These results extend our previously reported findings from this cohort showing that cigarette smoking represents an important environmental risk factor for RA among African Americans. These findings illustrate that the association of smoking with RA disease risk in African Americans may be impacted by genes encoding enzymes involved in the metabolism of smoking-related toxins. Based on our prior observations, we have estimated the attributable risk of RA due to heavy smoking to be 16% [12
], a risk that may be much higher in subgroups defined by specific DME genotypes. For instance, among African Americans carrying the NAT2 rs1208
polymorphism, the attributable risk of RA from heavy smoking approaches 30%.
While our approach is consistent with recent reports examining gene-smoking interactions in RA, it is important to acknowledge that some uncertainty remains regarding the most appropriate manner in which to model gene-environment interactions [45
]. In this study, we have examined measures of both additive and multiplicative interaction. Multiplicative interaction, often referred to as ‘statistical’ interaction, refers to the inclusion of a product term in regression analyses in order to generate an optimal ‘fit’ of the data in a given risk model. The absence of multiplicative interaction, such as in this study, does not exclude the existence of highly relevant biologic interactions [45
]. The present study suggests that one or more ‘pathways’ to RA development in African Americans involve the simultaneous presence of two risk factors in this case, heavy smoking and presence of a select DME gene polymorphisms.
Ours is the first study to date involving an African American population to investigate the role of DME
gene-smoking interactions in RA susceptibility. Perhaps not surprisingly, our findings share both similarities and differences with the few studies that have been conducted in populations of European ancestry. In their prior case-control study involving 82 individuals with RA, Pawlik and colleagues found that NAT2
alleles associated with the “slow acetylator” phenotype were ~five-fold more common in RA than in controls [23
]. This study did not report smoking rates nor were results provided specific to any possible gene-smoking interactions. In contrast to our results in African Americans showing no evidence of such an interaction, prior studies in populations of European ancestry have yielded evidence of an interaction between the GSTM1-null
genotype and smoking in RA susceptibility [22
]. More recently, however, investigators from the Nurses’ Health Study found no evidence of a GSTM1-null
-smoking interaction in disease risk in a population primarily of European ancestry [47
]. It is possible that the GSTM1-null
genotype, which is substantially less common in African Americans than in individuals of European ancestry, may exert a markedly varied impact across different study populations. Underscoring the potential importance that DME
genes might play in mediating RA risk secondary to smoking, Keenan et al recently reported evidence of significant gene-environment interactions between the GSTT1-null
genotype (a genotype not examined in the present study) and heavy smoking in their assessments of RA risk [47
It is important to point out that rather than selecting only SNPs with known biologic function, our approach involved the use of ht-SNPs with the broader goal of capturing the overall polymorphic nature of each gene of interest. Therefore, it is quite possible (even probable in select cases) that the SNPs examined and shown to interact with heavy smoking may have little functional biologic consequence. For instance, the rs9987109
SNP that showed a significant additive interaction with heavy smoking is found in the NAT2
intron and is thus unlikely to exert a functional role that increases disease risk. We anticipate that the association of this particular SNP with RA observed in heavy smokers likely relates to other regions in the gene that are in close linkage disequilibrium. It is noteworthy that the NAT2 rs1208
polymorphism is a missense polymorphism, one that is represented in many of the “slow acetylator” haplotypes that have been previously well characterized [48
]. A prior Japanese case-control study showed that NAT2
“slow acetylator” genotypes were significantly more frequent among individuals with systemic lupus erythematosus (SLE) compared to healthy controls [49
]. Results from this study also suggested a possible interaction between smoking and NAT2
with an AP of 0.50 (95% CI 0.12 to 0.88). The combination of cigarette smoking and a “slow acetylator” genotype conferred a more than 6-fold increased risk of SLE compared to the combination of never smoking and a “rapid acetylator” genotype.
These results may provide insight into the chemical compounds found in cigarette smoke that could drive heightened RA risk. Both NAT1 and NAT2 enzymes are known to be important in the metabolism of aromatic and heterocyclic carcinogens, both of which are known constituents of cigarette smoke. Moreover, NAT2 function has been implicated as a risk factor in arylamine-related carcinogenesis [50
]. Cigarette smoking represents a major environmental source of arylamine exposure in humans with urine levels that are at least twice as high in smokers as non smokers [51
]. Whether increased arylamine exposure in the context of reduced NAT2 function could help to explain the findings of this study remains uncertain.
In summary, we have found evidence of significant biologic interactions between heavy smoking and genetic variation in NAT2 in RA risk among African Americans. In addition to extending and replicating these findings in other populations including an independent cohort of African Americans, further work will be necessary to better understand the precise biologic mechanisms underlying these observed gene-smoking interactions in RA risk.