Disproportionate short-limb short stature is directly supportable from the original AOIII delineation as a "short-limb dwarfism syndrome" (PMID:2368807) and from the updated FLNB GeneReviews summary describing FLNB-AO3 as "severe short-limbed dwarfism" (PMID:20301736).
Large joint dislocations were missing from the file despite being a hallmark of the FLNB-AO3 phenotype. Primary mutation-series evidence supports joint dislocations in AOI/AOIII (PMID:16752402), and GeneReviews localizes these to the hips, knees, and elbows in AO3 (PMID:20301736).
Apnea is directly documented in the 1992 long-term survivor report together with severe tracheomalacia and tracheostomy requirement (PMID:1442028).
Motor delay is supportable only in a qualified way from the same survivor report, which states the infant survived more than one year "with motoneuronal developmental delay" (PMID:1442028). Because this may reflect secondary morbidity rather than a primary neurodevelopmental effect, the YAML description should stay cautious.
Existing phenotype claims retained
Rhizomelia, clubfoot, humeral hypoplasia, coronal cleft vertebrae, scoliosis, hypertelorism, flat/depressed nasal bridge, micrognathia, cleft palate, broad thumbs, and tracheomalacia are all directly supportable from the abstract of the Japanese survivor report (PMID:1442028).
Candidate claims not added or intentionally softened
Pelvic hypoplasia, rib hypoplasia, and generalized vertebral hypoplasia are mentioned in broader AOIII summaries (for example PMID:27258362), but the abstract language is too introductory and nonspecific to map confidently to narrower HPO terms without overclaiming.
Specific toe phenotypes were not added. PMID:1442028 notes that the feet had abnormalities similar to the short broad thumbs, but the abstract does not localize the finding clearly enough to justify Broad hallux or another toe-specific HPO term.
Intellectual disability / global developmental delay were not added. PMID:8008496 mentions prognosis for physical and intellectual disability only at a high level, without a clean disease-level assertion or count.
Frequency and onset qualifiers were omitted because none of the usable AOIII phenotype abstracts directly quantify occurrence or provide a cohort-based onset statement for the added entries.