Background　Sepsis　is　a　critical　and　life-threatening　clinical　syndrome.　Abnormal　plasma　lipid　levels　are　associated　with　disease　progression.　However,　causal　links　remain　uncertain　due　to　limitations　of　observational　studies.　Methods　This　study　conducted　a　comprehensive　two-sample　Mendelian　randomization　(MR)　analysis　utilizing　publicly　available　pooled　genome-wide　association　study　(GWAS)　data　on　179　plasma　phospholipid　subtypes,　731　immune　cell　phenotypes,　and　genetic　data　related　to　sepsis-related　mortality　outcomes.　The　causal　associations　between　plasma　lipids,　immune　cell　phenotypes,　and　the　risk　of　sepsis-related　death　were　investigated.　Results　MR　results　demonstrated　significant　causal　associations　between　five　subtypes　of　phosphatidylcholine　(PC)　and　one　sphingomyelin　(SM)　and　the　clinical　outcome　of　28-day　mortality　in　sepsis　patients.　Notably,　PC(16:0_18:2),　PC(18:0_18:1),　PC(18:0_20:4),　PC(O-16:1_16:0),　and　SM(d34:0)　demonstrated　a　protective　effect　in　reducing　the　risk　of　28-day　mortality(P　＜　0.05).　Conversely,　PC(16:1_20:4)　was　associated　with　an　increased　risk　of　28-day　mortality　in　sepsis　patients(P　＜　0.05).　Comprehensive　sensitivity　analyses　confirmed　the　robustness　of　these　results.　The　inverse　Mendelian　randomization　analysis　did　not　reveal　a　significant　causal　effect　of　sepsis　on　the　aforementioned　lipid　profile.　Mediation　analysis　identified　two　key　immune　cell-mediated　factors,　including　circulating　CD39　on　CD39　+　CD8br　cells　and　CX3CR1　on　CD14+　CD16+　monocytes,　as　mediators　of　the　reduced　risk　of　sepsis-associated　mortality　due　to　PC(O-16:1_16:0),　with　mediation　proportions　of　17.4%　and　13.6%,　respectively.　Conclusions　Our　study　supports　a　causal　relationship　between　plasma　lipid　metabolism　disorders　and　sepsis-associated　mortality　and　provides　new　insights　into　the　critical　role　of　immune　cell　alterations　induced　by　lipid　metabolism　disorders　in　sepsis　progression.　　©　2025　by　the　Shock　Society.