Table 5.
Factors associated with reported food-associated anaphylaxis among SAPPHIRE participants stratified by race and asthma status*
| Factors | All participants (n=6,949) | African Americans (n=4,150) | European Americans (n= 2,391) | Individuals with asthma (n= 5,609) | Individuals without asthma (n=1,340) | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| aOR (95% CI)† | P-value | aOR (95% CI)† | P-value | aOR (95% CI)† | P-value | aOR (95% CI)† | P-value | aOR (95% CI)† | P-value | |
| Age (per 10-year increment) | 1.03 (0.97–1.09) | 0.280 | 1.00 (0.93–1.07) | 0.963 | 1.10 (0.98–1.23) | 0.120 | 1.03 (0.97–1.09) | 0.373 | 1.10 (0.87–1.40) | 0.448 |
| Sex (female referent) | 0.81 (0.68–0.96) | 0.016 | 0.84 (0.68–1.03) | 0.096 | 0.63 (0.43–0.90) | 0.014 | 0.79 (0.66–0.95) | 0.011 | 1.06 (0.54–1.98) | 0.860 |
| Race (European American referent) | 1.99 (1.66–2.39) | 2.25×10−13 | -- | -- | -- | -- | 1.93 (1.60–2.33) | 7.63×10−12 | 3.85 (1.63–11.32) | 0.005 |
| BMI (per unit increase) | 1.01 (1.00–1.01) | 0.223 | 1.01 (1.00–1.02) | 0.252 | 1.02 (1.00–1.04) | 0.023 | 1.01 (1.00–1.01) | 0.179 | 0.99 (0.95–1.03) | 0.639 |
| Asthma status | 4.36 (3.23–6.02) | 1.28×10−20 | 3.94 (2.82–5.68) | 1.50×10−14 | 7.23 (3.27–20.50) | 1.62×10−5 | -- | -- | -- | -- |
| Proportion African ancestry (per 10% increase)‡ | ND | ND | 1.00 (0.90–1.10) | 0.936 | ND | ND | ND | ND | ND | ND |
SAPPHIRE denotes Study of Asthma-related Phenotypes and Pharmacogenomic Interactions by Race-ethnicity; aOR, adjusted odds ratio; CI, confidence interval; BMI, body mass index; and ND, not done.
Outcome was any reported food-associated anaphylaxis. Patient with symptoms consistent with NIAID/FAAN clinical criteria in at least two categories (see Methods Section) were categorized as having food-associated anaphylaxis. Race categories were based on participant self-report and asthma status was based on a documented clinical diagnosis with concordance by patient report.
Logistic regression was used to adjust for all of the variables shown. Odds ratio for age were reported for 10-year increments while the models adjusted for each year increase. Sex (female=0, male=1), race (European American=0, African American=1), and asthma status (no history=0, positive history=1) were all coded as dichotomous variables.
Proportion of African ancestry was assessed using existing whole genome sequence data on SAPPHIRE participants and the YRI and CEU reference populations from the 1000 Genomes Project. African ancestry was only assessed in individuals who identified as African American, and was only included in these stratified models. Odds ratios for ancestry were reported for 10% increments of African ancestry while the models adjusted for each percentage increase.