Lack of replication of higher genetic risk load in men than in women with systemic lupus erythematosus

Abstract

Introduction: We aimed to replicate a recent study which showed higher genetic risk load at 15 loci in men than in women with systemic lupus erythematosus (SLE). This difference was very significant, and it was interpreted as indicating that men require more genetic susceptibility than women to develop SLE.

Methods: Nineteen SLE-associated loci (thirteen of which are shared with the previous study) were analyzed in 1,457 SLE patients and 1,728 healthy controls of European ancestry. Genetic risk load was calculated as sex-specific sum genetic risk scores (GRS(s)).

Results: Our results did not replicate those of the previous study at either the level of individual loci or the global level of GRS(s). GRS(s) were larger in women than in men (4.20 ± 1.07 in women vs. 3.27 ± 0.98 in men). This very significant difference (P < 10(-16)) was more dependent on the six new loci not included in the previous study (59% of the difference) than on the thirteen loci that are shared (the remaining 41%). However, the 13 shared loci also showed a higher genetic risk load in women than in men in our study (P = 6.6 × 10(-7)), suggesting that heterogeneity of participants, in addition to different loci, contributed to the opposite results.

Conclusion: Our results show the lack of a clear trend toward higher genetic risk in one of the sexes for the analyzed SLE loci. They also highlight several limitations of assessments of genetic risk load, including the possibility of ascertainment bias with loci discovered in studies that have included mainly women.

Figure 1

Sum of genetic risk scores in women and men with systemic lupus erythematosus. The y-axis represents percentages of participants in each of the two groups with the sum of genetic risk scores (GRS_s) at the indicated intervals along the x-axis. This analysis was done with GRS_s obtained for each of the 1,247 women and 125 men with systemic lupus erythematosus (SLE) and 100% genotype success at the 19 SLE loci. The GRS_s of each patient with SLE is the sum of the products of the natural logarithm of the sex-specific OR by the number of risk alleles at each locus carried by the patient, as described by Hughes et al. [9]. Histograms and distance-weighted least-squares fitting lines for women (light gray bars and discontinuous line) and for men (dark gray bars and continuous line) are shown.