Postural instability gait disorder (PIGD) is associated with predominant gait dysfunction compared to the typical tremor dominant Parkinson’s disease (PD).¹ Imaging studies have suggested that the corpus callosum may be implicated in PIGD.^2-5 Diffusion tensor imaging (DTI) is a non-invasive imaging tool widely used to evaluate microstructural changes in brain white matter. We hypothesize that longitudinal neural tracts adjacent to the ventricles are more likely to be affected in patients with PIGD than in PD and healthy controls (HC). The objective is to conduct a diffusion tensor imaging (DTI) MR study to evaluate these periventricular longitudinal tracts in PIGD compared to PD and HC, and examine their clinical correlates.Patients clinically diagnosed with PD and PIGD by a movement disorders expert in a tertiary referral hospital were included in this study. Age and sex matched HC were also recruited. Only subjects without evidence of cognitive dysfunction based on the mini mental state score were included. The risk of falls was evaluated in all subjects using the Tinetti gait and balance score.⁶ All subjects consented and underwent MR brain imaging on a 3 Tesla scanner (Siemens Trio, Erlangen, Germany) using a standardized protocol. Circle regions-of interest (ROIs) of size 30 – 200 mm³ were drawn over the longitudinal periventricular tracts of the corpus callosum (CC, genu and body), anterior thalamic radiation (ATR), inferior longitudinal fasciculus (ILF), posterior limb of internal capsule (PLIC) and inferior fronto-occipital fasciculus (IFOF) bilaterally, and the DTI parameters recorded. The DTI values of fractional anisotropy (FA), mean diffusivity (MD), radial diffusivity (RD), and axial diffusivity (AD) from the right and left ROIs were averaged. The Student’s t-test was used to compare the continuous DTI variables between groups, and statistical significance defined at p < 0.05. Multivariate logistic regression analysis was performed to distinguish between PD and PIGD with the following independent variables: FA, MD, AD and RD for the 6 ROIs, patient demographics and Tinetti score.There were 60 subjects comprising 21 (17 men) patients with PD (aged 79 + 7 years), 19 (15 men) with PIGD (aged 81 + 5 years) and 20 (16 men) healthy controls (aged 79 + 5 years). There was no difference in the DTI values between the groups for the PLIC. For the CC, MD was highest in PIGD compared to PD and controls, and the difference was significant between PIGD and PD in the genu (p = 0.038) and between PIGD and HC in the body (p = 0.017). For the ILF, FA was lowest in PIGD compared to PD and controls, and the difference was significant between PIGD and PD (p = 0.0005) and between PIGD and HC (p = 0.01). Similarly, MD and RD were highest in PIGD and the difference significant between PIGD and HC (p = 0.011 and p = 0.005 respectively) and between PIGD and PD (p = 0.005) for RD only. In the ATR, FA was also lowest in PIGD and the difference significant between PIGD and HC (p = 0.025). RD was highest in PIGD and the difference significant between PIGD and HC (p = 0.038). For the IFOF, MD, AD and RD was highest in PIGD compared to PD and controls, and the difference was significant between PIGD and HC (p = 0.002, p = 0.005 and p = 0.05 respectively), and between PIGD and PD (p = 0.019) for AD and (p = 0.04) for RD. Multivariate Linear regression analysis revealed that a correlation exists between the clinical Tinetti score in PIGD and diffusivity measures in the ILF and ATR (Table 1). Our results show that besides the corpus callosum, other association and anterior projection fibers adjacent to the ventricles are also more affected in PIGD than PD or HC. The corticofugal fiber (PLIC) was not different between PIGD, PD or HC. Our findings correlate with other studies showing greater prevalence of nonspecific white matter hyperintensities in these longitudinal tracts in age-related gait decline.⁷ It has been postulated that abnormalities in these longitudinal tracts interfere with bidirectional transfer of information between key motor and cognitive cortical regions involved in executive function and visuoperception.⁸We demonstrated for the first time that the pathophysiology of the PIGD phenotype is likely multifactorial, involving the longitudinal tracts other than the corpus callosum, and that the clinical severity of PIGD correlates with diffusivity measures in the ILF and ATR. The potential of DTI parameters as surrogate markers in gait assessment of PIDG for neuroprotective or rehabilitative therapy needs further exploration.