BACKGROUND: Stiff person syndrome spectrum disorders (SPSD) are rare, disabling neuroimmunological conditions with no consensus diagnostic criteria, making diagnosis challenging. Misdiagnosis often occurs due to the limited awareness of atypical symptoms and presentations. This study aimed to identify key clinical and paraclinical features most predictive of classic SPS and SPS-plus diagnosis and misdiagnosis patterns. METHODS: We conducted a retrospective case-control study at Johns Hopkins SPS center, analyzing patient data from 1997-2021. A total of 154 classic SPS, 45 SPS-plus, and 66 control patients (evaluated for SPSD but given an alternative diagnosis) were included. Clinical assessments, autoantibody testing, electromyography (EMG), and other diagnostic studies were reviewed. Sensitivity, specificity, and diagnostic odds ratios were calculated, with logistic regression identifying the strongest diagnostic indicators of the SPS phenotypes. RESULTS: Torso/lower extremity symptoms, hypersensitivity triggers, paravertebral stiffness, and gait dysfunction were common in both phenotypes. Classic SPS was most specifically associated with high-titer GAD65 antibodies (98%), cerebrospinal fluid GAD65 positivity (100%), characteristic EMG abnormalities (>
90%), and hyperlordosis (87%). SPS-plus specificity was highest for cerebellar (98.5%) and brainstem (100%) signs/symptoms. High-titer GAD65 antibodies were the strongest independent diagnostic factor for both phenotypes. Misdiagnosis was most common in patients presenting with upper extremity, brainstem, or cerebellar involvement. CONCLUSIONS: Recognizing key diagnostic and misdiagnosis patterns may help clinicians make accurate and timely diagnoses of SPSD. This could help prevent misdiagnosis/overdiagnosis, ensure timely treatment, and assure appropriate patient populations are included in future interventional SPS clinical trials. Further studies are needed to validate these findings and refine diagnostic criteria.