Résumé : Glycogen storage disease type II (GSDII) is an autosomal recessive lysosomal storage disorder caused by acid alpha-1,4-glucosidase (GAA) deficiency. This study aimed to provide an in-depth description of a late-onset GSDII (LO-GSDII) cohort (n=36) and assess potential genotype-phenotype correlation. We performed a clinical record-based study, some patients (n= 19) were also followed prospectively. Phenotypes were highly variable. We focused our clinical assessment onrespiratory failure, as it is the most frequent cause of death in LO-GSDII. In addition to standard spirometric measures, in a subgroup of patients (n = 10) we utilized a new tool, optoelectronic plethysmography (OEP), to investigate the pathophysiology of respiratory muscle impairment.The GAA gene was sequenced in every patient, and pathogenic mutations were identified inall of them. Almost all (35/36) patients carried the same mutation on one allele, IVS1-32-13T>G, which was in compound heterozygosity with a variety of other GAA mutations. To investigate genotype-phenotype correlation, we divided the patient cohort in two groups, according to the severity of the mutation on the second allele. The respiratory function study focused on diaphragmatic weakness. According to the change in forced vital capacity in supine position (ΔFVC), we defined patients with ΔFVC>25% ashaving diaphragmatic weakness (DW) and those with ΔFVC<25% as without diaphragmatic weakness (noDW). We measured pulmonary function and chest wall volumes using OEP inboth groups. We found a good correlation between the supine abdominal contribution to tidal volume (%VAB) and ΔFVC. Patients showed reduced chest wall and abdominal inspiratory capacity and low abdominal expiratory reserve volume. In terms of genotype-phenotype correlation, we counted more subjects in the group with severe second mutations (n=21) who had severe motor disability and respiratory dysfunction. However, this finding remains preliminary because differences were not significant, likely because of small sample size. Finally, in two smaller substudies, we investigated the occurrence of urinary and fecal incontinence in LO-GSDII, and reported a possibly non-fortuitous association of LO-GSDII and hydromyelia in two individuals. Overall, this work 1) provided new insight into genotype-phenotype correlation in GSDII, suggesting that it is of complex nature; 2) refined the analysis of respiratory muscle impairment and showed the utility of OEP for respiratory assessment in this neuromuscular disorder, and possibly in others as well; 3) indicated some so far little studied phenotypic features of LO-GSD-II that deserve further investigation.