Introduction: Locomotion is a complex motor task integrating sensory-motor and cognitive inputs through central pattern generator control. Locomotion–respiration coupling (LRC), defined as the synchronization between gait and breathing events, has been described during running and walking. Patients with obstructive sleep apnea syndrome (OSAS) exhibit altered gait patterns during wakefulness, which appear to be partially independent of nocturnal symptomatic treatment. While intermittent hypoxia may induce long-term (“fixed”) central nervous system alterations, OSAS patients also display “reversible” adaptive mechanisms during wakefulness aimed at preventing upper airway (UA) collapse, involving changes in respiratory control and posture. Whether these adaptive mechanisms affect LRC and contribute to gait alterations remains unknown. Methods: Ten awake patients with controlled OSAS treated with a mandibular advancement device (MAD) and ten matched healthy controls performed a one-minute walking trial on an instrumented treadmill with motion capture. Ventilation was assessed via thoracic expansion, and spatiotemporal gait parameters were extracted from kinematic and force platform data. LRC was quantified using phase-locking values (PLV) between ventilatory events and sagittal-plane joint kinematics. OSAS patients were assessed both with and without their MAD. Results: Compared with controls, OSAS patients exhibited a longer double-stance phase and a shorter swing phase. Expiratory events occurred preferentially during double stance in OSAS patients, whereas controls showed a higher occurrence during swing. PLV analysis revealed increased synchronization between inspiratory events and hip and knee motion, as well as between expiratory events and trunk and hip motion in OSAS patients. These significative differences were no longer observed when patients wore their MAD. Logistic regression demonstrated that PLV metrics of inspiratory–hip and expiratory–trunk coupling effectively discriminated OSAS patients from controls, but not when UA stabilization was applied. Conclusion: Awake patients with controlled OSAS exhibit altered gait patterns characterized by increased locomotion–respiration coupling. This heightened synchronization likely reflects adaptive, reversible strategies aimed at limiting UA collapsibility during wakefulness rather than fixed hypoxia-related impairments alone. LRC may represent a functional marker of daytime motor alterations in OSAS and a potential target for interventions addressing UA mechanics beyond nocturnal treatment.
Background: Early relationships with caregivers shape emotional and cognitive development, colouring attitudes and behaviours across the lifespan. Secure attachment has consistently been associated with a lower risk of late-life depression (LLD). However, the extent to which parental (maternal or paternal) attachment contributes to late-life depressive symptoms and whether this association is influenced by cognitive impairment remains unclear. Objectives: We investigated whether mild cognitive impairment (MCI) moderates the association between parental (maternal/paternal) attachment style (secure vs. insecure) and late-life depressive symptoms. Methods: We analysed data from 120 participants enrolled in a cross-sectional study conducted at the Old-Age Psychiatric Service of Lausanne University Hospital. Retrospective parental attachment was assessed using the Attachment Multiple Model Interview (AMMI), from which maternal and paternal secure and insecure attachment scores were derived. Late-life depressive symptoms were measured with the Cornell Scale for Depression in Dementia using both self- and informant-reports. MCI was defined using the Clinical Dementia Rating (CDR), with a global score equal to 0.5 indicating MCI. Associations and potential moderation by MCI were examined using multiple linear regression models. Results: Lower secure paternal attachment was significantly associated with higher self- and informant-reported depressive symptom severity, but only among participants with MCI. No significant associations were observed independently of the MCI for maternal attachment or insecure paternal attachment. Conclusions: Early paternal bonds may shape emotional vulnerability differently in the context of cognitive decline. These results suggest considering both attachment history and cognitive status in interventions for late-life depressive symptoms.
BACKGROUND:Physical Training (PT), defined a planned, structured, and long-term exercise intervention, is recognized as an important non-pharmacological therapy in Parkinson's disease (PD). Beyond its symptomatic benefits, experimental evidence suggests that PT may also modulate pathological mechanisms involved in dopaminergic neurodegeneration. This systematic review summarizes the neurobiological and neurophysiological adaptations induced by PT in animal models of PD and in patients with PD, and proposes mechanistically driven recommendations for personalized exercise prescription. SUMMARY:MEDLINE and Web of Science were searched from January 1, 1974, to April 3, 2026. The PRISMA 2020 guidelines were followed, and two reviewers independently screened titles and abstracts to identify studies that reporting neurobiological or neurophysiological effects of PT performed for at least 4 weeks, in both animal models of PD or in patients with PD. Risk of bias were assessed using validated tools for animal studies (SYRCLE), and for randomized (RoB 2) or non-randomized controlled trials (ROBINS-I) in humans. A total of 78 studies were included, comprising 55 animal studies and 23 studies in patients with PD. In parkinsonian animals, PT consisted mainly of forced treadmill training, performed 20-60 min/day, 5 days/week, typically for 4 weeks. PT reduced oxidative stress and mitochondrial dysfunction, enhanced antioxidant defences, and regulated mitochondrial dynamics. Exercise also decreased α-synuclein accumulation, attenuated neuroinflammation, and increased neurotrophic factor expression. These adaptations promoted synaptic plasticity, neurogenesis, and preservation of nigrostriatal dopaminergic neurons. In patients with PD, aerobic training (AT), resistance training (RT) and balance training (BT) have been tested, and multimodal interventions have been investigated, most commonly involving 120-150 min/week for 6-12 weeks, sometimes combined with virtual reality or exergaming. PT induced structural and functional adaptations across the brain-muscle axis, including modulation of dopaminergic signalling, reorganization of cortico-striatal and cerebellar connectivity, improved cortical inhibitory control, enhanced corticospinal drive, and neuromuscular remodelling. AT mainly improved fatigability and cardiorespiratory fitness, whereas RT and BT appear to preferentially enhance muscle strength, rate of force development, gait automaticity, and balance control. KEY MESSAGES:Current evidence indicates that PT acts as a multimodal intervention capable of inducing adaptations ranging from molecular pathways to large-scale motor networks in PD. These findings support the implementation of personalized multimodal programmes combining AT, RT and BT, with progressive intensity and long-term adherence. Future studies should better define dose-response relationships, identify biomarkers of responsiveness, and determine whether exercise may exert disease-modifying effects on PD.
BACKGROUND:Lewy body dementia (LBD) is characterized by pervasive cognitive, motor, and neuropsychiatric symptoms. Although cognitive decline has traditionally shaped staging and prognosis, accumulating evidence indicates that gait and postural disturbances follow partially independent trajectories and may remain preserved even as cognition deteriorates. These observations challenge conventional assumptions about late-stage decline and suggest that locomotor function may complement cognitive measures when assessing late-stage transition in LBD. SUMMARY:This review synthesizes current evidence on gait disturbances across the LBD spectrum and their neural substrates, in comparison with Alzheimer's disease (AD) and Parkinson's disease (PD). Gait impairment in LBD arises from distributed network dysfunction involving dopaminergic and cholinergic pathways, fronto-striatal, parietal, and cerebellar contributions, resulting in greater gait variability, reduced automaticity, and heightened dual-task costs relative to AD and PD. Despite this vulnerability, preserved procedural memory, brainstem locomotor circuits, and cerebellar-cortical compensation may allow walking to persist longer than cognition in some individuals. In advanced disease, rapid gait deterioration has been described near the end of life, suggesting that gait trajectories may capture aspects of late-stage decline that cognitive measures, affected by fluctuations and attentional instability, may not reliably reflect. KEY MESSAGES:-Gait and cognition do not decline in parallel in LBD; in some individuals, walking may outlast cognitive abilities.-Gait disturbances reflect distributed network dysfunction rather than isolated motor system degeneration.-In late stages, gait performance may offer a clinically useful indicator of disease transition when cognitive testing becomes less reliable. -Integrating gait assessment into clinical evaluation may enhance prognostication and support palliative care planning in LBD.
