A genomic structural equation modelling study elucidates shared genetic architecture of polygenic traits associated with post-intensive care syndrome

Post-intensive care syndrome (PICS) is defined by persistent psychological, cognitive and physical impairments after critical illness, yet its shared genetic basis remains unknown. We applied genomic structural equation modeling (Genomic SEM) to large-scale GWAS summary statistics for major depressive disorder (N = 217,584), post-traumatic stress disorder (N = 199,213), cognitive function (N = 257,841), memory performance (N = 152,605) and hand grip strength (N = 461,089) to construct a latent genetic factor related to PICS component phenotypes. We then performed multivariate GWAS, Bayesian fine-mapping, MAGMA, sCCA-TWAS with FOCUS, pathway enrichment, cell-type analysis and spatial transcriptomic mapping. A single factor Genomic SEM showed good fit (CFI = 0.981; SRMR = 0.168). The factor GWAS identified 1,301 genome-wide significant SNPs and 590 largely independent lead variants, including 574 novel signals not genome-wide significant in any input trait. Fine-mapping and TWAS convergently prioritised loci and genes such as NTRK1, CACNA1C, SPG11, MLKL, TRIB3, WNT5B, C2orf88, TRIM38, MIEF1, MSTN, implicating neuronal plasticity, programmed cell death, immune-inflammatory regulation, metabolic stress and muscle wasting. Heritability was enriched in conserved coding regions, non-myeloid neurons, and embryonic brain, spinal cord, dorsal root ganglion, muscle and barrier organs. These findings provide empirical support for a shared polygenic architecture underlying PICS-related traits and offer a multilevel map of variants, genes, pathways, cell types and tissues that may shape long-term psychological, cognitive and physical vulnerability after critical illness.