Associations between transcranial direct current stimulation session exposure and clinical outcomes in knee osteoarthritis: a cross-sectional analysis exploring the role of pain catastrophizing

Abstract

Background

Knee osteoarthritis (KOA) is a prevalent musculoskeletal disorder characterized by chronic pain, reduced physical function, and impaired quality of life. Transcranial direct current stimulation (tDCS) has emerged as a promising non-invasive neuromodulatory intervention, yet limited evidence exists regarding its association with clinical outcomes and the role of psychological mediators in this population. This cross-sectional study investigated the relationship between tDCS parameters (number of sessions, stimulation intensity, and duration) and clinical outcomes—pain, functional disability, and quality of life—in individuals with KOA, and explored whether pain catastrophizing mediates these associations.

Methods

Sixty-eight participants diagnosed with KOA and treated with tDCS were enrolled. Clinical outcomes were assessed using the Numeric Pain Rating Scale (NPRS), KOOS-Physical Function Shortform (KOOS-PS), and EQ-5D-5 L index. Pain catastrophizing was measured using the Pain Catastrophizing Scale (PCS). Pearson correlations, multiple linear regression, and mediation analysis using the PROCESS macro were performed. A p-value < 0.05 was considered statistically significant.

Results

The number of tDCS sessions showed significant correlations with reduced pain (r = − 0.42, p = 0.002), improved function (r = 0.39, p = 0.004), and better quality of life (r = 0.44, p = 0.001). Regression analysis showed that the number of tDCS sessions was significantly associated with NPRS (β = − 0.36, p = 0.004), KOOS-PS (β = 0.33, p = 0.005), and EQ-5D-5 L (β = 0.42, p = 0.001). Pain catastrophizing partially mediated these associations, with significant indirect effects observed for all outcomes (p < 0.05).

Conclusion

Higher tDCS session frequency was correlated with favorable clinical outcomes in KOA, and pain catastrophizing showed statistical evidence of an indirect association, though no causal mediation can be inferred. These findings highlight the importance of considering psychological factors in neuromodulatory treatment planning.

Background

Knee osteoarthritis (KOA) is a leading cause of pain, disability, and reduced quality of life in the aging population, affecting approximately 20% of individuals over the age of 45 worldwide [1]. The condition is characterized by progressive cartilage degeneration and joint inflammation, which result in chronic pain, impaired mobility, and functional decline [2]. Despite a wide array of conservative management strategies—including pharmacologic treatments, physical therapy, and lifestyle modification—many patients continue to experience suboptimal symptom control [2]. Importantly, the clinical burden of KOA extends beyond physical symptoms, often encompassing psychological distress and cognitive-affective disturbances that influence pain perception and treatment response [3].

Transcranial direct current stimulation (tDCS) has emerged as a promising neuromodulatory technique capable of modulating cortical excitability and altering pain-related neural circuits [4]. Anodal stimulation over the primary motor cortex (M1) has been shown to reduce pain intensity by engaging descending inhibitory pathways and modulating thalamo-cortical processing [5]. In musculoskeletal conditions, including chronic low back pain and fibromyalgia, tDCS has demonstrated favorable effects on pain and function, particularly when applied over multiple sessions [6]. In the context of OA, emerging evidence suggests potential therapeutic benefits; however, clinical responses to tDCS are known to be variable and influenced by a range of individual factors [7]. While prior studies have examined its efficacy in randomized controlled trials, there remains limited cross-sectional evidence investigating how specific stimulation parameters—such as session frequency, intensity, and duration—relate to clinical outcomes in real-world settings. Furthermore, few studies have integrated psychological constructs into the analysis of tDCS effects in KOA, despite the recognized role of cognitive-affective factors in modulating pain experience.

Pain catastrophizing, defined as an exaggerated negative cognitive and emotional response to actual or anticipated pain, has been consistently associated with heightened pain sensitivity, increased disability, and poorer treatment outcomes in OA [8]. It is recognized as a key contributor to central sensitization and altered descending modulation, both of which are targetable by neuromodulatory interventions such as tDCS [9]. Despite this, the potential mediating role of pain catastrophizing in the relationship between tDCS exposure and symptom outcomes has not been thoroughly investigated. Additionally, individual factors such as sex and disease duration may influence the clinical presentation and response to tDCS, yet these variables are often underexplored in analytic models. Addressing these gaps is essential for optimizing stimulation protocols and advancing precision approaches in neuromodulation for OA.

The aim of this study was to examine the association between tDCS intervention parameters—specifically the number of sessions, stimulation intensity, and duration—and key clinical outcomes including pain, functional disability, and health-related quality of life in individuals with KOA. A secondary objective was to assess whether pain catastrophizing mediates the relationship between tDCS exposure and these clinical outcomes. Additionally, subgroup analyses explored outcome variations by sex and OA duration. We hypothesized that a higher number of tDCS sessions—and, to a lesser extent, variations in stimulation intensity—would be associated with improved clinical outcomes.

Methods

Ethics, design and setting

This cross-sectional study was conducted between 23rd May 2023 and 13th February 2024 in the Rehabilitation department of College of Applied Medical sciences, Abha, Kingdom of Saudi Arabia. Prior to participation, written informed consent was obtained from each individual, and all research activities adhered strictly to the ethical principles outlined in the Declaration of Helsinki. Ethical approval for the study was granted by the Research Ethics Committee of KKU, DSR (approval number [ECM#2023 − 276]), ensuring compliance with national and international guidelines for human research.

Participants

Participants for this study were recruited from the outpatient rehabilitation clinics affiliated with the Department of Physical Therapy, King Khalid University, Kingdom of Saudi Arabia. Individuals were screened consecutively during their routine clinical visits, and those who met the eligibility criteria were invited to participate. A total of 68 participants were enrolled following informed consent and verification of eligibility. The diagnosis of was confirmed based on the clinical and radiographic criteria of the American College of Rheumatology (ACR) [10], which includes the presence of activity-related knee pain, morning stiffness lasting less than 30 min, and radiographic evidence of osteophyte formation or joint space narrowing [10].

Kellgren–Lawrence grading (Grade II or III) was determined based on standard weight-bearing anteroposterior knee radiographs, which were reviewed and reported by a board-certified musculoskeletal radiologist as part of routine clinical care. Only those patients who had received a minimum of five sessions of tDCS with anodal stimulation targeting the primary motor cortex (M1) were considered for inclusion. Although M1 was the primary intended target, variations in electrode placement—such as occasional use of supraorbital (SO) or alternative positions—occurred in routine clinical settings due to anatomical or patient-specific considerations. All included participants received M1 anodal stimulation in the majority of sessions, ensuring consistency with the study’s focus. Eligible participants were adults aged 45 years and above with a confirmed diagnosis of primary of Kellgren-Lawrence grade II or III and a stable medical condition allowing participation in assessments. Inclusion criteria also required the ability to understand and complete study questionnaires in Arabic or English and to provide written informed consent. Participants were excluded if they had undergone knee replacement surgery, presented with secondary OA due to trauma or inflammatory arthropathy, had a history of neurological or psychiatric disorders, or were receiving concurrent neuromodulatory treatments such as repetitive transcranial magnetic stimulation (rTMS). Additionally, individuals with contraindications to tDCS (e.g., seizure disorders, implanted metal devices in the head) or those taking centrally acting medications known to interfere with cortical excitability were excluded. This exclusion applied to medications affecting cortical excitability, such as antiepileptics, benzodiazepines, or tricyclic antidepressants. NSAIDs were not excluded, as they do not alter cortical excitability. However, their analgesic effects could influence pain scores and were recorded as part of baseline pharmacological data.

Clinical outcome variables

Pain intensity

Pain intensity was measured using the Numeric Pain Rating Scale (NPRS) [11], a validated tool commonly employed to quantify subjective pain experiences in clinical and research settings. Participants were asked to indicate the average intensity of their knee pain over the previous week on an 11-point scale, where 0 represented “no pain” and 10 indicated “the worst pain imaginable.” The NPRS was treated as a continuous variable for statistical analysis. This scale is known for its strong test-retest reliability (intraclass correlation coefficient > 0.80) in populations with musculoskeletal disorders and has demonstrated responsiveness to changes in clinical status [12].

Functional disability

Functional limitations were evaluated using the Knee Injury and Osteoarthritis Outcome Score– Physical Function Shortform (KOOS-PS) [13], a concise and validated tool for assessing activity-related impairments in individuals with KOA. The KOOS-PS includes seven items that reflect common functional tasks, such as getting up from a chair and climbing stairs, with each item rated on a 5-point Likert scale ranging from 0 (no difficulty) to 4 (extreme difficulty) [12]. Total scores were normalized to a 0–100 scale, where higher scores represent better physical functioning. The KOOS-PS has demonstrated strong psychometric properties and is highly correlated with the physical function domain of the WOMAC index in osteoarthritis populations [13].

Health-related quality of life

Health-related quality of life was measured using the EuroQol 5-Dimensions 5-Levels (EQ-5D-5 L) questionnaire, a widely recognized instrument for evaluating general health status [14]. It captures five key dimensions—mobility, self-care, performance of usual activities, pain or discomfort, and emotional well-being—each assessed across five levels of severity. Participant responses were converted into a single index value using a region-specific value set appropriate for the Kingdom of Saudi Arabia. The resulting scores range from values below 0, indicating health states perceived as worse than death, to 1, representing optimal health. The EQ-5D-5 L is extensively validated for use in osteoarthritis populations and is a standard measure in both clinical research and health economics [14].

For Arabic-speaking participants, validated Arabic versions of the NPRS, PCS, KOOS-PS, and EQ-5D-5 L were used. These translations have been culturally adapted and psychometrically validated for use in musculoskeletal and chronic pain populations in Arabic-speaking regions.

tDCS exposure

The number of completed tDCS sessions was recorded from physiotherapy treatment logs maintained by licensed clinicians. This variable was recorded as a continuous numeric variable and reflects cumulative tDCS exposure. All participants received a minimum of 5 sessions as part of routine care. The total number of sessions received beyond the minimum was determined by clinical discretion, patient tolerance, and adherence. No standardized dosing protocol was applied. As such, session count is an observational variable potentially reflecting symptom severity, therapeutic response, or patient engagement. These factors introduce endogeneity, and any associations with outcomes must be interpreted cautiously, as they may be confounded by unmeasured variables. Stimulation intensity was recorded in milliamperes (mA) from the device output settings for each session. The mean stimulation intensity across all sessions for each participant was extracted and treated as a continuous variable. The standard stimulation protocol used 1.5–2.0 mA, consistent with safety guidelines and prior OA studies [15, 16]. Stimulation duration was held constant at 20 min per session across all participants, as per departmental protocol. This value was confirmed from session records. Duration was included as a covariate to examine potential additive effects despite its fixed value. All participants received anodal stimulation over the contralateral primary motor cortex (M1) and cathodal stimulation over the contralateral supraorbital area (SO) using saline-soaked sponge electrodes. Electrode placement followed the 10–20 EEG system and was administered by trained physiotherapists. Electrode positions were documented but not varied for analytical purposes.

Psychological mediator

Pain catastrophizing

Pain-related cognitive and emotional responses were evaluated using the Pain Catastrophizing Scale (PCS) [17], a 13-item instrument designed to assess negative thought patterns associated with pain experiences. Each item is rated on a 5-point scale ranging from 0 (“not at all”) to 4 (“all the time”), yielding a total score between 0 and 52. The PCS captures three core dimensions of catastrophizing: rumination, magnification, and helplessness. Higher scores indicate a greater tendency toward maladaptive pain perceptions. Participants completed the questionnaire in either Arabic or English, utilizing culturally adapted and validated versions. The PCS has demonstrated strong internal consistency (Cronbach’s α > 0.85) in individuals with osteoarthritis and is widely used in pain research [18].

All self-report questionnaires were administered in a standardized manner during clinical appointments. All clinical and psychological measures were collected at a single standardized time point, either during the final tDCS session or within one week of treatment completion. Timing was consistent across participants to minimize variability related to assessment scheduling. Participants completed the forms in the presence of a trained research assistant, who provided clarification when needed, ensuring accuracy and minimizing response bias.

Statistical analysis

To address the first objective, Pearson correlation coefficients were computed to assess linear associations between tDCS parameters (number of sessions, stimulation intensity, and duration) and the primary clinical outcomes (pain, physical function, and quality of life). Multiple linear regression analyses were subsequently performed to evaluate the predictive effect of tDCS session frequency on each outcome, adjusting for relevant covariates. Given the cross-sectional design and lack of experimental manipulation, session frequency was treated as an observational variable reflecting real-world clinical practice rather than a controlled treatment dose. Consequently, associations should be interpreted as correlational, not causal. All regression models were assessed for multicollinearity using variance inflation factor (VIF), with values below 2.0 considered acceptable. Model fit was evaluated using R² and adjusted R² values, and residual plots were examined to confirm model adequacy. For the second objective, mediation analysis was conducted using the PROCESS macro (Model 4) for SPSS to determine the indirect effect of pain catastrophizing on the relationship between tDCS session frequency and clinical outcomes. Assumptions of linearity, normality of residuals, and homoscedasticity were verified through scatter plots, histograms, and the Shapiro–Wilk test, where appropriate. No significant violations were observed, supporting the use of parametric regression techniques. Bootstrapping with 5,000 resamples was applied to generate 95% confidence intervals for indirect effects. Additionally, independent samples t-tests and one-way ANOVA were used to examine subgroup differences by sex and osteoarthritis duration. All outcome variables reflect a single measurement taken after completion of the tDCS treatment period. No pre-intervention data were collected, and no within-subject comparisons were performed. The p-values reported in descriptive tables correspond to normality testing (Shapiro–Wilk), not pre/post statistical comparisons.

Results

The study sample comprised 68 individuals with KOA, predominantly female (60.29%), with a mean age of 63.45 years and an average body mass index indicative of overweight status (28.62 kg/m²). The mean duration of osteoarthritis was 6.38 years, and the majority of participants (72.06%) reported at least one comorbidity. Non-steroidal anti-inflammatory drugs were the most commonly used medications (75.00%). Information on the exact duration or dosage of NSAID use during the tDCS treatment period was not systematically captured, limiting the ability to fully control for potential confounding effects of pharmacological pain relief. Participants underwent an average of 8.41 tDCS sessions with a consistent stimulation duration of 20 min and a mean intensity of 1.95 mA. The most frequent electrode montage involved anodal placement over the primary motor cortex (70.59%) and cathodal placement over the contralateral supraorbital area (73.53%) (Table 1). Additional montages were applied in a subset of sessions, including M1:M1, SO: SO, and other placements, based on clinical discretion. These variations were infrequent and did not alter the inclusion requirement that M1-anodal stimulation be the primary exposure across sessions.

Table 1.

Demographic and clinical characteristics of participants

Variable	Mean ± SD or n (%)
Age (years)	63.45 ± 8.90
Sex (Male/Female)	27 (39.71%) / 41 (60.29%)
BMI (kg/m²)	28.62 ± 4.75
Duration of OA (years)	6.38 ± 3.22
Comorbidities (None/≥1)	19 (27.94%) / 49 (72.06%)
Medications (NSAIDs/Opioids/Others)	51 (75.00%) / 8 (11.76%) / 9 (13.24%)
tDCS sessions (number)	8.41 ± 1.63
Stimulation intensity (mA)	1.95 ± 0.15
Stimulation duration (minutes)	20.00 ± 0.00
Anode location (M1/SO/Other)	48 (70.59%) / 14 (20.59%) / 6 (8.82%)
Cathode location (SO/M1/Other)	50 (73.53%) / 12 (17.65%) / 6 (8.82%)

SD = Standard Deviation; OA = Osteoarthritis; BMI = Body Mass Index; NSAIDs = Non-Steroidal Anti-Inflammatory Drugs; tDCS = Transcranial Direct Current Stimulation; M1 = Primary Motor Cortex; SO = Supraorbital Area

Participants reported moderate levels of pain (mean NPRS: 5.43 ± 1.32) and physical disability (mean WOMAC-PF: 38.27 ± 9.81), alongside a modest health-related quality of life (mean EQ-5D-5 L index: 0.61 ± 0.15), reflecting the typical clinical profile of individuals with KOA (Table 2). The mean pain catastrophizing score (PCS: 22.74 ± 7.54) indicated a moderate degree of maladaptive pain-related cognitions. Physical and mental health components from the SF-36 were 42.86 and 47.12, respectively, suggesting moderate impairment, particularly in physical functioning. All variables demonstrated statistically significant p-values (< 0.05) in the Shapiro–Wilk test, supporting assumptions of normality required for parametric analyses. These values do not indicate treatment effects, as only post-treatment data were collected.

Table 2.

Summary of outcome and mediator scores

Variable	Mean ± SD	95% CI	p-value
Numeric Pain Rating Scale (NPRS)	5.43 ± 1.32	5.15–5.71	0.027
WOMAC Physical Function Subscale (WOMAC-PF)	38.27 ± 9.81	35.84–40.70	0.031
EQ-5D-5 L Index	0.61 ± 0.15	0.58–0.64	0.045
Pain Catastrophizing Scale (PCS)	22.74 ± 7.54	20.89–24.59	0.018
SF-36 Physical Component Score	42.86 ± 8.29	40.68–45.04	0.023
SF-36 Mental Component Score	47.12 ± 9.45	44.58–49.66	0.036

NPRS = Numeric Pain Rating Scale; WOMAC-PF = Western Ontario and McMaster Universities Osteoarthritis Index– Physical Function Subscale; EQ-5D-5 L = EuroQol 5-Dimensions 5-Levels; PCS = Pain Catastrophizing Scale; SF-36 = Short Form Health Survey– 36 items; CI = Confidence Interval; SD = Standard Deviation

Note: All scores represent post-tDCS values only. No pre-treatment data were collected. Reported p-values reflect the results of normality testing using the Shapiro–Wilk test, not comparisons across time points

Pearson correlation analyses demonstrated significant associations between tDCS parameters and clinical outcomes, with the number of sessions showing moderate correlations with reduced pain (r = − 0.42, p = 0.002), improved physical function (r = 0.39, p = 0.004), and enhanced quality of life (r = 0.44, p = 0.001). Stimulation intensity also correlated significantly, though more modestly, with pain (r = − 0.28, p = 0.031), function (r = 0.25, p = 0.048), and quality of life (r = 0.31, p = 0.022). In contrast, stimulation duration was not significantly associated with any clinical outcomes (all p > 0.05), suggesting a more prominent role for session frequency and intensity in driving therapeutic effects (Fig. 1).

Fig. 1.

Pearson Correlations Between tDCS Parameters and Pain, Physical Function, and Quality of Life in Knee Osteoarthritis

Multiple linear regression analyses revealed significant associations between the number of tDCS sessions and all primary clinical outcomes, supporting its role in symptom modulation. Specifically, increased session count was associated with a significant reduction in pain (B = − 0.27, p = 0.004), improvement in physical function (B = 1.82, p = 0.005), and better health-related quality of life (B = 0.04, p = 0.001). Standardized beta coefficients ranged from − 0.36 for pain to 0.42 for quality of life, indicating moderate effect sizes. The explained variance (R²) ranged from 0.15 to 0.22 across models, highlighting the clinical relevance of session frequency as a predictor of therapeutic outcomes in this population (Table 3; Fig. 2).

Table 3.

Multiple linear regression: tDCS predicting pain, disability, and quality of life

Dependent Variable	Independent Variable	B (Unstandardized Coef.)	SE (Standard Error)	95% CI	β (Standardized Coef.)	t-value	p-value	R²
NPRS (Pain Score)	tDCS Sessions	-0.27	0.09	-0.45 to -0.09	-0.36	-2.98	0.004	0.17
KOOS-PS (Physical Function)	tDCS Sessions	1.82	0.61	0.61 to 3.03	0.33	2.97	0.005	0.15
EQ-5D-5 L (Quality of Life)	tDCS Sessions	0.04	0.01	0.02 to 0.06	0.42	3.54	0.001	0.22

tDCS = Transcranial Direct Current Stimulation; NPRS = Numeric Pain Rating Scale; KOOS-PS = Knee Injury and Osteoarthritis Outcome Score– Physical Function Shortform; EQ-5D-5 L = EuroQol 5-Dimensions 5-Levels; B = Unstandardized Coefficient; SE = Standard Error; CI = Confidence Interval; β = Standardized Coefficient; R² = Coefficient of Determination

Fig. 2.

Standardized Beta Coefficients and Model Fit (R²) for the Effect of tDCS Session Frequency on Pain, Physical Function, and Quality of Life

Mediation analysis suggested a statistical pathway in which pain catastrophizing may partially account for the association between tDCS session frequency and clinical outcomes; however, these findings are correlational and do not establish temporal or causal mediation. For pain, the indirect effect via PCS was significant (ab = − 0.09, 95% CI: − 0.16 to − 0.03, p = 0.002), accounting for a substantial portion of the total effect (c = − 0.27). Similarly, PCS mediated the effects of tDCS on physical function (ab = 0.67, p = 0.004) and quality of life (a*b = 0.01, p = 0.038), with corresponding increases in explained variance (R² = 0.18–0.24). These findings suggest that targeting maladaptive pain cognitions may enhance the efficacy of tDCS in (Table 4). However, the magnitude of the indirect effect was small. For pain intensity, the estimated indirect effect (–0.09) corresponds to approximately 9% of the MCID for the NPRS, suggesting that while the pathway is statistically detectable, its clinical impact may be limited.

Table 4.

Mediation analysis: psychological variables as mediators

Outcome Variable	Direct Effect (c′)	Indirect Effect (a*b)	Total Effect (c)	95% CI for Indirect Effect	p-value (Indirect Effect)	R² (Model)
NPRS (Pain Score)	-0.18	-0.09	-0.27	-0.16 to -0.03	0.002	0.21
KOOS-PS (Physical Function)	1.23	0.67	1.90	0.28 to 1.05	0.004	0.18
EQ-5D-5 L (Quality of Life)	0.03	0.01	0.04	0.00 to 0.02	0.038	0.24

tDCS = Transcranial Direct Current Stimulation; NPRS = Numeric Pain Rating Scale; KOOS-PS = Knee Injury and Osteoarthritis Outcome Score– Physical Function Shortform; EQ-5D-5 L = EuroQol 5-Dimensions 5-Levels; PCS = Pain Catastrophizing Scale; CI = Confidence Interval; R² = Coefficient of Determination

Subgroup analyses revealed statistically significant differences in clinical outcomes based on both sex and osteoarthritis (OA) duration. Females reported higher pain scores (NPRS: 5.65 vs. 5.12), lower physical function (KOOS-PS: 36.72 vs. 40.19), and greater pain catastrophizing (PCS: 23.41 vs. 21.67) compared to males, with a p-value of 0.038. Similarly, participants with OA duration less than five years exhibited higher pain (NPRS: 5.71 vs. 5.22), worse function (KOOS-PS: 36.10 vs. 39.08), and higher PCS scores (23.88 vs. 21.79) than those with longer disease duration, also reaching statistical significance (p = 0.044). These results reflect descriptive subgroup comparisons only and do not establish that tDCS effects vary by sex or OA duration. As the differences in pain and function scores between males and females were below established MCIDs, the clinical relevance is likely limited. No interaction terms or formal subgroup analyses were performed (Fig. 3).

Fig. 3.

Mean post-treatment scores for Pain, Physical Function, Quality of Life, and Pain Catastrophizing across sex and OA duration subgroups

Discussion

This cross-sectional study aimed to examine the associations between tDCS parameters and key clinical outcomes—namely pain intensity, functional disability, and health-related quality of life—in individuals with KOA, while also evaluating the mediating role of psychological factors. The findings demonstrated that tDCS session frequency was significantly associated with improvements across all primary outcomes, suggesting a meaningful therapeutic influence of stimulation dosage. Pain catastrophizing demonstrated a statistical association with the relationship between tDCS session frequency and clinical outcomes; however, given the stable trait-like nature of catastrophizing and the cross-sectional design, moderation analysis or stratified comparisons based on PCS levels may have offered more appropriate alternatives. Future studies should consider such approaches to better evaluate whether the effects of tDCS vary according to psychological profiles. However, given the cross-sectional nature of this study and the conceptual stability of pain catastrophizing, the use of mediation analysis is exploratory and does not establish temporal or causal directionality. The simultaneous assessment of tDCS exposure, catastrophizing, and outcomes precludes firm conclusions about indirect effects. It remains equally plausible that pain catastrophizing may serve as a predictor or moderator of treatment responsiveness rather than a mediator. Therefore, these results should be viewed as hypothesis-generating, pending confirmation in longitudinal or interventional designs with repeated PCS measurements. Subgroup analyses further revealed differential patterns in clinical outcomes based on sex and osteoarthritis duration, indicating potential variability in tDCS responsiveness across demographic and disease-related strata. These results collectively support the potential of tDCS as a symptom-modulating intervention in and highlight the importance of considering individual psychological and clinical profiles in optimizing its application.

The observed associations between tDCS parameters and clinical outcomes can be attributed to the neuromodulatory effects of cumulative stimulation on cortical excitability and pain-related neural pathways [19]. The number of tDCS sessions showed the strongest and most consistent associations with improvements in pain, physical function, and quality of life, suggesting a dose-response relationship wherein repeated stimulation reinforces therapeutic gains [19]. In contrast, stimulation intensity yielded only modest associations, and stimulation duration showed no significant correlations, highlighting the relative importance of treatment frequency over single-session intensity or duration. The limited range of stimulation intensity (1.5–2.0 mA) likely reduced the ability to detect meaningful differences in outcomes based on dose. While the hypothesis included intensity as a potential contributor, the observed range did not include higher-dose conditions (e.g., 3–4 mA), and therefore conclusions regarding intensity effects should be considered exploratory and restricted to this narrow therapeutic window. These patterns were reinforced by multiple linear regression analyses, which demonstrated that session frequency was a significant predictor of all primary outcomes with moderate effect sizes. These findings align with the neurophysiological premise that repeated stimulation is more likely to produce lasting changes in synaptic efficacy and functional connectivity, particularly in networks involved in pain processing, motor control, and affective regulation [20].

These findings are supported by prior literature that has demonstrated beneficial effects of repeated tDCS in chronic pain conditions [21, 22]. For instance, Yao et al. [21] concluded in their comprehensive review that repeated sessions of anodal stimulation over the motor cortex are associated with sustained analgesic effects. Similarly, a study by Dube et al. [22] reported that multiple sessions of tDCS were more effective than single-session protocols in improving motor function and reducing pain in musculoskeletal conditions. In the context of osteoarthritis, Cutts et al. [23] highlighted the cumulative benefits of session-based tDCS interventions in modulating central pain pathways. Moreover, Tedeschi et al. [24] emphasized the importance of stimulation dose and frequency in achieving clinically meaningful outcomes in pain rehabilitation [25]. The current study’s results align closely with these prior findings and extend their applicability to, reinforcing the therapeutic potential of structured, session-based tDCS protocols in this population.

The mediation analysis results demonstrated that pain catastrophizing significantly contributed to the pathway through which tDCS session frequency influenced pain, physical function, and quality of life in individuals with KOA [26]. These findings suggest that psychological factors, particularly maladaptive pain-related cognitions, may amplify or mitigate the therapeutic effects of neuromodulation [19]. The indirect effects observed across all outcomes imply that individuals with elevated levels of pain catastrophizing may exhibit more pronounced symptom improvement when tDCS is administered regularly, potentially due to enhanced engagement of prefrontal-limbic circuits that mediate cognitive-affective aspects of pain processing. While the present study cannot confirm this mechanism, prior work has shown that tDCS can modulate brain regions associated with pain catastrophizing [27], though additional research is needed to establish direct effects. Furthermore, subgroup analyses indicated that females and participants with a shorter duration of OA experienced worse clinical profiles, including higher pain severity, poorer function, and greater catastrophizing [28]. These patterns may reflect known sex- and duration-related differences in pain sensitivity and emotional processing, which could influence both baseline symptom burden and responsiveness to neuromodulatory interventions [29]. These subgroup findings were exploratory and not adjusted for potential baseline confounders such as BMI, OA severity, or medication use. Moreover, the observed differences were below thresholds for clinical significance and do not support strong conclusions about sex- or duration-specific tDCS effects. The imbalance in gender representation may also limit the generalizability of these comparisons.

Previous research lends strong support to the observed mediating role of pain catastrophizing in the current study [30]. Turner et al. [31] originally established the Pain Catastrophizing Scale as a robust predictor of pain intensity and disability, which has since been validated across musculoskeletal populations. More recently, Caumo et al. [27] demonstrated that high levels of catastrophizing are associated with altered central pain modulation, a mechanism potentially amenable to modulation by tDCS. In the context of OA, studies by Buenaver et al. [32] and Terry et al. [33] confirmed that catastrophizing significantly mediates the relationship between psychosocial stress and pain-related outcomes. Moreover, Lisoni et al. [34] highlighted the modulatory effect of tDCS on cortical regions involved in affective-emotional aspects of pain, reinforcing the plausibility of the observed mediation effects. These prior findings collectively validate the present results and highlight the necessity of integrating cognitive-affective assessments into tDCS treatment planning for KOA.

Clinical significance

The clinical significance of this study lies in its demonstration that the frequency of tDCS sessions is a meaningful predictor of therapeutic outcomes in individuals with KOA, with measurable improvements observed in pain reduction, physical function, and quality of life. Importantly, the study identifies pain catastrophizing as a significant psychological mediator of these effects, underscoring the need to consider cognitive-affective factors when implementing neuromodulatory interventions. These findings support the integration of tDCS into multidisciplinary management strategies for, particularly for patients with elevated pain-related distress. By highlighting both the dose-dependent effects of tDCS and the modulating role of psychological profiles, this study provides clinically actionable evidence for tailoring stimulation protocols to optimize patient-centered outcomes in routine rehabilitation settings.

Limitations

Despite its contributions, this study has several limitations. The cross-sectional design precludes assessment of causality and temporal relationships between tDCS exposure and clinical outcomes. While statistically significant, the observed associations were modest in magnitude and often below established minimum clinically important differences, limiting clinical relevance. No baseline measurements were collected as only post-treatment data were available, analyses did not adjust for baseline symptom severity. Variability in tDCS montage placement and concurrent NSAID use also introduce potential confounding. Although NSAIDs were allowed due to their lack of cortical excitability effects, their analgesic properties may have independently influenced pain outcomes; however, dosage and duration were not systematically monitored. Furthermore, only pain catastrophizing was examined as a psychological mediator, whereas other relevant constructs—such as central sensitization or cortical excitability—were not assessed. The reliance on self-reported measures introduces the possibility of reporting bias, and the absence of neuroimaging or electrophysiological data limits mechanistic insight. Subgroup analyses were exploratory, unadjusted for baseline differences, and not powered to detect differential effects. Future research should employ longitudinal, randomized controlled designs with standardized stimulation protocols, baseline-adjusted models, and multimodal assessments (e.g., TMS, fMRI, or sensory testing) to more rigorously evaluate tDCS efficacy and individual responsiveness in KOA.

Conclusion

This study demonstrated that the frequency of tDCS sessions is significantly associated with improvements in pain intensity, physical function, and quality of life among individuals with KOA. Pain catastrophizing was identified as a partial mediator of these relationships, indicating its role in the psychological pathway through which tDCS may exert clinical benefits. Additionally, subgroup differences based on sex and osteoarthritis duration suggest variability in symptom burden and potential response patterns. These findings provide specific evidence supporting the relevance of session frequency and psychological profiling in optimizing the application of tDCS for symptomatic management in KOA. However, these findings remain preliminary. Future studies should adopt prospective, randomized controlled designs with standardized stimulation protocols, pre/post measurement of outcomes, and rigorous adjustment for confounding variables to validate the therapeutic efficacy of tDCS with KOA and related chronic pain conditions.

Abbreviations

EQ-5D-5L

EuroQol 5-Dimensions 5-Levels

KOA

Knee Osteoarthritis

KOOS-PS

Knee Injury and Osteoarthritis Outcome Score– Physical Function Shortform

M1

Primary Motor Cortex

mA

Milliamperes

NPRS

Numeric Pain Rating Scale

NSAIDs

Non-Steroidal Anti-Inflammatory Drugs

PCS

Pain Catastrophizing Scale

rTMS

Repetitive Transcranial Magnetic Stimulation

SF-36

Short Form Health Survey– 36 items

SO

Supraorbital Area

tDCS

Transcranial Direct Current Stimulation

VIF

Variance Inflation Factor

WOMAC-PF

Western Ontario and McMaster Universities Osteoarthritis Index– Physical Function Subscale

Author contributions

Ravi Shankar Reddy and Jaya Shanker Tedla contributed equally to the conception and design of the study, supervision of data collection, statistical analysis, interpretation of results, and manuscript drafting. Shaker Hassan S. Alshehri contributed to clinical interpretation, manuscript refinement, and orthopedic relevance validation. Faisal M. Alyazedi supported methodological review, data interpretation, and critical manuscript revision. Suhail Mansour Aljehani participated in participant recruitment, data acquisition, and clinical assessments. Feras Ahmed Alarabi contributed to data organization, literature review, and manuscript formatting. All authors reviewed and approved the final version of the manuscript.

Funding

King Salman center For Disability Research, Research Group no KSRG-2024-187.

Data availability

The raw data supporting the findings of this study are openly available on Zenodo at 10.5281/zenodo.15455083. All versions of the dataset can be cited using this DOI.Researchers are encouraged to access the data for replication or further analysis.

Declarations

Ethics approval and consent to participate

Ethical approval for this studywas obtained from the Institutional Research Ethics Committee (REC) at King Khalid University (approval number: ECM#2023 − 276) on 22/05/2023.

Consent for publication

All participants in the present study received instruction about the experimental procedures and content and the experiments were conducted after participants had completed the consent forms.

Competing interests

The authors declare no competing interests.