This systematic review and meta-analysis was conducted in strict accordance with PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines (Multimedia Appendix 1) [32]. To ensure transparency and reproducibility, the study protocol was registered with PROSPERO (registration number CRD42024520169).

The selection criteria followed the PICOS framework: (1) Population: individuals officially diagnosed with PD; (2) Interventions: telehealth or telemedicine interventions delivered via telephone, internet, or other digital communication technologies [11]; (3) Comparison/Control: studies with an experimental group compared to control groups using standard care, routine care, conventional care, or waitlist control; (4) Outcomes: assessment of intervention effects on overall health or specific behavioral and psychological symptoms associated with Parkinsonism; (5) Study type: only randomized controlled trials (RCTs).

The exclusion criteria were as follows: (1) studies published in non-English languages, (2) incomplete studies, such as research protocols or ongoing studies, (3) studies where interventions were exclusively telehealth-based without a comparison group, or where no telehealth intervention was applied, (4) studies lacking sufficient details on relevant outcome measures, and (5) studies without adequate statistical data for analysis. No publication date limitations were applied.

Two researchers (MS and FT) systematically searched 5 English-language databases (PubMed, Embase, Cochrane Library, Scopus, and Web of Science) from database inception until June 21, 2024. The comprehensive search strategy combined subject headings and keywords related to three main topics: (1) PD, (2) telehealth, and (3) RCTs. There were no limits on publication status or dates. The detailed search methods are provided in Multimedia Appendix 2. Additionally, references of included studies were manually reviewed.

The reference management tool EndNote X9 (Clarivate Analytics) was used for data management. After removing duplicates, 2 reviewers (MS and FT) independently screened titles and abstracts based on predefined inclusion and exclusion criteria. Potentially relevant articles underwent full-text assessment to determine eligibility. Any discrepancies were resolved through discussion, and when consensus was unattainable, a third reviewer (Luomin) was consulted for a final decision.

Data extraction used a specifically designed form. Extracted information included authors’ names, publication dates, country, study design, sample size, demographic and clinical characteristics (average age, gender distribution, and disease attributes), type and duration of interventions and control groups, outcome measures, and key study findings.

Two researchers (MS and FT) independently assessed the methodological quality and potential bias of included studies using the Cochrane Collaboration’s Risk of Bias tool [33]. Disagreements were resolved through discussion with a third reviewer (Luomin) to achieve consensus. The tool systematically evaluates several dimensions of bias, including random sequence generation, allocation concealment, blinding of participants and personnel, handling of incomplete outcome data, and selective outcome reporting. Each study was assessed for potential selection, performance, detection, attrition, and reporting biases. Risks in each domain were categorized as low (unlikely to significantly affect results), high (likely to significantly undermine confidence), or unclear.

All meta-analyses were conducted using random-effects models, based on the conceptual assumption that true effects vary across studies due to inherent differences in populations, interventions, and settings, rather than on statistical metrics such as I². A dual analytical approach incorporating the Hartung-Knapp-Sidik-Jonkman (HKSJ) adjustment was used to obtain more accurate and conservative interval estimates: for studies providing a single effect size, the HKSJ method was used to calculate 95% CIs; for studies with multiple dependent effect sizes, a 3-level random-effects meta-analysis was performed in R software using the metafor package (version 4.5.1; R Foundation for Statistical Computing) using t-distribution–based inference. This model accounts for sampling variance (level 1), within-study variance (level 2), and between-study variance (level 3), with t-distribution inference being analogous to the HKSJ correction for multilevel models. Heterogeneity was quantified with the τ² statistic. Although 95% prediction intervals were initially planned to illustrate the expected range of true effects in similar future studies, they were not calculated due to the limited number of studies for each outcome (all k<10), as such intervals are unreliable with small samples. The I² statistic is reported for descriptive purposes only, acknowledging its limited pragmatic use in conveying the magnitude of true effect variation across settings. Prespecified subgroup analyses examined intervention type (digital vs traditional telehealth) and follow-up duration. Sensitivity analyses were conducted using leave-one-out elimination. Small-study effects (for which publication bias is one possible explanation) were assessed with funnel plots and multilevel Egger tests. Statistical significance was set at 2-tailed P<.05. Complete analysis code is provided in Multimedia Appendix 3.

Figure 1 shows the PRISMA flowchart detailing the study selection process for this systematic review and meta-analysis. The initial electronic database search yielded 649 records. After removing duplicates, 343 articles remained. Following title and abstract screening, 266 articles were excluded, leaving 77 articles eligible for full-text review. Among these, 20 articles (conference abstracts, reviews, or research protocols) were unavailable in full, leaving 57 articles for detailed evaluation. Ultimately, 15 articles met the inclusion criteria and were included in the final analysis.

This systematic review included 15 RCTs [20,28-31,34-43] with a total of 765 participants. The studies were conducted across various countries: 2 in Spain, 6 in the United States, 1 in Japan (Tokyo), 3 in Italy, 1 in Sweden, 1 in Brazil, and 1 in Australia. These studies were published between 2016 and 2024. All participants were diagnosed with PD and aged 18 years or older. Sample sizes ranged from 9 to 49 participants. All trials used telehealth technologies, including telehealth systems, telephones, applications, and tablets. Follow-up durations ranged from 1 to 12 months post intervention. Detailed study characteristics are provided in Table 1.

The methodological quality and potential biases of the included studies were assessed using the Cochrane Risk of Bias tool [33]. Figure 2 presents detailed findings. All 15 studies were confirmed as randomized, and 12 studies [20,28,30,31,34,36-41,43] clearly described methods of random sequence generation. Only 7 studies [28-30,36,38,39,41] detailed allocation concealment methods, categorizing them as having a low risk of selection bias. Regarding the blinding of participants and intervention providers, only 3 studies [36,39,41] reported adequate blinding procedures and thus had a low risk of performance bias. In contrast, 4 studies [28,37,40,43] without blinding were classified as having a high risk of performance bias. The remaining 8 studies [20,29-31,34,35,38,42] lacked sufficient information and were classified as unclear. Blinding of outcome assessors was reported in 10 studies [28,30,36-43], indicating a low risk of detection bias, while 5 studies [20,29,31,34,35] did not report this information. All studies provided complete outcome data, indicating low attrition bias. Assessment for other potential biases generally indicated low risk. Funnel plot analysis (Multimedia Appendix 4) revealed no significant publication bias among included studies.

This systematic review evaluated the effects of telehealth interventions on the QOL and associated health outcomes in patients with PD. Findings demonstrated that telehealth interventions significantly enhanced various dimensions of patient well-being, including QOL, depressive symptoms, anxiety levels, motor function, ADL, and cognitive abilities. These results differ from earlier reviews and meta-analyses [8,26,27]. The discrepancies might be due to accelerated advancements in telehealth and remote neurology after the epidemic, alongside improvements in telehealth service quality and increased research volume [22,44]. Although the effectiveness of telehealth interventions appears promising, additional studies are needed to establish more conclusive evidence.

Our meta-analysis identified complex patterns regarding telehealth’s effect on QOL in patients with PD. Interventions assessed by the SF-36/BBQ demonstrated a significant improvement, whereas those assessed using PDQ scales indicated only marginal benefit. This difference likely stems from the fundamental distinctions between scales: PDQ scales specifically measure PD-related deficits, whereas SF-36/BBQ assess general well-being [45,46]. Notably, telephone-based interventions substantially improved PDQ-based QOL, while digital interventions showed negligible effects. This distinction explains inconsistencies in prior meta-analyses [1,26,27] that did not account for intervention modality. Significant heterogeneity in PDQ analyses was primarily attributed to between-study differences, suggesting methodological variations. Sensitivity analysis highlighted substantial influence from Ramos et al [28], and potential small-study effects were noted. These factors necessitate cautious interpretation of PDQ outcomes. Conversely, SF-36/BBQ analyses exhibited minimal heterogeneity and robust sensitivity. Differences in results may also reflect variations in intervention type, therapeutic intensity, assessment timing, methodological quality, and sample sizes [27]. Follow-up duration did not moderate effects, contradicting assumptions that longer interventions yield superior outcomes. Instead, the intervention modality emerged as the critical moderator, highlighting the importance of direct human interaction in managing PD-specific QOL concerns. This aligns with telehealth’s capacity for dynamic therapeutic engagement [47,48], particularly beneficial for isolated patients [10], although our findings indicate that these advantages depend on modality.

Telehealth provides more dynamic, immersive methods for treatment, education, and counseling compared to traditional medical approaches, enhancing patient engagement and interaction [47,48]. Such enhancements assist patients with PD and their families in comprehensively understanding and managing disease-related challenges, thus promoting independence, motivation for self-care, and improved life quality [49]. Patients with PD require consistent engagement with health care teams for effective management of disease progression and treatment complexity [8]. A primary advantage of telehealth lies in serving patients in isolated or underserved areas, addressing health care provider shortages, and offering timely, high-quality care to improve patient outcomes [10]. Additionally, economic burdens and logistical difficulties substantially reduce patients with PD’s QOL [50,51]. By reducing health care–related costs and travel demands, telehealth can expand home-based medical services, further enhancing life quality for patients with PD [10,22].

Our meta-analysis demonstrated a significant antidepressant effect of telehealth interventions. Traditional approaches (telephone or cognitive behavioral therapy [CBT]–based) showed nearly 3 times the efficacy compared to digital interventions. This advantage aligns with neurobiological evidence linking depression in PD to dysfunction in serotonergic pathways and frontostriatal circuits [52], suggesting human-mediated therapies more effectively modulate emotional processing compared to automated digital tools. This finding notably diverges from earlier studies, which primarily emphasized motor symptoms and physical rehabilitation, often neglecting depressive symptoms [10,27]. The distinct focus of 3 specific studies [29,30,36] included in this meta-analysis may explain this difference. These studies emphasized cognitive and behavioral aspects of patient care, integrating telehealth with CBT, a method recognized for effectively reducing depression levels [53,54]. Moderate heterogeneity mainly resulted from between-study methodological differences, likely reflecting variations in measurement tools and intervention protocols. Contrary to expectations, categorical follow-up duration showed no significant moderating effect, although continuous analysis revealed a marginal positive association. This result suggests sustained engagement—particularly through telephone-based CBT [53,54]—might progressively reinforce neuroplastic changes in emotion-regulation networks. These findings reconcile previous contradictions in the literature [1,26,27]. Whereas earlier reviews primarily targeted motor symptoms, our analysis confirms telehealth’s antidepressant benefits when including behavioral interventions tailored to PD-related psychopathology.

Our findings align with those of Dou et al [55], indicating that telehealth interventions (tele-CBT and telerehabilitation training) significantly improve depressive symptoms in patients with PD. Feasibility has been verified internationally; for example, a cross-sectional study in Brazil showed effective telehealth implementation in resource-limited settings with high patient satisfaction [56,57]. Regarding specific methods, tele-motor training significantly enhanced patients’ motor function and indirectly alleviated depressive symptoms [55]. In contrast, tele-CBT directly targeted depressive and anxiety symptoms, showing greater effectiveness compared to other teleinterventions [31]. Teleconsultation had relatively limited efficacy in alleviating depressive symptoms but significantly improved access to medical resources [58]. From a neurobiological standpoint, PD and depression share common pathological mechanisms, including gut microbiota dysregulation, neuroinflammation, and reward-processing dysfunction [52]. Telehealth, especially through behavioral interventions such as CBT, may modulate these pathological processes and consequently alleviate depressive symptoms [59]. However, existing evidence suggests telehealth’s effectiveness may be weaker for chronic, nonepisodic mental disorders (eg, depression in PD) compared to primary depression [60]. To optimize telehealth potential, future research should investigate long-term outcomes, standardization of techniques, and cybersecurity considerations [61,62]. In summary, telehealth effectively reduces depressive symptoms in patients with PD, especially via tele-CBT, which overcomes geographical barriers and improves treatment accessibility. Nevertheless, individualized plans and sustained follow-up are necessary to achieve optimal therapeutic outcomes.

Our meta-analysis demonstrated robust anxiolytic effects of telephone-based telehealth interventions, with remarkable consistency across studies. This homogeneity suggests that telephone-delivered CBT provides a reliably standardized approach for managing PD-related anxiety. Notably, these benefits remained stable irrespective of follow-up duration, indicating sustained therapeutic effects without attenuation over 3-9 months. These findings resolve previous contradictions [14,26,27] by demonstrating that structured tele-CBT can effectively address PD-specific anxiety mechanisms, including fear-avoidance cycles and “off”-period distress resistant to conventional treatments. The negligible heterogeneity, with variance entirely attributable to sampling error, likely reflects 3 factors. First, interventions used standardized CBT protocols targeting PD-specific anxiety mechanisms such as hypervigilance toward motor fluctuations. Second, the uniform application of validated and sensitive scales (HAM-A/HADS-A) ensured measurement precision. Third, telephone delivery strengthened therapeutic alliances through real-time emotional interaction absent in purely digital interfaces. The integration of standardized protocols, precise assessments, and person-centered delivery resulted in methodological consistency across studies.

Although considerable evidence supports telehealth for anxiety in patients with PD, its exact mechanism and broader applicability require further investigation. Previous studies [63,64] showed comparable efficacy of telehealth and face-to-face interventions in reducing anxiety, depression, and stress scores, alongside improved heart rate variability. Anxiety reductions persisted long-term after telehealth interventions, confirming their noninferiority to in-person care. This effectiveness largely stems from multimodal interventions; for example, remote CBT overcomes movement-related barriers and, combined with exercise and biomarker monitoring, allows personalized care beneficial to underserved populations [55,64,65]. However, some research highlights intervention heterogeneity. A small study [66] indicated that telephone CBT effectively alleviated depression but not anxiety symptoms in patients with PD, suggesting anxiety may require more tailored strategies. Our analysis, in contrast, supports the long-term feasibility, effectiveness, and durability of telephone CBT effects. Earlier discrepancies might stem from small sample sizes or limitations of measurement tools. Although the revised Parkinson Anxiety Scale improved cultural adaptability, general scales (eg, HADS) may underestimate actual effectiveness due to limited sensitivity [29,67]. Future research should expand sample sizes, develop PD-specific anxiety interventions, and integrate multidimensional biomarker monitoring to improve telehealth precision and applicability.

Telehealth interventions significantly improved motor symptoms in patients with PD. This refined estimate may reflect advancements in methodological rigor involving multilevel analyses that account for independent effect sizes, an approach not consistently used in previous meta-analyses [1,26]. Due to the standardized use of MDS-UPDRS-III assessments and similar intensities of interventions, we observed remarkably low heterogeneity among studies. Notably, digital and traditional telehealth approaches demonstrated comparable effectiveness, indicating that essential motor rehabilitation components, such as amplitude training and balance exercises, effectively translated across different treatment platforms. The temporal stability of benefits further supported telehealth as a sustainable management option, with sensitivity analyses confirming robustness to study exclusion.

A primary therapeutic objective in PD involves improving motor symptoms, wherein treatment adjustments frequently depend on accurate motor assessments [68]. The telehealth framework enables improved and timely interactions between patients and health care providers compared to traditional face-to-face consultations, allowing for more individualized rehabilitation strategies tailored specifically to patients with movement disorders [69]. Multiple studies have confirmed the significant impact of telehealth interventions in alleviating motor symptoms in patients with PD. For instance, structured telerehabilitation programs, such as the Lee Silverman Voice Treatment BIG rehabilitation method, have effectively enhanced motor function, alleviated nonmotor symptoms, and improved the QOL for patients with PD [70]. Compared to teleconsultations alone, tele-motor interventions demonstrate superior efficacy in motor function improvement [55]. From a neuromechanism perspective, cueing techniques activate the motor cortex, thereby enhancing the stability of motor output, which provides scientific justification for using cue-based strategies in telerehabilitation [71]. Moreover, telerehabilitation is particularly suitable for patients with restricted mobility or those residing in medically underserved regions. Real-time video guidance ensures continuous rehabilitation training, effectively overcoming geographical limitations. Its safety and potential effectiveness in improving balance and functional activities have been confirmed by existing research [28,41,72]. Telehealth facilitates comprehensive monitoring of treatment effects through standardized scales (such as MDS-UPDRS) for assessing motor symptoms, combined with evaluation of nonmotor symptoms and QOL questionnaires [73,74]. Additionally, tele-motor interventions based on live-streaming have been proven feasible and safe, demonstrating high patient adherence (eg, twice a week) and thus confirming their practical use for continuous management of motor symptoms in PD [75]. Therefore, telehealth effectively enhances motor functions in patients with PD, offering advantages in personalized program design, activation of neural plasticity, and overcoming limitations in medical resource availability. With ongoing advancements in assessment instruments and technological integration, telehealth is anticipated to further improve long-term intervention outcomes.

The results of this study showed that telehealth interventions significantly improved ADL among patients with PD. This result aligns with previous research, reinforcing that remote health care interventions significantly enhance both ADL performance and motor symptoms in individuals with PD. Our analysis suggests that the significant improvements in ADL resulting from telehealth are due to multidimensional intervention strategies addressing the core symptoms of PD.

Relevant studies have shown that structured remote rehabilitation programs, delivered through real-time video instruction, enhance functional mobility and directly improve basic ADL tasks such as walking and dressing [57]. Simultaneously, high-intensity remote exercise interventions reduce motor sluggishness and freezing of gait, indirectly enhancing instrumental ADLs, such as complex daily activities like shopping and meal preparation [76]. The simultaneous improvements observed in ADLs and motor symptoms share clear pathophysiological connections; enhanced motor functions directly alleviate limitations in physical activity, enabling patients to execute daily routines more effectively [77]. Additionally, remote CBT improves executive functions, mitigating motor-related restrictions on complex ADL performance [77,78]. Telehealth frameworks achieve these synergistic effects by integrating 3 primary components: real-time video supervision ensures adherence to exercise regimens; home-based cognitive training modules restructure the prefrontal-limbic circuitry; and wearable sensors provide immediate feedback regarding movement quality [57,79,80]. Therefore, telehealth interventions positively and synergistically influence both motor symptoms and ADL performance in patients with PD. Future research should focus on optimizing intervention strategies and integrating motor and ADL training components comprehensively to further enhance the overall QOL for patients.

Preliminary evidence shows that telehealth interventions may enhance cognitive function in patients with PD; however, these findings should be interpreted cautiously. Although statistically significant, effect sizes exhibited substantial variability, ranging from negligible to considerable clinical improvement. This observed heterogeneity primarily stems from methodological differences among studies, potentially reflecting (1) the use of varied cognitive assessments (MoCA vs MMSE), each with differing sensitivities to PD-specific cognitive deficits, and (2) distinct intervention protocols within the limited scope of available evidence. Additionally, significant publication bias and insufficient data for sensitivity analyses further limit definitive conclusions.

Overall, the efficacy of telehealth interventions for enhancing cognitive functions in patients with PD has been established. These interventions significantly improve cognitive status, particularly executive functions and memory, as well as emotional and behavioral disorders, consequently enhancing the QOL for both patients and caregivers [78,81]. Among specific intervention methods, computer-assisted cognitive training has shown potential benefits for patients with PD accompanied by mild cognitive impairment, with feasibility confirmed for home-based training modalities [82,83]. Moreover, remote virtual reality applications (telehealth virtual reality) have shown promising results for improving cognitive task performance [31]. A recent network meta-analysis further supports the beneficial effects of remote interventions, including remote cognitive training, on cognition and other nonmotor symptoms [55]. The primary advantages of telehealth interventions include high accessibility (especially beneficial for patients with limited mobility or those residing in remote areas) and flexibility, with the patient’s cognitive reserve potentially enhancing treatment effect [84]. However, considerable heterogeneity exists within current evidence, aligning with our findings. This heterogeneity is largely attributed to variations in study design, inconsistencies in cognitive assessment tools, and diverse responses among patient subtypes [55,82,85]. In addition, the efficacy of telehealth interventions differed across cognitive domains. Therefore, these findings should be considered exploratory and interpreted cautiously. In conclusion, telehealth represents a promising cognitive management approach for PD with substantial potential; nevertheless, implementation barriers must be considered, strategies tailored to individual patient needs, and larger standardized trials conducted to further substantiate effectiveness.

This systematic review benefits from a rigorous methodological approach, using a meta-analysis grounded in RCTs and strictly adhering to established guidelines for systematic reviews. All analyses were conducted using random-effects models based on conceptual considerations, with HKSJ or t-distribution–based corrections applied to provide more accurate and conservative CIs. This significantly enhances the credibility of the findings. Furthermore, the review assesses the impact of telehealth interventions not only on QOL but also on multiple health-related domains such as depression, anxiety, motor function, ADL, and cognitive function in patients with PD, rather than restricting its focus solely to treatment modalities.

Nonetheless, several limitations of this review should be acknowledged. First, the limited number of RCTs included in the analysis may constrain the generalizability of these conclusions to broader populations. Second, due to the small number of studies per outcome (all k<10), prediction intervals were not calculated, which limits the interpretation of how the true effect may vary across different settings. Third, funnel plots and Egger tests were used to assess small-study effects, but these methods have reduced accuracy when fewer than 10 studies are analyzed per outcome, and they do not specifically measure publication bias. Finally, although we applied multilevel modeling to account for dependent effect sizes, residual heterogeneity, and variations in intervention protocols may still influence the results. Therefore, additional RCTs must be incorporated into future research to enhance the robustness and reliability of the findings.

Global disparities in medical resource distribution present substantial challenges to health care service advancement. This issue is particularly pronounced in neurological care, where specialist availability is limited, notably in suburban and rural regions. Consequently, many individuals with PD struggle to receive continuous medical support, resulting in significant declines in their QOL as the disease progresses. This situation places considerable strain not only on patients and their families but also on societal resources. The emergence of telehealth, however, offers a promising solution by providing innovative avenues for managing and treating PD. Telehealth has the potential to bridge existing gaps, enabling patients with PD who previously had limited or no access to receive essential health care services.

PD exerts substantial impacts on public health, prompting significant attention from the health care community toward preventative, diagnostic, and therapeutic strategies. As an innovative product of rapid technological advancement, telehealth represents a cost-effective, real-time, and secure platform for collecting patient data, significantly facilitating the diagnosis, monitoring, and rehabilitation of PD. Nonetheless, the efficacy of telehealth requires further validation through comprehensive and rigorous RCTs. Future research should not only evaluate functional recovery, cognitive enhancement, and health-related QOL but also examine aspects such as cost-effectiveness, patient satisfaction, and digital health literacy among older adults. Such investigations will facilitate more informed decisions and optimal tailoring of telehealth interventions for patients with PD.

Telehealth interventions have demonstrated the potential to significantly enhance various aspects of life among patients with PD, including alleviating symptoms of depression and anxiety, improving motor function, facilitating ADL, and enhancing cognitive performance. Despite these encouraging findings, there remains an urgent need for meticulously designed, large-scale RCTs to comprehensively evaluate telehealth’s effectiveness across the full spectrum of PD management.