����Estrogen hormone carries out its physiological functionsthrough its receptors. ESRa is expressed by many of the trigeminalneurons. The gene encoding estrogen receptor alpha, ��ESR1��localized in 6q25.1 locus is the most widely studied hormonalpathway gene in migraine. We analyzed two intronic ��intron 1��and two exonic ��exons 4 & 8�� polymorphisms of this gene.
����However, we could establish a significant relationship of a singleintronic polymorphism ��rs2234693�� in migraine with aura patients.
����The existing literature reveals conflicting reports [8,13]. Thenearby SNP rs9340799 was not associated with migraine. The lackof association of the exonic polymorphisms is previouslydocumented by several authors [8,9]. These polymorphisms donot result in any change in amino acid sequence. This furtherstrengthens our findings.
����On the contrary, a recent meta analysis of ESR1 polymorphismsreported significant associations of exonic polymorphisms .
����However, among six studies [5,8,9,10,11,13] on ESR1 rs1801132polymorphism, only two studies [5,11] reported significantassociations with migraine. Similarly, only a single study 
����among four reports on ESR1 rs2228480 polymorphism showedsignificant findings [9,10,11,12]. Thus, our results are inaccordance with majority of the published results. The discordancewith the results of meta analysis could be explained by thelarge variations in sample size of the included studies, which couldtilt the final outcome on the studies with larger samples.
����The intronic polymorphisms are at a distance of 50 bp andhence are in strong linkage disequilibrium. The polymorphismswere not in LD with other polymorphisms. Similarly, Colson et al.
���� did not find LD between exon 8 marker with intronic and exon4 markers. However, the authors reported LD between intron1and exon 4 markers. But the distance ����102 kb�� between themarkers could explain our findings.
����We were able to identify a particular haplotype group showingprotective effect of migraine. This particular group consists ofvariant allele ��G�� of one intronic polymorphism ��rs9340799��whereas the wild alleles of other polymorphisms. Hence, it showsthe antagonistic role of the two intronic polymorphisms. Literaturereview also reveals that XbaI*G ��variant�� and PvuII*C ��wild�� SNPsare associated with elevated serum estradiol ��E2�� production [6,7].
����Hence, these alleles may impart a protective effect. The results ofour study could be justified by the role of amino terminal portionin transcriptional activation and its association with various pathological conditions . In other words, we report that theeffect of ESR1 polymorphisms on migraine risk diminish fromamino to carboxy terminal.
����ESRb is mostly expressed by inhibitory neurons in cerebralcortex. Therefore, it regulates cortical excitability and spreadingdepression  involved in migraine pathophysiology. It isencoded by ESR2 gene located in 14q22-24. A promoterpolymorphism ��rs1271572�� has a functional effect on transcriptionalfactor binding and gene expression . The other selectedpolymorphism ��rs1256049�� is present in exon 5 which results in asilent change in amino acid. However, we were unable to find anassociation of these polymorphisms in migraine. No previousreports are available for these polymorphisms in migraine.
����We also studied the genetic interactions. Our study supports thesignificant interactions between CYP19A1 and ESR polymorphisms.
����Thus we infer that though individual polymorphisms haveno role but in the presence of other genetic variants, these could bea possible risk factor ��as in case of ESR1 intronic polymorphisms����
����We obtained antagonistic interaction of ESR1 rs9340799 with theCYP19A1 39UTR polymorphisms. This SNP showed an increasedrisk with rs10046 polymorphism whereas a protective effect withrs4646 polymorphism. It implies the role of CYP19A1 polymorphismsin defining the effect on migraine susceptibility. Similarly,we also analyzed gene �C gene interaction of haplotypes. In this casealso antagonistic role of ESR1 H haplotype was observed withCYP19A1 haplotypes. These findings further strengthen the effectof CYP19A1 polymorphisms on migraine susceptibility. This isfurther supported by the fact that haplotype H consists of both thevariants of intronic polymorphisms shown to have opposite effecton migraine susceptibility. So the individual effect of the haplotypeis more or less nullified. In other words, the same haplotype ofESR1 may show protective effect or risk depending on theCYP19A1 haplotypes. Hence, we conclude that CYP19A1 polymorphismsand haplotypes are the major causal factors formigraine susceptibility. However, their effect is enhanced ongenetic interaction with ESR1 polymorphisms and haplotypes. Noother group has analyzed or reported such genetic interactions.
����The variants of associated polymorphisms ��rs10046, rs4646 andrs2234693�� are found to affect estrogen levels ��either high or low����
����So, the bimodal effect of estrogen is implicated in migrainepathophysiology. Estrogen withdrawal is the most well acceptedtheory especially in migraine without aura. It is attributed to thefact that frequency of attacks increase with menstruation and useof oral contraceptives. In addition, migraine attacks subside during pregnancy and with menopause – two phases of high estrogenlevels in women. On the other hand, high estrogen levels influencegene expression and signaling, finally leading to inflammatory pain. Estrogen enhances cortical spreading depression involved inthe pathophysiology of migraine with aura . Hence abruptestrogen decline as well as abnormally high estrogen levels leads totrigeminal pain .
����Our study reports various significant findings. First of all, we arethe first group to report significant associations of CYP19A1 39UTR polymorphisms. Secondly, we identified protective haplotypegroups in ESR1 gene. Thirdly, we defined significant geneticinteractions between CYP19A1-ESR1 gene polymorphisms.
����Fourthly, we were also able to show significant genetic interactionof haplotypes. Finally, we were able to replicate all the significantfindings. Thus our results do not lack validation as is the case ofother association studies. In addition, the data of both the cohortswere pooled using statistical tests ��Fisher��s method and Mantel-Haenszel test���� So, in contrast to direct pooling, there is no need ofadditional adjustments or corrections. We are the first group to usesuch tests for validating data in defining role of hormone relatedgenes in migraine.
����In the era of genome wide association studies ��GWAS���� ourcandidate gene case-control study suffers from limitationsespecially in terms of sample size although our study achieved80% power.