Overall findings
One SNP, Arg16Gly, significant association and low heterogeneity were found between Arg16Gly polymorphism and salbutamol response in two models: heterozygous model (OR=1.470, 95% CI: 1.046-2.066,P =0.026, I2=38.6%,P H=0.122; Table 2 & Figure 2.5); and additive model (OR=0.668, 95% CI: 0.502-0.889, P =0.006, I2=0.0%, P H=0.470; Table 2 & Figure 2.6). This indicated that the patients with the AG genotype may be associated with a better response to salbutamol in comparison to those with the AA or GG+AA genotypes. However, statistically evident heterogeneity and non-significant association were detected in the dominant, recessive, allele and homozygous models (all P -value for OR>0.05). Thus, it is needed to perform further subgroup analysis and identify the sources of heterogeneity.
Concerning the ADRB2 Gln27Glu polymorphism, non-significant association with salbutamol response was found in any models (allP -value for OR>0.05) (Table 3 & Figure 3). In addition, no evidence of heterogeneity was detected between studies under five models (dominant model: I2=35.2%,P H=0.186; recessive model: I2=34.9%, P H=0.189, homozygous model: I2=47.7%, P H=0.105; heterozygous model: I2=0.0%,P H=0.452; and additive: I2=0.0%, P H=0.776), but not the allelic model (I2>50%,P H<0.1). (Table 3)