residual confounding by obesity or asthma severity. More studies with better adjustment capacity for obesity and asthma severity are needed to further examine the association between higher ICS doses and the risk of GD in women with asthma.