Deep brain stimulation in Alzheimer's disease

Glucose metabolism

Across studies, DBS-f consistently engaged neural networks central to the DMN and Papez circuit, underscoring that AD is a pathological disorder of large-scale neural network dysfunction rather than isolated structural loss. ⁷ By stimulating compromised pathways, DBS-f may transiently restore synaptic communication and enable activity to propagate through downstream regions. In Laxton et al., DBS-f drove rapid activation of the hippocampus, entorhinal cortex, and parahippocampal gyrus, followed by delayed recruitment of the posterior cingulate and precuneus, supporting trans-synaptic propagation through the DMN and an excitatory mechanism of action.^7,22 One-year FDG-PET showed persistent hypermetabolism in hippocampal-entorhinal-parahippocampal structures, suggesting sustained activation of this pathway despite continued disease progression. Similarly, Smith et al. confirmed these excitatory network effects, with 1-year FDG-PET revealing increased metabolism across frontal-temporal-parietal-occipital-hippocampal and frontal-temporal-parietal-striatal-thalamic networks, contradicting the hypometabolism of glucose traditionally observed in these regions. ³⁴ When the VC/VS is the target of stimulation, Scharre et al. observed increased orbitofrontal and dorsolateral prefrontal metabolism, further supporting the excitatory hypothesis for interpreting the mechanism of DBS.^19,22 Given that nearly 70% of glucose-derived energy is consumed during excitatory neurotransmission, increases in cerebral glucose metabolism on FDG-PET reflect the heightened energetic demands of synaptic activity. ³⁸ This interpretation is bolstered by astrocyte-neuron metabolic coupling, which demonstrates that glucose use mirrors the energetic demands of neuronal firing and synaptic transmission. ³⁹ Increased synaptic activity provides a plausible mechanism through which DBS-f could influence network coordination beyond the stimulation target. According to Murrow's synchronization hypothesis, DBS-f may recalibrate abnormal oscillatory activity across expansive memory networks, thereby restoring the synchronization required for coordinated network activity. ⁴

Findings from the ADvance trial further support excitatory network activation. ³⁵ Patients who received DBS-f ON demonstrated the greatest increases in glucose metabolism at 6 months, but these effects dissipated at 12 months. ³⁵ Metabolic increases in the temporal, parietal, motor, and cerebellar cortices support that DBS-f activates neurons within the fornix, leading to signal propagation across memory networks. Murrow's model proposes that DBS-f can activate both synchronized and unsynchronized pathways, which may explain downstream cerebellar and sensorimotor engagement observed in the ADvance trial.^4,35

However, DBS may not be purely excitatory. The metabolic increases reported by Laxton et al. in the hippocampus, parahippocampal gyrus, and entorhinal cortex were accompanied by decreases in the anterior cingulate and medial frontal regions during DBS-f. ⁷ These findings support an inhibitory effect, wherein stimulation produces action potentials out of phase with activity in competing executive-control networks. ⁴ Simultaneously, local excitation of neurons within the fornix likely propagates through efferent projections, resynchronizing network activity.^4,22 Thus, Laxton et al. support a model in which DBS-f restores oscillatory timing and network function through a balance of excitatory and inhibitory effects. ⁷ Notably, this study was the only study in which increases in metabolism were accompanied by simultaneous decreases in other regions.

Xu et al. extended this framework by emphasizing widespread cortical and subcortical circuit abnormalities in AD. ¹⁸ They further showed that NBM-DBS, through cholinergic modulation, influences circuits relevant to cognition. Although FDG-PET evidence remains limited for NBM-DBS, short-term improvements in cognition with low-frequency stimulation remain consistent with the excitatory hypothesis.^18,22

The clinical manifestations of these network effects reflect the functional specialization of each DBS target. By reactivating hippocampal-cortical pathways within the Papez circuit and DMN, DBS-f enhances information flow through memory networks, producing modest improvements in recall, orientation, and cognitive engagement.^5,7,35 NBM-DBS generated improvements in agitation, anxiety, and sleep, demonstrating that NBM-DBS influences cholinergic and limbic pathways that regulate arousal and emotional behavior.^18,37 Lastly, VC/VS stimulation engages frontostriatal circuits mediating motivation and executive control, thereby improving apathy and goal-directed behavior through heightened orbitofrontal and prefrontal metabolism. ¹⁹

Altogether, these patterns demonstrate that each DBS target modulates different neural systems that support memory, behavior, and motivation. The study by Xu et al. was the only study to directly compare NBM-DBS to DBS-f in patients with severe AD. Both stimulation targets generated early but short-lived cognitive gains. ³⁷ DBS-f enhanced memory by strengthening hippocampal communication, whereas NBM-DBS improved attention and engagement by increasing cortical cholinergic activity. ³⁷ These differences underscore the importance of future comparative studies to determine target-specific efficacy, optimize stimulation parameters, and identify patients most likely to benefit.

Neuroplasticity

Sankar et al. provided the first in-human evidence that DBS-f may induce structural neuroplasticity in patients with mild AD. ⁵ Over 12 months, hippocampal volume increased in two of six patients, accompanied by heightened glucose metabolism, whereas overall hippocampal atrophy progressed more slowly than in matched ADNI controls, suggesting a stimulation-related effect rather than normal variability. ⁵ Exploratory DBM analysis revealed clusters of local volume expansion in regions both within and beyond the Papez circuit, including the bilateral parahippocampal gyri, right superior temporal gyrus, left inferior parietal lobule, and bilateral precuneus. ⁵ The distribution of these changes aligns with the theory that electrical stimulation reactivates and synchronizes neuronal activity within memory-related circuits.^4,5 This interpretation is further supported by comparative data showing that the degree of hippocampal volume loss over 1 year was greater in ADNI-matched controls than in DBS-f patients, supporting a potential slowing of atrophy. ⁵

These volumetric changes correlated with increases in hippocampal glucose metabolism as well as structural changes in the fornix, mammillary bodies, and functionally unrelated regions traditionally not impacted by AD.^4,5 Mechanistically, DBS-f activates hippocampal neurons and their efferent pathways, transiently increasing glucose metabolism, restoring oscillatory timing, and rehabilitating activity across the Papez circuit to drive trophic and neuroplastic changes. ⁴ The 5%–8% increases in hippocampal volume observed in the two patients in Sankar et al. were sufficient to restore hippocampal size to near pre-disease levels, far exceeding the expected 4%–5% annual atrophy rate in early AD. ⁵ For the entire DBS-f group, hippocampal atrophy was slowed compared to matched ADNI controls. ⁵ The magnitude of these volumetric increases was too large to be explained by scan-rescan variability, suggesting that therapeutic efficacy may be influenced by hippocampal structure. Exploratory DBM analysis identified localized volumetric expansions in cortical regions beyond the Papez circuit, including the bilateral parahippocampal gyri, right superior temporal gyrus, left inferior parietal lobule, and bilateral precuneus. ⁵ Additional increases were also evident in the thalamus and superior frontal gyrus, regions traditionally unaffected by AD. ⁵ Although clinical evidence for DBS-induced structural neuroplasticity remains limited, MRI studies have shown that the hippocampus can enlarge under certain conditions, demonstrating a capacity for structural remodeling despite significant neurodegeneration. ⁵ These widespread effects imply that electrical stimulation may reactivate and synchronize neural activity within and beyond memory-related pathways. ⁵

Preclinical research helps explain how these structural changes may occur. For example, stimulation of the anterior nucleus of the thalamus and entorhinal cortex has been shown to increase neurogenesis in the dentate gyrus of the hippocampus in rodents, and stimulation of the nucleus accumbens increases dendritic complexity in the prefrontal cortex. ⁵ Further preclinical evidence suggests that DBS can modulate neurotransmitter signaling, increase neurotrophic factor expression to support dendritic and synaptic growth, reduce glial activation and neuroinflammation, and restore hippocampal theta oscillations that are critical for memory processing.^5,40 However, these mechanisms have not yet been demonstrated in humans and remain theoretical contributors to the structural and metabolic changes observed clinically. Together, these findings provide biologically plausible pathways through which DBS may facilitate short-term adaptive neural plasticity in memory circuits, while underscoring the need for further research to determine whether such effects hold true in clinical studies.

Cognitive outcomes

In Laxton et al., all six patients experienced cognitive improvement 1 month after DBS-f, evidenced by reduced ADAS-Cog scores. ⁷ Four of the six patients continued to show improvement in ADAS-Cog scores at 6 months, whereas only one patient sustained a lower ADAS-Cog score at 12 months. The investigators evaluated the relationship between disease progression and therapeutic response by comparing preoperative MMSE and ADAS-Cog scores with outcomes at 12 months. Patients with higher baseline scores responded more favorably to DBS-f, indicating that preserved functional integrity of the fornix-hippocampus circuit influenced the likelihood of benefit. ⁷

Sankar et al. did not find any statistically significant correlations between 1-year changes in ADAS-Cog scores and hippocampal volume. ⁵ Two patients demonstrated hippocampal volume increases that clinically translated to improved cognitive outcomes at the 1-year follow-up. ⁵ However, the patient with the second largest hippocampal volume increase still experienced cognitive decline during follow-up. The patient with the largest hippocampal volume increase continued to present with preserved hippocampal integrity 3 years after stimulation but also demonstrated progressive worsening of ADAS-Cog scores. This situation suggests that hippocampal integrity may not be sufficient to sustain cognitive function on its own as AD progresses. ⁵ The investigators noted that PET imaging before and after stimulation might have more effectively measured neural circuit activity compared to baseline volumetric changes.

More recent studies from Xu et al. extend these findings, demonstrating that both DBS-f and NBM-DBS can produce short-term improvements in ADAS-Cog, MMSE, MoCA scores at 3 months. ³⁷ Moreover, no significant differences were observed between the two stimulation targets. ³⁷ Smith et al. reported an average increase of two points on ADAS-Cog every 6 months, with a single patient showing a four-point decline at 12 months, again illustrating variable response to DBS-f. ³⁴ These findings are consistent with earlier trials in which DBS-f enhances cognition in the short-term but fails to ameliorate cognitive dysfunction in the long-term.

The larger phase II ADvance trial confirmed the absence of group-wide benefit, showing no significant differences in cognitive metrics between DBS-f ON and DBS-f OFF groups at 12 months. ³⁵ Younger patients experienced a more rapid decline in ADAS-Cog 13 and CDR-SB with DBS-f ON compared to DBS-f OFF. Targum et al. echoed these findings, reporting that approximately 50% of younger patients worsened cognitively compared to only 6.7% of older patients regardless of whether the patient was assigned to DBS-f ON versus DBS-f OFF. ³⁶ Given that younger-onset AD represents only 4% of AD cases and is marked by more severe network disruption, these results reflect disease severity rather than age and should not be generalized to the broader AD population.^6,35,36

Across stimulation targets, outcomes were influenced by patient characteristics, particularly disease stage and age of onset.^35,36 Patients with less aggressive disease, preserved hippocampal volume, and higher baseline cognitive function derived the most benefit from DBS-f, whereas younger-onset patients (<65 years old) tended to decline more rapidly despite active stimulation. In contrast, older patients (≥65 years old) demonstrated slower cognitive decline and greater increases in glucose metabolism, suggesting that preserved network integrity and disease stage may influence responsiveness.^6,35,36 These results emphasize that DBS outcomes are shaped by baseline pathology and individual patient characteristics. Ultimately, these findings underscore the demand for biomarker-driven and disease stage-specific trial designs to identify patient subgroups most likely to benefit from DBS.

Cerebral atrophy

Differences in clinical response to DBS appear to depend on the degree of preserved functional circuitry, which may account for age-related differences reported by Lozano et al. ³⁵ Post hoc analysis of preoperative PET scans revealed that younger patients had significantly lower baseline glucose metabolism in temporal and parietal regions, with deficits of 6%–11% compared to older patients. ⁶ These findings highlight more advanced and widespread neurodegeneration at the time of treatment in the younger cohort and may explain their poorer outcomes. Despite receiving stimulation, these patients declined faster, suggesting that DBS cannot restore communication in networks that are severely disconnected. ³⁵

Consistently, Laxton et al. reported that patients with milder atrophy and higher preoperative cognitive scores at baseline demonstrated a more favorable response to DBS-f. ⁷ In Sankar et al., all DBS-f patients showed slower atrophy than matched controls indicating that preserved structural and functional integrity of the hippocampal-fornix pathway may be necessary for stimulation effects to propagate through memory networks. ⁵ Notably, across all trials, DBS did not stop or reverse the pathological progression of AD. Together, these findings demonstrate that the ability of DBS to propagate stimulation effects declines as circuit deterioration progresses, underscoring the need for biomarker-based assessments of metabolic activity, structural connectivity, and cholinergic function when selecting candidates and designing trials.

Neuropsychiatric outcomes

Although DBS has produced transient cognitive improvements, recent studies highlight its potential to improve neuropsychiatric symptoms and reduce caregiver burden, particularly with NBM-DBS in patients with severe AD. In the first trial by Xu et al. in 2024, patients receiving NBM-DBS demonstrated significant short-term improvements in activities of daily living and reductions in caregiver burden largely driven by improvements in patient sleep quality, mood, and neuropsychiatric symptoms. ¹⁸ Across all follow-up assessments, NPI scores improved, specifically with respect to agitation, depression, anxiety, apathy, disinhibition, abnormal motor behaviors, and sleep disturbances. Similarly, HAM-D scores decreased substantially with patients improving from severe depression to mild. ¹⁸ Findings from Xu et al. corroborate these results, showing that patients experienced reductions in NPI scores along with significant improvements in anxiety (HAM-A) and sleep quality (PSQI) at 3 months after stimulation. ³⁷

Compared to DBS-f, NBM-DBS generated greater improvements in neuropsychiatric outcomes, likely by modulating emotional and behavioral responses through its connections with the limbic system and pre-frontal cortex. ³⁷ In summary, these studies indicate that NBM-DBS mitigates neuropsychiatric symptoms and reduces caregiver burden, both of which are critical dimensions of therapeutic efficacy in AD. Neuropsychiatric symptoms are particularly prominent in patients with severe AD, affecting nearly 80% of patients and encompassing delusions, hallucinations, depression, anxiety, apathy, and mania, which contribute to substantial distress for both patients and caregivers. ³⁷ Approximately 53% of AD patients presenting with neuropsychiatric symptoms were misdiagnosed with primary psychiatric disorders, compared to only 4% without such symptoms, resulting in delayed treatment and increased caregiver stress. ³⁷ Given that nearly all patients with AD will eventually progress to the severe stage, for which no preventive or disease-reversing interventions exist, the ability of DBS to alleviate mood, sleep, and behavioral disturbances represents a clinically meaningful contribution that other therapies fail to achieve. Moreover, these neuropsychiatric and functional improvements can indirectly enhance patient comfort and family interactions, supporting a broader role for DBS within comprehensive disease management. Addressing neuropsychiatric symptoms is important for both the patient and care-taking ecosystem, highlighting the multifaceted nature of DBS as a treatment approach in AD. ³⁷

Ethical and patient-centered considerations

Although the primary focus of this review has been on stimulation targets and mechanistic outcomes, it is equally important to consider the ethical and patient-centered dimensions of DBS in AD. These considerations reflect well-established patient-centered principles that promote a holistic approach to caring for older adults, ensuring treatment remains aligned with patients’ needs, goals, and overall well-being. ⁴¹ One example is the Age-Friendly Health Systems 4Ms Framework, which guides care according to individual priorities while supporting cognition, mobility, and safety, an approach that is particularly important when treating individuals with cognitive impairment. ⁴¹ The inclusion of patients with moderate to severe cognitive impairment introduces complex ethical challenges related to informed consent, autonomy, and the balance between potential benefit and risk of undergoing such a procedure. In their study, Xu et al. employed a rigorous and ethically sensitive approach to informed consent, striving to ensure that patients with severe AD could comprehend and participate in decision-making despite cognitive impairment. ³⁷ Communication strategies were tailored to each patient's cognitive level, with explanations repeated as needed. In certain cases, legal representatives were involved to ensure that decisions reflected patient values and preserved autonomy. Throughout the study, continuous dialogue among patients, families, and the research team was maintained to foster trust and transparency, consistent with oversight by an institutional ethics committee. ³⁷ In contrast, earlier DBS trials enrolling patients with mild to moderate AD obtained informed consent directly from participants, with surrogate input only when necessary. ⁷ Xu et al. were notable for prioritizing patient-centered decision-making and emphasizing quality of life as a primary study endpoint. ³⁷

Although DBS-related reductions in caregiver burden are noteworthy, the ethical foundation of this intervention must remain the direct relief of the patient. Reductions in caregiver strain were largely driven by improvements in patient mood, sleep, and behavior, highlighting that caregiver relief is meaningful because it stems from improved patient well-being.^18,37 These findings underscore that, although a decrease in caregiver burden is a meaningful secondary outcome, the ethical and clinical implications of DBS should remain patient-centric. This prioritization ensures that therapeutic development aligns with core ethical principles of autonomy, dignity, and patient well-being.

Beyond the consent process, the practical and ethical implications of surgery and device management also warrant equal consideration. The more durable benefits of NBM-DBS included better communication between patients and families, reduced reliance on psychotropic drugs, and minimized adverse events from concurrent medication use.^18,37 NBM-DBS ameliorated hallucinations, anxiety, depression, and sleep disorders, thereby improving both patient and care-giver well-being.^18,37 Although the cognitive gains from NBM-DBS may be modest and short-lived, its effect on neuropsychiatric symptoms offers a therapeutic advantage over pharmacological interventions.^18,37 Although all reviewed studies determined DBS to be safe and well-tolerated, it remains an invasive neurosurgical procedure performed in a medically-vulnerable population often affected by advanced disease and multiple comorbidities.^{5–7,18,19,34–37} NBM-DBS and DBS-f were considered safe and well-tolerated, yet six patients in the DBS-f cohort died during the long-term follow-up, events attributed to disease progression, comorbid illness, and postoperative infections rather than surgical complications. ³⁷ Among patients with mild to moderate AD, infections accounted for many adverse events across studies.^{5–7,18,19,34–37} Even minor complications, such as superficial infections or delayed healing, can meaningfully affect comfort and dignity in these patients, reinforcing the need to weigh procedural risk against potential gains in quality of life. For instance, qualitative research in movement-disorder populations has shown that patients living with DBS implants may experience changes in self-identity, reduced independence, or emotional challenges even as symptoms improve.^42,43

In summary, these findings underscore the need for future DBS research that integrates quality-of-life and psychosocial outcomes alongside conventional cognitive and neuropsychological endpoints. Although sustained cognitive improvement remains an important goal, current evidence suggests that these gains are often transient; therefore, future studies should place equal emphasis on quality-of-life metrics to capture the full therapeutic impact of DBS in AD.

Limitations and considerations for future research

Across these studies, several critical limitations must be acknowledged. Chief among them is the small sample size across all nine studies reviewed. Six of the nine reviewed studies enrolled fewer than 20 patients, and even the phase II ADvance trial included only 42 participants.^{5–7,18,19,34,35,37} Participant recruitment in clinical trials for AD remains a challenge due to the invasive nature of the procedure and the lack of insurance coverage for investigational neurosurgery. As a result, the current literature is dominated by feasibility studies, with a notable absence of large-scale, controlled trials. Such small cohorts reduce statistical power and limit generalizability.

Furthermore, Laxton et al., ^{7 Smith et al., 34 and Sankar et al. 5 each studied the same six patients, meaning that many of the same methodological limitations were repeated across trials. The volumetric changes in Sankar et al. are largely driven by two patients, highlighting the disproportionate impact of outliers in a small sample. 5 Nevertheless, outlier patients met all study inclusion criteria, including significant memory impairment at enrollment and preserved hippocampal volume at baseline. Moreover, hippocampal volumetric changes were not consistently correlated with clinical outcomes, suggesting that the effects observed cannot be dismissed as a result of patient enrollment. 5}

Another important limitation is the lack of a control group in the earlier studies, which makes it difficult to eliminate confounding variables. In the more recent studies, secondary endpoints related to quality of life and caregiver burden were heavily reliant on caregiver reports, introducing subjectivity and bias and emphasizing the need for more objective metrics.

The stimulation parameters used were largely standardized from PD protocols rather than tailored to the neurophysiology of AD. Chronic DBS-f settings were set to 130 Hz, 90 µs, and 3.0–3.5 V based on monopolar stimulation mapping to maximize safety and efficacy. In Laxton et al., increasing the voltage to 5–6 V elicited autobiographical recall in two patients, suggesting that higher thresholds may more effectively engage memory circuits. ⁷ However, high settings induced autonomic effects and increased the risk of adverse events. Preclinical studies indicate that overstimulation can impair memory, underscoring the likelihood that the limited clinical benefit could be due to suboptimal dosing. ⁷ Parameter optimization has also not been investigated for alternative targets such as NBM or VC/VS, underscoring the need for target-specific stimulation settings in patients with AD.

Only two of the reported trials required pathological or biomarker confirmation of AD.^18,19 For the rest of the trials, AD diagnosis was based primarily on psychological and performance-based evaluations, raising the possibility that some participants may have had other dementias. Clinical diagnosis alone has been shown to be inaccurate in up to 23% of cases. ⁴⁴ Without confirmation from amyloid-β or tau biomarkers, or consideration of genetic markers like APOE ɛ4, patient heterogeneity in pathology and disease stage may alter treatment outcomes. This limitation is particularly important for younger patients with earlier disease onset, who often show more aggressive disease and less favorable outcomes in response to DBS.^6,35,36 Lastly, the short follow-up windows across studies may not have captured delayed benefits or adverse events.

Although DBS can transiently modulate neural activity, its effects are constrained by the irreversible neural circuit damage and neuronal loss characteristic of AD. Nevertheless, short-term cognitive gains and sustained neuropsychiatric improvements observed across studies remain clinically meaningful, particularly when pharmacologic and noninvasive interventions fail to yield comparable benefits.^{5,7,18,19,34,35,37} Recent advances in antibody-based immunotherapies, such as bapineuzumab, gantenerumab, and lecanemab, have demonstrated the ability to reduce amyloid-β accumulation in the brain, but these therapies do not halt disease progression or provide sustained improvements in cognition. ³⁷ Similarly, novel anti-inflammatory agents, such as daratumumab, NE3107, and mastinib, have demonstrated encouraging results in mild AD, though evidence remains preliminary and unconfirmed in larger cohorts. ³⁷ As AD advances, pharmacologic efficacy declines, and patients often experience compounding cognitive and neuropsychiatric symptoms, requiring multidrug regimens that increase the risk of adverse interactions and systemic side effects. ³⁷

Given the absence of curative therapies and the poor prognosis associated with moderate to advanced AD, even transient symptomatic and cognitive improvements represent meaningful therapeutic advancements.^18,37 The overall safety and tolerability of DBS reported across studies further supports its potential as a viable adjunctive therapy.^{5–7,18,19,34,36,37} Larger, rigorously controlled clinical trials are therefore justified not only to validate preliminary safety and efficacy data but also to address the methodological limitations that have constrained interpretation to date. Collectively, these limitations highlight several areas of focus for future DBS research.

Controlled clinical studies are needed to test how variations in stimulation parameters influence clinical outcomes and to establish optimized parameters for AD rather than relying on those derived from PD. Future trials should require confirmation of abnormal amyloid-β, cerebrospinal fluid biomarkers, and relevant genetic markers to reduce pathological heterogeneity within participants because earlier studies have shown that disease stage, baseline neural integrity, pathology, and stimulation target strongly influence clinical responses.^6,35,36 Consequently, stratifying cohorts by these biological and clinical factors will be critical for identifying patients most likely to benefit from DBS. To better understand how DBS modulates neural circuit activity, future studies should incorporate objective measures such as FDG-PET, DBM, or other advanced imaging alongside objective assessments of cognition. These modifications will strengthen the mechanistic understanding by linking DBS-induced alterations in network connectivity, and neurophysiological activity to corresponding improvements in cognition and functional performance.

Beyond elucidating the link between DBS mechanisms and clinical manifestations, a key unresolved question is whether the stimulation target itself determines the magnitude or durability of the clinical benefit. Although Xu et al. offered the first study to compare NBM-DBS to DBS-f in patients with severe AD, the trial was neither randomized nor adequately powered to determine target-specific efficacy. ³⁷ Beyond this comparison, no randomized or systematically controlled studies have evaluated stimulation sites across disease stages. Consequently, it remains uncertain whether different stimulation targets, such as the fornix, NBM, and VC/VS differentially influence cognitive or functional outcomes. Future randomized, comparative, clinical trials are needed to evaluate target-specific non-inferiority, clarify mechanistic differences, and determine whether one stimulation site provides superior or more durable symptomatic benefit. Lastly, follow-up assessments should extend beyond 12 months to further understand the durability and longitudinal effect of DBS, thereby clarifying its long-term clinical value and overall risk-benefit profile. Given that cognitive gains are often transient, incorporating standardized quality-of-life, psychosocial, and clinician-rated measures will be essential to determine whether DBS confers meaningful, sustainable benefits for patients beyond temporary cognitive improvement. Equally important, future studies should integrate standardized quality-of-life, psychosocial, and clinician-rated outcome measures to determine whether neuropsychiatric and functional improvements observed in severe AD also extend to patients with mild to moderate disease. Incorporating these multidimensional endpoints will be essential for assessing whether DBS yields meaningful and sustainable benefits for patients.